to profit or not to profit: the commercial transformation of the nonprofit sector

To Profit or Not to ProfitThe Commercial Transformation of the Nonprofit Sector

Nonprofit organizations are becoming increasingly like private firms. The transforma-tion is bringing a shift in financial dependence from charitable donations to commercialsales activity, with little-recognized consequences. To Profit or Not to Profit is a coordi-nated set of studies of why fund-raising for nonprofits is mimicking that of private firmsand what consequences it is having. User fees and revenue from "ancillary" activities -those not contributing directly to the organization mission except for the profit gener-ated - are mushrooming, with each having important side effects. User fees may priceout of the market some of the nonprofit's target group. Ancillary activities may distractthe nonprofit from its central mission.

These issues are examined from two perspectives. One focuses on issues that applyto nonprofits generally: the role of competition, a framework for analyzing nonprofitorganization behavior, the effects of distributional goals and differential taxation of non-profit and for-profit activity revenue, the effects of changes in donations on commercialactivity, and conversions of nonprofits to for-profits. A second set of studies focuses onspecific industries: hospitals, universities, social-service providers, zoos, museums, andpublic broadcasting. The book concludes with recommendations for research and forpublic policy toward nonprofits.

Burton A. Weisbrod is John Evans Professor of Economics and Faculty Fellow of theInstitute for Policy Research at Northwestern University. From 1990 to 1995 he servedas director of that public-policy research unit, then known as the Center for Urban Af-fairs and Policy Research. Previously Professor Weisbrod was Evjue-Bascom Professorof Economics at the University of Wisconsin-Madison and cofounder of its HealthEconomics Training Program. He has held visiting professorships at Brandeis, Harvard,Princeton, and Yale universities, as well as the University of California-Berkeley, theState University of New York-Binghamton, the Australian National University, and theUniversidad Autonoma de Madrid, Spain. Professor Weisbrod is the author of two pre-vious books and numerous articles on nonprofits, including the landmark analysis TheNonprofit Economy (1988), and has authored, coauthored, or edited 14 other books andmore than 160 articles.

To Profit or Not to ProfitThe Commercial Transformationof the Nonprofit Sector

Edited by

BURTON A. WEISBROD

CAMBRIDGEUNIVERSITY PRESS

PUBLISHED BY THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGE

The Pitt Building, Trumpington Street, Cambridge, United Kingdom

CAMBRIDGE UNIVERSITY PRESS

The Edinburgh Building, Cambridge CB2 2RU, UK http://www.cup.cam.ac.uk40 West 20th Street, New York, NY 10011-4211, USA http://www.cup.org10 Stamford Road, Oakleigh, Melbourne 3166, AustraliaRuiz de Alarcon 13, 28014 Madrid, Spain

© Burton A. Weisbrod 1998

This book is in copyright. Subject to statutory exception andto the provisions of relevant collective licensing agreements,no reproduction of any part may take place withoutthe written permission of Cambridge University Press.

First published 1998

First paperback edition 2000

Typeset in Times Roman 10 pt, in Quark XPress™ [MG]

A catalog record for this book is available from the British Library

Library of Congress Cataloging in Publication dataTo profit or not to profit: the commercial transformation of the nonprofit sector /

edited by Burton A. Weisbrod.p. cm.

A collection of papers initiated, discussed, and refined at two workingconferences held by the contributors.

Includes bibliographical references and index.1. Nonprofit organizations - United States - Finance - Congresses.

2. Associations, institutions, etc. - United States - Finance - Congresses.3. Public institutions - United States - Finance - Congresses.4. Fund raising - United States - Congresses.5. Volunteerism - United States - Congresses.6. Profit - United States - Congresses. I. Weisbrod, Burton Allen 1931-HD2769.2.U6T6 1998338.7-dc21 97^5976

CIP

ISBN 0 521 63180 7 hardbackISBN 0 521 78506 5 paperback

Transferred to digital printing 2003

Contents

List of contributors page viiForeword by Kenneth J. Arrow ixPreface xi

1 The nonprofit mission and its financing: Growing linksbetween nonprofits and the rest of the economy 1Burton A. Weisbrod

Part I Basic issues and perspective

2 Competition, commercialization, and the evolution ofnonprofit organizational structures 25Howard P. Tuckman

3 Modeling the nonprofit organization as a multiproduct firm:A framework for choice 47Burton A. Weisbrod

4 Pricing and rationing by nonprofit organizations withdistributional objectives 65Richard Steinberg and Burton A. Weisbrod

5 Differential taxation of nonprofits and the commercializationof nonprofit revenues 83Joseph J. Cordes and Burton A. Weisbrod

6 Interdependence of commercial and donative revenues 105Lewis M. Segal and Burton A. Weisbrod

7 Conversion from nonprofit to for-profit legal status:Why does it happen and should anyone care? 129John H. Goddeeris and Burton A. Weisbrod

vi Contents

Part II Industry studies

8 Commercialism in nonprofit hospitals 151Frank A. Sloan

9 Universities as creators and retailers of intellectual property:Life-sciences research and commercial development 169Walter W. Powell and Jason Owen-Smith

10 Commercialism in nonprofit social service associations:Its character, significance, and rationale 195Dennis R. Young

11 Zoos and aquariums 217Louis Cain and Dennis Meritt, Jr.

12 Commerce and the muse: Are art museums becomingcommercial? 233Helmut K. Anheier and Stefan Toepler

13 The funding perils of the Corporation for Public Broadcasting 249Craig L. LaMay and Burton A. Weisbrod

Part III Overview, conclusions, and public-policy issues

14 Commercialism among nonprofits: Objectives, opportunities,and constraints 271Estelle James

15 Conclusions and public-policy issues: Commercialism andthe road ahead 287Burton A. Weisbrod

Appendix: IRS Forms 990 and 990-Tfor nonprofit organizations 306References 317Index 336

Contributors

BURTON A. WEISBROD is John Evans Professor of Economics, andFellow of the Institute for Policy Research, at Northwestern Uni-versity.

HELMUT K. ANHEIER is Associate Professor of Sociology at RutgersUniversity and Senior Research Associate at the Institute for Poli-cy Studies at Johns Hopkins University.

KENNETH J. ARROW is Joan Kenney Professor of Economics, Emer-itus, Professor of Operations Research, Emeritus, at Stanford Uni-versity, and recipient of the Nobel Memorial Prize in EconomicScience in 1972.

Louis CAIN is Professor of Economics at Loyola University of Chi-cago and Adjunct Professor of Economics at Northwestern Uni-versity.

JOSEPH J. CORDES is Professor of Economics and Director of theGraduate Program in Public Policy at George Washington Univer-sity.

JOHN H. GODDEERIS is Professor of Economics at Michigan StateUniversity.

ESTELLE JAMES is Lead Economist, Policy Research Department, atthe World Bank.

CRAIG L. LAMAY is Assistant Professor at the Medill School of Jour-nalism at Northwestern University.

DENNIS MERITT, JR. , is Associate Adjunct Professor of Biology atDePaul University and a past President of the American Zoo andAquarium Association.

JASON OWEN-SMITH is a graduate student in Sociology at the Uni-versity of Arizona.

WALTER W. POWELL is Professor of Sociology at the University ofArizona.

vii

viii Contributors

LEWIS M. SEGAL is Economist at the Federal Reserve Bank of Chi-cago.

FRANK A. SLOAN is J. Alexander McMahon Professor of Health Pol-icy and Management and Professor of Economics at Duke Univer-sity.

RICHARD STEINBERG is Professor of Economics, Public Affairs, andPhilanthropic Studies at Indiana University / Purdue University-Indianapolis.

STEFAN TOEPLER is Research Associate and Adjunct Professor at theInstitute for Policy Studies at Johns Hopkins University.

HOWARD P. TUCKMAN is Professor of Economics and Dean of theSchool of Business at Virginia Commonwealth University.

DENNIS R. YOUNG is Professor of Nonprofit Management and Eco-nomics at Case Western Reserve University.

Foreword

Kenneth J. Arrow

In every modern economy, no matter how strong the pledged and actual alle-giance to the ideology of the market, there is a significant portion in which pro-duction is not governed by the maximization of profit and there are no legalclaimants to any of the revenues. The decision-making entities in this nonprofitsector are, of course, operating in the context of a market economy. They haveto purchase inputs on the market, and in many cases (e.g., hospitals, universi-ties) they receive revenues from the users of their products. To finance the pur-chase of these inputs, they frequently receive, in addition to their user fees,donations. The government not only subsidizes these donations through taxlaws but also frequently gives these nonprofits direct subsidies. Finally, non-profit institutions profit from gifts of time from volunteer workers. Thus the re-sources that support the activities of nonprofits are received through a varietyof channels: user fees, direct and indirect government subsidies, and donationsof money and labor.

The role of nonprofit institutions in the United States is certainly consider-able and worthy of study. Their revenues make up about 10 percent of the GNP,and they own roughly 10 percent of all property. Their special role, particularlytheir tax status, is justified on the grounds that they supply uniquely valuableservices. However, the links between the commercial and nonprofit sectors aregrowing rapidly, and the lines that divide them are getting harder to define. Insome cases, profit-making and nonprofit firms directly compete - for instance,in consulting work for public policy. They also cooperate, as when private firmsgive grants to universities to do research in areas of interest to those firms; butthese grants and other donations to nonprofit institutions from the private sec-tor can entail conditions - such as not publishing research - that conflict withthe nonprofits' aims. The property-tax exemptions for nonprofit firms create ad-ditional tax pressures on private property owners.

Perhaps the most striking possibility of all is the commercialization of non-profit firms. When competing with for-profits, nonprofits may develop a largevalue; pressure to capitalize on this value may come from insiders, who usu-

x Foreword

ally wind up with a major part of the gains. The motive for commercializationmay also arise out of the desire, and perhaps even need, to expand. Such expan-sion requires capital obtainable only on the bond or equity markets, and thusout of reach of a nonprofit institution.

It is curious that there is a sense of pressure on nonprofits at a time whenthey have been expanding. Perhaps there is some problem with the measures.A reduction in outside support (relative to an expanding economy) may haveled to an increase in commercial enterprises to raise money (such as gift shopsin museums). This will appear as an expansion of the nonprofit sector eventhough the distinctive output of that sector may not have increased - only thecommercial services identified with it.

The economic analysis of the nonprofit sector has always suffered from thedifficulty of defining the objective functions of its units. The objectives evenof a business firm faced with investments whose returns depend on judgmentof the uncertain future - and run by managers whose interests are not necessar-ily those of the stockholders - suffer from some ambiguities that theory has notsucceeded even in formulating clearly. The objectives of nonprofit firms, how-ever, are generally multiple and, in any but the simplest cases, hard to define.Burton A. Weisbrod has been one of the very few scholars who has consistent-ly and over a long period of time brought to bear theoretical and empirical re-search tools in this complex field. Now he and his colleagues have focused ona phenomenon always extant but of growing significance: the commercializa-tion of the nonprofit sector. Almost all of what is known regarding this devel-opment is to be found in these pages.

Preface

This volume is not the literary equivalent of the proverbial camel - a horse puttogether by a committee. Rather, it is a genuine product of intense teamwork.Great effort was made by all the contributors to incorporate their work as partof a single, unified analysis of the forces causing the growing commercialismof nonprofit organizations, the forms that commercialism is taking, and the like-ly consequences.

The process of producing an integrated volume began with lengthy conver-sations with the prospective authors, to determine not only their interest in join-ing the team but also their willingness to agree to procedures that would im-pose some strictures in order to produce a cohesive book. The next stage wasa working conference lasting a day and a half, during which the proposed per-spective and approach were discussed, to give each participating author fullopportunity to propose modifications; time was then allotted to the discussionof the two-page outlines that each author had circulated prior to the conference.In this way, each participant learned about the plans of all the others, and thevolume's integration was advanced.

Nine months later a second working conference was held. First drafts of allpapers had been circulated in advance, and each of us served as a discussant ofanother's paper. As the organizer of the overall research project, I gave writtencomments and suggestions to every author and had subsequent conversationsto expand on the revisions that were proposed. The highly interactive processgenerated multiple linkages across chapters.

I want to thank every author not only for the fine research contribution, butfor the superb teamwork. It was a pleasure to work with authors who were ded-icated not only to scholarly excellence, but to producing a book that would bemore valuable than the sum of its parts.

This book would not have been possible without the support of the AndrewW. Mellon Foundation for the research and the working conferences. I thankthe Foundation and its president, William G. Bowen, for their confidence andassistance. I also thank Anne Fitzpatrick and Laura McLean for their valuable

xi

xii Preface

and wide-ranging administrative help - coordinating the work of the contribu-tors, handling organizational arrangements for the conferences, editing the en-tire manuscript - and for maintaining the good humor that made the processflow smoothly. I appreciate their assistance greatly. I also thank Elizabeth Sel-vin for her fine administrative support in the final stages of the publication pro-cess, Jeffrey Ballou for reviewing and providing helpful comments at the page-proof stage, and Michael Gnat for his valuable editorial help throughout thepublication production process.

CHAPTER 1

The nonprofit mission and its financing:Growing links between nonprofitsand the rest of the economy

Burton A. Weisbrod

Introduction

Massive change is occurring in the nonprofit sector. Seemingly isolated eventstouching the lives of virtually everyone are, in fact, parts of a pattern that islittle recognized but has enormous impact; it is a pattern of growing commer-cialization of nonprofit organizations:

Nonprofit hospitals are launching health clubs open to the public, withthe latest exercise equipment and Olympic-size swimming pools, gen-erating substantial profits and threatening the for-profit fitness centerindustry (Stone 1997).

Nonprofit museums are opening glitzy retail shops, generating revenuethat is now a larger percentage of operating income than that fromfederal funding or admissions and memberships (Dobrzynski 1997;Mullen 1997).

Nonprofit universities are engaging in research alliances with privatefirms and suppressing research findings that are unfavorable to thosefirms' profit prospects (Altman 1997; Kolata 1997).

Nonprofits in various industries are forming for-profit subsidiaries, en-gaging in joint ventures with private firms, and paying executivescompensation at "Fortune 500" levels (Farhi 1997; Gosselin and Zit-ner 1997;Abelson 1998).

Commercialism in the nonprofit sector sounds like a paradox: Nonprofits aresupposed to be different from private firms, for whom commercialism is theirvery lifeblood. To some people, though, the uniqueness of nonprofit organiza-tions is by no means self-evident; perhaps they are really not different from pri-

I would like to thank Jeffrey Ballou for his comments.

1

2 Burton A. Weisbrod

vate firms, but are just as influenced by business motives and opportunities forself-aggrandizement.

Late in 1997 two apparently unrelated events brought front-page headlines.One involved a contract between the American Medical Association and theSunbeam Corporation, a manufacturer of consumer electronic products, withthe AMA promising to endorse Sunbeam products, such as heating pads andvaporizers, in return for payments expected to yield millions of dollars ("TheAMA's Appliance Sale" 1997; Graham and Van 1997). The other involved thepurchase by Chicago's Field Museum of Natural History of "Sue," the largestcomplete Tyrannosaurus rex fossil in existence - the $8.3 million cost being fi-nanced largely by McDonald's and the Walt Disney Company. McDonald's willget several "life-size replicas of the ferocious dinosaur, one of the most wide-ly recognized dinosaurs and a powerful promotional tool" (Swanson 1997b: 9),and one of the replicas will be displayed at McDonald's DinoLand USA at-traction at Disney's new (1998) Animal Kingdom theme park. In addition, themuseum will display the original in its new McDonald's Fossil Preparatory Lab-oratory, and there is talk of miniatures, with the Field Museum name, being in-cluded in hamburger Happy Meals (Swanson 1997a).

Both the AMA and Field Museum cases involved nonprofit, tax-exempt or-ganizations contracting to receive multimillion-dollar payments from privatefirms. Both arrangements generated money but also criticism. The criticism ofthe AMA was so intense that its top leadership resigned and the AMA broke theagreement with Sunbeam, resulting in a $20 million breach-of-contract lawsuit(Collins 1997; Zuger 1997).

The nonprofit sector has been both criticized and acclaimed. The rationalefor its special tax treatment and subsidies rests on the belief that it providesservices that are materially different from, and preferred to, the services thatprivate enterprise provides. The theoretical case for expecting such differen-tial behavior has been examined elsewhere (Weisbrod 1975,1988; Hansmann1980; Rose-Ackerman 1986; Gray 1991). This book builds on the foundationof that literature, examining the ways that the mechanisms used to finance non-profits' activities affect, and are affected by, the nature of those activities.

In the United States, the past three decades have seen a tripling of the num-ber of nonprofits, from 309,000 in 1967 to nearly one million today. Their to-tal revenues, less than 6 percent of GNP in 1975, exceeded 10 percent in 1990(Herman 1995). From 1980 to 1990, nonprofits' paid employment grew by 41percent - more than double the overall growth of national employment (Sala-mon et al. 1996). As the nonprofit sector experiences such rapid growth, atten-tion to it is bound to increase - and indeed it has.

Why is the nonprofit sector growing? Special privileges for nonprofits arejustified largely on the grounds that they perform the kinds of public-type func-tions typically identified with government - helping the disadvantaged, provid-

The nonprofit mission and its financing 3

ing social services, and supporting museums, schools, environmental preser-vation, and medical research. When the government is able to provide theseservices in forms and amounts that voters want, little role exists for nonprofits.However, when populations are very diverse, services that satisfy the majoritymay leave many people severely undersatisfied; nonprofits are thus understand-able as an alternative mechanism for providing public-type services. The morehomogeneous a society is, the more similar are its citizens' preferences, and thesmaller the need for nonprofit organizations (Weisbrod 1975).1

This perspective on the relationship between nonprofits and government im-plies that the growth of nonprofit activity reflects an increase in the gap betweenperceived social need and government provision. From the perspective of anynonprofit organization, the need for its activity depends on the size of the ag-gregate problem and the degree to which government deals with it, in additionto the degree to which other nonprofits are active. If a nonprofit university seesthe need to modernize its scientific research facilities in order to pursue cutting-edge research, but grants from government and donations from private sourcesare either decreasing or rising slowly, the lure of commercial revenue is likelyto be powerful. Similarly, if the number of homeless people is growing despitewhatever actions governments are taking, nonprofits fighting poverty are like-ly to be attracted to commercial mechanisms to generate additional revenue.

Nonprofits' pursuit of additional revenue may reflect not only pressures toincrease output but also increases in the cost of producing a stable output. Asproviders of heavily labor-intensive services, nonprofits tend to face a "cost dis-ease." William J. Baumol and William G. Bowen, although not focusing on non-profits, identified the cost-increasing pressure on service providers generally;they showed that in the service sector the limited opportunity to use capital toincrease labor productivity drives costs upward (Baumol and Bowen 1966;Baumol 1967). In the nonprofit sector a museum, zoo, day-care center, or home-less shelter, for example, may have little potential for increasing productivity;hence, over time, overall economic growth will increase their costs and expandtheir search for revenue simply to sustain output. Whether this is equally truefor other nonprofit subsectors, such as hospitals and colleges, is less clear.

1 In countries with relatively homogeneous populations, such as in Scandinavia, government is suf-ficient to meet the wants of its citizens for the various collective-type services; thus we find thatgovernments are in fact considerably larger in those countries, while the nonprofit sectors are rel-atively unimportant. This helps to explain two phenomena that have been widely observed: first,the far greater importance of the nonprofit sector in the United States than in other countries; andsecond, the growing importance of nonprofits everywhere, as population migration and infor-mation flows through television and computers are having the effect of magnifying diversity incountry after country. Nonprofit organizations accounted for one-seventh of net new jobs inFrance during the 1980s and one-ninth in Germany and the United States (Salamon and Anheier1994, 31). Nonprofits' real expenditures grew by 240 percent in Hungary, 78 percent in Japan,and 55 percent in the United States (Salamon et al. 1996,22,28,40).


As the nonprofit sector grows, the debate over its proper role in a moderneconomy continues, periodically grabbing public attention: Private firms claimthey are victims of "unfair competition" from subsidized, tax-exempt, nonprof-its, pressuring lawmakers to restrict nonprofits' commercial activities (Rose-Ackerman 1990); the federal government considers revising the personal in-come tax in ways that will affect incentives to donate to charitable nonprofits;local governments introduce the latest device for extracting money from sup-posedly tax-exempt nonprofits.

In short, what brings the activities of nonprofits into the limelight is theirlinks with the rest of the economy. Contrary to the common view, nonprofitsare far from independent of private enterprise and government. They competewith and collaborate with these other organizations in countless ways in theirefforts to finance themselves, to find workers, managers, and other resourcesto produce their outputs, and to develop markets for those outputs.

As the nonprofit sector grows in size and commercial activities, is it becom-ing indistinguishable from the private sector? The fundamental issue is how, ifat all, does revenue source affect an organization's behavior? This question issignificant because at root is whether the organizations in the rapidly growingnonprofit sector deserve the subsidies and tax exemptions they receive fromindividuals and from federal, state, and local governments. This book demon-strates the enormous complexity of attempts to answer this question, and theequally great importance of doing so. U.S. nonprofits, and particularly the near-ly three-quarters of a million that are deemed "charitable" and "social welfare"nonprofits - those exempt from federal taxation under sections 501(c)(3) and(4) of the Internal Revenue Code - do not operate outside the overall economy.Their interactions with private enterprise and government are taking a dizzy-ing, ever-growing variety of forms. In some instances the drive for revenue isbringing nonprofits into headlong competition with private enterprise; this ismost apparent in the hospital industry, where increasing pressure to containhealth-care costs is bringing massive industry reorganization and intense pricecompetition (see Chapter 8). In other cases, nonprofits are collaborating withthese other sectors - as in biotechnology research, where universities and pri-vate enterprise are forming joint ventures, finding ever more complex ways ofcooperating in the interest of generating revenues and facilitating scientists'mobility between the academy and private firms (see Chapter 9).

The keys to understanding the interrelations of nonprofits with other partsof the economic system are the differential constraints on them relative to gov-ernment and private enterprise, and the differential goals. Constraints are moreeasily observed. Whenever "rules" - regulations, taxes, or prices - differ amongeconomic sectors, organizations can gain from exchanges between sectors. If,for example, prices differ, perhaps because of differential subsidies, then the un-subsidized sector will buy from the low-cost, subsidized sellers, other things be-


ing equal. Nonprofit and for-profit organizations face differential restrictions onwhether their income is taxed, differential availability of resources such as vol-unteer labor, differential restrictions on their freedom to lobby legislators (Aren-son 1995b), and differential access to private and public subsidies. Differentialgoals, while more difficult to observe, are also important, as we shall see.

Competition

Nonprofits are competing increasingly with for-profit firms, and in an amazingvariety of forms. The competition between nonprofits and private firms is madeeven more complex by the fact that at the same time that nonprofits are mov-ing into activities that have previously been the domain of for-profit firms, pri-vate firms are expanding into traditionally nonprofit areas. Thus, private healthclubs have entered an industry long the preserve of the nonprofit YMC As andYWCAs in the United States, and for-profit amusement parks "are encroach-ing on [nonprofit] museum turf by adding educational aspects to their enter-tainment products" (Becker 1995). Whenever private firms and nonprofit organ-izations become more competitive with each other, regardless of which entersthe industry later, the likely result is that the two types of competitors will be-come increasingly indistinguishable. This expectation, however, should be re-garded as a hypothesis to be tested, for even its theoretic basis is not firmly es-tablished. It is possible that competition could sharpen distinctions, leading toincreasingly different market niches for nonprofits and for-profits. The natureof competition between nonprofits and for-profits, as well as of competitionamong nonprofits, remains poorly understood (but is examined in greater de-tail in Chapter 2). Among the industries we examine in Chapters 8-13 there isconsiderable variation in the relative sizes and growth rates of the nonprofitand for-profit sectors: In higher education (Chapter 9), social service agencies(Chapter 10), zoos (Chapter 11), and museums (Chapter 12), there is little orno private-enterprise sector. In the hospital industry (Chapter 8), by contrast,the nonprofit sector, although quantitatively dominant, is confronting compe-tition from an increasingly aggressive private-enterprise sector. In televisionbroadcasting, moreover, "public" (i.e., nonprofit-sector) television sustains aniche market in an industry dominated overwhelmingly by private firms (Chap-ter 13).

Cooperation

The commercialism of nonprofits does not necessarily involve their compet-ing with private firms; it also occurs through cooperative mechanisms in whichnonprofits and private firms join forces. Between 1980-1 and 1987-8, for ex-ample, private-industry support for university research more than doubled in


real terms (Webster 1994). Virtually every major U.S. university has joinedforces in some manner with large multinational firms, especially pharmaceu-tical and chemical firms (Blumenthal et al. 1996): Harvard University has con-tracted with the German chemical company, Hoechst A.G.; Washington Uni-versity, with Monsanto Chemical Company; and Northwestern University, withDow Chemical (see Chapter 9).

Scientific research is not the only area in which nonprofit universities arebecoming allied with private firms. Athletics is another. The University of Mich-igan recently reached a multimillion-dollar agreement to advertise athletic shoesfor Nike; in a masterful obfuscation, Nike agreed to "help" the university de-sign their uniforms! Even charities are engaging in such alliances: The Marchof Dimes (MOD) recently teamed up with Kellogg's, the manufacturer of break-fast cereals; the MOD received $100,000 in return for what amounts to an en-dorsement of Kellogg's Product 19 - a cereal that contains folic acid, whichhelps prevent birth defects of the spine and brain. This is no isolated example;the nonprofit MOD even has a senior staff member with the title Director ofStrategic Alliances, National Promotions, to develop such money-generatingarrangements (Mclver 1995 a).

Nonprofits have discovered that there are massive potential financial bene-fits from such symbiotic relationships with private firms - benefits that are notoccasional and random, but are systematic consequences of powerful econom-ic forces. Universities and other nonprofits such as hospitals, museums, andcharities receive many subsidies that can be converted into something salableto private firms or consumers. How should we look at all these market-orientedchanges? Are they helping to strengthen nonprofits and make them more effec-tive contributors to meeting society's needs, or are they weakening nonprofits,diverting their efforts, leading them into economic behavior that is counter-productive - perhaps even undermining public support, financial and political?The tension between a nonprofit's focus on raising revenues and on its public-serving social mission is well illustrated by universities' partnerships and al-liances with private firms involving research in the life sciences, which are in-creasingly common, lucrative, and contentious.

Over a decade ago, A. Bartlett Giamatti, then president of Yale University,noted this tension and pointed to "the academic imperative . . . to seek knowl-edge objectively and to share it openly and freely, while the industrial imper-ative is to garner a profit, which frequently creates the incentive to treat knowl-edge as private property" (Peterson 1983, 123). The subsequent growth ofuniversity-industry research collaboration may be exacting a price in the formof pressure from private firms to suppress discoveries that are adverse to theirinterests and profits (Chapter 9).

Conflicts over the dissemination of adverse findings have occurred in a num-ber of recent cases, involving researchers at the University of California at San


Francisco (UCSF) and at the University of Washington at Seattle. At UCSF, aresearcher, operating under a grant from a private pharmaceutical firm, foundthat the effectiveness of a brand-name thyroid drug was no greater than that ofgeneric versions that were considerably less costly - a finding that led the re-searcher to withdraw the paper, under threat of a lawsuit by the funder, on theeve of its publication in the Journal of the American Medical Association (Alt-man 1997). At the University of Washington, research findings that were un-favorable for private firms - that a popular form of spine surgery might not beeffective, and that a popular drug for lowering blood pressure was associatedwith increased risk of heart attacks - were, according to the researchers, greet-ed by political pressure to eliminate the Federal agency that paid for the spineresearch and "harassment from drug companies and their academic consultants"(Kolata 1997). The "invasion of commerce into medical care" has been seen as"an epic clash of cultures between commercial and professional traditions . . ."(MeArthur and Moore 1997). Others, though, see nonprofits' emulation of, andinvolvement with, private firms as a sign of efficient resource utilization.

Interaction between nonprofits and government

The changes occurring in the nonprofit sector not only impact the relationshipsbetween nonprofits and for-profit organizations, but also influence the way non-profit organizations interact with the governmental sector. When nonprofits ex-pand, governments lose revenue. In the United States, the loss is particularlyimportant for local governments, whose major source of revenue is propertytaxation: Every parcel of land that moves out of the taxed, private-enterprisesector and into the untaxed, nonprofit sector entails further loss of tax revenue.At a time when governments are searching desperately for new sources of in-come to deal with the growing problems of crime and poverty and the loss ofrevenues from the federal government, the expansion of nonprofits is a seriousfiscal threat. The most recent report by the U.S. Treasury Department statedthat 10 percent of all property in the country was held by tax-exempt nonprofitsin 1977 (Arenson 1995c), and the figure is surely higher today: Indeed, the cityof Syracuse, New York, found that 59 percent of its real estate was tax exemptin 1993; for Buffalo, New York, it was 34 percent (Glaberson 1996).

Local governments are responding to this erosion of their tax base, develop-ing innovative ways to circumvent state laws that grant exemptions from localgovernment taxation, in order to collect revenue from nonprofit organizations.For example, cities are withholding zoning approval and building permits fornew buildings unless the nonprofit university, hospital, or symphony orchestraagrees to pay a "voluntary" tax (often termed payments in lieu of taxation, orPILOT). Tension between nonprofits and governments is becoming so intensethat lawmakers are searching for reasons to withdraw tax-exempt status alto-


gether. This tension has escalated as evidence has surfaced that certain non-profit executives were receiving levels of pay that were unseemly for charita-ble organizations. When Americans learned of compensation of over $500,000per year for nonprofit organization executives such as the president of UnitedWay of America, William Aramony, the chief executive of the PTL Ministry,the Reverend Jim Bakker, and the president of Adelphi University, Peter Dia-mondopoulos (Hancock 1996), most were shocked. Both Aramony and Bakkerwent to prison, but in the process the public faith in nonprofits was shaken(Gaul and Borowski 1993; Arenson 1995a). Nonprofits are increasingly beingseen not as public-spirited philanthropies but as self-serving entities that pur-sue the interests of their top officials and board members.

Another form of tension between nonprofits and government involves theapplication of antitrust laws to nonprofits whose activities have traditionallybeen regarded as benevolent. In 1991 the U.S. Department of Justice broughtan antitrust action against a group of nonprofit universities for price fixing intheir granting of financial aid to prospective students. One of the schools refusedto abandon the practice, contending that the price fixing - which it admitted -actually promoted social welfare because it increased overall opportunities forlow-income students ("MIT Wins a New Trial in Price-Fixing Case" 1993). Ineffect, the university was saying that its collusion with other schools was so-cially valuable, whereas collusion in the private enterprise sector was not. TheJustice Department disagreed. Neither side, however, could call upon solid re-search to buttress its case.2

Nonprofits and government are not always at odds; instances of cooperationbetween these two sectors have been documented as well. For example, non-profits in the United States have discovered money-making opportunities inworking with state governments to market automobile license plates. Thirty-six states now permit some nonprofits to benefit from specialty license platesthat advertise arts organizations, universities, environmental groups, and gar-den clubs. The University of California at Los Angeles (UCLA) is collaborat-ing with the state of California to sell a license plate that finances student schol-arships (Herman 1995). Such cases of what amount to joint ventures betweennonprofits and governments appear to be rare at present, but fiscal pressures onboth types of organization are bound to generate more such novel alliances inthe future.

2 Nonprofit hospitals, too, have fallen victim to antitrust laws, as proposed mergers have been heldto be anticompetitive. Again, public policy hangs on matters about which little is known: wheth-er nonprofit and for-profit organizations use monopoly power in materially different ways. Onerecent study, however, indicates that nonprofit hospitals that gain greater market power are like-ly to reduce prices, whereas for-profits are more likely to raise them (Lynk 1995).


Financing

Many nonprofits face increasing financial pressure because the gap betweentheir resources and what they see as social "need" is growing (see, for exam-ple, Wong [1997] on food pantries). This results from government-program re-trenchment, in part, but the situation varies substantially among segments of thenonprofit sector. For arts and cultural nonprofits, revenues from governmenthave decreased in real terms, increasing at an average annual rate of only 0.8percent in nominal dollars during 1982-92 - considerably below the rate of in-flation (Hodgkinson and Weitzman 1996,12). In the nonprofit health-servicessector, revenues from government have grown in real terms; yet need, as reflect-ed by the number of uninsured persons, has grown nonetheless, thereby increas-ing the pressure on nonprofits to find additional resources. "Need" is difficultto define and measure, but if nonprofits search for new revenues, they have fewchoices: to increase private donations and/or to increase income from the saleof goods or services - that is, "commercial" activity.

Increasing donations is not promising for most nonprofits, although it hasbeen rather successful for public television (Chapter 13). In the aggregate, indi-vidual giving has remained rather constant, at around 1.9 percent of personalincome, for over two decades; between 1973 and 1994 it has ranged between1.77 percent and 2.00 percent, with no apparent trend (AAFRC 1995,48). Al-though some new and imaginative forms of appeals for donations are being de-veloped, there is little reason to expect that donations can be increased signif-icantly unless tax laws are substantially changed. Such tax-law changes arebeing discussed, particularly the granting of tax credits to encourage charitabledonations, but their future is beyond the control of any nonprofit organization.

New commercial activities are the major path open to nonprofits to gener-ate additional revenue; and once nonprofits enter the realm of finding salableoutputs, they are in the domain of private enterprise, where selling goods andservices is the preeminent source of private-sector revenue. Nonprofits that pur-sue revenue in the same ways that private firms do are likely to emulate thosefirms, and by becoming more like them may undermine the fundamental justi-fication for their own special social and economic role. To what extent this oc-curs is important for public policy, since the justification for the subsidies andtax exemptions to nonprofit organizations hinges on nonprofits' differentiationfrom private firms. When nonprofits enter a new industry in search of revenues- for example, as hospitals are doing by establishing profit-seeking athleticclubs - it is natural to ask what distinguishes their weight rooms, saunas, andfitness clubs from those at traditional for-profit gyms and, hence, what justifiestheir tax advantages (Lambert 1997). It should be noted, though, that if a non-profit becomes more commercial in its pursuit of revenue, it does not neces-sarily imply a forsaking of "core" values or mission.


A significant source of income for many nonprofit organizations is user fees,charges levied on people who use elements of nonprofits' mission-related out-puts that can be sold. Universities can charge tuition, hospitals can charge pa-tient fees, museums and zoos can charge admission. As Chapter 4 emphasizes,however, user fees are a "double-edged" sword: The revenue helps with the fi-nance problem, but the prices also unintentionally keep away some personswhose use of the services constitutes one of the organization's social goals. In-creased tuition, patient fees, and admission charges inevitably price some in-tended consumers out of the market. One potential way to deal with this dilem-ma is to discriminate in price, charging lower prices or nothing at all to the targetpopulation of "deserving" consumers, while charging others higher prices.

In the process of instituting user fees, nonprofits are increasingly using tech-niques out of the private-enterprise "playbook." Universities are pushing up tu-itions (prices) at rates that have captured the front cover of Newsweek with theattention-getting "$1,000 a Week: The Scary Cost of College" (April 29,1996).They are increasingly engaging in price discrimination - charging differentialprices by fine-tuning student financial aid to be more competitive - and theyare negotiating with parents over those financial-aid packages (Asinof 1997).The percentage of freshmen receiving tuition discounts has grown at large col-leges, from under 50 percent in 1990 to 60 percent in 1996, and the increasehas been similar at small colleges (Passell 1997). Museums and zoos, which tra-ditionally charged no admission at all, so as to maximize access, have droppedthat policy wholesale, with few remaining free to everyone (Chapter 11).

User fees are not the only evidence of growing commercialism among non-profits. Universities, for instance, are reaching out to new markets - pursuingunconventional students such as senior citizens and people seeking weekendeducation (NPR 1997), and carefully scrutinizing opportunities to find otherthings they can sell profitably. They are even offering souvenir parts from sportsstadiums that are being torn down: Notre Dame is selling bricks from parapetsat its stadium, Indiana University has sold pieces from its basketball court forplaques, and Princeton is auctioning team benches and ticket-window signs(Fairclough 1997). Hospitals, too, are doing more than raising the daily hospi-tal fee (approaching, and even surpassing, $1,000 per day). They are also reach-ing out to new markets - not only expanding into activities such as the fitnesscenters mentioned above, which have little to do with traditional patient care,but moving into new patient care activities that they had chosen to forgo untilfinancial pressures mounted. Thus, hospitals are opening potentially highlyprofitable heart transplant units, and with speed that has generated charges of"unwise procedures," high mortality rates, and "potential malpractice" (Burton1996); and they are "racing to develop and market equipment for . . . mini-mally invasive heart operations, vying for . . . a lucrative market." At least onehospital has put up billboards advertising the new procedure (Winslow 1997a).


Commerciallike behavior by nonprofits is extending to labor markets as wellas to output markets. Thus, nonprofit arts organizations and symphony orches-tras have been accused of succumbing to union pressures to generate high sal-aries even at the expense of higher admission prices and reduced markets; thearguments have suggested that these nonprofits are acting like private firms rath-er than like public-serving organizations (Ritenour 1997).3 As we have alreadyillustrated, nonprofits are also engaging in a growing array of new, ancillary ac-tivities designed simply or primarily to raise money. In the process they appearto be emulating private firms, finding ways to attract more paying customers.Prospective contributors are being treated increasingly as objects of psycholog-ical analysis to discover their motivations (Mclver 1995b). Universities, hun-gry for applicants, are offering guarantees of jobs after graduation (Quintanilla1996). The winds of change and of commercialization are everywhere.

This increase in commercial activities in the nonprofit sector raises the ques-tion of whether nonprofit organizations are merely "for-profits in disguise"(Weisbrod 1988), as Chapter 14 posits. To consider them as such might well bea mistake, however, because there is some evidence (presented in other chap-ters) that the way nonprofits use their opportunities to sell services reflects theirsocial-service missions to reach particular target populations, not simply tomaximize profit. There is also evidence that when contrasted in the same indus-try, nonprofits and for-profits behave in systematically different ways. For in-stance, pricing, including price discrimination, by nonprofits can be used notsimply to generate revenue, but to achieve the organization's distributional mis-sion (Chapter 4). Thus, while nonprofits may charge user fees to generate rev-enue to support their preferred, mission-oriented activities, they may and do es-tablish price below marginal cost - even at zero - for certain consumers, suchas the homeless, schoolchildren at zoos (Chapter 11), or indigent sick peopleat hospitals (Chapter 8). It is not clear, though, whether universities are actingdifferently than would a profit maximizer when they offer price discounts tostudents who use services during off-peak hours (weekends, evenings, and sum-mers). It may be that they are operating at maximum profitability in these mar-kets, applying a form of marginal cost pricing; but they could also be operatingless profitably, or even unprofitably, in an attempt to appeal to certain under-served groups (Arenson 1996).

Also arguing against the for-profits-in-disguise theory is the evidence thatnonprofit and for-profit organizations behave in systematically different waysin markets where they coexist. In the day-care industry nonprofits have beenfound to have more experienced teachers, to be more trusted by consumers, andto encourage more parental involvement as volunteers in the classroom (Mauser

3 Why this type of behavior is occurring at this time is not clear. The direct pursuit of additionalrevenue, which may explain much of observed commercialism, does not seem to apply here.


1993). In the nursing-home industry and the mentally handicapped facilitiesindustry, nonprofits have been found to have significantly lower prices; to pro-vide more labor per patient in such forms as registered nurses, dieticians, andvolunteers; to make greater use of waiting lists; and to generate greater consum-er satisfaction (Weisbrod 1998). In a number of industries within the health-care sector it has been found that nonprofits provide greater access to personsregardless of ability to pay (Schlesinger 1998), although that differential appearsto be decreasing in response to fiscal pressure. In the hospital industry it hasalso been found that nonprofits have significantly lower yearly variation in prof-it rates, which has been interpreted as evidence that they pursue non-profit-maximizing goals, seeking instead to maximize organization utility subject toa profit constraint (Hoerger 1991).

Such findings of differential organization behavior suggest, but do not nec-essarily prove, that when financial constraints allow, nonprofits do behave in afundamentally different manner from for-profit organizations. Nonprofits do ap-pear to exhibit greater willingness to deviate from profit-maximizing choicesin the interest of other goals, such as serving particular target groups, engag-ing in basic research, or providing other public-type services.

The entwining of mission and revenue sources

Nonprofit organizations confront a dilemma, as does public policy toward them:how to balance pursuit of their social missions with financial constraints whenadditional resources may be available from sources that might distort mission.When the mission is to provide collective services of the sort typically identi-fied with government, but with taxing power not available, what are the conse-quences of using other forms of financing? When, as is increasingly the case,the finance mechanisms involve selling other things to finance the collectiveoutputs, what are the effects? Can nonprofits simultaneously emulate privateenterprise and yet perform their social missions? Are public-interest goals andprivate-enterprise money-raising measures compatible? Nonprofits are gener-ally well aware of the issues, as is clear from the industry studies in Part II. Thedilemma was voiced recently by I. Michael Hey man, Secretary of the Smith-sonian Institution, who, while discussing the decision to seek "corporate spon-sorships" to raise additional money, noted that maintaining the nonprofit's "in-tegrity" was important, but that "the costs are outweighed by the benefits" (NPR1996). Whether "costs" are seen in terms of adverse effects on an organization'smission or on prospective donors' perceptions of such effects on that mission,or both, is an open issue; that is, the actual impact of a particular fund-raisingmechanism on organizational behavior may differ from that perceived by out-siders.


The sources of financial support for all organizations, whether nonprofit, for-profit, or governmental, are crucial determinants of the organization's behavior- the kinds of goods and services it produces, the combinations of labor andcapital it uses to produce them, and the ways it distributes the outputs. An or-ganization that is completely dependent on revenue from sales of goods andservices - "commercial" activities - will produce only things it can sell profit-ably, will use production processes that minimize production costs, and will pro-vide outputs to all who are willing to pay more than marginal cost of produc-tion, giving nothing away unless that generates sufficient revenue from othersources. This characterizes the private firm.

An organization that depends entirely on revenue from taxation respondsto a different set of incentives. Thus, a government agency that is financed ful-ly through taxation can be expected to produce services that are determinedthrough the political process, not via private markets. Moreover, it can distrib-ute its outputs to the target consumers at a zero price, since it needs no addi-tional revenue.

A "pure" nonprofit organization may be thought of as dependent entirelyon yet a third source of revenue: donations, in the forms of contributions, gifts,or grants. In the extreme case, with donations coming exclusively without"strings" attached (except that they must be put to the nonprofit's tax-exemptpurposes), the organization can distribute output as it wishes. This is similar tothe distributional situation for government agencies, but the goals of the non-profit may well differ from those of government, and the constraints imposedon nonprofits tend to differ from those that the political process imposes on gov-ernment.

Nonprofits can also be expected to differ from private firms and possiblyfrom government in the way they combine labor and capital in the productionprocess. Nonprofits receive enormous amounts of volunteer labor: The vast ma-jority of the total of nearly nine million full-time equivalent volunteers went tononprofits, where they constituted the equivalent of 36 percent of total employ-ment in 1994, compared with 9 percent in government (largely in schools),whereas in private enterprise volunteers are a negligible percentage of total em-ployment (Hodgkinson and Weitzman 1996, 13). There is also evidence thatnonprofits are able to employ paid labor at lower prices than private firms; suchreduced labor supply prices constitute a kind of donation, one that because ofits form does not appear in data on charitable giving. As a result of such lowerlabor supply prices and volunteered labor, nonprofits can be expected to usemore labor-intensive production processes than would an otherwise compara-ble private firm.

To say that private firms, government agencies, and nonprofit organizationstend to rely on distinct sources of finance - sales for the firms, taxes for the


agencies, and donations for the nonprofits - is not to say that any of these typesof organization relies on a single source.4 Our focus on nonprofits is designedto highlight the extent to which large changes are occurring in the economy inways that affect nonprofits' ability to pursue their missions. Donations from pri-vate sources are of declining relative importance, as Table 1.1 shows: Where-as nonprofits in 1964 received private contributions equaling over 53 percentof their total operating expenditures, that number shows a large and continuousdecline to less than 24 percent in 1993. Volunteered time has also declined inrecent years, from 9.2 million full-time equivalent workers in 1989 to 8.8 mil-lion in 1993 (Hodgkinson and Weitzman 1996, table 2.13). With the change inrevenue sources come changes in organizational performance as nonprofits ad-just. It should be noted, though, that not enough is known to be certain aboutthe direction of causation - whether changes in contributions of time and mon-ey are the cause or the result, or both, of changes in nonprofit organizations'activities.

What finance options are open to nonprofits? The central distinction thatneeds to be emphasized between "donations" and "sales" is the degree to whichthe recipient organization must relinquish control over its activities in order tosatisfy the party offering the resources. Chapter 7 discusses this point, highlight-ing control as the key element of "conversions" of nonprofits to for-profit legalstatus. Any organization selling its output can succeed financially only by of-fering outputs for which buyers - be they individuals, firms, governments, ornonprofit organizations - are willing to pay. In that sense the seller's ability tosucceed in pursuing its goals is constrained by buyers' wants.

Similar constraints sometimes accompany "donations." The offer of a dona-tion, without restriction, to support a nonprofit's organization mission does notconstrain organization behavior in pursuit of its goals, but not all donations areunconstrained. Yale University, for example, recently returned a major "gift"because its purpose was to support a program that, in the end, was not endorsedby the university administration and faculty. Similarly, Grove City College, inPennsylvania, turned down the opportunity to receive financial help through thefederal government student loan program because the college is not willing toaccept at least some of the restrictions imposed by the seven thousand regula-tions delineated under laws authorizing student grants and loans (Ritter 1996);contributions from governments, just as from private sources, can restrict an

4 Henry Hansmann has distinguished between "donative" and "commercial" nonprofits, reflectingthe difference between those that are dependent primarily on one or the other source of revenue(Hansmann 1980). However, there are nonprofits at virtually all level of dependency on dona-tions from one source or another, and there are distinct differences among industries. For non-profits engaged in legal aid, 97 percent of their revenues have come from "contributions, gifts,and grants"; for civil rights nonprofits, 65 percent; welfare nonprofits, 43 percent; cultural non-profits, 27 percent; health nonprofits, 8 percent (Weisbrod 1988,76 [table 4.2]).


Table 1.1. Private contributions as percentage of all nonprofitoperating expenditures, 1964-93

Year % Year % Year %

1964 53.5 1974 34.7 1984 30.61965 52.8 1975 33.1 1985 29.71966 50.4 1976 33.8 1986 32.01967 48.8 1977 34.2 1987 28.51968 47.7 1978 33.0 1988 27.31969 45.3 1979 32.7 1989 26.91970 41.5 1980 32.3 1990 25.51971 41.5 1981 32.4 1991 24.51972 39.1 1982 31.0 1992 23.61973 36.8 1983 30.3 1993 23.6

Source: Hodgkinson & Weitzman, Nonprofit Almanac 1996-1997. San Francisco:Jossey-Bass, 1996, tab. 2.1.

organization's pursuit of mission. Moreover, the effects of such restrictions arenot limited to donations to private nonprofit organizations; public organizationsalso are affected. The University of California at Berkeley recently balked ataccepting a multimillion-dollar grant from a Taiwan foundation because it re-quired the university to commemorate a former president of Taiwan who hadbeen blamed for "repression of thousands of political dissidents" (Golden1996). The fact, however, that the university still considered whether to acceptthe funds, despite the constraint they carried, points up the vagueness of manynonprofit and public organization missions, which makes it easy to debatewhether the package of the funds and constraints would or would not be ac-ceptable.

In short, when there are strings attached to donations, those revenues are es-sentially akin to sales. Unrestricted donations can be used to advance the organ-ization's mission, and so they are preferred by the organization to revenues thatentail restrictions. Whether restrictions are imposed by prospective "donors"or by buyers of goods, however, the effect is the same. If an organization mustproduce some ancillary output for sale in ordinary commercial markets to gen-erate revenue (as when a hospital sells laundry or laboratory testing services)or if an organization must tailor its production to a donor's wishes (as when anarts organization must forgo politically controversial subjects in order to get agrant from the National Endowment for the Arts [NEA]), funding poses a prob-lem for the pursuit of "mission" as that is seen by the nonproflt's trustees andleadership. There is likely to be conflict over whether the advantages of having


the additional funds to pursue the mission exceed the disadvantages of the con-straints. (Of course, restrictions do not necessarily constitute constraints: Anorganization may wish to use even unrestricted marginal funds in the samemanner as that prescribed by restricted funds.)5 For private firms, benefits andcosts have a single metric - organizational profit; for nonprofits and publicagencies, the mission metric is more complex. For the NEA, for example, thechoice is how to deal with Congress's displeasure with some past grants, whichimplies that future funding - "contributions" - may depend on the willingnessof the NEA to alter its view of its mission (Fritsch 1996).

Nonprofit organizations are becoming more dependent on commercial activ-ity in one form or another, and less on contributions from either private or gov-ernment sources, but all industries have not been changing equally. Cost inter-dependencies between mission-related and ancillary activities, for example, canlead to varying commercial profit opportunities being available to nonprofitsin different industries. Table 1.2 shows considerable industry variation in thegrowth of sales revenue from "program services" in recent years, although pro-gram services encompass user fees from both mission-related and ancillary ac-tivities. For nonprofits in health care, the share increased to 89.3 percent of to-tal revenues in 1992 from 86.7 percent in 1987, and it soared to 35.5 percentfrom 16.1 percent for environmental/animal-related nonprofits. By contrast, thismeasure of commercialism actually decreased for nonprofits engaged in inter-national and foreign affairs, while it remained stable in the arts, culture, andhumanities sector.

The potential dilemma for a nonprofit organization involves its need for re-sources but the likelihood that the methods used to generate those resources willrequire sacrifice of control over their use. No source of revenue has unambig-uous effects. While some donations are unrestricted, others limit recipient con-trol, as illustrated above. At the same time, "sales" (or "program service" rev-enues), while highly restricted insofar as buyers must be satisfied or they willnot buy, can generate profit that can then be used freely in pursuit of a non-profit's mission. Thus, the distinction made in a number of the chapters betweena nonprofit organization's revenue from donations and from sales is a proxy:The conceptual distinction is between revenue that is unconstrained, facilitat-ing advancement of the nonprofit's mission, and revenue that is constrained,forcing the nonprofit to choose between forgoing the funds or distorting its so-cial mission.

5 Constraints on how contributions are used may actually augment the supply of funds. This couldoccur if informational asymmetries (i.e., donors not knowing how funds would be used) wouldcause donors to give less unless the use of funds was appropriately restricted. A university alum-nus, for example, might wish to earmark a gift for financial aid, to be certain that the money wouldnot go for a football-stadium Scoreboard. At the same time, the fungibility of money can havea net effect quite different from that reflected in the donor's restriction.


Table 1.2. Sources of funds, nonprofit sector industries, 1987 and 1992

Industry0

Arts, culture, and humanitiesEducationEnvironmental/animalHealthHuman servicesInternational, foreign affairsAll nonprofits*

Privatecontributions (%)

1987

38.314.235.83.8

18.254.211.6

1992

38.713.330.23.3

19.158.310.2

Governmentgrants (%)

1987

11.011.98.32.4

25.526.87.9

1992

10.511.06.52.3

24.426.67.5

Programservices (%)

1987

26.756.816.186.743.513.269.1

1992

26.560.835.589.346.4

8.373.5

Note: Row amounts do not sum to 100% for either year because of the exclusion of "other" sourcesof revenue."Numbers of organizations for each industry are not given in the source.includes industry sectors not included above, which total 6.0% of total revenues in 1987 and 5.8%in 1992.Source: Adapted from Hodgkinson & Weitzman, Nonprofit Almanac 1996-1997. San Francisco:Jossey-Bass, 1996, tab. 5.11.

The issues are subtle. The scope of nonprofits' missions is generally definedso broadly that it is often unclear whether a particular fund-raising mechanismwill or will not be constraining. Moreover, even when nonprofits engage in"unrelated" business activities, they often argue that these activities, while pro-viding needed funds, do not distract the nonprofit from its social mission (seeChapter 10).

The contrary view, difficult either to support or refute, is that the simultane-ous pursuit of a nonprofit's mission - to provide certain privately unprofitablebut socially desirable services - and the provision of profit-making ancillary ac-tivities poses inevitable conflicts. Such a claim is that the two processes - oneto produce and distribute the mission-related good, the other to raise the fundsto finance it - cannot be entirely separated; they are interdependent.

The likelihood is great that cost interdependencies will be present betweenthe mission-related and ancillary outputs. The reason is that the search for an-cillary services that can be sold at a profit is likely to involve services that usemany of the same resource inputs as do the mission-related activities. The sameworkers and managers may devote their time to both, and they are likely toshare the same capital - buildings, space, and equipment. For example, sharingof inputs is involved when a university uses its engineering laboratories for bothstudent instruction and commercial testing services for private industry. Chap-


ter 5 reports estimates, for example, that added ancillary outputs have little mar-ginal effect on total managerial costs, whereas added output of mission-relatedactivities has substantial effects.

Sometimes the mission-related and ancillary activities are linked not throughtheir resource inputs but through the consumers to whom outputs are market-ed. Museum shops cater to the people who visit the associated museums, anduniversity bookstores cater to the students whose education constitutes a sig-nificant component of the university's tax-exempt mission.

Private firms also produce and sell multiple goods, and when they do, thegoods are also likely to be related through either their production or sales pro-cesses. What is special about nonprofits is that their choices of goods can beexpected to reflect the tax-exempt status of some of their activities, but not ofothers, as well as the subsidized nature of many of their inputs. Tax consider-ations have implications for the allocation of the nation's resources between the"untaxed" nonprofit sector and the "taxed" private-enterprise sector; this is thefocus of Chapter 5. The competitive tax advantage nonprofits have over privatefirms is greatest when the capital-labor ratio in the organizations is highest,and in geographic areas where corporate tax rates and real property tax ratesare highest.

The differential tax treatment of nonprofits and private firms can bring quiteundesirable social consequences. The case of Bennington College, a nonprofitschool in Vermont, highlights the possibilities. In 1983 the president of Ben-nington College made a discovery: Since Bennington was a nonprofit organi-zation that paid no corporate profits taxes, it did not benefit from being allowedto charge depreciation as a cost of production; private firms, by contrast, didbenefit, since depreciation reduced their tax liabilities. The college had foundan opportunity to gain from exchange between sectors that differed in their taxrules. Soon Bennington agreed to sell all its buildings to an alumni group, for$3 million to $8 million annually, and lease them back for 99 years. "In effect,the college would get an interest-free loan and the alumni would reap tax ben-efits" (Biddle and Slade 1983). Both the college and the real estate purchaserbenefited - as inevitably occurs when organizations face differential prices forthe same commodity (Galper and Toder 1983).

Not everyone gained, however: Taxpayers lost, as tax revenues declined.In effect, tax benefits were being sold. Congress subsequently prohibited such"sale-leaseback" arrangements, which have no economic justification but to re-duce taxes. Nevertheless, the potential for mutual gain for the taxed and untaxedsectors remains. Other novel exchanges between the nonprofit and private en-terprise sectors can be expected. After all, since organizations in both sectorscan benefit from the differential rules under which they operate, the incentiveto find ways to overcome the barriers to trade across sectors remains.

The existence of taxed and untaxed sectors also provides an incentive for or-ganizations to begin as nonprofits and later convert to for-profit status. Although


generally restricted by law, the incentive can be enormous, since nonprofits arelegally precluded from distributing their profits and surplus to managers or di-rectors. If a conversion can be arranged - as has been occurring with dramaticfrequency in the hospital and HMO industries - the potential for private gain isgreat; the consequences with respect to the nonprofits' social mission, how-ever, is also critical, and the breadth of missions make it difficult to discernthe validity of a claim that commercial activity distorts nonprofits' behavior to-ward nonmission activities. Analysis in Chapter 7 of the conversion process andthe associated incentives shows how the massive private benefits can obscurechanges that have important consequences for nonprofits' missions.

Once it is recognized that nonprofits' social missions involve collective ser-vices that must be financed primarily out of some combination of donations(private or public) and revenue from sales, a number of elements of nonprofitorganizations' behavior become more clear. Consider, for example, the ques-tion of whether hospitals should continue to be granted nonprofit charitable sta-tus, with the accompanying subsidies, in light of the fact that the vast majorityof their revenue is raised through sales - providing care to persons who pay,either directly or, more commonly, through insurance, just as patients do in for-profit organizations (Chapter 8; see also Clark 1980). Similarly, nonprofit uni-versities provide educational services to students, many of whom pay full tui-tion. Consumers who pay more than the marginal cost are, in effect, subsidizingthose who do not.

However, the social mission of a nonprofit, in any industry, should not bethought of as to do what a private firm would otherwise do. In the case of hos-pitals, the mission is to care for the indigent, to undertake research that gener-ates widespread knowledge, and to provide such community services as edu-cation about drugs and maternal nutrition. In the case of universities, the socialmission is the provision of basic research, education to the poor, and dissemi-nation of information; these are the candidates for public encouragement.

From one source or another, such collective goods - socially valuable butprivately unprofitable - must be financed. As noted earlier, contributions fromthe private sector and government are important sources (Gr0nbjerg 1993).They are preferred insofar as they are unconstrained. Profit from sales to con-sumers - for instance, hospital patients or college students - who pay more thanthe marginal cost of their services is another. Nonprofits' tax exemptions areunderstandable as an indirect way of financing private provision of collectiveservices.

Public policy and the financing of nonprofits9 mission

Attention to nonprofits' large and growing commercial activity is necessary be-cause it highlights both the differences between what society expects from non-profits and private firms, and the similarity in the financial constraints that they


face. By focusing on the collective-good mission of nonprofits and their inabil-ity to levy taxes to finance that mission, this book points to a likely contradic-tion in public policy: The social goals of nonprofits may well be incompatiblewith the instruments society wants them to use to attain those goals.

More public pressure is being placed on nonprofits to increase their activi-ties to offset government cutbacks. At the same time, pressure is increasing onnonprofits to avoid commercial revenue-raising devices because they involveeither competing with private firms or joining forces with them, both of whichpose problems.

Society cannot have it both ways. We cannot expect nonprofits' outputs toincrease if their access to revenue is restricted. At present, public policy is per-missive in its regulations as they affect nonprofits' access to commercial mar-kets. However, the disadvantages of having nonprofits act increasingly like pri-vate firms are considerable. Public-policy makers need to understand how theability of nonprofits to achieve their social goals is related to their access to rev-enue and, more particularly, their involvement with the sale of private goods.The challenge for public policy may well be to find alternative revenue sourcesthat would make it possible for nonprofits to act less like private firms, at leastin terms of their choices of outputs and target populations, and to concentratemore on their mission-related activities.

This public-policy theme will recur in one chapter after another, with anoverview in Chapter 15. For now it may be sufficient to note that there are dan-gers of commercialism when provision of collective services is the objective ofpublic policy. Just as there "must" be a better way to finance political electionsthan to "sell" access to the White House Lincoln Bedroom, so too there mustbe a better way to finance great nonprofits such as the American Museum ofNatural History, in New York City, than by converting it, albeit only one morn-ing per week, "into a kind of Jurassic Gym, an exotic health club for an eliteband of mostly Upper West Siders who . . . power walk their way past the dio-ramas" - in return for a contribution of at least $100 (Sontag 1997).

Understanding nonprofits9 commercialism -the chapters ahead

The chapters that follow are grouped in three parts. Part I, "Basic Issues andPerspectives," deals with broad matters that apply to nonprofit organizationsgenerally. Part II comprises industry studies that portray the varied missiongoals and financial bases, and their changes. Part III provides an overview ofour findings and their implications for research and for public policy towardnonprofits.

More particularly, Chapter 2 sets the stage for examining the interplay ofnonprofit organization mission and revenue sources. It shows how commercial


activity by nonprofits typically involves increased competition with for-profitsand with other nonprofits. It also considers how competitive commercial activ-ity affects nonprofits' ability to pursue their social missions.

Chapter 3 sets forth our conceptual framework. Viewing nonprofits as hav-ing missions to provide certain socially preferred services, it examines the ques-tion of financing nonprofits at a theoretic level, highlighting the relationshipsbetween a nonprofit's social mission-related activities and its ancillary activi-ties undertaken purely as revenue providers. Chapter 4 explores the issue of userfees, examining the differences between the kinds of discriminatory pricing thata profit-maximizing firm would tend to employ and the kinds that a nonprofitwould use if it is concerned about social goals or reaching particular target pop-ulations, rather than profit. For example, a for-profit firm would be less likelyto admit large classes of unprofitable consumers for free - as many nonprofitmuseums do with school groups - or to cut prices for undergraduates who arehighly able but poor, when less able but wealthier students who would pay ahigher tuition price are turned away - as nonprofit universities commonly do.

Chapter 5 shows how the differential tax treatment of commercial activitiesthat are either "related" (and thus tax-exempt) or "unrelated" (and taxable) tothe nonprofit's mission creates incentives that are sometimes surprising and of-ten unintended. For example, it explains the anomalous situation in which non-profits, overall, actually report that they are suffering losses in the ancillaryactivities that, while unrelated to their tax-exempt mission, have as their onlyjustification that they are generating profits in support of the tax-exempt mis-sion! In addition, this chapter explores the forces at work when nonprofit andfor-profit organizations confront each other, but also when the clash betweensocial mission and the lure of profit occurs within a nonprofit organization thatengages in both mission-related and unrelated activities.

Utilizing the theoretic model of Chapter 3, Chapter 6 explores the causalconnections between unexpected, exogenous changes in a nonprofit organiza-tion's revenue from its preferred source, donations, and that organization's sub-sequent commercial activity. Data from the IRS for the period 1982-93 are uti-lized to estimate the effect of changes in donative revenue on subsequent levelsof sales revenues - whether from mission-related user fees or from the sale ofunrelated goods and services. This chapter examines a unique, decade-long pan-el data set for nonprofit organizations in the hospital, higher-education, and mu-seum industries.

Chapter 7, the final installment of Part I, focuses on what we have termedthe "ultimate commercialism of nonprofits": their conversion to for-profit form.Increasingly common in the hospital and HMO areas, such commercializationposes severe public-policy challenges. This tension between the opportunity forprivate gain and the public-policy mission for nonprofits - social gain - is ex-amined. A conversion to for-profit status may simply be privately beneficial to


insiders, or an efficient reallocation of resources, or both; the challenge is toavoid the former. Particular attention is directed to the difficulty of ensuring thatthe assets of the nonprofit that is converting are used in ways that are consis-tent with public policy, since tax benefits contributed to the amassing of thoseassets.

In Part II the framework presented in earlier chapters forms the foundationfor studies of commercialism in hospitals, universities, social services, zoos andaquariums, museums, and public television. It becomes clear that the depen-dence of nonprofit organizations on revenue from sales varies greatly, as doesthe extent to which commercialism is growing in various industries. As the anal-ysis in Chapter 3 suggests, opportunities for profitably expanding commercialrevenue differ greatly across industries.

Having focused on the constraints that commercial revenue can impose onnonprofits' public-serving mission - which is the ostensible justification for thesubsidies and tax exemptions - and having highlighted the tension between theneed for revenue to foster that mission and the danger that revenue-raising activ-ities will compromise the mission, we turn in Part HI to an assessment of whathas been learned. Chapter 14 considers two alternative interpretations of theactual role of nonprofits: one that emphasizes the importance of social mission,another that emphasizes the role of private self-interest. Chapter 15 concludeswith a summary of the book's findings and interpretations. It also identifiesneeds for future research and the public-policy implications of our findings.

PARTI

Basic issues and perspective

CHAPTER 2

Competition, commercialization, and theevolution of nonprofit organizationalstructures

Howard P. Tuckman

Introduction

Competition among nonprofits and between nonprofits and for-profits is apresent day reality. What is its impact on nonprofit behavior in the marketplace?To what extent does it lead to commercialization of nonprofit activities? Doesit cause nonprofits to alter their organizational structures or for-proflts to altertheir services to copy nonprofit behavior? What effects does it have on non-profits' ability to pursue their charitable missions?

This chapter analyzes competition among nonprofits in settings where onlynonprofits compete, and in markets where both nonprofits and for-profits co-exist. It begins with a definition of competition and then presents five com-petitive forces and uses these to examine competition in a market where onlynonprofits compete. The conditions under which commercialization is likely tooccur are discussed, and the nature of competition in settings where nonprofitscompete with for-profits is explored. Examples from the health-care industryare used to explore the evolving legal structures that nonprofits are employingin settings where they compete with for-profits. The chapter ends with a discus-sion of the challenge that these changes pose for public policy: to ensure thatthe evolving institutional changes, as well as the pressures toward commercial-ization, do not diminish the unique charitable role of the sector.

The nature of nonprofit competition

Competition, the pursuit of the same objective by two or more firms, creates ri-valry among nonprofits for capital, labor, customers, and/or revenues. This ri-valry is normally not motivated by a desire for personal acquisition of profits,

The author acknowledges useful ideas from Richard Steinberg and Carol Babb.

25

26 Howard P. T\ickman

although equity accumulation may become a goal of an organization (Changand Tuckman 1990; Brody 1996,491). Nonprofits vie with each other for rev-enues, board members, customers, contracts and grants, donations, gifts and be-quests, prestige, political power, and volunteers (Brody 1996,469). Their com-petition with other nonprofits for customers may involve the production ofhigh-quality products, contesting of market share, or a concerted effort to holdthe best reputation. Nonprofits also compete for alliances with for-profit andgovernment entities (Weisbrod 1996).

It is helpful to use the term basis for competition to refer to the way that com-petitors choose to compete to meet their customer/constituent needs. Basis forcompetition changes depending on the industry and the needs of the consumer.In the for-profit world, burger chains compete on the basis of who can producefood most quickly and on price; amusement parks on the latest and most excit-ing new rides; discount stores on price and variety of inventory; and conve-nience stores on location and speed in getting customers in and out of the store.In the exclusively nonprofit world, competition frequently is based on quali-ty of service, ability to meet constituent needs, and reputation. For example,disaster-relief agencies usually compete on speed of response to an emergency;nonprofit hospices on quality of comfort offered to patients; and advocacy or-ganizations on how well they articulate the position of their members, and onaccess to media and public officials.

Nonprofits exist in a variety of settings, some involving competition withother nonprofits and others competition with for-profits. Limited situations ex-ist in which nonprofits offering food to the poor face for-profit competitors; in-stead, competition is normally with other nonprofits. In contrast, nonprofit hos-pitals vie with for-profit competitors in many marketplaces. Though the relevantsetting can be the local marketplace, national data are helpful in illustrating thedifferent competitive settings faced by nonprofits.

Table 2.1, which uses the Independent Sector definition of nonprofit organi-zations and is compiled from several national sources, shows the wide varia-tion in the importance of nonprofit organizations in various industries, rangingfrom single-digit market share in legal services and radio-TV broadcasting toover 50 percent share in hospitals, museums, and job training. Note too that theimportance of nonprofits has increased between 1982 and 1992 in some mar-kets (e.g., nursing and personal care) and declined in others (e.g., child day-careservices).

Commercialization of nonprofits occurs when these organizatons decide toproduce goods and services with the explicit intent of earning a profit. The deci-sion to commercialize normally leads a nonprofit into a mixed-mode market-place where it competes with for-profit organizations. Such competition can re-sult either if nonprofits enter a market where for-profits already exist or if theyare successful at creating a product that for-profits later choose to offer. Non-

Competition, commercialization, and organizational structures

Table 2.1. The importance of nonprofit organizations in the independentsector: Selected industries, circa 1982 and 1992

27

Category

Nursing and personal care facilitiesHospitalsElementary and secondary schoolsColleges and universitiesReligious organizationsLegal servicesIndividual and family servicesJob training and related servicesChild day-care servicesCivic, social, fraternal organizationsRadio and TV broadcastingProducers, orchestras, and entertainersMuseums, botanical, zoological gardensFoundations

Pet. of organizationsin independent sector^

ca * 1992

28.451.423.546.0

100.01.1

79.769.831.1

100.03.3

23.486.6

100.0

ca. 1982

21.351.0

8.751.5

100.01.1

88.579.441.2

100.05.1

19.394.3

100.0

Pet. of revenuesin independent sector

ca. 1992

30.966.024.834.6

100.01.1

90.583.341.2

100.03.2

24.694.3

100.0

ca.1982

26.963.49.1

31.9100.0

1.394.384.052.1

100.04.3

25.097.5

100.0

aAs opposed to the for-profit and government sectors.^Indicates that the exact date of the data varies by category.Source: Adapted from Hodgkinson & Weitzman, Nonprofit Almanac 1996-1997. San Francisco:Jossey-Bass, 1996, tab. 4.7, p. 201.

profit hospital entry into the medical-supply business provides an example ofthe former, whereas the entry of private health clubs into markets once domi-nated by YMCAs and YMHAs illustrates the latter. We shall deal with institu-tionally mixed industries later in this chapter; the next section explores compet-itive issues in marketplaces populated exclusively by nonprofits.

Competition in exclusively nonprofit markets

Lack of literature on competition in exclusivelynonprofit markets

Much has been written about the competition among health-care nonprofits inmixed-mode marketplaces and its implications (e.g., Gray 1986,1991) but, withthe exceptions of Steinberg (1987) and Ben-Ner and Van Hoomissen (1991),researchers studying exclusively nonprofit marketplaces have paid scant atten-tion to the effects of competition on nonprofit behavior. This is particularly true

28 Howard P. Tuckman

in the social service, religious, and cultural areas where the literature focusesprimarily on fund-raising issues and the problems associated with revenue ac-cumulation. As Weisbrod (1996) notes, the lack of the power to tax means thatwhen nonprofits face increased demand, they also face pressures to expand theirresource base. Much of the available literature written about nonprofits oper-ating in exclusively nonprofit markets has been devoted to how to accomplishthis, and on the consequences of resource dependency (Gronbjerg 1992,1993).An important concern raised in the literature is that nonprofits can be heavilyinfluenced by the goals and objectives of their major funders and this can af-fect their missions. Although a few authors (e.g., Young 1987) have identifiednonmonetary objectives that cause managers of nonprofits to compete with eachother, little attention has been paid to how competitive forces affect the goalsof nonprofits, their financial condition, their behavior in the marketplace, andtheir ability to carry out their goals.

Nonprofits face many competitive forces similar to those that affect for-profit organizations. For example, both are subject to changes in both supplyand demand. Donations to nonprofits are determined, in part, by the value do-nors place on their services; nonprofits with highly valued missions offeringservices in scarce supply (relative to demand) may find fund-raising easier thando those with less highly valued missions - although evidence on this point isscarce, and measurement of "valued mission" independent of actual donationsis difficult. Similarly, when nonprofits demand large numbers of volunteers,scarcities for this type of labor can develop.

Exclusively nonprofit marketplaces

When will a marketplace be populated solely by nonprofits? Normally, this willoccur when a nonprofit supplies products or services that cannot be financedthrough exchange between buyers and private enterprise or government sellers.Weisbrod (1977) theorizes that nonprofits arise to fulfill the demand of partic-ular groups for the production of "public" or "collective" goods. These goodshave two key characteristics: They are nonrival, in the sense that one individ-ual's consumption of them does not affect the consumption by another, and theyare nonexcludable, in the sense that individuals who fail to pay for them can-not be excluded from their consumption. Nonprofit radio broadcasts, economicdevelopment efforts on behalf of a city or region, and scientific research pro-vide examples of collective services. For-profits find it difficult to exist in suchmarkets because the primary source of their funding would have to be dona-tions and gifts, and donors are reluctant to donate to these enterprises, recogniz-ing the potential for their contributions to flow into organizational profits anddividends rather than increased output (Hansmann 1980).

Competition between for-profits and nonprofits is also unlikely in situationswhere excludable goods or services are produced and individuals receiving the

Competition, commercialization, and organizational structures 29

intended output cannot afford to pay: Housing the homeless, health care for thepoor, and youth services to indigents are cases where competition is likely tobe exclusively among nonprofits. In these circumstances, nonprofits act as fi-nanciers, seeking funds from donors and other sources, often buying goods fromthe for-profit sector and distributing them as charity. United Way provides awell-known example of this, as do Catholic and Jewish charities. Trust issuesnormally preclude for-profits from playing a financier role, primarily becausefor-profits have an incentive to increase profits by diverting donated funds fromthe donor's intended use (Hansmann 1980).

A third market setting comprising solely nonprofits in competition may ex-ist where goods and services could be produced and/or supplied by for-profitsbut where, for various reasons, nonprofits enjoy competitive advantage: whereliability laws impose heavier penalties on for-profits than nonprofits; public pol-icies provide and encourage donations and grants solely to nonprofits; publicadvocacy is detrimental to for-profits; volunteer labor is more available to non-profits, and so on. Public policy can affect the mix of for-profit and nonprofitproviders in these markets through its legal treatment of the respective entities(Steinberg 1993b).

In markets populated exclusively by nonprofits, the basis for competition canbe affected by the preferences of individual donors and foundations, the moresof professional organizations, and other third parties that emphasis charitablepurpose. Shelters for battered women are almost exclusively the domain of non-profits, and this places primary importance on the quality of the counseling re-ceived in these settings, rather than on raising market share or customer-basedrevenue (Ferris 1993). How and which services are delivered, who providesthem, and for what periods of time are determined largely not by the standardof their contribution to profits, but by the professionals and volunteers who in-fluence the way a nonprofit is run.

Similarly, the values and goals of religious nonprofits affect fund-raising, re-cruitment of volunteers, service provision, and so on. Jeavons (1993,72) arguesthat "for religious organizations, the proverbial 'bottom line' is 'faithfulness.'The question is not just whether they provide a particular service and do so well;it is also whether they do so in such a way as to make the love of God and theirown vision clear and visible to others." Hodgkinson et al. (1995, 216) reportthat contributors to religious organizations were far more likely than other do-nors to report that commitment to their religion or spiritual life was a key "lifegoal."

Porter's jive competitive forces in exclusively nonprofit markets

At least five forces influence the nature and degree of competition, as well asits basis (Porter 1980): ease of exit and entry, bargaining power of customers,bargaining power of suppliers, threat of substitute products or services, and ri-


valry among competitors. Porter uses these factors to analyze marketplaces pop-ulated by for-profits only; here we adapt them for use in understanding compe-tition in markets populated exclusively by non-profits, and in those involvingcompetition between nonprofits and for-profits.

Ease of entry. The conditions governing entry into an exclusively nonprofitmarket can affect the types of product produced by nonprofits as well as prod-uct distribution. New nonprofits have an important impact on existing waysof providing customer service. An example is helpful: In the 1960s and early1970s, economic education was provided primarily by university-affiliated cen-ters under the purview of the Joint Council on Economic Education and fund-ed by the Ford Foundation. The emergence of Junior Achievement as a substan-tial competitor in both national and local markets, as well as of other nonprofiteconomic education providers, created an interesting competitive situation inan exclusively nonprofit marketplace. The Joint Council affiliates initially em-phasized mainline economic principles and delivered education to teachers pri-marily through university-affiliated faculty. In contrast, Junior Achievementaffiliates emphasized applied business applications and sent businesspeopleto teach in the classroom. Both organizations sought external funding from thebusiness community, and both competed for position within the schools. Com-petition between the two major rivals changed the product of both organiza-tions, causing Junior Achievement affiliates to develop instructional moduleswith greater economic content and the Joint Council affiliates to provide morereal-world applications, such as modules on housing, insurance, and credit. Al-though the two competitors continued to rely t .1 different delivery modes - inone case teacher training and, in the other, celivery by businesspeople - someJunior Achievement affiliates added teacher trainers to their staffs. The emer-gence of competition in this area of instructional delivery influenced the con-tent and delivery of economic concepts and ideas.

Barriers to entry affect nonprofits, just as they do for-profits. Such barrierscan take the form of high capital investment, as in the case of national consumertesting labs (Consumers Union), a strong speciality in delivering a service (Ed-ucational Testing Service), and/or success in attracting a national donor base(American Heart Association). In each instance, the entry of new competitorsis made difficult by the advantages enjoyed by one or more existing firms: econ-omies of scale, product differentiation, capital accumulation, cost advantages,and access to distribution channels. These limit the ability of new entrants tochange the basis of competition.

All nonprofits face a barrier to entry in the form of IRS regulations that af-fect their ability to qualify as 501(c)(3) organizations. This particular barrier isamenable to change through public policy. Beyond this legal hurdle, barriers toentry differ substantially by market. For example, local advocacy organizations


normally face paltry barriers to entry because of the low costs involved in es-tablishing a presence in the advocacy marketplace. This implies the potentialfor substantial competition among advocacy organizations. Similarly, nonprof-its providing housing services to the homeless likely face low barriers, givenlimited government regulation and few national organizations with which tocompete (Milofsky and Blades 1991). By contrast, nonprofit hospitals and nurs-ing homes encounter barriers imposed by government regulation, such as certif-icate of need and accreditation requirements, although these typically apply tofor-profits as well. In these areas, government policies play an important rolein shaping competition. National nonprofit organizations, on the other hand -such as the National Rifle Association, Red Cross, and American Cancer So-ciety - impose high barriers against new entrants because they take advantageof economies of scale, access to a large contributor base, and relatively readyaccess to distribution channels.

Bargaining power of buyers. Buyer ability to influence competition amongnonprofits increases when buyers are large and few in number, purchases occurin large volumes, products are available from many nonprofit producers, buy-ers of services are price sensitive, and buyers can become future suppliers ofservices. This is true whether buyers are other nonprofits, for-profits, or govern-ments, and these factors affect even exclusively nonprofit marketplaces. For ex-ample, a foundation that purchases mental-health services for the poor from anonprofit at a certain cost per patient has the potential ability to affect the lengthand frequency of the provided services. Similarly, a wealthy donor with fewcompetitors can choose among several providers of homeless services and ex-ercise considerable impact on both the type of accommodations provided andnonprofits' policies in caring for the homeless. As nonprofit hospitals are soldto for-profit hospital chains and their proceeds used to establish foundationsto provide health-care services to the poor (see Chapter 7), these organizationsinfluence competition in the markets in which they finance care (Lutz 1996).

Nonpaying beneficiaries, such as the poor, usually are not a force affectingcompetition. In exclusively nonprofit marketplaces, where finance is entirelyfrom third parties, their direct influence is likely to be limited. In situationswhere nonprofits compete for beneficiaries - whether to increase their marketshare (so that as "market leaders" they are more attractive to potential donors)or to use excess capacity - the beneficiaries may not remain long enough (orhave the interest) to influence competitive behavior. On the other hand, non-paying beneficiaries do appear to exert influence on volunteers by affecting theirbehaviors (Britton 1993)

Large suppliers of capital clearly affect the types of service nonprofits offerand the way they deliver them (Useem 1987). For example, the federal govern-ment affected the basis of competition among social-service and community-


development organizations by specifying which services it funded (Gronbjerg1993). Medicare and Medicaid programs affected competition in health care bycovering some services and not others, establishing payment schedules for dif-ferent types of illness, and creating eligibility standards for health-care provid-ers (Gray 1991). In this instance, public policy plays an important role in shap-ing competition.

Third parties - that is, entities other than the ones who either provide or re-ceive the services - are often the buyers for nonprofits, often through grants orcontracts for services. This is the case for services for the poor such as healthcare, mental health, housing, and job training. In health care, in particular, bothnonprofits and for-profits are increasingly affected by the trend toward large-scale purchase of their services by HMOs and insurance companies (see Chap-ter 8). A few U.S. ballet companies have been blessed with benefactors who,through their donations, have made it possible to compete based on nationallyknown dancers. These donors have considerable influence on who is hired andon the offerings of the companies. Mental-health services are often contractedby federal and state governments, whereas individuals are the primary buyersof religious services. In each case, the purchaser can, and does, exercise influ-ence over goals of the nonprofit sellers and the types of service offered. For non-profits, competition is a process of harmonizing goals involving provision ofcollective goods and goods to the poor with the need to finance them.

Bargaining power of the suppliers of nonprofit inputs. Suppliers affectcompetition in exclusively nonprofit markets through their impact on the pricesof inputs, speed at which services are supplied, quality and quantity of goodsand services supplied, and so on (Steinberg 1990). They affect competitionwhen they are few in number, when their products are differentiated, givingeach some monopoly power, and/or when they can offer a unique advantage tothe nonprofits they supply.

Examples of the impact of suppliers abound. Most nursing homes no longercompete on the basis of the quality of their internal pharmacies since prepack-aged drugs can now be supplied externally at a lower cost, and more conve-niently, by for-profit suppliers. The supply and price of housing available tononprofits affect the quantity and quality of the housing offered to the poor andthe amount of space that nonprofits can provide. Suppliers of computer soft-ware affect how effectively the clients of a legal service for the poor can be pro-cessed, and designers of museum displays affect the quality of the experiencethat patrons have when they visit an exhibition. In each case, the ability of non-profits to compete is influenced by the nature of their supply relationships.

The suppliers of the trained personnel that staff nonprofits can also haveconsiderable impact on the quality of services these entities provide: Nursing

Competition, commercialization, and organizational structures 3 3

schools affect the level of care provided at nonprofit clinics, hospitals, and nurs-ing homes. Likewise, the Catholic schools that produce teachers to staff non-profit universities in developing countries have an important influence on thecurricula taught at these institutions. The quality and quantity of trained volun-teers can also have a strong impact on service delivery. Public policy affectsthese suppliers to the extent that it specifies specific training requirements.

Nonprofits operating in exclusively nonprofit marketplaces can compete insupplier markets, particularly those involving human resources. Hall (1992)notes that the dramatic increase in the number of nonprofits in the past few dec-ades created a demand for competent trustees that far exceeds the supply of per-sons with trustee experience or with understanding of trustee values, leading tointensified competition for competent board members, although the evidence isanecdotal. In support of his view, Brody (1996, 488) suggests that this leadsto a "pecking order . . . with lesser boards serving as stepping stones for moreimportant (visible) boards." Competition can also exist in the labor markets forsenior management at nonprofits (e.g., CEOs of nonprofit hospitals), marketers,fund-raisers, and other skilled professionals (e.g., physical rehabilitation spe-cialists) (Steinberg 1990).

Availability of substitute products or services. Substitutes for services alsoaffect competition in exclusively nonprofit markets by influencing the degreeof monopoly power a nonprofit provider can wield. The fact that substitute ser-vices are available influences the price that nonprofit administrators can chargethird parties and the way they make their appeal to donors. Availability of closesubstitutes also puts pressure on the administrators to reduce program costs.Drug-abuse services are provided largely by nonprofits on both an inpatient andan outpatient basis; these substitutes limit inpatient providers' ability to obtainfunds by raising prices, lest payers shift to outpatient providers, and vice ver-sa. Similarly, the presence of low- (or no-)cost counseling from religious orga-nizations can affect, and is affected by, the prices and types of counseling fromother mental-health providers. In exclusively nonprofit markets, the availabili-ty of substitutes helps to ensure cost consciousness. Public policy can affect thisprocess by encouraging the development of substitute services, as in Medicarereimbursement for home health services.

When the funders of nonprofits are well informed, the existence of substi-tutes competing for support puts pressure on competitors to be efficient in pro-duction and distribution. The widespread TV broadcasts of Sesame Street, forexample, may have affected the type of education provided by such substitutesas early childhood education centers and producers of children's books. Sim-ilarly, the presence of competitive cable TV stations creates pressures for pub-lic TV stations to differentiate their products. In a different context, the avail-


ability of drug therapies influences the willingness of nonprofit and for-profitmental-health organizations to use large amounts of the alternative therapy,counseling, to solve behavioral problems. Awareness of substitutes creates anincentive for nonprofit managers to monitor developments by competitors, aswell as to rethink the value-added they offer. In exclusively nonprofit markets,this serves to keep nonprofits competitive in their service offerings. Public pol-icy can facilitate this process by augmenting the flow of information on alter-natives to consumers.

Rivalry among competitors. In both for-profit and nonprofit marketplaces,attitudes toward rivalry are shaped by many things. Rivalrous political or so-cial ideologies, such as the difference in beliefs between pro-lifers and pro-choice advocates, affect the way that information is produced and delivered byorganizations advocating these positions. This would also seem to be the ori-gin of the competition between the liberal and conservative think tanks in Wash-ington, D.C. Markets initially occupied by a few competitors in the 1950s and1960s (e.g., Brookings Institution, Urban Institute) now house a variety of newcompetitors, such as the CATO Institute, Heritage Foundation, and a host ofsmall policy organizations. The desire of donors to broaden ideological rep-resentation and input into the public-policy-research world has apparentlyspawned a vibrant new industry that bases its competition on access to policy-makers, quick turnaround, and a reputation for quality work. Although researchorganizations with different ideologies are unlikely to have a common base ofpotential and actual donors, they compete for foundation and government grantsand the attention of Congress, the federal agencies, the media, and the publicmind.

Competition also arises as a result of disagreements over how best to pro-vide a particular service, as in the case of nonprofit birthing centers that com-pete with nonprofit hospitals for the business of pregnant women. Differencesin ideas about how best to provide mental health, education, and youth servicesare in part responsible for the rich diversity of organizations (and fundingsources) in the nonprofit sector. Particularly where outputs are hard to measure,as in the case of environmental advocacy, different delivery methods can per-sist for extended periods.

Competition intensifies in the presence of one or a few rivals, a number ofequal-sized competitors all determined to grow, the desire to beat a particularcompetitor, slow industry growth, lack of product differentiation, high fixedcosts, excess capacity, and high exit barriers (Pearce and Robinson 1991). Pub-lic policy can impact these forms of rivalry in a variety of ways - for example,by restricting certain forms of competition, affecting the growth rate of an in-dustry, and/or promoting entry by new competitors.

Competition, commercialization, and organizational structures 3 5

Is nonprofit competition less intense than far-profit competition?

Competition is not necessarily gentler or less intense because it occurs amongnonprofits, although it can be. In the 1990s, for example, the high fixed costsassociated with the provision of inpatient hospital care, together with excess ca-pacity in the industry, made nonprofit hospitals fierce competitors, the religiousroots of many of these organizations notwithstanding (Gray 1991). Intense com-petition for patients occurred even when few or no for-profits were present, andit took a variety of forms as hospitals sought horizontal integration and thenpresence in related markets such as home health care. Intense competition alsoexists in the market for Sunday morning evangelical TV programs where non-profit religious entities go to great lengths (and spend substantial sums) to com-pete for viewers likely to become church members and donors. This compe-tition is fueled by the limited number of desirable time slots, existence of amarket for TV-based delivery of religious ideas, and belief that one denomina-tion has more to offer than another. For the most part, public policy has had alimited impact in this domain, primarily because of a reluctance to get deeplyinto the oversight of the activities of religious groups.

The religious entities that sponsor nonprofits in the counseling, education,health, human service, and religious areas often have multiple impacts on com-petitive conditions. First, they provide an important source of labor and an en-try point for individuals with entrepreneurship skills (James 1987). This en-ables them to affect the resource markets that supply nonprofit organizations,to influence the ideas and skills new entrants bring to the marketplace (impact-ing on service quality), and to provide a source of labor in areas that might oth-erwise have problems staffing nonprofits - for instance, supplying manpowerto teach in rural African and Asian areas. Their ability to enlarge the labor poolis illustrated by the fact that missionary work by the Mormon church has ledto the residency of a large number of foreigners in Utah.

Religious entities also use their access to capital reserves to finance produc-tion of new products and services (Gray 1991). To the extent that they chooseto underwrite an activity irrespective of its costs, competition within an affect-ed industry can become bitter, particularly for organizations that have less ac-cess to donations and volunteer services. The environment is similar to the onethat prevails when, for reasons of tradition or pride, family firms refuse to exita declining for-profit industry (Porter 1980). In these conditions, competitorscan assume a bunkerlike mentality in which each attempts to survive irrespec-tive of the losses involved.

These entities also affect competition through their desire to "best" anoth-er religious group, to reach or serve a specific market segment (often but notalways defined in terms of religious objectives), and/or to deliver a particular

36 Howard P. Tbckman

message. Examples abound of nonprofit hospitals built to enable one religiousgroup to compete for patients with the hospital of a competitive denomination,of nursing homes set up to enable the elderly of a particular denomination topray with their "own kind," of colleges originally created to ensure that educa-tion carries the "correct" religious message, and even of institutions built withthe intention to convert individuals of different persuasions. In Memphis, Ten-nessee, during the 1980s, for example, two major hospital chains - Baptist Hos-pital and Methodist Hospital - engaged in a rivalry that involved buying up hos-pitals in adjacent counties. An important motivation for the buying spree wasthe desire of each chain to acquire hospitals before the other competitor did.This triggered a bitter acquisitions campaign that ultimately led to the creationof two large rival hospital systems. Deep pockets and strong ideological com-mitments can increase the level of rivalry.

Religious ideas can also shape the basis for competition in an industry. Forexample, for many years hospice care was unprofitable; consequently, religiousnonprofits were the primary source of supply for these services outside thehome. The driving principle for these entities was to provide an alternate envi-ronment to the home or hospital where patients could live out their remainingdays in dignity. The basis of competition was the provider's ability to offer thepatient and the family peace of mind rather than highly specialized medicalservices.

Does competition give rise to the commercializationof nonprofits?

The above discussion suggests that competition exists in a variety of forms inmarkets populated exclusively by nonprofits and that, in some circumstances,it can become intense. Interestingly, the presence of competition in these mar-kets offers neither a necessary nor a sufficient condition for nonprofits to com-mercialize their products. Four conditions are likely to give rise to a nonprof-it 's successful decision to commercialize:

1 The nonprofit must feel a need for additional revenues and perceivethat the sale of its outputs will provide a viable means to realize itsgoals.

2 The nonprofit's governing board must decide that the pursuit of prof-its from sale of outputs is consistent with, or at least does not substan-tially interfere with, the mission of the organization.

3 The nonprofit must have products suitable for sale in the marketplace.4 Consumers must be willing to purchase the products offered by the

nonprofit.


When these conditions cannot be met - as might be the case of a nonprofitproviding services to abused spouses or an economic development agency of-fering advice to potential newcomers - a decision to commercialize is unlike-ly to result in sustained for-profit activity. When they can be met - as in thecase of a hospital offering diagnostic services or a museum offering laser shows- some products and services will be sold for the purpose of earning a profit.

Competition in mixed-mode marketplaces

Historically, nonprofits have both collaborated and competed with for-profits(Galambos 1993, 97). For example, for-profits have funded a variety of non-profit economic development organizations designed to attract new businessesto a community or region. In Richmond, Virginia, for instance, a group of for-profit companies sponsored the Greater Richmond Partnership in an effort tobuild the reputation and job base of the area. Similar efforts can be found inBoston, Miami, and many other cities (Kanter 1995). Such collaborative effortscan be beneficial both to the parties involved and to the larger community.

In their quest for additional revenues, some nonprofits have sought market-places previously occupied by for-profits. For example, the nonprofit Educa-tional Testing Service opened a for-profit subsidiary that capitalizes on its ex-pertise in testing student learning, and the nonprofit Consumers Union createda unit that publishes books on drugs, menopause, how to evaluate a doctor, howto deal with medical problems, how to eat better, and so on. On the other hand,private firms sometimes seek out, and find, profit opportunities in marketplacespreviously dominated by nonprofits. Such new for-profit ventures have extend-ed into traditionally nonprofit areas like day-care centers, upscale health clubs,hospitals, museums, and job training. These forays into new markets by for-profit and nonprofit organizations expand the situations in which mixed-modecompetition takes place.

Although third parties normally prefer to fund nonprofits, in some situationsthey finance both for-profits and nonprofits, and this too fosters mixed-modecompetition. Third-party finance of for-profits can be found in the Medicare andMedicaid programs, in proposals to fund voucher schemes for education, andin workfare programs. Huge for-profit hospital chains such as Columbia/HCAwould probably not have arisen without federal government willingness to fi-nance their service provision, and it is unclear that they could survive in theircurrent form if this source of finance were eliminated.

Mixed-mode competition in health care moved from the single hospital tohorizontal hospital chains (multiple markets in one industry), to ambulatorycenters, doctors offices, insurance companies, nursing homes, and home health-care networks involving multiple industries. Gray (1991, 322) refers to this as


the "movement toward a health care system that is oriented toward profitabil-ity, economic rationality, and cost containment.. . ." Examples of mixed-modecompetition abound outside of health care: among colleges and universities,health clubs, theaters, radio and television stations, and in the biotech industry.In some cases, competition is limited to single local markets, as in the case ofmental-health counseling; in others, it spans geographic and industry bounds,as is the case for hospitals such as Baptist Hospital in Memphis owning multi-ple hospitals, a nursing home, diagnostic clinics, after-hours emergency centers,a health center, and parking lots.

Mixed-mode competition exists at the level of the individual product. Fre-quently, nonprofits decide to compete in for-profit markets by offering a lim-ited number of products that are complementary to their mission and capableof generating revenue. For example, art museums sell posters that reproduceworks of art that hang in their galleries; their stores can and do compete withthe private-enterprise shops that sell art to customers. Similarly, science muse-ums offer sound and light shows that compete with rock concerts and with oth-er entertainment offered by for-profit enterprises. Whether a nonprofit competessuccessfully will depend on its ability to produce a higher-quality, low-cost, ordifferentiated product. Nonprofits are similar to for-profits in their need to posi-tion their products (Porter 1979).

Reputation is one area in which nonprofits can have a substantial competi-tive advantage. For-profits build their reputations and use them to gain an ad-vantage over their competitors. Nonprofits, however, may find it easier than for-profits to utilize the trust they have earned from their constituents, especially ifthey have a reputation for high-quality service delivery. Their challenge is tofind ways to capitalize on the public's trust in them when they enter a commer-cial setting. A nonprofit museum sponsoring its own tours to Peru or Africa maybe viewed as better able to provide an exciting educational tour than a commer-cial travel agency or tour group. Similarly, a nationally known nonprofit heartcenter may be able to use its reputation for quality treatment of patients to sellfor-profit preventive health services more effectively than a national for-profitchain, and a nonprofit educational testing service may be able to sell education-al training services better than its for-profit competitors based on its reputa-tion as an expert evaluator of educational services.

As in the case of for-profit enterprise, favorable cost structures can providecompetitive advantage to nonprofits. This is particularly true in industries wheresize enables competitors to gain low average costs of service provision. Com-plementarities can exist in production (e.g., a nursing-home cafeteria offeringcatering services) or in consumption (an educational institution offering elder-hostel services). Economies of scale can also provide competitive advantage,as in the case of a nonprofit opera company that builds scenery for its own pro-


ductions and then rents its sets to other companies. An excellent example of thisis the Virginia Opera, which in 1996 contracted for scenic construction withsuch diverse entities as Lake George Opera Festival, Busch Gardens, Glimmer-glass Opera, Utah Festival Opera Company, Encore Theater, Virginia Sympho-ny, Hampton Roads Black Media Professionals, and the Jewish CommunityCenter. Volunteer labor may also reduce the costs of production, in some cas-es significantly: The Daughters of the American Revolution have many annu-al reenactments of historical events, charging admission to attend. Their eventsare staffed by volunteers who receive no compensation for their activities, thusreducing the cost of producing an event and increasing the DAR's profits.

Product differentiation also provides opportunities for nonprofits to gaincompetitive advantage. A well-known nonprofit may earn income selling T-shirts with its logo imprinted on the front, or by capitalizing on the ambianceof its facility (e.g, art museums renting rooms for private parties). Museums sellposters that celebrate specific exhibits, relying on the distinctiveness of the sub-ject matter or the prestige of the event to attract poster buyers. Museum plan-etariums use their distinct environments to attract devotees of laser light shows,whereas zoos gain advantage from settings that offer unique entertainment - forinstance, Halloween-associated events. The success of product positioning ofthis type is contingent on the ability of nonprofits to turn their environments in-to desirable settings.

Nonprofits competing in mixed-mode product markets must deliver ser-vices efficiently if they wish to enhance their revenues. The subsidies that501(c)(3) organizations receive offer some cushion against competition, but itmakes little sense for nonprofits to subsidize their own for-profit activities if thegoal of such activities is to generate revenues to subsidize the nonprofit activi-ties (see Chapter 5). Thus, over time, the forces of competition are likely to pushnonprofits toward increased use of for-profit business techniques, at least in re-lation to the products and services they sell in mixed-mode settings. This addi-tional dimension of commercialization - the adoption of for-profit productionand delivery techniques - can have far-reaching effects on the operation of anonprofit, particularly if approaches learned in a mixed-mode setting are appliedto production in exclusively nonprofit markets.

Competition need not lead to commercialization in an exclusively nonprofitmarketplace; but in a mixed-mode setting where commercialization has alreadyoccurred, competition can lead to changes in the mix of goods and services thatnonprofits produce and in how they organize production and distribution. Thesechanges may take simple forms, such as the opening of museum shops in mallsand cities, or it may involve substantial structural changes in corporate form.The next section explores several structural adjustments that nonprofits havemade in response to changing competitive conditions.


Competition and organizational form

Consider first situations where competition involves a need for capital to beused for acquisition and expansion. For example, many of the original HealthMaintenance Organizations (HMOs) were organized as nonprofit subsidiariesof hospitals or health insurance plans, but enrollments in these plans mush-roomed and competition with for-profits increased. As nonprofits, these HMOsfound their access to capital seriously limited, and this led them to seek for-profit status. Once the profit motive became part of the mission of such orga-nizations, a recasting of their nonprofit form took place, and most new HMOsorganized as for-profits (Starkweather 1993). Similarly, many of the larger non-profit Blue Cross and Blue Shield companies are now shifting to for-profit sta-tus to compete more effectively with for-profit insurers and gain better accessto capital markets (see Chapter 7).

Between the extremes of a nonprofit organization that offers an occasionalsalable product and the nonprofit with a strong appetite for capital involved inmultiple markets, lie several organizational forms that have evolved in responseto competitive conditions in mixed-mode settings. A majority of these newmodels of nonprofits are found in the health-care sector, in part because thepressures for cost containment have intensified competition in the past decade.Starkweather (1993) provides a categorization of the strategies that nonprofitsuse in mixed-mode competition. His analysis comprises two categories: hold-ing companies and joint ventures.

Holding companies

Under an unrelated passive income strategy (Strategy I), for-profit activity hasthe sole purpose of producing income for the nonprofit parent, and the parenthas nothing to do with the operations of the subsidiary. Such strategies fit situ-ations where a hospital owns a series of parking lots or a university owns a realestate company. Under this holding company-subsidiary arrangement, a parentcompany is established and used to control one or more subsidiaries, usuallythrough appointment of the subsidiaries' directors. For example, a nonprofithospital restructures, creating a nonprofit parent, a nonprofit hospital subsidiary,and for-profit ambulatory surgery centers and diagnostic laboratories. These en-tities usually have different directors. Proceeds from the subsidiaries are trans-ferred to the organizational parent. Because for-profits can sell shares, they of-fer access to capital markets not available under the nonprofit form.

Strategy II is called passive investment, cost reducing, and its goal is to cre-ate competitive advantage by finding ways to lower the cost of delivering a ser-vice. Thus, a nonprofit might acquire a for-profit subsidiary to obtain services


from it at lower cost than if purchased on the open market: A hospital mightacquire an equipment supply firm or a pharmaceutical supplier, and a nursinghome a home health agency. Once again, the nonprofit does not actively involveitself in the daily operation of the subsidiary.

Strategy III, active investment, revenue producing, is one in which the non-profit involves itself in both development and operation of its acquired subsid-iary. A hospital might use its administrative staff to run a sports-medicine clin-ic, a wellness program, or a home health service, the goal being to increase orsecure market share or to diversify service offerings. The active interest of theparent is justified by the fact that the for-profit activity either is consistent withor a part of the parent's mission (e.g, the health club produces healthier clients),or is linked strategically to key aspects of the parent's competitive environment(e.g., the home health agency produces referrals to the home).

Joint ventures

Joint ventures take the form of partnerships between nonprofits and for-profits.Although nonprofits may not partner exclusively to make a profit, profit willbe an important consideration for the newly formed entity. In health care, suchventures have been undertaken to bind doctors or other staff to the nonprofitswith which they associate, to improve capital access, to secure referral sources,to provide products or services that could otherwise not be offered, and/or todiversify. Neither the for-profit partner nor any lenders of capital to the venturehas access to the assets of the nonprofit in the event of financial failure, but as-sets owned by the joint venture itself are available to secure funding. The ma-jor advantage of joint ventures is that they permit parties to come together informal legal alliance without changing their own organizational structures or,presumably, their respective missions.

To create a joint venture, the partners establish a new, for-profit legal entity- either a corporation or a limited partnership - which sells ownership shares.One venture commonly found in the health-care industry involves partnershipbetween a hospital and its physician investors; it includes a board of directorsrepresentative of both groups and contracts for provision of health-care serviceswith the hospital, physicians, and other suppliers.

In private-corporate form, a joint venture enjoys limited liability and certainother advantages. In a limited partnership, the owners have somewhat less lia-bility protection but can take advantage of tax shelters, tax credits, and otherpass-throughs. Limited partners are not liable for losses or liabilities of the jointventure above the amount they invest in the corporation. However, a limitedpartnership must have a general partner, and that individual or entity assumesliability for losses, as well as responsibility for management of the enterprise.

42 Howard P. Tiickman

The advent of a new form, the legal liability corporation, will likely give riseto new varieties of joint venture structured in this format.

Nonprofit health-care providers have made greater use of these mixed-modelegal structures than have other nonprofits. Starkweather (1993,124-5) createsa taxonomy for classifying the joint-venture vehicles in this sector. Ancillary-service ventures are partnerships between hospitals and physicians designed toprovide services that the partners share in common - for example, ambulatorysurgery and magnetic resonating imaging. Third-party contracting ventures areestablished to facilitate contracting with health insurance companies, employ-ers, HMOs, and similar organizations; such ventures normally discount servicesin return for increased patient volume. Leased-space ventures are used by hos-pitals to facilitate construction of medical office building facilities; the spaceis proximate to the hospital and is rented to practicing physicians who bringin business. Finally, speciality ventures partner hospitals with specialized for-profit firms that provide services to complement or substitute those the hospi-tal would otherwise offer, in fields such as family counseling or alcohol anddrug abuse; these ventures lease hospital space and provide on-premise servicesbenefiting both partners.

(Revolutionary arrangements

In the preceding section we addressed situations where nonprofits either createsubsidiaries or enter joint ventures with partners who would normally competewith them. Coevolution - "the notion that by working with direct competitors,customers, and suppliers, a company can create new businesses, markets, andindustries" (Byrne 1996,47) - opens the possibility that, in certain situations,nonprofits will indeed collaborate with direct competitors. Abundant examplesof this evolving method of doing business can be found in the for-profit tele-communications industry, where companies like AT&T and the "Baby Bells"compete in some markets and collaborate in others. Collaborative approachescan also be found in the aluminum industry, where companies like Reynoldsjoint venture with aluminum producers in places like China and Russia.

At present, literature on collaborative ventures among nonprofit competitorsand on their consequences is difficult to find. However, the hospital industryprovides evidence that commercial nonprofits are engaging in a number of col-laborative partnerships. (For examples involving nonprofit universities, seeChapter 9.) Table 2.2 describes deals pending between for-profit and nonprofithospitals as of June 1996; some of these have been accepted, and some arereported as rejected (Lutz 1996). The table provides several insights into theevolving nature of nonprofit hospitals. Note that the buyouts that are occurringare often by combination, either between two or more nonprofits or betweennonprofits and for-profits. Note too that many of the deals include shared equi-

Table 2.2. Multiple-partner hospital deals pending as of June 1996

Hospital for saleor development

Deals acceptedMemorial Hosp. (TX)Lincoln General Hosp. (LA)

Doctors Hospital of Stark County (OH)Massillon Community Hosp. (OH)

St. Lukes Med. Ctr. (OH)

Univ. of Louisville Med. Ctr. (KY)

St. Joseph Med. Ctr. (CT)

New Hosp. in Tusla (OK)Physicians Regional Hosp. (TX)

Deals rejectedItasca Med. Ctr. (MN)

Cookeville General Hosp. (TN)

No. ofbuyers

23

22

2

2

2

22

3

3

Investor-ownedbuyer

Principal Hosp. Corp.None

Quorum Health GroupColumbia/HCA

Columbia/HCA

None

None

Columbia/HCALake Pointe Med. Ctr.

None

None

Tax-exemptbuyer(s)

Trinity Mother Frances Health SystemsWillis-Knighton Med. Ctr.; GlenwoodRegional Med. Ctr.; St. Francis Med. Ctr.Summa Health SystemsSisters of Charity of St. Augustine HealthSystemsSisters of Charity of St. Augustine HealthSystemsJewish Hosp. Health Care Systems;Alliant Health SystemsSt. Vincent's Health Service;Stamford Health SystemsHillcrest Health Care SystemsPresbyterian Health Care Systems

Allina Health Systems; BenedictineHealth Systems; Duluth ClinicBaptist Hosp.; St. Thomas Hosp.;Vanderbilt Univ. Med. Ctr.

Price/Stake

Price not disclosed; 100%$10 mil. for 40%

Not disclosedNot disclosed

Not disclosed

15-yr. mgmt. contract;$242 mil. pledged over 5 yrSt. Vincent's, 51%; Stamford,49%; price not disclosedDeal in discussionTerms not available

$9 mil. for 100%

$61.3 mil.

Source: Adapted from Modern Health Care, June 21, 1996, pp. 146-8.


ty, joint operating agreements, and arrangements involving shared governance.The implications of these changes, as well as those above, are explored in thenext section.

Competition, commercialization, and the challenge topublic policy

It is not easy to provide satisfactory quantification of the impact of nonprofitson society. The proliferation of mostly small nonprofits, the inadequacy of IRSForm 990 (see Appendix) as an instrument for evaluation, the decentralizednature of the sector, and the broad nature of most nonprofit mission statementscomplicate the task of pinpointing impact. As a consequence, society has nothad the benefits of a data base of studies to demonstrate that the outputs of non-profits are meritoriously different from for-profit enterprise. The rapid growthof the nonprofit sector, continuing debate over the role of government, declin-ing government payments to nonprofits, and heightened emphasis on justifyingtax exemptions have increased the need for the sector to pay greater attentionto questions of measurement.

This need to identify impact is particularly strong with regard to commercial-ization. The increased commercialization of the nonprofit sector is likely to pro-ceed unevenly among industries, if only because there are many areas wherenonprofits are unable to produce goods and services salable in the marketplace.Although this could give rise to a desire for nonprofits in these industries toexpand their missions to enable commercial activity, such a prospect does notseem likely because a broadening of mission may not be feasible for many typesof charitable organization. What, then, might happen to these organizations ifthey are unable to commercialize? This is a serious question for public-policymakers interested in the future of the nonprofit sector to ponder. One possibil-ity is that some of these will be forced out of business or made to curtail theirservice delivery. Alternatively, if their revenue needs continue to grow, somecharitable nonprofits may look more intensely at unrelated business income asa source of additional revenue. The extent to which such income can providesignificant contributions to revenue remains to be seen.

Several challenges are posed by the evolving nature of the nonprofit organi-zational forms. The materials in this chapter suggest that Weisbrod (1996; andsee Chapter 1) is correct in his prediction that increased fiscal pressures will leadto new forms of commercial activities and new organizational forms that blurthe distinctions between for-profit and nonprofit organizations. It is difficult toevaluate nonprofits with general mission statements such as "to increase thewelfare of children," and the magnitude of the task increases substantially whenthree or more nonprofits with varying mission statements join in partnership.If a joint venture does include other nonprofits, issues arise as to how to evalu-


ate the effects of the venture on the nonprofit mission of the originally nonprofitentity. When for-profit partners also are involved, additional complications ariseas to how to determine whether, and to what extent, a dilution of nonprofit mis-sion occurs.

Evolution of organizational form and product content is both inevitable andhealthy in a society built on competition and responsive to its challenges. Theproblem is to see to it that laws and economic institutions recognize the effectsof these evolving changes, and that they not only adapt to them but also guaran-tee that nonprofits do not lose sight of the missions that brought them into exis-tence. Brody (1996,536) may be correct in her assertion that the formal legaland economic differences among nonprofits and for-profits "are more of degreethan of kind," but she is equally correct in noting that historical, political, psy-chological, religious, sociological, and legal factors create "real" differencesbetween the sectors. The challenge for public policy is to ensure that the evolv-ing institutional changes, as well as the pressures toward commercialization, donot diminish the unique charitable role of the sector. This will not be easy in anera where the quest for fresh revenues to secure institutional survival and fos-ter growth is seemingly paramount.

CHAPTER 3

Modeling the nonprofit organization as amultiproduct firm: A framework for choice

Burton A. Weisbrod

Introduction

To understand why nonprofits are exhibiting a stepped-up involvement in com-mercial activities - expanding into new areas of activity, entering increasinglyinto business arrangements with private industry, pricing and advertising moreaggressively, and, in general, doing things that are novel for nonprofits but com-monplace for private firms - we need to understand their goals, or missions, andthe financial constraints on attaining them. This chapter presents the theoreticframework with which we analyze nonprofit organization commercial activi-ties in the chapters that follow. Our focus is on private nonprofit organizations,although most of the analysis applies to government organizations as well.

The decision by a nonprofit organization to produce a particular output, makeit available to particular consumers, and finance it through specific financemechanisms, reflects choices: whether to use revenue-raising techniques typ-ically associated with private enterprise, such as entering new markets in orderto generate profits, charging fees that raise revenue but restrict access in the pro-cess, and generally acting like profit-maximizing private firms; or to act differ-ently from private firms, avoiding particular markets or revenue-raising oppor-tunities that might conflict with their goals. Nonprofit organizations and privatefirms can be expected to make different decisions about engagement in "com-mercial" activities - profit-oriented actions in forms and degrees associated withprivate firms - if they have different goals or different opportunities and incen-tives.

Incentives are important here. Nonprofit charitable organizations can receivetax-deductible donations, whereas private firms cannot; as a result, each has avastly different degree of dependence on donations. Similarly, the fact that non-

I thank Jeffrey Ballou, Louis Cain, Joseph Cordes, Ian Domowitz, John Goddeeris, Laura McLean,Lewis Segal, Richard Steinberg, and Dennis Young for helpful comments on an earlier draft.

47


profit organizations pay no corporate profits tax on their mission-related activ-ities - those deemed "substantially related" to their tax-exempt purpose - whilepaying full tax on profits from "unrelated" business activity, affects their incen-tives to engage in these differing kinds of activity. In contrast, for-profit firmsgenerally pay profits tax on all activities, and so have little or no tax-inducedincentive to choose one type of activity over another. (Tax laws do influencechoices, however, through such measures as inducements to locate in enterprisezones and tax credits provided for research and development.)

Goals also matter. Though it is an oversimplification, private firms can gen-erally be characterized as seeking to maximize profit in the interest of stock-holders. How to characterize the goals of nonprofits is less clear. Supporters ofnonprofits represent them as public-serving institutions that engage in activitiesunderprovided by private enterprise and government and that, more broadly,offer another avenue for individual expression of social concerns. Critics seemore crass motives behind the goals of nonprofits: taking advantage of publicsubsidies and public trust to support inefficiency. They believe nonprofits to benothing more than "for-profits in disguise (FPIDs)" (Weisbrod 1988). Doubt-less, the vast array of nonprofits includes organizations at both ends of the spec-trum. The analytic framework presented here is a structure for examining thebehavior of nonprofit organizations and identifying in what ways they are sim-ilar to, and different from, profit-maximizing private firms. It guides much ofthe descriptions and analyses in the chapters that follow.

Framework for understanding nonprofit organizationbehavior

The behavioral model of nonprofits that guides our analyses and descriptionsof nonprofit organization behavior is built on the foundation developed by Es-telle James (1983) and extended by Jerald Schiff and Weisbrod (1991). In thismodel, nonprofits are viewed as multiproduct organizations potentially produc-ing three types of good that contribute in various direct and indirect ways tothe organization's mission. The first type of good comprises the organization'soutput mission: This involves some type of collective good, such as basic sci-ence research, medical care for the poor, or preservation of endangered animalspecies or cultural heritage. It may also involve public "trust" that an organi-zation will not act opportunistically against ill-informed consumers. The othertwo types of good involve private goods, which are not in themselves the mis-sion but are potential sources of revenue for financing the primary, mission-related output. One of these is incidentally related to providing the mission-related good and has the potential to generate user fees; the other is unrelated,or "ancillary," to the mission.

Modeling the nonprofit organization as a multiproduct firm 49

In short, a nonprofit chooses to provide varying amounts of each of threegoods:

1 a preferred collective good, which is difficult to sell in private markets(e.g., basic research);

2 a preferred private good, which can be sold in private markets butwhich the nonprofit may wish to make available to some consumersindependent of their ability to pay (e.g., access to higher education);and

3 a nonpreferred private good, which is produced solely for the purposeof generating revenue for the preferred good (e.g., paid advertising onpublic television).1

For each of the three goods we assume that the nonprofit can determine theamount it will produce (including none at all) and, subject to the constraints ofcompetition, its price. Each type of good is considered below.

The classification of revenue sources into three categories - donations (con-tributions, gifts, and grants), user fees, and ancillary activities - proves to beuseful analytically; even though every source of revenue does not fall neatly in-to those categories, most do. A few examples from 1995 applications to the IRSfor tax-exempt, nonprofit, status under section 501(c)(3) of the Internal Reve-nue Code illustrate the applicability of this classification. One organization, Sui-cide Anonymous Treatment, indicated that its "sources of financial support"would consist entirely of "government grants, foundation grants and individ-ual contributions" (IRS form 1023, employer identification number [EIN] 95-4507672). The application by Contra Costa County Earth Day reported its chiefsources of financial support to be more varied, encompassing two of the threerevenue sources: "grants from corporations, foundations, and government agen-cies; revenues from concessions, booth rental, and the sale of other items; park-ing receipts" (IRS form 1023, EIN 68-0347329) - concessions, booth rental,and parking all being ancillary activities.2

A third new nonprofit, the Archaeological Society of Central Oregon (EIN93-1133518) reported obtaining financial support from all three sources weidentified: donations, raffles (ancillary commercial activity), and membership(user) fees.

1 Mancur Olson (1965) recognized more than thirty years ago that providers of collective goods,such as the American Medical Association in his example, confront a finance problem resultingfrom free-rider behavior. He also noted that such organizations may rely on by-product (whichwe term ancillary) services, such as professional journals and insurance, to generate profit forfinancing the collective activities.

2 That is, they are not central to the social mission. They may, however, not be taxed as unrelatedbusiness activities, for they could be deemed by the IRS to be substantially related to the missionor, under other provisions of the tax law, excluded from taxation.


Other essential elements of the model are as follows:

the degrees of organization preferences for, or aversions to, engaging ineach of the three kinds of activities;

the effects that revenue-generating activities may have on the willingnessof external funders to give donations (contributions, gifts, and grants);and

the effects that revenue-generating activities may have on the nonprofit'scosts of producing the mission-related output.

This framework is not to be seen as a rigid structure that has been thorough-ly tested. Rather, it serves as a guide as we examine nonprofits' commercialactivity and discuss the reasons it is increasing, the forms it is taking and why,and its potential effects. The framework identifies variables and relationshipsto which we direct attention.

Nonprofits' goals

Understanding nonprofit organization behavior in general, and particularly itsincreasing reliance on commercial methods - the sale of goods and services -is critically dependent on an understanding of organization mission. Many ofthe problems confronting both research on nonprofits and public policy towardthem hinge on some rather unusual characteristics of nonprofits' goals or mis-sions. One is that nonprofits - that is, their leadership - may have preferencesabout the kinds of activities in which they engage and the persons to whom theoutputs are made available. This stands in contrast to the textbook model of theprivate enterprise firm, which has no such preferences, but cares only about op-portunities for profit.

The assumption that nonprofits as a class can be characterized usefully by asingle set of goals and trade-offs is open to question. Managers (and directors)and their preferences can influence nonprofit organization behavior greatly be-cause there is only a weak functional equivalent of private firms' stockhold-ers - the IRS - and there is no threat of corporate takeover as there is in theprivate-enterprise sector. There are, of course, constraints on nonprofits. Someof these are imposed by law - on nonprofits' obligations and on the uses ofany "profits" - and some are imposed by sources of funds, in addition to techno-logical and input-price constraints. Given, however, the broad scope of manynonprofits' tax-exempt authority and the difficulty of defining and measuringoutputs, there is room for considerable managerial discretion. Managerial pref-erences thus may interact with organization constraints to produce a manage-rial sorting process that determines nonprofit organization objectives and be-havior. Nonprofits may behave as if they were individuals with unique utility


functions. They may also behave as if they were run by a committee of manag-ers and directors, each of whom has distinct goals.

Nonprofits' goals may be multiple and in conflict, and there is no simplemeasure of the efficiency of the trade-offs being made among goals. This toois unlike in private enterprise, where maximum profit reasonably characterizesan organization's goal. Thus, for a university, the goal of teaching all undergrad-uates in small classes led by senior professors conflicts (even apart from anycost considerations) with the goal of graduate training, which requires that grad-uate students obtain teaching experience. It also conflicts with universities' goalof expanding knowledge, which requires faculty to focus on cutting-edge schol-arship of the sort usually associated with graduate rather than undergraduatetraining. Further conflicts become evident when universities' intermediate goalsare considered: For example, a university may prefer to have faculty devotemore time to being interviewed by newspaper, magazine, and television report-ers who would bring favorable attention to the university, but this takes awayfaculty time that could be used to expand research.

Nonprofits' goals are not only numerous, they are typically vague. Thus, itis difficult for society, the regulatory authority - the IRS in the United States -or even the nonprofit itself, to determine the degree to which goals are beingrealized and, hence, the degree to which their achievement is being facilitatedor retarded by some particular activity. What does it actually mean for a univer-sity to provide educational opportunity for all students who can benefit from it,for a hospital to provide care for all who need it, or for a museum to preservea cultural heritage? How can an objective viewer determine any nonprofit's so-cial contribution? With such broad goals, how can a nonprofit be held sociallyaccountable - rewarded for being effective and punished for performing poor-ly (Herzlinger 1996)?

Society's goals are generally intended not for individual nonprofit units butfor the industry as a whole. The social goals for institutions of higher educa-tion - for example, to provide wide student access and choice of subject mat-ter - do not require that every college campus be open to students of all abil-ity levels and interests. Thus, although access to one particular facility may becritical to an individual student or family, society's planning problem and itsassessment of overall college-sector performance involve a broader, systemicperspective.

The goals of nonprofits are not easily characterized in terms analogous tothe essential goal of private firms to maximize profits; but some evidence canbe gleaned from observing their revenue-generating activities. Our focus onnonprofits' increased commercialism can shed light on those goals, althoughit should be noted that this commercialism has potentially important conse-quences regardless of what the rationale may be for pursuing them.


When a nonprofit has a goal that encompasses more than reaping profit, thataffects its provision of each of the three types of good identified earlier - pre-ferred collective, preferred private, and nonpreferred private. We examine non-profit activities in these three types of market within a framework in which thereare two kinds of organizational goal: to provide certain kinds of services and toavoid engaging in certain other kinds of activities. Together, these constitute anumber of hypothesized deviations from profit-maximizing behavior, with non-profits providing more of some goods ("preferred" or "mission-related") andless of others ("nonpreferred") than would a profit-maximizing private firm. Thenature of those two types of good (or service) is explored, with the hypothesisthat preferred goods correspond to those that are socially desirable but private-ly unprofitable and, hence, are underprovided in private for-profit markets. Non-preferred goods are hypothesized to be those that can be and indeed are sold,may be provided by private firms, and are potential sources of revenue for non-profits seeking funds to finance their mission-related activities.

We do not assert that nonprofits necessarily behave in the manner described.Rather, the following chapters explore evidence that nonprofits' growing com-mercialism reflects predictable responses to the tension between goals and rev-enue constraints. Relatedly, the chapters also shed light on the usefulness of thedistinction between nonprofits' preferred and nonpreferred activities, and therevenue and cost interrelations between them.

Preferred activities

Consider the case of a nonprofit with the central goal of providing socially valu-able but privately unprofitable services. Such a goal - which has been referredto elsewhere as bonoficing (Weisbrod 1988) - can involve provision of twotypes of good. One type involves "collective," or "public," goods. These in-clude, for example, basic research, as by universities and medical research or-ganizations (see Chapter 9); the preservation of endangered species, as by zoosand aquariums (Chapter 11), and of cultural heritages by museums (Chapter12); the provision of television programming that has social value but cannotbe sustained through the sale of commercial air time (Chapter 13); environmen-tal protection; and trustworthiness about hard-to-monitor dimensions of servicequality, such as whether any charitable organization claiming to aid the poor orto provide "tender loving care" to the elderly frail in a nursing home is actual-ly doing so. Another type of preferred output may involve certain private goodsor services targeted to the poor or other "socially deserving" groups. These in-clude, for example, hospital care for the indigent (as explored in Chapter 8) andaccess to social-service agency activities such as scouting (see Chapter 10).Both types of output - some collective, some private - may be encompassed inthe organization's mission.


Nonpreferred activities

A second goal or objective of a bonoficing nonprofit might be, but is not nec-essarily, to avoid engaging in certain types of activity. It may wish not to com-pete with private firms in the sale of conventional private goods or, in general,not to produce goods or services other than those that contribute directly to itsmission. In this model nonprofits derive positive utility from greater achieve-ment of their mission, but negative utility from activities that represent distrac-tions from that mission - which could include activities that generate funds thatcan be used to finance increased mission-related activities.3 Thus, some poten-tial revenue sources, such as unrestricted donations and, particularly, unsolicit-ed donations are preferred to others - not simply because they are obtained eas-ily and without restrictions, but because they require neither efforts that distractan organization's leaders from pursuit of the mission nor any compromise ofthat mission. By contrast, the production of commercially salable ancillary ser-vices may boost revenue but be disliked by managers and trustees. Thus, uni-versities dislike selling research services to private firms when that interfereswith the broad dissemination of knowledge; public television stations dislikebroadcasting "commercials" that advertise private firms; and academic healthcenters dislike shifting their focus from basic research on unsolved medicalproblems to treatment of children with common colds (as necessitated by thedevelopment of university-based HMOs).

Involvement in ancillary goods and revenue-generating activities can takemany forms. While we direct particular attention to nonprofits' decisions to pro-duce salable goods and services, the organizations may also choose to make"passive" investments, such as buying securities in profit-making firms. The hy-pothesized aversion of nonprofits to commercial activity may apply also to suchinvestments, which the nonprofit owns but does not operate; some nonprofitsmay want to avoid owning stock in tobacco or gambling firms, for example.Little is known, however, about the extent of such preferences. In any event, ifthey are present, they are unrelated to tax consequences: Income from passiveinvestments is not taxed, even though income from active investments that areunrelated to mission is subject to the Unrelated Business Income Tax (UBIT).

There is another plausible way to look at this. Instead of seeing some activity as entering theorganization's objective function negatively, one could think as follows: Nonpreferred goodsmight not enter the nonprofit's objective function at all; they do, however, generate revenue, andthus enter the constraint set. Nonprofits would therefore limit production of these goods not be-cause they dislike them but because such production diverts resources from activities that enterthe objective function positively. In this model nonprofits would maximize profit in ancillary-goods markets, whereas in the model presented above, nonprofits would prefer to produce lessthan would maximize profit. (I owe this point to Jeffrey Ballou.) Evidence presented later sug-gests that nonprofits do indeed underprovide relative to a profit maximizer.


Our model also suggests that nonprofits prefer to minimize the use of userfees - charges for private goods that are related to the organization mission, suchas college tuition, hospital patient fees, and museum admission charges. Apartfrom revenue and cost interdependencies that could make it unprofitable to em-ploy user fees, it is not obvious why nonprofits would want to avoid them. Whywould universities prefer to set tuition at zero? Why would museums and zoosprefer to have no admission fees? Why would hospitals and day-care centersrather not charge their patients and clients? One answer involves the output-distribution goal that many nonprofits have. Their mission is often not merelyto provide particular services but to make them available either to everyone(e.g., medical research and public television) or to particular consumer groups(e.g., food, shelter, and medical care to the indigent; scouting services to youth).The problem with user fees is that they inevitably have the side effect of dis-couraging consumers whom it is the nonprofit's mission to serve. Thus, thereis potential conflict between increasing user fees to raise revenue and maximiz-ing the nonprofit's contribution to its mission. Profit-maximizing private firmsin these fields, by contrast, have no such distributional mission, and so they un-derstandably place fundamental reliance on user fees as their source of revenue.

The disadvantage of user fees to nonprofits with distributional goals woulddisappear if the organizations were able to determine each consumer's willing-ness to pay. A nonprofit could, then, charge prices that would raise revenue with-out deterring consumption by any person in the mission-related target group(see Chapter 4, and Steinberg and Weisbrod 1997). With such full information,prices could be set differently for various consumers, in ways that maximizednot profit but achievement of mission. Some consumers might be charged lessthan marginal cost, or even zero, and less than their marginal willingness to pay;others, less central to the target group, could be charged higher prices, perhapseven more than marginal cost if competition among nonprofits were limited.

Nonprofits' dislike of ancillary activities, which though not mission relateddo provide revenue for cross-subsidizing the mission-related output, has a quitedifferent basis from the aversion to user fees. Provision of an ancillary outputmay have one or more direct negative effects - apart from any favorable indi-rect effect operating through the revenue it generates - on organization mission:

1 It may distract management, causing a weakening of its attention tothe organization's central mission.

2 It may cause mission displacement - sacrifice of some element of theorganization's goal in order to satisfy prospective purchasers of the an-cillary services who would impose restrictions that compromise thenonprofit's mission.

A good example, cited above, is when a university has the opportunity toraise revenue by performing research for a private firm that insists on propri-


etary rights to the results, whereas this is inconsistent with the university's mis-sion of full dissemination of research findings.

Thus, we model nonprofits as preferring to maximize production of mission-related goods and to minimize production of either of the two kinds of goodthat can generate revenue. Within this framework, the two generators of non-donative revenue - ancillary goods and user fees - are seen as being pursuedfor their financial contributions only. Moreover, if they bring disutility, they willbe undertaken at levels below what would maximize their monetary contribu-tions. In short, if this model characterizes nonprofits aptly, such organizationswill not act as profit maximizers in these revenue-raising commercial markets.

Revenue from donations

Donations, encouraged by income-tax deductibility, and volunteer labor, whichis an in-kind donation, may provide some support, but further resources are ac-cessible to a nonprofit only if it engages in revenue-generating activities that itmay prefer to avoid.4 Donations are nonprofits' major alternative revenue sourceto the "commercial" sources of user fees and profit from ancillary activities.Since the following chapters make extensive use of data on donations, it is im-portant to clarify the relationship between the theoretic concept of a donationand the operational definition as applied by the IRS and nonprofits in theirreporting of revenue on the Form 990 tax return. In particular, we distinguishbetween donative revenue that is exogenous, coming to the organization essen-tially regardless of its activities, and donations that are endogenous, or influ-enced by the nonprofit's activities. Exogenous donations, we suggest, are pre-ferred as a revenue source to either form of commercial activity. Obtaining themdoes not require the nonprofit to divert activities from its mission.5 This pref-erence stands apart from the fact that there is no resource cost required for fund-raising; that is, because such donations permit the nonprofit to focus entirely onits mission, a dollar of exogenous donations may be preferred to a dollar of netrevenue from any other source.

Of course, many nonprofits devote resources specifically to fund-raising.This is, in effect, an ancillary activity, because its main purpose is to raise rev-enue. The finding that nonprofits engage in less fund-raising than would maxi-mize their net revenue (Weisbrod and Dominguez 1986) leads to the possibil-

4 Henceforth we do not address separately the market for volunteer labor or other donations in kind.The presumption is that they respond, and are responded to, in the same manners as donationof money. However, that is not entirely clear; the market for volunteer labor deserves further at-tention - both its supply and demand sides (Menchik and Weisbrod 1987).

5 Whether any donations are entirely exogenous may be questioned. Still, there may be importantdifferences in the degree to which donations from various sources respond to a nonprofit's activ-ities.


ity that fund-raising may be a nonpreferred ancillary activity. This possibilityhas led to questions of whether fund-raising should be regarded as an "unrelat-ed business activity," since it is not, itself, a part of organization mission (Suhrke1996).

The hypothesized preference for donations over commercial revenue resultsfrom the assumption that donations are exogenous - obtained without the orga-nization having to engage in fund-raising efforts (possibly a nonpreferred an-cillary activity). Not surprisingly, donations have been found to be responsiveto fund-raising efforts (Weisbrod and Dominguez 1986), but they may also re-spond to other organization activities. Surely, charitable contributions respondto a nonprofit's mission-related outputs, to its reputation for efficiency and in-tegrity, and hence to its trustworthiness to use donated funds effectively, al-though there has been little research on these relationships.

Commercial activity responds to donative revenue

If it is correct to characterize nonproflts as we have, the extent of their involve-ment in commercial markets, through either ancillary goods or user fees, canbe expected to vary inversely with exogenous donations. That is, other thingsequal, donations crowd out commercial activity. An exogenous decrease in do-nations - for example, in government grants to academic health centers or thearts - may be expected to cause those nonproflts to expand commercial activ-ity, approaching, but not reaching, profit-maximizing levels because of the dis-utility from increased commercialism. The disutility would result from a neg-ative direct effect of the activity on mission.

The potential negative effect that commercial activity - involving quid proquo sales - can have on a nonprofit's mission can be illustrated by recent in-volvements of a nonprofit university and a public school system in selling ser-vices to private firms. Universities are increasingly contracting to sell researchservices to private industry (see Chapter 9), and the contracts sometimes requirethe university to keep the research results "under wraps." Such an arrangement,which conflicts with a university's mission of full dissemination of knowledge,is typical of contracts between universities and the Walt Disney Company forresearch on such things as special polymer materials for its audio-animatroniccharacters, such as U.S. presidents (Wu 1996). Similarly, many public schools,hard pressed financially by declines in government grant support, have turnedto ancillary, noneducational activities that have brought charges that studentsare being exploited in return for profit-oriented commercial advertising. Chan-nel One, a twelve-minute current events television program that includes twominutes of paid advertising, has been allowed into 12,000 schools in the Unit-ed States, where it is viewed by eight million students. In return for accepting


the commercials the schools have received television sets, VCRs, and a satel-lite link (Honan 1997). In another case, the Seattle School Board voted in 1996to permit corporate advertising in the schools, leading the district's chief finan-cial officer to speculate that the next step might be "the cheerleaders, broughtto you by Reebok." Presently, students in Colorado Springs ride buses carryingcorporate logos and walk hallways decorated with posters advertising Moun-tain Dew soft drink (Stead 1997).

Such compromising of mission in the interest of revenue may be termedmission displacement. When it occurs, it is likely to take subtle forms that arehard to observe. Rarely will an organization reject its mission outright, for evenif the nonprofit's leadership were willing to do so, such a rejection would havevast consequences for them in terms of their fiduciary responsibility, as well asfor the organization's tax-exempt status and its donative revenues. Nonetheless,the potential for mission to be compromised, albeit in less direct forms, exists,particularly so in light of the breadth of many nonprofits' missions, which canmake it difficult to define operationally when an action is inconsistent with themission.

A potentially important issue (but beyond the scope of this study) is whetherinvolvement in commercial markets causes nonprofit organizations to alter thekinds of leadership they employ. The question is whether increased commer-cialism leads to "managerial displacement," the substitution of managers anddirectors who are especially knowledgeable about markets where profit is themeasure of success, for leaders who are efficient in the preferred-goods mar-ket. Such managerial sorting might take the form of nonprofits hiring more ex-ecutives or directors with backgrounds in the private business sector. While thiswould reflect the intended shift in the organization's behavior, it also could setin motion a dynamic process in which the fundamental character of the non-profit is changed. Anecdotal evidence that such convergence of management isoccurring can be seen in the "revolving door" that is increasingly characteris-tic of managerial movement between universities and the private medical tech-nology sector. The head of Bristol-Myers Squibb's research enterprise recentlyreturned to Harvard University, and faculty from Baylor College of Medicine,Duke University, and Cambridge University have moved to private-sector man-agerial positions with corporate giants SmithKline Beecham PLC, Merck &Company, and Perkin-Elmer Corporation (Winslow 1996; also see Chapter 9).

Our behavioral model, in which nonprofits prefer to avoid both types of com-mercial activity - sale of ancillary goods and user fees - but want the revenuethat they can bring, has a number of testable implications. As suggested above,one is that when nonprofits do engage in the kinds of commercial activities justillustrated, they do less of it than would maximize profit from those activities,because the negative utility from engaging in the activities will be balanced


against the revenue generated. As a result of not maximizing net revenue, theywill also not be able to maximize output of their preferred, mission-related ac-tivities, because of the desire to avoid the nonpreferred activities.

Another implication of nonprofits' hypothesized dislike of commercial ac-tivity is that its use will respond to certain exogenous changes. These includenot only exogenous changes in donative revenue, the generally preferred source,but also to exogenous changes in what is equivalent: a change in the perceivedseverity of "need" for mission-related output. Thus an increase in poverty orin the number of uninsured hospital patients can be expected to increase non-profits' commercialism. If net revenues from user fees or ancillary activities arenot being maximized by service providers in those industries, there is latitudefor increasing revenues and, ultimately, mission-related output.

These two implications are illustrated by the case of the British Museum andits current financing problems. The British government has traditionally giv-en grants (donations) to the museum, which have constituted most of the mu-seum's $84.5 million annual budget. Late in 1996 the government announcedthat it would begin to reduce its contribution. Under impending fiscal stressfrom the cut in grants, the museum hired a consultant, who recommended im-position of an entrance charge, contrary to its historic policy. Such a revenue-generating fee could have been introduced earlier, but it had not, ostensiblybecause it was believed to be inconsistent with the museum's mission of openaccess. The museum had not been maximizing its revenue from admission fees,but the museum's director opposed the consultant's recommended admissioncharge of $8.50, claiming it "would fly in the face of the museum's missionwhile discouraging perhaps half its annual visitors from coming." Despite therevenue potential, the museum decided against the fee; yet the trustees apparent-ly recognized that the revenue shortfall could become serious enough to over-come their aversion to an admission fee, for they stated that the introduction ofcharges cannot be ruled out (Lyall 1997). Apparently a "sufficient" cut in gov-ernment grants could bring a user-fee response.

Donative revenue responds to commercial activity

Just as donations may affect commercial activity, the level and nature of com-mercial activity can affect donations. Donors may prefer that a nonprofit avoiduser fees, maintaining free access to its services; alternatively, they may pre-fer that the organization work to meet its own financial needs by charging feeswhere that can be done without compromising its mission. Regarding ancilla-ry goods, donors may prefer either that a nonprofit develop revenue-generatingancillary-good markets, or oppose it as an inappropriate diversion from the mis-sion. In any of these cases, donations could be affected either positively or neg-atively by the form and level of commercialism.


More complex model: Interdependencies among revenuesources and among production costs

Interdependence among revenue sources

If we consider all sources of a nonprofit's revenue to be interdependent, thenthe nonprofit might be maximizing its total revenue from all sources and, hence,maximizing the provision of its preferred, mission-related activity, even if itwere failing to maximize its revenue from any one source. As noted above, ifa nonprofit attempted to augment revenue from exogenous donations by in-creasing its commercial activity, that could have a negative impact on dona-tions; in other words, donations may well be at least partially endogenous tononprofits' commercial success. Donors might decide that their contributionsor grants are not needed, or they might want to display opposition to certaincommercial activities by withholding donations. There is some reason to be-lieve, for example, that rapidly rising tuition (user fees) at public universities isnot just the result of a shortfall of government grants, but also a cause of it, aslegislators have come to see that the university will sustain itself financiallywith less funding from government grants.

Donations and commercial revenues, whether from user fees or ancillary ac-tivities, may also be related positively; an increase in commercial revenue couldcause donations to increase. This would occur if donors - perhaps foundationsor individuals - wished to reward "self-help." In this case, the predictions ofour simple model, outlined previously, might no longer hold. If increased com-mercial financial success generated additional donations as well, these rewardscould outweigh any direct disutility from commercial activity, and the nonprofitmight act to maximize profit in its ancillary-goods or user-fee markets. Thesame result would hold if revenue sources were independent and there was nodisutility from commercialism. The nonprofit would then be expected to maxi-mize profit in commercial markets, so as to maximize, in turn, achievement ofits mission.

An illustration of revenue interdependence will suggest the breadth of formsthat they can take. In this case, which involves nonprofit academic medical cen-ters, a change in revenue from user fees, resulting from an exogenous changein competitive behavior, has affected efforts to increase revenue from donations;thus causation here runs in the opposite direction from the British Museum case,in which a change in donations was affecting policy toward user fees.

Academic medical centers have long cross-subsidized their basic science re-search - arguably their principal mission-related output - with revenue gener-ated by the user fees that their faculty member physicians charged patients. Thisrevenue source has provided nearly $1 billion a year to support research. Dra-matic change within the health-care sector, however, has increased the power


of managed-care organizations to put downward demand pressure on medical-care prices, which has reduced the user-fee revenue available for cross-subsidiz-ing university research (Winslow 1997b). "Large health-care purchasers agreethat research and doctor training are important medical-school functions," but(apparently recognizing the collective-good elements of these functions) theyseem to be unwilling to pay for them. This free-rider behavior, having broughtan exogenous reduction in revenue for the research and training missions, hasled to a campaign, launched by a group of scientists and with the support ofSenator Daniel Patrick Moynihan, to increase government grants for research,financed by a new tax.6

This illustrates more than the offsetting, or "crowding out," effects of onerevenue source on others. It also shows that "grants," particularly from govern-ment, can have the characteristics of ancillary activities, having to be "sold" (tolegislators, and sometimes to private donors) in ways that may differ only im-perceptibly from sales in private markets, and that are equally distracting anddisliked by nonprofits. Thus, it is noteworthy that the tax and the support forresearch, which could have been put forward in the past but was not, has nowbeen proposed in apparent response to the severe financial problems of the na-tion's academic medical centers. Were it to pass, it would suggest that govern-ment contributions from one source are not independent of government contri-butions from another. In any event, the proposal itself seems consistent with amodel in which nonprofits pursue a particular source of revenue more or lessvigorously depending on the availability of other sources. This would suggestthat the nonprofits were not previously maximizing revenue from the unpursuedsource, unless it were the case that the initial decrease in grants had the effectof increasing the probability that the new tax would be enacted.

Interdependencies among costs

Costs, as well as revenues, may be interdependent. Until this point, we have as-sumed implicitly that the costs of producing the preferred mission-related good(henceforth referred to simply as the mission-related good) and each of the non-preferred commercial goods are independent of one another. Thus, a change inthe rate of production of an ancillary good, for example, was assumed to haveno effect on the cost of producing the mission-related good. In fact, they arelikely to be related. They would, for instance, be related if, as is quite proba-ble, particular ancillary goods are selected because of their expected profitabil-ity, and that is most likely when the same resource inputs can be used for boththe mission-related and commercial good. (Chapter 5 presents some evidencein support of this expectation, showing that aggregate labor costs appear to in-

6 The proposal was to levy a 2 percent surcharge on health-insurance premiums.


crease little when ancillary goods production increases.) It is not surprising,then, that nonprofit hospitals are raising revenue by selling laboratory services,zoos are selling animal-breeding services, and public television stations arerenting access to their studio space as well as to broadcast time (i.e., "underwrit-ing"). Again, as with revenue interdependencies, the existence of cost interde-pendences makes it more difficult to specify the response expected from a non-profit confronted with an exogenous change either in one revenue source or inperceived social need. An increase in commercial activity could bring multipleeffects, not only on revenue from that activity but on other revenue sources andon marginal production costs for the mission-related and commercial goods,which could in turn have further effects.

Summary: Private choices and public policy

For any organization, revenue sources influence output decisions, and outputdecisions in turn influence revenue availability. In the context of nonprofit orga-nizations, it is useful to think of their making two choices that can be expectedto define both their output mix and its financing. The first choice involves thekinds of output to be produced and the consumers who will be targeted; the sec-ond is how it will finance itself. The framework adopted here shows the com-plex ways that these two choices are related.

The goods and/or services that are a nonprofits' output can be viewed asbeing of three types: Two are preferred, in the sense that they constitute the or-ganization's social mission and provide the justification for government subsi-dies and tax exemptions. A preferred, or mission-related, good may be eithercollective, such as basic research, preservation of the environment, or propaga-tion of endangered animal species; or private, aimed at a target population forwhich there is a consensus that the social value of particular services exceedsrecipients' ability to pay - for example, hospital care for the medically indigent,access of youth to community centers, and attendance of schoolchildren at zoosand museums. Although the latter type, preferred private goods, provide privatebenefits to the targeted individuals, they are still, in a sense, also collective inthat many people other than the recipients value the provision of the servicesto the target group. One essential characteristic of such private-type services isthe potential for charging user fees. Nonprofit organizations may also chooseto produce a third type of good, which we term ancillary. Ancillary goods areproduced only because of the organization's desire to raise revenue to financeits preferred goods.

There is a significant interrelationship between each type of good and thetax laws that determine whether its related activities are or are not taxable, de-pending on whether they are judged by the IRS as being substantially "related"to the nonprofit's tax-exempt purpose. This interrelationship is discussed more


thoroughly in Chapter 5. For now, it should be noted that the theoretic distinc-tions among the three types of good may not be the same as the regulatory-taxdistinction among them. For example, a particular treatment program in a hos-pital, such as an aggressively marketed heart transplantation unit, could be re-garded by the IRS as related to the hospital's exempt purpose even though theactivity is nonpreferred by the organization and would not have been provid-ed to the same extent, if at all, were it not for the financial stress from declin-ing revenues. Similarly, a nonprofit's goals may be such that certain activitiesare deemed important to fulfilling its mission despite an IRS ruling that theactivity is taxable. Thus, the nonprofit American Economic Association contin-ued its book publishers' display activity at its annual meetings, based on a be-lief that it was professionally valuable, despite an IRS ruling that it was "un-related" to the association mission and hence taxable.

Whether a specific good is preferred or nonpreferred by a nonprofit can de-pend on its particular characteristics. Although elaborate fund-raising galas areuntaxed, they are likely to be nonpreferred, whereas straightforward provisionof information to prospective donors about the organization's needs is likely tobe a preferred - or, at least, neutral - good, although both generate donations.Similarly, obtaining a donation that is given with few or no restrictions on itsuse is preferred, whereas procuring one that has constraining "strings" attachedis more usefully thought of as the sale of a nonpreferred ancillary commodity- the donor offering a sum of money in return for some consideration - eventhough conventional terminology, and the IRS, would count it as a donation.

Some nonpreferred activities are clearly unrelated to the organization's ex-empt, or charitable, mission. When a university sponsors a rock concert at itsfootball stadium, it is not engaging in an educational activity. When a hospitalor a museum provides cafeteria services to its staff and visitors, it is not provid-ing health care or cultural preservation. In these cases, however, the profit gen-erated is "excluded" from taxation, under the IRS code, for such reasons as thatit is engaged in only occasionally or is undertaken for the convenience of itsmission-related beneficiaries (visitors to the facility). The fact that the revenueis not taxed is relevant to nonprofits' decision making, but that is different fromthe question of whether the activity per se is preferred by the nonprofit's con-trolling authority (board of directors and top management).

There are many nonpreferred activities that could generate profit to increaseprovision of the preferred good. Hospitals may (and some have) become foodcaterers and suppliers of commercial laundry or laboratory services. Museumsmay (and do) establish retail sales stores, on and off premises, that sell T-shirtsand other items having nothing to do with their exempt purpose except to gen-erate revenue. Nonprofit associations such as the American Medical Associa-tion publish journals that sell advertising space, thereby generating "unrelatedbusiness income," profit from which is taxable. These activities may tend to be


avoided by nonprofits for either or both of two separable reasons: the taxation,which tends to cause the organization to turn to other, untaxed, revenue-raisingdevices, and the nonpreferred or ancillary character of the activity.

These examples serve to illustrate the differences between the theoretic con-cepts of preferred and nonpref erred outputs, on one hand, and the regulatoryconcepts of untaxed (related and excluded) and taxed (unrelated) outputs, onthe other. As nonprofits seek funds they may be driven into nonpreferred mar-kets and activities; it is then in their interest, however, to minimize the taxespaid. This, in turn, leads in two directions, both with potentially important im-plications for public policy.

First, one probable consequence of nonprofits' commercial expansion is anattempt to push outward the regulatory boundary between taxed and untaxedactivities. Because of the breadth of nonprofits' legal missions and the limitedresources available to the IRS to police those borders, the expectation is thatnonprofits will continually be edging them outward. Second, when nonprofitsdo engage in unrelated business activities, they are likely not to choose activi-ties randomly but to select those for which expected profit is greatest, while bal-ancing profitability against the possible disutility of engaging in the activity. Itis likely that this process will lead to the selection of activities that, while non-preferred, are complementary, in either production or consumption, with theirpreferred outputs. Nonprofits can then be expected to allocate increasing por-tions of joint costs to their taxable activities, even if most or even all of thosecosts would have been incurred anyway, as their production of taxed activitiesincreases. Such cost allocations minimize tax burdens and maximize net, after-tax revenues for cross-subsidization of the preferred collective goods - there-by also posing issues for the process of competition between nonprofits and for-profit firms in those private-goods markets.

Our framework recognizes that nonprofit organizations often have opportu-nities to charge user fees for private services that are complements to their col-lective good missions. However, although hospitals charge patients they treat,colleges charge their undergraduates, museums charge attendees, and Scouttroops and community centers charge members fees, such revenue is not equal-ly available to all nonprofits. For example, nonprofits that support medical re-search or aid the poor - public television stations are another case in point -have fewer opportunities to charge user (or viewer) fees and, hence, are morelikely to pursue ancillary-goods markets to supplement donative income. More-over, as was noted earlier, charging a fee when possible entails the disadvan-tage that some target beneficiaries will be excluded. Thus, tensions are likelybetween nonprofits' pursuit of socially preferred, collective outputs and theirengagement in commercial markets where revenue raising is the goal.

Reappearing throughout subsequent chapters are the choices that nonprofitsmust make among the three sources of revenue:


1 donations of money and time from private individuals, corporations,foundations, and government;

2 user fees (charges to people who receive mission-related private ser-vices); and

3 sale of ancillary goods and services (i.e., not mission-related).

The relative dependence of nonprofits on each source varies considerablyamong industries. So do the forms, but this framework helps to organize under-standing of the choices across quite diverse segments of the nonprofit sector.When, for example, an academic health center charges fees to patients who arebeing treated for illnesses that are the object of its research mission, it is doingessentially what a museum is doing when it charges admission fees to visitors.Likewise, when it contracts with a private pharmaceutical company to performclinical trials (as part of the process of meeting Food and Drug Administrationrequirements for determining the safety and efficacy of new drugs), this is com-parable to a museum opening retail stores in shopping malls, or a nonprofit uni-versity press publishing a book with little scholarly appeal but a substantial salespotential - both are justified primarily by their contribution to revenue.

Both user fees and sale of ancillary services involve nonprofits in chargingprices for salable private goods. In choosing how much to do of each, nonprofitorganization management and trustees can be expected to balance the desir-ability of raising additional revenue, which can be used to augment mission-related output, against two forms of undesirability: that some members of a tar-get population will be unintentionally excluded (a possible consequence of userfees) and that the mission will be compromised (a potential result of engagingin ancillary activities).

This framework of nonprofit choice among outputs and sources of revenueis employed in a variety of ways throughout the ensuing chapters to describe,interpret, and assess the growing commercialism of nonprofit organizations inthe United States, to understand its probable future, and to highlight choices anddilemmas for public policy.

CHAPTER 4

Pricing and rationing by nonprofitorganizations with distributional objectives

Richard Steinberg and Burton A. Weisbrod

Introduction

The growing commercial activity of nonprofit charities, hospitals, educationalinstitutions, arts organizations, day-care centers, nursing homes, and religiousorganizations has led many to question the legitimacy of the nonprofit desig-nation and the concomitant tax and regulatory advantages conferred upon thesector. As noted in Chapter 1, if nonprofit organizations are engaging in com-mercial activity, it is easy to regard them as simply "for-profits in disguise"(Weisbrod 1988); however, there are many varieties of commercial activity.This chapter focuses on the ways in which nonprofit organizations and privatefirms can be expected to differ in their use of various pricing and other mech-anisms through which their goods and services are distributed. We emphasizethe potential importance of distributional goals to nonprofits - that is, their con-cern about reaching certain target populations - and suggest that insofar as theyhave distributional concerns they will use allocation mechanisms (pricing prac-tices, waiting lists, and the like) differently and have different effects on clientsand others, compared with private firms.

Economists and other social scientists have looked at many aspects of non-profit organization behavior, but we are unaware of any research that has exam-ined the array of alternative allocation mechanisms employed by nonprofits, letalone their relationships to organization missions. Thus, our most importantpurpose is to provoke researchers, practitioners, and other readers to look acrossindustry boundaries and particular allocation methods to refine understandingof the distribution of goods and services by nonprofit organizations - not onlywhat nonprofits do, but for whom they do it. We leave to future researchers the

We thank Lisa Whitecotton and Daniel Delany for research assistance, Kirsten Gr0nbjerg, RobertCoen, Wesley Lindahl, and participants at Northwestern University's colloquium on the commer-cial activities of nonprofit organizations for helpful suggestions. Steinberg acknowledges supportfrom the Indiana University Center on Philanthropy and from the Andrew W. Mellon Foundation.

65

66 Richard Steinberg and Burton A. Weisbrod

task of making our hypotheses more precise and performing appropriate empir-ical tests, but we focus on hypotheses likely to prove testable in practice. Muchof what we have to say applies equally well to allocation decisions by govern-ment agencies, even though our focus is on private nonprofit organizations.

Nonprofit organizations and governmental agencies generally espouse goalsother than profit maximization, such as fairness, redistribution, trustworthiness,public-goods provision, advocacy, education, preservation of values, accessibil-ity, social innovation, empowerment, participation, and promoting individualexpression. Profit maximization could provide a means to these ends, perhapsguiding the allocation decisions of a museum shop that sells T-shirts, a fund-raising campaign, or other ventures designed to raise revenue for the nonprof-it's preferred, mission-oriented services. Sometimes these other purposes areaccomplished as a side effect of profit maximization. For example, for-profitfirms must consider their reputations for fair and honest dealings if they are tokeep their customers. Sometimes, competitive pressures are such that the onlyalternative to profit maximization is bankruptcy. However, nonprofit organiza-tions and governmental agencies, motivated by other goals, often appear to al-locate their outputs persistently in a manner that does not maximize profits.

Some organizations legally defined as nonprofit may seek to maximize prof-its derived from some but not all of their activities. Other organizations legal-ly defined as nonprofit may seek to maximize profits derived from all activities,and thus function as for-profits in disguise. The purpose of this chapter is tocharacterize nonprofit departures from profit-maximizing behavior.

In the next section we delineate the varieties of fee schedules, nonprice re-quirements, and use of purchasing incentives, and present hypotheses regard-ing differences in the use of these mechanisms between nonprofit and profit-maximizing organizations. In a later section we outline the effects of allocationmechanisms on an organization's clientele, in order to develop corollary hy-potheses about the consequences of the differential allocational or distribution-al processes.

The varieties of nonprofit allocation

How are an organization's outputs allocated among potential consumers? For-profit firms use price as the primary allocation mechanism: A price is posted,those willing and able to pay the price receive the good or service, and the priceis bid up or down to the point where every potential consumer willing to payfor it is able to purchase the product. Monopolistic for-profit firms select a high-er price and produce a lower quantity, but still satisfy consumers' desired pur-chases at the selected price. For-profit firms also employ various forms of pricediscrimination, complex pricing mechanisms, and, more rarely, nonprice ra-tioning mechanisms. The unifying principle governing all these allocation tech-

Pricing and rationing with distributional objectives 67

niques is the quest for profits. Profit maximization implies that a perfectly price-discriminating firm would charge each consumer his or her reservation price -providing that the firm's marginal cost can be covered - although the cost ofdetermining that price limits its use. Auctions, tournaments, intertemporal pricediscrimination, nonlinear pricing that varies with the quantity purchased, andrationing by waiting all represent profit-maximizing responses under imper-fect information. Do nonprofits use these mechanisms differently from privatefirms?

Nonprofit organizations certainly use prices, although perhaps in a differentfashion. Organizations in both sectors employ a variety of complex pricingschemes that include sliding-scale fees and other forms of price discrimination,multipart prices with separate access and usage charges, nonlinear prices withcaps and deductibles. They also use nonprice allocation mechanisms such aswaiting lists, restrictive eligibility requirements, recruiting and marketing tech-niques, lotteries, and quality dilution.

Table 4.1 catalogs the variety of allocation mechanisms we discuss and sum-marizes our hypotheses about differences between their use by nonprofit andfor-profit organizations. Because of the variety of goals, outputs, environments,and constraints facing different kinds of nonprofit organizations, our hypoth-eses should be understood as pertaining to nonprofits in general, although notto all of them. Variations within the nonprofit sector in the use of these output-distribution mechanisms would provide further guidance for future researchseeking to test our hypotheses.

The simplest distributional mechanism is uniform pricing, of the textbookvariety, where consumers are offered the opportunity to purchase any quantityof a good or service at the same price per unit and this price is the same for allconsumers. Nonprofit organizations often employ such uniform pricing, partic-ularly for those activities that generate "unrelated business income" (incomegenerated from the sale of goods and services unrelated to the organization-al mission). A natural working hypothesis would be that nonprofits employuniform pricing under the same circumstances and in the same way as profit-maximizing firms. For-profit firms employ it when, given informational andother constraints, this practice would maximize the financial surplus eventual-ly distributed to owners. Nonprofits may employ it for unrelated goods whenthis practice would maximize the financial surplus available to further the or-ganizational public-goods or redistributional mission, or the private agendasof those in control of the organization. Profits are a means to an end in bothcases, although the ends differ. As Chapter 3 shows, the public-serving non-profits may seek to maximize profits in markets for some goods in order to fi-nance their mission-related activities.

However, Schiff and Weisbrod's 1991 adaptation of James's 1983 modelsuggests that unrelated business activities are disfavored in organizational util-


ity, so that when uniform pricing is used, nonprofits will underproduce unrelat-ed goods and services relative to the profit maximizer (see Chapter 3). Thus,any empirical test of this hypothesis would need to specify which activities aredisfavored or nonpreferred by which nonprofit organizations.

Sliding-scale fee schedules are a form of price discrimination in which con-sumers pay different prices per unit based on observable characteristics of theclient (such as income, family size, age, or the severity of the client's dysfunc-tion) rather than on the particular costs of serving that consumer. Sliding scalesare used by both nonprofit and for-profit sellers, and they are common in daycare for children, mental-health services, church and synagogue membershipcharges, and professional-society dues. Similar practices appear in other indus-tries under a variety of names. For example, college financial aid granted on aneed basis creates a sliding scale for net-of-aid tuition. Diverse copayment rateson health insurance create price discrimination in net health-care costs to pa-tients, based partly on observable consumer characteristics. Special prices forstudents or retirees at museums and on public transit systems are other instancesof price discrimination equivalent to a sliding scale.

Nonprofit organizations that are concerned with the consequences of theirallocation mechanism would ordinarily choose an income-based fee schedule,with high-income consumers paying higher fees. Profit-maximizing organiza-tions would also offer income-based fees (when income is the best feasibleproxy for consumer willingness to pay), because this way they could raise pricesfor some without losing other customers who could be profitably served. In thisrespect, nonprofit and for-profit sliding-scale fees would look quite similar.However, the lowest price selected by a nonprofit that had distributional objec-tives would often fall below the marginal cost of serving this group, whereasprofit maximizers would select a lowest price at or above marginal cost (Stein-berg and Weisbrod 1997). The fact that nonprofit organizations provide servicesfree of charge to consumers who will never become paying customers supportsthis hypothesis, although one needs to distinguish voluntary free care from char-ity care required by regulators and from care purchased and paid for by thirdparties such as government.

Sometimes price discrimination by nonprofits is tied simply to consumers'stated willingness to pay more, and to no other characteristic. For example,many nonprofit organizations have membership categories (supporting mem-ber, patron, etc.) that charge different fees but provide essentially the same ser-vice (or provide token added benefits that do not appear to justify the higherfees). In effect, the consumer is mixing a donation and a purchase. Supportingmemberships do not count as donations under the tax laws insofar as there isan exchange of value, but because the value received by the consumer is un-altered by the consumer's selection of a membership category, these member-ships combine donation with purchase, thereby constituting what amounts to

Table 4.1. Nonprofit use of allocation mechanisms

Allocation mechanism Examples Hypotheses about nonprofit departures from profit-maximization

PriceUniform pricing

Sliding-scale fees(interpersonal price discrim.)

Voluntary price discrim. (wheneligibility for particular pricescannot be verified by the seller)Intertemporal price discrim.Noncash payments

Day care; Mental health care;Professional-society dues;Net-of-financial-aid tuition

Supporting-member dues; Donationsto arts orgs.; National PublicRadio; Volunteering

Free-entrance days

Habitat-for-Humanity pricing(fees plus "sweat equity")

NonpriceWaiting lists

Eligibility reqmts.

Externally imposed eligibilityreqmts.

Quality dilution andopportunistic quality sharing

Product bundling

Day care; Nursing homes;Colleges, universities

University admissions ("merit");Fraternal societies; Religious orgs.;Work-shelters for the handicapped;Food pantries

Meet govt. contracting reqmts.;Conform with tax/regulatory reqmts

Soup kitchens; Homeless shelters;Museums and zoos; Worker training

Museums; Colleges, universities

Recruiting target populations Hospital location;School field-trips to museums

Nonprofits and for-profits use uniform prices identically for those outputsunrelated to the nonprofit mission (except for disfavored activities).

Nonprofits more likely to price-discriminate and to choose a lowest price that isbelow the marginal cost of serving their customers.

Nonprofits use extensively. The practice is not generally feasible for for-profits(except in cases where the volunteer can control, at least in part, who benefitsfrom her contribution, as in for-profit day care).

Nonprofits more likely to use this for those who will never be profitable to serve.

Nonprofits may require partial payment in the form of labor or in-kind.For-profits stick to cash.

Nonprofits more likely to use waiting lists. For-profits more likely to react topersistent excess demand by expanding capacity or increasing price.

Nonprofits more likely: (1) to use requirements poorly or negatively correlatedwith willingness to pay; (2) to restrict eligibility to those who cannot pay.For-profits more likely: (1) to use requirements positively correlated withwillingness to pay; (2) to use eligibility requirements to establish a niche market.

Nonprofits more likely to construe requirements broadly for unprofitable clientsand narrowly for profitable clients.

Nonprofits more likely to hold excess capacity in order to avoid quality dilution.For-profits more likely to dilute quality in cases of contract failure.

Nonprofits more likely to bundle "merit goods" in pursuit of paternalisticobjectives. For-profits more likely to bundle in pursuit of profit.

Nonprofits more likely to target mission-related populations regardless ofexpected future profitability.


voluntary price discrimination. Donations in general constitute a form of volun-tary price discrimination because each donor determines the size of contribu-tion (Hansmann 1981a; Ben-Ner 1986; Spiegel 1995); the same holds for do-nations of volunteer time.

We hypothesize that sliding scales and voluntary price discrimination aremore common in nonprofit than for-profit organizations for three interrelatedreasons. First, there is more scope for a sliding scale when the range of consum-ers is broad, as it is for nonprofits that serve persons unwilling or unable to paymarginal costs. Second, consumers are more likely to reveal their willingnessto pay truthfully when dealing with a nonprofit because they know that whenrevelation causes them to pay a higher price, that price is supporting a missionthey support rather than lining a stockholder's pockets (Ben-Ner 1986). Third,the redistribution accompanying the use of sliding-scale fees sometimes consti-tutes the nonprofit's mission, rather than serving merely as a means for raisingrevenue to accomplish some other mission.

Other forms of interpersonal price discrimination are based on occupancyrates and other such production-cost factors rather than on consumer charac-teristics. In the for-profit world, this kind of price discrimination is commonwhenever marginal costs are markedly lower than average costs. Commercialairlines, hotels, and movie theaters discriminate by varying prices seasonallyor by day of week or time of day, in order to fill empty space, furthering theirprofits. The key question is whether nonprofits price-discriminate in order toachieve other, nonfinancial goals. If nonprofits employ first-come, first-servedopportunities or waiting lists rather than higher prices to adjust to peak-leveldemand, these forms of price discrimination would be evidence of nonfinancialgoals.

Nonprofits, like for-profit firms, employ various forms of intertemporal pricediscrimination. Performing-arts groups discriminate over time when they distin-guish advance ticket prices from at-the-door prices. We have no hypothesis re-garding sectoral differences in this practice. However, some nonprofit museumscharge admission fees but waive them intermittently on a regular basis, offer-ing, for example, Free Tuesdays. Totally free admission, even on an occasionalbasis, appears to occur more typically in the nonprofit sector. For-profits some-times offer promotional prices and free samples in an effort to convert theseconsumers to paying customers, but rarely do they offer free services targetedto those who will never pay the marginal costs of provision.

Sometimes nonprofit organizations require consumer or user fees in non-monetary forms, such as labor or a mixture of cash and other forms of payment.For example, Habitat-for-Humanity, a nonprofit devoted to providing low-income housing, requires its customers to pay in both cash and labor ("sweatequity"). This type of pricing is uniquely well suited to nonprofits. For-profitscould certainly benefit from labor supplied by their customers but prefer them


to pay exclusively in the form of cash because it can be used to hire more ef-ficient workers. In contrast, nonprofits may prefer this "inefficient" paymentscheme when it directly supports some part of the organizational mission (e.g.,encouraging self-help) as well as supplies resources (e.g., labor, materials, andmoney for advancing the housing-construction mission).

Nonprice allocation mechanisms, while sometimes combined with positiveprices, may also serve to allocate a good or service that is given away at a priceof zero. Most obvious is rationing by waiting, through either explicit queuingor the use of waiting lists. We hypothesize that nonprofits are more likely to usewaiting lists, whereas for-profits are more likely to react to persistent excessdemand by expanding capacity and/or increasing the price. Evidence support-ing this hypothesis for the nursing home and mentally handicapped facility in-dustries is presented in Weisbrod (1988, 1998).

When a nonprofit gives goods away or sells them at a price below that whichwould clear the market, it often specifies formal eligibility requirements thatserve to ration the available output. Consumers may be required to provide evi-dence (or at least assert) that they meet eligibility standards involving income,wealth, family size, gender, religious affiliation, political persuasion, age, em-ployment status, disability, or the like before they are eligible to receive thegood or service. Sometimes organizations define their own criteria for deter-mining eligibility for service; other times the eligibility criteria are defined bya contracting government agency. In either case, nonprofit and for-profit orga-nizations may differ in the forms of eligibility requirements and the ways theyare implemented, insofar as their organizational goals are distinct. We consid-er, in turn, the cases of internally and externally selected eligibility rules.

When the organization freely chooses its eligibility criteria, we hypothesizethat nonprofits are more likely to use requirements that are poorly (or negative-ly) correlated with the consumer's willingness to pay, and may even restrict eli-gibility to those who are unable to pay. Profit maximizers use eligibility require-ments to extract more consumer surplus through price discrimination: "Supersaver" airline fares restricted to those who stay Saturday night are an attemptto fill seats with those who would not pay higher fares, while allowing the air-line to charge more to those - largely business travelers - willing to pay more.Discounts by hotels and car-rental firms to American Association of RetiredPersons (AARP) members are another attempt to use eligibility requirementsto segment the market. Eligibility requirements are used by profit maximizersas screens for generally unobservable willingness to pay; they allow firms toprice-discriminate without incurring the public-relations backlash and threat ofantitrust suits that would accompany a more direct attempt at perfect price dis-crimination across individuals.

Nonprofits may also use eligibility requirements to raise additional revenueby capturing more consumer surplus - using the revenues to finance manage-


rial perks or to cross-subsidize other, mission-related outputs; or they may usesuch requirements to further the organizational mission directly, by reaching outto particular target populations. Thus, nonprofits, we hypothesize, differ fromfor-profits in the nature of eligibility requirements, defining their target clien-tele so as to further their social mission directly rather than to maximize prof-it. Nonprofits often provide services only to the poor, to persons of a particularreligion, ethnicity, or age, or to the most severely afflicted. Overall, insofar asthey differ from for-profits in their goals, they will employ restrictions that arewell correlated with mission but less well correlated with consumer willingnessto pay. Unfortunately, it is difficult to test this hypothesis cleanly because prof-it maximizers also might use requirements that are poorly correlated with indi-vidual willingness to pay. In the case of for-profits, standards poorly correlatedwith individual willingness to pay are not instruments for advancing a socialmission but part of a strategy to establish a niche market by appealing to snob-bery, class interests, ethnicity, and the like. Eligibility requirements that serveto differentiate the product provide some monopolistic power to the seller, andso enhance profits.

When government agencies contract with nonprofits or private firms to pro-vides services such as home health care or primary-school management, it istypically the agencies who specify eligibility requirements. For example, Smithand Lipsky (1993,142-3) cite the case of emergency shelters in Massachusettsthat are obligated by government contractors to serve the "most disadvantaged"first. Nonprofits and for-profits may differ, however, in the faithfulness withwhich they implement these requirements. This is important because it is diffi-cult and expensive for the government contractor to determine whether the cli-ents served by an external agent actually are the most disadvantaged or meetother (possibly subjective) contract criteria. We hypothesize that sectoral differ-ences in behavior will depend upon whether the target population can be servedprofitably under such conditions. If clients cannot be served profitably, nonprof-its pursuing social goals will be less likely than for-profits to take advantage oftheir informational superiority to interpret the contractually imposed eligibilityrequirements in a fashion that restricts access; if, instead, clients can be servedprofitably, nonprofits would be more likely to interpret eligibility requirementsin a fashion that permits access only to the contractor's intended target clien-tele. Profit maximizers will tend to use their inevitable discretion by operation -alizing eligibility requirements in ways that further profits, declaring "border-line" clients eligible if serving them would be more profitable than serving moredisadvantaged persons, declaring them ineligible otherwise. If a contract speci-fies more than one fee class, profit maximizers would place clients in the mostprofitable fee class they can get away with. For example, Medicare paymentsto hospitals depend on the Diagnosis Related Group (DRG) in which a patientis placed. Even before the system was introduced in October 1983, computer


software programs had been developed to determine the highest-paying DRGconsistent with a given patient's specific conditions and diagnoses. Through theyears that followed, the phenomenon of DRG creep - the drift of patient clas-sification by hospitals into higher-priced diagnostic groups - was observed(Steinwald and Dummit 1989). It remains unclear, however, whether oppor-tunism is the source of this phenomenon.

Whether this discretionary behavior differed between nonprofit and for-profit hospitals is also not clear. In general, nonprofits are thought to be moretrustworthy contractees, reducing the need for the contractor to monitor andenforce contract compliance (Ferris and Graddy 1991) or to engage in lengthynegotiations over contract terms (DeHoog 1984). On the other hand, the non-profit mission (serving clients) may conflict with the government-agency mis-sion (keeping costs down while appearing to the electorate to be handling thesocial problem), and in these cases nonprofits would be less appealing to thecontractor. In the case of Medicare, if a nonprofit hospital feels that certain cost-ly procedures are medically desirable for a particular patient, it might place thatpatient in a more lucrative DRG rather than provide those procedures at a lossor prescribe less costly treatment. For-profit hospitals might prefer to place thepatient in a lower-priced DRG that is more consistent with government guide-lines and not attempt to provide the more expensive procedures.

A third form of nonprice allocation is quality dilution. For example, a soupkitchen faced with unexpectedly large demand may thin its "gruel"; a homelessshelter may line the floors with somnolent souls; a hospital may put three pa-tients in a room designed for two, or place patients on gurneys in corridors; anda college may increase its student-faculty ratio. We hypothesize that nonprofitshave output quality as a goal, and so are more likely to hold excess capacity inorder to avoid quality dilution in times of unexpectedly high demand. This pre-diction is consistent with the theory presented in Holtman (1983). In contrast,for-profits may dilute quality even when demand is as predicted in order to cutcosts and increase profits. This occurs when there is contract failure - where,owing to either the circumstances under which the good is provided or the na-ture of the good, the consumer is unable to verify that the quality of the prod-uct is at the promised level (Hansmann 1980).

Some nonprofits employ product bundling, requiring purchasers of one prod-uct also to purchase other products. Thus, a homeless shelter may require res-idents to obtain psychological or addiction counseling; a parochial school mayrequire students to attend religious services, buy a school uniform, and do vol-unteer work. For-profits also employ product bundling, but do so in order toenhance profits rather than fulfill a social mission. James (1986) argues that adesire to inculcate values and tastes lies behind the decisions to found an edu-cational, health, or social-service agency and to organize it as a nonprofit. Sheuses the example of religious organizations:


[T]he object is not to maximize profits but to maximize religious faith or religious ad-herents, and schools are one of the most important institutions of taste formation or so-cialization. Similarly, hospitals are a service for which people will have an urgent needat times, and so constitute a good way for religious groups to gain entry. The nonprofitform is chosen because the main objective is often not compatible with profit-maximiz-ing behavior. For example, religious schools, set up to keep their members within thefold, may have to charge a price below the profit-maximization level in order to enticethe largest numbers to enroll, (p. 155)

Sometimes nonprofits go beyond product bundling, preferential pricing, andeligibility restrictions in order to solicit actively the participation of target popu-lations. Thus, nonprofit hospitals are more likely to locate in low-income areas(Pauly 1987), and museums negotiate with schools to arrange field-trip visits.

Effects of nonprofit allocation

In this section, we characterize the broader impact of nonprofit price and non-price allocation practices on customers and donors. As summarized in Table 4.2,we hypothesize that nonprofit behavior produces a different division of econom-ic value between consumers and producers and among consumers; a differentallocation of the risk that the product will cost more (or be of lower quality orquantity) than anticipated; a different self-selection among potential consum-ers; a different signal to consumers and donors; and different incentives to econ-omize on resource consumption.

Market activities generate economic value divided between that received byproducers (producer surplus) and consumers (consumer surplus). Producer sur-plus is the difference between revenues received and the (variable) cost of pro-duction; consumer surplus is the difference between consumers' willingness topay and their outlays of time and money for the product in question.

In the for-profit world, total surplus is determined and divided so as to pro-vide the largest piece of the "surplus pie" to the owners of the firm. Sometimes(as in uniform pricing by a for-profit monopoly), the pie is made smaller as aside effect of providing a larger piece to the producer, but the size of the produc-er piece (in absolute terms, not as a share) is maximized. Profit-maximizingfirms would be unconcerned with the distribution of either output or consumersurplus among different types of client unless consumers were to rebel strong-ly against perceived unfairness. This sort of consumer reaction would reducethe equilibrium size of the piece given to producers, but producers would stillact to maximize the size of their piece. In contrast, nonprofits often care aboutthe allocation of surplus even when sales revenue would be unaffected or neg-atively affected by the nonprofit's pursuit of its concept of equity. We hypoth-esize that nonprofits are more likely to expand consumer surplus at the expenseof producer (or even total) surplus.


Table 4.2. Effects of nonprofit distributional mechanisms

Effect of allocation Hypotheses about differing effects of nonprofit and profit-maximizingmechanism behavior

Dividing total surplusBetween consumers For-profits maximize producer surplus. Consumer surplus is providedand producers only insofar as it enhances producer surplus.

Nonprofits are more likely to expand consumer surplus at the expense

of producer (or even total) surplus.

Among consumers For-profits do not care except insofar as the distribution of consumersurplus affects producer surplus.Nonprofits are more likely to consider fairness and distribution in thedesign of their allocation, cost-sharing, and pricing mechanisms.

Allocating risks For-profits bear risk only when there is a compensating increase in return.Nonprofits are more likely to bear the largest feasible share of risk,regardless of compensating financial returns, in order to ensure that theirclients are not harmed.

Screening For-profits use mechanisms designed to uncover those with highestability and willingness to pay.Nonprofits are more likely to use mechanisms designed to uncover thosewith greatest "need" (as in addictions treatment or child-abuseprevention), those who are most worthy (say, because they appear to bepious), or those with high willingness but not ability to pay.For-profit mechanisms are also designed to deter those who would bemost costly to serve (cream-skimming).Nonprofit mechanisms are less likely to deter high-cost clients; on theother hand, nonprofits are more likely to use exclusionary screens fornoncost reasons (e.g., to deter those who are poorly motivated).

Providing a signal oftrustworthinessTo consumers Nonprofits are more likely to use waiting lists and highly selectiveand clients eligibility requirements to signal their quality.

For-profits are more likely to use high prices to signal quality.

To donors and Nonprofit use of fees depends upon the preferences of major fundersgrantmakers regarding provision of seed money vs. operating support.

Providing an incentive Nonprofits are less likely to use high prices and more likely to use non-to economize on price mechanisms to provide incentives for consumers to economize onresource consumption the use of nonprofit outputs.

Nonprofits often use token fees to signal the importance of economizing.

The possibility that at least some nonprofit social-service agencies occupya middle ground has been presented by McCready (1988), who considers theiruse of Ramsey pricing. This is a self-financing approach, in which user fees areset so as to maximize the size of the pie, and is therefore intermediate to prof-it maximization (expanding producer surplus at the expense of total surplus) and


some nonprofit notions of equity (expanding consumer surplus at the expenseof total surplus). It would finance reductions in price for some nonprofit out-puts by increasing prices for other, revenue-generating outputs. Whether theprice of a particular output is increased or decreased depends upon whetherthe quantities of desired purchases by consumers and desired sales by produc-ers are very sensitive to price. Price-sensitive outputs would be priced belowmarginal cost, whereas insensitive outputs would be priced above marginal cost.Ramsey pricing is designed neither to accomplish redistributional goals nor totreat some outputs as more mission-related than others; but it does differ fromprofit-maximizing pricing in that it results in a larger total of consumer plus pro-ducer surpluses.

Nonprofits may also be concerned with the distribution of surplus amongconsumers, seeking "fairness" for its own sake, apart from revenue implica-tions. We hypothesize that nonprofits are more likely than for-profit firms to ar-ticulate and implement notions of fair distribution in their choice of pricing andnonprice rationing mechanisms. One notion of fairness sometimes employedis the "fair share of cost" notion, usually implemented as equal division of thecosts of providing a collective good. Lohmann (1980) notes:

This method of fee collection is most appropriate for short-term projects or projects inwhich clients form a closely knit group whose unity is reinforced by such "sharing." ...[This method] may be of greatest interest to those involved in community organizationand group work activities, in which the collection of fees on an occasional basis for spe-cific projects may be the only fund-raising activity necessary. Senior citizens groups tak-ing bus tours, youth groups planning parties, and social action groups planning demon-strations or protest marches may best be handled using the fair-share-of-cost method,(p. 86)

Other notions for fair nonprofit pricing borrow from the literatures on taxequity and bargaining fairness. Prices, like taxes, may be thought fair if theyare proportional to either the benefits received or the consumer's ability to pay.Horizontal equity (equal prices for similarly situated consumers) may also bethought to be important and to constrain nonprofit patterns of price discrimi-nation and market segmentation. We are unaware of any nonprofits that specifi-cally seek to implement formal notions of bargaining fairness (see, e.g., Mou-lin 1988; Young 1994), but clearly the potential is there.

Although some nonprofits are particularly concerned with fairness, othersdistribute their outputs in a,manner contrary to their stated mission. Favoritism,nepotism, kickbacks, self-dealing, and other abusive power relationships cangovern nonprofit allocation, particularly when allocation criteria require sub-jective judgments on the part of the nonprofit employee. There is more scopefor such abuse in the nonprofit sector because nonprofits are not subject to finan-cially motivated takeover bids that limit abuse among for-profits. On the otherhand, insofar as nonprofit managers and board members are dedicated to the


organizational social mission, this provides an internal check on abuse (Handy1995).

The choice of fee (or other rationing) structures also affects the allocation ofrisk. For example, a daily fee - say, on hospital care - places all the financialrisk that the inpatient stay will be unexpectedly lengthy on the consumer, where-as a fee based on completion of a specified task (such as the DRG-based sys-tem for reimbursing hospitals for their Medicare patients) places all the risk onthe provider. It would be worthwhile to explore whether nonprofits differ fromother organizations in their proclivities to accept or shift risk. There is little ex-tant theory to turn to for predictions, but much would seem to turn on two fac-tors: First, differences in organizational goals are likely to prove important.Those nonprofits that desire to serve target populations regardless of ability topay would seem to be more likely to bear this risk than for-profits or other sortsof nonprofit. Second, budgetary flexibility is likely important. Those nonprof-its that receive substantial unrestricted donations would seem to be more ableto bear risk. At the same time, the higher costs of borrowing because of non-profits' inability to sell shares of stock might make them less willing to bearrisk (Hansmann 1981b).

Profit-maximizing organizations are not directly concerned with the compo-sition of clientele, caring only insofar as the client mix affects aggregate net rev-enues. Abstracting, for the moment, from any differences in the costs of serviceprovision to various clients, the most desirable are those with the greatest will-ingness to pay for the service in question (or those who can attract other clientsof this kind). When willingness to pay is not directly observable, which is typi-cally the case, the price mechanism can serve as a screen to distinguish thosewith highest demand. Thus, nonprofits may use pricing for the same reason thatfor-profits do: to generate revenue from those who can pay (Moran 1985).

Even with identical desires to distinguish among consumers on the basis ofwillingness to pay, nonprofits with distributional objectives would differ fromprofit maximizers in their choice of fee structures: Profit-maximizers wouldselect fees that are as close as possible to the consumer's reservation prices,whereas distributionally concerned nonprofits would screen the same consum-ers but generally charge them lower prices (Steinberg and Weisbrod 1997); suchnonprofits would forgo some revenue in order to advance their overall goals,balancing their pursuit of distributional goals with their revenue needs.

Consumer willingness to pay results from both the consumer's preferencesfor a particular good or service and that consumer's ability to pay. When twoconsumers have the same ability to pay, differences in willingness to pay re-flect preferences; however, in general one cannot distinguish whether someonewho is willing to pay a large sum is wealthy and relatively unconcerned withthe cost or is of modest means and greatly desires the good or service. Non-profits sometimes care about the intensity of want or need, apart from wealth,


whereas profit-maximizers always care about the willingness-to-pay compos-ite. Sliding-scale fees based on income allow nonprofits to screen on the basisof innate preferences by correcting for differences in purchasing power, as Os-ter(1995) explains:

It is common practice . . . for private schools to require that even the most indigent stu-dent pay at least a small fraction of tuition. This practice helps to sort among parentsand identify those families with the highest value on education, (p. 99)

Price serves a screening function whether or not the nonprofit wishes to servethose with the greatest willingness to pay. Soup kitchens, for example, are moreconcerned with hunger than ability to pay. If the need for a subsidized meal isobservable, price does not have to serve this screening function - the combina-tion of price, or some other rationing device, and eligibility restrictions wouldbe sufficient. However, if the nonprofit desires to serve those with an importantbut unobservable characteristic - for example, "true need" - price is at best animperfect screen, most useful if the unobservable attribute is correlated withwillingness to pay. At worst, price may be counterproductive: Any price that issufficiently low to allow low-income households to purchase the good or ser-vice will also attract higher-income households that do not need a subsidizedprice. As a result, absent stigma and other social/psychological factors, pricescreens will fail to distinguish the target population reliably. In this case, ad-ditional or alternative screens might be used - for example, locating the soupkitchen in a run-down neighborhood will keep away the nonpoor looking fora free lunch at the same time it provides a convenient location for the needy.

Having a waiting list, whether for admission to a nursing home or for sup-per at a soup kitchen, may be another effective device for sorting out consumerswhen the organization's goal is not profit maximization but aiding a particu-lar target population (Nichols, Smolensky, and Tideman 1971). People who arewilling to wait are showing the intensity of their preferences and the availabil-ity of alternatives, and the nonprofit may wish to serve persons for whom alter-natives are poor. The problem is to get people to reveal their true circumstances.A soup kitchen can give food away at no charge, but its budget constraint anddistributional goal require it to try to target the poor and only the poor; sincethey generally have low opportunity costs of time, waiting in line helps to sortout those who are most in need, for they are the ones most likely to be willingto wait. In the case of nursing homes, willingness to be on a waiting list for ad-mission provides some basis for sorting prospective patients who are willingand able to pay a higher price from those whose incomes are more limited.Thus, although either a waiting line or a waiting list is inefficient given a goalof profit maximization, they may be efficient insofar as nonprofits' objectivefunctions include distributional goals.


The possibility that the screening, or demand-revelation, function of pricewill be counterproductive is highlighted by two examples. Titmuss (1971) not-ed the problem with using consumer payments for blood to compensate blooddonors. Blood markets suffer from informational asymmetry problems: Evenpersons who had reason to believe their blood was tainted by hard-to-detectinfectious diseases - hepatitis was the focus at the time - could profit nonethe-less from the sale of their blood because buyers were unable to distinguish"good" blood from "bad." If, argued Titmuss, blood money were prohibited,then only altruists, who presumably would not wish to inflict their diseases onothers, would supply their blood. A second example of the limitation of priceas a screening mechanism involves drug and alcohol treatment centers and thosepersons who deny their addiction problems. Charging them fees would be coun-terproductive, all else held equal, because those in denial would be least will-ing to pay for treatment.

The effectiveness of the screening function of fees for organizations in ei-ther sector depends on the behavior of competitors, both non- and for-profit. Ineither sector, a high fee cannot be used to screen for high-willingness-to-payclients unless competitors follow suit. The same is true for nonprofits that usewillingness to pay as a screen for some other characteristic that concerns them.For example, a graduate department of economics concerned with output qual-ity, not monetary profit, might wish to admit the "best" students regardless ofwillingness to pay. Presumably those who could use a graduate education as aspringboard to a highly successful career would be willing to pay more than oth-ers, all else being equal, but aid offers from actively competing schools wouldbid up the required amount of aid, generate rents for the student, and therebyeliminate any department's ability to use high fees as a successful screen forthe student's supply price.

Another aspect of screening that deserves mention is cream skimming, whereorganizations seek to serve the most profitable clients. For example, when thegovernment contracts with a private job-placement and training agency, a prof-it maximizer paid on the basis of the number of persons it places has an incen-tive to screen out hard-to-place clients. Similarly, cream-skimming insurancecompanies try to screen out those who drive dangerously, and cream-skimminghospitals have a profit-related incentive to screen out poor and uninsured pa-tients. Nonprofits, if they pursue broader public-interest goals and especially ifthey have distributional goals that favor the poor, would differ from for-profitsin their propensity to cream-skim. It is not obvious, however, in which direc-tion they would differ. If the nonprofit hospital accepts more low-cost, insuredpatients and receives reimbursement that does not distinguish, within Diagno-sis Related Group, low- from high-cost patients, it obtains more net profit thatcan be applied to the treatment of other target-group patients or toward further-


ing some other element of the nonprofit mission, such as community health edu-cation. In any case, in unregulated-price markets, competitive pressures mayforce all providers to attempt to cream-skim, although this same competitionwill drive prices down to marginal costs, depriving all competitors of profitfrom low-cost patients.

The allocation mechanism chosen sends signals to consumers and others asto the quality of service. If an organization thrives after maintaining a higherprice than its competitors, it is signaling to prospective consumers, by its sur-vival, that other customers find the product to be of higher quality than the com-petitor's. We hypothesize that nonprofits are less likely to use price as a signalof quality because use of this kind of signal conflicts with their distributionalobjectives. Instead, nonprofits are more likely to use waiting lists or highly se-lective eligibility requirements as a signal. We have no reason to believe eitherapproach would more reliably convey information about quality to consumers,but do hypothesize that different modes of signaling are employed by the twosectors.

Allocation mechanisms also provide a signal to third-party funders (privatedonors, grant makers, and governments) about the social value of service pro-vision, as Dan Delany (personal communication, 1996) explains:The fee and wait-list combination [chosen by an alternative health center for people withAIDS] . . . signaled a level of legitimacy for a service that had not been viewed as suchfor some time. Clients willing to pay, even after waiting in excess of a year for service,was a powerful statement to fellow service providers, government officials charged withdelineating funding priorities, and other funding sources.

The choice of allocation mechanisms also signals to third-party funders theorganization's competence and philosophy. The choice of fee structures signalsthe funded nonprofit's managerial competence and degree of commitment tomove towards self-sufficiency, which many funders value. On the other hand,allocation mechanisms may signal an organizational insensitivity to need. Whena soup kitchen charges fees, funders may question the organization's commit-ment to serving the needy, the quality of its management, or the kitchen's long-term stability. Thus, we hypothesize that nonprofit use of fees depends upon thepreferences of major funders and the nature of the service provided. (The po-tential effect of user fees on donations was noted in Chapter 3.)

The ability of any organization to contribute to the public interest in effi-ciency and distributional equity depends on the actions of both the organizationand its customers, patients, or clients. Consumers should have reason to restricttheir use of costly services, to seek less costly alternative modes of service de-livery, to take preventive actions that reduce their need for services, and to de-vote effort toward being a "good" client on whom agency efforts will not bewasted. The high fees charged by profit maximizers provide consumers with areason to economize in all these ways, even though it is no part of the profit-


maximizer's profit-making mission to help society in this way. Nonprofitswhose distributional concerns make them averse to high fees would not havethis automatic beneficial side effect of solving these "moral hazard" problems.Thus, some nonprofits temper their distributional concerns and charge tokenfees, and others provide efficiency-inducing incentives through nonfee mecha-nisms. David E. Mason (personal communication, 1996) provides an example:

[T]he Laubach Literacy organization . . . published teaching and reading materials . . . ,trained volunteer tutors, and conducted adult literacy programs . . . in 17 countries. . . .Though the money paid by those who used the reading material was so insignificant thatthe central office reaped virtually no income from it, we always charged a token fee. Wefound the learners took better care of the material, and attached a value to it because thefee was required.

In the same vein, fees can induce effort on the part of a third-party payer rath-er than the client. Private education, where the parents are the third party, pro-vides an example (Oster 1995):

[Private-school fees] also help to insure that parents have some "stake" in their children'seducation, so that price plays an ideological role as well. Given the extent to which edu-cation has been increasingly seen as a collaborative effort between parents and schools,some pricing may be quite important, (p. 99)

Although user fees usually provide positive efficiency-inducing incentivesto clients and third parties, there are dangers. Lohmann (1980) points out that,in the field of mental health,

[t]he use of strong-arm tactics in the collection of past-due accounts . . . would be con-sidered inappropriate.... One mental health agency . . . has largely solved this problemby linking payment of fees to the clinical context: reluctance to pay fees is interpretedas a non-cooperative attitude on the part of the client, and the issue is taken up by theclinical staff. Such an approach, however, may strike some as taking unfair advantageof client vulnerabilities. Certainly, it can be seen that significant tensions can arise be-tween the "helping" thrust of an agency and its need for fee revenue, (pp. 81, 89)

Conclusion

Nonprofit organizations and for-profit firms both engage in commercial activ-ities, charging fees for services provided. Some analysts, policymakers, andmass media conclude from this observation that commercial nonprofits are sim-ply for-profits in disguise - a conclusion that is premature. Unlike for-profits,nonprofits have a variety of distributional and other "bonoficing" objectives,and they operate under different legal constraints, so that they may set theirprices (and design their nonprice allocation mechanisms) on a different basis.In this chapter, we have illustrated a wide variety of ways in which distribution-al goals might be pursued, each suggesting a testable implication. Our tests are


informal and, we hope, suggestive, but they call for formal testing and interpre-tation. A balanced assessment of the commercial activities of nonprofit organi-zations requires analysis of the magnitude of differences between nonprofits'and for-profits' pursuit of distributional objectives, a judgment regarding the so-cial value of pursuing distributional goals through nongovernmental organiza-tions, and evaluation of alternative means of pursuing these objectives.

We do not assert, nor have we demonstrated, that nonprofits differ from pri-vate firms in their concern for how outputs are distributed among beneficiaries.All nonprofits are not the same, any more than all private firms are. Nonprof-its may not behave the same way in all markets, perhaps acting differently inancillary-goods markets, where the focus is on revenue generation, than in theiruse of user fees for their target populations, where revenue generation may bebalanced with distributional considerations. The distinctions we have discussedin this chapter are intended as conjectures that call for testing. It is important todetermine whether, and under what conditions, nonprofits are effective mecha-nisms for expressing complex social goals involving output distribution.

CHAPTER 5

Differential taxation of nonprofits and thecommercialization of nonprofit revenues

Joseph J. Cordes and Burton A. Weisbrod

Introduction

As the nonprofit sector has grown in economic importance, so too has its reli-ance on income obtained from a variety of commercial activities - that is, activ-ities involving the sale of goods or services. Because commercial income earnedin the nonprofit sector is largely, if not entirely, tax exempt, this trend raises sev-eral questions: Are nonprofit organizations encouraged to pursue commercialventures that they would otherwise eschew because the income from such activ-ities is tax exempt? When nonprofit enterprises move into commercial markets,does their tax-exempt status give them an economic advantage? Does the pros-pect of earning tax-free income create incentives for nonprofit organizationsto divert their energies from core philanthropic activities in order to pursuecommercial ventures that are justified only because they generate revenue forthe core activities? How does the differential tax treatment of nonprofits affectchoices among commercial ventures?

To answer these questions, we must recognize some distinguishing attributesof nonprofits (Weisbrod 1977,1988; Hansmann 1980; James 1983; Steinberg1991; Eckel and Steinberg 1994). Unlike for-profit enterprises, nonprofit orga-nizations may regard the pursuit of profit in any market as a means of achiev-ing other "more important" organizational goals, rather than as an end in itself.In addition, donations, an important source of revenue for many nonprofits, maybe sensitive to these commercial activities. For example, donors may care abouthow nonprofits divide their efforts between profit-making activities and pro-ducing what donors believe to be the organization's "primary" output.

We are grateful to David Bradford, Charles Clotfelter, Susan Rose-Ackerman, and Bruce Daviefor commenting on earlier drafts; to Cecilia Hilgert, Margaret Riley, Michael Alexander, DanielSkelley, and Gary Jurevich, of the IRS, for help in obtaining and interpreting data on nonprofitorganizations; and to Cagla Okten for research assistance.

83

84 Joseph J. Cordes and Burton A. Weisbrod

For these reasons, nonprofit managers may view commercial ventures dif-ferently than would for-profit competitors, and these differences must be con-sidered when analyzing the effects of differential tax treatment of nonprofit en-terprises. For a nonprofit organization, decisions on whether to engage in anycommercial activity often quite rightly involve more than expected profitabil-ity. Following Chapter 3 and James (1983), our basic theoretical framework isone in which nonprofit organizations seek first and foremost to maximize theirability to provide a charitable good or service, while avoiding production or dis-tribution of ancillary activities that can hamper achieving the primary mission.The ability to produce the preferred output, however, depends on the revenuenot only from donations but also from user fees and - what is central in thischapter - from the production and sale of ancillary commercial outputs.

Nonprofit managers may view these ancillary outputs as a mixed blessing.Although profits from such activities provide needed income to supplement do-nations and allow increased provision of the primary output, the leadership ofthe nonprofit may dislike producing secondary output in principle, especially ifthat seems inconsistent with the organization's underlying objectives.

When nonprofits enter commercial markets to sell ancillary outputs, theyface different tax rules than do for-profit producers of the same outputs. Thischapter describes the nature of these tax differentials as well as implications fornonprofit organization behavior. We then consider the major effects of choicesmade by nonprofit organizations to engage in commercial activities, determin-ing whether these choices are intentional, efficient, and equitable, and consider-ing whether observed behavior of nonprofits in commercial markets is consis-tent with our predictions.

Differential tax treatment of nonprofit organizations

Income earned by for-profit enterprises is subject to tax in various forms. At thefederal level, and in most states, businesses pay profits taxes on their income;and at the state and local level, businesses are taxed on property values. Non-profit organizations, however, are subject to different tax rules that provide verybroad, if not complete, exemptions from these taxes. Nonprofit charitable orga-nizations are generally exempt from paying state and local taxes on propertyused "primarily" for the organization's exempt purpose; and nonprofits are ex-empt from federal and state corporate income taxes on income from activitiesthat, even if commercial, are deemed to be substantially related to the organiza-tion's primary exempt purpose.

In principle, neither the state nor federal tax exemptions are intended to ap-ply to taxes that would otherwise be incurred in the pursuit of commercial ac-tivities that are not related to the nonprofit's primary exempt purpose. Thus,since the 1950s, nonprofits have been subject to a federal Unrelated Business

Differential taxation, commercialization of nonprofit revenues 85

Income Tax (UBIT), which is meant to tax nonprofits on the same basis as for-profit corporations on income earned in a "trade or business" that is "regularlycarried on" by a nonprofit and not "substantially related" to the nonprofit's ex-empt purposes (Simon 1987; Hansmann 1989). States that levy corporate in-come taxes have followed the federal government's lead in this area by im-posing their own unrelated business income taxes; and most states require thatproperty be used primarily for an exempt purpose to qualify for property-taxexemptions.

The attempt by federal and state governments to tax unrelated commercialactivities is intended to limit the extent to which nonprofits can exploit their tax-exempt status to produce commercial outputs having little or no connection tothe organizations' primary "tax-exempt" output (Steinberg 1991). In practice,however, it has proved administratively difficult for federal, state, and local tax-ing authorities to differentiate taxable and nontaxable commercial activities(Simon 1987; Mikesell in press). Thus, tax experts widely agree that for manynonprofits, most, if not all, income earned from commercial undertakings is ef-fectively tax exempt. Moreover, opportunities to shift reported revenues andcosts between taxed and untaxed activities can allow nonprofits to avoid tax-ation even when an activity is clearly unrelated to its tax-exempt purpose.

Differential taxation and nonprofit behavior

How do these tax differentials affect choices that nonprofits make about boththe degree and the forms of their commercial ventures? Answering this ques-tion first requires discussing the kinds of investment decisions that nonprofitorganizations make, and how these may differ from those made by for-profitenterprises.

Modeling the investment choices of nonprofit organizations

We represent the nonprofit organization as an entity that seeks to maximize theutility from providing a preferred, "core," or mission-related nonprofit good orservice, subject to the constraint that spending on that good cannot exceed rev-enue (James 1983; Schiff and Weisbrod 1991). Revenue depends on donations,user fees (discussed in Chapter 4), and profit from commercial activities, whichis the focus of this chapter.

As in any organization deciding whether to invest in a business venture, thereis an opportunity cost to a nonprofit engaging in commercial activities. As Rose-Ackerman (1982) noted, although nonprofits can finance investments in com-mercial activities from "free cash" or accumulated reserves, the use of thesefunds is not free; every dollar invested in the commercial activity could insteadbe used to earn income from passive investments in securities. The nonprofit


could also borrow some or all of the funds needed to finance the commercialactivity, but at a cost. The key question is whether the nonprofit can realize agreater rate of return by investing in ancillary commercial activities rather thanin passive investments.

Following Schiff and Weisbrod (1991), we also assume that donations in-crease with the nonprofit's production of preferred, mission-related goods andwith its fund-raising expenditures as well. In addition, we assume that commer-cial activity may negatively affect donations if opposed by donors or positivelyif viewed as appropriate self-help.

Maximizing a nonprofit's utility subject to the financial constraint impliesthat nonprofits will invest in fund-raising up to the point where a marginal dol-lar spent on eliciting donations equals the cost of fund-raising. However, fund-raising in itself could also be seen as a nonpreferred, ancillary activity, in whichcase the organization would stop short of the amount of fund-raising that wouldmaximize revenue from this source (Weisbrod and Dominguez 1986).

A similar condition defines the required return, or hurdle rate, that an addi-tional dollar invested in a commercial activity must earn to justify the invest-ment. For nonprofits this hurdle rate does not depend exclusively on the finan-cial cost of funds, as would be the case for a for-profit enterprise, because theorganization or its potential donors might have a distaste for engaging in ancil-lary commercial activity.

When such distaste is present, the nonprofit will invest in a commercial ac-tivity to the point where marginal profit on an additional dollar spent on thecommercial activity equals the sum of the financial cost of funds,plus an addi-tional premium to compensate for the loss in utility and/or donor contributionsresulting from increased investment in the commercial activity. This hurdle rateis the starting point for analyzing what motivates nonprofits to invest their re-sources in commercial activities.

Nonprofit and for-profit behavior in the absence of taxation

We first consider briefly how nonprofit and for-profit organizations would beexpected to behave if all returns from commercial activities were untaxed. Thissetting is clearly unrealistic, but it provides a useful baseline for understandingthe effects of differential taxation.

In such a world, a typical for-profit organization would invest in a commer-cial activity until the marginal profit per dollar invested in the activity equaledthe opportunity cost of funds. Whether a nonprofit organization would use thesame investment criterion depends, as noted, on whether it or its donors had adistaste for engaging in commercial ventures. Were there no aversion to under-taking commercial revenue-generating activities, a nonprofit would act as aprofit maximizer in commercial markets regardless of its goals; that is, it would


behave just like a for-profit enterprise with regard to the decision about wheth-er and how much to invest in a commercial activity. Nonprofits would wantto maximize net revenue from commercial activities to maximize funding ofmission-related output.

Thus, if there were no aversion by the nonprofit or donors to commercial ac-tivities, nonprofits would behave in the same way as for-profits in the commer-cial market even though the pursuit of profit was not a goal for the nonprofit asa whole. Both types of enterprise would use the same criterion - profitability -to evaluate whether or not to invest in a given commercial activity. If nonprofitsreceived no special subsidies, they would compete on an even footing in capi-tal markets with for-profits, who could not then complain about being undercutby "unfair competition."

However, if either the nonprofit or its donors had a distaste for engaging incommercial activities, the financial test that a nonprofit would apply before en-gaging in a commercial activity would actually be relatively more stringent. Itwould not be enough for a commercial activity to earn a profit commensuratewith the financial opportunity cost of funds. The marginal profit would have toexceed the financial cost of funds by enough to compensate for the disutility ofthe activity and net loss of contributions from donors; thus, a nonprofit wouldengage in less commercial activity than would be profit maximizing.

Aversion to commercial activities could be strong enough to cause the non-profit to avoid certain ancillary commercial ventures entirely. If economic re-turn to a commercial activity were set by competition among for-profit pro-ducers, the marginal profit that could be earned by a nonprofit could not exceedthe financial opportunity cost of capital. In such a case, the nonprofit could notearn a marginal profit high enough to cover both the cost of funds and the "aver-sion premium" required for engaging in the nonpreferred activity.

Even so, nonprofits might still choose to engage in some commercial activ-ities if there were cost complementarities between these activities and the pre-ferred, mission-related output and if the nonprofit benefited from donations orsubsidies. This could allow the nonprofits to earn a greater marginal return onthe commercial activity than could for-profit competitors lacking such benefits.

Thus, in a world in which both nonprofits and for-profits were untaxed, thepropensity of nonprofits to seek out commercial sources of revenue would de-pend on whether the nonprofit and/or its donors saw pursuing profit as a meansto finance - that is, cross-subsidize - its mission. Commercial activities seen asconflicting with the mission, while regarded as economically profitable by for-profits, would not be worth undertaking by nonprofits. Such aversion could beovercome, however, if nonprofit production of preferred output brought suffi-cient cost advantages in ancillary markets. In the more general case, in whichindividual nonprofits have varying degrees of aversion for commercial venturesand differing abilities to exploit cost complementarities, we would expect en-


try into commercial markets to be concentrated in activities where aversion isleast - perhaps because the activity is "closest" to the nonprofit's mission - andwhere there was the most potential for exploiting cost complementarities be-tween commercial outputs and their preferred output.

Effects of differential taxation of nonprofit and for-profitenterprises

A central insight of our model is that when a nonprofit is averse to pursuingprofit in commercial markets, it will engage in commercial activities only if itexpects to earn a premium above the competitive rate of return that it could earnfrom passive investments. This implies that differential taxation of non- and for-profit enterprises, by itself, would not provide a financial incentive for non-profits to pursue otherwise objectionable commercial ventures.

The importance of tax-induced investment premiums

Consider a case in which nonprofits have no special tax incentive to engage incommercial ventures, despite exemption from taxes on ancillary business prof-its. Such a situation would arise if for-profit enterprises were subject to a neu-tral tax on business income, one that taxed income from all sources at the samerate and allowed interest on debt to be deducted (Rose-Ackerman 1982). Al-though one might think that the exempt nonprofit would have an incentive toseek out tax-exempt commercial activities under these circumstances, nonprof-its would earn no premium above the competitive return by doing so. If profitsfrom all business activities were taxed at the same rate, the marginal profit thata nonprofit could earn from investing in a commercial venture would equal theopportunity cost of funds, as it would in the no-tax case. Moreover, as in theno-tax case, if either the nonprofit or its donors had a distaste for engaging incommercial activities, the nonprofit would prefer investing passively if all itcould do is earn the same rate of return as would a for-profit competitor. Exemp-tion from tax would allow a nonprofit to earn a higher return from commercialactivity; however, it would also raise the financial (opportunity) cost of engag-ing in such a venture because the financial returns from passive investmentswould also be tax-exempt, whereas interest on debt would not be tax-deductible.Under a neutral income tax, these two features of the tax exemption would ex-actly balance each other so that the nonprofit and for-profit would face the sameincentives at the margin for investing in a commercial enterprise.

Effects of property and corporate tax exemptions

There are, however, several important instances in which differential taxationof nonprofits potentially allows nonprofits to earn premium returns from invest-


ing in commercial ventures that exceed the competitive return. One is the ex-emption from local property taxes for nonprofits. If for-profit enterprises mustpay these taxes, a for-profit would invest in a commercial activity up to the pointwhere the marginal profit per dollar invested was enough to cover the oppor-tunity cost of funds plus the property tax per dollar invested. Since nonprofitsdo not pay property taxes, in effect they would earn the same pretax total re-turn as a for-profit, without having to pay the property tax.1

Exemption from corporate profit taxes on commercial activities deemed tobe "substantially related" to the nonprofit's primary purpose creates another po-tentially exploitable tax premium (Cordes and Sheffrin 1981; Rose-Ackerman1982). This comes about because, unlike in the example discussed above, busi-ness income is not taxed at the same rate regardless of source; rather, corporateincome is taxed once at the corporate level and then again at the individual lev-el, whereas income earned by unincorporated enterprises is taxed only once atthe individual level. To compensate for the double taxation of corporate income,the pretax return to an additional dollar invested in a commercial activity under-taken by a corporation will come to exceed the interest rate. For-profit unincor-porated firms, however, would continue to invest up to the point where the mar-ginal return equaled the opportunity cost of funds.

This model implies that nonprofits have no particular tax incentive to in-vest in tax-exempt activities dominated by unincorporated enterprises, becausecompetition in capital markets will cause these activities to earn the same pre-tax return as could be earned by investing passively in bonds. (As Hansmann[1989] has noted, with a corporate tax, nonprofits would have a tax disincen-tive to hold corporate common stock, because as shareholders they could notavoid paying the corporate tax, collected directly from the corporation). Therewould, however, be a tax incentive for nonprofits to engage actively in tax-exempt related commercial activities dominated by corporate producers, andcompete with for-profits in those markets because they could earn a return ex-ceeding the competitive return from passive investments.

Tax treatment of unrelated business income

Not all business income earned by nonprofits, however, is tax exempt. As not-ed above, profits earned from activities deemed unrelated to the organization'stax-exempt purpose are subject to both federal and state Unrelated Business In-come Taxes (UBIT); and most states tax any property that is not used primar-ily for the organization's primary exempt purpose.

In principle, these taxes would appear to discourage nonprofits from invest-ing in unrelated commercial activities. If a nonprofit is already averse to earn-

1 In practice, some nonprofits may pay something like a partial property tax to the extent that theymake "voluntary" payments in lieu of tax (PILOTs).


ing income from such ventures, an investment subject to federal and state tax-ation would have to earn an even higher pretax return to make it worthwhile.

In practice, however, many nonprofits may be able to arrange their financialaffairs so that income that is nominally subject to federal and state UBITs maybe subject to little or no tax. For example, a nonprofit could deduct a portionof "joint" costs - costs of inputs that contribute to both its mission-related andtax-exempt commercial outputs - against profits from its unrelated, taxed activ-ity. Although such cost shifting would not by itself allow a nonprofit to escapepaying taxes on property used solely to produce unrelated business income,property taxes could be avoided to the extent that the property used to generatesuch income was also "primarily" used for mission-related, exempt purposes.

The presence of joint costs shared by related and unrelated activities not on-ly permits but encourages nonprofits to shift costs. Indeed, there is striking evi-dence that nonprofits take advantage of such opportunities. Of the 32,690 non-profits that filed UBIT returns in 1991, almost three out of five reported losses- negative taxable income - and for the UBIT filers as a group, total expensesallocated to providing unrelated activities exceeded total revenues, implyingthat, overall, unrelated activities are unprofitable (Riley 1995). Since nonprof-its presumably engage in unrelated activities only to fund their tax-exempt ac-tivities, truly unprofitable unrelated services make no sense. In reality, cost shift-ing allows reported profit to be zero or even negative, even though there maybe a substantial marginal profit from the unrelated activity that generates consid-erable revenue for the preferred, tax-exempt outputs.

Empirical evidence

Our framework predicts that the relative importance of commercial activitiesto nonprofits reflects three broad factors:

1 the abilities of particular nonprofits to exploit cost complementarities;2 the nonprofit's aversion premium, which in turn depends on the pref-

erences of both nonprofits and their donors; and3 the size of the pretax investment premium that can be earned on com-

mercial ventures.

The model also predicts, moreover, that, despite the disadvantages of earn-ing income from unrelated, and hence nominally taxable, commercial activities,nonprofits will undertake such business activities given a certain circumstance:The activities must allow the nonprofit to earn a pretax investment premiumover and above the opportunity cost of funds. A major contributing factor is theopportunity to use the same inputs in both mission-related and the unrelatedcommercial market; this provides the opportunity for the nonprofit to shift costsand possibly revenue between taxed and untaxed goods to minimize taxes onthe unrelated activity.


We examined some of these predictions with data from the Statistics of In-come (SOI) public-use sample of IRS Form 990 tax returns (see Appendix) filedin 1992 by nonprofit charitable organizations exempt from taxes under Section501(c)(3) of the Internal Revenue Code.

Measures of commercial activity

Determining whether variables suggested by the theoretical model explain vari-ation in the propensity of nonprofit organizations to engage in commercial ven-tures requires a means of measuring an organization's propensity to engage incommercial activities. One such measure is Commercial Share, defined as theshare of the nonprofit's total revenue derived from sources that can reasonablybe attributed to commercial activities that are exempt from taxation. We mea-sure Commercial Share as the share of total revenue derived from program ser-vices, which is widely used to gauge how much nonprofit organizations dependon revenue sources other than contributions, plus other income from excludedactivities and special events. To focus on commercial revenue sources likely tobenefit from nonprofit tax exemption, Commercial Share excludes componentsof program service revenue (PSR) reported as unrelated business income. An-other, dichotomous measure is whether the nonprofit had at least $ 1,000 in grossrevenue from the unrelated business activity, and so was required to file a UBITreturn.

Although PSR, which is an important component of Commercial Share, isa commonly used indicator of the extent to which nonprofits engage in business-like activities, it is not ideal because it includes income from activities such astuition payments by students, which may be neither business activities in thesense discussed above, nor viewed as nonpreferred sources of revenue. None-theless, because Commercial Share includes income from a wide range of com-mercial activities, variation in this variable is likely to reflect variation amongnonprofits in the propensity to rely on commercial sources of revenue.2

Measures of cost shifting

We also seek to examine whether nonprofits attempt to shift costs to reduce tax-able income subject to the UBIT. Froelich and Knoepfle (1996) note that non-profits enjoy fairly wide latitude in allocating wages and salaries that are partof management and general "overhead" among their various functional activ-ities. One option open to nonprofits reporting revenue from sales of goods is toinclude a portion of wages and salaries that would otherwise be included under

2 In principle, it would be desirable to distinguish between elements of program services that arebusinesslike and those that are not, but this is not possible in the public-use sample. The resultsreported in Table 5.1 were not sensitive to the use of alternative definitions of Commercial Share,some of which included unrelated business income.


"managerial and general expenses" as part of the cost of goods sold. As Froe-lich and Knoepfle (1996) point out, nonprofits may choose this strategy for avariety of reasons, not all of which are motivated by taxes; for example, non-profits may seek to minimize their reported management expenses to appearmore efficient in using charitable donations. Still, a nonprofit subject to theUBIT has an additional incentive to shift wage and salary expenses into the costof goods sold to reduce taxable income. As a by-product of reporting these ex-penses as costs of goods sold rather than as management costs, the nonprofitwould also report lower amounts of managerial and general salary expense.Therefore, we use as a measure of cost shifting the variable Management/Gen-eral Salaries, which is the ratio of wages and salaries reported as part of "man-agement and general expenses" to total expenses reported on the Form 990 re-turn. We hypothesized that when a nonprofit is subject to the UBIT, it will bemore likely to shift a portion of these joint costs to the Form 990 category of"cost of goods sold," and hence report lower amounts of total management andgeneral wages and salaries as part of its functional expenses. UBIT filers wouldthus be expected to have lower ratios of management/general wage and salaryexpense to total expenses than would nonprofits with otherwise identical at-tributes who were not engaged in commercial activity subject to the UBIT.

Measures of cost complementarities and aversion

It is difficult to construct suitable measures for an organization's aversion to un-dertaking commercial activities and for its ability to exploit cost complemen-tarities, but we can use two reasonable, if crude, proxies.

Entity size. It is likely that both opportunities for exploiting cost complemen-tarities and the degree of aversion to pursuing commercial ventures will varywith the size of the nonprofit. Larger nonprofits may well have more oppor-tunities than smaller ones to exploit cost complementaries. To the extent thatsmaller nonprofits are also likely to be relatively new, prospective donors mightalso be more likely to monitor the extent of their commercial activities. If so,aversion to commercial ventures would fall with an increase in the size of thenonprofit.

The implication of these two hypotheses is that the propensity to engage incommercial activities should increase with the nonprofit's size, as measured bythe variable Assets, the end-of-year assets reported by the nonprofit. This pre-diction is consistent with tabulations of aggregate data reported on Form 990returns, which indicate that larger nonprofits receive a larger share of their rev-enue from commercial sources (Hilgert and Arnsberger 1992). These tabula-tions, however, are presented for nonprofits as a whole, and thus do not controlfor the nature of the nonprofit's primary output, as reflected by its grouping in


the National Taxonomy of Exempt Enterprises (NTEE), which one might alsoexpect to affect the likelihood of engaging in commercial activities.

Source of donations. It also seems reasonable to assume that the aversion pre-mium could vary with type of donor - for example, private versus government.Although the model does not make specific predictions about which type ismore likely to oppose commercial undertakings by nonprofits, we construct thevariable Pet. Public Contributions, which is the share of total contributions re-ceived from government sources, to examine the possibly differential effect ofgovernment and private "contributions, gifts, and grants."

Type of activity. Both a nonprofit's aversion premium and its ability to exploitcost complementarities may also vary with the nature of its primary output. Uni-versities or hospitals, for example, may have greater opportunities to developcost complementarities than, say, human-service providers because of the vari-ety of their resource inputs. Thus we include a series of dummy variables thatcorrespond to the major NTEE categories of arts, education, health, and humanservices, which have sufficient sample size to generate meaningful results (seealso Chapter 6). The comparison group in the regressions comprises nonprof-its in various other NTEE categories (environment and animals, international,public, societal benefit, religious, mutual membership, and unknown).

Measures of tax premiums

Last, the model predicts that nonprofits should be more likely to invest in com-mercial activities that offer the prospect of capturing tax premiums exceedingthe competitive return. Although the tax premiums generated by differential tax-ation at the federal level do not vary among nonprofits in a single year, localproperty tax and state corporate tax rates vary. In theory, there should thereforebe some variation by state in the pretax investment premiums that can be earnedfrom commercial ventures.

Hansmann (1987) and Gulley and Santerre (1993) use interstate variation inproperty and corporate tax rates to examine whether aggregate market share ofnonprofits in selected industries is higher in states with relatively high prop-erty and corporate tax rates - where the nonprofit's advantage is greatest. Theyfind that, as the theory predicts, the market share of nonprofit organizations in-creases with the size of the tax differential.

These studies focus on what might be described as the aggregate outcomeof many individual decisions made by nonprofits to engage in commercial activ-ity. We use data from the Form 990 returns to examine the behavior of individ-ual nonprofits. To capture such variation we have added two dummy state-taxvariables, using data published by the Tax Foundation (1993/4) and Minneso-


ta Taxpayers Association (1996), to the individual observations in the 1992public-use sample. These variables were constructed to distinguish states whererelatively high local property tax rates and corporate tax rates would, in theo-ry, allow nonprofits to earn relatively large tax-induced premiums. The variableProptax = 1 in states where local commercial property tax rates ranked amongthe top ten in the country; the variable Corptax = 1 if the nonprofit is located inone of the nine states with a statutory tax rate in excess of 10 percent.

Results

Table 5.1 presents estimates of a regression of the variable Commercial Shareon proxies for a nonprofit's aversion to commercial ventures and ability to ex-ploit cost complementarities, as well as state-level tax dummy variables. Table5.2 presents the probability (logit regression) that a nonprofit files a UBIT re-turn (Form 990-T) as a function of the same variables. Table 5.3 presents esti-mates of a regression of Management/General Salaries on several characteris-tics of the nonprofit, including whether it was required to file a UBIT tax return.

The results are generally consistent with expectations. Table 5.1 shows thatthe industry of a nonprofit's primary output significantly affects its propensityto rely on revenue from commercial sales. The significant, positive coefficientson the NTEE dummy variables show that nonprofits with mission-related out-puts in the areas of arts, education, health, and human services all derive a larg-er share of funds from tax-exempt commercial activities than do nonprofits inthe comparison group. Holding size constant, nonprofit providers of health ser-vices receive roughly $0.75 of every revenue dollar from such sources, reflect-ing the major effect of hospitals, followed by nonprofit providers of educationand human services, which derive roughly $0.50 of every dollar from com-mercial sources.

The results also show that after controlling for the nature of the nonprofit'smission-related outputs, size (Assets) of the nonprofit is associated positivelywith revenue share from commercial sources. The negative coefficient of As-sets Squared indicates that although the propensity to engage in commercialactivities increases with size, it does so at a decreasing rate. This suggests that,after some point, a nonprofit's ability to exploit cost complementarities encoun-ters diminishing returns.

When size is measured in dollars as a continuous variable, increases in as-set size have a quite small effect on share of commercial revenue. The estimat-ed coefficients on Assets and Assets Squared indicate that for a nonprofit with$20 million in assets, doubling its size would increase the share of revenue de-rived from commercial activities by roughly two-tenths of a percentage point.

This relatively small effect may, however, reflect the fact that almost threeout of every four charitable organizations in the SOI sample has a "large" as-set size of $10 million or more. To examine this issue further, we therefore con-


Table 5.1. Dependent variable: Commercial Share = programservice revenue as a fraction of total revenue

Variable

Intercept

NTEE classificationArts(# = 583)Education(#=1,960)Health(# = 3,516)Human services(#=1,689)

No. of observationsAdjusted R2

F-statistic

Coefficient

.184**(.010)

.062**(.017).357**

(.012).554**

(.011).289**

(.013)

9,104.259

353.98

Variable

Assets(units of $10 mil.)Assets Squared

Pet. PublicContributionsProptax

Corptax

Coefficient

.0094**(.0002)

-1.46E-6**(3.OE-8)-0.050**

(.011).033**

(.007).042**

(.011)

Note: Results are not sensitive to the use of alternative definitions of Commercial Share,some of which included unrelated business income. Standard errors are shown in paren-theses.** Statistically significant at the .01 level.

structed a dummy variable equal to 1 if the nonprofit had assets of $5 millionor more, and zero otherwise. Though the results are not shown here, when wereplaced the continuous measure of size in the regression for Commercial Sharewith this variable, the coefficients of the other variables remained of the samesign and magnitude as shown in Table 5.1; and smaller nonprofits are quanti-tatively as well as statistically much less likely to rely on commercial revenues.The coefficient of the size dummy implies that the commercial revenue shareamong "small" nonprofits - those with less than $5 million in assets - is almost17 percentage points lower than the corresponding commercial revenue shareamong larger nonprofits.

Table 5.2 shows broadly similar patterns and effects of both the nonprofit'sindustry and size on its propensity to undertake taxable commercial activities.The odds ratios calculated from the estimated coefficients in the logit regres-sion imply that nonprofit providers of health services are about 2.4 times morelikely than the reference group to report revenue from program services subjectto the UBIT. (The odds ratio is the ratio of the predicted probability that the or-ganization files a UBIT return to the predicted probability that it does not.) Non-


profit organizations in the arts and education are both about twice as likely asnonprofits in the reference group to receive such revenue. Nonprofit human-service providers, however, are about 50 percent less likely than those in thereference group to engage in unrelated business activities subject to the UBIT.

Assets also affect the probability of filing a UBIT return. The estimated co-efficients of Assets and Assets Squared indicate that doubling the size of a non-profit with $20 million in assets would increase its odds of filing a UBIT re-turn by roughly 12 percent. Replacing the continuous asset-size variables witha size dummy also results in a much larger estimated effect of asset size: Non-profits with assets of $5 million or more are roughly five times more likely tofile a UBIT return than are other nonprofits. Whether size is proxied by a contin-uous or a dichotomous variable, however, it should be noted that larger nonprof-its are also more likely to file UBIT returns simply because the scale of theirunrelated activities is more likely to trigger the UBIT filing threshold of $ 1,000in gross income.

Our results indicate that the source of donations affects a nonprofit's com-mercial activities - as measured by both commercial revenue share (Table 5.1)and filing a UBIT return (Table 5.2). The greater the share of contributions thata nonprofit receives from public rather than private sources, the less likely thenonprofit is to rely on revenue obtained from commercial activities (tax exemptand taxable). The estimated coefficients of Pet. Public Contributions in Tables5.1 and 5.2 imply, respectively, that a ten-percentage-point increase in the shareof contributions received from public sources reduces the revenue share derivedfrom program services by half a percentage point, and lowers the odds that thenonprofit files a UBIT return by about 20 percent. A possible explanation is thatpublic donors are more likely than private donors to be averse to commercialactivities of nonprofits.

Tables 5.1 and 5.2 also show that nonprofits are more likely to rely on com-mercial revenue sources when located in states with relatively high taxes. Theestimated coefficients of both the property- and the corporate-tax variables arepositive and statistically significant in the regression for Commercial Share, in-dicating that nonprofits in high-tax states are more likely to depend on commer-cial sales for revenue. The magnitudes of the coefficients imply that locationin one of the ten states with the highest commercial property tax rates raisesthe nonprofit's commercial revenue share by about 3 percentage points; and lo-cation in a high-corporate-tax state raises commercial share by 4 percentagepoints. These results are generally consistent with previous research showingthat cross-state variation in property and business taxation helps to explain therelative importance of nonprofits in certain industries.

Table 5.2 also indicates that nonprofits located in relatively high-corporate-tax states are also more likely to undertake unrelated - taxable - business activ-ities. The odds ratios calculated from the estimated coefficient imply that non-

Differential taxation, commercialization of nonprofit revenues

Table 5.2. Likelihood that a nonprofit files a UBIT return (logitregression)

97

Variable

Intercept

NTEE classificationArts(N = 583)Education(N= 1,960)Health(N =3,516)Human services(N = 1,689)

No. of observations% correct predictionsChi-square

Coefficient

-1.886**(.080)

.727**(.121).633**

(.093).857**

(.085)-0.621**

(.118)

9,10474.3

846.80

Variable

Assets(units of $10 mil.)Assets Squared

Pet. PublicContributionsProptax

Corptax

Coefficient

.056**(.003)

-8.90E-6**(5.81E-7)-0.233**

(.083)-.010(.056).287**

(.075)

Note: UBIT = 1 if nonprofit required to file a UBIT return, 0 otherwise. Standard errorsare shown in parentheses.** Statistically significant at the .01 level.

profits located in high-corporate-tax states are about one-third more likely tofile a UBIT return than are nonprofits in states with lower taxes.

These results would be unexpected if unrelated business activities in thesehigh-tax states were actually taxed. However, the ability of nonprofits essential-ly to avoid taxation of profit by shifting costs and revenue between taxable andnontaxable activities tends to make unrelated business activities relatively moreattractive financially. In these cases, we would expect variation in state-level taxrates to have the same effect on a nonprofit's decision to engage in an unrelat-ed business activity as in a tax-exempt activity.

In this vein, the results in Table 5.3 provide more direct statistical evidencethat nonprofits take advantage of opportunities for reducing their taxable in-come by shifting costs when they are subject to the UBIT. To test the hypoth-esis that UBIT filers have an incentive to shift wage and salary expenses intocost of goods sold, we restricted our attention in the SOI sample to nonprofitsreporting at least $1,000 in gross receipts from sales, thereby giving them theopportunity to shift management and general wage and salary expenses.


As predicted, we find that nonprofit organizations that filed a UBIT returnreported lower ratios of management and general wage and salary costs to to-tal expense on their Form 990 returns. The estimated coefficient indicates thatthe ratio of such costs to total expenses among UBIT filers is about 1 percent-age point lower than for organizations not subject to tax.

The magnitude of the coefficient may seem to indicate only a moderate de-gree of cost-shifting, but the amount of wage and salary expense that is esti-mated to be shifted must be compared with the amount of potentially taxableincome. Among firms required to file a UBIT return, the 1-percentage-point dif-ference in the ratio of "shiftable expense" to total expense amounts to roughly$650,000. This figure may be compared with the amount of "gross profit fromsales" (gross receipts minus cost of goods sold) reported by UBIT filers - ap-proximately $2.2 million. In other words, the amount of expenses estimated tobe shifted from overhead wages and salaries is almost a third of "gross profit."

In addition to the dummy variable UBIT, we also included the variable Con-tributions, which is the ratio of revenue received from private direct and indi-rect contributions, plus special fund-raising events, to total revenue. This vari-able is intended to control for the degree to which the nonprofit cares about itsreported management expense for reasons other than taxes (e.g., any adverseeffect on donations). The implication is that organizations that depend more oncontributions would be more apt to care about the reported amount of manage-ment expense, and hence have an incentive to shift management/general wageand salary expense into cost of goods sold, whether or not they were subject tothe UBIT. Although tangential to our analysis, the coefficient of this variableis significant and of the expected negative sign, indicating that nonprofits thatdepend more on contributions are also more likely to report lower ratios of man-agement/general wage and salary expense to total expense.

To see whether the above results reflect cost shifting, we also estimated thesame equation as that presented in Table 5.3 for the subsample of nonprofits inthe SOI sample with little or no opportunity to shift expenses from elsewhereon the Form 990 return into the cost of goods sold because they had less than$1,000 in gross sales. Although the results are not reported here, neither theUBIT nor the Contributions variable was statistically significant in these regres-sions. This is what one would expect in the case where shifting wages and sal-aries out of management and general expenses into cost of goods sold was notan option, providing some further evidence that the estimated coefficients inTable 5.3 are indicative of cost shifting that is motivated by both tax and othermotives.

These results also suggest that increased commercial activity by nonprofitsmay indirectly affect public perceptions of the "effectiveness" of different orga-nizations. Nonprofits that can deduct cost of goods sold have opportunities toreport expenses on the Form 990 return in ways that make financial ratios of-


Table 5.3. Dependent variable: Ratio of management and generalwage and salary expenses to total expenses

Variable

Intercept

NTEE classificationArts(# = 280)Education(N = 432)Health(#=316)Human services(#=252)

No. of observationsAdjusted R2

F-statistic

Coefficient

.098**(.006)

.017**(.007)

-0.016*(.006)

-0.018*(.007)

-0.016*(.007)

1,476.045

9.59

Variable

Assets(units of $10 mil.)

Assets Squared

UBIT

Contributions

Coefficient

-0.0006**(.0000)

1.3E-6**(5.7E-7)

-0.0096*(.004)-.0205**(.006)

Note: Standard errors in parentheses.* Statistically significant at the .05 level.** Statistically significant at the .01 level.

ten used to gauge an organization's "fund-raising effectiveness" (e.g., the ratioof management and general expenses, plus fund-raising costs, to contributions)look more favorable.

In addition to examining the shifting of costs from managerial expense tocost of goods sold, we also investigated the extent to which nonprofits' unrelat-ed business activities are taxed very little because joint costs are allocated tothose activities on the UBIT return - IRS Form 990-T - thereby reducing oreven eliminating reported profit. When resource inputs are joint, determined es-sentially by the level of mission-related activity, the true marginal costs of en-gaging in unrelated business activities will be small, or even zero. How muchof the costs reported by nonprofits as attributable to taxable, unrelated businessactivity on the Form 990-T returns is simply a reallocation of joint costs? Is ittrue that commercial activity in unrelated business is as unprofitable as mostUBIT returns indicate?

To answer these questions we sought to determine the extent to which costsclaimed on UBIT returns are truly attributable to the unrelated activities, and towhat extent they would been incurred even if those activities were not under-taken. Specifically, we regressed a nonprofit's "total compensation" costs on the


magnitude of its activities in related and in unrelated markets, using an organi-zation fixed-effects estimator. The predictions were that

1 additional related activity, as measured by the gross revenue it gener-ates, would be associated with greater labor-compensation costs, but

2 additional unrelated activity, also measured by its gross revenue,would be associated with less, even no, additional compensation costs.

The different expectations for the cost effects of related and unrelated, untaxedand taxed, activities hinge on the joint-cost hypothesis.

Data are from the SOI samples of Form 990 returns for the four years 1990-3. We analyze separately the regression equations for each of three industrygroups. The number of organizations varied somewhat from year to year, butfor the four years included 6,662 observations for hospitals, 2,014 for univer-sities, and 572 for museums.

The findings are striking. In each of the industries we found, as hypothesized,that (1) additional gross income from untaxed activity had a positive, statis-tically significant, and substantial effect on total compensation, and (2) addi-tional income from taxed (UBIT) activity had no significant effect on totalcompensation (Table 5.4).3 It seems, not surprisingly, that nonprofits are select-ing unrelated commercial activities that use the same inputs required for themission-related activities, so that the marginal costs of added unrelated activ-ity are essentially zero.

How do these findings compare with the allocation of labor-compensationcosts to the unrelated, taxed activities on actual UBIT returns? We were ableto examine these returns for a small subsample of nonprofits to determine therelationship between gross revenue and the compensation costs that were de-ducted as attributable to the taxed activities. The picture is very different. Table5.5 shows regression estimates for the effect of additional gross unrelated busi-ness income on labor-compensation expense reported on the Form 990-T re-turn. Data were available only for 1990. An incremental dollar of unrelatedbusiness gross income was associated with an additional, and highly significant,twenty-two cents of reported compensation costs in hospitals, fourteen cents inhigher education, and five cents in museums. For this subsample, we also rep-licated the regression equations in Table 5.4, and with the same findings (notshown, but available from the authors): Increased income from untaxed activ-ities was associated with a positive and significant effect on total compensationcosts, as shown on the Form 990, but increased gross income from taxed, un-related activities was associated with no statistically significant increase in com-pensation costs. In short, increased unrelated business income appears to lead

3 Untaxed activities are of two types: related and excluded. The latter are not taxed even thoughthey are not related to the nonprofit's mission. Exclusions are given when activities are inciden-tal or occasional, or are undertaken "for the convenience of" persons such as hospital patientsand museum visitors. Our findings are affected little by omitting excluded income.

Differential taxation, commercialization of nonprofit revenues

Table 5.4. Dependent variable: Total labor-compensation costs, 1990-3(Form 990 returns)

101

Variable

Gross UBI

Related and excluded income

/?2

Hospitals(6,662)«

0.068(.057)0.100**(.0026)0.985

Higher education(3,260)*

0.048(.046)0.062**(.007)0.988

Museums(572)*

0.538(.901)0.270**(.058)0.971

Note: Standard errors are given in parentheses.^Number of data points over the four years, 1990-3.* Significant at .05 level.** Significant at .01 levelSource: Our computations from data in IRS-SOI data tapes for 501(c)(3) nonprofits, 1990-3.

Table 5.5. Dependent variable: Total labor-compensation costs, 1990(Form 990-T returns)

Variable Coefficients

Gross UBI

Constant

* 2

Hospitals (TV = 24)

0.224**(.037)

$42,743*(20,016)

0.736

Universities (N= 12)

0.143**(.011)

$35,517(20,177)

0.702

Museums (N = 7)

0.046*(.017)

$12,765(11,541)

0.178

Note: Standard errors are given in parentheses.* Significant at the .05 level.** Significant at the .01 level.Source: Random sample of nonprofit organizations filing IRS Form 990-T (UBIT) tax returns drawnfrom the larger SOI annual sample of nonprofit organizations.

to no increase in actual compensation costs, but does lead to an increase in thecompensation costs that are allocated to taxable income.

Commercialism in the form of unrelated business activity is growing . Thenumber of filers of UBIT returns has doubled between 1985 and 1995, from24,103 to 50,034 (Herman 1997). Between 1990 and 1993, the latest year forwhich data are available, gross activity (unrelated business income) increasedby nearly 35 percent, from $3.5 billion to $4.7 billion (in nominal dollars). At


the same time that gross income was soaring, opportunities to avoid taxationby allocating joint costs appear to have been broadly recognized, and so muchmore in expenses were deducted that taxable income actually fell. Aggregatenet taxable unrelated business income, -$631 million in 1990, dropped to -$1.2billion in 1993 (Riley 1995, 1997).

If further analysis sustains our findings, the consequences are dramatic.Most, if not all, of the labor-compensation costs reported on UBIT returns re-flect not true costs of the activities but only allocation of fixed costs that wouldhave been incurred even were there less or no unrelated business activity. Thisdoes not imply any illegalities, but it does raise questions about accounting rulesfor joint-cost allocations between taxed and untaxed activities.

Policy implications

The theory and the evidence presented above suggest that differential taxationof nonprofits and for-profits creates opportunities for nonprofits to earn above-normal returns on commercial ventures, thus helping them overcome a "natur-al" aversion to pursuing profits in ancillary markets. As Rose-Ackerman (1982)first observed in the context of a model assuming no such aversion, this meansthat changes in the taxation of for-profit firms, while not specifically applyingto nonprofits, can nonetheless raise or lower the tax premium that nonprofitscan earn from commercial activity in competition with for-profits. Indeed, eventhough the taxation of nonprofits has remained essentially unchanged for manyyears, changes during the 1980s that occurred in the tax treatment of their for-profit corporate competitors first decreased and then increased the tax premi-ums that nonprofits could potentially earn from commercial ventures.

In the early 1980s, changes in business taxation set in motion by the Eco-nomic Recovery Tax Act of 1981 (ERTA) significantly reduced the effective taxrate on corporate profits. The effect of this change was to reduce the gap be-tween the return that nonprofits could obtain on commercial activity in marketsdominated by private corporate producers and the return that nonprofits couldearn on passive investments. For the five years in which ERTA was in effect,differential taxation of nonprofits should, in theory, have reduced the relativefinancial incentives to engage in active commercial activity .

The Tax Reform Act of 1986 (or TRA 1986), however, raised the effectivefederal tax rate on corporate profits. This should have increased the return avail-able to nonprofits on commercial activity in markets dominated by corporateproducers, thereby creating stronger tax incentives for nonprofits to invest incommercial activities, particularly those "related" to their exempt mission.4

4 The same point can be stated differently. One can imagine three organizational forms that com-pete with each other: for-profit corporations, for-profit unincorporated businesses, and non-prof-it organizations. If the tax system works to the disadvantage of for-profit corporations, it will


The effect of TRA 1986 on the incentive to engage in unrelated business ac-tivity, however, is ambiguous. On one hand, because the UBIT parallels the cor-porate income tax, increasing the effective tax rate on corporate income wouldhave reduced the incentive of nonprofits to engage in unrelated business activ-ities if the income from such activities was actually taxed. On the other hand,raising the effective corporate tax rate could have the opposite effect of encour-aging nonprofits to engage in unrelated business activities if the income fromsuch activities was effectively tax exempt - because of either lax enforcementor cost shifting. This latter possibility seems to have been plausible, in light ofboth the small amount of aggregate UBIT taxable income reported by nonprofitsin 1991 and our empirical analysis of the effects of UBIT filing status on report-ed expense ratios of nonprofits.

Whatever the changes in incentives, however, competition between tax-exempt nonprofits and private firms poses numerous issues that were discussedin Chapter 1 - in particular, whether nonprofit commercial activities, whatevertheir goal, are efficient uses of resources and fair to the taxed organizations. Thissubject has received considerable attention, beyond the scope of this chapter(Rose-Ackerman 1982; Weisbrod 1988;Hansmann 1989; Steinberg 1991). Thepreceding discussion, however, highlights two points that have not been em-phasized in this literature: The first is that differential taxation of nonprofits islikely to provide the greatest financial incentive to pursue profit in activities thatare economically related to nonprofits' primary mission-related outputs. Thesecond is that such increased competition with for-profits may be economicallyefficient. It may encourage nonprofits to use existing resources to expand theoutput of ancillary goods that, without differential taxation, would be profitableto produce, but which would not be produced because of nonprofits' aversionto profit-making activities.

Conclusion

A central theme of our analysis is that the effects of incentives operating throughthe tax system on nonprofits' commercial activities are best understood in aframework that explicitly accounts for the interaction between differential tax-ation and the following: nonprofit executives who may be averse to commer-cial activity, donors whose giving may be sensitive to such activity by the non-profits, and cost complementarities between nonprofit core activities and the

presumptively work to the advantage of the other two organizational forms. The boundaries oftax-induced competition between these organizational forms will shift in response to changesin tax policy. ERTA shifted these boundaries in favor of for-profit corporations, and by impli-cation against unincorporated businesses and nonprofits, while subsequent tax changes had theopposite effect. We are grateful to David Bradford for suggesting this point.


secondary money-raising commercial efforts. Within this framework, differen-tial taxation encourages nonprofits to pursue commercial ventures they wouldotherwise avoid, by providing excess financial returns that nonprofits can ex-ploit because of their tax-exempt status. The opportunities to earn such excessreturns, however, are likely to be greatest in markets for ancillary goods thatare related via cost and revenue complementarities to the organization's prima-ry mission-related outputs. This limits and channels the extent to which differ-ential taxation encourages greater commercialism among nonprofits.

The framework also yields several predictions, which we have examinedempirically. As expected, we find that the propensity of nonprofit organizationsto undertake both tax-exempt and taxable commercial activities depends on thenature of their primary mission-related output and size (as proxied by NTEEindustry category and assets); on the relative importance of public versus pri-vate contributions; and on the size of the excess return created by differentialtaxation of nonprofit and for-profit business. We also find that organizationsthat engage in taxable commercial activities are more likely to apportion jointcosts so as to shift such costs from tax-exempt to taxable sources of income.

CHAPTER 6

Interdependence of commercial anddonative revenues

Lewis M. Segal and Burton A. Weisbrod

Introduction and overview

As Chapter 3 outlines, nonprofit organizations' principal sources of finance aredonations and sales. That is, they can generate revenue through some form ofcontribution, or they may sell products and services. Proxy measures for theformer are what the Internal Revenue Service terms "contributions, gifts, andgrants," and for the latter, what it terms "program service revenues" (PSRs),which include user fees and revenue from ancillary activites. Since we utilizethese data extensively, it is important to be clear that PSRs encompass all sales-generating activities, whether central to the nonprofit's mission or not. More-over, "mission relatedness" of an activity is determined by the nonprofit andalso by the IRS - and these two perspectives may differ. While some activitiesthat generate program service revenues will be treated by the IRS as taxable,because they are unrelated to organization mission, and other activities as re-lated and, hence, not taxable, the nonprofit's own view may differ. It may judgethat an activity is not central to its mission even though the IRS considers theincome generated nontaxable; similarly, the nonprofit may view an activity ascentral to its mission even though the IRS determines it to be unrelated.

In the short run, nonprofits have additional sources of revenue - they mayreceive interest and dividends from investments, and can borrow and drawdown assets; but in the long run their financial health depends on their abilityto generate donations and to sell services profitably. This chapter asks: Are thesetwo revenue sources - one philanthropic, the other commercial - interrelated?More particularly, does a change in donative revenue - for example, a cut ingovernment grants - influence nonprofits' commercial activity, or vice versa?

We thank the Aspen Institute Nonprofit Sector Research Fund for contributing to support of the re-search reported here; Ian Domowitz, John Goddeeris, and Richard Steinberg for helpful commentson earlier drafts; and Cecilia Hilgert for assistance with interpreting IRS data. The views expressedhere are those of the authors and not necessarily those of the Federal Reserve Bank of Chicago.

105

106 Lewis M. Segal and Burton A. Weisbrod

We examine the hypotheses that nonprofits' commercial sales activities aremechanisms for financing their principal, tax-exempt mission, and that non-profits prefer to avoid such activities, engaging in them reluctantly and thenonly for their profitability. If sales activities are "nonpreferred," there are sev-eral testable implications: One, on which we focus, is that those activities areundertaken to a degree that varies inversely with the nonprofit's revenue fromthe preferred source, donations. If there is aversion to commercial activity, de-spite its potential financial contribution to the organization's mission, then do-nations would "crowd out" commercial activities, with decreased donationscausing expansion of commercial activity and increased donations diminish-ing it.

A second testable implication of an aversion to commercial activity is thatnonprofits engage in it less than would maximize profit and, hence, less thanwould maximize revenue to finance their preferred, mission-related activities.Testing this hypothesis is beyond the scope of this chapter, but our results doshed a bit of light on it. One piece of prior evidence is the finding that nonprof-its spend less on fund-raising than would maximize their net profitability (Weis-brod and Dominguez 1986; but see also Steinberg 1986). It seems plausible thatan aversion to engaging in fund-raising would reflect a larger predisposition toavoid activities that are, themselves, not part of the organization's mission, apartfrom their net revenue.

If there is aversion to commercial activity, revenue from donations would bepreferred to an equivalent sum from profitable sales activity. A plausible conjec-ture, therefore, is that an exogenous decrease in donations would be offset, butless than fully, by an increase in net, after-tax revenue from commercial activ-ity. To test the magnitude of such crowding out, however, it is necessary to esti-mate the profitability of commercial sales, the fraction of increased gross rev-enue from sales that can be used to cross-subsidize mission-related activities.Little is known about the profitability of various nonprofit-sector commercialactivities, and meaningful data are unavailable (although Chapter 5 makes a for-ay into that area).

Theoretical relationships between revenues from donationsand from program services

The relationship between revenue from donations and from commercial ac-tivity can take a variety of forms, as Chapter 3 indicates. If nonprofits are in-deed averse to commercial activity, donations will crowd out commercial activ-ity, and there will be a negative relationship between the two forms of revenue.There could, alternatively, be a positive relationship, "crowding in," if increasedcommercial activity is viewed favorably by donors and leads to increased do-nations. In addition to such interrelations between revenue sources, there could

Interdependence of commercial and donative revenues 107

be interrelations between costs: Producing commercial goods and raising dona-tions could be subject to economies of scope. Finally, the two forms of revenuecould be independent and unrelated, as would be the case if nonprofits acted asprofit maximizers in the commercial markets in order to obtain as much net rev-enue as possible for support of their missions, and if commercial and mission-related activities were separable in their costs and revenues.

Crowding out of commercial activity by donations is most likely when it in-volves sale of goods and services that are viewed by the nonprofit's manage-ment as most unrelated to the nonprofit's mission - activities engaged in onlyfor their financial contribution. It may also apply, however, to activities that arerelated to the mission. For example, as Chapter 3 shows, nonprofits frequently,though not always, can demand user fees of the persons they serve in pursuitof their mission. Colleges can charge students tuition, hospitals can charge pa-tients fees, and museums and zoos can charge visitors admission. They may pre-fer to avoid such user fees, however, lest they inadvertently discourage utili-zation by persons in the target population (Chapter 4), which would make thisrevenue-raising source nonpreferred. For-profit firms, by contrast, are presumedto be indifferent among alternative money-generating activities, caring onlyabout the net profitability. Thus, while a private firm would be expected to pur-sue commercial market opportunities at maximum profit levels, if nonprofitsprefer to avoid commercialism they will engage in it reluctantly, whether or notthe IRS treats it as taxable and only if the organization cannot obtain "sufficient"donations - in short, only if the net revenue it generates is great enough not onlyto be profitable (after any tax) but to overcome the disutility of engaging in it.

Contemporary concern about escalating college tuition illustrates the issuesand trade-offs. Rising college production costs, not accompanied by offsettinggovernment grants or private donations, have forced choice between retrench-ment, on one hand, and raising additional commercial revenue either from userfees or ancillary activities. All are occurring, but along with sharp rises in tui-tion - "tuition and fees doubling from 1976 to 1994" - has come increasingconcern about the effects, including "pricing as many as 6.7 million studentsout of higher education" (Applebome 1997, A17). In terms of the analyticframework we utilize, public and private donations and grants are the preferredsources of revenue, but their shortfall may have led to increased use of a non-preferred source, user fees. As Chapters 3 and 4 explain, the aversion to userfees as a source of revenue is caused by their negative effect on the target pop-ulation - in this case, students. This is despite the increased use of price discrim-ination, in the form of student financial aid, in attempts to minimize the adverseeffect of increased tuition on the target population of students.

As the tuition example makes clear, the hypothesis that nonprofits have pref-erences among alternative revenue sources is separate from legal definitions of"related" and "unrelated" business activities and their differential tax treatment.


For a nonprofit organization there are no tax consequences of obtaining revenuefrom user fees rather than from donations, although there is a difference to thepayers, since donations are tax deductible, at least for itemizers. The nonprof-it's preference for donations is the consequence of the hypothesized negativeeffect of user fees on the organization's mission, not on tax considerations. Amission that encompasses distributional goals, as discussed in Chapers 3 and 4,would cause a nonprofit not to be indifferent between a dollar of revenue fromdonations and a dollar from user fees.

So, too, a nonprofit may not be indifferent between raising funds throughuser fees and netting an equal amount through after-tax profit from provisionand sale of other, non-mission-related outputs - termed ancillary in Chapter 3.Whatever their legal form, and whether or not the IRS applies the UnrelatedBusiness Income Tax (UBIT) to them, these are activities that are seen by thenonprofit as elements outside of its mission, and in some cases incompatiblewith that mission. Aversion to ancillary activities may reflect a view that com-mercialism - acting in ways identified with private firms, and thinking aboutmarkets and profitability in their terms - detracts from the spirit and goals ap-propriate to nonprofits. To the extent this is the case, as noted above, there isno necessary identity between what a tax authority, the IRS, treats as unrelatedand what the organization leadership regards as not contributing to its mission.Thus, a nonprofit hospital that decides to open a major heart-transplantationunit, which it expects to generate substantial profit, knows that this will be re-garded by the IRS as a mission-related activity, and hence nontaxable, even ifthe hospital would not be establishing that unit were it not for its profitability.

An aversion to commercial activity has the same effect as a tax on it, increas-ing the effective marginal cost of production and thus decreasing the effectivemarginal profit from commercial activity. This leads to our central hypothesis:

Commercial activity in any form - whether involving user fees or expansion intoancillary markets - can be expected to be sensitive to the level of donations,changing inversely to it.

This chapter examines the microeconometric empirical evidence to deter-mine whether there is a negative causal relationship between donations andcommercial activity. We do not study separately revenue resulting from UBITactivities and from untaxed activities, which are aggregated in program services(but see Chapter 5, which does examine some relationships between the two).

During the 1980s the nonprofit sector, as a whole, became increasingly reli-ant on sales revenue relative to donative revenue. As Table 6.1 shows, "contri-butions, gifts and grants" (donations) to charitable (501(c)(3)) organizationsfrom both private and government sources declined as a share of total revenuebetween 1982 and 1993, while the share from program services (sales) grew.


Table 6.1. Revenues from donations'1 and from commercial activities forcharitable 50J(c)(3) nonprofits, 1982-93b (in billions of constant dollars)c

Taxyear

19821983198519861987198819891990199119921993

Donations(CGG)

$42.7946.5951.8354.8554.3058.3861.7064.8063.7171.8070.78

Totalrevenues

$203.42231.78249.43266.86273.56299.79320.73332.25359.54383.15390.63

Component as % of total revenue

TotalCGG

21.020.720.820.619.819.519.219.517.718.718.1

PrivateCGG

11.210.911.312.911.710.910.810.99.8

10.29.9

Govt.CGG

9.89.89.47.68.28.68.58.67.98.58.2

Program servicerevenue (PSR)

63.465.962.664.368.267.568.470.670.371.071.3

aReported to the IRS as "contributions, gifts, and grants" (CGG).^Excluding 1984, for which data are not available.CA11 dollar values are deflated by the Consumer Price Index for All Urban Consumers (1982-4 =100).Source: Entries for 1982 and 1983 are derived from Hodgkinson et al., Nonprofit Almanac 1992-1993. San Francisco: Jossey-Bass, 1992, tab. 4.5. Entries for 1985-93 are authors' calculationsfrom the IRS, Statistics of Income Division, data tapes.

Of course, revenue, whatever its source, is not equivalent to profit. The truemargin of expected net revenue from any source is an important determinant ofa nonprofit's choice among revenue sources, but evidence is difficult to obtain.What we expect, but cannot test, is that organization mission will cause differ-ential preferences among potential sources, so that it will not be the case thatmarginal profitability from all sources will be equated, let alone equated withthe opportunity cost of capital.

In our theoretic framework we emphasize this distinction between the non-profit's preferred and nonpreferred activities: The organization prefers to en-gage in its mission-related pursuits; it becomes involved in revenue raising, anonpreferred activity, only as necessary to achieve its mission. Some donationsmay come to the organization with little or no effort, but such exogenous in-come may be judged by the nonprofit to be suboptimal and lead it to devote re-sources to increasing contributions or to profit-making commercial activities.Fund-raising to garner additional contributions and program services to garnersales revenues are quite similar in their goal, which is largely to generate re-


sources that would permit expansion of the mission-related activities. This alsois the intention behind user fees, as discussed above.

Seldom is a source of revenue entirely exogenous, entirely independent oforganizational effort. We judge, however, that there is a greater measure of suchindependence with respect to contributions, gifts, and grants (CGG) than forcommercial sales revenues. Changes in federal government funding of grantsto hospitals, universities, and arts organizations, for example, can dramaticallychange donative revenue, as has been the case recently with sharply diminishedfederal grants to the National Endowment for the Arts ("Republican Effort toEliminate Arts Agency Survives Challenge" 1997) and to public television (seeChapter 13).

In a later section we turn to empirical tests to determine to what extent, ifany, nonprofit commercial activity in the form of program service revenue is aresponse to changes in exogenous revenue from donations. Difficulties emergeas we look for operational counterparts to the concepts of preferred and nonpre-ferred sources of revenue, and the differential tax treatment of various forms ofrevenue creates another force affecting organization choice among alternativesources of revenue.1

Available data from the IRS Form 990 return (Appendix) provide proxymeasures that map (imperfectly) to the aforementioned theoretic concepts: CGGto the preferred, unconstrained revenue sources, and PSR to the nonpreferredrevenue sources that involve activities, including commercial sales, sometimesseen by nonprofits as digressions from their missions.

Our concept of a nonprofit's commercial activity encompasses all endeav-ors that involve a transaction with a consumer paying for output received. Pay-ment, of course, is a matter of degree. A "pure" nonprofit organization is onethat finances production entirely through unrestricted donations generated cost-lessly and distributes its charitable output, at a zero price, to all "deserving" per-sons: Output is simply given away. This polar case is the ideal for two reasons:

1 Only with a zero price can the organization be reasonably assured thatno deserving person is being denied access to its service, althoughsome nonprice form of rationing might be required if the total quan-tity available is insufficient.

2 Only by remaining free of commercial activity can the nonprofit con-centrate its managerial resources on providing services rather than ondebates over the appropriateness of engaging in the sale of other goodsand services in order to raise money.

1 "Contributions, gifts, and grants," as reported on the IRS Form 990 information return, some-times constitute what we would term sales - e.g., for the naming of a building at a hospital, col-lege, or community center. At the same time, some revenue reported as sales - "program servicerevenues" on the Form 990 - has a substantial donative element, as when consumers pay morefor a service provided by a nonprofit than they would pay to a for-profit firm.


In the contrasting case, typical of a private firm, the organization receives nodonations, gives away nothing, and charges a profit-maximizing price for itsoutput. The distinction between a nonprofit's reliance on donations and on rev-enue from sales is central to this chapter's focus on whether donative revenuecrowds out program service revenue.

Actual nonprofit organizations are arrayed on a continuum in terms of thepercentage of revenue they derive from sales. The purely donative nonprofit de-rives none of its revenue from sales (Hansmann 1980); at the opposite pole arenonprofits that derive all of their revenue from sales. One index based on thisfinance-source characteristic is the collectiveness index, the proportion of an or-ganization's total revenue derived from donations (see Weisbrod 1988 for dis-cussion and application). Table 6.2, based on IRS tax data for charitable non-profit organizations in 1987, displays a complementary metric: the proportionof total revenue that comes from commercial sources. Its value varies substan-tially by industry, with health organizations being the most commercial (withsales accounting for 89 percent of total revenue) and community-improvementorganizations being the least (only 11 percent).

The more use a nonprofit organization makes of prices to raise revenue, themore it is emulating the private sector; whether such commercial activity takesa form that involves competition with the private sector is another matter. Chap-ter 1 shows how nonprofits' commercialism takes both forms, and Chapter 9 il-lustrates well the growing collaboration between universities and private enter-prise in scientific research. The current chapter attempts to explain the forcesaffecting the commercialism of the nonprofit sector - its increased dependenceon revenue from sales - independently of whether it involves competition orcooperation with for-profit firms.

We have reviewed, within the context of the multiproduct model presentedin Chapter 3, building on James (1983), the reasons for expecting one or anoth-er relationship between the level of a nonprofit's commercial activity and itsrevenue from donations. One analytic perspective indicates that exogenous con-tributions and commercial activity will be either unrelated or inversely related,depending on the degree to which commercial activity itself, apart from the rev-enue it generates, detracts from the goals of the nonprofit; but other perspec-tives suggest positive relationships between donative and commercial revenues.We explore several of these possibilities, including the effects of a change inservice provision by other organizations, economies of scope in production,complementarities in revenues, and price rationing of output.

The section on the empirical relationships between donations and sales usesour newly developed panel of tax-return data covering the period 1985-93 forlarge nonprofit "charitable" - 501(c)(3) - nonprofit organizations. Much of theprior empirical research on the mix of revenue sources used cross-sectional dataand focused on a single industry, such as universities (James 1983) or social-service agencies (Schiff and Weisbrod 1991). We examine each of five indus-


Table 6.2. Percentage of charitable 50J(c)(3) nonprofits engaging inunrelated business activity0

NTEE classification and code% filingUBIT return^

PSRas%of No. oftotal revenue observations

Science and technology research institutes,services (U)Health - general and rehabilitative (E)Disease, disorder, medical disciplines (G)Arts, culture, and humanities (A)Educational institutions and related activities (B)Community improvement, capacity building (S)Religion related, spiritual development (X)Human services - multipurpose and other (P)Recreation, sports, leisure, athletics (N)Mental health, crisis intervention (F)Philanthropy, voluntarism, and grant making (T)Housing, shelter (L)Youth development (O)

9.39.26.36.14.94.64.53.13.02.52.31.91.2

46893627561154415043246869

1283,369167666

2,040488234893183179492702185

aThe data are an asset-based sample, and the estimates presented are weighted averages based onthe sampling weights. However, differences in asset distribution across industries may skew theestimates.bThe UBIT return, Form 990-T, is required for all nonprofits that receive at least $1,000 in grossunrelated business income in a given year.Source: Authors' calculations from the IRS, Statistics of Income Division, data tapes.

tries, using a variety of techniques. The panel data permit us to consider timelags and organization-specific effects.

Our findings may be summarized as follows:1. Aggregate trends for the panel show that although sales revenues are in-

creasing faster than donative revenue, both have increased in real terms. We at-tempt to determine whether the slower rate of growth in donations caused anypart of the increase in commercial activity. We begin by documenting the posi-tive cross-sectional correlations between contributions and sales. Not surpris-ingly, there is a scale effect such that larger organizations tend to receive larg-er amounts of revenue from each type of source, donative and commercial. Thepanel nature of the data lets us control for organization-specific attributes of thenonprofit that do not change over time, including the scale of the organization,its geographic location, and its efficiency.

2. Controlling for organization-specific (fixed) effects, for the full panel ofnonprofits, we find a significant negative relationship between donations and


sales revenues, consistent with the theory that nonprofit organizations have anaversion to commercial activity.

3. Taking a longer view, we also examine the relationship by using thechange in three-year averages, 1985-7 compared with 1991-3, and again finda strong negative relationship.

4. All of these approaches assume that donations are exogenous, an assump-tion we relax in two ways. First, we use fund-raising expenditures to remove,or instrument out, whatever portion of donations is within the organization'scontrol. These expenditures prove to be an important influence on donations:Current and lagged values of fund-raising expenditures, in conjunction withfirm-specific effects and lagged contributions, explain about a third of the year-to-year variation in donations. Second, we use lagged donations, rather than cur-rent donations, as a regressor to explain the extent of commercial activity. Theseproduce some changes but do not alter the overall tenor of our findings. Regres-sion analyses based on log values of the dependent and independent variables,however, reveal substantially smaller results, near zero and even positive forsome industries.

5. Finally, we employ a statistical test of causality that allows for a feedbackeffect whereby variation in sales revenues causes changes in donative behav-ior, and that also allows both donations and commercial activity to be simulta-neously affected by unspecified and unmodeled forces, such as governmentspending, changes in the economic environment, or changes in need. The per-spective presented in Chapter 3 and summarized above portrays nonprofit or-ganizations in two dimensions: the types of services they produce and theirsources of revenue, with the two being related. Thus, anything altering the do-native revenue that a nonprofit realizes without its having to produce the non-preferred output would affect the level of commercial activity in the oppositedirection. A drop in government grants, such as has affected many of the non-profit subsectors examined in Part II of this volume, is likely to bring expan-sion of nonpreferred commercial output, other things equal. That prediction ofa negative relationship could be reversed, however, if, as is also noted in Chap-ter 3, there were more complex interdependencies, so that a change in commer-cial activity had further effects.

In addition to direct effects of donations on commercial activity, there couldalso be reciprocal effects, of commercial activity on donations. These would bepositive if donors favored nonprofits' self-help, revenue-raising efforts and re-warded them, or negative if donors disapproved of commercial activity.2

In addition to revenue interdependencies, there could be cost interdependen-cies. If the production processes are complementary, an increase in ancillary-

2 In the latter case, this effect would strengthen the expectation of a negative relationship betweendonations and sales.


good production will decrease costs for the mission-related outputs, thereby re-ducing revenue needs. Because of such interdependencies of costs and revenues- about which little is known - an exogenous decrease in donations could havea positive, negative, or zero effect on commercial activity, even if the directeffect we hypothesize is negative. In such cases of interdependent revenues orcosts, the effect on the nonprofit of an exogenous change in donations wouldbe to shift the donations function or the cost function for the preferred good.Thus, reality may be more complex than our simple model. No strong predic-tion can be made.

Little is known regarding whether managers or donors are indifferent as tothe nature of the nonprofit's revenue-generating tactics, and whether they gen-erally regard commercial activity as detracting from the organization mission,apart from the revenue generated. Anecdotal evidence, however, suggests thatnonprofits do sometimes forgo opportunities to obtain added revenue becausethey prefer to avoid engaging in those activities. For example, recently OxfordUniversity turned down a gift of $34 million, to establish a new school of busi-ness, apparently because it reached a judgment that a business school was in-compatible with its mission - that is, "with the historical values of the univer-sity"; the view was expressed that "education should prepare people for publicservice, not profit" (Ibrahim 1996). It is not clear, however, whether such deci-sions reflect a sense of organization mission or a recognition of effects on oth-er donors.

It is hard to say which set of assumptions about organization objectives, rev-enue interdependencies, and cost interdependencies is more likely. Moreover,we conjecture that the more likely set differs across industries: Donors to artsorganizations, hospitals, colleges, and antipoverty activities need not respondin the same way to increased commercial activity in ancillary markets. Similar-ly, opportunities for nonprofits to utilize their preferred-good inputs to expandproduction in ancillary markets may well be more limited in some industriesthan in others. Day-care centers, for example, employ inputs that are quite spe-cific to use by young children and are in use for much of the day and even eve-nings, and so appear to have relatively little potential for utilization in ancillaryrevenue-generating activities. Hospitals are more promising loci for productionof ancillary outputs, as they provide services that involve a wide variety of in-puts for inpatients and outpatients, health-care related but also involving foodpreparation, laundry, television rental, and so on, and that have substantial ex-cess capacity.

Changes in provision of the preferred service by other suppliers

The multiproduct model is extendable in several directions that provide furtherinsight into the relationship between contributions and commercial activities.


We consider the case of a nonprofit that is concerned with the aggregate levelof provision of the preferred good, not simply with its own production. An ex-ogenous decrease in production by others would raise the marginal utility sched-ule for the nonprofit's own production, and its desire to increase production.

Would this cause the organization to engage in more commercial activity?Again, the answer depends on whether the commercial activity is a neutral ora nonpreferred activity. A decrease in provision by government agencies or byother nonprofits has essentially identical implications for a specific nonprofit or-ganization as we described for a fall in the organization's revenue from dona-tions. In all these cases the effect is to increase the marginal utility of the non-profit's preferred output and, hence, to cause it to pursue added revenue moreaggressively. Diminished donations, ceteris paribus, reduce production of thepreferred good, moving the organization back along its marginal utility sched-ule to a greater marginal utility of output. A fall in exogenous provision shiftsthe entire schedule outward, again increasing marginal utility from added out-put of the preferred good. The result in both cases is an increase in the margin-al utility of the preferred output, which could be great enough to overcome thedisutility from additional ancillary commercial activity. Thus, we expect suchcommercial activity to be correlated negatively with a decline in provision byother agencies and organizations. Again, however, this result relies on commer-cial activity being nonpreferred, and it could be strengthened or weakened byrevenue and cost interdependencies, as explained above.

Analogous to a change in other providers' output of the preferred goodwould be a change in the overall magnitude of social "need" for the nonprof-it's services. To see this, consider a nonprofit social service agency that observesan increase in poverty. This would be equivalent to a decrease in other sup-pliers' - nonprofit or governmental - provision of antipoverty services, and itwould increase the willingness of a nonprofit organization to initiate or expandproduction of profit-generating ancillary goods. As we have noted, it could alsoincrease private donations. If such a simultaneous increase in both sources ofrevenue occurred, it would not reflect a change in one causing the change in theother, but the simultaneous effect on both revenue sources of the exogenouschange in need or in production by others of the preferred good. The point,which becomes important when we turn to the empirical work, is that findingsof correlation or, for that matter, lack of correlation, between donations andcommercial activity can reflect many plausible scenarios and, hence, are diffi-cult to interpret.

Price discrimination in the provision of preferred goods

Nonprofits' commercial revenue from "sales" can come not only from sales ofan ancillary, nonpreferred good, but also, as we pointed out earlier, from fees


charged to consumers of the preferred, mission-related good. Thus, a positivecorrelation between donative and commercial revenue may also result from theincreased use of price as a rationing mechanism for preferred goods. Access topreferred goods that are of the private-good type can be rationed by varyingprice according to recipients' perceived ability to pay, while also taking accountof the nonprofit's target consumers - those it seeks to help the most.

Thus, a decrease in, say, government grants could cause nonprofit hospitalsto charge private-pay patients more; and colleges could increase their full tui-tion, collecting more from students with the greatest demand while increasingfinancial aid to those with more elastic demand, possibly leaving their net pricesunchanged. Such price discrimination might be interpreted as a form of non-preferred activity, engaged in reluctantly, at less than profit-maximizing levels,thus being responsive to exogenous changes in revenue. Nonprofits that havedistributional goals, seeking to provide the preferred output to particular bene-ficiaries, can choose to use additional donations to cut prices below marginalcost to their target consumers. (For further analysis of price discrimination bynonprofit firms with distributional goals, see Chapter 4 and Steinberg and Weis-brod 1997.) Recognizing the role of prices and price discrimination of preferredgoods is important, for nonprofits' commercial (i.e., private-enterprise-like) ac-tivity is not limited to markets for ancillary goods.

Price discrimination cannot be used, however, for preferred goods that arecollective in nature. Universities can discriminate in their pricing of education,a private-type preferred good, but cannot do so in their provision of collectivegoods, such as basic research. Zoos can discriminate in admission fees to view-ers, but not in their provision of the collective good, animal species breeding.(For an analysis of commercialism in zoos and aquariums, see Chapter 11.)Nonprofits whose missions are to provide collective goods have little opportu-nity to charge user fees, and so their options are more limited.

In light of the complexities we have discussed, it is not surprising that ourfindings, reported in the next section, indicate no simple and consistent pattern.There is considerable variation among industries, some showing negative ef-fects of donations on their commercial activity, but others showing positive ef-fects, and still others showing none. In addition, findings are not entirely robustto alternative specifications. Negative relationships between donations and com-mercial activity are found in some specifications, although by no means uni-formly. The industry studies in Chapters 8-13 further demonstrate the variationin provision of ancillary outputs and in dependence on user fees.

The empirical relationships between donations and sales

A crude measure of a nonprofit's ancillary activity is its engagement in money-raising activities that the IRS regards as unrelated to their tax-exempt mission,


thereby generating income that is subject to the UBIT. Table 6.2 (see the pre-vious section) shows, for 1987, wide variation in the percentage of nonprofitsin each industry that reported such income, ranging from 1.2 percent in theyouth-development sector to 9.2 percent in the health sector. Our empirical anal-yses in the next subsection suggest a pattern, but one that varies among indus-tries.

We assembled the panel data used for this study from Internal Revenue Ser-vice Form 990 tax returns collected annually between 1985 and 1993. The sam-pling scheme used by the IRS Statistics of Income (SOI) Division is based onthe asset level reported on the nonprofit's tax return. The composition of the SOIsample varies from year to year, with larger (greater asset) nonprofits beingmore likely to be included than smaller ones. Although the sample varies overtime, all organizations with more than $50 million in assets (not controllingfor inflation) are included each year. Our analysis uses the "balanced" panel ofthe 2,679 nonprofit firms observed in all nine years. These large organizations,though only a small percentage of all nonprofits, represent a substantial portionof the total assets and revenue activity in the nonprofit sector.

There are two sources of possible bias from the use of this panel of nonprof-its. First, in addition to the fact that nonprofits with low asset levels are under-represented, only "surviving" organizations are included; that is, organizationsthat cease to exist or, for any other reason, do not file a return in any of the sam-ple years are not in the panel. Similarly, nonprofits that are "large" (over $50million in assets) in some year but not in others are excluded from the panelif data for them are not available in the years when their assets fall below thethreshold. Nonetheless, the relationship between organizational revenue fromsales and from donations (referred to in the IRS Form 990 returns as "contribu-tions, gifts, and grants") is worth examining for this important subset of non-profits. Our findings will not necessarily be generalizable, however, to the fullpopulation of nonprofit organizations, most of which are small. There is, on theother hand, a positive dimension to the limited sample: The focus on large or-ganizations provides a degree of homogeneity that is useful given the limitedinformation available from the tax returns.

Two-thirds of our panel consists of nonprofit organizations involved in healthand education activities, as characterized by Independent Sector using theirNational Taxonomy of Exempt Entities, which has been adopted by the IRS.Table 6.3 presents the industrial distribution of the organizations studied. Weanalyze first the combined sample of all organizations, and then the nonprofitsin each of the five industries for which the panel includes over a hundred orga-nizations - specifically, hospitals (NTEE code E22), which are approximately70 percent of the health category (code E), and universities (code B43), whichare about 40 percent of the education category (code B), along with housingand shelter, human services, and arts, culture, and humanities.


Table 6.3. Industrial composition of balanced panel of50J(c)(3) tax returns,1985-93, based on the National Taxonomy of Exempt Entities (NTEE)

NTEE classification and code Frequency % of sample

Health - general and rehabilitative (E)Hospitals, general (E22)

Educational institutions and related activities (B)University or technological institute (B43)

Housing, shelter (L)Human services - multipurpose and other (P)Arts, culture, and humanities (A)Philanthropy, voluntarism, and grant making (T)Community improvement, capacity building (S)Science and technology research institutes, services (U)Youth development (O)Disease, disorder, medical disciplines (G)Religion related, spiritual development (X)Mutual, membership benefit organizations, other (Y)Animal, related (D)Mental health, crisis intervention (F)International, foreign affairs, and national security (Q)Public, society benefit - multipurpose and other (W)Recreation, sports, leisure, athletics (N)Medical research (H)Employment, job related (J)Environmental quality, protection, and beautification (C)Unknown/unclassified (Z)Social sciences (V)Public protection (I)Food, nutrition, agriculture (K)Public safety (M)

Totals

1,121(779)643(265)155143128928937312827212020191917161413126422

41.8(29.1)24.0(9.9)5.85.34.83.43.31.41.21.01.00.80.70.70.70.70.60.60.50.50.40.20.10.10.1

2,679 99.7«

Source: Authors' calculations from the IRS, Statistics of Income Division, data tapes.aDoes not equal 100% because of rounding.

The upper portion of Table 6.4 (rows 2-5) presents descriptive statistics onthe organizations in the panel at the beginning of the time interval, 1985.3 Allvalues are in millions of constant dollars, adjusted by the Consumer Price In-dex (1982^- = 100). In 1985 the average nonprofit in the panel received $6 mil-lion in contributions, gifts, and grants while generating more than five times that

3 Data are also available for 1982 and 1983, although not for 1984, when the IRS did not take asample. Because of the missing year, we examine data only for the period beginning with 1985.

Table 6.4. Descriptive statistics for balanced panel of 50J(c)(3) tax returns, 1985-93

1. Number of organizations

2. CGG (mil. of 1982-4$)3. Program service revenue

(mil. of 1982-4$)4. Assets (mil. of 1982-4$)5. Pet. filed a UBIT return

6. Real growth of CGG revenue7. Real growth of program

service revenue8. Real growth of assets9. Increase in fraction filing

UBIT return

Fullbalancedpanel

2,697

University ortechnol. inst.NTEE code B43

265

1985 average ($)6.0

32.487.923

24.4

53.6222.5

34

1985-93 (% change)21

6153

15

18

3819

25

Hospital,generalNTEE code E22

779

1.3

57.073.336

38

6343

24

Housing,shelterNTEE code L

155

0.6

4.118.05

44

1622

0

Human services -multipurpose & otherNTEE code P

143

3.9

5.620.210

17

509

-1

Arts, culture,humanitiesNTEE code A

128

7.9

3.648.226

9

4742

10

Abbreviations: CGG, Contributions, gifts, and grants; NTEE, National Taxonomy of Exempt Entities; UBIT, Unrelated Business Income Tax.Source: Authors' calculations from the IRS, Statistics of Income Division, data tapes.


- $32.4 million - in program service revenues. The size and relative importanceof these revenue sources varies widely by industry: The average hospital is farmore commercial, receiving only $1.3 million in contributions but having $57million in sales revenues. At the other extreme, the average arts organizationin our panel received more than twice as much in contributions as in sales rev-enues. Row 5 of the table displays the fraction of firms filing a Form 990-T taxreturn, reflecting their receipt of revenue from unrelated business activities. Thismeasure again discloses great variation among industries in their involvementin commercial activity, in this case with unrelated ancillary activities: More thana third of hospitals and universities filed a UBIT return, compared to only 5percent of the housing and shelter organizations.

Such variation in commercial activity is consistent with our prior view thatbecause of the differing technologies of production in the various mission-focused activities, there would be variation among industries in their potentialnot only for developing profitable markets for ancillary goods, whether legallyrelated or unrelated, but also in their opportunities to generate revenue from userfees for mission-related activities. At the same time, these findings are also con-sistent with other models, such as that organization objective functions, includ-ing aversion to ancillary activities and to user fees for mission-related activi-ties, differ across industries.

Between 1985 and 1993 the revenue from each of the two principal sources,contributions (i.e., CGG) and sales, increased substantially in constant dollars.Contributions, gifts, and grants to nonprofits in our panel grew by more than21 percent (row 6), but sales grew 61 percent (row 7), nearly three times asmuch. There has, indeed, been increasing reliance of nonprofit organizations oncommercial activity. Consistent with this view, the fraction of nonprofits filinga tax return for unrelated business activity increased from 23 percent in 1985to 38 percent eight years later.

This upward growth trend, however, masks variation over time: Althoughaverage real contributions grew continuously for our panel of nonprofits, therewas variation from year to year in that growth. Moreover, in any particular yearmany nonprofits did not experience growth, with 40-60 percent of the sampleexperiencing a decline.

The discussion of the earlier section, "Theoretical relationships between rev-enues from donations and from program services," suggested that a decline inexogenous donations to a nonprofit organization may cause it to turn, if reluc-tantly, to increased commercial activities in order to cross-subsidize its preferredactivities, implying a negative "crowding out" relationship between donationsand commercial activity (see also Chapter 3). Other parts of the analysis, how-ever, suggested a positive, "crowding in" relationship, due to such additionalforces as the possible effects of commercial activity on donations and the abil-


Table 6.5. Regression analysis of balanced panel of50J(c)(3) taxreturns, 1985-93

Specification0 Estimate

Effect on program service revenue of:1. Contributions2. Contributions, w/ firm effects3. Contributions, w/ firm and time effects4. Contributions, w/ firm effects instrumenting for the possible

endogeneity of contributions5. Lagged contributions, w/ firm effects

Effect on log program service revenue of:6. Log contributions^7

7. Log contributions, w/ firm and time effects*7

Long-run regression analysis avg. (1991-3) - avg. (1985-8):8. Effect on program service revenue of contributions9. Effect on log program service revenue of log contributions^

Long-run regression analysis omitting zero values,avg. (1991-3) - avg. (1985-8):10. Effect on program service revenue of contribution11. Effect on log program service revenue of log contributions*7

Statistical tests for the direction of causality0

12. Significance of lagged contributions on program service revenue13. Significance of lagged program service revenue on contributions

0.76**-0.09**-0.13**

-0.13**-0.43**

0.10**-0.02

-1.05**-0.02

-0.73**-0.02

0.010.40

aAll dollar values are deflated by the Consumer Price Index for All Urban Consumers(1982-4= 100).^Log-log regression models use zero in place of log(0).cTests are based on a vector autoregression with firm fixed effects, and two lags of con-tributions, program service revenue, and fund-raising expenditures. Longer-lagged val-ues are used as instruments.* Statistically significant at the .10 level.** Statistically significant at the .05 level.Source: Authors' calculations from the IRS, Statistics of Income Division, data tapes.

ity of the organization to utilize the resources from its preferred goods in theproduction of ancillary goods.

To test these relationships we apply a more disaggregate, microlevel analy-sis. We begin by describing the methodology and results for the full panel oforganizations in Table 6.5. The panel data permit us to examine changes overtime while controlling for unobserved organization characteristics that do notchange over time; that is, we employ a "fixed-effect" regression model that


takes into account characteristics of each organization that, although not direct-ly observable, are specific to the organization over time.

Our findings indicate that unobserved heterogeneity among organizations isof major importance. Rows 1 and 2 of Table 6.5 present the regression estimatesof the effect of a one-dollar increase in contributions on the contemporaneouslevel of sales, with and without inclusion of a firm-specific component; the re-sults differ significantly and substantially. The regression without the fixed ef-fect estimates that a one-dollar increase in contributions is associated with aseventy-six-cent increase in sales activity - no crowding out effect here. Webelieve, however, that this estimate reflects scale differences across organiza-tions such that larger organizations receive more revenue, both commercial anddonative, than smaller ones. The regression estimate reflects a positive correla-tion due to scale but does not reflect causality. Inclusion of fixed effects, whichwe believe to be important, results in a statistically significant negative coeffi-cient on contributions. As implied by the cross-subsidization model with aver-sion to commercial activity, changes in total contributions are negatively cor-related with changes in gross sales revenue.

The negative result for the fixed-effect model in the pooled sample of orga-nizations is robust to a variety of alternative specifications, including the addi-tion of time-trend variables to capture aggregate economic conditions (row 3),the inclusion of fund-raising expenditures in an attempt to deal with the possi-ble endogoneity of donations (row 4), and the use of a one-year lagged inde-pendent variable instead of contemporaneous donations to deal with endogene-ity as well as effects over time (row 5). Controlling for the endogenous portionof contributions somewhat increases the negative relationship between dona-tions and sales. The coefficient displayed in row 5 of Table 6.5 suggests that aone-dollar decline in exogenous donations causes an increase in commercial ac-tivity of forty-three cents, although the profit available to cross-subsidize pre-ferred activities is surely less.

As a further robustness check we estimate the model in logs rather than lev-els, with (row 7) and without (row 6) fixed effects. These results are much clos-er to zero, suggesting that large values - which receive relatively less weight inthe logarithmic formulation - have significant influence on our estimates, eventhough the data sample is already limited to organizations with large amountsof assets, and even though we used fixed effects. The log-log model with fixedeffects again produces a negative coefficient, but it is not statistically signifi-cant. The interpretation of the estimate in row 7 is that a 1 percent decrease indonations causes a 0.02 percent increase in commercial activity.

The effect of changes in contributions on sales activity might well dependnot on single-year, transitory changes, but on changes expected to be "perma-nent." The single-year changes might measure fluctuations that are smoothedacross time, which may induce fluctuations in assets rather than sales. More-


over, data for any given year can be misleading, since year-end donations maysometimes be reported in the next year. To address these concerns we use three-year averages of the data in the regression model. That is, for each nonprofitwe form the average of gross sales and of contributions for 1985-8 and, again,for 1991—3; then we compute the change in each variable between the two pe-riods, thereby smoothing effects that might result from, say, a large amount ofdonations arriving either shortly before or shortly after the end of a nonprofit'sfiscal year. Row 8 of Table 6.5 presents these estimates based on a specifica-tion in levels. The result, again, is strongly negative and statistically significant.However, the log-log specification (row 9) continues to produce a statisticallyinsignificant result.

The difference between the level and log specifications led us to explore thedata further. We noted that there were particularly large fluctuations involvingtransitions to or from zero contributions. For example, the returns for the NewEngland Medical Center Hospital (EIN 042374071) report contribution above$20 million per year for 1985-8, zero contributions for the next two years, andthen contribution levels above $40 million for the next three years. Perhaps suchzeros are reporting error; perhaps they reflect large transitions of organizationsundergoing radical change - we do not know how to explain such anomalies.Moreover, although the Internal Revenue Service scrutinizes tax returns for in-ternal consistency, it does not examine the consistency of the returns over time.In order to reduce the effect of large transitions involving zero contributions,we repeated the long-run analyses, omitting from the estimation of mean contri-butions any reported zero. Thus, if an organization had two years with positivedonations followed by a year with zero, we used the two positive values butexcluded the year with the zero contributions. As we expected, the estimatesfor the effect of contributions on sales are somewhat smaller in absolute valuethan had been found when the zero observations were included, but they remainlargely consistent with the earlier analysis.

Overall, we find somewhat mixed relationships between an organization'sdonative income and its revenue from sales. For the aggregate sample, Table6.5 shows that a number of formulations, particularly those employing firm-specific effects, indicate significant, negative relationships (lines 2-5), where-as others, based on log values, show effects that are negative but insignificant(lines 7 and 9) or, in one case, positive and significant (line 6). The differencesbetween relationships observed in levels and in logs are noteworthy. For rea-sons not well understood, there seem to be larger absolute responses, but small-er percentage responses of sales activity.

Negative relationships in Table 6.5 generally show effects less than unity. Atthe outset of this chapter we conjectured that when nonprofits use commercialactivity to replace lost CGG, they replace them only partially. Moreover, sincecommercial revenue is only partially profit, the replacement rate is even lower


than is indicated by the coefficients on sales. This finding is consistent with thehypothesized aversion to commercial activity, although we cannot be certainas to the influence of cost and revenue interdependencies.

The structural regression models employed thus far assume that causalityruns from donations to sales. Explanations put forth in the theory section, how-ever, suggest that even if causation does run in that direction, it could simulta-neously run in the reverse direction - fluctuations in commercial revenue af-fecting donative revenue - or that both variables could be responding to a thirdvariable omitted from the analysis. One way to detect such relationships is touse a system of simultaneous equations to determine whether fluctuations in onevariable precede fluctuations in the other, after controlling for all other avail-able information. A statistical model that provides a test of such so-calledGrainger causality is the panel vector autoregression described in Holtz-Eakin,Newey, and Rosen (1988). Using it, we relate current sales activity to the pre-ceding two years' data on donations, fund-raising, and sales revenue. Separate-ly, there is an equation with donations as the dependent variable and the sameset of regressors. All of the data arc first-differenced - that is, differences be-tween years are used - to remove the firm-level effect we identified as impor-tant in the earlier analyses. This transformation, combined with the presence ofa lagged dependent variable, produces a correlation between the error term inthe regression and the transformed right-hand-side variable that is eliminatedusing earlier data as an instrument. That is, we regressed the change in salesactivity in year t on the changes in sales, contributions, fund-raising, and assetsin years t—\ and t-2 while using the data prior to t-2 as instruments.

Using this technique on the data for all industries, we find evidence that con-tributions statistically cause - that is precede - changes in sales, but sales donot cause contributions. The causality tests in row 12 of Table 6.5 reports thesignificance level of a joint F-test measuring the degree to which lagged contri-butions are a significant predictor of current sales after controlling for the orga-nizational fixed effect and prior values of all the variables.

Industry-specific effects of donations on sales revenue

Results for five industry subsamples are presented in Table 6.6, which we havelimited to the most important specifications. These five subsamples constitute55 percent of our aggregate sample; the balance consists of nonprofits (shownin Table 6.3) in industries for which samples are quite small.

Our findings for the housing/shelter and arts/culture sectors largely match theaggregate estimates - in our preferred, fixed-effect, formulations (rows 2 and4) - in both the nonlog and log estimates. The estimates from the log-log spec-ification, insignificant for the combined sample, are negative and statisticallysignificant for the two industries (row 4).

Table 6.6. Regression analysis by industry of balanced panel of 50J(c)(3) tax returns, 1985-93

University or Hospital, Housing, Human services - Arts, culture,technol. inst. general shelter multipurpose & other humanities

Specification* NTEE code B43 NTEE code E22 NTEE code L NTEE code P NTEE code A

Effect on program service revenue of:1. Contributions 0.92**2. Contributions, w/ firm and time effects 0.15**

Effect on log program service revenue of:3. Log contributions^ 0.21**4. Log contributions, w/ firm and time effects^ -0.02

Long-run regression analysis, avg. (1991-3) - avg. (1985-8)5. Effect on PSR of contributions 0.34**6. Effect on log PSR of log contributions^7 -0.01

Statistical tests for the direction of causality0

7. Significance of lagged contributions on PSR 0.328. Significance of lagged PSR on contributions 0.06

2.54*-0.01

0.08*0.02*

0.78*0.03

0.220.30

0.67**-0.38**

0.18**-0.22**

-0.43-0.27**

0.010.27

1.63**1.21**

0.27**0.14**

2.95**0.19

0.010.01

0.10**-0.15**

0.33**-0.10**

-0.23**-0.06

0.810.36

aAll dollar values are deflated by the Consumer Price Index for All Urban Consumers (1982^ =100).^Log-log regression models use zero in place of log(0).'Tests are based on a vector autoregression with firm fixed effects, and two lags of contributions, program service revenue, and fund-raising expenditures.Longer-lagged values are used as instruments.* Statistically significant at the .10 level.** Statistically significant at the .05 level.Source: Authors' calculations from the IRS, Statistics of Income Division, data tapes.


For the other three industries, however, findings are quite different. Betweencontributions and sales for universities and human-service organizations, wefind relationships that are both positive and statistically significant, but the re-lationship is essentially zero for hospitals (row 2). The log estimates (row 4)are also positive and significant, except for in the case of universities, wherethere is no significant effect. These three industries comprise 44 percent of ouraggregate sample. The fact that we find positive relationships between contri-butions and sales for them, but negative relationships for the aggregate, indi-cates that there are many organizations in the panel that exhibit negative rela-tionships in industries where samples were too small to analyze.

The association of increased donations with increased, rather than decreased,commercialization in an industry is not consistent with the simple multiprod-uct organization model in which commercial activities are disliked. It is con-sistent, however, with that model if, in those industries, there are economies ofscope in production - that is, cost complementarities or complementarities inrevenue. As we explained earlier, there is no reason to believe that such inter-dependencies are equally important across industries. We also noted other cir-cumstances that could elicit a finding of a positive relationship between dona-tions and sales, even if commercial activity is disliked; for example, a reductionin government grants could cause a simultaneous increase in private contribu-tions and commercial activity. There is a clear need to learn more about bothcost and revenue interdependencies in nonprofit organizations before it will bepossible to state with confidence the effect on commercial activity of changesin some source of donative revenue.

The results of tests of causality from commercial activity to donations alsovary by industry. Rows 7 and 8 indicate that fluctuations in donations affectcommercial activity in the housing sector, while in the university sector it ap-pears that the driving force is the reverse, from sales to donations. In two indus-tries, hospitals and arts, there is no evidence of causation in either direction,suggesting that both donations and sales are being influenced by a third factor.In another industry, human services, there is evidence of significant causationrunning in both directions.

Concluding remarks

The appropriate model for nonprofit behavior remains far from settled. So, too,is an understanding of how nonprofits come to have particular combinations ofrevenue from donations and from sales of goods and services, and particularcombinations of revenue from mission-related and unrelated activities. How-ever, a multigood production model, in which nonprofits determine the extentof activity in several sectors, is promising. Using data for 2,679 nonprofits ob-served from 1985 to 1993, we estimated a set of structural and reduced-form


models that suggest that exogenous declines in donations yield significant in-creases in commercial activity for some industry sectors but not for others. Theresults vary considerably by industry, a finding that is understandable in lightof differential organization goals as well as differences in interdependenciesamong revenue opportunities and in cost interdependencies between mission-related and ancillary goods production. Our findings indicate strongly the needfor more analyses of output and revenue determination in specific industrieswhere nonprofits play a major role.

Several questions raised during the analysis require further research. First,within our model of nonprofit behavior, contributions are considered exoge-nous. Relaxing this simplification involves determining the extent to which or-ganizations can influence contributions, including the effect of a nonprofit'scommercial activities on willingness of individuals and organizations to give itdonations, and the extent to which other unspecified events affect both sales anddonations. For example, changes in sources of sales revenue, such as throughthe adoption of Medicare and Medicaid in the health industry, can affect or-ganization behavior and, in turn, the supply of donations (including volunteertime). Second, a dynamic model of behavior suggests that nonprofits may drawdown assets in response to exogenous transitory fluctuations in donations.

Debate over whether government spending on social welfare programs hascrowded out private donations, so that current and prospective reductions ingovernment support will lead to increased private donations, illustrates the im-portance of these issues. In addition, our findings that reduced donations havequite different effects on commercial revenue-raising activity in the higher-education, hospital, arts, housing, and human-services components of the non-profit sector highlight the danger of generalizing about the responses of the sec-tor, as a whole, to exogenous changes in donations.

CHAPTER 7

Conversion from nonprofit to for-profitlegal status: Why does it happen andshould anyone care?

John H. Goddeeris and Burton A. Weisbrod

Introduction: Conversion as the ultimate commercialism?

If to behave commercially is to act like a for-profit firm, then the ultimate ex-pression of commercialism for a nonprofit is to convert its legal status to thefor-profit form. Conversion is increasingly common, most notably in healthcare, and is now attracting considerable public attention. Some observers be-lieve that nonprofits and for-profits inevitably behave in fundamentally differ-ent ways, and question whether conversions can ever serve the public interest.Others are more concerned about the terms, arguing that public assets must re-main devoted to the purpose for which nonprofit status was originally granted,and not be redirected toward private gain.

This chapter seeks to advance understanding of nonprofit conversions andtheir public-policy implications. To set the stage, we introduce some key issuesby reviewing briefly conversion activity in health care. We then consider theconcept of conversion more closely, discussing various means of transferringcontrol of nonprofit assets. Next we explore the possible motives for conver-sion, and speculate about the reasons for the rampant conversion activity inhealth care. Finally, we discuss at some length the important public-policyquestions raised by conversions. In our view, the most important are the fol-lowing:

1 Under what circumstances is conversion appropriate?2 How should the nonprofit's assets be valued? and3 What should happen to the financial assets that remain after the con-

version?

The authors thank Daniel Fox, Bradford Gray, James Schwartz, Steven Rathgeb Smith, and Wil-liam White for helpful comments, and Elizabeth Selvin for valuable research assistance.

129

130 John H. Goddeeris and Burton A. Weisbrod

Conversions in health care

Conversions in the health maintenance organization (HMO) industry have beencommon and increasingly controversial. Nonprofit firms - the only type eligi-ble for federal subsidies under the Health Maintenance Organization Act of1973 - initially dominated this industry. When direct federal subsidies endedin the early 1980s, however, market share began to swing heavily toward for-profits, partly through entry of new for-profit HMOs and their growth, but alsopartly from conversion of nonproflts (Iglehart 1984). From 1981 to 1995, thepercentage of HMOs classified as nonprofits plummeted from 82 percent (ac-counting for 88 percent of overall membership) to 29 percent (41 percent ofmembers) (Claxton et al. 1997, 12).

One of the earliest conversions was the transformation of the HMO of Penn-sylvania into U.S. Health Care in 1981. Some successful nonprofit HMOs fol-lowed this lead to become publicly traded corporations in the 1980s, with rel-atively little scrutiny. However, when managers and directors of Health Net- one of California's largest HMOs - offered to buy it for $108 million in 1991,consumer groups and other HMOs noticed the seeming undervaluation of thenonprofit's assets. When California's Department of Corporations finally ap-proved the conversion in 1992, it required a vastly higher payment: $300 mil-lion plus an 80 percent ownership interest in the new for-profit HMO. Theseproceeds were given to the nonprofit Wellness Foundation. Despite the massiveincrease in payment, some Health Net insiders reaped very large personal finan-cial gains. For that reason, and because the foundation's stock ownership madeit dependent on the financial success of Health Net, these arrangements havealso been criticized as inadequately protecting public assets (Hamburger, Fin-berg, and Alcantar 1995).

Actual and proposed conversions of various Blue Cross health-insuranceplans into for-profit entities have also provoked controversy. In June 1994 thenational Blue Cross and Blue Shield Association decided for the first time thatfor-profit firms could affiliate with the organization (Freudenheim 1994), pav-ing the way for several conversions. Even earlier, some plans had created sub-stantial for-profit subsidiaries. The most prominent example is California BlueCross, which in 1992 restructured so that 90 percent of its assets went into Well-Point Health Networks, a for-profit subsidiary. To Blue Cross, this was not aconversion - the parent entity continued to exist and owned 80 percent of Well-Point - but consumer groups and the Department of Corporations maintainedthat this arrangement potentially diverted public assets toward private ends. Ul-timately, WellPoint agreed to purchase all of the assets of California Blue Cross(including use of the name) in 1996, completing the conversion and providing$3 billion in cash and WellPoint stock for the creation of two new charitablefoundations (Hamburger et al. 1995; Schaeffer 1996).

Conversion from nonprofit to for-profit legal status 131

Accelerating conversion activity in the hospital industry has been propelledlargely by the aggressive action of Columbia/HCA, the nation's largest chainof for-profit hospitals (Kuttner 1996a,b). Nationwide, thirty-four conversionsof nonprofit hospitals to for-profits took place in 1994, fifty-eight in 1995, andmore than a hundred government or nonprofit hospitals were expected to affil-iate with or convert to for-profits in 1996 (Langley and Sharpe 1996). Interestin conversions extends to prestigious academic medical centers: The universi-ty hospitals at George Washington University, Columbia University, the Univer-sity of California at Irvine, the Tufts University New England Medical Center,and state universities in Oklahoma and South Carolina are among those beingsold to, or at least holding discussions with, for-profit hospital chains (Freuden-heim 1997). As the number of conversions has grown, so has scrutiny fromcommunity groups and state attorneys general and other regulatory bodies. Pri-mary concerns are whether a greater focus on profit will lead to a loss of un-profitable but valued community services, and whether hospital assets accu-mulated through community investment in nonprofits will be redirected towardprivate gain.

In a case that cuts across the hospital and health-insurance industries, Colum-bia/HCA agreed to buy 85 percent ownership of Blue Cross and Blue Shield(BCBS) of Ohio for $299.5 million (Tomsho 1996). As in the California case,BCBS of Ohio argued that because it would continue to exist, the deal wouldnot be a conversion. Moreover, BCBS of Ohio had already converted in 1986from a nonprofit to a mutual company owned by its policyholders. This compli-cated the question of its charitable obligations, because a mutual has no legalobligation to provide charity care or any other unprofitable service, whateverits social value. Among the controversial features of the Columbia deal wereover $15 million in severance packages, characterized as consulting fees andagreements not to compete, which would go to three BCBS of Ohio executives.Ohio's Attorney General sued to block the transaction, and the national BlueCross and Blue Shield Association voted to revoke the use of its trademark ifthe deal went through (Kuttner 1996b).

Factors peculiar to the health-care sector surely contribute to the drive forthese conversions, but increasing pressures to engage in commercial activity arecommon throughout the nonprofit sector. Greater access to capital is argued tobe a major advantage of the for-profit institutional form in health care; but ifthat is important for explaining conversions, it is presumably also an advantagein other industries such as colleges, museums, and social-welfare organizations.If, as critics claim, a major attraction of conversion is the opportunity for non-profit managers and board members to benefit privately, a similar opportunitysurely exists in other industries as well. It therefore seems prudent to think aboutthe phenomenon of conversions and its public-policy implications more broad-ly. It is also important, from a public-policy perspective, to develop appropri-


ate legal processes for determining the conditions under which a conversion issocially desirable, and to establish incentives that will encourage desirable anddiscourage undesirable conversions.

What is a nonprofit conversion?

On the face of things, "conversion to for-profit status" suggests that a nonprofitentity simply reorganizes as a for-profit firm. However, even a clear conversionof form is complicated by the fact that the nonprofit's assets do not belong toindividuals. If the nonprofit ceases to exist, where should these charitable as-sets go? Just as important, various other interactions between a nonprofit firmand the for-profit sector are possible (including for-profit subsidiaries and jointventures), which, while stopping short of full conversion, accomplish at leastsome of the same ends.

It is useful to think of any change in control over assets or responsibility forliabilities from the nonprofit to the for-profit sector as a kind of conversion. Ajoint venture between a nonprofit and a for-profit is a conversion if it alters thelocus of control over core assets of the nonprofit - as it has, for example, inarrangements between Columbia/HCA and various nonprofit hospitals. Conver-sion may occur even when assets are sold for a "fair market value": If the as-sets are deployed differently by the new private owners, their market value maybe insufficient to replace the collective goods that will be lost. Even making apassive investment in a private firm involves converting some assets, in thesense of relinquishing some control.

A nonprofit may also contract with a for-profit firm for management services.Whether such a contract involves significant public-policy issues depends onhow much control is transferred, the degree to which the for-profit firm is con-strained, and on the reward structure for the for-profit. Such transfers of man-agerial control have been made increasingly in hospitals and academic medicalcenters, as well as in public school and prison systems. Schools in Minneapo-lis, Hartford, and elsewhere have been put under private-sector management byfirms such as The Edison Project and Apollo Group, Inc., and prisons have beenincreasingly privatized through management contracts with firms such as Cor-rections Corporation of America and Behavioral Systems Southwest. In thesecases the nonprofit (or public) organization, rather than being sold, transferscontrol to a private firm for a finite period of time, with a variety of contrac-tual constraints spelled out. However, if nonprofit institutions are most valu-able in cases where some socially important dimensions of performance arevery difficult to define in a contract and to monitor (Weisbrod 1988), any trans-fer of managerial authority to a for-profit may threaten the nonprofit's originalgoals.


Although we shall generally speak of only two organization types, nonprofitand for-profit, there are in reality various legal forms of nonprofit, each facingsomewhat different constraints, just as there are different organization typesin the for-profit sector (e.g., incorporated and unincorporated businesses). Thenonprofit mutual-benefit corporation, for example, shares with the for-profitform the fact that its assets are privately owned, in this case by the organiza-tion's subscribers or members. Thus, a change in organizational form from char-itable nonprofit to mutual-benefit nonprofit, as has occurred for a number ofBlue Cross plans in the past decade, is in an important sense a conversion.

Despite such complexities, we focus in this chapter on the issues involvedwhen the resources of a nonprofit entity are substantially shifted to for-profitcontrol, and the nonprofit ceases to exist in its previous form. Cases of this type- such as the conversion of California Blue Cross to WellPoint Health Net-works, or the proposed purchase of most of BCBS of Ohio by Columbia/HCA- are attracting the most public notice. For purposes of public policy, however,we should remember that conversion is a matter of degree. Given the increas-ingly bright searchlight of public attention focused on total conversions, thegoals motivating conversions will likely be pursued through more subtle means.We expect more joint ventures, partial buyouts, and complex interorganizationcontracts that preserve the apparent independence of each organization whileactually shifting the locus of control.

What motivates conversions?

Modeling choice of institutional form

Determining why managers and trustees of a nonprofit organization would seekto convert it to a for-profit requires asking why the nonprofit form was choseninitially, and then what changed to make it less desirable. It is often useful tomodel organization behavior as the maximization of an objective function sub-ject to constraints. From that perspective, the choice of organizational form,nonprofit or for-profit, is a choice of constraints, because these differ for the twotypes of institution. The particular form is chosen because it is the most con-ducive to attaining the founders' goals.1

Constraints are imposed both by the legal system and by the market. Legalconstraints that differ between for-profits and nonprofits include tax liability onincome, real property, sales, and so on; access to tax-deductible contributions;access to equity capital; restrictions on "profit" distributions; the availability of

1 The problem is most appropriately viewed as a dynamic one, with the firm looking at long-runobjectives and the possibility that constraints may change in the future, and recognizing possi-ble change in organizational form at a later date.


government grants and subsidized loans; and regulatory requirements regard-ing product offerings, pricing, and access. Nonprofits possess certain tax advan-tages but lack access to equity capital (because that would imply private own-ership shares), and they are legally restricted by the nondistribution constraint(Hansmann 1980) from distributing any profits or surplus to managers or boardmembers.

The market also constrains behavior. For-profit and nonprofit firms are con-strained by demand conditions, although not necessarily identically. In somecases, buyers might regard nonprofits as more trustworthy and prefer to pur-chase from them. Trust is particularly important for goods with essential attri-butes that are difficult to monitor directly - for instance, the "tender loving care"provided in a nursing home or day-care center. If at least some buyers regardnonprofit status as a signal of trustworthiness, demand will differ between thetwo types of firm, even if all observable product characteristics are identical.

Differences in market-imposed constraints might also affect the supply oflabor or other resources. Volunteer time is more readily supplied to nonprofitfirms, and even paid workers sometimes accept lower wages to work for non-profits (Weisbrod 1983; Preston 1989; Roomkin and Weisbrod 1998; but cf.Goddeeris 1988), perhaps because they identify with a nonprofit's goals and de-rive satisfaction from working to achieve them.

For-profit and nonprofit firms coexist in a number of important industries:Health insurance, hospitals, nursing homes, and day-care centers are leadingexamples. Nonetheless, there may be important differences in behavior acrossorganizational type, even within an industry, and different choices about orga-nizational form may reflect different managerial objectives. Suppose, for exam-ple, that the founding managers of a medical clinic care about two things - theirown monetary compensation and providing subsidized health care for the poor- and must decide whether to organize as a nonprofit or for-profit. There is like-ly to be some trade-off between managerial compensation and the amount ofcollective goods that can be supported in either sector, once chosen. Maximalmonetary compensation may be higher in the for-profit sector, if the nondistri-bution constraint actually limits compensation of nonprofit managers; but larg-er collective-good output may be possible in the nonprofit sector because of taxadvantages and the ability to attract donations. Managers who place less valueon the collective goods would thus organize as for-profits. Those with a strongenough preference for providing such services would pick the nonprofit sector.

Why do conversions occur?

Within this framework, how can we explain conversions? A useful working hy-pothesis is that changes in legal constraints or market opportunities are the pri-mary cause. If, for example, nonprofit firms are exempt from a property tax that


other firms must pay, a cut in the tax rate reduces one advantage of the nonprof-it form. Such a policy change could induce an organization with a sufficientlyweak preference for the nonprofit form to switch status (Chapter 5). Althougha shift of managerial goals in the direction of personal financial gain could alsoaccount for a conversion, economists generally find it more fruitful to modelgoals or preferences as stable, and to predict behavioral changes on the basisof changes in constraints. Constraint changes - particularly changes in tax lawsor regulations, but also the availability of private donations and governmentgrants and contracts - are relatively easily observed, and they affect all similar-ly situated firms in the same way. Moreover, without a model that predicts sys-tematic changes in goals that simultaneously affect many organizations, andonly in certain sectors, such an "explanation" does not elucidate today's patternof conversions.

At the same time, organizational goals or preferences are not entirely in-dependent of constraints. A nonprofit organization's goals are determined byits managers and board, since it has no stockholders or other owners, and thisset of key decision makers often changes over time, sometimes in response tochanges in constraints. Goals may change endogenously as the nonprofit en-counters budgetary difficulties in the pursuit of its original goal, and the orga-nization takes on board members who represent sources of actual or potentialfinance. In the early 1980s, for example, a nonprofit food bank in Phoenix withlocal volunteers as directors witnessed a struggle in which food manufactur-ers, such as Beatrice Foods and Kraft, and food marketers came to dominatethe board (Birnbaum 1982). In the course of time, an organization's goals maychange considerably, as managers with greater concern about finance replacethose whose visions were of public-service outputs. As the altered goals growless compatible with traditional nonprofit constraints, conversion may becomeappealing.

A decision to convert can thus be viewed as a rethinking of the original deci-sion about choice of organizational form, which may be triggered by changingconstraints or managerial goals. There are, however, important differences be-tween a conversion decision and the initial choice of form. A desire to convertmay evolve naturally over an organization's life cycle, for example, even if theexternal environment and goals do not change. For a newly established non-profit firm with few assets, the nondistribution constraint may not matter, as thefirm has little or nothing to distribute. As the market value of assets grows, how-ever, the firm's managers and board will be tempted to distribute some of thatvalue to themselves. If public policy permits conversions to occur in a mannerthat effectively subverts or even weakens the nondistribution constraint, a con-version may occur. Some organizations could even pursue a long-range strat-egy of forming as nonprofits and converting to for-profit status if successful.

On the other hand, if a conversion occurs and the firm's behavior shifts to-


ward maximizing market value, the collective goods it was providing as a non-profit are not necessarily lost in the process: The income from the accumulatedassets could be redirected toward their continued provision (as could the in-creased taxes, if any, that the firm will now pay as a for-profit entity, althoughtheir disposition is outside the control of the firm's decision makers). Laws andregulations about the disposition of assets after conversions will thus have animportant impact on incentives to convert, as we explore in more detail below.

The Bennington College case discussed in Chapter 1 illustrates another mo-tivation for the conversion of control over at least some assets (Galper and To-der 1983). The college sold its buildings to a private firm and then leased thebuildings back. The difference in tax rules applied to the nonprofit college andfor-profit private firm - tax deductions for depreciation on the buildings wereof no value to the college - made this advantageous for both, though nothingchanged in the use of the buildings. The gains to the participants came at theexpense of other taxpayers, and the IRS ultimately prohibited this particularform of legal legerdemain. Even so, differences in rules for different organiza-tional forms continue to create incentives for transactions between them that areprivately advantageous but of no social value. These incentives always exist,but they may become stronger or weaker with changes in legal constraints ormarket conditions, and they may interact with other motives for conversion.

Although we have emphasized conversions from the nonprofit to the for-profit form, nothing in the theory we have outlined requires that conversions goonly in that direction. If some firms were near the margin when initially select-ing the for-profit sector, a change in constraints that increased the attractivenessof the nonprofit sector might tip them toward conversion. A possible asymme-try between conversions in the two directions is that to the extent conversionsare part of a long-term strategy motivated primarily by private gain, we mightexpect the more successful nonprofits to seek conversion to for-profit status (toescape the nondistribution constraint), whereas for-profits that seek conversionwould be the relatively unsuccessful ones. For these for-profits that have doneless well than their initial expectations and are only marginally viable, nondis-tribution may not be an important issue, and the prospect of tax advantages orsubsidies available to nonprofits may look increasingly attractive.

Recent studies of the hospital industry document that conversions amongpublic, nonprofit, and for-profit institutions occur in all directions (Ferris andGraddy 1995; Needleman, Chollet, and Lamphere 1997). Needleman and col-leagues find that between 1990 and 1995 alone, fifty-six for-profit hospitals con-verted either to nonprofit or public institutions.

Understanding recent health-care conversions

Just why conversions are so widespread in the health-care sector today requiresanalysis beyond the scope of this chapter. In brief, however, there is reason to


believe that changes in the legal and regulatory environment, as well as in thenature of market competition (i.e., constraint changes), have made it more dif-ficult for nonprofit health insurers, and nonprofit hospitals, to survive in com-petition with for-profits.2 Indeed, several policy changes have made it moredifficult for health-care nonprofits to finance collective goods successfully. ForHMOs, the end (in 1983) of federal grants and loans for development reducedthe nonprofit advantage; this change appears to coincide with and perhaps stim-ulated the shift toward for-profit status. For Blue Cross and Blue Shield plans,too, tax advantages have eroded over time: Exemption from the federal corpo-rate income tax was ended in the Tax Reform Act of 1986, and many states havelimited the special tax advantages of Blue Cross plans. Changes in tax policyin the 1980s also reduced advantages of nonprofit hospitals over for-profits inthe issuance of tax-exempt debt (see Chapter 8 and Sloan 1988).

Simultaneously, the nature of competition in the health-care marketplace hasbeen changing. Research and medical care for the indigent (and, to a lesser de-gree, training of medical students and residents) are difficult to finance throughdirect sales. Historically, they have been financed out of a combination of gov-ernment grants and profits from government and private insurance payments formedical care, payments that exceeded the costs of care and left room for cross-subsidization. However, an era in which third-party payers passively reimbursedhospitals for their self-reported costs or charges has given way to one of stan-dardized payment rates (led by Medicare's adoption of pricing based on Diag-nosis Related Groups [DRGs]) and even active price competition.

The health-care financing system and government policies of the past (in-cluding the Hill-Burton Act of 1946, which subsidized construction of hospi-tal beds) encouraged an enormous buildup of hospital capacity in the post-World War II period. As hospital admissions began to fall in the early 1980s(due to technological changes but also to DRG-based payments and the growthof managed care), average occupancy levels fell and have remained low, cur-rently around 65 percent.

The interest of nonprofit health-care organizations in conversion coincideswith a drive toward consolidation in the industry.3 Some analysts have stresseda link between excess capacity in health care, particularly in hospitals, and bothof these movements, conversion and consolidation. As the market for hospitalservices became more price competitive and the level of overcapacity more ap-parent, hospitals often came to look to affiliation with larger organizations, of-ten for-profits, as vital (Coye 1997; Hollis 1997).

2 The March-April 1997 issue of Health Affairs, devoted to hospital and health plan conversions,is a good source of additional discussion.

3 Consolidation can bring social benefits, to the extent that it creates efficiencies through the real-ization of economies of scale and elimination of excess capacity. Consolidation in health care,however, is surely also motivated by a desire for greater market power, which benefits organi-zations that achieve it at a cost to the rest of society (Fuchs 1997).


The debate over conversion of hospitals to for-proflts is often entwined withone over local versus more distant control (Langley and Sharpe 1996). One mo-tivation for preferring to affiliate with for-profits is the perception that a for-profit company will find it easier to take bold actions necessary for the organi-zation's survival, such as layoffs or elimination of services - a perception thatat least one former hospital executive claims is "always present" in nonprofithospital conversions (Coye 1997). This may equally be a basis for local oppo-sition to such a conversion by employees and other input suppliers.

As government and employers have sought to become more prudent pur-chasers of health care, health insurers similarly have needed to compete morefor enrollees on the basis of price and to embrace managed-care principles. Inaddition, access to capital has come to be seen as critical to survival for healthinsurers, and as favoring for-profit firms over nonprofits, which cannot sell eq-uity interests.4

Conversions in health care may thus be a natural adaptation to changes ingovernment policies and in the market environment wherein the organizationsoperate, changes that have been more pronounced there than in other industriesin which nonprofits are significant players. Such conversions thus appear to il-lustrate the revenue-substitution process - presented abstractly in Chapter 3 andexamined quantitatively in Chapter 6 - carried to an extreme. At the same time,as we discuss in a later section, it seems that financial enrichment of key non-profit decision makers has often been an incentive for the conversion of non-profit HMOs, and in at least some proposed conversions of Blue Cross plans.The possibility that it is private financial gain, through subversion of the nondis-tribution constraint, that primarily has motivated health-care conversions is dif-ficult to rule out.

Public policy and conversions

In this section we examine the three related public-policy questions identifiedat the outset of this chapter:

1 Under what circumstances is a conversion of a nonprofit firm to for-profit status socially appropriate?

2 How should the nonprofit's assets be valued?

4 It is interesting to ask why access to equity capital would have increased in importance, andwhether this situation is temporary or permanent. The health-insurance industry is apparently un-dergoing a transitional period of high capital investment, in information technology and networkbuilding, associated with the adoption of more aggressive managed-care techniques. Part of theinterest in access to equity capital is for financing acquisitions, as the health-insurance industryconsolidates. Either of these motives suggests that if for-profits have an advantage in this regard,its importance is temporary. Capital investment would be a stable share of revenues for a "ma-ture" insurer that is not expanding rapidly, and could be financed largely with internal funds.


3 What should happen to the financial assets remaining after the conver-sion?

Public policy ideally should be structured so that conversions of nonprofitfirms to for-profit status occur when, and only when, they advance "social wel-fare." Defining and operationalizing this rule, however, is a significant chal-lenge. Addressing question 1 is an essential first step to thinking clearly aboutpolicy. If conversions can improve society's welfare, we should be concernednot only with preventing conversions that are inappropriate or on the wrongterms, but with creating an environment that does not unduly hinder the bene-ficial ones. Questions 2 and 3 are thus closely related to each other and to ques-tion 1. Whether a conversion is deemed appropriate depends upon all the termsof the deal, including the valuation of the assets to be transferred and what hap-pens to those assets and to the nonprofit's other obligations and responsibilities.How questions 2 and 3 are answered in practice also affects the incentives oforganizations to seek conversions.

As we have noted, we shall focus on conversions in which the assets of anonprofit entity are substantially shifted to for-profit control. Nonetheless, weare mindful that the more restrictive public policy is toward complete conver-sions, the greater will be the incentives to shift control of assets in more com-plex ways involving "partial" conversions.

When is a conversion socially appropriate?

This question involves important issues of both economic efficiency and distri-butional equity. With regard to efficiency, the concern of conventional benefit-cost analysis, we may say that a conversion is appropriate if the assets will pro-duce greater social value under a for-profit form of organization. Social valueis ordinarily conceived of as the aggregate of what members of society are will-ing to pay for what the organization produces (Haveman and Weisbrod 1975).

It is difficult to argue that conversion of a nonprofit firm to for-profit can nev-er advance social welfare. Because of the stronger incentives for minimizingcosts and responding to consumer demands that come with private ownershipof assets, as well as possibly greater flexibility associated with access to equi-ty capital, the same resources could plausibly produce greater value in the for-profit than in the nonprofit sector. Were it not so, we must question why the for-profit is the primary form of organization in most industries. Even if we acceptthat privileges associated with nonprofit status were originally granted to ad-vance some social objective, in particular cases that decision may come to beviewed by public-policy makers as an error. Circumstances may also change,and the evolution of market forces may make it increasingly difficult for non-profits to compete. There may also be a social decision to use different means


to promote social objectives once attended to by nonprofit firms, calling intoquestion their reason for existence. For example, New York State now requiresother insurers to share Empire Blue Cross and Blue Shield's past responsibil-ity as "insurer of last resort," while at the same time it has abolished subsidiesthat helped Empire support unprofitable business. Empire is now seeking toconvert to for-profit (Freudenheim 1996; Rosenthal 1996a).

Conversions of nonprofits bear some resemblance to corporate takeoverswithin the for-profit sector. It is frequently argued that such takeovers must beefficiency enhancing or they would not occur. The argument in brief is that ifone firm or management group is able to offer enough for the assets of anoth-er to wrest control, it must be because the new management will do a better jobwith those assets, will deploy them in a way that increases the value that theyproduce. This argument is appealing because in many circumstances within thefor-profit sector, activities that increase private market value also produce so-cial gains. This presumption - immortalized in Adam Smith's metaphor of an"invisible hand" guiding resource allocation efficiently - is strongest for com-petitive markets, with firms operating under the same sets of rules, producingordinary private goods.5

We should be cautious, however, about using market value as a guide to so-cial value when we compare institutions that operate under different legal con-straints (see Chapter 5). Even within the for-profit sector, whenever firms oper-ate under different rules, comparing market values of assets can be misleading.A case in point is the U.S. corporate income tax, which falls differentially onreturns to capital employed in the corporate sector. As Harberger (1974) dem-onstrated long ago, such a tax leads to an equilibrium where similar assets earnthe same private return (have the same market value) everywhere but have high-er social value in the more heavily taxed sector.

An organization's desire to convert from a less heavily taxed (nonprofit) toa more heavily taxed (for-profit) form creates at least some presumption thatthe conversion would enhance efficiency because the organization would takeon added tax liabilities by converting. An even more fundamental issue, how-ever, is that some outputs have social value, yet are not profitably provided byprivate enterprise. Health care for the poor and basic medical research are exam-ples. Relying on for-profit firms exclusively, without public subsidies, will leadto inefficiently low provision of these collective - or "public" - goods due to

5 In some circumstances this premise is questionable even for takeovers involving only for-profitentities. Suppose, for example, that a corporate raider is able to increase his or her target profit-ability by taking away promised health-care benefits from retired workers and thereby loweringcosts, which the old management was unwilling to do. This reduction in welfare of the retireesmay lead to a higher market value for the firm, but in a proper benefit-cost analysis the cost toretirees should also be taken into account. For an interesting discussion, see Krugman (1994,chap. 13).


familiar free-rider problems. Nonprofit organizations may provide more of thesegoods than for-profits operating in similar markets, although the evidence isnot entirely clear, and there is some indication of lack of difference (see Chap-ter 8 regarding hospitals). Public policy toward conversions of nonprofits shouldnot overlook the impact on provision of collective goods.6

On what terms should conversions be permitted?

What should determine the price paid by a for-profit firm to gain control of anonprofit's resources? An obvious suggestion is a process of competitive bid-ding. There is surely merit in this idea, but such a process can be costly, timeconsuming (Schaeffer 1996), and surprisingly complex. Consider the ambigu-ity of precisely what is to be transferred from the nonprofit to the for-profit. Asale would naturally include the items that appear on the organization's balancesheet, its assets and financial liabilities - but what of the organization's formalor informal "obligations" to provide collective goods? Suppose a nonprofit hos-pital that has been providing some charity care for poor and uninsured patientsis considering sale to a for-profit hospital chain. Would the sale, if it took place,require the new owner to provide such charity care, and at the same level?

If competitive bidding is used, with a requirement that the highest bid be ac-cepted, it is very important to define terms carefully and to make sure that of-fers are truly comparable. This is one argument for preferring "complete" to"partial" conversions. Different suitors will undoubtedly prefer different termsfor a joint-venture deal, making offers difficult to compare. Columbia/HCAhas,for example, participated in joint ventures in which it receives in addition to itsownership share a management fee. Clearly, a higher management fee increasesthe price it would willingly pay for its share. Even in the context of a completeconversion or outright sale, difficulties of specifying all the relevant terms ofthe contract, and in enforceable terms, are considerable. Conversion agreementshave frequently specified that the for-profit buyer will maintain the level ofcharity care formerly provided by the nonprofit hospital (Claxton et al. 1997),but because charity care is notoriously difficult to define, such provisions aredifficult to monitor and enforce.

Another reason for not necessarily accepting the highest bid is the social in-terest in maintaining competitive pricing in product markets. In a community

6 Although there has been little research on the consequences of conversions in any industry, thereis considerable literature comparing empirically the behavior of nonprofit and for-profit organi-zations in industries in which they coexist. Among the industries studied are day care (Knapp1989; Badelt and Weiss 1990; Krashinsky 1993; Mauser 1993), university research (Blumenthalet al. 1996), nursing homes (Weisbrod and Schlesinger 1986; Weisbrod 1988; Hirth 1997), long-term psychiatric care facilities (Schlesinger and Dorwart 1984), and hospices (Christakis and Es-carce 1996).


with just two hospitals, hospital A may be in a position to offer the highest bidfor hospital B, because the combined hospitals would be able to engage in mo-nopoly pricing, an advantage that other potential buyers of hospital B would nothave. Hospital A's advantage does not come from an ability to use resourcesmore efficiently, but from being better positioned to extract payment from con-sumers. In this case, antitrust laws might rule out hospital A as a potential buy-er, but the issue exists in less extreme situations. The point is that the highestbid is not the socially most efficient bid if it results from an expected exploita-tion of monopoly power.

A key question for public policy is whether the postconversion provisionof collective goods, such as charity care or medical research in the case of ahospital, is best handled through contractual agreements with the acquiring for-profit, or purchased using the assets transferred to a successor foundation. Thelatter approach appears to have certain advantages. Contractual mandates toprovide such goods are likely to lead to extensive legal wrangling over precisedefinitions and measurement. Profit-oriented firms seek to minimize activitiesthat generate financial losses. A for-profit hospital has an incentive to definecharity care, for example, in such a way as to meet its legal obligations whiledeparting as little as possible from what would otherwise maximize profits. Notobligating for-profits to provide collective goods should maximize bids, andmay make competing bids easier to compare.

A conversion should proceed only if the social value of the resources is high-er in the for-profit sector. An open and competitive bidding process should elic-it bids that are close to actual private valuations. Even the highest bid, howev-er, is likely to be a lower-bound estimate of social value, for it would not reflectthe social value of unprofitable collective goods such as indigent care. If a pri-vate purchaser were required to provide such services, the costs of providingthem would be negatively reflected in its bid, whereas the positive social valueof the services would not be. Moreover, expected tax payments would detractfrom the bid of a private for-profit bidder, even though, from a social perspec-tive, tax payments have positive value.

To approximate social value in the private for-profit sector, we should addto private bids the present value of future tax liabilities, as well as an adjust-ment for the social value of community obligations. A private bidder would sub-tract losses on obligatory community services - that is, their production costminus any associated receipts - before identifying its maximum bid. We couldtherefore add to private bids a positive amount equal to expected losses due tocommunity obligations. In particular cases this adjustment could be too largeor too small. It would be too large if the community-service obligation was it-self inefficient - that is, if the required services had a social value below themarginal cost of providing them. On the other hand, it would be too small if thesocial value exceeded the cost of production. This is reasonably clear concep-


tually, but estimating the marginal social values of the required unprofitableservices is typically extremely difficult in practice.

Even more difficult is evaluating the social value of the firm in its nonprofitform. Theoretically it is the present value to society of the services that thoseassets would deliver over the remainder of their life, if they remain under non-profit control, less any net subsidies the firm would use in delivering those ser-vices. Private-market valuations of the nonprofit's assets, either in total, as elic-ited through bids to purchase the firm and convert it to for-profit, or piecemeal,as obtained through appraisals of the value of individual tangible assets, are notdirectly relevant. These reflect the value of the assets if they were deployed inthe for-profit sector, where they can be expected to provide different kinds ofservices or to different beneficiaries. Bids from nonprofit suitors provide at leastsome information about the firm's value in that sector. For comparison with bidsfrom for-profits, these should be reduced by the present value of net subsidiesthat the nonprofit could expect to receive (just as the for-profit bids should beincreased by the value of tax liabilities). Even rough valuations of the expect-ed flow of collective goods, if the organization retained its nonprofit form,would be useful for comparison with the value in the for-profit sector.

In addition to attempting to make such difficult comparisons of value direct-ly, public policy should also concern itself with the process by which conver-sions are proposed. A necessary, though generally not sufficient, condition fora conversion to occur is that the nonprofit's leadership desires it. These deci-sion makers are arguably best informed about what their organization can ac-complish as a nonprofit and what would be lost if the nonprofit was to disap-pear or be replaced by a private firm. Ideally, they would not seek a conversionunless they saw it as the best way to advance the social objectives that the non-profit was established to serve.7 They should, from a social-efficiency perspec-tive, compare the financial resources that would be realized from a sale (andwhich could then be devoted to charitable purposes) with the value of what theorganization could accomplish if it remained nonprofit. If the objectives ofthese decision makers, as they consider conversion and compare alternatives,are well aligned with social goals, then the invisible hand may yet work reason-ably well. Unfortunately, nonprofit board members and top executives do notalways have the knowledge and motivation to carry out their public responsibil-ities effectively.

7 We should not suppose that decision makers have perfect information about the firm activities.In a recent case of alleged serious fiscal mismanagement of a nonprofit hospital, one trustee admit-ted, "I'm a layman - what do I know?" A former official of the hospital added, "The board waslike a bunch of sheep.... they would raise their hands yes to everything Ward [the CEO] asked"(Langley 1996).


If the nonprofit decision makers are to act in their legal fiduciary capacityas agents for the public in promoting only efficient conversions, their decisionsshould not be biased by personal financial gain. The possibility of moving to ajob with the new for-profit at higher pay does not necessarily bias an individ-ual's decision, if the rate of pay in the for-profit is market based; but if a man-ager or trustee is rewarded significantly for facilitating a transfer of control andis also involved in deciding its terms, private reward may improperly influencethe decision. For facilitating a conversion, officials of a nonprofit may be re-warded with a job with the purchasing firm at a wage far exceeding market op-portunities - or more subtly, at a seemingly competitive wage but with various"golden parachute" provisions. There could also be lucrative consulting ar-rangements, or payments for agreements not to compete, as in the Columbia/HCA offer for Blue Cross and Blue Shield of Ohio.

It is also objectionable for officials of the original nonprofit to buy it them-selves - as happened with Health Net - at least in the absence of a workablecompetitive-bidding process. The temptation to undervalue the nonprofit inthese circumstances is simply too great. In many HMO conversions insiderswere allegedly able to purchase ownership shares at low prices before the con-verted organizations were valued in the stock market; they then profited enor-mously when the organizations "went public" and a true market value was es-tablished. Hamburger et al. (1995) claim that in California in the early 1980s,directors and managers of nonprofit HMOs that converted to for-profits becamemillionaires "overnight" as a result of capital gains on stock in the new com-panies (see also Fox and Isenberg 1996). Leonard Schaeffer (who as CEO ofCalifornia Blue Cross was himself the engineer of its conversion to WellPointHealth Networks) stated the point emphatically in a published interview (Igle-hart 1995):

Before the conversion of WellPoint, the value of every single company that convertedto for-profit status was significantly underestimated... .Almost all of the value createdwent to the management and boards of these companies. .. . FHP International, Foun-dation Health, Pacificare, TakeCare, you name it. These are companies that today areled by multimillionaires who achieved that status by virtue of receiving stock that wasdramatically undervalued at the time of conversion.

Some states have begun to pay close attention to potential conflicts of inter-est involved in conversion proposals (Allen 1997; Matzke 1997). The IRS is al-so taking a closer look. Marcus Owens, the director of its Exempt OrganizationDivision, has stated that sales, conversions, and joint ventures of hospitalswould be getting more attention, focused on "the level of private benefit thatmight be occurring in the transaction," among other things (Internal RevenueService 1997).

One possible response to potential conflicts of interest is to prohibit pay-ments of any kind - including compensation for employment - from a for-


profit firm that buys a nonprofit to any high official of the acquired nonprofit.This idea could even be extended to conversions of organizational form with-out an outright sale; the board and top managers might be required to resignand to refrain from any financial association with the new entity. A somewhatanalogous prohibition on senior federal officials' switching sectors already ex-ists: They may not be paid by firms with which they had dealt in their govern-mental roles, for varying periods of as long as five years.

Although we believe that the potential for conflict of interest is very real,such an absolute prohibition would probably be counterproductive. The pros-pect of a personal financial loss, and perhaps loss of prestige as well, might mo-tivate managers and board members to oppose socially desirable conversions.In particular situations, it would be efficient for the leaders of a nonprofit to playsimilar roles in (or at least serve as consultants to) the new for-profit. These of-ficials have knowledge and experience - specific human capital - concerningtheir organization's operations and its relationships with its customer base, thevalue of which could not be economically replicated by the firm.8

A final issue about the terms of conversion is whether the converted organi-zation should be allowed to use the name of the former nonprofit. The nationalBlue Cross and Blue Shield Association agreed to that possibility in 1994, butvoted to revoke the use of its trademark by BCBS of Ohio if its deal with Co-lumbia/HCA were completed. In markets where consumers have difficulty as-sessing product quality, a trusted name can be a very valuable asset; whetherthe name is transferred may therefore significantly affect the offer price. If thename is not transferred and is therefore no longer used, there would appear tobe a destruction of that value, which may seem inefficient. It is surely a loss ofprivate value for the for-profit purchaser, but is it a loss of social value? Not ifthe surviving organization is to act differently in important ways, and the useof the name of the predecessor might convey that there has been no change.

The social value of allowing the new firm to retain the name of the nonprof-it is probably far lower than its private value to the organization, and it may noteven be positive. Allowing the firm to retain the name has positive social val-ue insofar as the name conveys accurate information, reducing uncertainty forbuyers, and facilitates better matches between sellers and buyers (insurer andenrollees, or hospitals and patients). However, there is a negative social valueif the new firm derives benefits from the brand name by virtue of misleadingconsumers into having greater confidence than is warranted by the firm's be-havior.

8 Bradford Gray suggested to us the Empire Blue Cross case as one where it seems appropriate thatthe leaders of a nonprofit remain in place following a conversion of form. The conversion pro-posal appears to be well structured to avoid problems of conflict of interest.


What should happen to the assets?

Consideration of incentive effects has led us to conclude that conversion shouldnot be a vehicle for subverting the nondistribution constraint. Thus none of thevalue of the nonprofit's assets should be used for the financial benefit of individ-uals closely connected with the organization. All of the proceeds of a sale shouldbe transferred to an approved social purpose. For nonprofits organized for acharitable purpose, the law in most states and the common-law cy pres doctrinerequires that the assets be dedicated to purposes as close as possible to the orig-inal ones (Hamburger et al. 1995; Kane 1997). For example, a foundation creat-ed from the sale of a nonprofit hospital might be dedicated to financing care forthe indigent or promoting community health. Economic efficiency argues thatthey be directed to the use where they produce the most value, although identi-fying high-value uses for funds allocated to charity is difficult. Policymakersand regulators must prevent new foundations from becoming vehicles for avoid-ing the nondistribution constraint, with funds devoted to exorbitant salaries orother uses of little social value.9

Concerns that incentives for proposing conversions be efficient and that pub-lic assets not be unfairly appropriated for private gain both argue for separatecontrol of the new organization and the one that receives the assets of the for-mer nonprofit. If managers or board members of a purchasing firm also sit onthe board of the new foundation created with the proceeds of the sale, the foun-dation is more likely to act in the firm's interests. Even without direct overlapin control, if the foundation holds a major ownership share in the for-profit orga-nization (e.g., one of two foundations created in the California Blue Cross con-version was endowed with stock of the for-profit WellPoint), their fortunes re-main closely entwined. To separate them as much as possible, the proceeds ofa sale generally should be paid in cash rather than in equity or debt issued bythe purchasing firm or successor for-profit. If equity shares of the for-profit mustform part of the sale proceeds (because providing its full equity value in cashwould be too damaging to the for-profit in the short term), the foundation shouldagree to sell them within a reasonable time frame.

Concluding remarks

Nonprofit and for-profit organizations operate under different rules and con-straints - financial and other. Whenever individuals are free to choose underwhich set to operate, they will be influenced by the costs and benefits of thealternatives. Moreover, they may want to switch from one set of constraints to

9 Jaffe and Langley (1996) relate a story of a former board member of a small town hospital whobecame director of a foundation formed through its sale, at a salary of over $200,000 per year.


another if exogenous changes - in, say, tax policy or government grants - alterthe relative attractiveness of the respective constraints, or if changes within theorganization refocus its objectives.

Certain legal constraints, such as lower taxes and greater subsidies, favor theaccumulation of assets by nonprofit organizations: Many nonprofits have sub-stantial assets that were amassed through contributions, gifts, and grants as wellas untaxed income from the sale of program services and investments. Otherconstraints favor private firms, however, such as access to equity capital. For-profits are surely less restricted in the ways they can deploy assets and distrib-ute surplus revenues.

Conversions transfer assets from the more-restricted uses permitted for non-profits to the less-restricted private enterprise form. If the assets of nonprofit or-ganizations could be converted into a form that was not subject to restricted dis-tribution, the incentives for conversion among successful organizations wouldbe enormous - consider an art museum with collections worth hundreds ofmillions of dollars. Nonprofits would often be only transitional organizations,formed to take advantage of public and private subsidies and then eventuallyabandoned - converted to private firms - so that the assets could be used forprivate gain. Of course, if it became widely known that such transfers were per-mitted, any trust that consumers place in the public-spiritedness of nonprofitswould be severely undermined, which would in turn diminish donations and thesocial role that nonprofits can play.

Among the difficulties of determining whether a specific conversion is so-cially efficient rather than merely privately profitable are the following: Howdo we determine the social value of the nonprofit, in order to make comparisonsto private bids? How do we specify which assets, liabilities, and obligations noton the balance sheet are being sold, and ensure that those obligations are met?How do we prevent conflicts of interest that encourage managers and directorsof a nonprofit to undervalue its assets for private gain?

Conversions clearly hold the potential for abuse - for inefficient change thatserves little or no social goal but enriches insiders at public expense. Informa-tional problems seriously impede establishment of rules to guide regulators. In-siders are generally better informed than regulators about the magnitude anddistribution of both private and social benefits. We can be confident that agree-ments between the managements of a nonprofit and a for-profit will generateprivate benefits to the leaders and owners of both, but the overall social bene-fits are more doubtful. It does not follow, however, that conversions are inevit-ably inefficient. Changing circumstances can reduce the importance of the so-cial mission that first led to the development and growth of nonprofits in someindustry, and can make reallocating resources efficient. Important as entry ofnonprofits may be, so too is it important that there be mechanisms for facilitat-ing their exit.


The best public policy would establish rules that discourage nonprofit offi-cials from proposing conversions for the wrong reasons (private gain), yet donot unduly inhibit efficient redirection of resources into private control. Howhigh public policy sets the barriers - for instance, the stringency of restrictionson financial arrangements between the nonprofit's insiders and the purchasingfor-profit - should, as usual, reflect a balancing of marginal gains and losses.The potential losses from an overly restrictive policy - that is to say, social gainsforgone by efficient conversions denied - are not yet well understood; nor, forthat matter, are the potential losses from a permissive policy. A better under-standing of the motives for and consequences of current conversions in healthcare is a natural, though surely challenging, target for further research. Equal-ly important is applying that understanding to nonprofit organizations outsideof health care, so as to anticipate prospective pressures and be prepared to dealwith them. The potential for conversions to generate private benefits but socialcosts is great - and by no means limited to the health sector.

PART II

Industry studies

CHAPTER 8

Commercialism in nonprofit hospitals

Frank A. Sloan

Introduction

Most U.S. hospitals are organized as private, not-for-profit organizations. Al-though the share of such hospitals is declining, as of 1994,60 percent of nonfed-eral, short-term general hospitals were nonprofits; only 12 percent were propri-etary, and the rest were government hospitals (AHA 1995). Private nonprofithospitals, measured in terms of GNP, constitute by far the largest category ofnonprofit institutions in the United States.

The hospital industry is undergoing massive change. Partly because of mod-ifications in payment practices, demand for hospital inpatient care is shrinking,leading to hospital closures and mergers as well as some conversions to for-profit status. Hospitals are integrating not only horizontally but also vertically,the latter including combinations with other types of health-care providers and,mainly through contracting, the establishment of links to health-insurance plans.Entry of capitated health plans has introduced a dose of competition into thehealth-care sector, which in turn is adding to competitive pressures on hospi-tals and appears to have made them more commercial than heretofore.

The changes that are occurring raise several important public-policy con-cerns. The social expectation is that even persons who are disadvantaged be-cause of their health, low income, or other factors, such as race or ethnicity, haveaccess to needed care of high quality. Some hospitals have engaged in teachingand research, supported in part by profits from patient service and explicit sub-sidies. By virtue of their ownership status, it is also expected that at least somenonprofit hospitals will engage in these noncommercial activities. Important forpublic-policy purposes is the question whether nonprofit hospitals are becom-ing less committed to such "noncommercial" activities and whether these ac-tivities cease almost entirely when nonprofits convert to for-profits.

I wish to thank the participants in Northwestern University's colloquium on commercialism in non-profit organizations for their initial feedback.

151

152 Frank A. Sloan

Conversions of hospitals from nonprofit to for-profit status have attractedconsiderable attention of the media in the mid-1990s, as was discussed in Chap-ter 7. They seem to be important to the public even though conversions fromeither nonprofit or government to for-profit status constituted only about a quar-ter of the ownership changes occurring between 1990 and 1993 (Needleman,Chollet, and Lamphere 1997).

To illustrate media interest: When Baptist Hospital in Nashville, Tennessee,was considering a switch from nonprofit to for-profit status in early 1996, thehospital claimed that it would no longer contribute about $30 million a year infree care and other community benefits if it relinquished its nonprofit status; itwould, however, pay additional taxes. According to the hospital's administra-tor, "The question is, would this hospital function better if we didn't have thisadditional $15 million to $19 million burden?" (Bell 1996). Judging from thearticle's tone, readers were to answer "no." The potential benefit of the extratax revenues to the various communities the hospital serves was not mentioned.

To the extent that increasing commercialism in the hospital sector is judgedto be undesirable, public-policy makers have a limited number of instrumentsfor taking corrective action. States or communities can oppose nonprofit conver-sion. In most states, the attorney general's office can provide oversight of trans-actions involving transfer of a substantial share of assets from a nonprofit toa for-profit organization; actual authority of attorneys general varies widelyamong states (Butler 1997; Hollis 1997). Another instrument is certificate ofneed (CON); but since the mid-1980s, states have had the option of droppingCON, and some, especially those with relatively competitive health-care mar-kets, have done so (Delaware Health Care Commission 1996). Moreover, CONonly reviews entry, changes in facilities and services, and new investment thatmay accompany reorganization. Antitrust agencies may oppose mergers or ac-quisitions on grounds that they are anticompetitive, although failure to performnoncommercial activities is not a violation of antitrust law. Except for remov-ing tax-exempt status or refusing to issue tax-exempt securities on its behalf,there is little government can explicitly do to change the behavior of a nonprof-it (Morrisey, Wedig, and Hassan 1996). Removal of tax-exempt status is polit-ically risky and hardly, if ever, occurs.

This chapter addresses two fundamental issues. First, what are the forces thatare pushing nonprofit hospitals into sameness with their for-profit competitors?Second, to the extent that nonprofit hospitals are eliminated and/or forced intosameness by market forces and public budget cuts, what will have been lost?Should public-policy measures be developed to slow or even reverse the trendin conversions to for-profit status that seems to be evolving?

To add focus to the analysis, the chapter excludes consideration of teachinghospitals. These institutions are very complex, and the public-policy issues re-lating to medical education and unsponsored clinical research are particular to

Commercialism in nonprofit hospitals 153

these organizations. A very small proportion of hospitals engage in much edu-cational research. Also, this chapter excludes consideration of some activitiesthat, if performed, may create positive externalities, such as contracting withessential community providers and reporting bad clinical practices to appro-priate authorities (Gray 1997), or health promotion and screening initiatives(Young, Desai, and Lukas 1997). Such benefits are excluded from this study notto downplay their potential importance, but rather because there is no empir-ical evidence that nonprofits perform differently in these cases.

The next section presents reasons various authors have suggested for thedominance of the nonprofit form in the hospital sector in this century. The fol-lowing one further describes changes that have occurred in U.S. health care dur-ing the past decade and their impact on the hospital industry. The ensuing sec-tions discuss (respectively) the new empirical evidence on profitability, pricingand cost shifting, uncompensated care, quality and innovation, capital funds,and investment. These sections emphasize recent work; descriptions of earlierempirical research can be found elsewhere (see, e.g., Institute of Medicine 1986;Sloan 1988). A final section offers discussion, conclusions, and implications forpublic policy.

Explanations of why the nonprofit hospital is dominant

Several explanations have been offered for the dominance of the nonprofit formin the U.S. hospital sector. Kenneth Arrow (1963) explained this dominance asa response to uncertainty and incomplete markets for risk in markets for med-ical care. Nonprofits, because they presumably are not merely profit seekers,would not fully exploit their market power. Thus, price would be lower and out-put higher than in the for-profit case. Quality of care might also be higher. Fur-ther, rather than being paid to shareholders, profits may be distributed in theform of "free" care, and cross-subsidize unprofitable care, unsponsored re-search, and medical education. The nonprofit form is more consistent with thefiduciary relationship between patients and providers, an extension of Arrow'sargument, and between donors and recipients of donations. Implicit subsidiesin the form of internal transfers of funds are more likely to occur in nonprofitorganizations. (Although explicit subsidies may be preferred, society cannot betrusted to provide a socially optimal level of such subsidies; hence, it is betterto hide them.) Even though the notion of local control is conceptually distinctfrom ownership form, in practice they are often related since the choice may bebetween a for-profit hospital company operated from a distant site and a free-standing local nonprofit hospital.

Other explanations are less charitable to nonprofits. The nonprofit form facil-itates cartelization by physicians in the community (Pauly and Redisch 1973).Physicians may find it easier to divide profits of a nonprofit than a for-profit

154 Frank A. Sloan

corporation: In the latter case, physicians have to share profits with equity hold-ers, and some profits are distributed to the community as tax payments (Clark1980). A possibility suggested by Hansmann (1996) is that the nonprofit formhas become increasingly anachronistic and is protected by entry regulationstrongly biased against for-profits.

Nonprofit hospitals have consistently earned profits (see issues of the Amer-ican Hospital Association's Hospital Statistics); their "nonprofit" status refersto the way that hospital profits are used, not to the absence thereof. Accordingto Hansmann's (1980) classification, nonprofit hospitals are "entrepreneurialnonprofits."

One justification for the high degree of similarity between for-profit and non-profit hospitals has been that competitive pressures from for-profits have forcedthe nonprofits to act as profit seekers. Without the pressure from profit-seekingcompetitors, it has been argued that nonprofits would have acted in a more so-cially responsible manner.

Major changes in the U.S. health system since the mid-1980s

Two major changes in the U.S. health delivery system since about the mid-1980s have had and are having major impacts on the hospital sector, irrespec-tive of ownership form: implementation of Medicare's Prospective PaymentSystem and the growth of competitive medical plans.

Implementation of Medicare Prospective Payment System

Before the 1980s, hospitals were paid on a retrospective cost or charge basis.Under retrospective cost, the share of hospital cost attributable to services pro-vided to insured patients was recovered after services were delivered. Hospi-tals with higher cost were paid more (see, e.g., Sloan and Steinwald 1980). Anexplicit payment to profit was sometimes made as a surcharge over cost. Un-der retrospective charge reimbursement, hospitals were reimbursed a percent-age of billed charges; the remainder of the charge was collectable from patients.Generally, a hospital had both cost- and charge-paying patients. Depending onhow the programs were structured, hospitals had little incentive to be efficient.

Various prospective payment schemes were designed to encourage efficien-cy by fixing the payment per unit of output however defined, in advance of pro-vision. The Medicare Prospective Payment System (PPS), first implemented in1983, affected more hospitals than any of the prospective payment programsoperated by individual states. PPS set the payment as a fixed price per discharge,with price varying by Diagnosis Related Group (DRG). (Outpatient cost contin-ues to be reimbursed by Medicare on a retrospective basis.) In recent years,Medicare has accounted for around one-third of hospital revenue; yet this shareof revenue sufficed to produce dramatic effects on treatment patterns of patients


more generally and on overall hospital operations (see, e.g., Antel, Ohsfeldt, andBecker 1995).

Conceptually, a program like PPS should have the following impacts:

1 Under retrospective cost reimbursement, a dollar saved by the hospi-tal meant a fraction of a dollar lost in revenue, where the fraction de-pended on the share of the hospital's patients with such insurance. Bycontrast, under PPS, a dollar saved by the hospital results in a fractionof a dollar extra profit; thus, hospitals have a greater incentive to beefficient. Prior to PPS, profits could be earned by manipulating hos-pital accounts (Danzon 1982) rather than by gaining efficiency.

2 Absent competition from other hospitals, PPS may give an incentiveto cut quality. Still, with nonprice competition on various dimensionsof quality, quality may rise until profits are eliminated (Pope 1989). Infact, quality of care appears to have risen after PPS was implemented(Keeler et al. 1992). If quality competition is market driven rather thandetermined by professional norms, there is no reason to expect differ-ential behavior by for-profits and nonprofits.

3 It might appear that profits would rise under PPS, but this depends onthe extent to which nonprice competition erodes quality and how wellMedicare price updates keep up with increases in factor prices paidby hospitals.

4 As demand for inpatient care is reduced, hospitals may expand out-puts for which reimbursement is unchanged, such as ambulatory care,nursing-home care, home health, and hospice care. Various hospitalservices besides acute hospital inpatient care did indeed expand rapid-ly as the inpatient sector shrunk (Delaware Health Care Commission1996), but this incentive was faced by hospitals of all ownership types.

Growth of competitive medical plans

The second important trend is the growth of competitive medical plans (CMPs).These include health maintenance organizations (HMOs) and preferred provid-er organizations (PPOs).

Growth of CMPs has two important impacts on hospitals in general, plus athird development that is particularly important for nonprofit hospitals:

1 Demand for inpatient care declines dramatically, reinforcing the down-ward pressure on demand from PPS. In relatively mature CMP mar-kets, rates of hospitalization are far below those where indemnity in-surance prevails (see, e.g., Sokolov 1995) and have fallen more inmarkets dominated by CMPs, such as in California (Robinson 1996).This has resulted in excess hospital capacity as well as hospital merg-ers and closures. Interestingly, between 1983 and 1993, the for-profit

156 Frank A. Sloan

sector in California shrank at a faster rate than did the nonprofit sec-tor. At the same time, there were some conversions of nonprofit hos-pitals to membership in for-profit hospital systems.

2 The growth of HMOs in particular has increased vertical integrationof inpatient care with other sites of care (see, e.g., Shortell et al. 1996).Beginning with California's selective contracting law (Zwanziger andMelnick 1988), which is now commonplace among states, insurerscould direct patients to hospitals with which they had contractual rela-tionships. Typically, hospitals are not owned by HMOs, but rather arepaid a per-diem rate, fee-for-service, often at a discount, or a paymentper case (ProPAC 1995, 79; Robinson 1995). Although quality maybe a factor in HMOs' choices, price unquestionably is a factor. Thus,competition among hospitals has been transformed from exclusivelynonprice competition (Robinson and Luft 1987) to competition onprice and quality.

3 Enrollments in for-profit HMOs have far surpassed those in nonprofitHMOs in the 1990s (VHA and Deloitte & Touche 1996,14). Recent-ly, several Blue Cross and Blue Shield plans, which traditionally wereall nonprofits, have converted to for-profit organizations. (For an anal-ysis of the public-policy implications of such conversions, see Chap-ter 7.) The main justification offered by insurers in public testimonybefore regulatory agencies is the need to raise equity capital. Since thehospital sector is shrinking, hospitals do not face this same motive toconvert to for-profit status, although by consolidating, hospitals dorealize internal efficiencies (Lynk 1995) and may be able to borrowfunds at more favorable rates (Grossman et al. 1993). It would seemplausible that a for-profit HMO would expect the organizations fromwhich it purchases care to behave like a for-profit too.

Exogenous changes and the future shape of the nonprofit-hospital sector

Economic research on hospital behavior can shed some light on how nonprofithospitals have and are likely to respond to important changes in health-care pur-chasing practices. Also, such research can reveal whether the nonprofit hospi-tal, as it currently operates, is worth preserving as an organizational form. Thisliterature review emphasizes research published during the past decade.

Hospital profitability

Overall, profitability of hospitals, irrespective of ownership, as measured by to-tal margin (total revenue minus total expense as a percentage of total revenue),


1984 1986 1990

Figure 8.1. Trends in hospital total margins: NFP: not-for-profit; FP, for-profit;ALL includes public hospitals. (Source: Prospective Payment Assessment Com-mission 1995,56)

Table 8.1. Distribution of total margins by hospital ownership, 1993

Hospitalgroup

NonprofitsFor-profitsAll

Percentile

25th

0.3(1.2)0.1

Median

3.64.93.7

75th

6.810.67.4

Percent withneg. margin

22.229.023.8

declined from levels attained immediately after implementation of PPS but hasremained fairly constant since then (Figure 8.1). Nonprofit hospitals had posi-tive total margins in all years, and in some had higher margins than for-profithospitals. By the early 1990s, for-profit margins had climbed above nonprofitmargins, and the differential appeared to be widening; even at its greatest ex-tent, however, it was slightly more than two percentage points.

Although for-profit mean and median profitability was higher than for non-profits, there was more variation in hospital profits for for-profits than for non-profits in 1993 (Table 8.1). In that year, 29 percent of for-profit hospitals hadnegative total margins versus 22 percent for nonprofits.

This descriptive evidence is consistent with Hoerger's (1991) analysis. Heconducted two tests of the hypothesis that nonprofit hospitals behave differ-ently from for-profit hospitals. First, a profit-variability test determined whether

158 Frank A. Sloan

profits of nonprofits exhibited less intertemporal variation than did profits of for-profit hospitals, as he expected. This would occur if nonprofits maximize util-ity subject to a profit constraint, as in Newhouse's (1970) model, or maximizeincome per physician on the medical staff, subject to a zero-profit constraintfor the hospital, as in Pauly and Redisch's (1973) model. Second, he examinedwhether nonprofit profits are, as he supposed, less responsive to changes in suchexogenous factors as Medicare payment levels. Both his hypotheses were sup-ported by his empirical analysis.

From this evidence, it appears that, for those institutions that survive, non-profit hospitals will on average do what it takes to earn a nonnegative total mar-gin, although for-profit hospitals do not like negative margins either. In the pro-cess of maximizing profits, for-profits may undertake riskier activities.

One logical way to maintain profitability is to shift service mix away fromunprofitable and toward profitable services. Unfortunately, direct national em-pirical evidence on changes in service mix by ownership form is not available.Even if it were, one would additionally need to know the relative profitabilityof specific hospital services.

Some empirical studies have investigated diffusion of technology in hospi-tals; unfortunately, well-controlled empirical evidence is mostly available onlyfor the pre-PPS period. A large variety of innovations have been examined, in-cluding process and (more often) product innovations (Romeo, Wagner, and Lee1984; Sloan, Valvona, and Perrin 1986; Sloan, Whetten-Goldstein, and Wilson1997). With rare exceptions, there were no statistically significant differencesin adoption rates between nonprofit and for-profit hospitals. Both nonprofit andother hospitals face pressure to shift to more profitable services.

Waters (1992) assessed profitability of cardiac catheterizations and coronary-bypass surgery units in Florida during the 1980s. Particularly noteworthy werethe procedures' contribution margins - additional to total hospital profits attrib-utable to the service - which indicate that a few procedures may account formuch of a hospital's total profitability. The best data on profitability were forcardiac catheterization units; these came both from Florida's Health Care CostContainment Board (HCCCB) and from certificate-of-need (CON) applicationsfiled by individual Florida hospitals. The mean contribution per patient to totalhospital profit was over $1,200 (1990$). In 1989, the profit projected for thesecond year of operation in CON applications filed by hospitals was $420,000(1990$). Not surprisingly, 104 CON applications were submitted by hospitalsduring the 1980s requesting permission to open cardiac catheterization units.Despite the efforts to constrain the growth of capacity, the number of cardiaccatheterizations performed in Florida increased fivefold during the 1980s.

Unfortunately, HCCCB did not collect comparable data on service-specificrevenue and cost for the coronary-bypass surgery units. However, judging fromCON applications, the margin was over three times that for catheterizations.


This case study illustrates the resourcefulness of hospitals. There are otherexamples: For instance, because PPS did not cover mental health and rehabil-itation, both nonprofit and for-profit hospitals made great efforts to expand intothese areas. Those hospitals that will survive hospital downsizing may be ex-pected to develop new, more profitable product lines as the need arises.

Hospital pricing patterns: Cost shifting in the face ofdemand shifts?

Various health-care experts have argued that when public agencies reduce theamounts they pay for inpatient care, hospitals respond by raising prices to pri-vately insured patients. This behavior is called cost shifting (see, e.g., Morrisey1994 for further discussion), although charge shifting may more accurately de-scribe this behavior (Phelps 1986). Prerequisites for cost shifting are (1) suffi-cient market power (i.e., hospital ability to set price) and (2) unexploited mar-ket power - that is, the hospital cannot be profit maximizing before the pricedecrease by the public payer. Since for-profit hospitals plausibly are profit max-imizers, there is a strong conceptual argument that they do not shift cost.

Cost shifting and other forms of non-profit-maximizing behavior of nonprof-its can be analyzed within the following framework. Analysts have often spec-ified a utility function for the nonprofit hospital with profit and something else(X) as arguments (Pauly 1987). Profits earned by the nonprofit can be spent inpursuit of objectives of the hospital's management. The X may be quantity orquality of service, or uncompensated care. Because the nonprofit hospital likesX, it does not operate at the profit-maximizing level. However, if it suffers a de-crease in product demand or an increase in factor prices, it may move towardthe profit-maximizing price by reducing levels of X. To the extent that such costshifting occurs in response to various pressures, one could say that these hospi-tals are becoming more commercial.

Cost shifting is often confused with two-part price discrimination. Underprice discrimination, a decrease in the price paid by a public payer would re-duce the price charged in the market where the hospital sets price. Under costshifting, by contrast, such a decrease would raise the private price. As notedabove, to shift cost, the hospital must have market power that was unexploitedprior to the negative shock. Under price discrimination, it is only necessary tohave market power.

Whether or not hospitals shift cost cannot be deduced, but rather must bedecided by the empirical evidence. To the extent that CMPs reduce hospitals'market power, one would expect cost shifting to be declining, if in fact it everexisted. To noneconomist experts on health, however, the issue of nonprofit (andperhaps even for-profit) cost shifting is beyond question: They have no doubtthat it occurs.

160 Frank A. Sloan

Table 8.2. Hospital payment-cost ratios forMedicare, Medicaid, and private payers,1980-93

Year

19801981198219831984198519861987198819891990199119921993

Medicare

0.960.970.960.970.981.011.010.980.940.910.890.880.890.89

Medicaid

0.910.930.910.920.880.900.880.830.800.760.800.820.910.93

Private

1.121.121.141.161.161.161.161.201.221.221.271.301.301.29

Viewing the time series of payment-cost ratios (Table 8.2), it would appearat first glance that hospitals do in fact shift cost. Through the 1980s there wasa secular decline in the amounts paid by Medicare and Medicaid relative to cost,and a simultaneous secular increase in the ratio of payment from private sourcesto cost. Such evidence, however, is not conclusive. First, the information is forall hospitals, and there is no reason to believe that a for-profit hospital wouldengage in cost shifting. Second, and potentially more important, changes in rev-enue relative to cost may arise because of shifts to more profitable products rath-er than from price increases. Third, in the early 1990s, the payment-cost ratiofor Medicaid patients rose appreciably, whereas the ratio for the private payersfell very slightly.

Some of the more rigorous empirical studies of cost shifting have found em-pirical evidence that hospitals do shift cost (see, e.g., Dranove 1988); yet Mor-risey's (1994) very comprehensive review of the empirical literature conclud-ed that the evidence overall suggests very little cost shifting occurs. In a sense,this controversy is more of historical than of current interest. Even if hospitalscould have cost-shifted at one time, there must be a limit to their unexploitedmarket power. Also, with CMP growth and increased hospital dependence onpublic payers, it is likely that they will also no longer have sufficient marketpower - and both are needed for cost shifting to occur.


Uncompensated care

One public-policy concern is that an increase in commercial practices of non-profit hospitals will place the poor and the uninsured at an even greater disad-vantage than previously (Mann et al. 1995; see also Chapter 4). However, dif-ferences between for-profits and nonprofits in provision of uncompensated careand in provision of care to publically sponsored patients were very small, evenbefore implementation of PPS and the growth of managed care.

Sloan et al. (1986), using national data on hospitals for 1978-83, assessedamounts of charity care and bad-debt care provided by community hospitalsof various ownership types. Charity care was defined as care rendered to pa-tients whom the hospital adjudged unable to pay; bad-debt care was the valueof care delivered to patients adjudged by the hospital as able to pay but whodid not do so. For hospitals located in urban areas in 1982, Sloan and coauthorscomputed bad-debt care plus charity care as a percentage of hospital charges asfollows: 3.7 percent for nonprofits; 3.0 for for-profits; and 8.6 for public hos-pitals. For rural hospitals, the corresponding numbers were 4.0,4.2, and 5.3 per-cent, respectively.

Uncompensated-care data for 1994 showed nonprofits at 4.5 percent of rev-enues and for-profits at 4.0 percent (ProPAC 1996,84). Morrisey et al. (1996)compared uncompensated care provided by nonprofit hospitals in California in1988 and 1991 to the tax subsidies such hospitals received. For all but 20 per-cent of the hospitals, uncompensated care exceeded the tax subsidies. Earlierresearch had also found the shares of revenue from Medicare and Medicaid tobe similar between nonprofit and for-profit ownership types (Pattison and Katz1983; Sloan and Vraciu 1983; Becker and Sloan 1985). Several more recentnon-peer-reviewed studies have concluded that for-profit hospitals provide lessuncompensated care, but at least some of these studies were conducted to sup-port public advocacy positions of nonprofits (see Kuttner 1996a).

Using a national sample for 1981, Norton and Staiger (1994) found thatwhen for-profit and nonprofit hospitals are located in the same area, they servean equivalent number of uninsured persons. However, for-profit hospitals in-directly avoid provision of care to the uninsured by locating more frequentlywhere better-insured persons live.

Young and coauthors (1997) examined changes in uncompensated care, asa percentage of hospital gross patient revenue (billed charges), for seventeenCalifornia hospitals that converted from nonprofit to for-profit status. Theyfound that acquisition of nonprofit hospitals by for-profits did not lead uniform-ly to less uncompensated care among acquired hospitals. They concluded thattheir results are consistent with the prior research by Norton and Staiger indicat-ing for-profit hospitals provide no less uncompensated care than do nonprofits,given their locational choices.

162 Frank A. Sloan

Frank and Salkever (1991) modeled provision of uncompensated care bynonprofit hospitals. They considered two alternative models of nonprofit hos-pital behavior. In one, the "pure altruism" model, hospitals have profit and un-met need in the community as arguments in the utility function. In this model,increases in the supply of uncompensated care by other hospitals in the com-munity "crowd out" provision of such care by the hospital in question (unlessthe income or endowment effect is very strong). In their alternative formula-tion, the "impure altruism" model, the nonprofit hospital competes for publicgoodwill by providing free care. Here crowding out is less likely because hos-pitals compete for patients who do not pay. Using data from Maryland nonprofithospitals for 1980^, the authors obtained mixed results for both the pure andimpure altruism models. In a market dominated by CMPs, it would seem unlike-ly that they, the new customers, would be particularly impressed by a hospitalproviding a lot of uncompensated care.

Thorpe and Phelps (1991) evaluated the effect of change in New York State'shospital rate-setting program on provision of uncompensated care. As in otherstudies, they found evidence of crowding out. In particular, nonprofit hospitalsprovided less uncompensated care when public hospitals were present in theirmarkets.

New York and Maryland are at one end of the regulatory spectrum: NewYork was the first state to implement all-payer rate setting of hospitals, andMaryland was the only state to retain it. At the other extreme is California,where price competition among hospitals began. With data from that state, Gru-ber (1994) tested the hypothesis that increased competition among hospitalsreduced hospital prices and profits, which in turn led to decreased provision ofuncompensated care. He argued that the potential for shopping by PPOs andHMOs was likely to be greatest in those market areas that had ready substitutesavailable for a given hospital. Availability of substitutes was measured alterna-tively by a Herfindahl index of patient concentration and a distance-based mea-sure. His empirical analysis of the 1984-8 period revealed a large decrease inprices and profits in the California markets with the least concentration of out-puts, and a relative drop in provision of uncompensated care in these same mar-kets. The result on concentration is consistent with Thorpe and Phelps's (1991)finding.

In sum, the issue is not so much that nonprofits have played or play a spe-cial role in the provision of care to disadvantaged populations. Their past rolecan be debated, even with reference to the same empirical evidence. Rather thefocus of policy should be oriented toward the future. Changes in hospital pay-ment, increased competition, and cuts in the growth of public budgets will re-duce provision of such care by all types of hospitals at the same time that thenumber of uninsured persons in the United States is rising.


Quality of care

Another variable is quality of care. It has been hypothesized that nonprofit hos-pitals provide higher-quality care than is demanded, and hence higher than for-profit hospitals are likely to provide (see, e.g., Newhouse 1970). The most rig-orous large-scale empirical study of quality that permits comparisons betweenquality offered by nonprofit and for-profit hospitals is by Keeler and coauthors(1992), who assessed quality based on information from hospital charts. Thesample included fourteen thousand Medicare beneficiaries hospitalized with oneof five diseases in five states.

Keeler and his colleagues found no difference in quality between nonprofitand for-profit hospitals on two quality indicators; public hospitals fared worseon both criteria. On a third measure, however, there was a statistically signifi-cant difference between the quality of care of nonprofit hospitals and that of for-profit and public hospitals, in favor of nonprofits. These authors appear to havebeen more persuaded by the results on the first two indicators, concluding that"nonprofit and for-profit hospitals provide similar quality overall" (p. 1712).They added that "[t]he slightly better quality of for-profit nonteaching hospitalsmakes up for the much higher percent of (high-quality) nonprofit hospitals"(p. 1714). Previous work by Hartz et al. (1989) had found that mortality washigher in for-profit than nonprofit hospitals, but they had used fewer covariatesfor their mortality adjustment.

In their national study of 981 hospitals in 1983^, Shortell and Hughes(1988) found no difference in quality of care by ownership. On structural mea-sures of quality, such as percentage of hospitals with Joint Commission of Hos-pital accreditation, or with cardiac care and intensive care, the two organiza-tional forms are quite similar (Herzlinger and Krasker 1987).

Capital funds

One of the distinguishing features of nonprofit hospitals is that they are pre-cluded from obtaining capital funds from the sale of equity. All their equity isgenerated either externally in the form of philanthropic contributions or inter-nally in the form of retained earnings. Like for-profits, nonprofits have accessto debt capital. Both also have some access to tax-exempt industrial revenuebonds, but only nonprofits can borrow in other tax-exempt bond markets.

Historically, Blue Cross (in some areas) and Medicare paid nonprofit and for-profit hospitals differently, paying for-profits an explicit return on equity. Med-icare no longer makes this distinction, and it is unlikely that many Blue Crossplans do so either. All cost payers cover payment for interest and depreciationexpense.

164 Frank A. Sloan

Wedig and coauthors (1988) found that, given this payback guarantee, hospi-tals highly dependent on revenue from cost-based payers took on more debt;hospitals could recover interest cost from such payers. In the same study, theeffect of the share of revenue the hospital derived from cost-based sources wasgreater for for-profit than for nonprofit hospitals. Other than the effect operat-ing through the cost-based payer share, there were no differences in hospitalcapital structure according to hospital ownership. Perhaps, as my discussionswith persons in the hospital industry suggest was so at the time, for-profit hos-pitals were more adept at maximizing reimbursement. The system rewardedthe maximization of reimbursement rather than the minimization of cost.

Wedig, Hassan, and Sloan (1989), assessing the effects of variations in cost-recovery policies of insurers on hospital location, found that for-profit hospitalmarket shares were higher in states where payment policies were more gener-ous. The converse was true for nonprofit hospitals: The nonprofits satisfied pa-tient demand by being in areas where the for-profits were not willing to serve.

In principle, access to voluntary contributions should be a competitive ad-vantage for nonprofit hospitals. In practice, such donations are an unimportantsource of funds for hospitals as a percentage of revenue. For example, in 1984,only 5 percent of total spending on hospital construction was funded by philan-thropy (Levit et al. 1985).

Dependence on philanthropic funds has diminished over time. Sloan and co-authors (1990) studied the decline in hospital philanthropy both theoreticallyand empirically. In their time series, real donations for medical-facility con-struction rose from $115 million in 1935 to a peak of $3.2 billion in 1965, thendropped to $1.0 billion in 1981 (1996$).

One plausible reason for the decline is the growth of health-insurance cover-age. However, as they showed, the direction of effect of increased coverage ondonations cannot be determined theoretically. On the one hand, increased cov-erage crowds out donations by providing an alternative source of revenue. Onthe other hand, insurance also raises patient willingness to pay for hospital out-put, which the donor plausibly likes; thus, increased coverage allows a dollar'sworth of donations to buy more output. The patient, in effect, matches part ofthe donor's gift. Lump-sum subsidies should reduce donations, but matchinggrants may increase them. (See Chapter 3 for further discussion of the variedpotential effects of increased revenue from program services on revenue fromdonations.) Using both time series and cross-section analysis, the authors foundstrong evidence that increased insurance coverage crowds out private giving tohospitals. Although there has been some erosion in insurance coverage in re-cent years, this is slight when viewed over the entire century. Having access tophilanthropy was a far greater competitive advantage for nonprofit hospitals in1925, when there was essentially no health insurance, than in 1985, when thevast majority of Americans had some coverage.


Given that the only source of private external equity to nonprofits is philan-thropy, it is difficult to know what their cost of capital is. Several authors haveargued that the cost of nonprofit equity equals the return on equally risky tax-free securities (Conrad 1984, 1986; Silvers and Kauer 1986; Herzlinger andKrasker 1987); but determining "equally risky" is not an easy task. One solu-tion used by Sloan et al. (1988) was to take the /3, a measure of riskiness, forpublic for-profit companies; compute the unlevered (without debt) /3; computea levered /3 for nonprofit hospitals with the debt-equity ratio for nonprofits; andthen adjust for the fact that nonprofits do not pay the corporate income tax. Inthis manner they found the cost of equity capital to be from 2.5 to 1.1 percent-age points lower for nonprofit than for for-profit hospitals. The weighted costof capital (cost of debt and external equity capital) was from 1.0 to 0.5 percent-age points lower for the nonprofits - not a very large competitive advantage.During the period analyzed, the weighted cost of capital for for-profit compa-nies was in the 6-9 percent range (nominal estimates). Using a sample of Cali-fornia nonprofit hospitals, Morrisey and coauthors (1996) calculated that, onaverage, interest-rate subsidies for nonprofit hospitals in that state amounted to$437,000 per hospital in 1991, which was about one-quarter of the total tax sub-sidy such hospitals obtained. Corporate income tax and property income taxsubsidies accounted for the rest. Lower cost of capital does not give nonprofitsmuch of a competitive advantage that they might in turn use for the communi-ty's benefit.

Investment

Behavioral differences between nonprofit and for-profit hospitals should plau-sibly be reflected in investment behavior:

1 Their objectives may differ. To the extent that the "bosses" of nonprof-it hospitals value such objectives as quality beyond that valued by themarket (Newhouse 1970) or engage in conspicuous consumption ofhospital inputs (Lee 1971), this should cause such hospitals to pur-chase more plant and equipment.

2 The constraints surely differ. Differences in tax and insurer policieshave already been noted; but perhaps particularly important is the for-profits' access to equity markets, which contrasts with the very limit-ed "market" for philanthropic funds open to nonprofits.

3 As Hoerger (1995) argued, lenders, because of information asymme-tries, may be especially reluctant to lend funds to nonprofits. (This isthe most speculative difference of the three.)

Hoerger (1995) studied individual hospital investment decisions during the1980s in California, Florida, and Tennessee. His most noteworthy finding on

166 Frank A. Sloan

ownership was that cash flow, the sum of the hospital's accounting profits andaccumulated depreciation, had a much greater positive effect on investment fornonprofit than for for-profit hospitals. This result is plausible because nonprofithospitals cannot issue stock.

Discussion, conclusions, and implications for public policy

Not-for-profit hospitals have been transformed from the almshouses of the nine-teenth century to some of the larger locally run enterprises in the communitiesthey serve. In large part, the rapid growth of nonprofit hospitals in particularand acute-care hospitals in general reflects the massive infusion of funds fromboth private and public health insurance that occurred after World War II. Al-ready by the 1970s, nonprofit and for-profit hospitals were much more alikethan different. The nonprofit hospital of the 1990s fits Henry Hansmann's(1980) definition of the entrepreneurial nonprofit better than ever before.

Hospitals of nonprofit and for-profit ownership are similar in provision ofuncompensated care, in quality of care, and in adoption of technology. There isa greater difference in provision of uncompensated care by public hospitals onthe one hand and either type of private hospital on the other. Conversions ofnonprofits to for-profit hospitals do not appear to reduce provision of uncom-pensated care.

These generalizations reflect averages. As the comparisons between uncom-pensated care and tax subsidies revealed, not all nonprofit hospitals pay theirway. Those that do not, merit attention by policymakers; yet such hospitals mayprovide community benefits other than uncompensated care. Also, hospital con-versions from nonprofit to for-profit status may often be in the public interest,or at least neutral; but sometimes, the main motive may be personal enrichmentwithout demonstrable community benefit. The public sector has an appropriaterole in scrutinizing such changes.

Relatively few hospitals have made major commitments to medical educa-tion and biomedical research. This review has deliberately excluded such hos-pitals, although they are reflected in summary statistics for nonprofit hospitals.Society will need to decide among alternative mechanisms for funding such ed-ucation and research. Keeping nonprofit hospitals afloat as a general strategyfor supporting these worthy objectives would be a very inefficient way to ac-complish this important social objective.

More recently, during the 1990s, a major shift is occurring from a model ofcare focused on inpatient treatment of illness episodes to one that emphasizesprimary care, prevention, and integration of a range of services. The hospitalwill no longer be the traditional hub of the health-care system (Shortell, Gil-lies, and Devers 1995). The concept of community is expanding to include careof whole populations. Whether or not hospitals will be able to adapt is still an


open question, as is whether for-profits and nonprofits will differ in their sur-vival abilities.

To the economist, the fact that hospitals are organized in alternative waysprovides a laboratory for studying effects of ownership type and organizationalincentives on behavior. Interestingly, the major behavioral differences relate tohospital financial and location decisions. For-profit hospitals have been morecareful to locate in market areas where families have the ability to pay for hos-pital care, and they have avoided states in which reimbursement has been low.Nonprofit hospital investment decisions are more influenced by availability ofinternal funds.

In the future, hospital-location decisions will reflect the business strategiesof integrated networks. To the extent that such networks find it profitable toserve locations that have been previously underserved, such as inner cities ofmajor metropolitan areas, they will see to it that there is adequate supply of ser-vices. These networks will have no interest in serving individuals who have nohealth-insurance coverage.

The public seems to have little concern about who owns their hospitals. TheKaiser Family Foundation (1995) conducted a survey of American's percep-tions about for-profit and not-for-profit health care in 1995. Particularly rele-vant to the discussion here are responses to the survey's questions. Asked tocompare nonprofit with for-profit hospitals on a number of criteria, respondentspreferred nonprofits in two respects: "more helpful to the community," 65 to 31percent; and "cost you less," 73 to 22 percent. However, for-profits were pre-ferred in three other ways: "are more responsive to customers," 54 to 42 per-cent; "provide better quality care," 57 to 34 percent; and "are more efficient,"59 to 35 percent. Given the empirical evidence that shows very little difference,the citizens' perceptions about quality, cost, and efficiency seem wrong or atleast exaggerated. The difference in helpfulness to the community may be real.In fact, when asked, "Which do you think is better for your community - a hos-pital run and owned by a local for-profit organization or a hospital run andowned by a for-profit national chain?" 70 percent said that the former wouldbe more responsive. This seems to be a matter of local control rather than for-profitness. When asked, "Please tell me which type of health care organizationyou would trust to provide you with high quality health care at a reasonableprice - would it be a for-profit insurance plan, HMO, or hospital, or a nonprof-it health insurance plan, HMO, or hospital, or do you think there is not muchdifference?" almost half the respondents said "not much difference"; the restwere split evenly between the two ownership types.

If local control is a primary benefit of the nonprofit hospital, the next ques-tions is, What is the nature of this good, and who within the community bene-fits? Two likely stakeholders are hospital employees and physicians on the med-ical staffs of these hospitals. Benefits of local control to consumers of hospital

168 Frank A. Sloan

care are less apparent, unless one counts the mere presence of a hospital as aconsumer benefit.

The issue of control appears to be at the heart of opposition by regulatory,physician, or community groups to conversion of nonprofit and public hospi-tals to for-profit chain status (see Chapter 7 and Kuttner 1996a,b). Many of theconcerns are legitimate, such as who controls the foundations established at thetime of the sale and whether hospital trustees or executives receive excessiveinducements to promote it. Another issue is how funds controlled by the newfoundations will be allocated. There is no empirical evidence on this. The for-profit chains may argue that it is the foundations that were established to funduncompensated care, and that therefore the chains themselves have no contin-uing obligation to fund such care out of operating funds as they have done pre-viously.

Will the nonprofit hospital wither? It is not clear what the future holds forthe nonprofit hospital sector. According to one journalistic report, "we will seeeither a growing convergence in the behavior of nonprofits and for-profits ora sharper delineation between institutions with a community purpose and thosedriven by the bottom line" (Kuttner 1996b, 450). More likely, pressure to pre-serve the bottom line will be felt by almost all hospitals. If nonprofits are to dis-tinguish themselves by their community focus, they will have to do two things:define what this focus means in operational terms and secure sources of fund-ing other than patient dollars.

CHAPTER 9

Universities as creators and retailers ofintellectual property: Life-sciences researchand commercial development

Walter W. Powell and Jason Owen-Smith

Introduction

Over the past decade there has been a remarkable shift in the division of laboramong universities, industry, and the federal government. In our view, thistransformation is most pronounced in the life sciences and the commercial fieldsof medicine, Pharmaceuticals, and biotechnology. The post-cold war focus offederal science and technology policy on "competitiveness" has been noted bymany observers, and the intensified interest in basic research and collaborativeproduct development by large private corporations in various high-technologyfields has been widely studied. However, the accompanying change in the man-date of research universities toward a greater focus on commercializing researchfindings is much less understood. Our goal in this chapter is to enhance the un-derstanding of the ways in which the relationship between universities and theprivate economy has changed, particularly in the life sciences. We highlight theprimary forces that have blurred the traditionally distinct roles of the academyand industry, illustrating these trends with data from the life sciences. We re-view several explanations for this transformation, and conclude with a discus-sion of its consequences both for public policy and for the institutional role ofuniversities as generators and disseminators of basic knowledge.

Throughout much of the post-World War II era there was a relatively cleardistinction between basic and applied research, with the former the domain ofthe university and the latter the turf of business. The federal government, out-side of defense-related research, supported the creation of an infrastructure forbasic research and, through the National Science Foundation (NSF) and the Na-

We thank Avner Ben-Ner, Allen Buchanan, Christian Freuh, Laurel Smith-Doerr, and Sarah Soulefor comments on an earlier draft. Powell acknowledges the research support of the Aspen Insti-tute Nonprofit Sector Research Fund and NSF grant 9710729.

169

170 Walter W. Powell and Jason Owen-Smith

tional Institutes of Health (NIH), funded individual scientists as well. Univer-sity research was basic in the sense that it aimed to understand phenomena ata fundamental level. The NSF defines basic science as research whose objectiveis a fuller knowledge or understanding of the subject under study, rather thana practical application thereof; but this focus, as Rosenberg and Nelson (1994)suggest, has never meant that basic research was inattentive to the pull of im-portant technological problems and policy objectives. Indeed, Geiger (1986) hasshown that the "knowledge-plus" focus of U.S. universities has been a "quiet"reality for much of this century, and not just a post-WWII phenomenon. More-over, Ben-David (1977) has observed that U.S. universities have long had amore practical orientation than universities in the United Kingdom or Germany.

Still, universities focused more on the "R" side of the R&D (research anddevelopment) continuum, while much of industry eschewed basic research be-cause the payoffs were either too long-run or difficult to appropriate. The greatbulk of industrial R&D was focused on shorter-term problem solving.1 To besure, basic research flourished in a handful of large corporations, such as AT&T,Kodak, Du Pont, and IBM, whose dominant market positions cushioned re-search budgets from market pressures. However, recent changes in governmentregulation and intensified competition have ended the era of great corporatelabs, and the centralized corporate R&D lab may never actually have been aswidely employed as typically assumed (Rosenbloom and Spencer 1996).

Dasgupta and David (1987,1994) argue that the realms of science and tech-nology are separated more by their social organization and reward structure thanby the actual character of their work. Despite the similarities in the methods oftheir work, scientists and technologists enter their respective realms "precom-mitted" to different norms and rules of the game. For scientists, priority of dis-covery is the goal, and publication the means through which new knowledge isshared in a timely fashion (Merton 1957). The public nature of scientific knowl-edge encourages its use by others, and in so doing increases the reputation ofthe researcher (Merton 1988; Stephan 1996). In contrast, patents are the coinof the realm in the technologist's world. Rewards are pecuniary, and the incen-tive to divulge new information quickly is not as potent.

Our argument is that the separation of the realms of science and technologyno longer holds in the life sciences. The formerly independent, if fragile, sys-tem is today fully interdependent as universities have become much more ori-ented to the commercialization of research. Chapter 1 cautions that "[n]onprofitorganizations confront a dilemma, as does public policy toward them: how tobalance pursuit of their social missions with financial constraints when addition-al resources may be available from sources that might distort mission." With

1 For an excellent historical survey of the relationship between universities and industry, seeRosenberg and Nelson (1994).

Universities as creators and retailers of intellectual property 171

respect to research universities, we argue that income-generating activities thatwere formerly ancillary are taking on much greater salience. This shift is not,we argue, driven by an urgent need for new sources of financing but due tochanging incentives that favor increased efforts at the commercialization of re-search.

This remaking of the institutional division of labor between universities andthe private economy is both recent and profound. We begin our survey of thebroad contours of this changed landscape by focusing briefly on key policychanges at the federal level. We then turn to the corporate sector, noting the sig-nificant alterations in the structure of the firm, particularly with regards to howaccess to new knowledge is organized. The role of universities is considerednext, and we document the increasing commercialization of research, especial-ly in the life sciences, and the new status of knowledge as intellectual proper-ty. We think it is important, however, to recognize that the relationship betweenuniversity research and commercial technology development varies consider-ably across academic fields and industries; our analysis of the life sciences isnot necessarily generalizable to other areas. We thus heed Mowery and Rosen-berg's (1993,53) caution that "no single model or description of the constraints,advantages, and disadvantages of such collaboration is likely to be accurate forall university-industry collaborations."

A changed landscape

Government policy

With the ending of four decades of rivalry and conflict with the Soviet Union,the rationale for federal science and technology expenditures has been reorient-ed toward programs that enhance economic "competitiveness" (NAS 1992; Co-hen and Noll 1994; Slaughter and Rhoades 1996). There is a growing federalview that research universities can and should play a larger and more direct rolein assisting industry and promoting national competitiveness. Universities arebeing urged by the federal government to seek a more direct partnership withbusiness in the development and commercialization of new technologies.

The shift in federal policy - from basic research to increased concern overits application - is reflected in legislation, funding plans, and joint agreements.Most notable on the legislative front were the 1980 Patent and TrademarkAmendments (Public Law 96-517), also known as the Bayh-Dole Act. This leg-islation allowed universities, nonprofit institutions, and small businesses to re-tain the property rights to inventions deriving from federally funded research.In the words of Congress, "it is the policy and objective of the Congress to pro-mote collaboration between commercial concerns and nonprofit organizations,including universities." The 1984 Public Law 98-620 expanded the rights of


universities more broadly by removing restrictions in Bayh-Dole and permit-ting universities to assign their property rights to others.2 The Stevenson-Wydler Act of 1980 (PL 96-480) and its 1986 amendments; the CooperativeResearch Act of 1986 (PL 98-462); the National Competitiveness TechnologyTransfer Act of 1989 (PL 101-189); and the Clinton administration's 1993 "de-fense conversion initiative," in which formerly off-limits defense-related re-search was opened to commercialization - these have all followed the path laidby Bayh-Dole to remake federal policy to enhance the commercialization ofresearch at universities.3

These legislative changes sparked a considerable upsurge in licensing, aswell as rapid growth in the number of university-industry research centers(UIRCs), cooperative research and development agreements (CRADAs), feder-ally funded research and development centers (FFRDCs), and industry-univer-sity research consortia. Accompanying these initiatives has been a significantchange in funding policies.4 At the NSF, numerous programs have been devel-oped to promote university-industry collaboration, and funding in some engi-neering and science and technology fields requires that NSF-supported centershave an industrial component. Stigler (1993, 172) suggests that the NSF hasfound it "easier to explain large-scale projects and research centers to Congressthan to argue convincingly for the diffuse benefits of a broad-based funding ofindividual projects.. . ." The impact of these UIRCs is considerable: Cohen,Florida, and Goe (1994) estimate that 19 percent of university research is nowcarried out in programs that involve close linkages with industry.

Cohen and Noll (1994) argue that the competitiveness rationale for feder-al research support has two strong consequences:

1 There is greater privatization of both the selection (the review pro-cess) and the results (intellectual property rights) of research. The re-view process is increasingly "privatized" by eligibility criteria that re-quire corporate participation, by assigning industry responsibility forthe evaluation of technical merits, and by including the feasibility ofcommercialization and marketing plans as key points of evaluation.

2 Jonathan Cole, Provost of Columbia University, writing in a 1993 Daedalus issue on "The Amer-ican Research University," terms Bayh-Dole "prescient" and notes that "annual revenues frompatents and licenses (at Columbia) have risen from roughly $4 million to $24 million over thepast five years . . . and over the next decade, we could see these figures grow to as much as $75million a year" (Cole 1993, 31). In congressional hearings reviewing the consequences of theBayh-Dole Act, university presidents and officials testified to the success of the legislation in pro-moting technology transfer and generating jobs, particularly in biotechnology (U.S. Congress,Senate, hearings before Subcommittee on Patents, Copyrights, and Trademarks, April 19,1994).

3 Lee (1994) provides a detailed survey of legislation fostering technology transfer.4 For example, the federally funded national laboratories are now expected to generate more of their

operating budgets through the sale of technology (Schriesheim 1990-1; Roessner and Wise 1994).


2 There is now increased collaboration between U.S. firms and researchorganizations, and federal research funds now go to a wider array ofresearch organizations than just research universities.

Business strategy

The embrace of the competitiveness rationale at the federal level reflected awidespread perception among policymakers that U.S. corporations were faringless well in international competition. Although the United States was widelyregarded as the world leader in scientific research, U.S. industry in the 1980slooked vulnerable indeed. Universities and industry were chided for their fail-ure to transfer basic research into commercial development. Within the largecorporation, there was growing recognition that firms had become much lessself-sufficient in their ability to generate the science and technology necessaryto fuel economic growth (Von Hippel 1988; Nelson 1990). As a result, there isnow much greater reliance on external sources of R&D (Badaracco 1991; Ham-el 1991; Hagedoorn 1993; Saxenian 1994). The causes of this transformationare myriad, involving an indissoluble combination of corporations' need to ac-cess sources of expertise located outside organizational boundaries, pressuresto compete with global rivals that do little research internally but are quick toexploit the developments of others, and the tremendous scale of investmentrequired to commercialize new technologies (see discussions in Powell 1990,314-18; Powell, Koput, and Smith-Doerr 1996, 116-22).

In many rapidly developing areas of technology, research breakthroughs areso broadly distributed across both disciplines and institutions that no single firmhas all the necessary capabilities to keep pace (Powell 1996). Consequently, insuch fields as advanced television systems, biotechnology, computers, optics,and semiconductors, firms are turning to cooperation with former competitors,as well as to partnerships with universities and government institutes. Rosen-bloom and Spencer (1996, 70) capture these developments aptly: "What wasonce a race has become more like a rugby match." They anticipate a "diminish-ing role for corporate laboratories as the wellspring of innovation," and suggestthat the "seeds of new technological advance will probably sprout more oftenin university or government laboratories" (pp. 70-1). The private firms that arebest able to exploit such new developments are those with both the most exten-sive external connections and the strongest internal capabilities for evaluatingthe quality of research done elsewhere (Powell et al. 1996).

Many observers have noted that this transformation of corporate researchis most pronounced in the biopharmaceutical field, where there is a complexintermingling of government and university research, small-firm initiative, andlarge-firm development and marketing muscle. Federal research funding hassupported much of the basic science underlying the new biotechnology, and top


researchers in universities and in the intramural branches of the NIH have madethe pioneering discoveries. Science-based small firms, either spun off from uni-versities or with extensive academic linkages, have played a key role in newproduct development. Alliances among universities, small startups, and estab-lished pharmaceutical corporations have proven to be viable vehicles for thecommercialization of new medical treatments.

Universities

Changes in both federal policy and corporate practice have consequently re-shaped the external environment and incentive structure of research universi-ties. No longer are universities just the providers for industry of basic researchknowledge and a trained labor force, skilled in the newest technologies; theyhave become, in the words of NSF Division Director Daryl Chubin (1994,126),"creators and retailers of intellectual property." High-quality research, Chubinsuggests, has multiple components: a basis for scientific knowledge, innovativetechnology available for transfer, and the expertise embodied in faculty and re-search staff. Etzkowitz and Webster (1995,480-1) observe that "science andproperty, formerly independent and even opposed concepts referring to distinc-tively different kinds of activities and social spheres, have been made contin-gent upon each other through the concept of intellectual property rights." Thiscommercialization of knowledge makes universities key contributors to eco-nomic innovation, both as, in Chubin's (1994,126) language, "resource and cat-alyst."

Not only is university research now evaluated more extensively for its com-mercial application, but universities themselves are increasingly viewed as "en-gines of economic development" (Feller 1990). The successes of university-industry affiliations have played a key role in the development of such high-technology-based industrial districts as Silicon Valley, Route 128, Austin, Texas,and the Research Triangle in North Carolina (Smilor, Kozmetsky, and Gibson1988; Rosegrant and Lampe 1992). Numerous analysts have observed that inno-vation has become dependent on a region's technological infrastructure (Romer1986; Krugman 1991; Jaffe,Trajtenberg, and Henderson 1993; Feldman 1994;Feldman and Florida 1994). Thus the success of a relatively small number ofuniversities in contributing to local economic development has changed the ex-pectations for nearly all universities. Now every governor wants the next Sili-con Valley in his or her backyard.

In turn, some universities have decided they can play a more aggressive fi-nancial role than just being an incubator of new knowledge: They seek to sharein the income that may be generated by their new discoveries. Hence, we arewitnessing the growth of universities as venture capitalists (Matkin 1990). Theprogression from incubator of ideas, to patent licensing and technology trans-


fer, to science parks, and finally to equity ownership is based in part on the rec-ognition that in fast-developing fields patents do not easily "capture" intellectu-al property (Merges and Nelson 1990). Negotiating for equity in start-up com-panies is often easier for universities to do than licensing their research findings;but the most compelling factor to universities is the prospect of large financialreturns.

The life sciences represent the cutting edge of these developments, in partbecause of the large number of start-up companies created with the assistanceor direct involvement of academic researchers (Zucker, Darby, and Brewer1994). The biotechnology industry is remarkably clustered in but a few areas -San Diego, the Bay Area, Boston, Seattle, and to a lesser extent, Philadelphiaand Houston - with close proximity to major universities, research-orientedhospitals, and cancer-treatment facilities.5 Physical, intellectual, and economicintegration between firms and universities is so pronounced that they constitutea common technological community (Powell 1996).

The life sciences represent the leading edge

We stress that the nature of industry-university-government relationships arelargely idiosyncratic to particular disciplines. Indeed, one reason that the lifesciences are rather unusual is that biotechnology represents a novel case of anindustry that was developed inside the university. The initial discoveries weremade by university scientists, who then also played a leading role in the in-troduction and development of the new ideas. Biotechnology has thus largelycollapsed the distinction between basic and applied science: Fundamental newdiscoveries, such as gene therapy or the identification of a fat gene, have imme-diate scientific and medical importance as well as enormous commercial rele-vance. Of course, while the life sciences may be the leading edge of commer-cial ventures, there are novel efforts underway in all branches of universities toexploit their "assets" in a manner to enhance revenue generation. Athletics arethe most visible example, but selling seats at poetry readings and all manner ofprofessional education, conferences, and outreach programs are other signs ofgreater attention to commercial activities.

We document the blurring of the division of labor between universities andindustry, marshaling a range of evidence to demonstrate the convergence of thepublic and private sectors. We first present summary data on university patent-ing and on royalties derived from university licensing. We then look at the in-

5 Although university faculty played a key role in biotechnology's emergence, the continuing in-volvement of academic researchers does not appear to be geographically localized. In researchwe are doing, we find that networks of collaboration are not locally based (Koput, Powell, andSmith-Doerr 1997), and Audretsch and Stepan (1996) report that biotech scientist-firm linkagesare typically nonlocal.


creased role of industry funding of university research in the health-sciencesfields, illustrating this trend with a snapshot of the burgeoning field of genom-ics. Next we examine the notable changes in the labor market for biologicalscientists, and see how a common scientific community has evolved that spansuniversities and industry. Finally, we briefly look at two universities at the fore-front of research and its commercialization in the life sciences: Stanford andJohns Hopkins.

Patenting

One way to examine university commercialization efforts is to analyze univer-sities' propensity to patent. Table 9.1 illustrates the limited activity of universi-ties in the period 1969-73, but shows that by 1989-94 university patenting hadincreased dramatically. Note that all three patent classes where university ef-forts were most significant are in the life sciences. Henderson, Jaffe, and Trajt-enberg (1995) have noted that from 1965 to 1992, total U.S. patenting grew byless than 50 percent; our analysis of NSF data for the period 1969-94 showsan increase on the part of universities of over 1,100 percent. Moreover, a muchgreater number of universities is now involved: In 1965,96 patents were issuedto 28 universities; in 1992 nearly 1,500 patents were granted to 150 universi-ties (Henderson et al. 1995). In short, university patenting has grown rapidlywhile industry activity has increased modestly, and the life sciences have beenat the forefront of university efforts.

A 1992 General Accounting Office report (GAO 1992) suggests that thesechanges in patenting propensity reflect greater university focus on commercial-ly relevant technologies and increased industry funding of university research.More specifically, the increased drive to patent is a direct response to the Bayh-Dole legislation that facilitated university retention of property rights to feder-ally funded research. A GAO survey of thirty-five universities found that tech-nology developed in whole or in part with NSF and NIH funding accounted for73 percent of the $113.1 million that the universities received in license incomein 1989 and 1990. The typical licensees were small U.S. businesses in biotech-nology or pharmaceutical corporations.

Licensing and corporate support ofR&D

University revenues derived from patenting and licensing have been growingannually, reaching nearly $255 million in 1994, and a sizable percentage of li-censing revenues are derived from life-science applications. For example, MITis very active in this area: A recent survey (Guterman 1996) of technology trans-fer there reports biotechnology constitutes roughly half of MIT's exclusive li-censes - despite its not having a medical school. The commercial benefits of

Universities as creators and retailers of intellectual property

Table 9.1. Top patent classes for university patenting

111

Class

1969-7373

128435424204514

56423324310

/989-9443551442412825053032420436473

Description

Measuring and testingSurgeryChemistry: Molecular biology and microbiologyDrug: Bio-affecting and body treating compositionsChemistry: Electrical and wave energyDrug: Bio-affecting and body treating compositionsHarvestersChemistry of inorganic compoundsElectricity: Measuring and testingElectrical generators and motors

Chemistry: Molecular biology and microbiologyDrug: Bio-affecting and body treating compositionsDrug: Bio-affecting and body treating compositionsSurgeryRadiant energyChemistry: Peptides or proteinsElectricity: Measuring and testingChemistry: Electrical and wave energyElectrical computers and data processing systemsMeasuring and testing

No. of patents

56323131282726252323

863789518306241229192188178171

Source: NSF National Science Board, Science and Engineering Indicators: 1996, pp.252-3.

licensing are obvious - successful product development by a firm generates roy-alties to a university and to the inventor in question - but the rationales forlicensing a discovery are diverse, and create very different challenges for uni-versities.

Consider the following three motivations, each of which characterizes an im-portant aspect of university licensing of innovative research:

1 A diagnostic test - such as for Wilm's tumor, which affects only a fewhundred young children - has limited commercial value; however, itcan be licensed broadly to speed its clinical development while free-ing the faculty member to continue basic research.

2 In contrast, some licensing is exclusive and entrepreneurial: Produc-tive faculty convert discoveries into the platform for companies thatthey spin off from their university research; the university then li-censes those discoveries to the companies. Such an arrangement pre-


vents the loss of valuable faculty to industry. Indeed, in our researchon the biotechnology industry, we have encountered numerous exam-ples of faculty at leading research universities who have developedtwo or more companies while maintaining their university positions.At the same time, this process does involve a partial "loss" of faculty,insofar as their energy and creativity is shifted to firm-related work.

3 Licensing allows discoveries to be utilized as magnets to draw spon-sored research into the university. Clearly this has been a powerfultrend in the life sciences, generating enhanced funding for faculty, pro-viding employment opportunities for graduate students, and promot-ing a new dialogue between business and industry - feedback thatmight otherwise be absent from both basic science and corporate prod-uct development.

The first example of broad licensing allows a professor to continue his or herresearch; the second - involving exclusive licensing - enables a professor, butnot necessarily the university, to have the best of both worlds, so to speak; andthe third attracts corporate funding to the university.

The growing industry role in health-related research is illustrated in Figure9.1. Note that since 1980 both industry and federal support for health researchhas grown rapidly; around 1988, however, industry spending surpassed the con-tribution of government. To be sure, the great bulk of medical industry R&D isstill done "in-house," but in the context of declining corporate-research spend-ing across U.S. industries, the biopharmaceutical sector is a striking counter-point. Figures for 1995 suggest federal funding for medicine and health on theorder of $13 billion, and an industry contribution of $15 billion (NIH Econom-ics Roundtable, Executive Summary, Jan. 1996). Looking from the viewpointof a single university, we can see just how significant corporate support fromthe biopharmaceutical sector is. At the University of Arizona in 1995, the topfifty corporate sponsors of research and development gave $25 million to theuniversity. Nearly a third of these sponsors are biotech or pharmaceutical com-panies (University of Arizona 1996). In addition, a ranking of all U.S. firms interms of R&D expenditures as a percentage of sales shows that eight of the topten spenders on R&D are biopharmaceutical firms, with two biotech companies- Genentech and Amgen - at the very top (NSF 1996, 121).

Genomics is the broad label for one of the hottest fields in the life sciences.The race to map the sequence of the human genome has triggered a parallel raceto determine gene function. The emerging area of functional genetics includesefforts to pinpoint genes for specific diseases, to correct genetic defects via genetherapy, and to discover tools (ranging from combinatorial chemistry to bio-informatics) to manage the flood of genetic information. Industry-universitycollaborations in this area are both commonplace and so complex that a week-


-#- Federal Government

^r Other Government

Industry

•©. Private

CQ

1970 1980 1990Year

Figure 9.1. National support for health research and development, by source,1960-93. (Source: Provided to authors by NIH Division of Planning and Eval-uation, Planning and Policy Research Branch)

ly scorecard is needed to keep track of who is working with whom. Knowledgeis advancing at an explosive rate in this area, and appealing hypotheses for in-tervention in unmet medical needs are abundant, albeit highly uncertain. Oneattraction of genomics and related new technologies is that they offer a power-ful means to accelerate the drug discovery process. Large pharmaceutical com-panies, fearful of being left behind as new technologies for rapid high-volumescreening of new chemical entities are developed, have either acquired, takenequity positions in, or joined in collaborations with many of the leading smallcompanies. The appeal of the smaller biotech companies is based on both theirintellectual property related to genomics and their closeness to university-basedresearch. Larger, established biotech firms, such as Amgen and Chiron, areheavily involved in genomics as well. For example, Amgen was willing to payas much as $90 million, of which $20 million was up front, to Rockefeller Uni-versity for the rights to a recently discovered gene that may play a key role inobesity. Chiron licensed the exclusive rights to a gene mapping and sequenc-


ing technology developed at New York University that could vastly speed theanalysis of whole human genomes.

A recent report in an excellent industry newsletter (BioWorld FinancialWatch, Sept. 25, 1995) notes that the most active players in genomics are theindependent, "little" biotechs - firms such as Millennium and Sequana Thera-peutics - that have "far-reaching collaborations with academic institutions."Included among their various partners are university researchers, research clin-ics and hospitals, nonprofit institutes, and medical foundations. This broad ar-ray of partners spanning institutions typifies the cutting-edge areas of life sci-ences, illustrating how the boundaries of the research community are beingredrawn.

Changes in the division of labor

We have stressed that academics played a critical role in biotechnology's emer-gence, that many leading professors have formed companies to advance andcommercialize their research, and that biotech companies are closely alignedwith university research. However, researchers also routinely move back andforth between universities and biotech firms, professors take sabbaticals at com-panies, and most established biotech firms run postdoctoral fellowship pro-grams. Cutting-edge research is now performed by intellectually and institu-tionally heterogeneous groups, and researchers at for-profit companies play akey role in the basic science. The 1993 Nobel Prize in chemistry went to KaryMullis for work done at a biotech firm. Powell et al. (1996,140-1) present datathat illustrate the research clout of biotech firms. Table 9.2 ranks leading con-tributors to the literature in molecular biology and genetics, as measured by ci-tations per publication. There are two notable features of the list: the relativeabsence of universities and the presence of two commercial firms, Genentechand Chiron, as numbers 4 and 5, respectively. Many of the most critical publi-cations, as measured by citations, are now coming from nonacademic organi-zations, and in some cases, private firms. In short, the labor market for life sci-entists has been greatly expanded, and the cross-traffic between universities andindustry is now so extensive that it is fair to consider biotech firms and univer-sities as part of a common technological community.

Two universities at the forefront

Further evidence of this blurring or redefinition of boundaries is apparent froma brief review of activities at two leading universities, Stanford and Johns Hop-kins. Both play a preeminent role in life-science scholarship and its commercial-ization. Stanford has been notable for the key contribution its faculty played inthe development of gene splicing, which made biotechnology possible, and in


Table 9.2. Top ten most visible institutions in molecular biologyand genetics, 1988-92

Institution

Salk InstituteCold Spring Harbor LabsWhitehead InstituteGenentechChironInstitute Chemie BiologiqueFred Hutchinson Cancer CenterMITPrincetonMRC Lab Molecular Biology

Cites per publication

41.640.839.733.132.831.827.125.824.023.7

No. of papers

403359392225200261413

1,060369430

Source: Powell, Koput, & Smith-Doerr (1996, 141), based on data from the Institutefor Scientific Information.

the large number of linkages between Stanford faculty and biotech companies.Johns Hopkins has attracted attention for its novel efforts to exploit commercial-ly the research prowess of its faculty and the inventive financial arrangementsit has negotiated with commercial entities.

In 1973, Stanley Cohen of Stanford and Herbert Boyer of the University ofCalifornia at San Francisco (UCSF) created the first recombinant-DNA clone,thus making genetic engineering practical. They worked out a clear way totransplant genes from different organisms into bacteria, which then could begrown in large quantities. Boyer went on to play a founding role in the establish-ment of Genentech, the first widely known biotech company, and one that at-tracted enormous attention in October 1980 when it went public in a frenziedstock offering. Cohen remained at Stanford, where the university obtained apatent on its gene-splicing technique - a valuable piece of intellectual prop-erty that earned Stanford and UCSF $66.3 million each between 1980 and Au-gust 1996 (Puzzanghera 1997, 12A). Etzkowitz and Webster (1995, 489) re-mark that Stanford created the organizational arrangements that promoted theview of high-quality science and generating money as complementary. At Stan-ford's Office of Technology Licensing, Neils Reimers persuaded first Cohenand then Boyer to patent their work. Cohen (quoted in Etzkowitz and Webster1995,489) reports having been somewhat reluctant at first:

My initial reaction to Reimers' proposal was to question whether basic research of thistype could or should be patented and to point out that our work had been dependent ona number of earlier discoveries by others.. . . Reimers insisted that no invention is madein a vacuum and that inventions are always dependent on prior work by others.


Reimers later advised both MIT and the University of California on the setupof their technology-transfer offices (Matkin 1990).

At Johns Hopkins Medical School, an internal venture-capital fund has beencreated to bankroll promising lines of research and move them toward commer-cialization. Adopting the view that the university can play the role of venturecapitalist with greater acumen and not have to share the gains with outsiders,faculty compete internally for venture funds. Johns Hopkins has also "pushedthe commercialism envelope" in other ways, such as licensing prior to discov-ery. For example, the Medical School approached Oncor Inc., a cancer-basedbiotech firm, and persuaded it to fund research in a newly established labora-tory of a leading researcher in return for first rights to discoveries that mighteventuate from the research. Moreover, of the more than $1 million that Oncorpaid the university over a three-year period, about 20 percent was in Oncorstock. Johns Hopkins has aggressively pursued corporate funding of campus-based research. In the process of this commercial expansion, the university hasalso replaced Bethlehem Steel and Westinghouse Electric as Maryland's larg-est private employer.

Key factors driving these developments

The changes underway in life-science research are far-reaching. Most partici-pants are surprised, even astonished, at the speed and extent of these transfor-mations. The life sciences are going through a profound intellectual revolution,just at the time federal, corporate, and university policies are changing in re-sponse to new economic and political conditions. Consequently, developmentsin the life sciences are playing an important role in the process of creating newinstitutional mechanisms to respond to these changes.

Expanding opportunities, not resource scarcity

In contrast to many other areas of academic research, the life sciences do notsuffer from declining resources. Recall that Figure 9.1 showed a steep rise inthe outlay of both federal and industrial support for health-science research. Thelife sciences presently receive more than 55 percent of all federal research sup-port. The NIH is now by far the largest supporter of academic research, and be-tween 1981 and 1990, its budget increased by 50 percent (in constant dollars),a figure two-thirds greater than the increase in total federal outlays. At the endof the 1980s, the government began to fund the Human Genome Project at therate of $3 billion over fifteen years. In fiscal year 1996, while most branches ofgovernment limped along on partial and continuing funding extensions, the NIHreceived a firm commitment from Congress and the president, with a 5.6 per-cent increase, the largest in the federal budget; and, more recently, the 1997 ap-


propriations bill gave the NIH an increase of 6.9 percent to a budget of $12.7billion (Marshall and Lawler 1996).

Increased budgets do not, of course, necessarily keep pace with the costs ofdoing research, especially in fast-developing fields. Moreover, the number oflife scientists competing for funds has certainly increased, intensifying the com-petition. Still, the overall financial picture in the life sciences looks promisingin contrast to other areas of basic research. The motivations in this field are notdriven by a lack of resources, we submit, but rather by the opportunity to ex-pand the pool of available research funds: the chance to develop a new line ofresearch more rapidly, to pursue a promising idea in order to speed the develop-ment of a new treatment that will either improve or save patients' lives. Indeed,there is growing evidence that commercial and basic funding go hand in hand.David Blumenthal (1992) and his colleagues (Blumenthal et al. 1996) reportthat scientists receiving support from biotechnology companies are the same re-searchers who also receive federal funds. Rather than a partitioning of the fieldinto scientists with industry support on the one hand and those with federalmonies on the other, we see the Matthew effect in operation: To those that al-ready have, more shall be given (Merton 1968).

At an institutional level, we find - not surprisingly - the same pattern. Theuniversities most involved in the new biotechnology are among the wealthiestand most prestigious research universities. They are precisely where the cen-ters of excellence in life-science research are located. Our research indicatesconsiderable overlap between the universities holding the largest number of for-mal contractual agreements (including licensing, long-term research partner-ships, and large-scale clinical trials) with dedicated biotechnology companies,and those with the most accomplished research programs in the life sciences.Table 9.3 lists the universities with the most collaborations with biotechnologyfirms, and shows that these schools are also those with the strongest graduateprograms in molecular biology.

We think this convergence is reflective of a change in the nature of knowl-edge, one that has not been sufficiently stressed, nor its ramifications fully rec-ognized. Thoughtful commentators such as Drucker (1993) and Nonaka (1994)have argued that a transformation, equivalent in scope to the industrial revolu-tion, is underway in which the leading edge of the economy is more and moredependent on the production of knowledge. We make a less grand claim, butone with broad implications. The life sciences represent an area in which estab-lished conceptions of knowledge and existing institutional arrangements nolonger adequately fit with the current methods for the production and commer-cialization of science. We see this change in the collapse of the distinction be-tween basic and applied science, in the growing diversity of sources of expertknowledge, and in the complex interdisciplinary and multi-institutional teamsneeded to pursue cutting-edge research. As Powell and Smith-Doerr (1994)

Stanford UniversityHarvard UniversityMITJohns Hopkins UniversityUniversity of California^

53192,6,10


Table 9.3. Universities with the most formal contractual agreements withbiotechnology companies, 1988-95, and NRC/NAS ranking in molecularand general genetics, 1995

Institution NAS/NRC ranking Collaborations

1611111010

aThe University of California (UC) system is treated as a single entity because contracts from allcampuses are signed with the Board of Regents. UC-San Francisco was ranked second, UC-SanDiego sixth, and UC-Berkeley tenth in the NAS/NRC rankings.Sources: Powell, unpublished data derived from Bioscan, 1988-95; National Research Council/National Academy of Sciences rankings of leading U.S. graduate programs by field of study, 1995.

have argued, knowledge is increasingly located in networks of relationships,and access to such networks is a key to competitive survival. Recognition ofthis change in the location and organization of expertise means that it is inappro-priate and misleading to focus only on changes in the organization of universi-ties, without examining how changes in universities dovetail with simultane-ous changes in private-sector R&D, federal policy, and the evolving structureof science. In short, we must view cutting-edge science as the product of a co-evolutionary system of interlinked institutions.

Gibbons and colleagues (1994) describe a shift in the production of knowl-edge from a traditional disciplinary basis, which they term Mode I, to a morediffuse, interdisciplinary project, labeled Mode II. Well-understood extensionsof existing knowledge, with eventual applied commercial applications, typifyMode I. Several of the earliest university commercial endeavors - for exam-ple, agriculture, chemical engineering, mining and metallurgy - are exemplarsof Mode I. Gibbons et al. argue that Mode I knowledge is organized aroundestablished disciplines, where intellectual development is linear. Commercialapplications occur "downstream," and innovations involve breakthroughs inscale. Mode II knowledge, they suggest, is exemplified by biotechnology, high-energy physics research in the areas of superconductivity, and the marriage ofcomputers, software, and telecommunications being played out today on theWorld Wide Web. These forms of knowledge span disciplines, are more com-monly organized through networks than collegial hierarchies, and are character-ized by rapid, often nonlinear development. Consequently, the internal dynam-ics of science have generated a new system of knowledge production, in whichgreater interdisciplinarity and more collaboration are key features.


Although the Mode I-Mode II distinction is a broad characterization, we findit is an apt description of the revolution in the nature and use of knowledge inthe life sciences. Recall that the changes in the most active areas of universitypatenting were from harvesters and measurement devices to the life sciences(see Table 9.1). Remember also that leading biotechnology firms, as well asnonprofit research institutes, now play a significant role in producing and pub-lishing cutting-edge research. At the same time, industry research in estab-lished pharmaceutical companies has become much more theory based, mod-el driven, and deductive than earlier methods of trial-and-error search andinductive reasoning. Breakthrough research involves the collaboration of mul-tiple disciplines and heterogeneous institutions. Consider the 1992 Oliver etal. paper in Nature that reports the sequencing and characterization of an en-tire yeast chromosome. (Yeast has many genes that are homologous to humansand has a fairly "simple" structure that makes it highly useful for inserting for-eign genes for amplification.) This exceedingly laborious and important workinvolved 141 scientists from more than thirty-five nonprofit, public, and pri-vate institutions from around the globe. Alternatively, examine any one of anumber of publications in Science on BRCA1, the gene that plays a critical rolein familial breast and ovarian cancers. For example, the 7 October 1994 arti-cle, "BRCAJ Mutations in Primary Breast and Ovarian Carcinomas," hastwenty-seven authors located at the NIH, a new biotech company, a new de-partment of medical informatics at the University of Utah medical center, aSwedish university, Sloan-Kettering Cancer Center, and two departments atDuke University. These papers vividly illustrate the changing locus of knowl-edge, and how intellectual progress depends on collaboration across discipli-nary and institutional boundaries. (Such cooperation between nonprofit organi-zations [here, universities] and for-profit firms was highlighted in Chapters 1and 8, where it was identified as an important avenue for nonprofits' pursuitof commercial revenue and as contributing to the breakdown of the distinctionbetween the two sectors.)

We contend that the boundaries between universities and firms in the life sci-ences are crumbling. These new organizational arrangements blur the distinc-tion between academic research and commercial development. Life-sciencesresearch not only spans disciplines and organizations but increasingly fusesknowledge and property. Stephan (1994) has shown that the success of initialpublic offerings of biotech firms is significantly related to the reputation of uni-versity scientists affiliated with the firm. At the same time, the annual reportsof university research offices look like documents prepared for the private sec-tor, whereas biotech firm annual reports could pass as grant reports to the NIH.The strong precommitment to different norms and rewards, stressed by re-searchers in the economics and sociology of science (Dasgupta and David 1987,1994; Merton 1957,1968,1988), also appears to be declining as the realms ofscience and technology become inseparable.


Consequences of new mandate for universities

As university management of intellectual property becomes more aggressiveand far-reaching, and as universities become regarded as key contributors toeconomic growth, they become enmeshed in a wide array of political and eco-nomic relationships. This broader role of universities engages many more po-litical constituencies and interest groups and, in turn, creates new expectationswhile making it much more difficult for universities to maintain their legitima-cy. Nobel Laureate Philip Sharp, former Director of MIT's Center for CancerResearch, worries that the desire to harness the commercial potential of univer-sities has led to earmarking, or the award of special-purpose funds by Congresswithout the use of a peer-review process. He cautions, "As universities becomemore identified with commercial wealth, they also lose their uniqueness in soci-ety. They are no longer viewed as ivory towers of intellectual pursuits and truth-ful thoughts, but rather as enterprises driven by arrogant individuals out to cap-ture as much money and influence as possible" (Sharp 1994, 148).

We focus first on how the new mandate for universities has triggered theearmarking of federal research funds to particular local institutions, so that po-litical decisions threaten to reduce the influence of merit review. The pursuit ofnew revenues has also led to significant organizational changes in the structureof universities. There is growing evidence of an academic arms race, resultingin a division between the haves and have nots. Many commentators have ques-tioned whether commercial activities compromise scientific impartiality byintroducing the profit motive into research (Krimsky 1991; Brooks 1993). Westress a less direct, second-order effect of increased commercialization: the po-tential for university and faculty interests to be at odds with one another. Final-ly, we discuss how changes in the mandate of universities have led to disputesover the values and the very culture of university life. We illustrate these issueswith a discussion of the different issues raised by patenting and publishing.

Politics of funding

The trend of shifting research funds away from scientific peer review towardthe more overtly political field of congressional earmarking is suggested by thegrowth of earmarked funds noted in Table 9.4. The expansion of such funds hasbeen dramatic, from $11 million in 1980 to $708 million in 1992, coincidingwith the birth and development of biotechnology. Merit review plays no role inthe earmarking process. Traditionally, earmarked funds were used for the con-struction of university facilities, but by 1992, 42 percent of earmarked fundswent for university research. Put differently, Congress provided almost $300million to university researchers in 1992 without any attempt at intellectual as-sessment (Pennisi 1993).

Universities as creators and retailers of intellectual property

Table 9.4. Congressional earmarking foruniversities and colleges: Total funds by year,1980-93

187

Year

19801981198219831984198519861987198819891990199119921993

Totals

No. of earmarks

70913639384872208252279499

NA1,422

Amount

10,740,0000

9,370,99977,400,00039,320,000104,085,000110,885,000163,305,000232,392,000299,026,000247,976,333470,279,499707,989,031763,000,000

335,768,862

Source: NSF National Science Board, Science and Engineer-ing Indicators: 1993, p. 139.

Debates over earmarking are typically loud and bitter. Critics argue that theprocess is dominated by special-interest politics and extensive lobbying, and re-sults in lower-quality science while simultaneously denying merit-based fund-ing to more worthy projects. Supporters of earmarking, such as former BostonUniversity President John Silber, defend their large payments for lobbying ac-tivities, claiming earmarking represents a way to "level the playing field."6 Sil-ber charges that the peer process is a "tight knit old-boys network," and "an oli-gopoly."7 Earmarking, Silber opines, "is the ante that gives places like BostonUniversity a seat at the table, and the last thing those [elite] institutions wantis an aggressive new player in their game." Boston University received a totalof $56.5 million in earmarked funds in 1984,1988, and 1992, to build research

6 Boston University paid Cassidy and Associates, the most renowned academic lobbying firm, $7.9million over the period 1981-94. Northwestern University, also highly successful in obtainingearmarked funds (portions of which went to support a biotechnology center), paid Cassidy andAssociates $3.7 million during 1984-94 (U.S. House of Representatives 1993, 389, 198).

7 Partially stemming from criticisms raised at the congressional earmarking hearings, the GAOconducted a detailed study of peer review at the NSF and NIH in 1993, finding little evidence ofbias (Marshall 1994).


centers in physics and engineering. In turn, BU's success in obtaining compet-itive grants increased by "647 percent" (U.S. House of Representatives 1993,231). The grants also brought important economic benefits to the local commu-nity, in construction jobs and subsequent employment, a point BU, as well asother universities such as Northwestern, emphasized in their congressional tes-timony in support of earmarking.

Organizational responses

The pursuit of new sources of revenue, from either the commercial marketplaceor the political arena, has led to a number of organizational adaptations withinuniversities. The 1992 GAO report on university research noted that thirty-fourof the thirty-five surveyed institutions, having substantially expanded their pat-ent and licensing programs since 1980, had established a technology-licensingoffice, whereas in 1980 only twenty-two had had such an office. The spread oftechnology-transfer offices, the fees paid to lobbying firms, and the legal ex-penses associated with defending intellectual property are all costs associatedwith new forms of academic entrepreneurship.

Earlier we noted that the commercial firms best able to exploit new techno-logical breakthroughs are those with the most extensive external connectionsand the strongest internal capabilities for evaluating research done elsewhere(Powell et al. 1996). The same process appears to hold for universities, with sev-eral consequences. One effect is an increasingly sharp division between aca-demic haves and have nots. Research funding is already highly concentrated.Schultz (1996,133) reports that 85 percent of federal research funding - about$13 billion in 1992 - goes to a hundred research universities, and 21 percent ofthat funding goes to just ten universities. These same universities are in the fore-front of commercialization efforts. A second-order effect is found in the consid-erable efforts on the part of universities to develop institutional arrangementsto foster external linkages (e.g., industrial liaison programs, contract researchagreements, research parks, clinical trials programs) and internal administrativecapabilities to facilitate them (e.g., offices of sponsored research, technologytransfer, patent administration, institutional development, large legal depart-ments). Matkin (1990) evaluates the technology-transfer offices of four researchuniversities, finding marked differences in how their "commercialization arms"are organized. We know little at this point about the commercial efficacy of dif-ferent forms of university organization, but we suggest that a third-order effectof enhanced commercial efforts is to change the calculus by which political andeconomic leaders evaluate universities.

The various on-campus offices and off-campus programs set up to promoteand process commercial endeavors are not inexpensive to operate. Intellectual-property law is a burgeoning and expensive field, and patent applications incur


considerable costs and time to file, and even more so to enforce. There seemsto be a wholesale rush in this direction by all universities, triggering a form ofstatus-based competition in which universities show the symbols, if not thefruits, of commercial efforts. However, even though many universities are at-tempting to develop internal competencies at commercialization, we suspectthat the lion's share of the results will go to a small handful of universities withthe strongest basic-science research portfolios. As the academic race for com-mercial support heats up, a winner-take-all market results (Frank and Cook1995). Those universities not in the vanguard will find themselves comparative-ly much poorer, losing out in competitions for new facilities, up-and-comingfaculty, promising graduate students, and research funding.

Conflicting interests

Commercial activities do more than generate revenues. Universities in the fore-front of life-sciences research will undergo heightened scrutiny and criticismas they attempt to chart a course in unfamiliar waters. The reality, however, isthat there is no turning back to a less complex and contradictory era. The emerg-ing discoveries in the life sciences are so powerful and dramatic, and the med-ical and material rewards so considerable, that each new scientific discoveryrepresents yet another commercial opportunity. There have been only limitedefforts to assess the consequences of these developments on the internal cultureof universities. Most attention has focused on the individual faculty membersand whether the integrity of their scientific research is compromised by com-mercial efforts (NAS/NAE/IOM 1992). A few observers recognize that the newenvironment creates competing claims: research for knowledge, research fortreatment, and research for competitive advantage (Trias 1996). Indeed, someuniversity administrators argue that "conflicts of interest aren't bad; they'regood." Craig Heller, Stanford's associate dean of research, argues that conflictsmean "you have an entrepreneurial environment. It has to be recognized andmanaged" (Puzzanghera 1997, 13A).

Universities, aware of the potential conflicts between advancing knowledgeand generating revenues, are struggling to develop rules of conduct that simul-taneously - some would say incompatibly - safeguard faculty from commer-cial pressures, mandate disclosure of possible conflicts of interest, and provideincentives to faculty working in areas with commercial potential. What, howev-er, protects faculty when the university represents its own interests in externalcommercial negotiations rather than those of the researcher whose work is be-ing marketed? There are numerous disputes between faculty members who pre-fer an accessible, open license for their discovery, which would maximize thebreadth of knowledge dissemination, and universities that seek a more lucra-tive, exclusive license. Who is the fiduciary when universities convert a profes-


sor's discovery into equity ownership in a company, and that company is sub-sequently sued for patent infringement? A current legal case in California isillustrative of the possible conflicts between universities and their faculty. Twoprofessors won a jury award of $2.3 million from the University of Californiaat San Francisco (UCSF) after claiming it defrauded them by licensing their pat-ents to other companies at a discount in exchange for sponsored research sup-port from these companies. Thus for all the attention to possible faculty con-flicts of interest, a parallel caution is needed regarding the growing opportunityfor faculty and university interests to diverge.

The broadest ramification of the new conception of knowledge as propertymay be its capacity to change the culture of academic life. In a back-page es-say in Scientific American, a professor of pathology at NYU Medical Centerobserves that "when today's professors hit the big time, they have to read theirprofessional literature and Business Week, write scientific papers and patent ap-plications, teach, give seminars, and sit on the scientific advisory boards of vari-ous corporations" (Zolla-Pazner 1994,120). The author goes on to remark that:

The academic scientist finds herself taking a crash course in business and law. The de-mands of negotiating agreements and writing patents drain time and energy. Some re-search activities are redirected from basic science toward more immediately practicalgoals. The promise of continuing industrial support is seductive but inevitably tied tocommercial products and the bottom line. The lab may find itself focused on an agendaset by the company. The basic research that sparked the initial effort may lie fallow. Thespontaneity of scientific pursuit, so prized by those lucky enough to have investigator-initiated government research grants, may be restricted. The speed with which the pro-fessor can share data or new reagents may be slowed. The result, in the worst scenario,would be deleterious for the lab, harmful for science, bad for society.

We doubt that such a wholesale appropriation of a scientist's research agen-da is likely, particularly when there is evidence that scientists who are success-ful in one arena, such as federal grants, also fare well in commercial areas; butthe focus of research can be shaped in subtle ways by commercial exigencyrather than scientific curiosity. Feller (1990) and Cohen and Noll (1994) arguestrongly that academic research is now more directed toward questions whoseanswers constitute "patentable" or commercial outcomes. Feller goes on to sug-gest that faculty working on the "newer" scientific questions are more intensive-ly involved in the postdiscovery stages of research; in one respect, this meansthat faculty now exert more control over the terms and means by which aca-demic research moves to the marketplace. Many would regard this developmentas positive, but few would be satisfied if its consequence was that the merits offaculty research were judged according to its commercial significance.

In an important new study, Packer and Webster (1996) analyze the emer-gence of a patenting culture in British universities, and stress that the makingof claims for patenting departs notably from traditional conceptions of novelty


in science. Patenting involves demonstrating to an imaginary person - the le-gal fiction of a "person skilled in the art" - that a discovery would not have beenan obvious extension of existing knowledge or practice. The universe of pat-enting, Packer and Webster observe (p. 438), is a virtual one in which no onehas "membership" status even when one participates in it. Patenting does notjust involve different rules from academic publishing, it is a different game alto-gether. Writing papers and building an audience for one's ideas involves enroll-ing other academics in a collective project; patenting has more to do with con-trolling others. Put differently, patenting involves claims staking; publishingentails claims making. "Holding patents is not so much a means of enhancingthe credibility of scientists in their research world but a means of defending pri-or investment in the area" (p. 441). For centuries, scientists have done theirresearch virtually unfettered by patent constraints. Now, worries Stanford pro-fessor and Nobel Laureate Arthur Kornberg, "Every one of us working in a lab-oratory . . . have to wonder whether anything we do may have been protectedby a patent and whether we will be sued for it" (Carlton 1995, B4).

Conclusion

We have argued that a profound blurring of the roles of universities and privateindustry is developing in the life sciences, and that this has broad consequencesfor research universities. We contend that these changes are, in large part, irre-versible because they reflect a significant transformation in the nature of knowl-edge. We agree with Hicks and Katz (1996, 394) that "research collaborationamong geographically separated institutions will become the normal way ofconducting research - the rule, not the exception." We add that in the life sci-ences such collaboration spans the academy, private industry, nonprofit researchinstitutes and hospitals, and government laboratories, and we see these institu-tions as coevolving in ways that make them organizationally more similar.

These developments offer ample opportunities for universities to diversifytheir funding base and to contribute to both the advancement of life-sciencesresearch and the development of powerful new medicines that will be of consid-erable benefit to society. At the same time, the ramifications of the growing par-allels in the organization of the academy and industry, and the role of federalpolicy in promoting these trends, are poorly understood. Already Feller (1990)has observed that many academic research teams have the character of "quasi-firms" as scientists eagerly pursue R&D programs aimed at commercial appli-cation. To the extent that the norms of open science are fragile (Dasgupta andDavid 1994), universities may well be endangering their distinctive reward sys-tems. Changes in the reward system (such as tying salary increases or tenureto success at research commercialization) could result in a loss of legitimacy,or speed the movement of scientists to other institutions; either development


would impair the educational mission of research universities. Moreover, theloss of researchers to industry could result in fewer basic-science discoveriesby universities, harming their research function. As the cross-traffic betweenuniversities and industry increases, it becomes less apparent what will remaindistinctive and appealing about a career of university-based research.

We also caution that universities are not well equipped as organizations todeal with the growing status of knowledge as property. Matkin (1990, 382-3)contends that universities typically see the problems associated with commer-cialization as "isolated instances of bad judgement or bad luck, and call for adhoc solutions," rather than comprehending that these problems are outcroppingsof a deeper transformation. We add that of all the participants in the new high-ly linked universe of knowledge development and commercialization, it is uni-versities whose established routines of operations will be most transformed.If practices fundamental to the traditional mission of universities are alteredin a piecemeal fashion, without recognition of the deeper and more systemicchanges we describe, the potential for distortion of the goals of universities isconsiderable. We also stress that changes in operating routines made to accom-modate technological developments in the life sciences might have unanticipat-ed consequences when applied to other sectors of the university. Given that theareas of knowledge that we characterize as Mode II - where intellectual break-throughs have immediate commercial relevance - represent only a limited por-tion of university research, the life sciences may not represent a viable modelfor all university technology-transfer activities.

Enhanced efforts at commercialization typically lead universities to devolvefinancial responsibility to lower levels, allowing individual research units auton-omy and responsibility for their own funding. As this process unfolds, market-based criteria become the dominant logic in resource-allocation decisions. Cur-rent trends clearly point in the direction of this more instrumental focus onresources, even though funding opportunities that are abundant today may beless plentiful in the future. In contrast to the long-term steady support for re-search provided by the federal government, future sources of funding are like-ly to be much more variable and dictated by commercial need. In the life sci-ences, we note that market criteria do not discriminate between medical andnonmedical goals. To date, the great bulk of biomedical research has focusedon unmet medical needs; but there are many areas, such as obesity or smallstature, where highly lucrative research applications might be based more oncosmetic considerations.

The changes underway at research universities are the result of multipleforces: a transformation in the nature of knowledge and a redefinition of themission of universities by both policymakers and key constituents. These trendsare so potent that there is little chance for reversing them - nor necessarily arationale for doing so. Nevertheless, without recognition of the confluence of


forces that are spearheading increased efforts at commercializing research, uni-versity responses are likely to be incremental and inadequate. It must be clear-ly recognized that the conditions that have given rise to the commercializationof basic research in the life sciences are idiosyncratic to this field. Moreover,changes that entail the use of market-based criteria (e.g., more focus on patent-ing and licensing) to evaluate the "merits" of research may, in unanticipatedways, lead to a corrosion of the mission of research universities, undercuttingpublic trust in them. Such developments could, then, erode the very features thathave made U.S. research universities unparalleled contributors to both intellec-tual and commercial advance.

CHAPTER 10

Commercialism in nonprofit social serviceassociations: Its character, significance,and rationale

Dennis R. Young

Introduction

Overall, income from sales of services now constitutes the largest and fastest-growing source of revenues for private, nonprofit organizations in the UnitedStates, proportionally larger than charitable contributions or income from gov-ernmental sources. Salamon (1992) estimates that fees and charges constitutedsome 51 percent of total gross income of nonprofit public-benefit organizationsin 1989, and that such income accounted for 55 percent of the growth in non-profit revenues between 1977 and 1989. Although reliance on income fromsales varies widely among subsectors, there is little doubt that these trends arenow pervasive throughout the U.S. nonprofit sector.

This chapter investigates the nature of this development in selected segmentsof the social services subsector- in six national, social service "umbrella" asso-ciations encompassing youth services, health charities, and services for olderAmericans, as well as among affiliates of a seventh association in the field ofcommunity and recreational services. First, a conceptual framework is estab-lished that relates commercial income to the missions of nonprofit organizations

I wish to thank the staffs of the national associations that provided information and data for thischapter, including Allan Finkelstein, Robert Fischer, and Edward Kagen of the Jewish Communi-ty Centers Association, John Garrison and Bea Lyons of the American Lung Association, John Sef-frin and Dawn Hardwick of the American Cancer Society, Mary Rose Main and Florence Corselloof the Girl Scouts of the USA, Russell Weathers and Stewart J. Smith of Camp Fire Boys and Girls,David Liederman of the Child Welfare League of America, and Joan Wise, Steve Zaleznick, andHorace Deets of the American Association of Retired Persons. I especially want to thank my re-search assistant, Fatih Kocan, for his steadfast help with the statistical analysis. Thanks also toLaura Leete for her advice, and to Anne Standley for the support provided to me through the re-search program of the Mandel Center for Nonprofit Organizations. Finally, I am indebted to mem-bers of the project on nonprofit commercialism, especially Richard Steinberg and Helmut Anheier,for their candid criticism and helpful guidance.

195

196 Dennis R. Young

and to changes in their economic environments. Second, the variety and mag-nitude of commercial income sources, and the strategic issues associated withthese sources, are described through case studies of the six national associations.Third, the commercial income of affiliates of the Jewish Community Centers(JCC) Association is statistically analyzed to determine its relation to other in-come sources, including external grants and contributions and membership fees,and to financial performance.

These analyses raise several important public-policy questions concerningthe regulation and taxation of nonprofit organizations. As nonprofit commercialincome grows, how much of it should be taxed? If commercial activities canundermine the integrity of nonprofits or divert them from their social missions,what limits are required to maintain tax-exempt status? If commercial fundingundermines the confidence of donors or otherwise discourages giving, how im-portant is the loss of donated revenue and what can be done to maintain donorincentives? This study reveals that nonprofit social service associations endeav-or to maintain consistency between mission and commercial activity and alsoto minimize tax liability, and are aided in doing so by missions framed in verybroad terms. These organizations also try hard to avoid commercial initiativesthat pose serious risks to their reputations. There is also evidence that donationsare reduced as sales revenues increase. All told, the study suggests that althoughnonprofit social service organizations deal conscientiously with the challengesand uncertainties posed by commercial funding, tax and regulatory policies mayrequire reexamination in light of the economic environment in which nonprofitorganizations now operate.

Conceptual framework

Studies of enterprising behavior by nonprofit organizations indicate that com-mercial practices are related, directly or indirectly, to the promotion of organi-zational missions (Skloot 1988). Some scholars argue that reliance on partic-ular income streams can co-opt or subvert missions, or displace organizationalgoals (DiMaggio 1986b); the present study, however, reveals that mission is anoverriding consideration in decisions to undertake commercial activities, thoughthe connection to mission is sometimes indirect and subtle. Indeed, missions ofnonprofit organizations, especially national umbrella groups, are often articulat-ed in very broad terms, so that mission relevance is difficult to specify with pre-cision. Nonetheless, association leaders explicitly reference their missions inmaking decisions about commercial activity.

Estelle James's (1983) model of nonprofit-organization behavior, refinedby Schiff and Weisbrod (1991), illuminates how nonprofit commercial activ-ities relate to mission. James distinguishes among nonprofit services that yieldfavored outputs, which are related to mission and valued by managers; neutraloutputs, which are unrelated to mission and to which managers are indifferent;

Commercialism in nonprofit social service associations 197

and disfavored outputs, which may impair the mission or are distasteful to man-agers (Young and Steinberg 1995). In this model, nonprofits may choose to pro-duce favored services at a financial loss, and use financial surpluses from neu-tral or disfavored services to subsidize them. (See Chapter 3 and its discussionof preferred and nonpreferred outputs.)

Commercial activities may contribute to mission in one of three ways:

1 Some activities constitute mainstream (favored) mission-related ser-vices that can be financed, at least in part, by user fees. For example,the Child Welfare League of America (CWLA) provides its memberorganizations with technical consultation. In so doing, CWLA's prin-cipal decision is what to charge, not whether to provide the service.

2 A nonprofit may undertake an (neutral or disfavored) activity solely togenerate surplus revenue. Pure fund-raising strategies, such as rentingout facilities for private parties or hosting bingo games, fit this descrip-tion. Here, the nonprofit organization presumably designs its fee struc-ture so as to maximize profits, and must decide if the activity gener-ates sufficient profit to offset any indirect costs associated with damageto its reputation or image. (See the discussion of "ancillary" activitiesin Chapter 3.)

3 Some (favored or neutral) commercial activities, while optional, areexplicitly intended both to generate surplus revenues and to contributedirectly to mission. For example, the American Association of RetiredPersons (AARP) receives royalties on licensed insurance programsthat serve its members, and the Girl Scouts sell cookies - an activitysaid to provide girls with an experience that helps build character andimpart business skills. In these cases, the organization must decideboth whether it is worthwhile to offer such services and what to chargefor them. Fees may be designed to ensure that operations break evenor run a surplus, but not necessarily to maximize profits if doing sowould substantially reduce mission-related benefits. (See Chapter 4.)

All commercial nonprofit services may be thought of as falling along a con-tinuum within the third category: services with both mission-related and finan-cial consequences but also having a wide range of (positive or negative) weightsattached to generating revenue versus mission-related benefits and costs. Mis-sion-related benefits and costs include (1) direct benefits associated with provid-ing the service to constituents of the organization, and (2) damage to the orga-nization's reputation resulting from involvement in an activity that is viewed asbeing unsavory, in conflict with the organization's mission, excluding worthybeneficiaries who cannot pay, or causing other harm. As considered below, con-cern over reputation is particularly salient in connection with so-called cause-related marketing ventures undertaken with business corporations. These ac-tivities, which can provide nonprofits with considerable visibility to reach the

198 Dennis R. Young

general public through commercial channels, involve "a relationship which tiesa company, its customers and selected products to an issue or cause with thegoal of improving sales and corporate image while providing substantial incomeand benefits to the cause" (American Cancer Society, 1996a, p. 2).

To determine its optimal policy for undertaking a given commercial activ-ity or imposing a specific fee structure, a nonprofit would offset the aforemen-tioned mission-related costs and benefits against indirect financial benefits andcosts: (1) financial profits to the organization that it can use to promote its mis-sion, and (2) financial losses from reductions of donated income that may oc-cur in reaction to the undertaking of commercial activity. The latter effect,known as crowding out, has been studied in connection with both governmentand commercial funding of nonprofit organizations (see Steinberg 1993a; King-ma 1995). Crowding out occurs when donors feel that their contributions areless needed because of increases in other revenues, or where the character ofsuch revenues make the organization seem less attractive.

The foregoing theory establishes a number of prior expectations for empir-ical examination of the commercial activities of nonprofit social service asso-ciations. We expect:

1 the commercial initiatives of these associations to contribute in vary-ing combinations both to generating net revenues and to accomplish-ing the organization's mission;

2 that commercial initiatives will be approached with caution becauseof potential risks to reputation and possible losses of other sources ofincome; and

3 that commercial initiatives will be encouraged by financial pressuresand by the decline of other streams of income, such as grants and do-nations. (See also Chapter 6.)

For the specific national associations examined here, some nuances may beadded to these expectations. First, some of these associations put a particularvalue on membership, over and above the revenues generated from member-ship dues. For AARP and CWLA, membership rolls represent political clout,whereas for affiliates of the JCC Association, membership is a measure of thedegree to which these organizations are succeeding in maintaining Jewish com-munity life. Rather than maximize revenues from membership, these organi-zations try to enlarge membership for its own sake by keeping dues modest(Chapter 4). Hence, the net effect on commercial initiatives of changes in mem-bership revenues is the result of two factors: (1) the inclination of nonprofitmanagers to generate new commercial income to compensate for any declinein membership revenues, and (2) the fact that membership growth from lowerdues may increase the market in which the organization can sell its commercialservices. Indeed, most commercial sales in these associations appear to be made


to their own members. The relationship between membership income and com-mercial income is thus intrinsically ambivalent: If membership revenues slow,commercial activity may increase because managers will attempt to compen-sate for the loss, though they may fail if the (membership) market is shrinking.Alternatively, if membership income increases, managers may choose to reduce(unfavored) sales income, yet that income may nonetheless increase if the un-derlying (membership) market is increasing. We speculate that change in mem-bership levels rather than the price of membership (dues) is the stronger driverof membership revenue, and hence that commercial sales and membership rev-enues will be directly rather than inversely related to one another.

Finally, since the national associations that we examine (except for AARP)encompass semiautonomous organizational affiliates that operate in different lo-cal contexts, we expect the strength of the relationships postulated in the mod-el to vary among differing contexts. For example, the degree to which the lo-cal managerial culture is entrepreneurial, or the local community is charitableor tolerant of commercial practices in nonprofits, may vary regionally, betweenlarge and small cities, or for different sizes and types of nonprofit organization.

Association case studies

Here we describe the commercial practices of six prominent national nonprofitsocial service associations - the character, variety and significance of com-mercial income as a source of revenue for these associations and how this ischanging over time; the relationship of commercial initiatives to the missionsof these associations; and the strategic and managerial issues association lead-ers face in undertaking commercial initiatives. The cases allow us to investi-gate several implications of the conceptual framework presented in Chapter 3and elaborated above, including whether

1 reliance on commercial income is increasing as traditional revenuesources, such as donations and government funding, become more re-stricted;

2 commercial initiatives are being undertaken cautiously and in a man-ner sensitive to preserving associations' abilities to address their basicmissions; and

3 there are particular tensions in such areas as partnering with businesscorporations, where associations are vulnerable to tarnishing their rep-utations among donors, or in the pricing of services, where associa-tions may confront trade-offs between member benefits and revenueconsiderations.

A general questionnaire inquiring about the nature, extent, and profitabilityof commercial activities was directed to the national executives of four youth-

200 Dennis R. Young

oriented associations and four health charities that were participating in a larg-er study of national association structure at the Mandel Center for Nonprofit Or-ganizations of Case Western Reserve University in 1995-6. The instrument wasalso sent to three other national associations that appeared to have particularlyinteresting commercial practices. Of the eleven associations contacted, six re-sponded and four actually filled out all or part of the survey form; the remain-ing two respondents provided reports or narratives instead. Officials of five ofthe respondent associations were interviewed in person, and all respondentsprovided follow-up correspondence and reports.

The six associations do not constitute a scientific probability sample, and thesurveys and interviews did not provide sufficient data for comparative statisti-cal analysis. Still, the responses were relatively detailed and offer a rich pictureof commercial practices. For easy comparison, the following vignettes are pre-sented in successive pairs of similar types of association: two health charities,two youth-serving associations, and two associations that advocate for the inter-ests of large segments of the American population - children and older people,respectively.

American Lung Association

The mission of the American Lung Association (ALA) is to promote researchand public knowledge of lung disease and to advocate for its prevention, treat-ment, and cure. According to annual reports, the budget of the ALA national of-fice has grown approximately 40 percent in current dollars between 1989 and1995, from $33.3 million to $46.6 million. The composition of its revenues haschanged as well: Revenue from donations and government grants has declinedfrom almost 35 percent of revenues to just over 20 percent of the total, whilerevenue from sales has risen from 60 percent to 75 percent. The latter includesgross sales of supplies and services to state and regional affiliate Lung Associ-ations, as well as external sales income composed largely of revenues from sci-entific journals and conferences. The latter component is responsible for the rel-ative rise in sales income: External sales have increased from 13 percent to 26percent of revenues between 1989 and 1995, while internal sales have remainedflat around 48 percent.

For 1995, the ALA reports commercial practices in all three of our concep-tual categories: activities integral to its mission, those undertaken primarily toraise funds, and those that contribute both to fund development and mission. Inthe first category, the ALA publishes two medical journals on lung disease thatearn small profits on gross revenues of $4 million in revenues. These revenuescome substantially from advertising, on which the ALA pays Unrelated Busi-ness Income Tax (UBIT). The association also sponsors an annual internationalscientific conference on lung disease, which earns a profit that is used to offset


the costs of the American Thoracic Society, the medical section of the ALA.In addition the ALA prepares and sells educational materials to its affiliateLung Associations - an activity intended to break even but that in fact has yield-ed small profits or losses (under 5 percent of total national-office revenue) overthe past six years.

Activities undertaken primarily for fund-raising purposes include the leas-ing of mailing lists, which brings in $100,000 annually to the national associa-tion. The ALA also helps affiliates raise funds through gift annuities and plannedgiving, for which it charges fees to cover substantially all its expenses. Someof the ALA's more interesting commercial ventures, contributing both to mis-sion and to fund-raising, involve partnerships with businesses. Golf-privilegecards, which may be purchased from an affiliate Lung Association and entitlebearers to discounts at local golf courses, generate $8 million per year. Thesecards both help the ALA financially and promote the benefits of healthful out-door activity. A scan of the Web sites of affiliate Lung Associations reveals oth-er such practices, including the sale of ski-privilege books by the Western Mass-achusetts Lung Association; of fitness passports to local health clubs and gymsand of restaurant-discount cards by the Lung Association of New York; and ofsmoking-cessation and weight-control programs, as well as carbon-monoxideand radon detectors, by the Lung Association of Ohio.

At the national level, the ALA allows a pharmaceutical company to adver-tise its brands in national magazines along with statements of support for theALA and its public-health work: Essentially, the company provides free adver-tising for the ALA and more than a million dollars per year in payments in ex-change for the credibility it achieves by identifying with the ALA. In addition,the pharmaceutical company's sister brands support the ALA's "Children's ALAChristmas Seal Design Contest," bringing national attention to Christmas sealsand, in turn, providing income for Lung Association affiliates.

American Cancer Society

The mission of the American Cancer Society (ACS) is to eliminate cancer asa major health problem and to diminish suffering from the disease, through re-search, education, advocacy, and service. Until recently, the ACS had limitedits commercial involvement to the publication of two medical journals, forwhich it receives royalties of almost $1 million annually (0.2 percent of its $420million in total revenue in 1995) under contract with a commercial publisher.ACS's publishing policies are now under active review (American Cancer So-ciety 1996b), and financial performance has become one of several criteria con-sidered in evaluating publication initiatives; others include whether publicationsaccurately reflect ACS policy, are scientifically and clinically correct, and aredistributed effectively to target (mission-relevant) audiences. A recent report of

202 Dennis R. Young

ACS's Publications Subcommittee recommended that "as a long run goal, theACS publications program as a whole should either generate positive net rev-enue or be revenue neutral" (American Cancer Society 1996b, 6). Though pub-lications remain integral to ACS's mission, the association is grappling with thedegree to which they should be self-supporting or profitable.

ACS's overall policy toward commercial revenue is clearly changing. Ac-cording to its 1995 annual report, ACS is "forging a new breed of partnershipwith corporations, raising money for cancer-fighting programs while reachingout to new audiences with life-saving messages" (p. 11). This policy came tofruition in 1996, when the association undertook its first two major initiativesin corporate sponsorship - with SmithKline Beecham Consumer Healthcare(SBCH) and the Florida Department of Citrus (Freudenheim 1996b). The part-nership with the Florida Department of Citrus (the state association of growersand producers) involves a nationwide education campaign through local gro-cery stores calling attention to the role of good nutrition in preventing cancer.The Department funds the campaign and also makes a $1 million annual grantto the ACS for research, education, and patient service programs. Under theSBCH arrangement, the ACS logo appears on boxes of NicoDermCQ nicotinepatches and Nicorette nicotine gum, calling attention to ACS's education pro-gram on smoking cessation. SBCH also provides ACS with $1 million per yearin unrestricted grants. ACS views these initiatives not as product endorsementsbut as partnerships to create a national awareness program on the issues ofsmoking cessation and proper nutrition (American Cancer Society 1996a). ACSleaders insist that the mission-related benefits of these efforts dominate any fi-nancial considerations, and that there is a good fit between the mission and theproducts of the corporate sponsors with which they work. Nonetheless, SBCHand the Florida growers clearly receive benefits equivalent to product endorse-ments from ACS.

Girl Scouts of the USA

The stated purpose of Girl Scouts is "to inspire girls with the highest ideals ofcharacter, conduct, patriotism, and service that they may become happy and re-sourceful citizens." This general mission is addressed through a variety of spe-cific educational and recreational activities for girls of different ages, from pre-schoolers to teenagers, administered by local Girl Scout councils affiliated withthe national organization, Girl Scouts of the USA (GSUSA). The national orga-nization reports a number of commercial services considered integral to its mis-sion, including a profitable national equipment service that supplies uniformsand other materials to participants, and various fee-bearing but loss-making pro-grams and training events for councils. Affiliated councils also administer sev-


eral mission-focused commercial activities, including fee-generating programsand activities that run losses; profitable "girl product sales" (products sold byGirl Scouts) including Girl Scout Cookies; and local equipment sales, the profit-ability of which varies by council. GSUSA establishes operational guidelinesand standards for product sales by local councils. Any given troop (a local sub-group of a council) is permitted to participate in no more than two product-salesactivities in a given year, and only one of these may be a cookie sale. (IRS rulesfor avoiding the UBIT on such activities require that they not be carried out ona continuous basis.)

Commercial mission-relevant services represent substantial portions of GirlScout organization budgets. Revenue from net national equipment sales contrib-uted approximately 34 percent of the $36 million 1994 operating budget of thenational organization. Meanwhile, net revenues from product sales constitutedapproximately 57 percent of the $413 million 1994 aggregate operating bud-gets of affiliated councils (GSUSA 1995).

In the category of commercial activity intended primarily for fund-raising,GSUSA reports only income from rental of its facilities. The national organiza-tion rents its conference center in Briarcliff Manor, New York, to outside groupsand uses the income to offset the costs of operating this center for its own train-ing programs. Similarly, local Girl Scout councils generate income from therental of their program sites to outside groups. Generally, facility rentals to out-side groups are priced at prevailing market rates, although this service is alsoseen by Girl Scout officials as providing visibility and good community rela-tions for the organization.

Girls Scouts reports little involvement in cause-related marketing, althoughGSUSA does sponsor a prepaid telephone card and has participated in GIFT-USA, a service in collaboration with the Reader's Digest Association and itssubsidiary, QSP, Inc., wherein people called an 800 number to order gifts offlowers, chocolates, perfume, jewelry, and other items. In this arrangement,QSP increased its sales, and a percentage of the profits benefited Girl Scouts.Some councils are also reported to involve themselves sporadically in their owncoventures and cause-related marketing on the local level; but the national or-ganization provides guidelines for such ventures, to which councils must con-form.

Although cause-related marketing has not yet become a significant financialfactor in Girl Scout operations, GSUSA appears to be approaching this optionseriously and cautiously as a possible way to diversify local council finances,which are now heavily dependent upon product sales. However, GSUSA guide-lines put relatively strict bounds on councils' cause-related marketing, includ-ing limiting such projects to no more than one annually and stipulating the fol-lowing for agreements with partner corporations:

204 Dennis R. Young

Girl Scouts is to be portrayed as a recipient of largesse or commenda-tion, and not an endorser of corporate products.

Exclusive relationships between Girl Scouts and a particular corpora-tion are to be avoided.

Girl Scouts is to have approval rights over all copy and artwork.Councils should dissociate themselves from materials urging the pur-

chase of particular products.Promotion of prizes or incentives encouraging Girl Scout participants

to purchase specific corporate products is to be avoided.

The guidelines also recommend minimum payments made to Girl Scouts fromany promotional activity.

GSUS A also requires councils to obtain its permission prior to entering intoany agreement that might be construed as a commercial endorsement. For ex-ample, GSUSA rejected an offer by a McDonald's fast-food restaurant to helpa local Girl Scout council by advertising its cookie sales on a discount couponfor purchases at the restaurant. In contrast, GSUSA approved the proposal of aPhiladelphia bank to run newspaper ads saluting Girl Scouting.

Overall, net commercial income has been a fairly stable part of Girl Scouts'revenue over the past five years. Such income has declined from 49 percent to45 percent of the national operating budget while increasing from 55 percentto 56.5 percent of affiliates' budgets. To date, Girls Scouts of the USA and lo-cal councils have not found it necessary to pay the UBIT on any of these rev-enues because this income has been offset by corresponding expenses or hasqualified, by IRS rules, as related to the Girl Scout mission.

Camp Fire Boys and Girls

The stated purpose of Camp Fire is "[t]o provide, through a program of infor-mal education, opportunities for youth to realize their potential and to functioneffectively as caring, self-directed individuals, responsible to themselves, andto others; and, as an organization, to seek to improve those conditions in so-ciety which affect youth" (1995 annual report). This broadly framed mission,like that for Girl Scouts, is manifested in specific service activities of localcouncils.

Camp Fire reports an array of commercial activities similar to Girl Scouts,all characterized as integral to its mission. For the national organization, thisincludes the sale of goods associated with Camp Fire operations, such as uni-forms, T-shirts, program books, and supplies. The national organization alsoprovides technical assistance to local councils on such matters as governance,fund-raising, planning, and programming, for which councils pay an annualcharter fee calibrated to the size of a council's operating budget. The charter fee


covers use of Camp Fire's intellectual property, including its name, insignia, andprogram materials. The sale of merchandise and technical assistance servicesare reported as profitable for the national organization. Charter fees contributed54 percent and net profit from merchandise sales 13 percent to the $3.9 millionnational budget in 1994. The national organization also receives fee revenue(2 percent of its budget) from its (mission-related) conferences and workshops.

Affiliate councils also have two principal sources of fee revenue: They sellancillary goods, including chocolate, and they charge user fees for services suchas clubs, camps, child care, and other mission-related programming. These ac-tivities are all reported to be profitable. Program fees constituted about 30 per-cent of the $49.3 million affiliate operating budgets in 1994, and net revenuesfrom chocolate sales alone represented approximately 8.5 percent.

Camp Fire chocolate sales are very similar to Girl Scout cookie sales. Theyare said to contribute to a sense of belonging to the organization and to buildlife skills, and are undertaken in an episodic manner that avoids classificationas taxable unrelated income. However, Camp Fire also acknowledges hazardsof the chocolate sales, such as potential product-liability problems and reputa-tional effects stemming from the sale of a product that is nutritionally question-able. While program fee revenues for Camp Fire councils have increased dra-matically in recent years, by roughly two-thirds in current dollar value from1990 to 1994, chocolate sales have declined by approximately 33 percent overthat same period, continuing a downward trend in that revenue stream since1983. However, the latter is apparently related to Camp Fire's operational diffi-culties with its chocolate-sales operations rather than to any explicit strategicdecision to deemphasize it as a revenue source.

The financial statement of the national organization also cites a small compo-nent of net rental income (1.5 percent of its operating budget). Some local coun-cils have such income as well. Camp Fire reports no significant cause-relatedmarketing activity, but the national organization has recently undertaken to ex-plore potential partnerships with corporations.

Child Welfare League of America

Child Welfare League of America is a national membership association of morethan nine hundred diverse public and nonprofit agencies in the United Statesand Canada "that are devoted to improving life for abused, neglected, and other-wise vulnerable children and young people and their families." CWLA servesprincipally as an advocacy, standard-setting, and technical-assistance organi-zation.

CWLA is substantially engaged in commercial activity, all of which is re-ported as integral to its mission, including a major publications operation, a con-sultation program, and a conference program. In each of these activities, CWLA

206 Dennis R. Young

leadership says it would like to run a profit, but only within the bounds of meet-ing service obligations. The publications and training programs are reported tobreak even, and the conference programs run a modest surplus. Consultationactivity runs at an overall loss; however, some consultation is included free inbasic membership fees.

Gross consulting income represented approximately 15 percent of the 1994budget of $8.5 million, gross publication sales 13 percent, and gross conferenceincome 7 percent; CWLA also collected miscellaneous program service fees(0.5 percent). Membership fees charged to affiliates accounted for 32 percentof CWLA's total revenues, support from endowment was 5 percent, and mostof the rest (28 percent) derived from grants and contributions. CWLA also re-ports 1.5 percent of its budget from cause-related marketing activities. Theseare not separately identified in its financial statements but presumably countedas (corporate) contributions or manifested as in-kind (nonfinancial) benefits (seebelow).

Goods and services sold by CWLA are purchased by members and nonmem-bers at different prices. Members receive publications free of charge and dis-counts to attend conferences, whereas nonmembers are charged "full price" forthese services. Some consultation is provided free to members, but most is paidfor in addition to membership fees; nonmembers are charged for all consulta-tion. In keeping with its service orientation, CWLA charges substantially lessthan prevailing for-profit consulting rates - a maximum rate of $1,000 a day,compared to private rates of $2,000 a day or more for equivalent value (as esti-mated by CWLA management). Overall, CWLA received approximately $1.2million in gross consulting income in 1994 while spending slightly more thanthat on this program.

CWLA, and many of its affiliates, actively engage in cause-related market-ing, which CWLA claims provide financial benefits as well as valuable visibil-ity for the organization in its efforts to promote child welfare. CWLA's leader-ship sees such benefits as hinging on affiliations with "quality" companies, andrisks as connected to poor choices of corporate partners. For example, CWLAavoids companies whose products, such as tobacco or alcohol, are potentiallydamaging to children. Sometimes the risk is less obvious: CWLA cites its asso-ciation with department store J. C. Penney's line of Pee Wee Herman clothes,just two years before that actor's arrest on a morals charge.

CWLA's cause-related marketing ventures include:

"Oshkosh B'Gosh: Tips for Tots for Parents," a program that distrib-utes CWLA flyers in Oshkosh B'Gosh stores and advertises CWLAbooks in the store's catalogs;

an arrangement with Aveda Hair Products in which CWLA flyers aredistributed in beauty parlors featuring Aveda products, with cus-


tomers asked to donate to CWLA as one of five designated char-ities; and

Citibank Dollars under which people receive "CitiDollars" by usingtheir credit cards and are given the option to donate these points toCWLA, also as one of five designated charities.

CWLA affiliates engage in a wide variety of cause-related marketing activitieswith franchises such as Wendy's and Burger King, local supermarkets, and oth-er businesses. In the state of Washington, an affiliate makes more than $200,000a year selling "Washington Apples"; in California, an affiliate makes and sellsa Monopoly® game of their own community; and in New York City, the Chil-dren's Aid Society holds a "Day at the Circus," for which it receives contribu-tions.

CWLA's commercial revenue streams have been reliable, modestly grow-ing sources of income over the past decade. While total revenues in currentdollars rose 140 percent from 1986 to 1994, from $3.6 million to $8.7 million,gross publication sales fluctuated between 13 and 15 percent of that total; grossconsultation income rose from 12 percent of the total in 1990, when it was firstseparately reported, to 15 percent in 1994; and gross conference income (alsofirst reported in 1990) varied between 6 and 8 percent of the total between 1990and 1994 as well. CWLA pays only a few thousand dollars in UBIT, on rev-enues from magazine advertising and rental of mailing lists.

Reliance on membership dues varied between 41 and 32 percent of total rev-enue over the 1986 to 1994 period, representing a growth in absolute terms buttrending downward percentagewise because of overall revenue growth. Reli-ance on grants and contributions has varied substantially in recent years, oscil-lating in the 20-32 percent range between 1986 and 1994.

American Association of Retired Persons

Unlike the 501(c)(3) associations reported here, the American Association ofRetired Persons is classified as a 501(c)(4) social-welfare organization by theIRS because of its strong emphasis on public-policy advocacy. AARP's purposeis to promote the interests of older Americans (age 50 and over) through ad-vocacy, supportive services, and the provision of information and research. Al-though AARP has some four thousand local chapters and is considering waysto decentralize its operations, it operates primarily as a unified national corpo-ration, and its chapters are not autonomous or separately incorporated. The as-sociation has also recently established a 501(c)(3) foundation that can accepttax-deductible donations for specific charitable purposes.

AARP's operating revenues have grown steadily from $150 million in 1985to $416 million in 1995 (Arthur Andersen LLP 1986-96). Approximately 32

208 Dennis R. Young

percent of AARP's 1995 operating revenue came from membership fees paidby individual members, varying from 32 to 39 percent over the period 1985-95.These fees are kept deliberately low ($8 per year) in order to make member-ship widely accessible. Members receive, free of additional charge, the organi-zation's magazine, Modern Maturity, and its Bulletin covering current events.AARP attracted some $55 million in advertising revenue in its publications in1995, but these revenues are said only partially to defer the total costs of its pub-lications program, which also includes specialized newsletters for volunteers,older employees, and other subgroups, and a series of educational brochures onissues such as consumer affairs, crime prevention, and retirement planning, allavailable free of charge. AARP pays full postage on the proportion of its pub-lications that represents advertising and receives a nonprofit postal rate for therest. The association is also subject to UBIT on profits from its advertising rev-enues, but has not paid any tax because "applicable deductions for editorialcosts offset the revenue derived" (AARP 1995, 11).

AARP operates several service programs under its own auspices that it con-siders intrinsic to its mission and for which it charges fees. These include edu-cational programs, such as preretirement and driver-improvement programs, aswell as on-line information services. Other sources of AARP income are moreexplicitly fund-raising in character, including program and royalty fees (13 per-cent of total revenues) and administrative allowances from group insurance pro-grams (23 percent of revenues) undertaken with corporate partners. These pro-portions have all remained relatively stable over the past decade as the overallbudget has tripled. The group health-insurance program with Prudential Insur-ance Company of America offers Medigap coverage to association members.AARP helps market the program, monitors the insurance company, and servesas an ombudsman for member complaints; in exchange, AARP receives an al-lowance from Prudential of almost 3 percent of member contributions (about$102 million in 1994). AARP's program and royalty fees involve the licensingof its name to commercial providers who receive, in effect, AARP's "seal ofapproval" as an asset with which to market their services.

AARP is also engaged in a number of start-up ventures, including a life-insurance program with New York Life Insurance Company, and an annuityprogram with American Maturity Life. In addition, AARP is entering the fieldof managed health care by licensing its name to health maintenance organiza-tions (Freudenheim 1995). AARP also owns a taxable subsidiary called AARPFinancial Services Corporation, which operates credit card and mutual fundprograms (in partnership with Bank One and Scudder, Stevens & Clark, respec-tively) that generate net revenues on which corporate income taxes are paid.Currently, the payment of taxes on AARP's other service-income-generatingprograms is in dispute between AARP and the IRS over the issue of whetherparticular forms of income are taxable (AARP 1995, 12).


A characteristic common to all of these program fee, allowance, and roy-alty arrangements with corporations is that they provide services of interestto AARP members with assured quality, availability, and reasonable cost. Forexample, Medigap insurance is made available with no underwriting and nopreexisting-condition exclusion, and the group's automobile insurance has veryrestricted cancellation policies. These programs also generate net revenues tar-geted to specific AARP services, including the Grandparent Information Cen-ter and Legal Counsel for the Elderly. Hence, these activities both contribute tomission and generate profits to support other organizational services that do notpay for themselves.

AARP's new foray into licensing of managed-care providers highlights someof the risks and potential conflicts of such arrangements. This initiative wasprompted by the declining fee-for-service market for medical care and thegrowth of managed care. However, as the New York Times reports:AARP will be lending its name to businesses that at times limit or even refuse medicalservices, instead of just making sure that claims are paid. That could create a backlashin a constituency that has generally looked upon the association as a vigorous advocatefor its interests. (Freudenheim 1995, C22)

This new initiative, perhaps more than other AARP ventures, highlights thefine line the association, as all nonprofit organizations, must walk between ad-vocating and promoting the interests of its members or other constituents andresponding to commercial opportunities that promise lucrative financial returns.

Review and assessment

The six case studies reveal substantial reliance by these associations on - andtheir increasing interest in - commercial sources of income. The variety of com-mercial activity, the growth of fee income from sales of services, and the grow-ing involvement in cause-related marketing arrangements with business corpo-rations are particularly noteworthy. The vignettes reveal that the value receivedby the associations from commercial initiatives is twofold, as our conceptualframework suggests: both direct, mission-related benefits and surplus revenuesthat can be used to promote the organization's mission-related work. In cause-related marketing ventures, private corporations also benefit by obtaining ac-cess to the markets represented by the constituents of these associations andfrom a "seal of approval" that gives the companies privileged access to thesemarkets. Thus, the associations trade on their reputational and membership as-sets to expand sales of products and services that become identified with theirnames, so as to promote their causes and raise funds. In so doing, they are be-coming more sophisticated about the need to preserve and protect those assetsby entering collaborations only with reputable companies and into activities thatwill not harm their constituents. Nonetheless, as cause-related marketing be-

210 Dennis R. Young

comes more prominent, nonprofits will continue to encounter tension and crit-icism, as recent AARP and American Cancer Society experiences have shown.

In all cases, broadly framed organizational missions provide the touchstonesthat association leaders reference in assessing the advisability and design of par-ticular initiatives. In almost all cases, commercial initiatives bear some directlogical connection to mission; for instance, health charities link with compa-nies providing health-related products or services. Although these connectionshelp sell products, they also commonly contribute to mission, directly throughprovision of valued services to constituents or increased visibility of a cause,and indirectly through net revenue generated for the nonprofits. Associationsseek to avoid commercial initiatives in conflict with missions or posing risksfor their constituents, though some of their recent initiatives are not without con-troversy. Nonprofit social service associations are still struggling with the fineline between commercialism and mission orientation, and the fineness of thisline ultimately raises questions about whether changes in regulatory or tax pol-icies are needed to ensure that these organizations continue to act in the publicinterest.

Most of the associations (AARP, CWLA, GSUSA, Camp Fire) reviewedhere look to their own members as the markets for the services they sell. In ad-dition, national leaders of associations such as Camp Fire and GSUSA recog-nize that their affiliates have increased their reliance on commercial fundingover time, in part because they have been less able to rely on traditional sourcesof local charitable donations such as United Way. The dynamics of decline inexternal sources of support and the cultivation of members as markets are is-sues examined more closely in the following analysis of the affiliates of the JCCAssociation.

The Jewish Community Centers Association

The Jewish Community Centers Association is a national organization whoseaffiliates are Jewish Community Centers and YM-YWHAs in local communi-ties in the United States and Canada. These affiliates (hereafter called JCCs)provide health, recreation, and other community services. The JCC Associationpublishes annual Budgetary Trends reports (e.g., Kagen 1995) that provide de-tails on the income and expenditure patterns of individual affiliate members.1

JCCs derive their incomes from several sources, including program incomefrom fees and tuitions for nursery schools, day camps, child care and other ser-

1 Although affiliates are autonomous and vary in their accounting practices, they are all asked toprovide the same data to the national organization. Because compliance is high, the data set isrelatively uniform and complete, permitting examination of the factors associated with a JCCsreliance on income from sales, including the influence of donated and membership revenues.


vices; basic membership fees; additional membership fees paid for health-clubprivileges; individual contributions; revenues from fund-raising events; invest-ments; facilities rentals; grants from federated sources including United Wayand Jewish federations; grants and contracts from government; foundationsgrants; and other miscellaneous sources. Over the years, the composition ofJCCs' revenues has changed. While total revenue for these centers has grownsubstantially, most sources of income, including foundation and governmentgrants, charitable contributions, and membership fees, have grown much lessrapidly. The major component of revenue growth has been the increase in userfees charged for participating in programs, which have grown from just over30 percent of the total in 1980 to more than 45 percent in 1994. Overall, inter-nally generated income derived from sales of services and dues has increasedfrom 60 percent to almost 80 percent of the total over this period, while exter-nal income derived from grants and contributions has declined from 40 percentto a little over 20 percent.

Below, we offer a preliminary statistical model for explaining the changesin various components of commercial income (called "activity" or "service" feeincome by the JCC Association) that JCCs received between 1993 and 1994.We hypothesize that:

The level of commercial income in a given year depends on basic membershiprevenues in that year, external donative revenues in that year, the operatingsurplus or deficit in the previous year, and various environmental or "control"variables, such as organization size and geographic location, that may affectcommercial revenues for a given JCC.

It is difficult to estimate this model directly because commercial income,membership income, and external income are all correlated with the total rev-enues of the organization, leading to colinearity problems. Thus, it is helpfulto transform the model into first difference form, by subtracting the equationrepresenting this relationship in one year from that of the next year. Thus, forthe years 1994 and 1993, our model is the following:

Change in Commercial Income from 1993 to 1994 = Jc(change in basic Mem-ber Revenues from 1993 to 1994; change in External Income from 1993 to1994; change in operating Surplus or Deficit from 1992 to 1993).

Note that the control variables are presumed not to change between 1993 and1994, and hence drop out of the equation. Nonetheless, in our analysis we re-peated our estimation of this model for different subsets of JCCs, by size andlocation, to determine whether the management processes among those subsets

212 Dennis R. Young

vary in a manner that affects the parameters of decision making on commer-cial income generation.2

Various measures of the dependent variable Commercial Income were esti-mated, including

1 Program Fees, for services such as day care and summer camps;2 Unrelated Income, encompassing revenues from special fund-raising

events and facilities rentals; and3 Total Sales, which is the sum of Program Fees, Unrelated Income, and

fees for health-club services.3

The sum of these three components reflects the overall package of options avail-able to a JCC that wishes to generate sales income.

As noted, the foregoing model assumes that an executive of a JCC decideshow aggressively to pursue commercial income strategies, contingent on theexpected levels of other "exogenous" income streams (including regular mem-bership dues as well as grants and contributions received from outside sources),and the JCC's financial performance in the previous year. An operating surplusfrom the previous year, or any increase in membership and external revenues,will reduce the pressure to generate commercial income; an operating deficit ora decrease in those revenues will increase the incentive to generate commercialincome. Thus, we expect the signs associated with External Income and Sur-plus to be negative. Commercial income may reflect promotion of less favoredactivity, or instituting or raising prices, hence curtailing demand for preferredactivities. All other things equal, our framework suggests that managers prefernot to pursue such policies and are inclined to increase commercial income onlywhere it helps to compensate for expected losses in other revenue streams or tostanch operating losses (see Chapter 6).

The matter of membership revenues is more nuanced, given the dynamicsof membership fees. An increase in membership revenues reflects either an in-crease in the number of members, or an increase in membership fee levels, orboth. If the number of members increases, this represents an enhanced base ofcustomers to whom supplementary fee-based services can be sold. Most earnedincome services, such as day care, nursery schools, summer camps, or healthclubs, are indeed sold primarily to individuals who are JCC members. Thus, we

2 The consecutive years 1993 and 1994 were chosen for three reasons: first, the reporting formatfor the data has changed over time, and these two years were recorded consistently with one an-other. Second, these are the most recent years for which these data were available. Third, the mod-el presumes that allocation decisions are made on a year-to-year basis, with this year's decisionsdepending partially on last year's financial performance. Thus, analyzing changes in consecutiveyears most closely tracks the character of the decision-making process.

3 Note that health-club fees are supplementary to regular membership fees. Many JCCs do not havehealth clubs, and they are an optional service to members where they do.


may expect a positive relationship between membership fee income and com-mercial income. The net effect of an increase in membership revenues on com-mercial income will be the difference between the conventional negative influ-ence (crowd-out) and the positive influence of an increased membership base.

We need to justify the assumptions that membership revenues, external in-come, and surplus are indeed exogenous (i.e., essentially given to JCC manag-ers and independent of their commercial revenue decisions). External donativeincome is most easily understood to be exogenous: These revenues are provid-ed by governments, foundations, United Way, and local Jewish federations. Al-though allocations of these funds to individual JCCs may be influenced some-what by the energy or creativity of their managements, they are probably muchmore heavily determined by the condition of the local economy, trends in giv-ing and government expenditure, and competing demands on charitable funds.

Membership revenues are conceivably more influenced by management andless clearly exogenous; however, memberships serve a different objective forJCCs than do program fees or other sources of internally generated funding. Anoverall goal of JCCs is to provide for the cultural needs of local Jewish com-munities. Maintaining Jewish identity, community traditions, and cohesion areoverriding imperatives to maximize participation in JCC programs. Therefore,membership enhancement rather than revenue generation serves as the primaryinfluence in centers' membership drives, despite pressures from some fundersto increase member revenues; JCCs are thus disposed to keep membership feeslow enough to maintain current members and attract new ones. For purposeshere, membership fee policies are taken as given in management's general de-liberations on revenue strategy, although this may not be entirely accurate. Fi-nally, we have assumed that an operating surplus or deficit in a given year servesas a signal to a JCC executive as to whether sales activities should be pursuedmore or less aggressively in the next year. Since the previous year's deficit ishistorical fact and cannot be affected by current-year decisions, we can treat itas exogenous.

Data on a sample size of eighty-one JCCs were available for the years 1992,1993, and 1994. Both linear and log-linear regression models were estimatedfor the entire sample and for subsets of organizations in different regions of theUnited States and different-size cities. The JCC Association classifies the urbanareas in which each tabulated JCC is located into one of four categories - inter-mediate, large intermediate, large, and metropolitan - using a combination ofthree factors: Jewish population size, total expenditure budget, and number ofmembers. Hence the categories are as much an indicator of JCC size as of thesize of the city in which it is located. We did separate analyses for the thirty-five JCCs classified as large and metropolitan, and for the forty-six cases classi-fied as intermediate and large intermediate, to see if the operations and manage-ment processes in larger-city JCCs were different from those in smaller ones.

214 Dennis R. Young

Each center was also classified by region of the country (and Canada) as de-fined by standard grouping of states (Lincoln Library of Essential Information1966). We performed separate analyses for the thirty-six centers located in theNew England and Mid-Atlantic states and for the forty-three centers located inthe other U.S. regions to see if differences in managerial cultures or traditionsmanifested themselves by region.

Although significant results were found for several different estimations, themost consistent and significant were obtained for the log-linear regressions us-ing the forty-six JCCs in the smaller-city subset.4 These estimations, displayedin Table 10.1, highlight the following findings:

Change in External Income is significantly and negatively related tochange in Program Fees revenue, consistent with theoretical expec-tations. Elasticity is -31 percent for this relationship; that is, a 10percent decrease in External Income is associated with a 3.1 per-cent increase in revenue from Program Fees, all else held equal.

Change in Surplus is positively and significantly related to change inUnrelated Income, with a modest elasticity of 1.4 percent. The signof this relationship is counter to expectations of our model. Ratherthan slacken Unrelated Income activities when they run surpluses,JCCs may be mildly inclined to pursue them further, perhaps tohedge against risk, accumulate capital for future investment, or forother reasons, as suggested by Chang and Tuckman (1990).

A positive and significant relationship exists between Total Sales andMember Fees, and a significant and negative relationship betweenTotal Sales and External Income. The elasticity of change in TotalSales revenue with respect to change in External Income is mod-est, approximately -2.5 percent. However, elasticity of change inTotal Sales with respect to change in Member Fees is quite large:An increase in Member Fees yields a proportional increase in To-tal Sales almost twice as large.

These results are generally consistent with theoretical expectations: Wheresignificant, external income exhibits an inverse relationship with commercialincome (as was also hypothesized in Chapter 3). Membership revenues, whichprobably reflect changes in underlying membership rolls, exert large direct ef-fects on total sales income because services such as health-club privileges are

4 It is not obvious why these results emerge significantly for the smaller-city subset and, except forthe membership revenue effect, not for the data set as a whole. Although there are no dramaticdifferences in the distribution of funding sources of JCCs in the smaller versus larger cities, thesmaller JCCs may nonetheless feel more vulnerable to changes in external income from grantsand contributions. Hence, a reduction in external income may be more likely to prompt compen-satory action to increase sales income in these organizations.


Table 10.1. Log-linear regression analyses of factors affectingcommercial income in smaller-city Jewish Community Centers,1993-4

External IncomeMembership RevenuesSurplus

ConstantSignificance (F)Adjusted R2

Sample size

Program Fees

-0310*0.818

-0.00904

0.4793.005*0.134

46

Unrelated Income

0.00377-0.0657

0.0138**

3.26**5.01**0.236

46

Total Sales

-0.0245*1.95**0.00723

-3.19*9.091**0.384

46

Note: All variables are logged differences of 1994 values minus 1993 values, except forSurplus, which is the logged difference of (1993 Surplus) - (1992 Surplus). Surplus ismeasured as the difference between total revenues and total expenditures in a given year.*Significant at .05 level**Significant at .01 level

purchased by members. However, operating surplus is found to exert an unan-ticipated small direct effect on unrelated commercial income.

The fact that commercial income is inversely related to external income sug-gests that policymakers need to examine whether new donor incentives can becreated to maintain contributions in the context of increasing reliance on salesrevenues. Moreover, since government grants and contracts are a component ofexternal income for the JCCs, the analysis suggests that government also needsto examine how much its own direct funding cutbacks exacerbate reliance ofnonprofits on commercial income.

Conclusion

The case studies of national associations and the statistical analyses of JCC af-filiates support our conceptual framework for understanding the pursuit of com-mercial income by nonprofit organizations. In particular, new sales initiatives,imposition of fees for mainline services, and collaborations with business allappear to be driven by a combination of desires to promulgate favored (i.e.,mission-related) services and to generate surplus revenues. It appears to be therare initiative that does not contain elements of both of these motivations. In-deed, few if any of the commercial initiatives of the studied associations areeasily characterized as undertaken solely for the revenue, or as "disfavored" inthe sense of detracting from the organization's mission. Even where initiatives

216 Dennis R. Young

seem largely designed for their revenue-producing potential, associations gener-ally go to some pains to demonstrate how these activities are supportive of theirorganizations' purposes, constituents, and capacities. In virtually every case, as-sociations can identify some direct contribution that an initiative makes towardits mission. Moreover, the six national associations appear to exercise signifi-cant caution to ensure that their commercial activity damages neither the orga-nizations' reputations nor their abilities to promote their work among constit-uents - the two principal assets upon which these associations rely. Indeed, itis these nonprofits' reputations that make them attractive to their corporate busi-ness partners.

Our results suggest that nonprofits are moved to increase sales income byfinancial considerations. This is consistent with the theoretical expectation thatnonprofit managers undertake sales initiatives or raise prices only reluctantlywhen the welfare of their enterprises is threatened by cutbacks in governmentgrants or private giving. The JCC analysis specifically confirms the apparentsubstitution effects between external (contributed) revenue and reliance onincome from sales. It must be noted, however, that the directionality of thiscrowd-out effect of contributions on commercial income requires further in-vestigation. Though it seems likely that reductions in exogenous donated rev-enues lead to at least partially compensatory increases in JCCs' sales activity,it may also be the case that commercial venturing and the raising of fees andcharges by the JCC affiliates lead charitable and governmental sources to re-duce their grants to these organizations - a matter about which little is known.

Some other financial considerations that might be expected to influence theundertaking of commercial initiatives appear to be relatively unimportant. Inparticular, UBIT appears only to be a minor concern for the six national asso-ciations; they pay little such tax, and appear to design their sales programs toavoid tax vulnerability (Chapter 5). Similarly, the JCC analysis suggests thatfinancial performance, as measured by the previous year's operating deficit orsurplus, does not influence commercial income generation in a major way.

Finally, the JCC analysis also makes clear that considerable variation in prac-tices exists among local nonprofits in their approaches to commercial income.In particular, important differences exist between larger- and smaller-city JCCs.The six national associations also exhibit a wide variety of commercial prac-tices. In all cases, however, it is fair to conclude that commercial income is be-coming more important over time as a source of revenue, and all of these orga-nizations continue to examine their options and strategies for generating moresuch revenue.

CHAPTER 11

Zoos and aquariums

Louis Cain and Dennis Meritt, Jr.

Introduction

Those who have not visited a zoo or aquarium since LB J was president will finda quite different place. The menagerie mentality, cages and tanks arranged fordisplay purposes, has disappeared. The best zoos are no longer those with themost different kinds of animals. As part of the ecological awakening that culmi-nated in the passage of the Endangered Species Act of 1973, and part of theirhistoric mission of exhibiting animals, zoos and aquariums transformed theirdisplay space from cages and tanks to habitats designed to promote normal be-havior. By helping to create an environment in which breeding could take place,this investment in new capital helped promote a second mission, one to whichthese institutions have rededicated themselves - species preservation.

There have been other changes as well: To facilitate research, laboratoriesand libraries have been expanded. To provide education, classrooms have beenconstructed and teachers hired. All the while, zoos and aquariums have tried toremain financially accessible to low-income citizens. What has changed littlesince the founding of the American Zoo and Aquarium Association (AZA) inthe 1920s is the mission of these institutions:

to promote and advance zoological parks and aquariums; to aid in the exchange and im-portation of zoological specimens; to provide exhibits for scientific, educational and rec-reational purposes; and to aid in the preservation of wild life.1

What has changed are the weights attached to each of these goals and the meansadopted to accomplish them.

The Endangered Species Act of 1973 reflects a recognition that human ac-tion has imperiled the survival of an increasing number of species. By one esti-

We thank Neal David, Joseph Feme, Craig LaMay, and Jan Schweitzer for their comments; SupriyaMathew and Gina Germane for their research assistance.1 The earliest citation we could locate for these goals is Doolittle (c. 1932,100).

217

218 Louis Cain and Dennis Meritt, Jr.

mate, 20 percent of all species resident in rain forests will become extinct overthe next quarter century. By another, it takes ten million years to replenish whatis lost after a wave of extinction. As Edward Wilson has observed, "Humani-ty is now destroying most of the habitats where evolution can occur" (Wilson1993, 29).2

As implemented, the goal of the federal legislation is to minimize the prob-ability of extinction, but that legislation has been concerned primarily with wildpopulations. Zoos and aquariums, the institutions responsible for captive popu-lations, have adopted an identical goal, yet they play almost no role in this legis-lation - nor in the resulting policies.3 All concerned have strong preferences forattempting to preserve species in the wild, but it will take more than habitat ac-quisition and protection to prevent extinction (see, e.g., Randall 1986; Williamsand Nowak 1993; see also Tober 1981). It has been estimated that, in the nexthalf century, at least fifteen hundred species will be so endangered that theirsurvival will require captive propagation (Foose 1993, 163).4 Species such asPrzewalski's Wild Horse and the California Condor survive today only in cap-tivity.5 Consequently, zoos and aquariums have undertaken a coordinated pro-gram to preserve endangered species. In 1980 a report entitled "A Species Sur-vival Plan for AAZPA" (Meritt 1980) was presented at the association's annualmeeting.6 A Species Survival Plan (SSP) is a cooperative agreement among ac-credited AZA members and similar institutions internationally to preserve thegene pool of endangered species through controlled breeding for genetic diver-sity. Such plans have led to the development of a large computerized data base,the International Species Inventory System (ISIS), that tracks the animal collec-tions of the major institutions (see Seal 1976). In short, once zoos and aquar-iums renewed their dedication to this mission, modern technology was broughtto bear in a short amount of time.

There is a heightened sense of urgency. Until the 1960s, scientists estimat-ed the total number of animal and plant species at approximately three million.Advances in both taxonomy and statistics over the next two decades caused thatestimate to be raised to ten million. This explosion in the number of species sug-gested the possibility that many more were endangered than originally believed.

2 The purpose of the Endangered Species Act of 1973 is to provide a means for conserving the hab-itats in which endangered species reside, as well as to conserve the species themselves.The U.S.first passed wildlife conservation laws around 1900; endangered-species conservation legislationwas passed in 1966 and 1969 prior to the Act of 1973.

3 The entire spectrum of legislation in this area is discussed in Bean (1983). Federal monies havebeen allocated to purchase critical habitat, but the public funds that have been made availableto zoos and aquariums have come largely from local governments.

4 Over the same period, it has been estimated that the efficient utilization of resources will enablezoos and aquariums to house approximately a thousand species (Conway 1986).

5 There are plans under way to reintroduce these species into their native habitat.6 Prior to 1994, the association was called the American Association of Zoological Parks and

Aquariums (AAZPA).

Zoos and aquariums 219

The first official federal list of endangered species, the "Redbook," appearedin August 1964 and included only sixty-three vertebrate species.

Captive propagation programs such as SSPs are a short-term solution to asurprisingly long-run problem:

The primary objective of captive propagation, however, is the reinforcement, not re-placement, of wild populations. Captive populations are not an end in themselves. Evenif the worst occurs in the immediate future, we must be optimistic that in several centu-ries, perhaps as long as 500 to 1,000 years, the human population will stabilize and de-cline. It should then be possible to restore or even reconstruct ecosystems, but only ifwe have kept all the parts. (Foose 1993, 150)

Each organism contains a genetically coded blueprint, one that we still donot understand all that well. The only way we have of retaining the informa-tion from, say, a rain forest is to protect it, to preserve it as it is. To quote Ed-ward Wilson:

The surviving biosphere remains the great unknown of Earth in many respects. On thepractical side, it is hard even to imagine what other species have to offer in the way ofnew pharmaceuticals, crops, fibers, petroleum substitutes and other products. We haveonly a poor grasp of the ecosystem services by which other organisms cleanse the wa-ter, turn soil into a fertile living cover and manufacture the very air we breathe. (1993,29)

The transformation of zoos and aquariums from menageries to habitats ispart of the attempt to resolve this uncertainty. Whereas exhibition was once pri-mary, species preservation (including the attendant research and education) hasbecome at least as important. That transformation involved a considerable in-vestment in plant and technology, one that severely strained the financial con-straint of many institutions.

Zoological parks and aquariums are often cited as examples of merit goods,goods that the public demands but is unwilling to purchase at prices that covercosts.7 The principal reasons people go to zoos and aquariums reflect a differ-ent set of motivations than those of the people running them. Demand and sup-ply was ever thus.

The hypothesis of this chapter is that, to meet the increase in costs associat-ed with the re weighing of their goals, zoos and aquariums have faced a tight fi-nancial constraint and have increasingly turned to commercial activity. As ex-plained in Chapter 3, the revenue-generating activities are divided into thoseconnected with the organization's preferred private goods (or PP-goods; e.g.,animal exhibition, for which it can charge user fees) and the preferred collec-tive goods (or PC-goods; e.g., species preservation, education, and research,for which it cannot). The PC-goods have been the motivation for the capital im-provements that have been made, even though there clearly are important spill-

7 This follows from the definition in Musgrave (1959). See, for example, Netzer (1978), 16-18.It is clear that zoos are externality-producing collective goods; see Throsby (1994), 23.


overs into the provision of the PP-good. As we shall demonstrate, there has beenno change in the structure of demand for the PP-good. More money could beraised through increased admission fees (the sale of the PP-good), but zoos andaquariums have been restrained in their use of that approach lest they becomeinaccessible to poor families. (See Chapter 4 for more on such distributionalgoals.) The Reagan-era tax-rate reductions, which increased the price of chari-table giving, led to a reduction in donations.8 Consequently, zoos and aquari-ums have turned increasingly to the sale of nonpreferred private goods (or NP-goods; e.g., stuffed animals, T-shirts, food), the sale of which has been growingin importance. Still, the PP-goods must not be neglected for they are what keepsthese institutions in the hearts and minds of their consumers.

In order to examine our hypothesis, we study zoos and aquariums that aremembers of the AZA.9 Because of the limits on what the statistics collected bythe AZA reveal, we conducted two case studies:

the relatively small Glen Oak Zoo, owned and managed by the Peo-ria (IL) Park District, which has an annual attendance that has re-mained at 125,000 for several years; and

the large Lincoln Park Zoo in Chicago, with a reported four millionvisitors annually, which is managed by the nonprofit Lincoln ParkZoological Society and subsidized by the Chicago Park District.10

Most U.S. zoos are like Glen Oak. They are in the public sector, customar-ily part of a park district for budgetary purposes. Most have a private supportgroup, a zoological or aquarium society, that provides supplementary funds.Although most can be characterized as "public," this merely describes featuressuch as ownership or primary source of funds.11 There are no substantive differ-ences between the goals or operations of public and private nonprofit zoos andaquariums. Glen Oak's goals and how it would like to accomplish them are sub-stantially the same as Lincoln Park's.

Statistical overview

Zoo finance is complicated by the fact that there are usually multiple budgetauthorities to which a single institution reports. Animal care typically is as-

8 Many public subsidies are tied to private donations through some degree of matching arrange-ment.

9 As a first step, we sought and gained the endorsement and encouragement of the Board of Di-rectors of the then-AAZPA. They were extremely supportive because no one, from any disci-pline, has ever attempted this analysis.

10 The zoo was owned and managed by the Chicago Park District until 1 January 1995, when man-agement was turned over to the Zoo Society under a thirty-year operating agreement.

1! There are a few AZA members that are private, for-profit enterprises, but they are not includ-ed in this study.


signed to the zoo or aquarium's budget, but maintenance and repair might bethe responsibility of the park district in which the zoo is located. Developmentoperations (membership, gift shops, and the like) are often the responsibility ofa private zoo society.

Our limited sample is based on annual reports and budgets from eighteenzoos and aquariums in the United States. There is great variety in the mannerin which the information is reported, which limits the analysis.12 For assess-ment purposes we divided the fifteen-year period between 1975 and 1989 intothree consecutive five-year periods. The zoos and aquariums represented inTable 11.1 were those for which we had both income and expense informationfor at least two years within two of these three periods. Hence this table repre-sents only nine institutions, although its basic contour is unchanged if the infor-mation from the other nine is included.13

Table 11.1 is an attempt to sort out some common denominators. It was as-sembled as follows: The information for each institution was reconfigured intocommon categories (i.e., membership revenue) for each year. The revenue ineach category was expressed as a percentage of the total. We then calculated,for each institution, the simple average of those percentages for each five-yearperiod. The table reports weighted means of the average percentages over allinstitutions.14 Table 11.1 also provides a preliminary look into the question ofwhether organizational size mattered. The institutions have been divided be-tween "large" (six) and "small" (three) institutions to see if there were any dif-ferences. We defined "large" as an institution with annual revenue (or budget)greater than $1.5 million.15

The growing commercialization hypothesis implies zoos have become moredependent on commercial revenue (sale of the NP-good) over time. There isa substantial decrease in the percentage of revenue derived from donations be-tween the late 1970s and early 1980s. In all three groupings, fewer total dollarswere received from donations in the late 1980s than a decade earlier. Given thisdecrease, all the other percentages increase between the late 1970s and early1980s, with the exception of admissions (sale of the PP-good) in small zoos.

12 There may well be some sample selection bias: We believe the majority of U.S. zoos and aquar-iums operate at a deficit, yet of the ten institutions for which both revenue and cost data areavailable, only three reported an operating loss. The remaining eight sent us information on their(budgeted) expenditures.

13 A few, too few, institutions sent us information for earlier years. The data for the years before1975 are consistent with what is reported in these tables for the period 1975-9.

14 The weights are the proportion of each zoo's total income or total expenses to the pooled totalfor all the zoos.

^ The choice of $1.5 million was purely arbitrary. There are two institutions with average rev-enues within 10 percent of that figure. Most were either in excess of $3 million or less than$1 million.


Table 11.1. Overview of zoo finance

All institutionsTotal revenue

DonationsAdmissionsConcessionsMembershipsOther

Total expense

Large institutionsTotal revenue


Total expense

Small institutionsTotal revenue


Total expense

1985-9

$4,972,3660.20810.45200.22960.06010.0502

$5,224,022

$6,851,0910.21430.45540.22270.06040.0472

$7,218,020

$1,214,9170.13830.41390.30760.05660.0837

$1,236,025

1980-4

$3,235,8490.28490.42980.19620.05450.0346

$3,359,051

$4,450,7360.29340.43280.18550.05480.0335

$4,643,504

$806,0740.19130.39590.31430.05180.0467$790,145

1975-9

$1,868,6280.58190.29360.09230.01120.0209

$1,876,734

$2,746,5660.60240.27860.08870.01100.0193

$2,785,931

$551,7210.42880.40590.11950.01320.0327$512,939

Note: Dollar amounts are annual averages. Other numbers are proportions of total rev-enues.

The growth of expenses reflects commitments, particularly to increases inPC-goods, that zoos and aquariums made in the 1970s, when they would haveanticipated the composition of their revenue stream to contain a larger share ofdonations than was realized in the 1980s. Though they might have anticipatedthat donations would not grow as fast as required revenues, it is doubtful theyanticipated the drop that occurred. Both large and small institutions began sell-ing memberships, which increased from slightly more than 1 percent of rev-enues to around 5.5 percent. Both increased the proportion coming from con-cessions by about 2.5 times. Large zoos and aquariums increased the portioncoming from admissions, but this remained relatively constant for the smallones. To understand the search for replacement revenue, to see why all zoos andaquariums turned to sales of the NP-good as well as to raising the price of thePP-good, it is necessary to take a closer look at all potential revenue sources.


Donations

The decline in the share of total revenue coming from donations was true forboth large and small institutions, although the small ones received approximate-ly two-thirds the percentage of revenues from donations as large institutions(and no more than 15 percent of the absolute amount of donations) in each peri-od. This is attributable, in part, to the fact that, among the zoos in our sample,local government subsidies were more often accounted explicitly by large zoos,but this does not explain the entire one-third difference. Small zoos have alwaysbeen less dependent on donations than have large zoos, as the smaller percent-age for these institutions attests.

From our analysis of the Lincoln Park Zoo we find that its individual dona-tions largely have been for capital purposes. The 1996 fiscal year, which end-ed on March 31, is the first complete year under privatization, and in that yearreported contributions were 42.8 percent of revenue.16 With privatization, therewas a major need to solicit donations to help build an endowment.17 Corporatedonations are intermingled with corporate sponsorships (also accounted as do-nations) - the Smith Company building rather than the Mr. Smith building. Thisis largely an untapped area at Lincoln Park, as well as at other zoos and aquar-iums.

Approximately 10 percent of the Glen Oak Zoo's revenue comes from dona-tions. This is, in part, the result of an Adopt-an-Animal program that parallelsone offered by Lincoln Park. In part, the donations are the result of the effortsof the Friends of Glen Oak Zoo, a support group that has helped provide capi-tal funds and, in recent years, about $12,000 per year in operating funds throughthe solicitation of donations, but also through the sale of memberships and NP-goods at concessions stands. A generally unrecorded form of donation towardoperating costs comes as food for the animals: Glen Oak receives occasionalusable donations from grocers; Lincoln Park gets substantially more.18

Glen Oak provides an excellent example of a second type of unrecorded do-nation that goes toward capital costs. The zoo transformed its old lion cage intoan education room and constructed a new, outdoor lion enclosure with volun-

16 This contrasts with a much lower percentage for large zoos in the most recent period reportedin Table 11.1. During this time, the zoo was nearing completion of a major capital-fund drive.The accountants created estimates for the 1995 fiscal year in which "contributions" were aneven larger percentage of larger reported revenues.

17 Endowment income helps smooth the annual variation in ordinary income, especially for thoseinstitutions charging admission fees. An event such as privatization represents a "hook" to so-licit such donations, but it was not fully realized by Lincoln Park. The zoo recently received athree-year grant to develop a planned-giving program.

18 When Lincoln Park acquired koalas, it needed a source of eucalyptus. It receives a donation bymeans of which several species of eucalyptus are grown in Florida and airlifted to Chicago ona regular basis.


teer labor (including the architect) and materials at (or below) cost. The totalout-of-pocket cost was roughly $50,000, as opposed to $300,000 if, as wouldbe true at a large zoo, the factors were purchased at market prices.

Admission fees

The Lincoln Park Zoo has never charged an admission fee. As part of privatiza-tion, the Chicago Park District pledged it will provide a subsidy to the zoo soit can continue to operate without one. The monetary value of this pledge in1996 was $5.5 million, approximately a third of reported revenues; whether thatwill suffice to ensure free admission in the future remains to be seen.19 The GlenOak Zoo, under its agreement with the Peoria Park District, is currently respon-sible for raising 56.7 percent of its budgeted revenues. Presently, 50 percentof its $600,000 annual budget comes from admission fees ($3.25 for adults and$1.50 for children), slightly higher than the 41 percent reported for small insti-tutions in Table 11.1.

To evaluate admission fees (i.e., sale of the PP-good) as a potential revenuesource, we collected data from 111 nonprofit, AZA-accredited institutions thatreported in 1994 and the 85 institutions in this group that had also reported in1968. The average adult admission charge had increased from $0,210 in 1968to $3,934 in 1994 - by a factor of almost 19. There are at least two reasons,besides general inflation, why prices jumped so dramatically over this period.

First, the proportion of institutions charging admission has increased. A sym-posium at the 1965 AZA Annual Meeting debated the question of admissionfees; the consensus was that a modest such fee - one that was not set to raiserevenue - helped to minimize problems such as loitering and littering, yet keptthe zoos and aquariums accessible to low-income families ("Symposium onFree versus Admission Zoos" 1965, Z26-33, Z73-5). The average adult admis-sion in 1994 ($3.93) can be considered "modest" in comparison to alternativeattractions. It also can be argued that the average adult admission overstates thetrue price: Zoos and aquariums customarily admit members without additionalcharge every day, have well-publicized "free days" (often a designated day eachweek), and admit school groups and the like without charge for educational pro-grams.20 The symposium did not discuss what impact raising fees might haveon revenues, nor how much it might cost to collect an average admission feeof $0.21. In any event, it is clear that many of the institutions that had been freein 1968 were charging admission in 1994.

19 For the following seven years, the Park District contribution will change with the CPI, then be-come frozen in nominal terms.

2 0 Haeseler (1988,173) reports that, in a sample of thirteen zoos, free admissions accounted foran average of 29 percent of total attendance.


Second, in the 1980s, zoos raised admission fees in an attempt to raise rev-enue, particularly to help defray the cost of constructing new habitats. The av-erage adult admission fee in real terms (1982-4 = 100) increased from $0.60 in1968 to $2.65 in 1994, an average annual increase of 5.9 percent over twenty-six years. Although it is not completely clear why these institutions adopted amore aggressive pricing policy, the timing suggests that the absolute decline indonation dollars between the late 1970s and early 1980s may have been an im-portant stimulus.21

The available data on attendance are also an overestimate of what an econo-mist would like to have. The AZA reports the Annual Attendance claimed byeach accredited zoo; in some cases this is only an educated guess on the part ofthe institution - many zoos (especially the free ones) do not have an accuratecount. Nonetheless, even if accurate attendance figures were available, it wouldstill be true that many of the attendees did not pay the full adult admission.Mindful of data limitations, regressions were run for 1968 and 1994, using thisquantitative information as the dependent variable, to address the question ofwhether the structure of demand had changed in any way that would enablezoos and aquariums to rely more heavily on admission fees.

The Price variable is the institution's adult admission rate. As noted, this isan overestimate of the average amount collected; in fact, we believe it to bemore highly overestimated than the attendance data. For this reason, the esti-mated coefficient, showing the relationship between price and the level of atten-dance, will be biased upward.

Other independent control variables include the Population of the metropol-itan area in which the zoo is located and the median Income of families for thatmetropolitan area. Both variables are taken from the Census of Population for1970 and 1990, respectively.

Visitor surveys suggest that those most likely to attend zoos and aquariumsare families with children.22 They also suggest that older people are unlikely toattend because of the amount of walking involved. Consequently, the percent-ages of the metropolitan population aged Under Eighteen years and those OverSixty-Five are included. These also come from the two Censuses.

Given that these institutions reputedly set price with reference to the goal ofretaining access by low-income residents, a variable is included for the percent-age of Low Income persons in the population. The data used are from the Coun-ty and City Data Book and reflect the percentage of low-income residents inthe central city defining the metropolitan area (reported as "low income" in the1972 book [1969 data] and "poverty income" in the 1994 book [1989 data].)

2 1 A second important stimulus, to be discussed below, was a renewed effort to attract members,and free admission is one of the normal benefits of membership.

2 2 A number of visitor surveys have been published in the Proceedings of the Annual Meetingof the AZA. In addition, we solicited such studies from all AZA-accredited institutions.


The number of species and specimens, both of which are reported by AZA,are indicators of institution size, statistical measures of the depth and breadthof a collection. They are highly correlated, however, and only the Number ofSpecimens is used in our statistical analysis.23 There are some difficulties withthis measure. For example, a zoo specializing in large mammals is likely to re-quire a larger space, yet have fewer species and specimens than a zoo special-izing in invertebrates. Casual empiricism (confirmed by the data in this case)suggests a strong, positive correlation between the size of an institution and an-nual attendance.

According to the visitor surveys, the modal zoo visitor unit during the weekis a mother and her children; the father is present on the weekend. Since theseinstitutions represent one of the few places a family can take young childrenfor an outing, the Number of Competitors is defined as the number of AZA-accredited institutions in the area.24

One dummy variable, Aquarium, was added as a reflection of the differencesbetween zoos and aquariums; each institution that is exclusively an aquariumwas noted.

A multiplicative specification of the functional form has been selected; thevariables appear in logarithmic form, with the exception of the three measuredin percentages and the one dummy variable. One can interpret the coefficientof the logarithmic variables as elasticities. Further, it is assumed these institu-tions are providers of a strictly private (PP-)good: the opportunity to view ani-mals. Almost all of them receive some type of subsidy, if only the ability to usepublic land on favorable terms; they are not dependent solely on admission feesfor revenue.25

The results of the estimations appear in Table 11.2. The coefficient on Pricehas the expected negative sign and lies between -1 and 0 in both regressions,although it is not statistically significant in 1994. As noted earlier in this sec-tion, it is biased away from zero. The fact that zoos and aquariums operate inthe inelastic portion of their demand curves suggests they do not set price tomaximize profits.26 The price inelasticity of demand implies zoos and aquar-iums could raise additional revenue through an increase in price.

Increases in Population lead to increased Annual Attendance. A 10 percentincrease in population is associated with a 2.9 percent increase in attendance.

2 3 Similarly, both the number of full-time employees and the annual budget are highly correlatedwith the number of specimens. The number of specimens is the least correlated with annualattendance.

2 4 Commonly cited alternatives to such a trip are visits to parks and shopping malls.2 5 In almost all cases, subsidies for operating expenses are independent of other revenue sources.

Subsidies for capital expenses are often on a matching basis.2 6 Given that both price and attendance are overestimated, it seems unlikely that the "true" coef-

ficient is in the elastic portion of the curve.

Zoos and aquariums

Table 11.2. Estimates of the demand for zoos and aquariums

227

In Price

In Population

In Income

% Under 18

% Over 65

% Low Income

1968

-0.6650*(-2.5066)

0.2868**(3.3302)0.5680

(0.7318)0.0086

(0.2482)-0.1135**

(-2.6576)0.0067

(0.0257)

1994

-0.0838(-0.7672)

0.2916**(3.7945)0.2010

(0.3578)0.0044

(0.1827)-0.0237

(-0.8433)-0.0050

(-0.4063)

In No. of Specimens

Aquarium

No. of Competitors

Constant

Observations(Aquariums)

1968

0.6099**(6.9672)-1.2633**

(-4.7808)-0.0098

(-0.0906)0.8861

(0.1272)

0.78085(5)

1994

0.5138**(9.0958)-0.3598*

(-1.4636)-0.1528*

(-2.2641)3.9288

(0.6714)

0.667111

(8)

Note: The dependent variable is the logarithm of Annual Attendance; f-values are in parentheses*Significant at .05 confidence level, two-tail test.**Significant at .01 confidence level, two-tail test.Sources: AAZPA, Zoos and Aquariums in the Americas, August 1968 and 1970; AZA, ZoologicalParks and Aquariums in the Americas, 1994-95; U.S. Bureau of the Census, Census of Popula-tion, 1970, vol. 1, and Census of Population, 1990, vol. 1; County and City Data Book, 1972 and1994.

This may be a reflection of the fact that our best-known zoos and aquariumsare in larger cities and attract more out-of-town visitors. Of the five institutionswith a 1994 attendance in excess of two million, all are located in metropoli-tan areas with populations greater than two million.27 Since the 1994 regres-sion includes several smaller zoos that either did not report or did not exist in1968, the mean annual attendance for 1968 was, in fact, larger than that for1994 (see Table 11.3).

The coefficient on Income is positive in both years, although it is not statis-tically significant. Zoos and aquariums are normal goods. The hypothesis thatprices have been kept low to keep the institution affordable by the poor sug-gests an income elasticity between zero and one, and this is what we observe.

The three demographic variables add little; only one of the six coefficientsis statistically significant. The percentages Under Eighteen and Over Sixty-Five both have the expected signs, but that of percentage Low Income changes

2 7 There are ten institutions reporting attendance greater than or equal to 1.5 million. Of these nineare in areas of greater than two million in population, the exception being the Monterey BayAquarium. At the other extreme, of the eight institutions reporting attendance less than or equalto 100,000, four are in areas of less than 250,000 and three in areas of 250,000-500,000; onlythe Brandywine Zoo is in an area of more than 500,000.


Table 11.3. Means of variables

Variable

PricePopulationIncome% Under 18% Over 65% Low IncomeAquariumNo. of SpecimensNo. of CompetitorsAnnual Attendance

No. of SpeciesSpecimens/Species

1968

1.46804,621.4

10,183.1734.519.03

10.540.059

795.491.42

490,345.6

211.353.76

1987

5.181,049,441.7

36,797.4725.8011.6518.360.072

1,183.811.68

461,999.8

217.015.46

Sources: AAZP\,Zoos and Aquariums in the Americas, Au-gust 1968 and 1970; AZA, Zoological Parks and Aquariumsin the Americas, 1994-95; U.S. Bureau of the Census, Cen-sus of Population, 1970, vol. 1, and Census of Population,1990, vol. 1; County and City Data Book, 1972 and 1994.

from positive to negative. The reason appears to be the result of a North-Southdifferential: On average, Southern zoos have a smaller annual attendance andare located in metropolitan areas where the percentage of low-income familiesis larger.

The coefficients on the variable representing the size of the zoo, Number ofSpecimens, have the expected positive sign and are statistically significant. Justas the more popular zoos and aquariums are in the largest cities, they also tendto be larger than the average. The five institutions with an annual attendanceover two million report an average of over 4,109 specimens and 552 species,in comparison to an average of 1,184 specimens and 217 species for all 111 in-stitutions. It is also worth noting the increase in the number of specimens perspecies, consistent with the shift from menageries to habitats (see Table 11.3).

The coefficient on the variable Number of Competitors is negative in bothyears, but it is statistically significantly different from zero only in 1994. Thisis consistent with the hypothesis that the presence of substitutes in the area re-duces attendance. At the mean 1994 attendance, one additional institution inan area will reduce attendance by approximately 65,000. The coefficient on theAquarium variable is always negative and statistically significant in both 1968and 1994. At the mean 1994 attendance, aquariums drew roughly 140,000 few-er people.


A Chow (F-test) was run to test for a change in the structure of demand be-tween these two years. As noted, even though these years bracket tremendouschanges in the nature of zoos and aquariums, the conclusion is that there hasbeen no statistically significant change in demand as measured by the set ofvariables in Table 11.2.28 Zoos are beginning to take advantage of the inelastic-ity of demand by charging higher admission fees, but, between these two years,there was no significant change in demand that would make such fees a moreattractive revenue source.

Memberships

As Table 11.1 reports, revenue from this source has increased from 1.1 percentin the late 1970s to around 6 percent, but it remains a relatively small percent-age of total revenue. Memberships are a mixture of donation and concessions.The normal membership package includes benefits such as free admission. Mostzoos and aquariums offer NP-goods (a stuffed animal or T-shirt) as an incen-tive for making the initial commitment that are not offered for renewals. Mem-berships in the Lincoln Park Zoological Society generate (subtracting direct ac-quisition and retention costs) about sixteen cents per dollar the first year andabout eighty-three cents per dollar thereafter for the zoo's general revenue fund.Both the Lincoln Park Zoological Society and the Friends of Glen Oak Zoo plancontinued efforts to increase memberships, but report relatively stable member-ship over the past few years. Although some increase is possible, Lincoln Parkbelieves this is a "mature" revenue source that is not going to be a source ofsignificant revenue increase. Glen Oak suspects the stability of membershipthey have observed is the result of a Friends group that has not been as activeor as effective as would be desirable. They believe they could increase theirmarket penetration.

Concessions

The final category of revenue, concessions, represents the sales of NP-goods.As Table 11.1 documents, concessions now generate approximately 22 percentof the revenue at large zoos and almost 31 percent at small zoos. These percent-ages were considerably smaller in the late 1970s. Although the percentage hasincreased steadily for the large zoos, it jumped dramatically for the small zoosin the early 1980s, then remained constant. It appears to be the case that, for

2 8 The statistical test is based on the residual sum-of-squares (RSS) for the individual regressionsversus a pooled regression. The RRSs are 17.961 for the 1968 regression and 31.685 for the1994 regression; that for the pooled regression is 54.652. The resulting value is /MO,186 = 1 595.The critical value at the 5 percent confidence level is /M0,200 = 2.56.


small institutions, the installation of carnival rides and boutiques selling T-shirtshas proved to be a more profitable strategy than the organized appeals to thecharitable impulses of local government, business, foundations, and individualconsumers that have proved successful for the large institutions.

Most of the financial information we received did not provide a detailedbreakdown of what is included in the "concessions" category. Those few casesfor which information is available suggest that greater diversification is takingplace. One zoo reported that "retail sales" accounted for roughly 67 percent and"parking" for 30 percent of gross concessions revenue in the late 1970s, but thatin the late 1980s "retail sales" were only 15 percent, "parking" was 45 percent,and "rides" were 40 percent. Another zoo, which reported that "rides" com-prised nearly 50 percent of gross concessions revenue in the late 1970s, nowreports "rides" at roughly 25 percent; the difference was that its "gift shop" hadincreased from 5 percent to 35 percent. It is clear that these institutions con-tinue to search for creative ways to enhance their revenue and that this processinvolves the development of new sources of commercial revenue (NP-goods)as well as an extension of existing sources.

Although gift shops generally have excellent potential as a revenue source,they need to be of sufficient size and strategically placed. Those at the LincolnPark Zoo are not; but even so, "visitor services," which includes food service,souvenirs, parking, and the like, accounted for 5.8 percent of net revenues infiscal year 1996. There is a need to move shops (and restaurants) toward the pe-riphery, where they can be open even when the zoo is not. The public is under-served in an area where the yield (profit) is close to 20 cents on the dollar.

Concessions account for 30 percent of the Glen Oak Zoo's gross revenues,which is about average for small zoos in Table 11.1. The space devoted to theyear-round gift shop has been increasing, and its revenue has been on the rise.Food service and pony rides, available seasonally, also provide commercial in-come. Phase 1 of the zoo's master plan calls for construction of a new gift shop;phase 2 for construction of a new concession compound, in recognition of itsrevenue-generating potential.29 Moreover, a large area to the side of the outdoorexhibits is being developed for rental purposes with the expectation that corpo-rations and other groups will rent it for evening functions.30 Another way in

2 9 In addition, phase 1 will construct outdoor habitats for those primates still in indoor cages. Phase2 will address the center core, with the main highlight being a giraffe exhibit in which the ani-mals can be seen at several different levels.

3 0 Special events (e.g., a corporate "day at the zoo" with special tours) ordinarily are one-dayevents that, if successful, can net as much as $25,000 for a larger zoo, less for a smaller one.In the overall scheme, this is a relatively small amount, but such events are being encouraged,as are fee-based programs.


which such spaces are being used by zoos and aquariums is as sites for outingsby conferences meeting in the city.31 Like most small zoos, Glen Oak is aggres-sively pursuing new ways to generate revenue through commercial channels.

Concluding remarks

As is true of many other nonprofit institutions, zoos and aquariums increasing-ly have found it desirable to cast a broader net in their search for revenue. Thecause has been the reweighing of organizational goals and the more costlymeans adopted to attain them. Costs have increased as zoos converted theirplant and equipment toward space more compatible with their increased empha-sis on the goal of species preservation (a PC-good in the context of the modelpresented in Chapter 3). While admission fees are a feasible source of revenuewith respect to the customary mission of animal display (a PP-good), beneficia-ry user fees are not feasible with respect to species preservation. The inelastic-ity of demand we found for viewing animals implies that additional revenuesto fund the collective good can be obtained by raising admission fees. Since thevast majority of these facilities are nonprofits, however, those fees typically areset not to maximize profit (or even revenue), but at lower levels to achieve moresocially oriented goals, such as facilitating attendance by the poor. Private do-nations and public subsidies, the other potential sources of funding for collec-tive goods, have decreased, thereby increasing the pressure to find new revenuesources. This has led zoos and aquariums to look to new and expanded com-mercial activities (NP-goods). These activities, from selling stuffed animals toleasing space for private parties, lie outside the stated goals of these institutions,but they are understandable as complementary with unrestricted donations asmeans to finance provision of PC-goods.

What are the consequences? The increased reliance on concessions has notled zoos and aquariums astray as yet. Most of these institutions are headed byindividuals trained in animal management, but there is a need for this group todevelop a stronger sense of the institution as a business. Financial constraintsundoubtedly will prove more binding in the future as these institutions play anincreasingly active role in species preservation, but it is doubtful that this wouldevolve to where financial concerns dominate animal ones. Gift shops and res-taurants will remain secondary interests, in many cases outsourced to privatefirms. Raising admission fees will remain a viable way in which to increase rev-enues, but, given the fact that so many of these institutions are in the public sec-

3! Additional revenue-generating activities include music performances on the grounds (jazz andcountry are the most common, with at least one zoo offering opera) and the sale of liquor, wine,and beer.


tor, there will be weekdays when admission will be free. Low-income familieswill be inconvenienced, but not priced out of the market. Of greater conse-quence is the question, What happens if zoos and aquariums are successful inpreserving species? Are the profits from stuffed animals and hamburgers anequitable way to cover the costs of deceleration on the road to mass extinction?

CHAPTER 12

Commerce and the muse:Are art museums becoming commercial?

Helmut K. Anheier and Stefan Toepler

Introduction

In this chapter, we examine the commercialization of art museums by lookingat trends and patterns of revenue and expenditures as well as other financial in-dicators over time. Commercialization involves greater reliance on earned in-come relative to other forms of revenue, such as donations and governmentgrants. In the case of museums, earned income includes fees, sales, and chargesthat are related or unrelated to the charitable purpose of the organization. Wefirst explore possible reasons for, and characteristic elements of, commercializa-tion, and then look for patterns that might help explain whether (and if so, why)art museums have become more commercial over time. For the main part ofthis analysis, we make use of the tax returns (Forms 990) of the largest U.S.museums at five different points in time between 1982 and 1992.

Over the past three decades the museum field has undergone major changes,as museums slowly transformed from secluded temples of culture to more pop-ular institutions with considerable public appeal (d'Harnoncourt et al. 1991,54-5). Beginning in the 1960s, both the number of museums and the numberof visitors to them increased considerably. Attendance at art museums rose fromtwenty-two million in 1962 to forty-two million in 1975 to seventy million in1986 and to over seventy-five million in 1988 (Heilbrun and Gray 1993,170).The number of art museums expanded as well. Indeed, one out of two Ameri-can art museums was founded after 1970: 13 percent were founded in the 1980s,and 37 percent in the 1970s, whereas 18 percent were created during the 1960s,and only 31 percent of all art museums were founded before that decade (AAM1992, 64).

We would like to thank Eric Crutchfield and Virginia Hodgkinson of Independent Sector for mak-ing the data for this study available to us, as well as the participants in the commercialism work-shops held at Northwestern University.

233

234 Helmut K. Anheier and Stefan Toepler

The general expansion of the field was accompanied by a concomitantchange in the resource structure of museums. Until midway through the twen-tieth century, art museums predominantly relied on endowments and individ-ual patrons to support their ongoing operations and acquisitions of new artobjects. Since the 1960s, however, new funders - namely government, founda-tions, and corporations - have entered the field. This shift placed new demandson museums: The new funders, while supporting the general expansion of mu-seum activities overall, were less inclined to help finance ongoing operationsand current expenditures. Instead, they preferred to support special activitiessuch as exhibitions and outreach programs or specific acquisitions. This fre-quently resulted in a mismatch between the broader array of museum functionsand the usually more narrow range of objectives pursued by funders. Finan-cially, this mismatch could make it difficult for museums to meet operatingexpenditures, which, over time, may lead to chronic underfunding of some coreareas. As Temin (1991,179) remarked, some museums "appear to be in a stateof perpetual deficit."

Despite the growth in public and private donative support to the arts formuch of the past three decades, museums have nonetheless faced often latent,sometimes overt, commercialization pressures. As one observer holds, "in thelast fifteen years or so many of those charged with governing nonprofits aremore market-oriented than mission-oriented. . . . Mission gets blurred or is soelastic it can be construed to accommodate whatever might be the latest income-producing technique.... The trend is now more than evident in the museumworld" (Malaro 1994,113). This is not to say that museums have never engagedin commercial activities in the past. To the contrary, the Metropolitan Museumof Art in New York, for instance, opened a museum shop as early as 1908 (Weis-brod 1988,109), and started a mail-order operation in 1921. However, such ear-ly ventures remained the exception rather than the rule. In contrast to other artsorganizations, such as theaters, orchestras, operas, or dance companies, whereearned income typically accounts for half or more of total operating expendi-tures (DiMaggio 1986a, 84), most art museums, by virtue of their charters,could not easily raise earned income by charging higher admission (user) fees.

Nonetheless, it now appears that museums have increasingly become inter-ested in generating income from fees, charges, and sales. One indication of thisgrowing interest is a burgeoning of marketing-related articles that has occurredin museum trade journals since the mid-1980s, both in the United States andabroad (Rentschler 1996; see also Blattberg and Broderick 1991). Another indi-cator is the expansion of museums shops, both on-site and off-site (typically inshopping malls). Moreover, many museums are either establishing, remodeling,expanding, or upgrading cafeterias and restaurants to make them more attrac-tive to a wider clientele.

Commerce and the muse: Art museums becoming commercial? 235

The business professionalization of museums may also cause internal ten-sions between curators, the traditional guardians of art museums on the onehand, and museum directors and administrators responsible for the financial vi-ability of the institution on the other. Historically, museum directors were cho-sen from the ranks of curators. Recently, however, this unity of artistic and ad-ministrative responsibility has been weakening, as business background andorientation have become primary qualifications for museum directorships. In-creasingly, museums are experiencing difficulties recruiting art historians aschief executives willing to take on the fund-raising and financial responsibil-ities the position entails (Riley and Urice 1996).

The commercialization trend of museums raises important public-policyquestions. Critics argue that the increasing business orientation among muse-ums may deflect their original mission and, in the long run, endanger the basisfor their preferential tax treatment. This point is forcefully put forward by MarieMalaro, who states that "growing commercialism . . . is changing the whole em-phasis of what museums do, and it has all sorts of very serious ramifications.. . . It calls into question . . . why is this organization a nonprofit and why shouldit retain many of its privileges" (NPR 1996b).

Still, how pronounced is this commercialization trend? Is it a general trendor does it affect only some types of museums? What are the forces behind com-mercialization, and what forms does it take? To answer these questions, we firstlook at the funding environment in which museums operate.

Commercialization pressures

Compared to other areas of nonprofit activity, the reasons for a trend towardcommercialization among art museums are less obvious. Unlike health care, forexample - the most commercial field of nonprofit activity - museums have notbecome dependent on government as their major financial customer; after all,government does not reimburse museums for services rendered to visitors. Un-like science and technology museums, art museums do not directly competewith commercial theme parks and similar organizations that increasingly shifttheir focus from pure entertainment to "edutainment" (Mintz 1994). Since highculture is typically prone to market failure, competition between art museumsand for-profit firms is typically absent.

As mentioned in the preceding section, until the mid-twentieth century, vir-tually all museums depended for support, both in terms of operating costs andacquisition budgets, largely on individual private patrons. Foundations had nosignificant presence in arts funding until the late 1950s, when the Ford Founda-tion and a few other large grant makers established special programs to supportthe arts and culture. The federal government followed in 1965 with the estab-


lishment of the National Endowment for the Arts (NEA), which in turn fosteredsimilar programs at state levels. Moreover, the Institute of Museum Serviceswas established in 1976 to provide direct operating support to all types of mu-seum, albeit to a limited extent.

The 1970s, and even more so the 1980s, saw considerable growth in supportfor the arts from a number of sources, particularly government, foundations, andcorporations. This influx of new financial resources into the arts sector has beensubstantial: Private giving increased from approximately $0.5 billion in the mid-1960s to about $10 billion (current dollars) in the mid-1990s (AARRC 1997).By the same token, NEA appropriations increased from less than $3 million in1966 to a high of $176 million in 1992, leveraging substantial additional stateand local funding. Total government support in the mid-1990s was estimatedat more than $1 billion (Cobb 1996).

How do the increases in government and foundation support to the arts sincethe 1960s relate to possible commercialization pressures among art museums?There are four major reasons why commercialization may have grown amongmuseums as the overall funding base broadened:

First, the availability of new resources frequently supported capital invest-ments, specific acquisitions, and special projects rather than general funds to-ward operating costs. The result was a significant expansion of the museumfield: Not only did the number of museums increase during this period, but ex-isting museums expanded as well. The 1988 Museum Survey of the AmericanAssociation of Museums reports that close to half (47 percent) of all art muse-ums reported major renovations, and nearly 13 percent moved to new buildingsites (AAM 1992,76).

Second, art museums are subject to so-called cost disease. Using the per-forming arts as an example, Baumol and Bowen (1966) suggest that the arts area rising-cost industry, in which increases in labor cost are not easily compen-sated by gains in productivity. Art museums also fit that description. Becausetechnological advances can substitute for labor only to a limited extent, muse-ums may face income gaps over time. Of course, new technologies have had,and are having, considerable effects on different aspects of museum operations,but personnel costs remain a major part of operating expenses, particularly forlarger museums. As Robert Buck, Director of the Brooklyn Museum, has putit: "We are nine-tenths salaries. To keep the actual operation going with thenumber of educators we need, the number of curators, the number of guardsis enormous . . . , " and private patrons often prefer to support named collectionsand buildings rather than "secretary salaries and supplies" (NPR 1995).

Third, the objectives of funders may not be in tune with the perceived finan-cial needs of museums. In a study on the influence of funders on museum exhi-bitions, Alexander (1996) suggests that the goals of individual donors, founda-tions, corporations, and government agencies frequently run into conflict with


those of the museum staff, most notably the curators. Some exhibition projectsfavored by curators fail to attract sufficient interest among private and institu-tional donors, which in turn forces museums to finance such exhibitions inter-nally, at least in part. Moreover, staging exhibitions may be the most visible,but it is only one of five major activities museums perform: the other four - col-lection, conservation, research, and interpretation - are typically more difficultto support financially (Noble 1970). (At least the latter two are the collective-type activities highlighted in Chapter 3.) For instance, foundation funding for"collection and conservation" amounts to an insignificant share of total foun-dation giving to the arts, and has even declined over the past decade (Weberand Renz 1993, 89). Accordingly, despite a greater availability of donativeresources, museums may still face difficulties attracting outside funding forcertain core activities, forcing them to develop, and manage more effectively,internal resources including commercial income. This, in turn, affects the man-agerial behavior of the museum administration. Indeed, noncuratorial staff withbusiness training (and business-oriented attitudes) is becoming increasingly fre-quent in museum management and administration, changing them from "schol-arly to managerial organizations" (Alexander 1996, 95).

Fourth, it now appears that government support for the arts has reached itspeak. After significant growth in the 1970s, appropriations for the NEA, whichhad been crucial in stimulating both private and other governmental spending,became relatively stagnant during the 1980s and early 1990s, then were sharplycurtailed in 1996 (NEA 1996). At the level of state government, art appropri-ations continued to grow until the late 1980s, before declining in the recessionof the early 1990s (NASAA 1992). Starting in 1994, however, state support be-gan to rebound (Institute for Community Development and the Arts 1997).

Thus, four factors - increased operating costs due to past expansions, inher-ent difficulties to control costs, the divergence between funding needs and avail-able support, and recently declining government funds - may have led muse-ums to follow the general trend toward greater commercialism that has beenobserved in other parts of the nonprofit sector. What are the different commer-cialization options that museums may have adopted over the past decade?

Commercialization options

There are several basic options museums can pursue to increase commercialrevenue. These range from admission fees, museums shops, parking garages,and sale of reproductions, exhibition books, catalogs, and postcards, to block-buster events around special exhibitions that focus on famous artists, populargenres, or en vogue periods. We briefly discuss some of the major commer-cialization options under the headings "Admissions" and "Ancillary Opera-tions."


Admissions

The setting of admission (user) fees remains a major problem for art museums.Perhaps most important, there are equity considerations involved. From the be-ginning, public education has been the main rationale for the establishment ofart museums, and their basic mission was to make the fine arts more accessibleto social groups otherwise excluded from art consumption, thereby improvingcultural tastes and promoting a general moral uplift (Zolberg 1986,184-5). Inkeeping with that mission, most art museums have traditionally not chargedadmission fees at all, fearing "that substantial entrance fees would prevent therelatively poor from partaking" (Heilbrun and Gray 1993,179).1 (See Chapter4 for analysis of pricing policy and distributional goods.) Indeed, in a surveyconducted by the American Association of Museums, only 36 percent of artmuseums reported charging any admission fee, the lowest proportion amongmuseums in general (AAM 1992,133).2 Moreover, many museums that chargesuch fees prefer suggested contributions over fixed prices, thereby voluntarilyrestricting their ability to exclude customers neither willing nor able to pay fullfees.3 Other museums, while charging principally fixed or suggested admissionfees, waive them either on specific days or for specific groups, such as school-children or senior citizens. In recent years, the fees and suggested donationshave been increasing (Wade 1997).

However, restrictions in setting fees do not necessarily apply to special exhi-bitions. Beginning in the late 1970s, museums have gradually come to rely onblockbuster exhibitions, that is, exhibitions with high popular appeal, to gener-ate additional income. For example, according to its 1995 annual report, thePhiladelphia Museum of Art yielded record income from admissions as well asfrom museum shop and restaurant sales through the greatly successful exhibi-tion "From Cezanne to Matisse: Great French Paintings from The Barnes Foun-dation." In fact, museum finances have partially become subject to what hasbeen termed blockbuster economics, whereby museums apply profits from theblockbuster toward deficits in their current operating budgets (Parker et al.1991,61). This is precisely the kind of commercial revenue cross-subsidizationthat the model presented in Chapter 3 predicts.

1 A recent survey among nonvisitors suggests that admission fees are indeed a major subjectiveentrance barrier, especially among low-income groups and for those living outside major urbancenters; see Kirchberg (1996).

2 The AAM survey in 1988 included both public and private museums. Since public museums areoften free by charter, the percentage of nonprofit art museums charging entrance fees may actu-ally be higher. The percentage of all types of nonprofit museums charging admission is close to62 percent (AAM 1992, 134).

3 According to the AAM survey, of those art museums that have entrance fees, roughly 43 percenthave fixed fees and 17 percent suggested fees. The remaining nearly 40 percent of museums withadmissions did not indicate whether admissions are fixed or suggested (see AAM 1992, 133).


Ancillary operations

Some of the major activities museums have undertaken to increase commercialincome are expansions of museums shops, licensing arrangements with com-mercial firms, arrangements with travel agencies, and solicitation of corporatesponsorships (Malaro 1994,113-14). Other activities include the expansion ofrestaurants and cafeterias, renting out of museum facilities for special events,and parking operations.

Judging from the literature, however, the main thrust of commercial activ-ity has been on retailing and merchandising, as museums not only increasedstore space but also moved off-site to shopping centers and malls. As of 1997,the Metropolitan Museum of Art, for instance, operates fifteen off-site stores ineight states and nine countries abroad. Similarly, the Art Institute of Chicagooperates three off-site stores, and the Museum of Fine Arts, Boston, has five.Moreover, some of the largest museums have recently moved into the catalogbusiness, following the early example of the Metropolitan, which led the wayin 1921. The Art Institute of Chicago started its catalog business in 1982 withfive thousand copies; ten years later it mailed over two million (Trescott 1994).

It seems as if merchandising has become a major economic obligation, par-ticularly for larger museums. Such sales operations, for instance, accounted forclose to 46 percent of total revenue of the Metropolitan Museum in 1995, andfor nearly 60 percent of the Museum of Fine Arts, Boston. However, revenuesimply costs. In the case of the Boston museum, merchandising revenue did infact become more expensive in recent years: Although total revenues increasedby about 25 percent between 1991 and 1995, and merchandising revenue by 33percent, merchandising cost jumped by 43 percent during the same period (1996annual report). In other words, in examining commercialization trends amongart museums, it is important to examine the weight of such commercial activi-ties in the context of other cost and revenue categories over time (see Chapter5). We follow up on this question in the next section.

Are museums becoming more commercial?

Are we experiencing a massive trend toward commercialization among Amer-ican art museums? Before we answer this question, let's take a look at the dataused for our analysis.

Data sources

The data are based on a multistage sampling procedure that begins with the IRSBusiness Master File. From this data set, the IRS selects the IRS Form 990 TaxReturn File of Nonprofit Organizations, which includes all nonprofit organiza-


tions filing Form 990, the principal tax-return form used for tax-exempt enti-ties. This data set forms the basis for the IRS-SOI Sample, which includes or-ganizations with assets of $10 million or more that file Form 990 tax returns,and then a random sample, based on asset size, from the rest of the nonprofitorganizations in the IRS Form 990 Tax Return File. Assets sizes are groupedinto seven classes ($10 million or more; $5 million to under $10 million; $2.5million to under $5 million; $1 million to under $2.5 million; $500,000 to un-der $1 million; $250,000 to under $500,000; and under $250,000). The prob-ability of being sampled is relative to the asset bracket of organizations. Non-profit organizations in lower asset-size brackets are less likely to be sampledthan those in upper asset brackets.

The IRS-SOI Sample of tax-exempt organization formed the basis for thecompilation of the data on American museums in this analysis. From the IRS-SOI Samples for 1982, 1987, 1989, 1990, 1991, and 1992 we selected all artmuseums, which numbered in the range 190-200 for each period. Each samplerepresents the bulk of economic activity of American museums in the givenyear. Using 1992 as an example, the 198 organizations in the sample accountfor 60 percent of all donative income, 90 percent of assets, and 80 percent ofall expenses for the total of all 2,003 art museums in the IRS Form 990 TaxReturn File of Nonprofit Organizations (Hodgkinson and Weitzman 1996,247,tab. 5.7).

With the help of these six annual samples, we generated a consolidated "pan-el" data set that included all museums that were part of either the 1992 or the1991 samples plus two other additional prior periods. Thus, we excluded mu-seums that were present neither in the 1991-2 samples, nor reporting on twopre-1991 periods. This yielded a sample size for the consolidated panel data setof 263 museums. With the help of the consolidated panel data set, it is possibleto track changes in patterns and trends over time more precisely than would bethe case in a cohort design, let alone cross-sectional analysis.

Commercialization trends?

Our data suggest that art museums have not become significantly more com-mercial in recent years. This does not mean, however, that no commercializa-tion trend can be identified at all; in fact, as we shall see, half of all museumsincreased their share of either related business income or excluded unrelatedbusiness income, and every sixth museum increased unrelated business incomebetween 1990 and 1992.

However, these commercialization trends take place against the backgroundof relative stability in the overall patterns of total museum revenue and expen-diture. As Table 12.1 shows for the sum total of all 263 museums, the relative


Table 12.1. Revenue sources for art museums, as share (%) of total revenue,1982-92

Revenue category

Direct public supportIndirect public supportGovernment grantsProgram revenue^Membership duesInterest on savingsDividendsNet rental incomeOther investment incomeSales of assetsNet income fund-raisingGross profit from salesOther revenue

Total revenue (mil. curr. $)N =

1982

32.00.56.9

13.77.63.7

12.80.30.29.30.17.06.0

77871

1987

29.90.48.3

15.17.12.2

10.70.40.59.70.39.75.6

1,522185

1990

35.40.5

12.912.96.11.9

12.00.20.35.40.58.93.0

1,872201

1991

34.40.5

12.612.25.62.0

10.80.30.19.40.48.13.5

2,107199

1992

34.80.6

11.613.35.31.3

10.40.30.2

10.90.48.22.7

2,320198

^Includes admission fees.Source: Consolidated data set based on IRS-SOI samples.

shares of revenue sources have changed little over the ten years between 1982and 1992. Direct public support (i.e., private grants and donations), accountedfor about one-third of total revenue for each period, and changes are usuallywithin the 1- or 2-percentage-point range. The same can be said for programrevenue (admissions, royalties, etc.) and sales: Between 1982 and 1992, thesetwo revenue sources that are most closely linked to commercialization grewsubstantially in nominal dollars, but changed very little in relative terms (Table12.1). Program revenue accounted for 13.7 percent of total museum revenue in1982 and roughly the same share a decade later; gross profits from sales repre-sented 7.0 percent of total revenue in 1982 and 8.2 percent after ten years. Gov-ernment grants are the only revenue source that shows a general and significantincrease between 1982 and 1992, moving up from 6.9 to 11.6 percent. Where-as in 1982, government grants represented about half of the relative size of pro-gram revenue, both sources are about equal in their financial weight by 1992.

The financial size of the museums included in the sample increased substan-tially during this period: Between 1987 and 1992, total revenues rose by 65percent from $1.5 billion to $2.3 billion (Table 12.1). This means that govern-ment grants funded the growth in museum revenue disproportionately to other


Table 12.2. Expenses of art museums by major category, as share (%) oftotal expenses, 1982-92

Expense category

Program servicesManagement, generalFund-raisingPayments to affiliatesTotal expenses(mil. current $)N =

1982

76.820.32.80.0

55977

1987

74.721.9

3.20.2

1,121185

1990

74.122.2

3.60.1

1,487201

1991

74.621.9

3.60.0

1,645199

1992

76.020.2

3.80.0

1,783198

Source: Consolidated data set based on IRS-SOI samples.

sources. By implication, the expansion in total revenue was not the result of rel-ative increases in commercial income. Although museums took in more revenueoverall, including all forms of commercial income, only government grantsgrew both absolutely and relatively. For the period under consideration, this in-crease in government grants reflects the growing engagement of government infunding the arts.

How can we explain this pattern of stability and the general absence of pro-nounced commercialization trends? The expenditure side of museum operationsis one area to look for answers. Perhaps the structure of museum costs has re-mained relatively stable as well, making it less likely for museums to venturemassively into new forms of revenue. Indeed, as Table 12.2 suggests, museumsas a whole experienced virtually no change in the composition of major costitems: Program services, management and general administration, and fund-raising costs still account for the same relative cost shares in 1992 as they didin 1982 and the years in between, even though expenditures rose from $1.1 bil-lion in 1987 to $1.8 billion in 1992. No major shifts in expenditure patternsseem to have propelled museums to seek additional or alternative revenue fromcommercial market sources.

A second possibility for the observed stability can be found in changes inthe relationship between revenue and operating expenditure. We could assumethat a relative decline in revenues compared to growing expenditures wouldpush museums to pursue more commercial strategies over time. Indeed, set-ting the cost and income levels in 1982 to unity, we see that revenues increasedslightly more rapidly than operating expenditures until 1989, but have laggedbehind since then, with increases in expenditures outpacing growth in revenue.At the same time, the revenue-expenditure ratio declined from 1.62 in 1982 toabout 1.32 for the period 1991-2, with the biggest drop occurring between 1989


and 1990. In other words, whereas museums had, on average, $162 in revenuefor $100 in operating expenditure in 1982, this buffer was cut in half a decadelater. Even though 1992 revenue-expenditure ratios still show an aggregate rev-enue surplus, they nonetheless mean that museums have fewer dollars left foracquisitions and collections, as well as for investments in buildings and othercapital expenditures. Moreover, museum curators have less funds available in-house to stage exhibitions independent of special support from outside.4 Thus,although museums do seem to be facing a somewhat tighter financial situation,it is not one tight enough to pressure them to rethink, at fundamental levels, theirrevenue and expenditure patterns in view of the alternative of generating great-er commercial income.

A third reason for the stability in revenue shares can be found in the rela-tionship between assets and liabilities over time. We could argue that currentaccount categories remained stable because there was little change in assetsand liabilities that would have propelled museums to restructure their revenuesources and expenditure patterns. As Table 12.3 shows, assets did increasesubstantially between 1987 and 1992, from $4.8 billion at the beginning of 1987to $9.2 billion five years later; liabilities, however, grew even faster, more thandoubling from $639 million in 1987 to $1.5 billion in 1992.5 Taken together,the trends in assets and liabilities mean that museums became more wealthy andmore indebted at the same time, as the median values indicate: The median end-of-year asset value (not shown in Table 12.3) jumped from close to $12 millionin 1987 to $18.5 million in 1992. Likewise, median liabilities grew even morein relative terms, from $0.7 million to $1.3 million. Thus, although there is ageneral trend toward increased assets, they seem to have been increasingly fi-nanced through some form of liability: Asset-liability ratios declined overallby nearly 3 points from 8.8 to 6.2. Though most museums have accumulatedmore assets in recent years, and even more in terms of liabilities, our data showthat this trend is most pronounced among the larger museums.

What does this trend mean for commercialization pressures? For one thing,growing liabilities translate into payments at some stage, most likely reducingthe amounts available for operating and capital expenditures. If this is the case,museums may have to seek additional sources to increase revenue even furtherin the future. However, we should keep in mind that during the same period inwhich museums grew in terms of assets and liabilities, only government grantsshowed a significant increase in their relative and absolute share of total reve-nue; all other sources remained stable in relative terms. With government grants

4 The finding that the financial "buffer" is crucial for museums is also supported by Alexander's(1996,60) study of thirty large art museums. She reports that about 80 percent of all art exhibi-tions are internally funded; only 20 percent involve outside support from foundations, corporatesponsors, or government agencies.

5 Figures in Table 12.3 are adjusted for inflation by applying a price deflator for 1989 and onward.


declining in the late 1990s, we should therefore expect that the picture of rela-tive stability portrayed in Tables 12.1-12.3 may no longer prevail in the future.Thus, commercialization pressures among art museums are not only the resultof likely government cutbacks; they are also closely related to a relatively longperiod of expansion of museum operations and holdings.

Changes in income-producing activities

Although the revenue and expenditure structure of museums has shown remark-able stability during the growth years between 1982 and 1992, there are signsthat changes are under way. Before analyzing what these changes are, howev-er, it is useful to consider the stability even in the types of income-producingactivities in which museums typically engage. Table 12.4 shows that relatedbusiness income declined somewhat, from 28.0 percent in 1990 to 26.4 per-cent in 1992; excluded business income grew more substantially from 21.6percent to about 25 percent, whereas unrelated business income remained sta-ble. Museums primarily engage in equal shares of related and excluded busi-ness income, and only to a very limited extent in taxable activities raising rev-enue unrelated to the charitable purpose of the organization. In fact, less than$2 of every $100 in museum revenue comes in the form of unrelated businessincome (UBI). As was pointed out in Chapter 5, however, reported UBI mayunderstate substantially the actual income distribution of unrelated business ac-tivities.

The composition of income-producing activities associated with commer-cialization also shows remarkable stability in terms of the various income cate-gories (Table 12.4). Percentages have changed little over the three years from1990 to 1992. Some areas often associated with a general trend toward com-mercialization have even experienced slight declines: Membership income inthe form of dues and related charges dropped somewhat in relative terms, asdid sales, both perhaps as a result of the 1991-2 recession. Other categories,namely fund-raising income and program services, show slight increases.

Of course, these data are at a relatively high level of aggregation, since theyencompass both increases and decreases in the various income-producing cate-gories across a large number of museums. If, however, we look at these changesat the organizational level, we find that the relative frequency of increases in allforms of business income is higher than that for decreases. Table 12.5 indicatesthat slightly more than half of all museums increased their share of total rev-enue from related business activities during 1990-2, while about a third report-ed decreases in this income category. For the great majority of these cases,changes are less than 10 percent, or under 3 percent annually for the three yearsunder consideration. That is, not surprisingly, museums rarely became drasti-cally more commercial or less commercial in a single year; rather, they changed

Table 12.3. Assets and liabilities of art museums, 1987-92

Net assets, beginning of year(% of 1982)

Excess/deficit for the year(% of 1982)

Other changes(% of 1982)

Net assets, end of year(% of 1982)

Total assets, end of year(% of 1982)

Total liabilities, end of year(% of 1982)

Assets-liabilities ratioN =

1987

4,770,468,772(100)

402,274,264(100)

(156,581,926)(100)

5,016,161,114(100)

5,655,656,297(100)

639,495,179(100)

8.84185

1989

5,345,997,894(112)

465,317,730(116)

(27,235,113)(17)

5,784,080,511(115)

6,544,895,984(116)

760,815,249(119)

8.60196

1990

5,730,700,781(120)

365,142,027(91)

(36,213,323)(23)

6,059,629^41(121)

7,048,938,775(125)

989,309,124(155)

7.13201

1991

6,399,563,864(134)

447,564,029(111)

136,484,076(-87)

6,983,611,967(139)

8,373,676,594(148)

1,390,065,302(217)

6.02199

1992

7,204,921,033(151)

527,087,798(131)

6,357,254

M)7,738,366,076

(154)9,221,135,696

(163)1,482,769,682

(232)6.22

198

Note: Figures are adjusted for inflation.Source: Consolidated data set based on IRS-SOI samples.


Table 12.4. Art museums' related, unrelated, and excluded income-producingactivities, selected sources, 1990-2

Program servicesUnrelatedExcludedRelated

GovernmentUnrelatedExcludedRelated

MembershipUnrelatedExcludedRelated

Fund-raisingUnrelatedExcludedRelated

Pet. of total revenue

1990

12.80.20.8

11.80.20.00.00.26.10.00.25.9

0.50.00.20.3

1991

12.10.20.8

11.10.20.00.00.25.60.00.25.5

1.00.00.80.3

1992

13.20.20.8

12.20.20.00.00.25.30.00.25.11.00.10.20.7

SalesUnrelatedExcludedRelated

OtherUnrelatedExcludedRelated

Income-prod.UnrelatedExcludedRelated

yv =

Pet. of total revenue

1990

8.90.10.58.33.01.51.10.5

activs. 51.31.8

21.628.0

201

1991

8.00.10.47.53.51.31.01.3

52.41.6

24.426.5

199

1992

7.40.10.47.02.91.31.20.5

53.01.6

25.026.4

198

Source: Consolidated data set based on IRS-SOI samples.

gradually. A similar pattern holds for excluded business income, but with a sub-stantially higher proportion (8.3 percent) reporting increases of 10 percent ormore, reflecting the more frequent expansion of revenues from cafeterias, park-ing facilities, and the like.

Table 12.5 also indicates that the pattern for unrelated income is very differ-ent from the other two revenue sources involving sales. Most museums - overthree-quarters - did not change the relative proportion of UBI during the three-year span, and those that did all reported increases of less than 10 percent. Al-though about twice as many museums increased as decreased their share ofunrelated income, we should recall that the overall share of UBI is less than 2percent of total revenue. Thus, even if museums increase unrelated business ac-tivity, the net revenue impact is small.

Conclusion

This chapter explored the commercialization of American museums. Specifical-ly, we looked at the patterns and trends of commercialization, and the reasonsthat lie behind it. In the period under consideration here (1982-92), art muse-


Table 12.5. Changes in income-producing activities for museums, 1990-2

Change, 1990-2

Reduction > 10%0.1% < Reduction < 9.9%No change0.1% < Increases 9.9%Increase > 10%Total

Share of business

Related

4.131.214.146.54.1

100.0

income (%)

Unrelated

0.07.7

76.316.00.0

100.0

Excluded

4.730.016.640.4

8.3100.0

Note: N = 169.Source: Consolidated data set based on IRS-SOI samples.

urns as a whole experienced a major expansion in terms of revenue, expendi-ture, assets, and liabilities. However, overall composition of museum revenuehas been relatively stable, with few changes in the relative shares of variousincome sources. The one exception is government support, the only revenuesource that showed disproportional growth during this period. Against this back-ground, our analysis suggests that no general and pronounced trend towardcommercialization of museum operations has taken place, although some indi-cations of increased commercial activities are nonetheless identifiable. Muse-ums seem to become slowly and gradually more commercial over time, if at all.This is especially true with regard to excluded and unrelated business income,the latter generally being negligible and of minute importance. The subtle com-mercialization trend that seems to be occurring now, however, does so on theheels of a major expansion of the museum field in the 1980s: Every second mu-seum increased its related-business-income share of total revenue during 1990-2, though, for most museums, these changes are less than 3 percent annually.

All in all, our analysis suggests that commercialization in the museum fieldis not as grave an issue at present as it is often perceived in the field and in theliterature. How can we reconcile this difference between the general perceptionand what is borne out in the financial data? One explanation would be that com-mercialization on a grand scale is more prevalent among very large museums.Some forms of commercialization, such as blockbuster exhibitions or extensiveretailing, require substantial financial investments; many smaller museums maybe unable to afford the high start-up costs involved. Since larger museums andtheir activities have higher visibility in artistic and art-policy circles, increasedcommercial activities among these art museums may have invited generaliza-tions to the field as a whole.


However, there might also be other reasons for the perceived commercial-ization of the museum field - reasons not necessarily reflected in the financialdata that formed the basis of our analysis. For example, Malaro (1994) pointsto the possibility that an increased internal business orientation, rather than theuse of commercial means of income generation alone, is at the heart of the com-mercialization of art museums. Similarly, Alexander suggests that the influenceof management on defining the mission of museums made them "less elitist,more popular institutions that are increasingly geared toward income and fund-ing" (Alexander 1996,95). With the traditional balance of power between cura-tors and administrators gradually shifting in favor of the latter, vital decisionsare increasingly made in view of economic viability rather than based on artis-tic merit and quality. Seen this way, commercialization in the museum contextdoes not necessarily mean that museums charge higher fees for their core ac-tivities or engage in new activities to increase sales, as is the case in other non-profit industries; what it could mean is that museums change their core servicesto make them more attractive to a broader range of clients and funders.

Based on our analysis, however, we can suggest that the rise of commercialmuseum management seems to coincide with the emergence of the second ma-jor shift in arts funding in less than four decades. The first, which took place inthe 1960s and 1970s, involved a shift from individual philanthropy to founda-tion and government support. In the second, which has merely begun, muse-ums are likely to put more emphasis on earned income from admissions, mer-chandising, and other ancillary activities. Art museums are embarking on thiscourse somewhat hesitantly, and appear frequently torn between "commerce"and the "muse" as seemingly contradictory options. Nonetheless, many art mu-seums are likely to face financial difficulties, as expenditures have risen fasterthan revenues, and increases in liabilities have outpaced those for assets. Thiswould explain why commercialization has evolved as a major policy option andconcern for cultural institutions in the latter part of the 1990s.

CHAPTER 13

The funding perils of theCorporation for Public Broadcasting

Craig L. LaMay and Burton A. Weisbrod

Introduction and perspective

On November 5, 1967, President Lyndon Johnson signed into law the PublicBroadcasting Act, creating the Corporation for Public Broadcasting (CPB). Thepurpose of the act, Johnson said, was to "renew the promise of the vision in theword 'television.'" Johnson described the new CPB as "the people's corpora-tion. Fully independent. Broadly representative. Supported by public and pri-vate money. Free from interference by private interests, government agency,or political party" (Johnson 1967).

"Public" radio and television are industries that illustrate well the interplayof outputs and mechanisms to finance them. The collective-goods aspects oftheir broadcast outputs are unusually large; the process of production and distri-bution of programs is such that the incremental costs of serving additional lis-teners or viewers is essentially zero, and it is costly, though not impossible, tolimit listener and viewer access (Owen and Wildman 1992).

These are the precise conditions of production under which financial prob-lems emerge. With user fees to listeners being infeasible, dependence on contri-butions, gifts, and grants becomes great. Because grants from government havefallen in recent years, public broadcasting has had to pursue all of its principalfinance options: It has moved aggressively to expand private donations, espe-cially through "memberships"; and it has worked to increase revenues fromeach of its two potential sources of sales - user fees from its advertisers and"underwriters," and ancillary services that are complements of the broadcastprocess.

Every potential revenue source - government grants, private donations, pri-vate sales, or anything else - represents a willingness to pay for some partic-

We thank Michael Janeway, James G. Webster, and Steven S. Wildman for comments on an ear-lier draft.

249

250 Craig L. LaMay and Burton A. Weisbrod

ular output, and the ability of its provider to tap into that willingness. Thus thedecision by a provider of public radio or television programming to reach outto a particular source of funds implies that programming will be of the type thatappeals to that source. Decisions on the nature of programming cannot be madeindependently of the availability of funding for that package of service. Thecharacter of the services being produced and the success of a public broadcast-er to generate funding for those services are inextricably entwined; dependingon the nature of the "outputs," it can be easy, difficult, or even impossible to fi-nance them through sales to "consumers" - in this case, listeners and viewerson one hand, and commercial advertisers on the other.

The history of public radio and television makes clear that their goals wereof a particular sort: (1) to provide programming that was "different" from whatprivately owned stations were providing, and (2) to broadcast them differently- without "commercial" interruptions. That is, the concept of public broad-casting rested on the belief that the private market, though extensive, had failedto provide certain kinds of outputs for which at least some people were willingto pay. How actually to generate the financial support, however, was and is thekey problem.

Potential revenue sources in public broadcasting

The multiproduct framework, set forth in Chapter 3, for analyzing nonprofits'responses to cutbacks in government contributions, will guide our examination.As was explained there, any nonprofit organization can generate revenue fromeither of two basic sources, donations and sales. The collective-type characterof radio and television broadcasting essentially precludes selling services to lis-teners and viewers or charging user fees. Some readers will immediately objectthat a license fee - a tax on television and radio receivers - has traditionallybeen the finance mechanism for most public broadcast systems in Europe. Sucha fee accounts for 80 percent of the revenues of the world's most well-knownand well-regarded public broadcaster, the British Broadcasting Corporation(BBC), and those revenues are substantial - approximately $2.5 billion of theBBC's $3.2 billion budget in 1996 ("Time to Adjust Your Set" 1995). Howev-er, the fee is on receivers, it must be remembered, not programs; thus althoughfees are technically possible they are inherently inefficient. Moreover, no mat-ter where the fee is set, it will be too high for some people, thus excluding thoseviewers from watching programs despite the cost of serving additional viewersbeing zero.

In the United States, the dominant system of broadcasting has always beenprivate and for-profit, so license fees do not exist. For public broadcasters, then,remaining revenue options include donations from (1) government and (2) pri-vate sources; and commercial revenues in the forms of (3) user fees from orga-

The funding perils of the Corporation for Public Broadcasting 251

nizations purchasing access to broadcast time - that is, advertisers - and (4) saleof other, ancillary services that would be profitable.

Government grants

Government grants have been critical for public radio and television not onlyin the United States but throughout the world, and the fiscal pressure resultingfrom government cutbacks is also quite widespread. In Canada, for example, in1994, before the latest round of budget slashing, nearly 80 percent of the Cana-dian Broadcasting System (CBC) budget came from the federal government;the rest came largely from the sale of commercial advertising time. In 1996 acut of some 12 percent - $127 million - occurred, and by 1998, when the cutsalready announced will have taken place, grants will have dropped by about25 percent in current dollars, even as production and distribution costs increase(DePalma 1996). In the United States, the federal government has been consid-erably less central to public broadcast funding.1 In fiscal 1995, federal appropri-ations of $285.6 million constituted about 14 percent of CPB's $1.8 billion rev-enue from all sources ("The Struggle Over Federal Aid,. . ." 1996). A reviewof public broadcasting's income over nearly two decades, from 1972 to 1989,shows that tax-based income from all sources (federal, state, and local) declinedfrom 71 percent in 1972 to 46.8 percent in 1989. The federal contribution topublic broadcasting reached a maximum of 27.3 percent of revenues in 1980;since then, federal revenues have declined, as shown in Table 13.1.

Historically, public broadcasting has argued that the federal appropriation isessential to leveraging all other revenue sources. The steady lessening of feder-al appropriations has forced public broadcasters to adjust by some combinationof decreases in broadcast outputs and intensified efforts to increase revenuesfrom other sources.2

Private donations

Another potential revenue source, private donations, has limited promise. Thefamiliar difficulty of relying on private giving applies to all collective goods;

1 For comparison purposes, government spending in the United States is $ 1.06 per capita for pub-lic television. Japan spends $17.71 per head for NHK, Canada spends $32.15 per head for theCBC, and the United Kingdom spends $38.56 per head for the BBC.

2 Federal funding of public television has always commanded bipartisan support, and normallyfunding has come in the form of three-year authorizations and two-year appropriations. Appropri-ations do not always match authorizations: The actual appropriation for 1993, for example, was$253 million, well short of the $285 million Congress authorized. In 1993, Congress authorizedsubstantial increases in federal funding for the years 1994-6: $310 million for 1994 (appropri-ation was $255 million); $375 million for 1995 (appropriation was $285 million); and $425 mil-lion for 1996 (appropriation was $275 million). In fiscal 1997, authorized funding dropped to$260 million, and in 1998, to $250 million; but the authorization for 2000 is $300 million.


since the broadcasts are available without a user fee, obtaining private dona-tions involves an act of altruism that runs headlong into free-rider behavior. Itis no wonder that public television and radio stations estimate that only some10 percent of viewers and listeners contribute (Rowland 1993, 182).

Airtime revenues

Revenue from the sale of broadcast time to private firms for commercial an-nouncements - a form of user fee - constitutes the financial lifeblood of pri-vate, for-profit radio and television; historically, however, it has been anathemato public stations. While private broadcasters have focused on the sale of air-time to business interests, public broadcasters have avoided this market. Pub-lic radio's and television's goals seemingly precluded the kinds, frequency, andtiming of commercials that are most attractive to business advertisers. Therewas some, but a limited, market for public stations' sale of commercial broad-cast time insofar as stations developed a reputation for trustworthiness, perhapsbecause of their nonprofit character, and some private firms wished to "pur-chase" an identification with that nonprofit broadcasting organization, even ifthe sponsorship was constrained in its form, content, and time positioning (e.g.,only before or after a program).

This form of commercial revenue surely holds the greatest financial poten-tial as public broadcasters search for revenue to offset loss of governmentgrants, but it is also the most problematic. The more public broadcasters' fi-nances come to mimic those of private firms, the weaker the nonprofits' abil-ity to carve out a unique social role and, hence, to justify any special treatmentthrough government grants, private donations, or tax exemptions and subsidiesas nonprofit organizations.

Taken together, private donations and the sale of broadcast time (in the formof program underwriting and corporate sponsorship) comprise more than halfof all income for public broadcasting, as shown in Table 13.1. Between 1972and 1989, private donations rose from 28.5 percent of CPB's annual income to53.2 percent. The meaning of this change is not certain, but two arguments standout. One is the argument made by former Public Broadcasting Service (PBS)President Lawrence Grossman, who has said that private donations, whetherfrom individual subscribers or large corporations, help to make public televi-sion truly "public" - not dependent on government. In this view, declining tax-based support and rising public support may pose operational difficulties, butnonetheless serve to confirm and enhance public broadcasting's mission. Thesecond argument is that private support by definition comes with strings at-tached, and that those strings affect public broadcasting's mission, perhaps ad-versely. Public broadcasters who seek program support from businesses andfoundations do in fact make compromises in program content, and anecdotal

The funding perils of the Corporation for Public Broadcasting

Table 13.1. Tax-based revenues and private revenues as sources ofU.S. public-broadcasting income, 1972-89 (in millions of dollars)

253

FY

197219731974197519761977197819791980198119821983198419851986198719881989

Federal

Total

59.855.667.192.3130.1135.3160.8163.2192.5193.7197.6163.7167.0179.2185.7243.0247.5263.9

(%)

(25.5)(21.8)(23.1)(25.3)(30.0)(28.1)(29.1)(27.0)(27.3)(25.2)(23.4)(18.2)(17.1)(16.3)(16.4)(18.8)(18.1)(17.3)

Nonfedera1

State & localtax-based

107.7127.3139.1156.6175.9191.3218.2245.5271.6277.5301.0318.3334.5358.4378.8389.2415.8451.3

(%)

(46.0)(50.0)(47.9)(42.9)(40.6)(39.7)(39.5)(40.7)(38.5)(36.1)(35.6)(35.4)(34.3)(32.7)(33.4)(30.1)(30.4)(29.5)

Private

66.871.984.3115.9127.3155.6173.4194.7240.7297.7346.6417.1472.8558.7569.5662.3704.1812.3

(%)

(28.5)(28.2)(29.0)(31.8)(29.4)(32.3)(31.4)(32.3)(34.2)(38.7)(41.0)(46.4)(48.5)(51.0)(50.2)(51.2)(51.5)(53.2)

Total

Sources

234.3254.8290.4364.8433.3482.1552.3603.5704.9768.9845.2899.2974.2

1,096.31,134.01,294.51,367.41,527.6

(%)

(100.0)(100.0)(100.0)(100.0)(100.0)(100.0)(100.0)(100.0)(100.0)(100.0)(100.0)(100.0)(100.0)(100.0)(100.0)(100.0)(100.0)(100.0)

Source: Corporation for Public Broadcasting.

evidence says that many worthwhile projects are scuttled, others never pro-posed. In practice, broadcasters make a determination that a program is worth-while, then identify sympathetic funders. Again, however, whether and how thatprocess affects a public broadcaster's mission is difficult to measure, primarilybecause, in the United States particularly, public broadcasting's mission is fa-mously ill-defined. Some years ago, a U.S. public-broadcasting executive, de-spairing over the steady schedule of British imports and nature programming,described the U.S. service as "English people talking and animals mating, andoccasionally vice versa." Given this problem of vagueness, it is very difficultto know whether public broadcasting's mission would change, or whether anychange would affect programming decisions, if public broadcasting's revenuesremained the same but simply came from a different mix of sources, or evenfrom sources that were entirely tax based. Even public broadcasting's most ar-dent defenders admit that the mission of the institution lacks clear and opera-tional articulation (Rowland 1993,166-7). Further complicating this problem


of mission and its implications for programming is that Englishmen and ani-mals are also regular fare, though at a price, on several cable channels.

Revenues from ancillary goods and services

A final source of expanded revenue involves sale of ancillary goods and ser-vices, other than broadcast time. In every industry the inputs used in the pro-duction of its preferred, mission-related activity have some potential for use inother activities, and these can be a source of revenue for cross-subsidizing themission-related effort. In this case the mission is to provide particular types ofprograms, without commercial interruptions and without commercial influenceover program content. The potential for profitable production and sale of ancil-lary goods varies among industries, as was discussed in Chapter 3 and as wasreflected by the variation, shown in Chapter 6, among industries in the extentof their unrelated business activity. Later in this chapter (see "Alternative fi-nancing for public broadcasting") we shall bring to bear some evidence on howpublic television is responding to its financial stringency, including the expan-sion of its ancillary activities.

These four revenue sources - two involving donations (from governmentand private sources) and two involving sales (user fees for the mission-relatedoutput and revenue from ancillary outputs) - do not have equal potential for ex-pansion. They also have unequal consequences for contributing to organizationmission. Table 13.1 shows how the distribution of public-television revenuesfrom government and private sources has changed over time.

The dilemma of commercial activity

The "special" character of public radio and television - which includes not onlythe programming but also the nature of commercials - creates at least the pos-sibility that stations can find other salable services that may not be available tothe private-enterprise sector. These nonprofit organizations - 629 public radiostations and 350 public television stations in the United States as of 1998 - havesought to sell "memberships," which bundle sale of the collective good, publicbroadcasting, with sale of such private goods as monthly program magazinesand "premiums," such as "free" dinners and umbrellas with the station logo.

Public stations have also moved aggressively into other ancillary commer-cial markets. They sell music cassettes and transcriptions of discussion pro-grams; they sell advertising of business services in newsletters and membershipmailings; they produce broadcast programs for sale to other stations; they con-tract with private firms to market "educational" goods such as books and com-puters. Most of these activities capitalize on the stations' reputations for "qual-ity," trustworthiness, and public-serving, not profit-seeking, goals.


There is nothing to prevent private broadcasters from mimicking the fund-raising methods of public broadcasters, just as these nonprofits can mimic theirprivate-enterprise counterparts. Private broadcasters are less likely, however, tohave equal success in attracting donations of the premiums that public stationsare giving to member-donors, or in attracting people to "join" as members, orin attracting volunteers to help with fund-raising. Both tax-deductibility consid-erations and differential attitudes toward donating money or time (volunteer-ing) to nonprofit and for-profit broadcasters confront these two types of organi-zations with differential revenue-generating opportunities.

It is not surprising, therefore, that private broadcasters have chosen not topursue "memberships" with the same vigor shown by the nonprofit broadcast-ers (though some cable channels, such as the Discovery Network, market them-selves in similar ways). Similarly, it is not surprising that public broadcastershave steered away from the sale of airtime to business interests. As this chap-ter will show, however, external financial pressures - particularly the decreasedavailability of unrestricted governmental funding - have presented nonprofitbroadcasters with a dilemma. They can retain the focus on the twin-missiongoals of "quality" - as they define it - and avoidance of the commercial fund-ing techniques used by private broadcasters; they can, that is, restrict their ac-tivities to what can be financed through donated, presumably unrestricted, rev-enue. Alternatively, they can compromise one or both goals in some measureby increasing their reliance on commercial revenues.

It is tempting for public broadcasters to believe there is no dilemma - thatresort to new revenue-generating sources will not necessitate a compromise inquality. Achieving their goal of quality programming is an admittedly ambig-uous concept, as both critics and supporters of public broadcasting emphasize;thus the impact on mission of emulating private stations' commercials is diffi-cult to assess. One intermediate position has been widely adopted by publicradio stations - "selling" what amount to personal, rather than business, "com-mercials" by mentioning donors' names and a brief message concerning theirbirthdays, anniversaries, and so on. We show, however, that a far greater com-promise of the nonadvertising goal has been evolving at a rapid rate; indeed,some public broadcasters appear to have dropped completely the goal of avoid-ing commercials, although at this point still restricting their timing so as not tointerrupt programs.

Peering ahead, this chapter's examination of the growing commercialism ofpublic broadcasters is intended to shed light on the question of whether they arebecoming indistinguishable from private stations and are thus vacating their dis-tinct historic position. Will the process of competition for revenues continue toblur the distinction, or will public broadcasting contract or even wither away?What is not clear is whether nonprofit broadcasters have sufficient advantagesin the market for private donations - whether because of tax deducibility or al-


truistic support - to sustain their current level and form of activity, or whetherthey can be sufficiently innovative in overcoming free-rider problems in waysthat will sustain their social role. A key question is whether an exogenous dropin government grants causes even a partial, let alone a full, offsetting increasein private donations. Theory suggests that there will be only a partial offset, fora "crowding out" of private donations by public funding is a general phenom-enon in markets for collective goods.

To summarize our perspective, we see any organization as pursuing a pol-icy that optimizes the balance of revenue from varied sources, taking into ac-count the effect of each source on the organization's achievement of its goals.This process involves balancing an organization's multiple goals with one an-other and with the availability of revenues from each source. It also involvesrecognizing that the pursuit of revenue is a process that not only raises fundsbut can also affect achievement of the organization's goals. In short, revenuegeneration has both a direct effect on organization behavior, through its impacton the funds available to pursue the organization mission, and an indirect ef-fect, through the compromises in organization outputs that may be required toobtain those funds.

In the case of broadcasting, the unique mission of public, nonprofit stations,never entirely clear, is becoming even less so. This is particularly true in televi-sion, as the technology of cable has greatly expanded the variety of program-ming, appealing to smaller, specialized audiences. Still, survival is a powerfulmotivator, and nonprofit stations are clearly being driven at an increasingly ur-gent pace to seek funding to offset loss of government grants, and in the pro-cess are becoming more like the private broadcasters whose behavior they wereestablished explicitly to avoid. To understand the dilemma confronting publicbroadcasting today - involving the trade-off between social mission and accessto revenue - it will be useful to review its history.

The goals of public broadcasting in historic perspective

When President Johnson signed the Public Broadcasting Act, he set forth an ob-jective for the new Corporation for Public Broadcasting: "It will have a singlepurpose: to demand the best in broadcasting and to deliver it for the bettermentof our people" (Johnson 1967). Such an ill-defined mission provides little guid-ance to decision makers. It is difficult, indeed, for any public broadcasting sta-tion, or for the system nationally, to judge whether pursuing or forgoing anyrevenue-generating opportunity is or is not inconsistent with the social mission,even though it recognizes that the availability of support from any particularsource is tightly entwined with decisions on how the funds will be used. Theheightening debate over whether commercial advertising should be permittedon public TV - and if so, in what forms - illustrates the problem:


Managers of some of the biggest PBS-member stations want the right to carry ads on-air. Corporate underwriting already provides about 16 percent of public broadcasting'soverall revenues, and those managers believe that they could replace much of the lostfederal aid if the corporate donors got real commercials instead of restricted underwritercredits. . . . Most managers appear to oppose commercials, however . . . more than 80public TV station leaders officially opposed advertising in September 1995; noticeablyabsent were executives from 20 of the 30 largest cities. ("Should Public BroadcastingGo More Commercial?" 1996)

Advertising revenue opportunities are certainly more limited in smaller cit-ies, but it is unlikely that there would be opposition to commercial advertisinganywhere were it not that the social mission for public TV is so ambiguous op-erationally.

In many ways, Johnson's 1967 remarks echoed those made only a few yearsearlier by President Kennedy's young chairman of the Federal CommunicationsCommission (FCC), Newton N. Minow. Speaking before the annual meetingof the National Association of Broadcasters in 1961, Minow had chastenedbroadcasters for their seemingly single-minded devotion to the needs of adver-tisers and called television "a vast wasteland" (Minow 1961). The speech ener-gized the nonprofit broadcasting community, then supported primarily by theFord Foundation, and led the Carnegie Corporation of New York to launch aninvestigation into the prospects for a new kind of national television service. Asthe Carnegie Commission foresaw it, the service would be commercial-free andthus offer distinctly different programming than the three existing commercialbroadcast networks. To describe this new service, the commission chose theterm "public television." The choice was intended to imply something broaderthan "educational television," a service that already existed and that, in Car-negie's view, conjured up images of classroom instruction. As the commissiondefined it, public television would include "all that is of human interest and im-portance which is not at the moment appropriate or available for support byadvertising, and which is not arranged for formal instruction" (Carnegie Corpo-ration of New York 1967, 11).

In its clear rejection of advertiser-supported programming, the commissionwas identifying the essence of U.S. broadcast television economics: When aviewer turns on a TV program, he or she is not the consumer but the product.The real consumer, insofar as financing is concerned, is the advertiser; the pro-grams are merely bait. It was this system, Newton Minow had charged, that ledto the incessant dumbing-down of all television programming (Minow 1961).

By rejecting advertising, Carnegie reasoned, public broadcasting would befree to do better. At the same time, however, it recognized that some fundingmechanism was necessary, and while it was not averse to government funding,"we all agreed that the federal government was a last resort," commission mem-ber Lee DuBridge said later. "If we could see any other way to get adequate fi-


nancing, we would be glad to take it" (Robertson 1992). The funding sourcethe commission recommended was an excise tax on television sets, similar tothe license fee used in Europe. The initial tax would be set at 2 percent of pur-chase price, to rise to 5 percent over time (Carnegie Corporation of New York1967,68-9). To augment funds from this new tax, the commission further sug-gested that public television seek an endowment of $25 million from "industryand private foundations" with which to insulate itself from political pressure.3

When President Johnson signed the Public Broadcasting Act into law, how-ever, the excise tax was missing. It had been removed from the bill at the objec-tion of the Electronics Industries Association, a trade group whose membershipincluded manufacturers of television sets and radios.

The financing problem in the age of TV abundance

Public broadcasting was thus launched without a secure financial base and witha relatively small commitment of government funding. The largest single sourceof public-broadcasting revenues is the public itself, whose contributions ("mem-berships") account for about 21 percent of annual revenues (see Figure 13.1).

Membership drives are a regular part of the programming, occurring sev-eral times a year. No other major public broadcasting service in the world usessuch a revenue-raising device, and many people within CPB and outside it thinkthe effort undignified. The issue of whether such fund-raising is appropriate ordignified is irrelevant to a profit-maximizing firm; for a public-television broad-caster with vague social objectives, however, such criteria can cause intensedebate. This is particularly so when the broadcaster's collective-good nature(which precludes charging viewers) combines with the policy of sharply re-stricting commercial advertising (which limits revenue from selling broadcasttime to private firms) virtually to eliminate conventional user fees as a revenuesource. With little access to revenues from direct beneficiaries - viewers or ad-vertisers - finance options are very restricted. The other potential forms, though- private donations, government grants, and sales of ancillary goods, which isdiscussed later - have the apparent attraction of insulating fund-raising from de-cisions on program quality and content; or so it seems.

The linkage between revenue source and broadcast content has, over theyears, been a volatile political subject. Its volatility has been fueled in part bythe original Carnegie Commission's belief that, somehow, noncommercial pro-gramming would be better programming. Better than what the commission didnot say; nor did it provide any clear test of social mission that would focus to-day's debate over what market niche remains for public television in the newtechnological world of cable and satellite TV, VCRs, video CDs, and the Inter-

3 The first private contribution to CPB was a $1 million gift from CBS.


Subscri21.50%

State Colleges& Universities8.30%

Figure 13.1. All sources of U.S. public-broadcasting revenue, 1994. In 1994,the last year for which full data are available, U.S. public television's total in-come was reported at $1.38 billion, of which the total federal appropriationwas $255 million; the rest came from membership subscriptions ($297 mil.),state and local governments ($294 mil.), businesses ($225 mil.), foundations($78 mil.), and colleges and universities ($34 mil.). (Source: Corporation forPublic Broadcasting)

net. The public-policy question is whether that niche justifies public subsidiza-tion of a service that is watched by only a small minority of viewers. On aver-age, public broadcasting's nationally broadcast programs earn Nielsen ratingsof about a 2 (or about 2 million homes) during the weekday prime-time view-ing hours between 7 P.M. and 9 P.M. - a figure that would drive private commer-cial broadcasters to ruin, but which is almost identical to the ratings earned bypublic broadcasting's cable competitors, such as the Discovery or Arts & Enter-tainment networks.4 The only specific programming recommendations in the1967 Carnegie Commission report that come close to defining the source of

4 Public television's Nielsen rating varies with its national program schedule, and some programs,such as the 1991 documentary The Civil War, received Nielsen ratings of 20 and higher - on parwith major network sports and entertainment events. For an average rating, see A. C. Nielsen'smonthly reports, PTV Cumulative Audience Ranking by Station Market. Weekly Nielsen figuresare also reported in the industry trade magazine Broadcasting & Cable.


private-market failure that might justify public television, publicly financed,were for children's programming and public-affairs programming that wouldgo beyond the "bare reporting of incidents" (Carnegie Corporation of New York1967, 96). Beyond that, there was little attempt to operationalize importantquestions about public broadcasting's mission: Who is the public? What are its"needs'? These questions have bedeviled both CPB and the federal governmentever since. Complicating matters further, the communications core of publicbroadcasting's enterprise raises, in the United States, legitimate questions aboutthe constitutionality of its receiving any government support.

What does public financing really buy?

Another common criticism of public broadcasting is that it is "inefficient" and"unorganized." In a significant sense it is, and was designed to be. This is an-other aspect of the vagueness of its social goal. Efficiency is a matter of achiev-ing goals, and whether public broadcasting is or is not well organized hingeson how closely those goals are being realized. CPB is a private, nonprofit corpo-ration authorized by Congress to develop noncommercial radio and televisionservices. Public broadcasting, however, is an enormously decentralized andfractious enterprise. PBS itself is a station-owned cooperative organization es-tablished in 1969 to acquire and distribute, but not produce, programming. Thenation's 350 public broadcasting stations are independently owned and oper-ated by 175 different licensees, with each making its own decisions about pro-gramming and scheduling; PBS is not a network in the operational sense.

By law, federal funding goes directly to CPB, which then allocates it eitherto individual radio and television stations or to fund programming. Almost halfof CPB's 1995 expenditures - $143 million - went to the nation's 350 individ-ual public television stations in the form of "community service grants." An-other $44.6 million went to radio community service grants. A relatively smallsum - $67 million - actually went to fund radio and television programming(CPB 1996). That funding may go to stations, to PBS or the Independent Tele-vision Service (ITVS), or directly to producers.

Responses to diminished federal financing

With each cut the pressure mounts to find alternative sources of revenue. Theprincipal focus has been on increasing individual and corporate contributions.Such "donations," however, especially from corporations, have had a signifi-cant element of a sale, for these "sponsorships" often include on-air identifica-tion of the sponsor, as well as a listing in a catalog that is distributed to "mem-bers." Thus, in the context of the multiproduct nonprofit organization, publicTV stations have been engaging in two principal fund-raising activities: Devel-


oping new methods for generating individual donations, and selling the outputthat has the greatest private market value, commercial broadcast time - althoughthe latter has stopped somewhat short of the forms of commercials acceptedby private broadcasters. The differences in finance sources between private andpublic stations have decreased but not disappeared.

Reform of public broadcasting's financing has become unavoidable as a re-sult of the diminished federal government grants and, in many cases, weakenedstate and local support too. In 1990, PBS enlisted the Boston Consulting Group(BCG) to appraise its future. The resulting report, Strategies for Public Televi-sion in a Multi-Channel Environment, recommended that to secure donations,especially corporate donations, stations should move away from local program-ming and toward the "safely splendid." In BCG's view, local programming wasnot cost effective because it did not encourage individual memberships or at-tract corporate underwriters. Local stations, on the other hand, were good atraising money. In the end, BCG recommended against a centralized nationaltelevision service, but urged that production resources be shifted toward na-tional programming (Boston Consulting Group 1991).

More recently, the report of the Twentieth Century Fund Task Force on Pub-lic Television urged in 1993 that federal funding of stations' operations be elim-inated entirely and the resources earmarked for national programs. Once thiswas done, the task force said, federal funding should be increased. Specifical-ly, "national funding of public television should come from new non-taxpayersources of funding such as possible spectrum auctions or spectrum usage fees"imposed on commercial broadcasters. "Commercialism," the report said, shouldbe resisted, particularly that which targeted preschool children, one of publicbroadcasting's core audiences (Twentieth Century Fund 1993, 3-6).

Still another proposed response to diminished federal funding has been toprivatize CPB (Carter 1992; Jarvik 1992). This would convert these nonprofitbroadcasters into for-profit firms that might or might not survive in competitionwith private broadcasters. The public-policy issues go beyond the matter ofwhether public subsidies are efficient, as the analysis of conversions showed inChapter 7: Whatever the motivation may be for converting - and whether weare considering public broadcasting, hospitals and HMOs, or any other segmentof the nonprofit sector - conversions have many effects. As that chapter shows,they alter not merely the finance mechanism for an organization, but also itsgoals, the constraints on it, and hence the character of its outputs. Conversionsalso redistribute wealth, shifting assets that were financed in large part throughgovernment and private donations - in the present instance, to improve publicbroadcasting, vague though that term may be - into private ownership, raisinga question of equity. Additionally, public broadcast channels are licensed assuch and can make no such conversion independently; they must have the ap-proval of the FCC. Finally, many public broadcast licensees - about one-third


of the total of 175 - are colleges and universities, state or local educational au-thorities (Ward 1993,143). Many of these entities, some of them predating thecreation of CPB, see educational service as their mission - up to and includingclassroom instruction - and believe that privatization is neither desirable norlikely (Davis 1996).

Alternative financing for public broadcasting

Our theme is that with diminishing federal funds, and further cuts almost a cer-tainty, public broadcasters are turning their attention to finding new sources ofrevenue, thereby blurring the distinction between public and commercial broad-casting. It is not the first time that alternatives to government funding have re-ceived major attention. Over the years a variety of funding mechanisms havebeen proposed, beginning with the follow-up Carnegie Commission on the Fu-ture of Public Broadcasting (Carnegie Corporation of New York 1979). Its re-port urged a spectrum fee on commercial broadcasters as a supplement to gen-eral tax revenues, but the recommendation was ignored.

More recently, in 1982, President Reagan's FCC Chairman, Mark Fowler,called U.S. commercial broadcasters' claims that they serve the public interest(as the statutory terms of their licenses require) a "fiction," and urged that in-stead of receiving free and exclusive use of their channels they pay a spectrumfee and be freed from government content regulation. The resulting revenues,Fowler said, could be used to support public broadcasting (Fowler and Brenner1982,207). The decision to levy such fees, of course, lay with Congress, whichdid nothing. In 1996 this debate arose again, led by Senate Majority Leader BobDole, who held up passage of the Telecommunications Act when he realizedthat it gave broadcasters, without charge, additional spectrum for advanced dig-ital broadcasting. Dole estimated the value of that spectrum at $70 billion, andsuggested that for both economic and First Amendment reasons broadcasterspay for it, though he said nothing about where the resulting revenues might go(Minow and LaMay 1996). Several public broadcasters urged that a portion ofthem, along with revenues from the sale of several public broadcast televisionstations in small markets around the country, be used to create a $4 billion trustfund for public broadcasting (Boot 1995).

The spectrum and trust-fund proposals are but the latest in a long line ofalternative financing proposals advanced over the years. None of the earlierproposals - surtaxes on residential electricity and telephone bills; governmentmatching of private contributions; taxes on license transfers when commercialstations are sold; and taxes on the advertising revenues of commercial broad-casters and cable operators - have come to fruition (Ward 1993, 150).

Because they have not, public broadcasters have sought to generate theirown sources of revenue. This process began in earnest in 1981, when the Rea-


gan administration threatened to end federal funding entirely, and, in response,Congress amended the Communications Act such that "Each public broadcast-ing station shall be authorized to engage in the offering of services, facilities orproducts in exchange for remuneration." The door to commercialism, enteringnew markets where one or another form of output could be sold profitably, wasopen.

Stations can now rent out their studio space, accept advertising in their pro-gram guides, and sell books, music tapes, transcripts, and classroom materialsbased on PBS programs. They do all of these, and more. PBS itself leases timeon its satellite transponders through PBS Enterprises, a wholly owned, for-profit subsidiary that develops services for use with new communications tech-nologies. PBS Video sells videocassettes of public-broadcasting programs. PBShas even created its own home video label - PBS Home Video - and allows allpublic stations to buy its tapes at 52 percent off retail, then resell them at consid-erable profit.

With the expanded opportunities, together with the fiscal pressure resultingfrom cutbacks in government grants that encourages stations to pursue thoseopportunities, public broadcasters are developing an ever-widening array ofcommercial ventures to raise revenue for their preferred, mission-related out-puts. In the fall of 1996, for example, CPB gave four "strategic business de-velopment consultation" grants to public radio and television stations, essen-tially to develop new commercial ventures. Station WHYY-FM in Philadelphialaunched a book-club service linked to the highly popular NPR program FreshAir with Terry Gross; WHRO-TV in Norfolk, Virginia, developed a World WideWeb consulting service for businesses; several small Illinois public televisionstations established a teleconferencing business for citizens, businesses, andgovernment agencies in rural west-central Illinois; and television station KCETin Los Angeles created a Center for Professional Development to offer multi-media and Internet training to teachers and private industry (CPB 1995).

The grants, awarded in November 1995, were the second round of the CPBBusiness Development Initiative; the first round gave grants to WPSX-TV atPennsylvania State University to develop CD-ROM and other electronic pub-lishing; to KTCA/KTCI-TV in Minneapolis-St. Paul to develop on-line, CD-ROM, and teleconferencing services; to KQED-TV in San Francisco to devel-op adult-learning services; and to KNPB-TV in Reno to create a retail operationsimilar to LearningSmith or the Store of Knowledge stores found in Boston,Los Angeles, and Chicago (CPB 1995).

Many other public television stations are well ahead of the curve in findingnew commercial revenue sources. WTTW-Chicago, for example, partners inthe production of programs with MTV Networks, Lifetime cable, and Walt Dis-ney's Buena Vista television unit. In Seattle, KCTS-TV has produced the high-ly successful syndicated science program for young people, Bill Nye the Sci-


ence Guy, for the Disney Channel. WTTW in Chicago, WGBH in Boston, andKCET in Los Angeles have all launched chains of retail stores with for-profitpartners that supplied the initial capital, either LearningSmith Inc. or Store ofKnowledge Inc. Public radio stations are just as active in business innovation.In St. Paul, Minnesota, for example, the twenty-nine-station Minnesota PublicRadio Network raises about $150 million a year from its mail-order-catalogbusiness (Schatz 1996).

The profitability of all these activities is not clear. WGBH, for example, re-ports that net income from its for-profit ventures amounts to no more than about5 percent of its annual operating budget (Ward 1993,158). Reported profits onunrelated business activity, however, are likely to be huge understatements, aswas explained in Chapters 3 and 5. Nonprofits, including public broadcasters,have the incentive to select ancillary business activities that use inputs jointlywith their tax-exempt, mission-related activities, and then to allocate as muchof the joint costs as possible to the taxed sector, thereby showing little or noprofit in that sector. Indeed, for all nonprofit organizations, the total reportedprofit from unrelated business (UB) activity is actually negative. Real losses,however, are not credible, and even less credible is evidence that gross UB rev-enue is growing rapidly but losses are getting larger (Chapter 5). In publicbroadcasting, as elsewhere in the nonprofit sector, creative accounting in theallocation of costs and revenues between the organization's untaxed (mission-related) activities and its taxed (i.e., UB) activities doubtless distort and un-derstate the actual contribution of the UB activities to organization resources.

Further, many people in public broadcasting doubt that these enterprises canbe greatly expanded without risking congressional wrath and the hastened endof all federal monies. Many for-profit businesses have attacked nonprofits' in-creasing engagement in commercial business activities as constituting unfaircompetition (see Chapter 3). Public broadcasters have not been singled out forcriticism, but their commercial expansion is particularly rapid, and so they arelikely to be under growing challenge as political forces bring pressure on theIRS to restrict nonprofits' UB activities as well as to expand the scope of whatis defined as unrelated (SBA 1983, GAO 1987, Emshwiller 1995).

Despite concerns, public broadcasters have pushed on, more worried aboutcongressional threats to eliminate funding. PBS predicts a $22 million revenueincrease in 1997 from the sale of videotapes and other educational media, anda $10 million increase from the sale of satellite services. Additionally, PBS hasentered into a $75 million "program partnership" with Reader's Digest. Anoth-er deal that PBS announced in March 1995, a $15 million investment from MCIto develop an on-line service, has so far not gone forward (Behrens 1996). Notonly is public broadcasting entering a vast uncharted sea of commercial activ-ity, but it is doing that increasingly in partnerships with private firms. The bar-riers between nonprofit and for-profit organizations are falling.


It must be noted that this is true not only in the United States, but also in Eu-rope, Canada, and Japan. Because of the growing number of channels availableeverywhere, the competition for talent, film, and major events, such as sports,has driven their price up well above the inflation rate. This is a particular prob-lem for public broadcast systems dependent on a license fee for TV sets, be-cause the fee is typically pegged to inflation or lags behind it. Public broadcast-ers have thus turned to new sources of revenue, including advertising, to makeup the shortfall. The BBC, for example, in 1994 created BBC Worldwide, a di-vision dedicated to developing commercial revenues. In 1995 the BBC launchedtwo new commercial services outside Britain - BBC World, a news channel,and BBC Prime, an advertiser-supported entertainment channel; both were un-dertaken in partnership with private firms ("Time to Adjust Your Set" 1995).

Corporate underwriting and advertising

In the debate about how to finance public broadcasting, concerns about adver-tising - under whatever name it goes - have always been central. Americanbroadcasting's earliest pioneers had been nonprofit institutions - most notablyuniversities, churches, and labor unions - all of which viewed radio broadcast-ing as essential to their public mission and many of which believed advertisinga necessary, if not entirely desirable, adjunct to that mission (Barnouw 1968,chap. 1). Today, advertising on public broadcasting is in the form of corporateunderwriting of programs. That support accounts for about 17 percent of pub-lic television's total budget and about 30 percent of the national programmingbudget (Ward 1993, 149) (see Figure 13.1).

In 1981, the role of advertising in public television was examined more care-fully when Congress and the FCC authorized a Temporary Commission on Ad-ditional Financing (TCAF) in that year's Public Broadcasting Amendments Act.The TCAF project was an eighteen-month experiment in which ten public tele-vision stations around the country were allowed to broadcast advertisementsunder restricted conditions: Advertisements were to appear only between pro-grams, never during (and then only for two minutes); and no advertising at allwas to appear in company with children's programs.

According to the FCC report issued after the experiment, the results werepromising. Viewer contributions during the experiment either stayed constantor increased. Interviews done with ten thousand public-television viewers afterthe TCAF ended found that few discerned any harmful effects from limited ad-vertising; and so long as advertisements did not interrupt programs, viewers didnot object to them. Viewers also said they preferred advertisements to on-airfund drives (FCC 1984).

Despite these findings, the FCC expressed concern that whatever benefitsmight accrue from advertising would be outweighed by its presumed invidious


effects on the mission of public broadcasting. The commision also worried thataccepting commercials would be a net money loser because it would requireadded production costs, renegotiated labor contracts, taxes on revenues notdeemed "charitable support," and the loss of viewer support. As a compromise,it urged what it called "enhanced underwriting," a carefully regulated form ofsponsorship, and not a terribly successful one (FCC 1984). Advertisers oftenconclude that underwriting spots are ineffective, and withdraw support. In 1991,for example, corporations provided a total of $17 million to be exclusive under-writers of twenty PBS programs; by 1993, there were only fourteen programsexclusively underwritten, with total support of $6.3 million (Window to theWorld Communications, Inc., Chicago, IL, personal communication, 1993). In1994, PBS announced publicly that it would reconsider its underwriting guide-lines with the intent of liberalizing them. Jonathan Abbott, a PBS senior vice-president, told the Wall Street Journal, "We want to make sure we're a desti-nation more companies are comfortable considering" (Jensen 1994, A3).

Despite the narrowing difference between commercial television advertisingand the forms adopted by public television, and however public broadcasting'sunderwriting guidelines are finally redrawn, its use of advertising will remaincontroversial. One of the most heated debates in public broadcasting today isthe uncertain conflict between the revenue that advertising, in any form, cangenerate and the public-serving mission that requires financial independence.This debate is well under way in Europe, where advertising has increasinglyfound its way into public-broadcasting finance. Noting this, some of the largeprogram-producing public broadcasting stations in the United States have urgedCongress and the FCC to experiment further with liberalized commercial ad-vertising as a revenue source (LaMay 1994, A25). Enthusiasm for advertisingis not universal, however. Opposition can be summarized by the views of Shar-on Percy Rockefeller, president of WETA in Washington, D.C.: "Commercialswould be the death knell of public television. We already have a vast commer-cial broadcasting system in this country; we need the alternative. And it's thealternative because it's noncommercial. It's so obvious. We put programs onthe air for the purpose of the content, not for the purpose of making money fromcommercials" (Mifflin 1996).

"Experiments" cannot resolve the debate over the social efficiency of allow-ing nonprofit organizations, public broadcasting or other, to emulate the finan-cial approaches used by private firms. One half of the experiment is eminentlyfeasible: to determine whether, and to what extent, commercial advertising orsome other measure can actually increase the flow of revenue. The "rub": Howwill the organizations' social mission be affected? That requires understandingwhat the mission is in ways far more operational than by reference to "the pub-lic interest." Without this, experiments will do little to resolve disagreements


over the appropriate role for advertising or, for that matter, other fund-raisingtechniques.

Concluding remarks

The public-broadcasting sector of the nonprofit economy illustrates the tensionbetween nonprofit organizations' social goals, which differ from those of pri-vate firms and are the basis for subsidies and tax-exemptions to nonprofits, andthe increased commercialism of nonprofits, their use of revenue-raising tech-niques that emulate those of private firms. It is clear that, vague as public broad-casting's social goals may be, the organizations that comprise it do fit the frame-work in which some revenue-generating measures and some forms of outputare preferred over others, and the use of commercial-type mechanisms that in-volve selling services - a nonpreferred approach - is responsive to exogenouschanges in contributions, gifts, and grants.

As government grants continue to decline, and as competition for private do-nations becomes more intense among all participants in the nonprofit sector,public broadcasting's mission will come under further scrutiny. This is true notonly in the United States, but in public broadcasting systems worldwide. Typi-cally, however, other public broadcasting systems have detailed public-servicecharters specifying programming and service requirements, as well as an orga-nizational structure designed to meet them. U.S. public broadcasting has nei-ther of these things, in part because it was created as a nonprofit alternative toa predominantly for-profit, commercial broadcast system. However, particular-ly with public television, it is no longer clear what "alternative" the service pro-vides.

As the United States moves toward digital broadcast television delivery, thequestion of public television's mission is likely to become even more con-fused. Digital transmission permits a broadcaster to operate six channels in thespectrum space currently used to operate one, and the explosion of new profit-seeking channels will further segment the television audience while also in-creasing the competition among all programmers for the limited supply ofexisting video content. Further, the transformation to digital broadcasting willrequire all broadcasters to make huge capital investments in equipment and fa-cilities. Many broadcasters, public broadcasters included, are expected to re-coup some of that cost by using their extra channels for nonbroadcast opera-tions such as Internet service or telephone paging. The manner in which publicbroadcasters respond to these new competitive and structural challenges willhave a great deal to say about their mission in the multichannel marketplace.

PART III

Overview, conclusions, andpublic-policy issues

CHAPTER 14

Commercialism among nonprofits:Objectives, opportunities, and constraints

Estelle James

Introduction

Earlier chapters in this book provide a comprehensive picture of growing com-mercialism among nonprofit organizations. Commercialism is defined to meanthe degree of reliance on sales revenues rather than donations or governmentgrants, the production of goods for sale that compete with goods produced byfor-proflt organizations, collaborations and partnerships with for-profits, and, ul-timately, conversion into for-profits. The book attempts to answer the questions:Is commercialism increasing? If so, why and so what? Should it be encouragedor discouraged by public policy? Is growing commercialism good because itallows nonprofits to flourish and grow, or bad because it makes them more likefor-profits? Here I contrast the arguments set forth in this book with a contraryhypothesis about nonprofit behavior and why it appears to be changing. I re-view the evidence provided to see which view of the nonprofit world is mostconsistent with the facts, and discuss some of the public-policy implications.

The book depicts nonprofit organizations as having different and more altru-istic objective functions than for-profits, which lead them to engage in commer-cial activities marginally and reluctantly, in order to cross-subsidize their pre-ferred noncommercial activities, when other revenue sources such as donationsand subsidies are not available. In fact, this is a point of view that I have es-poused in the past, and I believe it is consistent with long-standing behavior insome organizations, such as universities.

However, the contrary hypothesis that I shall examine here is stimulated byempirical observations in this book, indicating that the most dramatic growthin commercialism has occurred in situations where institutional or legal con-straints on sales revenues were lifted, revealing large new opportunities. Non-

Editor 's note: After the studies in the preceding chapters were completed, we invited Estelle James,a distinguished contributor to the literature on the nonprofit sector, to write an overview chapterpresenting her perspective on our work.

271

272 Estelle James

profits responded enthusiastically to the new opportunities, without much evi-dence of reluctance. It is unclear whether the large "potential profits" therebygenerated were used to cross-subsidize loss-making, public-consumption goodsor were simply plowed back as investments to generate more profit-makinggoods; the contrary hypothesis predicts that the latter effect will dominate inmany cases. Declines in donations, in some industries analyzed in this volume,seem to be a consequence rather than a cause of the increased reliance on sales.These empirical observations suggest that constraints and large opportunities,rather than altruistic objectives and cross-subsidization, play an important rolein explaining the limited commercialism of nonprofits in the past and its strik-ing emergence in some industries in recent years. When constraints are removedand large new opportunities for profit-making present themselves, nonprofitsmay behave very much like for-profits.

The theory of nonprofit organizations andcross-subsidization

The starting point for my discussion is the assertion (which I call the conven-tional hypothesis) that nonprofits have nonpecuniary altruistic objectives. Theyexist to provide public (collective) goods and to distribute quasi-public goodsin ways that are different from market distribution - that is, to redistribute realincome via free or low-price access to goods and services that generate exter-nal benefits. These objectives lead to pricing policies that differ from those offor-profits. (See Chapter 3 for a theoretical discussion of possible differences.)

How are these goods financed, given that consumers cannot be excludedfrom the consumption of public goods, and nonprofits do not want to rely onthe price mechanism even in cases where it is technologically feasible? In partthrough donations, by recipients of the consumer surplus or the external bene-fits who receive a "warm glow" by acting altruistically. Donations, it is wellknown, are limited by the free-rider problem and by the difficulty potential do-nors have in monitoring the effects of their donations. Nonprofit status (i.e., thenondistribution constraint) helps solve the monitoring problem by reassuringdonors that their contributions will, in fact, be spent on services rather than sim-ply on increasing profits that are distributed to the firm's owners. The choiceof nonprofit managers whose objectives coincide with those of the donors pro-vides additional reassurance. Tax deducibility of donations helps solve the free-rider problem, by decreasing the after-tax cost of the donations.

Government subsidies to nonprofit organizations are important additionalsources of revenue. When government grants and private donations decline, thenonprofit is forced to become commercialized, to survive. Commercializationis said to be nonpreferred by nonprofits, because it involves producing goodsthey dislike, or charging prices that limit the consumption of goods they do like(Chapter 4). Thus, when other discretionary funds are available, commercializa-

Commercialism among nonprofits 273

tion stops below the profit-maximizing point. The theory of cross-subsidizationhypothesizes that when other revenue sources become less available, nonprof-its will become more commercialized in order to generate a larger profit, whichthey can use to finance their provision of preferred public and quasi-publicgoods. This book argues that commercialism has been increasing in the non-profit sector over the past two decades for this reason, and examines in detailseveral industries dominated by nonprofits to document this claim.

In evaluating a theoretical argument, it is useful to have a countertheoryagainst which to test it. As a countertheory, then, we might hypothesize thatpecuniary rather than altruistic objectives dominate the decisions of many non-profits. These "false" nonprofits - which Weisbrod (1988) referred to as "for-profits in disguise" - may maximize profits that they then distribute in disguisedform (as higher wages and perks), or they may maximize revenues that lead topower and prestige for their managers. They are lured into the nonprofit sectorby the tax and subsidy advantages that they gain therefrom. If these advantagesdecline, the disadvantages (loss of access to equity capital) may dominate andlead to an exodus from the nonprofit sector.

Further, nonprofit managers and donors may have different interests. If non-profit managers are primarily interested in revenues, they may limit their com-mercial activities because they fear that donations will fall if they appear to beself-supporting. The operative factor here is not the disutility of commercial ac-tivities, but their negative impact on other sources of funds. If large new oppor-tunities open up - for example, due to technological, institutional, or legalchanges that enable the price system to work - the gain in sales revenue mayfar outweigh the possible loss in donative revenues, and nonprofits may eager-ly embrace the new opportunities. We may then see commercialism increasedramatically and nonprofit behavior that is very much like that of for-profits.

Thus, whereas the conventional, cross-subsidization hypothesis emphasizesthe differences in objectives between for-profits and nonprofits, views commer-cialism as a source of disutility, and sees increases in commercialism as a re-sponse to declining donations and grants, the counterhypothesis emphasizesthe similarities in objectives between nonprofits and for-profits, downplays theimportance of disutility, sees donations as responding to sales rather than viceversa, and underscores the key role played by the removal of institutional andlegal constraints in opening up new sales opportunities and expanding commer-cialism. The next section of this chapter examines the empirical evidence to seewhich of these views receives more support.

The evidence

The great value of this book is that it provides detailed empirical evidence frommany industries that allow us to explain and evaluate the growing commercial-ism among nonprofits.

274 Estelle James

In Chapter 6, one of the chapters that cuts across industries, Segal and Weis-brod ask: Did the slower growth rate of donations over the past decade causeincreased reliance on sales revenues? Their first answer is positive. Using pan-el data (a pooled cross section with lags) from the Forms 990 of nonprofits for1985-93, they demonstrate that (next year's) sales revenues increase signifi-cantly when the growth rate of donative revenues declines. However, as theynote, this specification assumes that donations are exogenous. When they usea vector autoregression model that allows for simultaneous determination ofdonations and sales, they find that sales influence donations, rather than viceversa. When the data are disaggregated by industry, the correlation betweensales and donations is positive for some - most notably hospitals and universi-ties - and negative for others.

On the whole, I find this unconvincing evidence of cross-subsidization basedon preferred versus nonpreferred goods. Rather, the simultaneous-determinationmodel is consistent with the idea that revenue-maximizing nonprofit managerswill charge a lower price than they would if they viewed donations as exoge-nous; that is, their reluctance to rely on sales revenue may stem more from itsexpected impact on donations than on its disutility to them - a possibility not-ed in Chapter 3. The zero or positive relationship between sales and other rev-enue sources seems to exist primarily in multiproduct industries, where dona-tions may be targeted to goods where price finance still does not work andcross-subsidization from the more profitable goods is not expected.

Chapters 7 and 8 describe in detail the decline in economic and legislativeadvantages for nonprofits that have led to increased commercialism in the healthindustry. Initially nonprofits had a secure place in this industry, justified by theexpectation that they would provide charity care, better-quality service, and oth-er socially useful activities through donations and cross-subsidization. Healthcare has long been given as a key example of an industry where consumer in-formation is limited, trust is a substitute, and nonprofits were more likely to betrusted by consumers than were for-profits. As a result, the industry has beendominated by nonprofits, which have enjoyed numerous tax and subsidy advan-tages. For example, according to the Health Maintenance Organization Act of1973, only nonprofits were eligible for federal subsidies, and nonprofit hospi-tals have long benefited from special tax exemptions. However, direct federalsubsidies to HMOs ended in the early 1980s, and the tax advantages of BlueCross and Blue Shield and hospitals gradually eroded over time, especially dur-ing the 1980s.

Even more important - and possibly explaining this change in public pol-icy - the growth of health insurance, through private coverage, Medicare, andMedicaid, ensured that most people could pay for medical services. In otherwords, institutions developed that enabled price finance to work, even for thepoor, thereby diminishing the redistributive rationale for nonprofits. The advent


of medical insurance, which was strongly supported by hospitals, makes cross-subsidization and the provision of charity care by nonprofits less critical. Notsurprisingly, donations fell drastically in response to insurance, evidence thatas institutions change so that the good in question can be privatized, donors willtake their money elsewhere. This diminishes the comparative advantage of thenonprofit sector. Total revenues of hospitals increased, however, implying thatthe gains from insurance and sales far outweighed the loss from donations.(Consistent with this observation, for-profit hospitals tend to avoid areas whereuninsured people live, whereas nonprofits are more likely to survive in theseareas).

In recent years, employers and government have sought cheaper and morepredictable health-care arrangements. Competition among both nonprofits andfor-profits - in the context of prospective payment systems, the setting of pricelimits by Medicare and Medicaid, the increasing use of HMOs and preferredproviders, and negotiations between insurance companies and hospitals - meantthat although costs were covered, there was little surplus left for discretionaryspending and cross-subsidization.

As a result of these forces, numerous studies show that (with the exceptionof the small group of teaching hospitals) nonprofit and for-profit hospitals co-exist and behave in very similar ways with respect to quality of service, degreeof uncompensated care (minimal), and efficiency. Indeed this coexistence andsimilarity, largely stemming from a common reliance on third-party price fi-nance, may have led to the diminution of public-policy privileges for nonprofitsdescribed above. Consequently, the rationale for remaining nonprofit appearsto be disappearing - that is, the disadvantages of the nonprofit form (lack ofaccess to equity capital) outweigh the advantages - and we are currently ob-serving a wave of conversions, asset sales, and contracting-out arrangements,from nonprofits to for-profits, among the Blues, HMOs, and hospitals. Nor haveI seen much evidence of reluctance as to conversion in these organizations;in fact, the managers are often major beneficiaries - evidence for the counter-theory.

The other major change in balance between nonprofits and for-profits has oc-curred in the biotech area, where the lines between nonprofit-university researchand for-profit applications have become increasingly blurred. Chapter 9 de-scribes the many changes in federal policy that set the stage for these develop-ments. Legislation in 1980 allowed universities to retain the patent rights to dis-coveries stemming from federally funded research, a right that had previouslybeen denied to them. After 1984 these property rights could be transferred (sold)to others, including for-profits. Legislation in 1986 permitted universities andfirms to collaborate without fear of antitrust litigation. In 1989 the opportuni-ties for licensing of university research and other intersectoral collaborationwere further expanded. In 1993 defense-related research was opened up to com-

276 Estelle James

mercialization. New NSF programs encourage or require university-industrycooperation. Corporations are used to evaluate the technical merits of propos-als, including the feasibility of commercialization and marketing as key pointsin the evaluation process.

In this case, as in the health industry discussed above, public-policy changesprobably were not exogenous; rather, they were probably an endogenous re-sponse to changing economic circumstances. After the cold war ended, and asinternational trade and capital movements expanded, public policy toward re-search has been driven by the belief that U.S. corporations were losing out inglobal economic competition, and that we could avoid this by developing newtechnologies, which requires a shift from basic to applied research. In the past,with the United States dominating the world economy, it was expected that do-mestic firms would capture most of the benefits of basic research; but as othercountries have grown in economic power, this assumption is no longer valid.Instead, Japan, Korea, and other emerging countries may be expected to appro-priate some of the benefits of our basic research. Under these conditions theU.S. government has less incentive to invest in basic research, and more incen-tive to shift its focus to the applied level, where the benefits are narrower andmore likely to be captured by domestic firms; and this is exactly what has hap-pened, as legal constraints changed to permit and encourage applied research.

Fortuitously, the computer and biotech (genetics and Pharmaceuticals) indus-tries were on the verge of major breakthroughs, poised for commercial develop-ment of scientific discoveries. As research in these industries generated hugepotential profits, universities seized upon the new opportunities for pecuniarybenefits. The result: a proliferation of alliances between industry and the acad-emy, an upsurge in licensing, cooperative R&D agreements, and joint researchconsortia. Nineteen percent of university research and a much larger share inthe biotech area is now carried out in close linkage with industry. Key facultymembers move back and forth between industry and university laboratories,and small start-up firms, spun off from universities, are often led by present orpast faculty members. Universities get an increasing share of their income frompatents, licenses, royalties, equity in these start-up companies, and researchagreements with private industry. A large and growing share of university re-search is now financed by private rather than public sources. In the life sciences,private support is greater than public support; but in this case public and pri-vate funding sources appear to be complements rather than substitutes. For ex-ample, the upsurge in university patenting and university share of total patentsthat has occurred over the past decade has attracted private funds, but most ofthe discoveries originated in NSF- and NIH-funded research projects.

Three key points are especially worth noting: First, the changing universitybehavior is due to changing public policies - in particular, the removal of con-


straints that closed off key avenues of commercialization to nonprofit univer-sities, at a critical moment in time, permitting large new opportunities to openup. Second, these changing public policies are themselves an endogenous re-sponse to changing perceptions of the (domestic) economic benefits from ba-sic versus applied research. Third, the universities responded enthusiastically.Though some concerns have been raised about the conflict between the academ-ic value of free and open information versus the commercial motive for keep-ing information secret, on the whole universities (particularly the life-scienceparts of universities) aggressively pursued for-profit alliances and modes of be-havior once the legal barriers came down. There is little evidence of disutilityhere, nor any indication that private funding of the life sciences allowed otheruniversity resources to be allocated elsewhere, as a form of cross-subsidization(e.g., profits from applied biotech research likely has not found its way to hu-manities libraries). I regard this as further evidence for the counterhypothesis.

In contrast to health and higher education, in other industries commercial-ism is moving at a much slower pace and with fewer changes in underlying eco-nomic conditions and legal constraints. Chapter 12 finds that, among art muse-ums, sales revenues have increased, but only in proportion with other revenues.As government and foundations have supported capital investment, special ex-hibitions, and outreach programs, the maintenance and normal operating ex-penses that they generate must be covered from elsewhere, and museums haveturned to admission fees and sales from auxiliary operations for this purpose.Still, the proportion of revenues from sales has remained roughly constant, ascommercial and donative sources have increased at similar rates.

Chapter 10 describes complementarity between membership contributionsand sales, but possible substitution between sales and donations, in the social-service industry. Although many managers of social-service nonprofits feel apressing need to generate additional resources because government funding forthese activities has been cut, commercial activities are approached with cau-tion owing to a possible loss of reputation and donations. Are these activitiesfrowned upon because they generate intrinsic disutility, or because they mayreduce other funding sources; that is, does a utility-maximizing or revenue-maximizing objective shape the nonprofits' behavior? We cannot respond withcertainty, given the possibilities of cross-subsidization on the one hand and neg-ative interactions among funding sources on the other hand, as Chapter 3 not-ed. However, we do observe that to minimize crowd-out, managers look forcommercial activities that are a natural outgrowth of their primary missions:The American Lung Association and the American Cancer Society accept spon-sorship from companies selling products that seem consistent with good health,AARP endorses (for a fee) insurance companies that are said to provide goodservice to senior citizens, and so on.

278 Estelle James

We also observe that membership fees for social-service nonprofits are keptlow - to increase member numbers and hence the organization's political cloutand its market for commercial sales. In this case, membership fees and salesrevenues are complementary because most sales are to members. Again, this be-havior is consistent with maximization of either revenue or utility.

Econometric evidence provided in Chapter 11 seems most consistent withthe utility-maximization and cross-subsidization hypotheses. Zoos charge pricesbelow the profit-maximizing levels; that is, they operate in the inelastic partof their demand curves, to keep admissions high and reach low-income groups.Sales of nonpreferred, ancillary goods (animal rides, souvenirs, food, parking)by zoos has grown as donations have decreased since the 1970s. Although sim-ilar behavior might be predicted for revenue maximizers in the face of strongand increasing crowd-out, zoos generally seem to fit the conventional model.

Further evidence both of cross-subsidization and the importance of con-straints is found in Chapter 13 on public broadcasting. Initially public TV wasnot supposed to engage in commercial activity; indeed, freedom from the de-mands of advertisers was its raison d'etre. However, since 1981 public broad-casting stations have been permitted to sell services and products, as new legis-lation removed old constraints and opened the door to commercialism. Ruleson corporate sponsorship have eased, and further easing is under consideration.Simultaneously, government grants to public broadcasting have declined. Asa result of these forces, we observe growing partnerships with for-profits, fund-raising and membership drives, sales of broadcast time to corporate sponsors,and provision of auxiliary services (CDs, transcripts). It is difficult to disen-tangle whether the removal of constraints or the increased financial pressurefor cross-subsidization is the greater motivating force for commercialism; prob-ably the two interacted, creating greater opportunity and greater need at thesame time.

As in the social-service industry, public broadcasters fear that advertisingand other forms of commercialism may crowd out public and private donations,and they shape their sponsors' messages and other sales techniques to minimizethis negative effect. I suspect we shall see much more commercialism in thisindustry as legal restrictions continue to be lifted.

As in the health and education industries, changing public policy in publicbroadcasting may be seen as an endogenous response to changing technologyand institutions - in this case the advent of cable TV, which enabled the pricemechanism to work better, hence undercutting the rationale for governmentgrants and private donations to public TV. In the days before cable, pricingcould not be used because exclusion of viewers was not technologically feasi-ble, and commercial TV was financed by advertisers interested in maximizingconsumption of their products. Strong preferences for a particular type of pro-


gramming by small groups of people were ignored by these advertisers becauseit would not increase their sales. Public broadcasting, financed by governmentgrants and private donations, was supposed to overcome this market failure.

After the advent of cable TV, this small group could finance its preferred typeof programming by paying a high fee to a cable producer, undercutting the ra-tionale for grants and donations. Cable TV allows strong specialized tastes tobe catered to, since it converts a nonexcludable good into an excludable goodfor which intensity of preference can be registered via the price people are will-ing to pay. Thus, the economic rationale for subsidizing public TV and for lim-iting its commercialism has declined in the past two decades, and the subsidiesand constraints have likewise declined. The quest for organizational survival hasled public TV to seek other (nonprice, noncable) financing sources, and theseare likely increasingly to rely on commercialism directed toward a niche au-dience.

In sum, the industries where commercialism has been most rampant, as man-ifested in partnerships, mergers, and conversions, are also industries where (1)institutional or technological change has made exclusion and price financingmore feasible and/or (2) tax and subsidy advantages to and constraints uponnonprofits have been lifted since the early 1980s. The removal of constraintsmakes it possible for nonprofits to act more like for-profits; the reduction in taxand subsidy advantages makes it advantageous for nonprofits to convert to for-profit status; and the feasibility of price finance enables both legal and behavior-al conversion between the two sectors. Indeed, we observe all these instances,particularly in the health industry and the university-biotech arena. Moreover,these occurrences are all consistent with the counterhypothesis, which viewsmany (but, of course, not all) nonprofits as quite willing to react opportunisti-cally to changing incentives and constraints in order to maximize revenues ordisguised profits, rather than as being driven by nonpecuniary altruistic objec-tives reluctantly to undertake commercial activities. Furthermore, in all thesecases public policies may be viewed as an endogenous response to changingeconomic conditions and institutions that make it desirable and feasible to shiftthe balance from public to private goods.

At the same time, less dramatic increases in commercialism observed inmuseums, social-service organizations, zoos, and public broadcasting are moreconsistent with the conventional hypothesis that nonprofits commercialize andcross-subsidize when they are driven to do so by a reduction in public or pri-vate donations. In these cases, underlying institutions and constraints have notchanged so as to produce discontinuous changes in behavior. Commercialismin these industries continues to be slowed by the disutility it creates and by thefear that sales revenues may drive out donations: It is impossible to disentan-gle these two effects.

280 Estelle James

Public-policy implications of commercialism

Why do we care about commercialism among nonprofits? Why is it worth writ-ing a book about this topic? Should special privileges given to nonprofits bewithdrawn as nonprofits become more like for-profits? To answer these ques-tions it may be useful to recount the reasons why public policy creates a spe-cial legal category for nonprofits and gives them special tax and subsidy advan-tages in the first place.

Tax and subsidy advantages: Why do they exist andshould they be removed?

Historically, most nonprofits were religious organizations that started schools,hospitals, and mutual-benefit organizations to serve their members better andmaintain their loyalty. From a political-economy perspective, they may have re-ceived special tax treatment because of their political power and (especiallyin the United States) because of our legal strictures to maintain a separationbetween church and state.

As society became more secular, so too did these organizations; but they re-mained concentrated in areas (education, health, social-service charities) whereexternalities are thought to exist and where the price mechanism could not berelied upon to achieve optimal results. Indeed, this book asserts - but does notprove - that nonprofits exist to provide public goods and to expand consump-tion of quasi-public goods to groups that otherwise could not afford to purchasethem. Subsidies and tax privileges to nonprofits may be given to encourage do-nations and to overcome the free-rider problem that exists where exclusion andpricing are not possible.

In other countries direct government payment for these services is common,but in the United States indirect subsidy via tax advantages is more likely - pos-sibly because as a heterogeneous society we cannot reach a consensus on whichgoods and services to subsidize. Thus we decentralize that decision and subsi-dize via matching grants to organizations that receive "votes" in the form of do-nations.

Public policy earmarks these privileges only for organizations that agree notto distribute their revenues to private owners in the form of dividends or capi-tal gains, because the government cannot observe the detailed behavior of manydecentralized organizations. Just as the nonprofit form may be taken as a proxyof trustworthiness by private donors, so too it may be regarded as less likely toabuse its subsidies (thereby causing a political scandal) by public donors. How-ever, organizational form is only a second-best proxy for what we would real-ly like to observe. Indeed, if "for-profits in disguise" are enticed into the fieldby the existence of tax privileges and subsidies, it may be a third-best proxy.


Given this background, increased commercialism of or sales by nonprofitsenables them to pursue their altruistic mission better by providing revenues,which should please both public and private donors. However, balanced againstthis are a number of problems it creates or new situations it underscores. First,it may signal that the service provided by nonprofits can now be financed bythe price mechanism, due to new technologies and institutions, so public andprivate donations are no longer needed. In this case, private donations are like-ly to diminish over time, as donors shift their money to other areas where theyare less dispensable. Since donations are diminishing, tax exemption for philan-thropy does not pose a new public-policy problem, but other tax privileges maymerit reconsideration. If benefits can be captured by prices, why should income-and property-tax exemption be offered?

Second, the increased emphasis on sales revenues may attract more "for-profits in disguise" into the sector - and may lead "real" nonprofits to hire newtypes of managers, with training, background, and objectives similar to thosein the for-profit sector. These managers may be equally likely as for-profit man-agers to cheat the consumer or donor with respect to output characteristics thatare not readily observable. In effect, true nonprofits may be turned into "for-profits in disguise" as a result of the managerial selection process, without fullcognizance of this fact by the organization and without a deliberate decisionhaving been made. Further, the psychological theory of cognitive dissonancesuggests that attitudes follow behavior - so even if values were not pecuniaryto begin with, they would gradually become so as managers are expected tomeet monetary goals and are evaluated according to their success in doing so.In that case, the generation of commercial revenues may become an end in it-self, and nonprofit status is no longer a reliable signal for trustworthiness or forthe desire to "do good" - which were the rationales for tax privileges and pub-lic subsidies in the first place.

Additional policy questions are raised in this volume about the justificationfor and possible abuse of income-tax exemptions for nonprofits when auxiliary-good production is large. Chapter 5 shows that nonprofits with unrelated busi-ness income (UBI) report lower general and administrative expenses on theirForms 990 than those without UBI. This suggests that a large share of their gen-eral expenses are attributed to their unrelated business (UB) activities, therebysheltering the UBI from taxes even though these activities are not supposed tobe tax exempt. One might expect that, for similar reasons, buildings with largedepreciation potential would be treated as costs of the UB rather than of the non-profit. (The nonprofit does not need these costs to shelter its regular income,which is automatically tax exempt.) These costs may be true joint costs that can-not be disentangled between the nonprofit and UB, or they may be costs thatcould in principle be disentangled but cannot readily be monitored. In the lat-ter case, this attribution means that the government gets less tax revenue from

282 Estelle James

the UB activities than it "should," and other firms or households must be taxedmore as a result.

Nonprofits also benefit from property-tax exemption. Chapter 5 shows thatnonprofits tend to engage in commercial activities disproportionately in stateswhere these tax rates are high, giving them a large comparative advantage overfor-profits. For example, property used partially for UBI generation, particular-ly property without large depreciation potential, may be counted as nonprofitproperty that is exempt from property tax, thereby enabling nonprofits to under-cut for-profits producing similar goods, distorting the choice of organizationalform, and further reducing tax revenues.

In general, when one organization engages both in profit-making and non-profit activities, producing both public and private goods, it is natural to tryto segment the income and cost flows between the two so as to maximize thevalue of the tax exemptions. This is particularly problematic for public policyif the nonprofits cannot be counted on to use the resulting revenues, via cross-subsidization, for socially useful purposes. Since the value system in nonprof-its may itself change as commercialism increases, for the reasons given above,it may be that such exemptions should gradually be withdrawn once the com-mercial activities of an organization exceed a specified point.

(When) Should special constraints apply to nonprofit behavior?

Besides these questions as to the tax privileges of nonprofits, this book raisesquestions about whether and under what circumstances special legal constraintsshould apply. This issue is most clearly exemplified by the case of biotech re-search at universities, where constraints on commercial behavior were previ-ously strong but have recently been substantially weakened. The rationales forthe constraints were, presumably, the potential conflict between profit-makingbehavior and public-good-maximizing behavior, and an unwillingness to relyon nonprofit status alone to resolve this conflict in the public interest.

In the past, when federally funded basic research was carried on at universi-ties, results could not be patented and were regarded as open information, pub-lished in scientific journals and freely discussed at scientific meetings, as soonas (or even before) they were validated. In contrast, privately funded appliedresearch was more likely to be carried on at industrial laboratories, and resultswere patented before they were made public. As the constraint on universitypatents was removed, the line between these two types of research has becomeblurred, and the public-private good conflict has reemerged.

Patents are a compromise between the efficiency value of making informa-tion readily accessible once it is exists and the need to provide incentives toincur the costs that generate this information in the first place. The institutionof patents may increase research funding, but it also directs that funding into


areas where benefits can be appropriated and marketed within a reasonablyshort time, even if other areas such as basic research are more promising fromthe long-run, global perspective. Government grants to universities, togetherwith the nonprofit university value system, solved the investment problem with-out patents and therefore allowed the efficiency gains from more open accessand a broad, long-run perspective. Increased scarcity of government grants (rel-ative to demand) combined with removal of constraints on patenting rights in-troduced this conflict into academia. Although the commercialization of sci-entific discoveries via patents makes more resources available for research, itdirects the nature of the research toward the short and medium term, and runsthe risk of limiting the utilization of research results.

Moreover, under the new set of incentives, the allocation of university re-sources is likely to be strongly influenced by the profitability of outputs. Forexample, disciplines with commercial prospects may be given priority overthose without, rather than basing these judgments on educational criteria. Alarger share of university resources and faculty time may be devoted to the busi-ness rather than to the science of research and development. Faculty memberswhose research may lead to profitable patents or licenses, or collaborations withprivate industry, may be more likely to be hired and tenured than others; there-fore, they have an incentive to concentrate on these areas. Although access togrants has always been a factor in resource-allocation decisions at universities,the grants supposedly were awarded according to basic scientific merit; but nowcommercial profitability plays a stronger role. In effect, the removal of con-straints on the commercial activities of universities has shifted the trade-offbetween public and private research goods in favor of the latter. Have we cho-sen the right mix? Is this shift in values good for academia and for the econ-omy in the long run? This is the public-policy issue raised as legal constraintsare lifted and commercialism increases.

Conversions: How can society's interests be protected?

The most extreme form of commercialism occurs, as Chapter 7 showed, whennonprofits convert to for-profit status, as has been occurring most notablyamong HMOs, hospitals, and the Blues in the health industry. Presumably con-version occurs when the advantages of for-profit status outweigh the advantagesof nonprofit status. This may be socially desirable if price finance is now sowidely feasible that a public good has been turned into a private good, if accessto equity capital is important in order to expand facilities, and if donations areno longer a large potential source of revenue. The danger is that the decisionmakers may take into account only the gains to them personally, rather than thegains and losses to society as a whole. For example, if a group of insiders pur-chases at a low price nonprofit assets that have been accumulated out of tax-

284 Estelle James

exempt donations over the years, and if they cut the public services of the neworganization in order to maximize their profits, this may be a case where soci-ety as a whole has lost. Alternatively, if a foundation is set up with the proceedsof the asset sales, and management is turned over to people without a socialcommitment but who earn high salaries and perks, this may be another exam-ple of a social loss. Several instances of this sort are reported in this volume.Conversions thus raise issues around who should be permitted to buy and un-der what procedures (e.g., Should competitive open bidding be required?), howthe nonprofit's assets should be valued, how society should be compensated,and who should manage the compensation funds.

Conclusion

Over the past two decades subsidies to many nonprofit organizations have fall-en while the need for their services has grown due to government cutbacks. Thegrowing commercialism among nonprofits is often attributed to these two fac-tors: The conventional theory of nonprofit behavior spelled out earlier predictsthat they will rely increasingly on sales revenues to cross-subsidize their pre-ferred public goods as public and private donations fall. This volume providesseveral empirical examples from the museum, social-service, zoo, and publicbroadcasting industries consistent with this hypothesis. Though commercialismcan lead to abuses of tax privileges (e.g., through cost shifting), these may beminor and correctable in comparison with the gains in nonprofit services.

However, in the health and university-biotech industries, where the most dra-matic examples of commercialism have occurred, including extensive partner-ships with and conversions into for-profit firms, the driving force seems to havebeen more basic changes in underlying institutional and economic conditions.Moreover, perhaps as a consequence of these underlying changes, the wholeset of legal privileges and constraints faced by nonprofits has changed in theseindustries. Widespread medical insurance permits full pricing for hospital ser-vices, and special legal privileges for nonprofits thereafter decline; global com-petition puts greater emphasis on applied over basic research, and this leads tothe removal of legal proscriptions on the rights of universities to patent the re-sults of federally funded research. These basic changes alter in a discontinuousway the incentives of organizations to enter or stay in the nonprofit sector andthe behavior of those organizations that do remain. In particular, we see littleevidence of a reluctance to commercialize or of the use of sales revenues forcross-subsidization in these cases.

It appears that marginal and global changes in commercialism may havequite different causes and effects. A diminution in other revenue sources or anincrease in perceived need may explain marginal changes in commercialism,given the underlying set of economic, institutional, and legal constraints - con-


sistent with the cross-subsidization hypothesis. However, they do not explaindiscontinuously large changes in commercialism. These may be due, instead,to dramatic changes in constraints and the opening up of new sources of salesrevenues; the disutility from profitable activities and the cross-subsidization ofloss-making activities become far less important to nonprofits under these cir-cumstances. This suggests that pecuniary objectives may never have been farfrom the surface. When faced with large new opportunities for commercialism,many nonprofits seem quite willing to shed their altruistic cover and assume thevalues and behavior of for-profits.

Given that sales revenues appear to crowd out donations, tax exemptions forcontributions automatically become less important under these circumstances.However, public-policy issues are raised concerning:

1 the use of an organization's legal status to determine eligibility forincome- and property-tax exemptions and other privileges in situationswhere nonprofits and for-profits produce similar goods and/or sharesimilar values; and

2 the conditions for converting from nonprofit to for-profit status onceassets have been accumulated out of tax-exempt revenues.

In both cases the policy goal should be to ensure that public resources areindeed being used to provide public rather than private goods and to promotesocial rather than personal welfare - even though these categories are admitted-ly ambiguous and imprecise.

CHAPTER 15

Conclusions and public-policy issues:Commercialism and the road ahead

Burton A. Weisbrod

Introduction

Should anyone care, we asked in Chapter 1, whether the nonprofit sector be-comes increasingly commercial? The answer is yes, many people should care:taxpayers, who are affected by the subsidies to nonprofits; consumers of theseorganizations' services; governmental policymakers, as stewards for the publicinterest; owners of private firms, which are affected by nonprofits' activities;and nonprofit managers and directors, who are responsible for their organiza-tions. We have surveyed the landscape of the nonprofit sector, examining whyrevenue-generating sources are changing, what forms they are taking, and whatthe consequences are of their use.

We turn next to an overview of our findings, focusing on the building blocksof our analyses: goals (or mission) and constraints. That discussion, w.iich con-stitutes the majority of this chapter, is followed by a section identifying a num-ber of issues that call for targeted research, and another on what our findingsimply for the development of sound public policy toward the nonprofit sectorand its financing. In short, we now take stock of what has been learned and whatits implications are - for research and for public policy. A final section peersinto the future of the nonprofit sector in light of its search for revenue.

What has been learned about nonprofits9 commercialism?

Overview

The causes and consequences of commercialization are not abstract issues. Non-profit organizations' commercial activities are bringing revolutionary changes

I thank Jeffrey Ballou, Louis Cain, Estelle James, Laura McLean, and Dennis Young for helpfulcomments on an earlier draft.

287


in traditional behavior, and in the process they are blurring the distinction be-tween nonprofits and private firms. Prior chapters have identified many formsof nonprofits' commercialism, but new examples appear almost daily. A recentaccount involves major museums, such as Chicago's Art Institute, Shedd Aquar-ium, and Field Museum of Natural History: They have begun holding after-hours cocktail parties, "competing with establishments on the other end of thecultural spectrum: bars" and giving rise to their characterization as "meat mar-kets" (Caro 1997). Revenue is produced through admission fees and drink sales,but the question of whether such activities undermine the museums' cultural-preservation and educational missions remains.

In the hospital industry, a quite recent instance of aggressive marketing ofprofit-making activities is the decision by the largest nonprofit hospital in Nash-ville, Tennessee, "to build and operate a $15 million, 18-acre office and train-ing-field complex that it will rent to the Houston Oilers [professional footballteam]," which moved to Tennessee in the fall of 1997. The hospital's chairmanproudly reported that "Baptist Hospital will be on national TV. . . . When a play-er is hurt, a golf cart will rush onto the field with Baptist's name all over it."The CEO unabashedly spotlighted the market-oriented philosophy: "We arethe pioneers of a nonprofit hospital competing with investor-owned hospitals"(Langley 1997).

Commercialism by nonprofits should not just be dismissed as inefficient andcounterproductive. It offers real advantages, despite the problems it poses. Wehave found evidence of significant scientific advances resulting from cooper-ation between universities and private-sector firms (Chapter 9). We have alsofound evidence, in the higher-education, hospital, and museum industries, thatincreased commercialism in the form of unrelated business activity is efficientin the sense that it imposes little marginal cost, given the resources alreadyavailable for production of mission-related activities (Chapter 5); thus, it wouldappear to be inefficient to discourage, let alone prohibit, use of those resourceseven for business activities that are unrelated to tax-exempt missions.

Earlier chapters have documented not only the many fascinating new formsof profit-oriented activities in various industries, but also the complexity of de-termining whether use of those finance mechanisms, which mimic private enter-prise, is consistent with social missions. Our analytic framework (Chapter 3)implies that nonprofits select particular revenue-generating activities deliberate-ly, reflecting such variables as their ability to take the labor and capital used fortheir central, mission-related activities and use them to produce other, ancillaryoutputs. Choices also reflect the flexibility of the IRS in deciding which activ-ities will be taxed as being unrelated. Thus, forms of money-raising activitiescan be expected to evolve over time, in response to changes in technology orin IRS administrative practices, in addition to further cutbacks, should they oc-cur, in government support.

Conclusions and public policy issues: The road ahead 289

Nonprofits can be expected to increase efforts to push outward the marginsof the kinds of revenue-producing mechanism they use, continually probingIRS regulatory restrictions. These restrictions have a number of dimensions,including

1 whether an activity will be subject to taxation as an unrelated business,2 what accounting methods nonprofits may use to calculate their taxable

income - including the rules on how nonprofits are permitted to allo-cate joint costs so as to minimize their tax liability (Chapter 5)1 - and

3 which activities are regarded as inconsistent with the nonprofit's mis-sion and, hence, as justification for withdrawal of tax-exempt status.

With nonprofits' commercialization bringing both revenue needed to financetheir unprofitable, collective-goods mission and behavior that may be inconsis-tent with that mission, difficult choices result. The dilemma is real. Althoughour analyses generally reflect concerns about the balance between the drive forrevenue and adherence to social goals, the precise nature of sacrifices that non-profits make to raise revenue is seldom clear-cut. Increased commercializationis not ideal. The more important question, though, is whether its many forms,encompassing user fees and various ancillary outputs, are preferable to the alter-natives. Those could include greater dependence on government - throughgrants, tax subsidies, or tax encouragement of private donations - or diminishedoutput of collective goods.

Mission: Its vagueness and its effects on nonprofits

If it were entirely clear that certain commercial activities were in conflict withnonprofits' pursuit of their tax-exempt missions, the IRS would be revoking tax-exempt status far more often. What is more commonly observed, however, isambiguity as to the effects - ambiguity that reflects the typically broad scopeof mission. Thus, we have found that in light of such mission vagueness, it isunderstandable that nonprofits typically claim there is no conflict - that com-mercial sales activities are not simply generating money but are simultaneous-ly advancing organization mission. The Girl Scouts, for example, hold that saleof cookies is not merely a commercial activity to raise money, but an exercisethrough which young girls gain experience that helps build their character andimpart business skills (Chapter 10). What, then, are the limits to the goods andservices that might be sold in pursuit of that mission, or of AARP's mission ofpromoting the interests of older Americans (age 50 and over) through advo-

1 The issues are actually more complex. A nonprofit would wish not merely to minimize taxation,but to maximize aggregate revenue net of both taxes and any disutility from engaging in non-mission-related activities (as discussed in Chapter 3).


cacy, supportive services, and the provision of information and research (Chap-ter 10)?

The contrast with a profit-making firm is stark. The relative simplicity ofgauging a firm's profitability stands in sharp contrast to the complex, multi-dimensional social goals that, whether attained or not, often characterize non-profits. The complexity is not a weakness of nonprofits but rather a source ofconsiderable difficulty in identifying and measuring their success or failure,and, thus, in making them "accountable."

As our research proceeded it became increasingly apparent that missionvagueness is fundamentally important in predicting and evaluating nonprofit-organization behavior: It makes the operational definition of an "unrelated"business activity inherently problematic; it explains why nonprofits can under-take an ever-widening array of revenue-enhancing activities that the organi-zation can argue are related to mission and that the IRS finds difficult to term"unrelated" and, hence, taxable.

Mission vagueness permits managers and trustees to alter behavior in re-sponse to changing financial constraints while - at least arguably - continuingpursuit of the same goals. Means and ends are easily confused when goals arevague, and the wider the scope of mission, the greater is the challenge to orga-nization leadership, prospective donors, and regulators to determine whethera profitable commercial activity is mission related.

Narrowness of mission focus is not inherently desirable, nor is wide scopeinherently undesirable. Breadth permits latitude for organizations to adjust tochanging external conditions, and such flexibility is necessary for efficient useof resources. Drawing a distinction, however, between vagueness of missionand flexibility of resource use is both difficult and important for two reasons:to predict nonprofits' responses to exogenous changes in regulatory, financial,and technological constraints, and to assess the accomplishments of nonprofitsin achieving their goals.

Broad mission scope poses a severe regulatory challenge: to determine whenthe "border" is being crossed between a true mission-focused activity and amere money-raiser. This is evident throughout our industry studies (Chapters8-13). It is also explicit in the admission of the hospital CEO, cited above: "Noquestion," he reportedly admitted; in the selection of revenue-generating activ-ities, "we push the envelope" (Langley 1997).

A question posed in many chapters is whether such aggressive commercial-ism is causing nonprofits to sacrifice subtle elements of their social, collective-good, mission - for which the subsidies and tax exemptions are given - in theinterest of generating revenue. In the case of university-industry research col-laboration, universities have agreed to delay or even avoid publication of re-search findings in order to protect proprietary secrecy, actions that conflict with


their traditional objective of full dissemination of knowledge. We have seen re-peatedly how broad goals make it difficult to determine whether a commercialactivity that garners revenue brings any adverse side effect with regard to mis-sion. Without that determination, the true extent of nonprofits' contributionscannot be identified.

Nonprofits' pursuit of mission: Effects on private firms

To assess the overall contribution of nonprofits it is not enough to focus on thoseorganizations. In addition to the impact of their commercial activities on theirown revenues and on achievement of their organization goals, there are effectsof those activities on other sectors of the economy - private enterprise, for one.We have seen that nonprofits are expanding into activities that compete with pri-vate firms. This may or may not be socially efficient, but it is clearly adverseto the firms. Such competition varies greatly among industries: Food pantriesand homeless shelters use technologies that offer few opportunities for com-mercial expansion, but museums can open retail shops (Chapter 12), and pub-lic television stations can air announcements for commercial firms (Chapter 13).In the hospital industry, where the competition for patients who pay is especial-ly keen, the legitimacy of subsidies to nonprofits is increasingly being ques-tioned in light of evidence that their provision of medical care to the indigentclosely resembles that of for-profit hospitals (Chapter 8). In the mapmakingindustry, the nonprofit National Geographic Society has been charged withhaving an unfair advantage in its nascent cartographic competition with the for-profit firm Hammond Maps (Hays 1997).

We have also found evidence, for various industries, that nonprofits' tax ad-vantages, although ostensibly given to support mission-related activities, real-ly extend beyond them to unrelated activities. Thus, most of nonprofits' taxableactivities actually go untaxed, as nonprofits are typically successful in allo-cating enough joint costs to the taxable activities to eliminate reported profit(Chapter 5). Thus, even when nonprofits expand into commercial activities thatare clearly unrelated to their missions, they generally have cost advantages overprivate-enterprise competitors.

Nonprofits' commercialism is not inevitably detrimental to private enter-prise. Sometimes the effects are favorable, involving not competition but col-laboration. For example, as Chapter 9 documented, universities are increasing-ly engaging in scientific research in cooperation with for-profit biotechnology,chemical, and pharmaceutical firms. Clearly, both the universities and the firmsexpect to benefit from joint ventures or other collaborations. In such cases, theissue is not any adverse effect of nonprofits' commercialism on their privatepartners, but of effects elsewhere in the economy.


Nonprofits' pursuit of mission: Effects of commercialism ongovernmental tax revenues

One cause of social concern is that the gains to nonprofits and private firms maybe at the expense of third parties: For example, government tax receipts mayfall because nonprofits find ways to sell their tax advantages to private firms.One instance of such "tax arbitrage" across sectors occurred in higher educa-tion when, as we recounted in Chapter 1, a nonprofit college sold all its build-ings to a private firm and then leased them back, thereby "selling" their depre-ciation. Such an allowable expense has no value to a nonprofit organization,which pays no tax on its profit, but depreciation is a valuable deduction to a pri-vate firm that can use it to reduce taxable profit.

Such a sale of nonprofits' tax advantages is not necessarily socially undesir-able, for it is a source of revenue for nonprofits. However, unrelated businessincome is also a source of revenue, and yet Congress has seen fit to tax it. Theissue is what the limits should be on how nonprofits may use their resources,including tax advantages, to expand income. Currently, there is a more gener-al public policy against transactions that have no justification other than to re-duce taxation - which explains why the IRS disallows them, even when non-profits are involved.

Nonprofit organization goals and outputs

How problematic any form of commercialism is cannot be determined withoutassessing the organization's "outputs," its success in achieving its social goals.Because this is so difficult, it is common for the IRS to examine procedural evi-dence rather than the harder-to-measure outputs. It may explore, for example,whether charitable assets are being used for private or "noncharitable" pur-poses; but this input-oriented indicator, while useful, tells little about what theorganization is accomplishing overall. Similarly, it may attempt to distinguishpursuit of social mission from officers' pursuit of personal gain, but this too isdifficult. Clearly, there are potential conflicts, but they are hard to observe anddo not directly reflect outputs. Charges of overcommercialization, such as thosemade by critics of Nashville's Baptist Hospital - "When marketing and person-al glory dwarf charity care [in a nonprofit hospital], the mission has changed,"and "A nonprofit hospital shouldn't cross the line to spend millions of dollarsfor blatant advertising and promotion" (Langley 1997) - cannot be assesseduntil there is better understanding of nonprofits' outputs. Until operational mea-sures of nonprofits' outputs in various industries are developed and standard-ized, the debate will continue over how to operationalize such allegedly nega-tive influences as the pursuit of "personal glory" and "blatant advertising" ona nonprofit's outputs.


Similar issues, involving new and expanded commercial activities, havebeen seen in every industry we examined - hospitals (Chapter 8), universities(Chapter 9), social-service agencies (Chapter 10), zoos (Chapter 11), art muse-ums (Chapter 12), and public broadcasting (Chapter 13). The bold pursuit ofprofit has brought to the surface of the debate not only tax-related questionsbut, more broadly, the consistency of commercialism with a nonprofit's social,collective-good goals. Our earlier evidence that missions are often stated in ex-tremely broad terms implies that debate will continue to grow over whether aparticular nonprofit's commercialism has gone "too far." The IRS, with verylimited resources for enforcement, can be expected to rely heavily on organiza-tion self-reports, with each nonprofit deciding how far it can go in terming itsactivities related, and only occasional IRS disputation.

However, whether commercialization is, on balance, enhancing or impedingachievement of nonprofits' social goals involves much more fundamental mat-ters than the size of the IRS budget for enforcement. The vagueness of manymission statements implies an inevitably large element of judgment to definewhich specific activities are or are not "substantially related" to the mission.

Revenue and cost constraints

Revenues limit what any organization, nonprofit or other, may do, and the na-ture of those limitations is not simple if revenue sources are interdependent andresponsive to the organization's decisions. We have seen that nonprofits maybe concerned, for example, about the effects of their decisions to increase rev-enue from one source or another - user fees or ancillary activities - on theirrevenue from yet another source - contributions, gifts, and grants (CGG). Suchincreased self-help could lead either to decreased or increased CGG, depend-ing on whether prospective donors favor or oppose those efforts. Because ofsuch revenue interdependencies, a nonprofit would not necessarily maximizeits total revenue by trying to maximize revenue from each individual source;an increase in revenue from one source could diminish revenue from anothersource even more.

There are still other fiscal interdependencies and associated constraints. De-cisions to produce more of certain salable outputs can affect both revenues andalso costs. Thus a nonprofit might be enabled to expand its mission-related out-puts not only by increasing revenue but also by decreasing costs, and increasedproduction of some ancillary activity could reduce production costs for the or-ganization 's mission-related outputs (Chapter 3).

Resource interdependencies can take other, complex forms. Increased com-mercial activity intended to generate revenue to expand production of mission-related collective goods and services can affect the types of manager employed,the types of director selected, and thereby the organization's decisions on how


it pursues its mission. This can be seen in the nonprofit National GeographicSociety's recent restructuring and greatly expanded emphasis on commercial-ly profitable activity. In order to increase revenue for its expeditions and edu-cational activities, it is producing full-length feature films, developing commer-cial cable television partnerships with NBC, and producing, in the fall of 1997,the first National Geographic Road Atlas. To do this it is bringing in new man-agement, not from other nonprofits but from private business. The effects onthe quality of research scholarship for the magazine articles are uncertain, butthere is already evidence of pressure to diminish the time allotted for field re-search (Hays 1997). Whether the trade-off is socially efficient cannot present-ly be resolved, but the balance clearly is changing.

All these interdependencies among revenue sources, costs, and organizationbehavior add a significant dimension to understanding nonprofits' financialconstraints. It has long been understood by economists and argued by nonprof-its' leadership that private donations are affected by changes in the individualincome-tax law, such as reduction in marginal tax rates or increases in require-ments for itemizing deductions for charitable giving (Clotfelter 1990). What hasscarcely been noted, however, is that those effects can be expected to triggeradditional forces, especially commercial activity, so that the final effects on non-profits' revenue may be quite different. Our evidence is that there is great vari-ability among industries in their offsetting of losses in contributions, gifts, andgrants by increasing commercial activity from "program services" (Chapter 6),and this suggests that nonprofit-organization behavior is more complex thanhas been commonly recognized. At this point, however, little is known aboutthe causes of this behavioral heterogeneity across industries. Opportunities toexpand commercial markets profitably may differ greatly, as may organizationalwillingness to adopt or increase user fees or enter new ancillary markets.

Potential conflict between revenue source and goal attainment

Both forms of commercialism, user fees and ancillary activities, have the po-tential to generate revenue, and both involve private goods, which can be with-held from any consumer. The key difference is that those persons served by an-cillary goods are of no special concern to the nonprofit's mission, whereas userfees are charges made to consumers who constitute at least a portion of the or-ganization's target population. We have highlighted the potential conflict be-tween pursuing revenue via user fees and achieving a nonprofit's mission whenthat mission is to reach some specific population, such as the poor or all school-children (Chapter 4). It is certainly feasible to impose charges on patients ata research hospital, where care is related to the clinical research activities, oron undergraduate students, who constitute an undergraduate college's targetpopulation, or on museum visitors, whose education is at least one element of


the nonprofit's mission. However, barring perfect price discrimination, we haveshown the danger that user fees will price out of the market at least some mem-bers of the target population. We have seen how increasing revenue through userfees, as through the institution of admission fees at museums and zoos, trou-bles nonprofits because of the negative effect on attendance and, hence, on theireducation mission (Chapter 11). In this way too, efforts to relax resource con-straints can have adverse side effects on the achievement of organization goals.We have emphasized, moreover, that the organization's view of which activ-ities are and are not central to its objectives may differ from the distinction madeby the IRS, under the tax law, between related (untaxed) and unrelated (taxed)activities.

Were it not for the links between goals and the revenue-related activities usedto attain them, a nonprofit would act as a profit maximizer in those product mar-kets where they seek simply to raise revenue. The model presented in Chapter3, however, implies that nonprofits may not wish to maximize profits in suchmarkets because they view the activities themselves as inconsistent with themission; in at least some industries, such as zoos and public television, it seemsclear that admission fees and sale of broadcast airtime are intentionally restrict-ed because of a sense that generating more revenue from those sources wouldbe inconsistent with mission. At the same time, however, it remains very diffi-cult to distinguish such goal incompatibility from the effects of revenue inter-dependencies in which some form of commercial activity is disliked by pros-pective donors, causing them to decrease contributions (Chapter 14). In suchsituations a nonprofit that did seek to maximize profit in commercial markets,in order to subsidize its collective-goods outputs, would not act to maximizeprofit in each specific market where it can charge prices. The organization, rec-ognizing negative marginal effects across markets, would balance the effect ofincreasing profit from one commercial activity against the loss in maximum at-tainable profit from other sources.

Behavioral differences among industries

Many of the goals, constraints, and interrelationships we have identified as char-acteristic of nonprofits appear to differ systematically across industries.2 Evenif we have accurately assessed the basic forces operating on nonprofits - thatis, if we have identified the essential character of their objective functions, thenature of trade-offs among revenue sources, and the production-cost interde-pendencies - there can be large differences in their effects among industries.

2 They may also differ between small and large nonprofits, between younger and older nonprofits,or in other ways not explored here explicitly.


In fact, not only can there be differences, they can be expected. For example,the potential for realizing profit from ancillary activities, which do not contrib-ute directly to the organization's mission but can generate profit to be used toaugment mission-related activities, will depend on the production technologyused for the pursuit of mission. That technology, which varies among industries,is likely to differ widely in its profit potential: For example, a nonprofit dedicat-ed to perform ballet would appear to have far fewer opportunities for ancillary-goods profit than a hospital.

Such differences in technology may help to explain the substantial variationwe found among industries in the effects of changes in revenue from donationson commercial activity from "program service revenues" (Chapter 6). For someindustries we found strong negative relationships between receipt of revenuefrom donations and subsequent revenue from program services, as the simplecrowding-out hypothesis suggested. For other industries, however, there was norelationship; in still others, a positive one. One possible explanation is that dataavailable from the IRS Form 990 return, the principal current data source, donot distinguish the two sources of commercial revenue, user fees and revenuefrom ancillary activities. A second complexity is that the simple crowding-outrelationship may actually operate in all industries, yet be offset by interdepen-dences among revenue sources or by cost interdependencies.

Tax-related constraints: Joint-cost allocations betweentaxed and untaxed activities

As we observed earlier (see "Nonprofits' pursuit of mission: Effects on privatefirms"), it appears that when nonprofits do report engaging in activities that areunrelated to their tax-exempt mission - and most do not, although the ranks aregrowing rapidly - they generally allocate so much cost to the taxed activitiesas to eliminate all tax liability. This is true despite our evidence that costs areheavily "joint." In this way, the actual incremental cost of the unrelated busi-ness activity is essentially zero.

From an examination of the relationship between a nonprofit's taxed and un-taxed activities it is a short step to recognizing the effects of nonprofit commer-cial activity on private firms. Although the specific effects on private enterpriseof the nonprofit sector's commercial expansion have not been a major focus ofthis volume, our examination of cost allocations between nonprofits' taxed (un-related) and untaxed (related and excluded) activities has shed some light onthem. As a result of the joint-cost allocations, unrelated business (UB) activi-ties, which presumably have no justification except to generate profit for cross-subsidizing the mission, are reported to be negative by some two-thirds of non-profits that file Form 990-T returns reflecting taxable activity. Moreover, these"losses" are so great that they exceed the modest profit reported by the others,so that aggregate nonprofit-sector profit from all organizations reporting UB


activity is negative! It is striking to note that while nonprofits are reporting in-creased total UB activity over time, they are also reporting increased losses.

These findings appear to be the result of accounting cost allocations, not gen-uine unprofitability. In contrast with data reported on Form 990-T returns show-ing that costs of UB activity increase more than revenues, we found, in aggre-gate data and in the hospital and higher-education industries, that actual costsdo not increase at all; by this reckoning, UB activities are extremely profitable.Thus, the generally negative profits reported by most UBIT filers make clearthat a great deal of UB activity carries no tax liability. Cost allocations betweennonprofits' untaxed and taxed activities are not merely technical issues; theyhave broad consequences on resource-allocation neutrality across ownershipsectors.

Altering constraints: Converting a nonprofit to afor-profit

What we have termed "the ultimate commercialism" is the conversion of a non-profit organization to for-profit status (Chapter 7). A conversion may or may notimply a change in organizational goals. It does so insofar as nonprofits pursuegoals that involve provision of goods and services that are socially valuable butprivately unprofitable, whereas for-profit firms do not - a view that finds expres-sion and evidence in many chapters. On the other hand, whatever social policymay intend, nonprofits may actually differ little from private firms. This is ar-gued for the hospital industry (Chapter 8); for conversions of hospitals andHMOs to for-profit status, where private gain to nonprofit organization exec-utives seems to have motivated at least some recent activity (Chapter 7); andmore generally, in an overview (Chapter 14).

Our analysis of conversions demonstrated the difficulty of discerning whena conversion is a reallocation of resources from a less efficient to a more effi-cient use, and when it represents merely a private benefit to insiders. Privategain does not preclude social efficiency; neither, though, does it ensure it. Sincethis basic issue of distinguishing efficient reallocation of resources from sim-ple redistribution of wealth is the same across all nonprofit-organization in-dustries, the current prominence of conversions in health care - hospitals andHMOs - suggests that conversions may occur, with the same attendant prob-lems, in other industries. Which industries will be the next to move aggressive-ly to convert nonprofits to for-profits is not clear, just as widespread conversionswere not foreseen for the hospital and HMO industries even a decade ago. Thereare two questions, not dealt with in this volume but nevertheless important:First, from an efficiency perspective, which industries will experience the threatof extinction of nonprofits unless they have greater access to capital? Second,from a wealth-distribution perspective, which industries offer the greatest po-tential opportunities for private gain from conversion? Speculation suggests thatcapital-market constraints on nonprofits may become increasingly severe in


nursing homes, where an aging population will present opportunities for profit-able expansion into regional, if not national, chains. Regarding opportunities forprivate benefit, the massive wealth in nonprofit colleges, universities, and mu-seums continues to offer financially tempting targets.

Assessing our analytic framework

The framework adopted in Chapter 3 focused on nonprofits' choices among out-puts and among revenue sources that determine organization budget constraints,as well as on the interdependencies among outputs and among revenue sources.It proved to be useful as a base for structuring industry studies and posing re-search questions. Still, as with any relatively simple characterization, its smallnumber of variables cannot capture fully the great variety of behaviors exhib-ited by the more than a half-million charitable nonprofit organizations.

The goals-constraints structure assumed that nonprofits have goals that areclear enough to constitute meaningful objectives. That is often not the case,however, as we have repeatedly noted. In the absence of the functional equiv-alent of stockholders pursuing profitability, nonprofits' objectives may or maynot be translatable into terms analogous to profit maximization. Moreover, be-cause of mission vagueness, the actual goals pursued by a nonprofit may reflectmanagerial preferences, and, through a managerial sorting process, those pref-erences can be expected to reflect the (revenue and other) constraints faced byprospective managers. Thus, an organization's constraints and its actual goalsmay not be independent.

Implications for research

Though we have learned much about the forces influencing nonprofits' increas-ing commercial activities, some important matters remain unclear. This sectionpoints to the principal directions that our work has identified for future research.

The search for generalizations as to nonprofits' behavior remains a challenge.So, too, is the identification of differential patterns of behavior across industries.The following five dimensions, involving elements of nonprofit-organizationgoals and fiscal constraints, deserve increased research attention. Whether theinterest is in predicting nonprofits' behavior or in developing wise public pol-icy, it is important to increase understanding of the following:

Nonprofit-organization objective functions

1 The breadth and multiplicity of goals of individual nonprofit organi-zations, and the differences among industries in the trade-offs beingmade among goals;


2 the importance to various nonprofits of not simply providing certaingoods or services but of reaching particular target populations; and

3 the degree of willingness to engage in production and distribution ofancillary goods in order to raise funds for subsidizing production oftheir mission-related goods, as well as the willingness to charge userfees and to adopt commercial approaches to expand donations.

Budgetary constraints

4 The opportunities - technological and market - available to variousnonprofits to enter private-goods markets and compete successfullywith private firms and other nonprofits, or to benefit from joint ven-tures with private firms; and

5 the interdependencies among nonprofits' various revenue sources andamong costs of producing various mission-related and ancillary out-puts.

Chapter after chapter focused on the importance of revenue constraints fromvarious sources, but the framework did not deal explicitly with uncertaintyabout the promise of each prospective source. That is, the revenue constraintslimiting what a nonprofit can do are not fully predictable. As a result, nonprof-its' decisions on how much to pursue each of a variety of revenue sources mayreflect not only the expected revenue from each alternative but the differinguncertainty as well. A particular source of revenue may appear to be quite un-certain yet turn out to be quite stable or otherwise predictable; alternatively, asource that seems highly predictable may turn out not to be. Study is needed,therefore, of uncertainty, both as anticipated and as realized. Governmentgrants, for example, have come to be increasingly uncertain, as anticipated inan era of cost cutting and rethinking of the role of the federal government. Infiscal 1996, to illustrate, legislative delays caused funding agencies such as theNational Science Foundation and the U.S. Geological Survey virtually to dis-continue grants to university scientists for at least four months. According toone account, "field studies and laboratory experiments were being canceled.. . . Admissions to graduate programs [were] also being affected because of theuncertainty regarding the amount of funds available for research assistantships"(Menke 1996). Eventually appropriations restored the funding. Still, the gener-al questions regarding uncertainty are as follows:

1 Do nonprofits perceive various revenue sources as differentially reli-able?

2 Have some sources actually proven to be more dependable than oth-ers?

3 How do nonprofits respond to differences in predictability of revenuefrom alternative sources?


With respect to all the issues in this section, there is need to learn more abouttheir importance and influence not only for the nonprofit sector as a whole butalso for individual industries. Disaggregation is very much needed, given ourfindings of distinct interindustry behavioral patterns that are as yet poorly un-derstood. At the same time, while there are limitations to a one-size-fits-all mod-el of nonprofits, there are also disadvantages to emphasizing their diversity, forthat limits identification of common characteristics and, hence, the ability togeneralize across industries.

An industry perspective also highlights the influence of competition on non-profits' commercial activity. It would be useful to learn whether greater com-petition, whether from other nonprofits, private firms, or government agencies,sparks commercial activity of nonprofits.

Implications for public policy

Our analyses of nonprofit commercialization do not lead inexorably to propos-als to introduce major reform into the nonprofit sector, its activities, or its rev-enue sources. Nevertheless, they do suggest some directions for change.

It is apparent that public policy affects the ways nonprofits are financed, thekinds of output they produce, and the mechanisms through which outputs aredistributed. The instruments of public policy are far more complex than is com-monly recognized, going far beyond tax deductibility of donations, tax exemp-tion of nonprofit-organization surplus or profit, and outright subsidies. Publicpolicy also encompasses incentives for nonprofits to produce profit-generatingproducts, regulations that affect the use of accounting procedures that minimizeor eliminate taxable profit, and rules that affect commercial alliances with, orconversions into, private firms. In this section we highlight a number of waysin which our findings link with public policy toward nonprofits.

Our findings that revenues from donations, user fees, and ancillary activitiesare likely to be interdependent - a change in one affecting the availability ofrevenue from one or more of the others - have major implications for public-policy development. A government policy that alters the availability to nonprof-its of one source of revenue can have unintended and unexpected side effects.To illustrate: As Congress and the administration consider cutting governmentgrants for the arts and the humanities, as well as for antipoverty programs, thedebate has focused on whether private donations will offset governmental cuts.The interdependence of government grants and private donations to nonprofitsis certainly relevant to these organizations' budget constraints. However, ouranalysis has focused on another revenue interdependence: the effect of such adecrease in government grants on nonprofits' pursuit of commercial activity inthe forms of increased user fees or production of ancillary goods. We havefound some evidence that reduced donations (from all sources) lead to increased


commercial activity. Industry disaggregation, however, disclosed substantialvariation among industries, with reduced donations being associated with in-creases in commercial revenue in some industries but with decreases in others(Chapter 6). Such varied responses of nonprofit organizations to cuts in contri-butions, gifts, and grants were not obvious a priori, but in light of our analysisthey are not surprising: Interindustry differences in organizational goals (in-cluding preferences for and aversion to particular activities), in interdependen-cies among costs, and in technologies and market demands that present vary-ing commercial opportunities, can each influence responses to an exogenouschange in revenue from any particular source.

Recommendations on improving data about nonprofits

Sound public policy toward nonprofits requires information about performance.Data with which to analyze nonprofit-organization behavior and changes overtime are very limited. IRS Form 990 returns - the major source of data, whichwe utilized in a number of chapters - do not, and cannot be expected to, coverthe entire range of issues on which information is needed. That form, valuableas we have found it to be, is inadequate in several respects. It needs to be supple-mented with data on organization output and on physical quantities of inputs,to aid in the assessment of nonprofits' efficiency and social contribution.

The following recommendations are made regarding data development andreporting:

The IRS Form 990 return should be reconsidered, with the goal of eliminatingquestions that have proven to be of little use to the IRS, researchers, or public -policy makers, and adding others, especially focusing on "outputs."

Agreement on what constitute outputs and how to measure them is elusive;but without it, the goal of nonprofit accountability is unattainable. Thus:

A process should be initiated in which researchers, nonprofit-sector leaders,government officials, and private-sector representatives of industries affectedby nonprofits' commercial activities develop consensus measures of nonprofitoutputs, including the distribution to target groups. Outputs should be industryspecific, and selected to identify measures evidencing success or failure.

Makers of public policy should confront the fact that a growing fraction ofnonprofits are reporting unrelated business income, as well as the fact that ex-tremely little taxable income results. Public policy should reflect better under-standing of the growing complexity of nonprofit-organization activities, as moreof them produce outputs clearly unrelated to their exempt purpose. To advancethis process:


Data from the Form 990 return and the Form 990-T return should be coordi-nated and published. The Form 990-T data, though confidential, can be aggre-gated in ways that would preserve confidentiality yet expand understanding ofthe connections between related and unrelated activities.

Nonprofits compete with private firms and government agencies in many in-dustries. Unless nonprofits' activities are viewed in an industrywide context,there will be little basis for determining in what ways they either accomplishgoals not attained by other types of organization, or affect decisions of thoseorganizations.

Data on nonprofit-organization activities should be supplemented by analogousdata for the private-enterprise and governmental organizations in the industry.This is not to say necessarily that this should be a responsibility of the IRS, letalone that it should be related in any way to the Form 990 return, but only thatindustrywide data are needed to assess the role of nonprofit components.

Recommendation on public policy toward nonprofits'commercialization

Public policy directed at nonprofits and their financing has effects that extendfar beyond the nonprofit sector. Changes in personal income-tax rates or in gov-ernment grants, for example, have more complex consequences than have gen-erally been realized. Reductions in tax rates or grants initially reduce nonprof-its' revenues, which can cause nonprofits to alter their commercial activity; this,in turn, will have additional effects on nonprofits' revenue if donations respondto commercial activity. In addition, changes in commercial activity can alterthe way a nonprofit's mission is seen and pursued by its leadership, as we havealready noted. Because such reverberative effects have significant implicationsfor the direction of public policies affecting nonprofits:

When cuts are considered in federal and state government support for collegesand universities, hospitals, the arts, and so on, attention should be directed toside effects. Such cuts - in grants, contracts, or tax subsidies - may drive non-profits to increased commercialism. In the process, private firms would be af-fected by the extended competition; and government (at various levels) couldbe affected insofar as commercialism drives nonprofits into alliances with pri-vate firms in ways that reduce tax payments.

Recommendation on encouraging private contributions

The principal revenue alternative to commercial activity is increased contribu-tions, gifts, and grants. There is, in short, an alternative to the choice between


increased commercial activity and retrenchment: The drive for commercial rev-enue can be redirected. Worthy of attention, though not devoid of problems, arechanges in tax laws to encourage private donations.

Ways to increase private donations should be explored. Mechanisms worth ex-amining include making it easier to take tax deductions for charitable giving,instituting limited tax credits, and permitting taxpayers to earmark a portion oftheir tax payments for specific nonprofits.

Recommendation on regulating conversions of nonprofits tofar-profits

Conversions of nonprofits to for-profits deserve greater public-policy attention.They are sometimes necessary, as is capital mobility more generally, to encour-age efficient use of resources. At the same time, conversions should protectassets built up with the assistance of charitable donations and government sub-sidies, and should avoid wealth transfers beyond what are needed to motivateefficient reallocations. Should there be a rush of nonprofits in asset-rich in-dustries other than health care that seek to change to for-profit enterprises, andthereby become free of the nondistribution constraint, the same public-policyissue would arise: How to balance concerns about efficient resource use againstconcerns about equity toward donors and avoidance of unfair enrichment ofinsiders.

There is no basis for predicting where pressures for conversions may arisenext, just as there had been little recognition over the past decade of the impend-ing sea change in the nonprofit health-care sector. At the same time, it is note-worthy that the vast wealth in nonprofit colleges, universities, and museumspresents tempting opportunities for private gain through conversions. Thus:

Public policy toward conversions should include development of better mech-anisms to distinguish efficient resource reallocations from self-serving privategain. These should include requirements for arms-length transactions and thefunctional equivalent of open competitive bidding. Attention should also be di-rected to identifying industries likely to see major efforts to convert.

Recommendation on relating public policy to the diversity ofnonprofits

One of our major conclusions implies that great caution is required before im-plementing any sweeping reforms for the nonprofit sector as a whole. Industrycomponents of the nonprofit sector differ greatly in their dependence on dona-tions relative to commercial revenue, and in the revenue opportunities available


to them. Any simple "solution" to the nonprofit-sector finance problem can havevastly varying effects across the wide expanse of nonprofit organizations. Thus,guidance for flexible public policy toward nonprofits seems justified by ourfindings:

Avoid "sweeping" reforms of nonprofit finance mechanisms. Because of differ-ences among industries in nonprofits' access to capital, access to user fees, thesuitability of their technology for the production of ancillary goods, and theirability to overcome free-rider problems, any broad reform is likely to have quitedifferent effects among industries having nonprofit components.

Commercialism: The road ahead

For decades the nonprofit sector has been small enough, and its methods of rais-ing revenue innocuous enough, to be largely out of sight. Success has changedthis; with growth has come visibility. The increased public awareness of non-profits reflects growth in their numbers, in their share of the gross domesticproduct, and in their use of commercial revenue-raising instruments tradition-ally identified with private enterprise. In light of nonprofits' receipt of publicsubsidies, demands for restricting their activities and for holding them more ac-countable for demonstrating their social contribution and its value are increas-ingly evident. Attacks on their tax exemptions, a widening array of penalties("intermediate sanctions") now available to the IRS for various regulatory vio-lations, and decreases in government funding, or shortfalls relative to "need,"all demonstrate growing public concern.

Because nonprofits often see their role as performing many elements of theunfinished job of governments, they tend to see their missions as forcing themto become more creative and imaginative in generating revenue. Competitionfor resources, however, drives all organizations - nonprofit, for-profit, and gov-ernmental - to search for new markets, and markets new to one type of organi-zation are likely to be already occupied by one or more others. When nonprof-its seek opportunities to raise revenue by producing goods or services they cansell profitably, they enter the domain of the private, for-profit firm. There, thedrive for profit demands attention to costs and revenues and, hence, to avoid-ance of activities that, however desirable they may be from a societal perspec-tive, do not generate profit. What happens when nonprofits' pursuit of revenuedrives them to act like private firms?

The answer, as previous chapters have shown, is not simple. There are dan-gers of goal displacement, as the social mission slips from sight in the drive forrevenue. Aggressive marketing and merchandising produce almost inevitableconflict, sometimes forcing organizations to choose between "capitalist appe-tites and . . . integrity" (Farhi 1997, 13). Peering ahead, it is likely that:


missions of nonprofits engaged in commercial activities will grow more ambiguousthrough time. New demands on senior management to pay attention not only to nonprofitbut also to for-profit goals, the adoption of new structures such as joint ventures that cre-ate mixed missions and messages for participating entities, and the tendency of seniormanagement to look at activities from the perspective of their contribution to revenuesmay create an environment in which nonprofits must work especially hard to keep theircharitable mission in daily focus. Increased responsibility will likely fall on the Boardsof Directors of commercial nonprofits to insure that a dilution of charitable mission doesnot occur. . . . (Howard P. Tuckman, personal communication, 1997)

Whether the future finance bases and commercial activities of the nonprofitsector as a whole, and of specific industries within it, will be extensions of his-torical paths is not predetermined. Much depends on nonprofits' success in find-ing new sources of revenue. Barring unlikely changes in government grantsor tax-law encouragement of private donations (of money and labor), increaseduser fees and, especially, expanded commercial activities are the most promis-ing revenue sources. If nonprofit-organization commercialism continues to in-crease in magnitude and in scope, nonprofits will become even more entwinedwith private enterprise and less distinct from it.

APPENDIX

IRS Forms 990 and 990-T for nonprofitorganizations

Appendix: IRS Forms 990 and 990-T 307

990 Return of Organization Exempt From Income TaxUnder section 501 (c) of the Internal Revenue Code (except Mack lung benefittrust or private foundation) or section 4947(a)(1) nonexempt charitable trust

Note: The organization may have to use a copy of this return to satisfy state reporting requirements.

OMB No. 1545-0047

This Form isOpen to Public

Inspection

A For the 1996 calendar year, OR tax year period beginning , 1996. and ending

B Check if:I I Change of address[~1 Initial returnCD Final returnQ Amended return

(required also forState reporting)

G Type of organization—• • Exempt under section 501 (cK ) < (insert number) OR • Q section 4947(a)(1) nonexempt charitable trustNote: Section 501 (cH3) exempt organizations and 4947(a)(1) nonexempt charitable trusts MUST attach a completad Schedule A (Form 990.

Pleas*use IRSlabel orprint ortype.See

SpecificInstruc-tions.

C Name of

Number a

City, towr

organization

nd street (or P.O box

. or post office, state,

if mail is not delivered to Street address) ^oom/suite

and ZIP+4

D Employer identification number

E State registrati

F Check • •

ion number

if exemption applicationis pending

H(a) Is this a group return filed for affiliates? D Yes • No

(b) If Yes," enter the number of affiliates for which this return is filed;. . •

(c) Is this a separate return fileo by an organization covered by a group ruling? I I Yes I I No

I If either box in H is checked "Yes," enter four-digit group

exemption number (GEN) •

J Accounting method: O Cash O AccrualD Other (specify) >

Check here • Q if the organization's gross receipts are normally not more than $25,000. The organization need not file a return with the IRS; but if it receiveda Form 990 Package in the mail, it shoulo file a return without financial data Some states require a complete return.

Note: Form 990-EZ may be used by organizations with gross receipts less than $100.000 and total assets less than $250,000 at end of year.

Revenue, Expenses, and Changes in Net Assets or Fund Balances (See Specific Instructions on page 9.)

1 Contributions, gifts, grants, and similar amounts received:' la

1ba Direct public support

b Indirect public support

c Government contributions (grants) I 1c

d Total (add lines 1a through 1c) (attach schedule of contributors)

(cash $ noncash $ ) .

2 Program service revenue including government fees and contracts (from Part VII, line 93)

3 Membership dues and assessments

4 Interest on savings and temporary cash investments

5 Dividends and interest from securities

6a Gross rents

b Less: rental expenses

c Net rental income or (loss) (subtract line 6b from line 6a)

7 Other investment income (describe •

8a Gross amount from sale of assets other ^ securities

than inventory

b Less: cost or other basis and sales expenses.

c G a i n o r ( l o s s ) ( a t t a c h s c h e d u l e ) . . . .

d Net gain or (loss) (combine line 8c, columns (A) and (B))

9 Special events and activities (attach schedule)

a Gross revenue (not including $

6a

6b

8c

9a

9bcontributions reported on line 1a)

b Less: direct expenses other than fundraising expenses

c Net income or (loss) from special events (subtract line 9b from line 9a)

10a Gross sales of inventory, less returns and allowances

b Less: cost of goods sold 10b

c Gross profit or (loss) from sales of inventory (attach schedule) (subtract line 10b from line 10a)11 Other revenue (from Part VII, line 103)12 Total revenue (add lines Id , 2, 3, 4, 5, 6c, 7, 8d, 9c, 10c, and 11) .

13 Program services (from line 44, column (B))

14 Management and general (from line 44, column (C))

15 Fundraising (from line 44, column (D))

16 Payments to affiliates (attach schedule)17 Total expenses (add lines 16 and 44, column (A)) .

18 Excess or (deficit) for the year (subtract line 17 from line 12)

19 Net assets or fund balances at beginning of year (from line 73, column (A)) .

20 Other changes in net assets or fund balances (attach explanation) . . . .21 Net assets or fund balances at end of year (combine lines 18, 19, and 20) . .

For Paperwork Reduction Act Notice, see page 1 of the separate instructions. Cat. No. 11282Y 1 9 9 0 (1996)

308 Appendix: IRS Forms 990 and 990-T

Page 2

Statement ofFunctional Expenses

All organizations must complete column (A). Columns (B), (C), and (D) are required for section 501(c)(3) and (4) organizations

and section 4947(a)(1) nonexempt charitable trusts but optional for others. (See Specific Instructions on page 13.)

Do not include amounts reported on line6b, 8b, 9b, 10b, or 16 of Part I.

(B) Programservices

(C) Managementand general

(D) Fundraising

Grants and allocations (attach schedule)

(cash $ noncash $

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

Specific assistance to individuals (attach schedule)

Benefits paid to or for members (attach schedule).

Compensat ion of officers, directors, etc. . .

Other salaries and wages

Pension plan contr ibut ions

Other employee benefits

Payroll taxes

Professional fundraising fees

Account ing fees

Legal fees

Supplies

Telephone

Postage and shipping

Occupancy

Equipment rental and maintenance . . . .

Printing and publ icat ions

Travel

Conferences, conventions, and meetings . .

Interest

Depreciation, depletion, etc. (attach schedule)

Other expenses (itemize): a

Total functional expenses (add lines 22 through 43) Organizationscompleting columns (B)-(D), cany these totals to lines H-15

37

44

Reporting of Joint Costs.—Did you report in column (B) (Program services) any joint costs from a combinededucational campaign and fundraising solicitation? • D YesIf "Yes," enter (i) the aggregate amount of these joint costs $ ; (ii) the amount allocated to Program services $(iii) the amount allocated to Management and general $ ; and (iv) the amount allocated to Fundraising $

Statement of Program Service Accomplishments (See Specific Instructions on page 16.)

D No

What is the organization's primary exempt purpose? •All organizations must describe their exempt purpose achievements. State the number of clients served,publications issued, etc. Discuss achievements that are not measurable. (Section 501{c)(3) and (4) organizationsand 4947(a)(1) nonexempt charitable trusts must also enter the amount of grants and allocations to others.)

Program ServiceExpenses

(Required for 501 (c) (3) and(4) orgs., and 4947(a)(1)

trusts; but optional forothers.)

(Grants and allocations $



(Grants and allocations $e Other program services (attach schedule) (Grants and allocations $f Total of Program Service Expenses (should equal line 44, column (B), Program services).


Page 3

Balance Sheets (See Specific Instructions on page 16.)

Note: Where required, attached schedules and amounts within the descriptioncolumn should be for end-of-year amounts only.

(A)Beginning of year

(B)End of year

45 Cash—non-interest-bearing46 Savings and temporary cash investments .

45

47b

51a

51b

47a Accounts receivableb Less; allowance for doubtful accounts .

48a Pledges receivableb Less: allowance for doubtful accounts .

49 Grants receivable50 Receivables from officers, directors, trustees, and key employees

(attach schedule)51a Other notes and loans receivable (attach

schedule)b Less: allowance for doubtful accounts . .

52 Inventories for sale or use53 Prepaid expenses and deferred charges .54 Investments—securities (attach schedule)55a Investments—land, buildings, and

equipment: basisb Less: accumulated depreciation (attach

schedule)56 Investments—other (attach schedule) . .57a Land, buildings, and equipment: basis . .

b Less: accumulated depreciation (attachschedule)

58 Other assets (describe •

59 Total assets (add lines 45 through 58) (must equal line 74) .

49

55a

55b

57a

57b

60 Accounts payable and accrued expenses .61 Grants payable62 Deferred revenue63 Loans from officers, directors, trustees, and key employees (attach

schedule)64a Tax-exempt bond liabilities (attach schedule)

b Mortgages and other notes payable (attach schedule)65 Other liabilities (describe • )

54

66 Total liabilities (add lines 60 through 65)

65

Organizations that follow SFAS 117, check here • • and complete lines67 through 69 and lines 73 and 74.

67 Unrestricted68 Temporarily restricted69 Permanently restrictedOrganizations that do not follow SFAS 117, check here • D and

complete lines 70 through 74.70 Capital stock, trust principal, or current funds71 Paid-in or capital surplus, or land, building, and equipment fund . .72 Retained earnings, endowment, accumulated income, or other funds73 Total net assets or fund balances (add lines 67 through 69 OR lines

70 through 72; column (A) must equal line 19 and column (B) mustequal line 21)

74 Total liabilities and net assets / fund balances (add lines 66 and 73)

68

73


Page 4

Reconciliation of Revenue per AuditedFinancial Statements with Revenue perReturn (See Specific Instructions, page 18)

Reconciliation of Expenses per AuditedFinancial Statements with Expenses perReturn

a Total revenue, gains, and other supportper audited financial statements . . •

b Amounts included on line a but not online 12, Form 990:

(1) Net unrealized gainson investments . . $

(2) Donated servicesand use of facilities $

(3) Recoveries of prioryear grants . . . $

(4) Other (specify):

(1)

Total expenses and losses peraudited financial statements . . •Amounts included on line a but noton line 17, Form 990:Donated servicesand use of facilities *

(2) Prior year adjustmentsreported on line 20,Form 990 . . . . *

(3) Losses reported online 20, Form 990 . 1


Add amounts on lines (1) through (4) •

c Line a minus line b •d Amounts included on line 12,

Form 990 but not on line a:

(1) Investment expensesnot included on line6b, Form 990 . . . I


Add amounts on lines (1) through (4)^c Line a minus line b •d Amounts included on line 17,

Form 990 but not on line a:

(1) Investment expensesnot included on line6b, Form 990. . . I


Add amounts on lines (1) and (2) •Total revenue per line 12, Form 990(line c plus line d) •

Add amounts on lines (1) and (2) •Total expenses per line 17, Form 990(line c plus line d) •

List of Officers, Directors, Trustees, and Key Employees (List each one even if not compensated; see SpecificInstructions on page 18.)

(B) Title and average hours perweek devoted to position

(C) Compensation(If not paid, enter

-0-.)

(D) Contributions toemployee benefit plans *deferred compensation

(E) Expenseaccount and other

allowances

75 Did any officer, director, trustee, or key employee receive aggregate compensation of more than $100,000 from your

organization and all related organizations, of which more than $10,000 was provided by the related organizations? • CD Yes CH No

If "Yes " attach schedule—see Specific Instructions on page 18.


Form 990 (1996) Page 5

Other Information (See Specific Instructions on page 19.)

78a

Did the organization engage in any activity not previously reported to the IRS? If "Yes," attach a detailed description of each activity .Were any changes made in the organizing or governing documents but not reported to the IRS? . . .If "Yes," attach a conformed copy of the changes.Did the organization have unrelated business gross income of $1,000 or more during the year covered by this return?.

b If "Yes," has it filed a tax return on Form 990-T for this year?79 Was there a liquidation, dissolution, termination, or substantial contraction during the year? If "Yes," attach a statement80a Is the organization related (other than by association with a statewide or nationwide organization) through common

membership, governing bodies, trustees, officers, etc., to any other exempt or nonexempt organization? . . .b If "Yes," enter the name of the organization • _

and check whether it is • exempt OR • nonexempt.81a Enter the amount of political expenditures, direct or indirect, as described in the

instructions for line 81 181a Ib Did the organization file Form 1120-POL for this year?

82a Did the organization receive donated services or the use of materials, equipment, or facilities at no chargeor at substantially less than fair rental value?

b If Yes," you may indicate the value of these items here. Do not include this amountas revenue in Part I or as an expense in Part II. (See instructions for reporting inPart III.) l»2b|

85d

83a Did the organization comply with the public inspection requirements for returns and exemption applications?b Did the organization comply with the disclosure requirements relating to quid pro quo contributions? . .

84a Did the organization solicit any contributions or gifts that were not tax deductible?b If "Yes," did the organization include with every solicitation an express statement that such contributions

or gifts were not tax deductible?85 50i(c)(4). (5), or (6) organizations.—a Were substantially all dues nondeductible by members?

b Did the organization make only in-house lobbying expenditures of $2,000 or less?If Yes' was answered to either 85a or 85b, do not complete 85c through 85h below unless the organizationreceived a waiver for proxy tax owed for the prior year.

c Dues, assessments, and similar amounts from members 185cd Section 162(e) lobbying and political expenditurese Aggregate nondeductible amount of section 6033(e)(1)(A) dues notices .f Taxable amount of lobbying and political expenditures (line 85d less 85e) . . 135ffg Does the organization elect to pay the section 6033(e) tax on the amount in 85f?h If section 6033(e)(1)(A) dues notices were sent, does the organization agree to add the amount in 85f to its reasonable

estimate of dues allocable to nondeductible lobbying and political expenditures for the following tax year?. . .86 501(c)(7) organizations.—Enter, a Initiation fees and capital contributions included on

line 12 L?6a_b Gross receipts, included on line 12, for public use of club facilities 86b

87 501(c)(12) organizations.—Enter: a Gross income from members or shareholders 87ab Gross income from other sources. (Do not net amounts due or paid to other

sources against amounts due or received from them.) 187b88

89a

At any time during the year, did the organization own a 50% or greater interest in a taxable corporation orpartnership? If "Yes," complete Part IX501(c)(3) organizations.—Enter: Amount of tax paid during the year under:section 4911 • _ _ _ _ _ ; section 4912 • ; section 4955 •

b 501(c)(3) and 501(c)(4) organizations.—Did the organization engage in any section 4958 excess benefittransaction during the year? If "Yes," attach a statement explaining each transaction

c Enter: Amount of tax paid by the organization managers or disqualified persons during the year undersection 4958 • _

d Enter: Amount of tax in 89c, above, reimbursed by the organization • _90 List the states with which a copy of this return is filed •91 The books are in care of • Telephone no. • ( L

Located at • ZIP + 4 •92 Section 4947(a)(1) nonexempt charitable trusts filing Form 990 in lieu of Form 1041—Check here . . .

and enter the amount of tax-exempt interest received or accrued during the tax year . . • 1 92 I

84b

85b I

• D


Page 6

Analysis of Income-Producing Activities (See Specific Instructions on page 22)

Enter gross amounts unless otherwiseindicated.

93 Program service revenue:

Unrelated business income

(A)Business code

(B)Amount

Excluded by section 512. 513. or 514

(C)Exclusion code

(D)Amount

(E)Related or

sxempt functionincome

f

9

94

95

96

97

a

b

98

99

100

101

102

103

Fees and contracts from government agencies

Membership dues and assessments . . .

Interest on savings and temporary cash investments

Dividends and interest from securities . . .

Net rental income or (loss) from real estate:

debt- f inanced property

not debt- f inanced property

Net rental income or (loss) from personal property

Other investment income

Gain or (loss) from sales of assets other than inventory

Net income or (loss) from special events

Gross profit or (loss) from sales of inventory .

Other revenue: a

104 Subtotal (add columns (B), (D), and (E)) . .

105 Total (add line 104, co lumns (B), (D), and (E))Note: (Line 105 plus line 1d, Part I, should equal the amount on line 12, Part I.)

Relationship of Activities to the Accomplishment of Exempt Purposes (See Specific Instructions on page 23.)

Explain how each activity for which income is reported in column (E) of Part VII contributed importantly to the accomplishmentof the organization's exempt purposes (other than by providing funds for such purposes).

Information Regarding Taxable Subsidiaries (Complete this Part if the "Yes" box on line 88 is checked.)Name, address, and employer identification

number of corporation or partnershipPercentage of

ownership interestNature of

business activitiesTotal

incomeEnd-of-year

assets

PleaseSignHere

Under penalties of perjury, I declare that I have examined this return, including accompanying schedules and statements, and to the best of my knowledgeand bplief, it is true, correct, and complete. Declaration of preparer (other than officer) is based on all information of which preparer has any knowledge(See General Instructions on page 8.)

Signatu Type or print name and title.

PaidPreparer'sUse Only

Firms name (or kyours if self employed) m •and address f

©

Check ifself-employed • LJ

Preparer's SSN


990-T Exempt Organization Business Income Tax Return(and proxy tax under section 6033 (e))

For calendar ye«r 1996 or other tax yev beginning 1996, and ending ,19 .• See separate instructions.

OMB No 1545-0687

•I address changedB Exempt under section

D 501 ( )( )or

D 4O8(e)

PleasePrint orType

C Book value of all assetsat end of year

Name of organization

Number, street, and room or suite no. (If a P.O. box. see page 5 of instructions.

City or town, state, and ZIP code

D Employer identification number(Employees' trust, see instructions for Block 0

on page S.j

E Unrelated business activity codes(see instructions for Block E on page 5.)

F Group exemption number (see instructions for Block F on page 5) •6 Check type of organization. • D 501 (c) Corporation D 501 (c) Trust D Section 401 (a) trust D Section 408(a) trustH Describe the organization's primary unrelated business activity. •

I During the tax year, was the corporation a subsidiary in an affiliated group or a parent-subsidiary controlled group? . . • • Yes • NoIf "Yes," enter the name and identifying number of the parent corporation. •

J The books are in care of •Unrelated Trade or Business Income

Telephone number • (

I c Balance •1a Gross receipts or salesb Less returns and allowances

2 Cost of goods sold (Schedule A, line 7)3 Gross profit (subtract line 2 from line 1c)4a Capital gain net income (attach Schedule D)b Net gain (loss) (Form 4797, Part II, line 20) (attach Form 4797)c Capital loss deduction for trusts

Income (loss) from partnerships (attach statement) . . .Rent income (Schedule C)Unrelated debt-financed income (Schedule E)Interest, annuities, royalties, and rents from controlledorganizations (Schedule F)Investment income of a section 501(c)(7), (9), or (17)organization (Schedule G)Exploited exempt activity income (Schedule I)Advertising income (Schedule J)Other income (see page 6 of the instructions—attach schedule)TOTAL (combine lines 3 through 12) .. (combine lines 3 through 12) | 13 [ | | | |

Deductions Not Taken Elsewhere (See page 7 of the instructions for limitations on deductions.)(Except for contributions, deductions must be directly connected with the unrelated business income.)

141516171819202122232425262728293031323334

Compensation of officers, directors, and trustees (Schedule K)Salaries and wagesRepairs and maintenanceBad debtsInterest (attach schedule)Taxes and licensesCharitable contributions (see page 8 of the instructions for limitation rules) .Depreciation (attach Form 4562) I 2 1

Less depreciation claimed on Schedule A and elsewhere on returnDepletionContributions to deferred compensation plansEmployee benefit programsExcess exempt expenses (Schedule I)Excess readership costs (Schedule J)Other deductions (attach schedule)TOTAL DEDUCTIONS (add lines 14 through 28)Unrelated business taxable income before net operating loss deduction (subtract line 29 from line 13).Net operating loss deductionUnrelated business taxable income before specific deduction (subtract line 31 from line 30) . .Specific deductionUnrelated business taxable income (subtract line 33 from line 32). If line 33 is greater than line 32,enter the smaller of zero or line 32

22b

34

For Paperwork Reduction Act Notice, see page 1 of separate instructions. Cat No 11291) l 9 9 0 - T (1996)


Form 990-T (1996) Page 2

Organizations Taxable as Corporations (see instructions for tax computation on page 9).Controlled group members (sections 1561 and 1563)—check here • . See instructions and:Enter your share of the $50,000, $25,000. and $9,925,000 taxable income brackets (in that order):(1) 1$ L _ J (2) \* L _ J (3) 1$ I IEnter organization's share of: (1) additional 5% tax (not more than $11,750) I* I(2) additional 3% tax (not more than $100,000) I*Income tax on the amount on line 34

_L

36 Trusts Taxable at Trust Rates (see instructions for tax computation on page 10) Income tax onthe amount on line 34 from: D Tax rate schedule or • Schedule D (Form 1041) . . . . •Proxy tax (see page 10 of the instructions). . •Total (add line 37 to line 35c or 36, whichever applies)

Tax and Payments39a Foreign tax credit (corporations attach Form 1118; trusts attach Form 1116) .

b Other credits, (see page 10 of the instructions)c General business credit—Check if from:

• Form 3800 or • Form (specify) •d Credit for prior year minimum tax (attach Form 8801 or 8827) . . .e Total (add lines 39a through 39d)

40 Subtract line 39e from line 3841 Recapture taxes. Check if from: • Form 4255 • Form 8611 . .42 Alternative minimum tax43 Total tax (add lines 40, 41, and 42)44 Payments: a 1995 overpayment credited to 1996

b 1996 estimated tax paymentsc Tax deposited with Form 7004 or Form 2758d Foreign organizations—Tax paid or withheld at source (see instructions)e Backup withholding (see instructions)f Other credits and payments (see instructions)

39a39b

44b

44f

4546474849

Total payments (add lines 44a through 44f).Estimated tax penalty (see page 3 of the instructions). Check • • if Form 2220 is attached .Tax due—If line 45 is less than the total of lines 43 and 46, enter amount owed •Overpayment—If line 45 is larger than the total of lines 43 and 46, enter amount overpaid . . . •Enter the amount of line 48 you want: Credited to 1997 estimated tax • f Refunded •

35c

38

40

45

47

49Statements Regarding Certain Activities and Other information (See instructions on page 11.)

1 At any time during the 1996 calendar year, did the organization have an interest in or a signature or other authorityover a financial account in a foreign country (such as a bank account, securities account, or other financial account)If "Yes," the organization may have to file Form TD F 90-22.1. If "Yes," enter the name of the foreign countryhere •

2 During the tax year, did the organization receive a distribution from, or was it the grantor of, or transferor to, aforeign trust?If "Yes," see page 12 of the instructions for other forms the organization may have to file.

3 Enter the amount of tax-exempt interest received or accrued during the tax year • $SCHEDULE A—COST OF GOODS SOLD (See instructions on page 12.)Method of inventory valuation (specify) •1 Inventory at beginning of year2 Purchases3 Cost of labor4a Additional section 263A costs

(attach schedule)b Other costs (attach schedule)

5 TOTAL—Add lines 1 through 4b

6 Inventory at end of year. . . .

7 Cost of goods sold. Subtract line 6from line 5. (Enter here and online 2, Part I.)Do the rules of section 263A (with respect toproperty produced or acquired for resale) applyto the organization?

PleaseSignHere

Under penalties of perjury, I declare that I have examined this returnbelief, it is true, correct, and complete Declaration of preparer (other

including accompanying schedules and statements, and to the best of my knowledge athan taxpayer) is based on all information of which preparer has any knowledge.

Signature of officer

PaidPreparer'sUse Only

Preparer's Lsignature W

Firm's name (or yours,if self-employed)and address

Check ifselfemployed • I I

Preparer s social security number


Form 990-T (1996)

SCHEDULE C-RENT INCOME (FROM REAL PROPERTY AND PERSONAL PROPERTY(See instructions on page 12.)

1 Description of property

(D(2)

(3)

Page 3

LEASED WITH REAL PROPERTY)

2 Rent received or accrued

(a) From personal property (if the percentage of rentfor personal property is more than 10% but not

more than 50%)

(D(2)

(3)

(4)

Total

(b) From real and personal property (if thepercentage of rent for personal property exceeds50% or if the rent is based on profit or income)

Total

Total Income (Add totals of columns 2(a) and 2(b). Enterhere and on line 6, column (A), Part I, page 1.) . •

3 Deductions directly connected with the income incolumns 2(a) and 2(b) (attach schedule)

Total deductions. Enterhere and on line 6, column(B), Part 1, page 1 . . •

SCHEDULE E—UNRELATED DEBT-FINANCED INCOME (See instructions on page 13.)

1 Description of debt-financed property

(1)

(2)

(3)

(4)

4 Amount of averageacquisition debt on or

allocable to debt-financedproperty (attach schedule)

(1)

(2)

(3)

(4)

Totals

5 Average adjusted basis ofor allocable to

debt-financed property(attach schedule)

2 Gross income from orallocable to debt-financed

property

6Column 4divided bycolumn 5

%

%

%

%

•Total dividends-received deductions included in column 8

3 Deductions directly connected with or allocable todebt-financed property

(a) Straight line depreciation(attach schedule)

7 Gross income reportable(column 2 x column 6)

Enter here and on line 7,column (A), Part 1. page 1.

•

(b) Other deductions(attach schedule)

8 Allocable deductions(column 6 x total of columns

3(a) and 3(b))

Enter here and on line 7,column (B), Part 1, page 1.

SCHEDULE F—INTEREST, ANNUITIES, ROYALTIES, AND RENTS FROM CONTROLLED ORGANIZATIONS(See instructions on page 14.)

1 Name and address of controlled organization(s)

(1)

(2)

(3)

(4)

5 Nonexempt controlled organizatic

(a) Excess taxable income

(D(2)

(3)

(4)

whichever is large

2 Gross income

organization(s)

ns

a),r

Totals

divided byColumn (b)

%

%

%

%

3 Deductions of controllingorganization directly

connected with column 2income (attach schedule)

4 Exempt controlled organizations

(a) Unrelatedbusiness taxable

income

6 Gross income reportable

column 5(c))or

Enter here and on line 8,column (A), Part 1, page 1.

(b) Taxable income computedas though not exempt undersec 501 (a), or the amount incol. (a), whichever is larger

7 Allowable deductions

column 5(c))

(c)column (a)divided bycolumn (b)

%

%

%

%

Enter here and on line 8column (B), Part 1, page 1.


Form 990-T (1996) Page 4

SCHEDULE G—INVESTMENT INCOME OF A SECTION 501(c)(7), (9), OR (17) ORGANIZATION(See instructions on page 14.)

1 Description of income

(1)

(2)

(3)

(4)

Totals •

2 Amount of income

Enter here and on line 9,column (A), Part 1, page 1.

3 Deductionsdirectly connected(attach schedule)

4 Set-asides(attach schedule)

5 Total deductionsand set-asides (col. 3

plus col. 4)

Enter here and on line 9,column (B), Part 1, page 1.

SCHEDULE I—EXPLOITED EXEMPT ACTIVITY INCOME, OTHER THAN ADVERTISING INCOME(See instructions on page 14)

1 Description of exploited activity business incomefrom trade or

3 Expensesdirectly

connected witrproduction of

unrelatedbusiness incom

4 Net income(loss) from

unrelated tradeor business

(column 2 minuscolumn 3). If again, compute

cols. 5 through 7

5 Gross incomefrom activity thatis not unrelated

business income

6 Expensesattributable to

column 5

7 Excess exemptexpenses

(column 6 minuscolumn 5. but not

column 4).

Column totals .

Enter here and online 10, col. (A),Part I, page 1

Enter here and online 10, col. (B),Part I, page 1

Enter here and>n line 26, Part II,

page 1

SCHEDULE J— ADVERTISING INCOME (See instructions on page 15.)^ ^ ^ B Income From Periodicals Reported on a Consolidated Basis

1 Name of periodical

(1)

(2)

(3)

(4)

Column totals (carry to Part II,ine(5)) •

2 Grossadvertising

income

3 Directadvertising costs

4 Advertisinggain or (loss) (col.2 minus col. 3) Ifa gain, compute

cols. 5 through 7.

, ««,

5 Circulation 6 Readershipcosts

7 Excessreadersh p costs(column 6 minus

column 5, but notmore thancolumn 4).

'daftm&»niiiil

Income From Periodicals Reported on a Separate Basis (For each periodical listed in Part II, fill incolumns 2 through 7 on a line-by-line basis.)

(5) Totals from Part I

Column totals, Part II

Enter here and online n , col (A),Part I, page 1

Enter here and online 11, col, (B),Parti, page 1

SCHEDULE K—COMPENSATION OF OFFICERS, DIRECTORS, AND TRUSTEES (See instructions on paqe 15.)1 Name 2 Title

3 Percent oftime devoted to

business

%

%

%

%

Total—Enter here and on line 14, Part II, page 1 •

4 Compensation attributable tounrelated business

References

Abelson, Reed. 1998. "Charities Use For-Profit Units to Avoid Disclosing Finances,"New York Times, February 9, pp. A1, A12.

Alexander, Victoria D. 1996. Museums and Money: The Impact of Funding on Exhibi-tions, Scholarship, and Management. Bloomington and Indianapolis: IndianaUniversity Press.

Allen, Richard C. 1997. "The Massachusetts Experience," Health Affairs 16(2): 85-8.Altman, Lawrence K. 1997. "Drug Firm, Relenting, Allows Unflattering Study to Ap-

pear," New York Times, April 16,pp.Al,A12."The AMA's Appliance Sale." 1997. Editorial, New York Times, August 14, p. A14.American Association of Fund-Raising Counsel (AAFRC). 1995. Giving USA 1995.

New York: AAFRC Trust for Philanthropy.1997. Giving USA 1997. New York: AAFRC Trust for Philanthropy.

American Association of Museums (AAM). 1992. Data Report from the 1989 Nation-al Museum Survey. Washington, DC: AAM.

American Association of Retired Persons (AARP). 1994. Partnerships for the Future,Annual Report, Washington, DC.

1995. "Statement of the American Association of Retired Persons Before the SenateFinance Committee's Subcommittee on Social Security and Family Policy of theUnited States Senate," Washington, DC, June 20.

American Cancer Society (ACS). 1995. Annual Report, Atlanta.1996a. "An Open Letter from the American Cancer Society," September.1996b. "Strategic Publishing Plan White Paper," Atlanta, February 20.

American Hospital Association (AHA). 1995-6. Hospital Statistics. Chicago: AHA.American Lung Association (ALA). 1990-5. Annual Reports, New York.Antel, J., R. Ohsfeldt, and E. Becker. 1995. "State Regulation and Hospital Costs," Re-

view of Economics and Statistics 77(3): 416-22.Applebome, Peter. 1997. "Rising College Costs Imperil the Nation, Blunt Report Says,"

New York Times, June 18, pp. Al , A17.Arenson, Karen W. 1995a. "Large Charities Pay Well, Survey Finds," New York Times,

September 5, p. A7.1995b. "Law Would Tighten Limits on Political Advocacy by Charities," New York

Times, August 7, p. A9.1995c. "A Small College's Tax-Exempt Status Challenged," New York Times, Septem-

ber 26, p. A14.1996. "Struggling Campuses Will Try Off-Peak Pricing Experiment," New York

Times, September 28, p. Al.

317

318 References

Arrow, Kenneth. 1963. "Uncertainty and the Welfare Economics of Medical Care,"American Economic Review 53(5): 941-73.

Arthur Andersen LLR 1986-96. "American Association of Retired Persons, Consolidat-ed Financial Statement." Washington, DC: Arthur Andersen LLP.

Asinof, Lynn. 1997. "In Paying College Costs, Parents Discover They Can Negotiateabout Financial Aid," Wall Street Journal, April 16, p. Cl.

Audretsch, David B., and Paula E. Stepan. 1996. "Company-Scientist Locational Links:The Case of Biotechnology," American Economic Review 86(3): 641-52.

Badaracco, Joseph. 1991. The Knowledge Link. Boston: Harvard Business School Press.Badelt, Christoph, and Peter Weiss. 1990. "Specialization, Product Differentiation and

Ownership Structures in Personal Social Services: The Case of Nursery Schools,"Kyklos 1:61-81.

Barnouw, Erik. 1968. The Golden Web: A History of Broadcasting in the United States,1933-1953. New York: Oxford University Press.

Baumol, William J. 1967. "Macroeconomics of Unbalanced Growth: The Anatomy ofUrban Crisis," American Economic Review 57: 415-26.

Baumol, William J., and William G. Bowen. 1966. Performing Arts - The Economic Di-lemma: A Study of Problems Common to Theater, Opera, Music and Dance. NewYork: Twentieth Century Fund.

Bean, Michael J. 1983. The Evolution of National Wildlife Law, revd./expd. ed. NewYork: Praeger Publishers for the Environmental Defense Fund.

Becker, E., and Frank A. Sloan. 1985. "Hospital Ownership and Performance," Econom-ic Inquiry 23(1): 21-36.

Becker, T. J. 1995. "Opening Door to Synergy," Chicago Tribune, May 16, §5, pp. 1,7.Behrens, Steven. 1996. "PBS Predicts Budget Gains," Current, February 12, p. 1.Bell, Julie. 1996. "Hospital Firms Court Baptist: It's an Option for Non-Profits," Nash-

ville Tennessean, March 15, pp. IE, 4E.Ben-David, Joseph. 1977. Centers of Learning: Britain, France, Germany, and the Unit-

ed States. New York: McGraw-Hill.Ben-Ner, Avner. 1986. "Nonprofit Organizations: Why Do They Exist in Market Econ-

omies?" in Susan Rose-Ackerman (ed.), The Economics of Nonprofit Institutions:Studies in Structure and Policy. New York: Oxford University Press, pp. 94-113.

Ben-Ner, Avner, and T. Van Hoomissen. 1991. "Nonprofit Organizations in the MixedEconomy," Annals of Public and Cooperative Economics 62(4): 519-50.

Biddle, Wayne, and Margot Slade. 1983. "Ideas and Trends: Bennington's Mutual AidIdea," New York Times, April 10, §4, p. 9.

Birnbaum, Jeffrey H. 1982. "Charity That Delivers Surplus Food to Needy Is Split byAccusations," Wall Street Journal, October 25, pp. Al , A20.

Blattberg, Robert C , and Cynthia J. Broderick. 1991. "Marketing of Art Museums," inMartin Feldstein (ed.), The Economics of Art Museums. Chicago and London:University of Chicago Press, pp. 327-46.

Blumenthal, David. 1992. "Academic-Industry Relationships in the Life Sciences: Ex-tent, Consequences, and Management," Journal of the American Medical Asso-ciation 268(23): 3344-9.

Blumenthal, David, Nancyanne Causino, Eric Campbell, and Karen Seashore Louis.1996. "Relationships between Academic Institutions and Industry in the Life Sci-ences -An Industry Survey," New England Journal of Medicine 334(6): 368-73.

Boot, Max. 1995. "Competing with Big Bird," Wall Street Journal, 23 October, p. A27.Boston Consulting Group. 1991. Strategies for Public Television in a Multi-Channel En-

vironment. Washington, DC: CPB.

References 319

Britton, N. R. 1993. "Permanent Disaster Volunteers: Where Do They Fit?" Nonprofitand Voluntary Sector Quarterly 20(4): 395^15.

Brody, Evelyn. 1996. "Agents without Principals: The Economic Convergence of theNonprofit and For-Profit Organizational Forms," New York Law School Law Re-view 40(3): 457-536.

Brooks, Harvey. 1993. "Research Universities and the Social Contract for Science," inL. M. Branscomb (ed.), Empowering Technology. Cambridge, MA: MIT Press,pp. 202-34.

Burton, Thomas M. 1996. "Heart Transplant Team Is Torn by Allegations of UnwiseProcedures," Wall Street Journal, December 12, pp. Al, A6.

Butler, Patricia A. 1997. "State Policy in Nonprofit Conversions," Health Affairs 16(1)(March-April): 69-84.

Byrne, John A. 1996. "Strategic Planning," Business Week (August 26): 46-52.Camp Fire Boys and Girls. 1995. A Tradition of Lighting the Way, Annual Report, Kan-

sas City, MO.Carl ton, J. M. 1995. "Roche Brings Leading Institutions into Lawsuit over Patent

Rights," Wall Street Journal, May 25, p. B4.Carnegie Corporation of New York. 1967. Public Television: A Program for Action.

New York: Bantam Books.1979. A Public Trust: The Report of the Carnegie Commission on the Future of Pub-

lic Broadcasting. New York: Bantam Books.Caro, Mark. 1997. "Museums as Meat Markets: After-Hours Events Find Singles Look-

ing for Love amid All the Glass Cases," Chicago Tribune, June 6, p. 1.Carter, Bill. 1992. "Conservatives Call for PBS to Go Private or Go Dark," New York

Times, April 30, p. Al.Chang, Cyril F., and Howard P. Tuckman. 1990. "Why Do Nonprofit Managers Accu-

mulate Surpluses, and How Much Do They Accumulate?" Nonprofit Manage-ment and Leadership 1(2): 117-35.

Child Welfare League of America (CWLA). 1986,1988,1990,1992,1994. Annual Re-ports, Washington, DC.

Christakis, Nicholas A, and Jose J. Escarce. 1996. "Survival of Medicare Patients afterEnrollment in Hospice Programs," New England Journal of Medicine 335 (July18): 172-8.

Chubin, Daryl E. 1994. "How Large an R&D Enterprise?" in D. H. Guston and K. Ken-iston (eds.), The Fragile Contract: University Science and the Federal Govern-ment. Cambridge, MA: MIT Press, pp. 118-44.

Clark, Robert. 1980. "Does the Non-profit Form Fit the Hospital Industry?" HarvardLaw Review 93 (May): 1416-89.

Claxton, Gary, Judith Feder, David Shactman, and Stuart Altman. 1997. "Public Poli-cy Issues in Nonprofit Conversions: An Overview," Health Affairs 16(2): 9-28.

Clotfelter, Charles. 1990. "The Impact of Tax Reform on Charitable Giving: A 1989 Per-spective," in Joel Slemrod (ed.),Do Taxes Matter? The Impact of the Tax ReformAct of 1986. Cambridge, MA: MIT Press, pp. 203-35.

Cobb, Natalie. 1996. Looking Ahead: Private Sector Giving to the Arts and Humanities.Washington, DC: President's Committee on the Arts and Humanities.

Cohen, Linda R., and Roger G. Noll. 1994. "Privatizing Public Research," ScientificAmerican 271(3) (September): 72-7.

Cohen, Wesley, Richard Florida, and Richard Goe. 1994. University-Industry ResearchCenters in the U.S. Pittsburgh: Carnegie-Mellon University Press.

320 References

Cole, Jonathan R. 1993. "Balancing Acts: Dilemmas of Choice Facing Research Univer-sities," Daedalus 122(4): 1-36.

Collins, Glenn. 1997. "AMA Chief Quits, Citing Sunbeam Deal," New York Times, De-cember 5, p. A17.

Conrad, Douglas A. 1984. "Returns on Equity to Not-for-Profit Hospitals: Theory andImplementation," Health Services Research 19(1): 41-63.

1986. "Returns on Equity for Not-for-Profit Hospitals: A Commentary and Elabora-tion," Health Services Research 21(1): 17-20.

Conway, William G. 1986. "The Practical Difficulties and Financial Implications of En-dangered Species Breeding Programs," International Zoo Yearbook, 210-19.

Cordes, Joseph J., and Steven M. Sheffrin. 1981. "Taxation and the Sectoral Allocationof Capital in the U.S.," National Tax Journal 36: 419-33.

Corporation for Public Broadcasting (CPB). 1995. "CPB Leads Public Radio and Tele-vision to New Ventures," November 30 press release. Washington, DC: CPB.

1996. Public Broadcasting Revenue, Fiscal Year 1995. Washington, DC: CPB, Sep-tember.

Coye, Molly Joel. 1997. "The Sale of Good Samaritan: A View from the Trenches,"Health Affairs 16(2): 102-7.

Danzon, Patricia M. 1982. "Hospital 'Profits': The Effects of Reimbursement Policies,"Journal of Health Economics 1(1): 29-52.

Dasgupta, Partha, and Paul David. 1987. "Information Disclosure and the Economics ofScience and Technology," in G. R. Feiwel (ed.), Arrow and the Ascent of Mod-ern Economic Theory. New York: New York University Press, pp. 519-^-2.

1994. "Toward a New Economics of Science," Research Policy 23(5): 487-521.Davis, Bill. 1996. "Congressional Funding for Public Broadcasting: Will There Be a

Public Broadcasting Trust Fund?" Current, July 24, 1-3.DeHoog, Ruth Hoogland. 1984. Contracting Out for Human Services: Economic, Polit-

ical and Organizational Perspectives. Albany, NY: State University of New YorkPress.

Delaware Health Care Commission. 1996. Evaluation of Certificate of Need and Oth-er Health Planning Mechanisms. Report of the Cost Containment Committee,April.

DePalma, Anthony. 1996. "Canada's Public TV System Is Facing Deep Budget Cuts,"New York Times, September 23, p. C2.

d'Harnoncourt, Anne, Paul J. DiMaggio, Marilyn Perry, and James N. Wood. 1991."The Museum and the Public," in Martin Feldstein (ed.), The Economics of ArtMuseums. Chicago and London: University of Chicago Press, pp. 35-60.

DiMaggio, Paul J. 1986a. "Can Culture Survive the Marketplace?" in Paul J. DiMag-gio (ed.), Nonprofit Enterprise in the Arts: Studies in Mission and Constraint.New York and Oxford: Oxford University Press, pp. 65-92.

(ed.). 1986b. Nonprofit Enterprise in the Arts: Studies in Mission and Constraint.New York and Oxford: Oxford University Press

Dobrzynski, Judith. 1997. "Art (?) To Go: Museum Shops Broaden Wares, at a Prof-it," New York Times, December 10, pp. Al, A16.

Doolittle, Will O., ed. [1932?]. Zoological Parks and Aquariums: An Annual Assem-blage of Information and Facts by the American Association of Zoological Parksand Aquariums. Wheeling, WV: AAZPA.

Dranove, David. 1988. "Pricing by Non-Profit Institutions: The Case of Hospital Cost-Shifting," Journal of Health Economics 7(1): 47-58.

Dranove, David, and Richard Ludwick. 1997. "Competition and Pricing by Nonprofit

References 321

Hospitals: A Reassessment." Working Paper, Kellogg Graduate School of Man-agement, Northwestern University.

Drucker, Peter. 1993. "The Rise of the Knowledge Society," Wilson Quarterly 17(2)(Spring): 52-71.

Easley, David, and Maureen O'Hara. 1983. "The Economic Role of the Nonprofit Firm,"Bell Journal of Economics 14 (August): 531-8.

Eckel, Catherine, and Richard Steinberg. 1994. "Tax Policy and the Objectives of Non-profit Organizations in a Mixed-Sector Duopoly." Working Paper, Indiana Uni-versity Center on Philanthropy, Indianapolis.

Emshwiller, John R. 1995. "More Small Firms Complain about Tax-Exempt Rivals,"Wall Street Journal, August 8, p. B1.

Etzkowitz, Henry, and Andrew Webster. 1995. "Science as Intellectual Property," inS. Jasanoff, Gerald E. Markle, James C. Peterson, and Trevor Pinch (eds.), Hand-book of Science and Technology Studies. Thousand Oaks, CA: Sage, pp. 480-505.

Fairclough, Gordon. 1997. "On the Bench: A Stadium Brawl at Princeton Winds Up inCourt," Wall Street Journal, April 16, p. Bl.

Farhi, Paul. 1997. "Mapping Out a Greater Society: National Geographic Is Pushingbeyond Its Yellow Borders into New Market Territory," Washington Post, Janu-ary 27, pp. 12-14.

Federal Communications Commission (FCC). 1984. Report of the Temporary Commis-sion on Alternative Financing. Washington, DC: FCC.

Feldman, Maryann P. 1994. The Geography of Innovation. Boston: Kluwer.Feldman, Maryann P., and Richard Florida. 1994. "The Geographic Sources of Innova-

tion," Annals of the Association of American Geographers 84(2): 210-29.Feller, Irwin. 1990. "Universities as Engines of R&D-based Economic Growth: They

Think They Can," Research Policy 19: 335-48.Ferris, James M. 1993. "The Double-Edged Sword of Social Service Contracting," Non-

profit Management and Leadership 3(4): 363-76.Ferris, James M., and Elizabeth Graddy. 1991. "Production Costs, Transaction Costs,

and Local Government Contractor Choice." Economic Inquiry 29: 541-54.1995. "Hospital Industry Dynamics and the Nonprofit Role in Health Care," Inde-

pendent Sector Research Forum Working Papers, March. Washington, DC: In-dependent Sector, pp. 453-70.

Foose,Thomas J. 1993. "Riders of the Last Ark: The Role of Captive Breeding in Con-servation Strategies," in Les Kaufman and Kenneth Mallory (eds.), The Last Ex-tinction, 2d ed. Cambridge, MA: MIT Press, pp. 149-78.

Fowler, Mark S., and Brenner, Daniel L. 1982. "A Marketplace Approach to BroadcastRegulation," Texas Law Review 60: 207-57.

Fox, Daniel M., and Phillip Isenberg. 1996. "Anticipating the Magic Moment: The Pub-lic Interest in Health Plan Conversions in California," Health Affairs 15(1): 202-9.

Frank, Richard G., and David S. Salkever. 1991. "The Supply of Charity Services byNonprofit Hospitals: Motives and Market Structure," Rand Journal of Econom-ics 22(3) (Autumn): 43-5.

Frank, Robert H., and Phillip J. Cook. 1995. The Winner-Take-All Society. New York:Free Press.

Freudenheim, Milt. 1994. "Blue Cross Lets Plans Sell Stock," New York Times, June 30,p. Dl.

1995. "AARP Will License Its Name to Managed Health Care Plans," New YorkTimes, April 29, pp. Al , C22.

322 References

1996a. "Empire Blue Cross Seeking to Become For-Profit Group," New York Times,September 26, p. Al .

1996b. "Marriage of Necessity: Nonprofit Groups and Drug Makers," New YorkTimes, August 20, p. C2.

1997. "For Teaching Hospitals, Multiple Complications," New York Times, May 20,p .Dl .

Fritsch, Jane. 1996. "Choosing Pragmatism over Confrontation to Preserve Arts Grants,"New York Times, December 10, pp. Bl , B2.

Froelich, Karen A., and Terry W. Knoepfle. 1996. "Internal Revenue Service 1990 Data:Fact or Fiction," Nonprofit and Voluntary Sector Quarterly 25: 40-52.

Fuchs, Victor. 1997. "Managed Care and Merger Mania," Journal of the American Med-ical Association 277: 920-1.

Galambos, L. 1993. "Nonprofit Organizations and The Emergence of America's Cor-porate Commonwealth in The Twentieth Century," in D. C. Hammack and D. R.Young (eds.), Nonprofit Organizations in a Market Economy. San Francisco:Jossey-Bass, pp. 82-104.

Galper, Harvey, and Eric Toder. 1983. "Owning or Leasing: Bennington College andthe U.S. Tax System," National Tax Journal 36: 257-61.

Gaul, G. M., and N. A. Borowski. 1993. "In High-Level Jobs at Nonprofit, CharityReally Pays," Philadelphia Inquirer, April 22, pp. A18-20.

Geiger, Roger. 1986. To Advance Knowledge: The Growth of American Research Uni-versities, 1900-1940. New York: Oxford University Press.

General Accounting Office (GAO). 1987. Tax Policy: Competition between TaxableBusinesses and Tax-Exempt Organizations. Washington, DC: GAO, Report GAO/GGD-87-40BR.

1992. "University Research: Controlling Inappropriate Access to Federally FundedResearch Results." Washington, DC: GAO, Report GAO/RCED-92-104, May.

Gibbons, Michael, Camille Limoges, Helga Nowotny, Simon Schwartzman, Peter Scott,and Martin Trow. 1994. The New Production of Knowledge. London: Sage.

Girl Scouts of the USA (GSUSA). 1995. Annual Report, New York.Glaberson, William. 1996. "In Era of Fiscal Damage Control, Cities Fight Idea of 'Tax

Exempt,'" New York Times, February 21, pp. Al , C17.Goddeeris, John H. 1988. "Compensating Differentials and Self-Selection: An Appli-

cation to Lawyers," Journal of Political Economy 96: 411-28.Golden, Tim. 1996. "Universities Find Donors Sometimes Impose a Price," New York

Times, December 9, pp. Al , A10.Gosselin, Peter G., and Aaron Zitner. 1997. "Pay Package Boomed at Kennedy Non-

profit," Boston Globe, June 15, pp. A1, A12.Graham, Judith, and Jon Van. 1997. "AMA Chief Quits Over Product Deal," Chicago

Tribune, December 5, pp. 1, 22.Gray, Bradford. 1986. For-Profit Enterprise in Health Care. Washington, DC: National

Academy Press.1991. The Profit Motive and Patient Care. Boston: Harvard University Press.1997. "Conversion of HMOs and Hospitals: What's at Stake?" Health Affairs 16(2)

(March-April): 29-47.Gr0nbjerg, Kirsten A. 1992. "Nonprofit Human Service Organizations: Funding Strate-

gies and Patterns of Adaptation," in Y. Hasenfeld (ed.), Human Services as Com-plex Organizations. London: Sage Publications, pp. 13-91.

1993. Understanding Nonprofit Funding: Managing Revenues in Social Service andCommunity Development Organizations. San Francisco: Jossey-Bass.

References 323

Grossman, M., F. Goldman, S. Nesbitt, and P. Mobilia. 1993. "Determinants of InterestRates on Tax-Exempt Boards," Journal of Health Economics 12(4): 385^410.

Gruber, Jonathan. 1994. "The Effect of Competitive Pressure on Charity: Hospital Re-sponses to Price Shopping in California," Journal of Health Economics 13(2)(July): 183-212.

Gulley, David O., and Rexford E. Santerre. 1993. "The Effect of Tax Exemption on theMarket Share of Nonprofit Hospitals," National Tax Journal 46: 477-86.

Guterman, Sonia K. 1996. "Licensing Biotechnology from Universities," BioPharm(May): 38-42.

Haeseler, John K. 1988. "How Much Is a Visitor Worth?" AAZPA Annual Proceedings,172-7.

Hagedoorn, John. 1993. "Understanding the Rationale of Strategic Technology Partner-ing," Strategic Management Journal 14: 371-85.

Hall, P. 1992. Inventing the Nonprofit Sector and Other Essays on Philanthropy, Volun-tarism, and Nonprofit Organizations. Baltimore: Johns Hopkins University Press,pp.135-206.

Hamburger, Eleanor, Jeanne Finberg, and Leticia Alcantar. 1995. "The Pot of Gold:Monitoring Health Care Conversions Can Yield Billions of Dollars for HealthCare," Clearinghouse Review (August-September): 473-504.

Hancock, LynNell. 1996. "All the President's Perks," Newsweek, January 22, p. 56.Hamel, Gary. 1991. "Competition for Competence and Inter-Partner Learning with-

in International Strategic Alliances," Strategic Management Journal 12: 83-103.

Handy, Femida. 1995. "Reputation as Collateral: An Economic Analysis of the Role ofTrustees of Nonprofits," Nonprofit and Voluntary Sector Quarterly 24(4): 293-306.

Hansmann, Henry B. 1980. "The Role of Nonprofit Enterprise," Yale Law Journal 89(5)(April): 835-901.

198 la. "Nonprofit Enterprise in the Performing Arts," Bell Journal of Economics 12:341-61.

1981b. "The Rationale for Exempting Nonprofit Organizations from the CorporateIncome Tax," Yale Law Journal 91: 54-100.

1987. "The Effect of Tax Exemption and Other Factors on the Market Share of Non-profit versus For-Profit Firms," National Tax Journal 40: 71-82.

1989. "Unfair Competition and the Unrelated Business Income Tax," Virginia LawReview 75: 605-35.

1996. The Ownership of Enterprise. Cambridge, MA: Harvard University Press.Harberger, Arnold C. 1974. Taxation and Welfare. Chicago: University of Chicago Press.Hartz, A. J., H. Krakauer, E. M. Kuhn, M. Young, S. J. Jacobsen, G. Gray, L. Muenz,

M. Katzoff, R. C. Bailey, and A. A. Rimm. 1989. "Hospital Characteristics andMortality Rates," New England Journal of Medicine 321: 1720-5.

Haveman, Robert H., and Burton A. Weisbrod. 1975. "Defining Benefits of Public Pro-grams: Some Guidance for Policy Analysts," Policy Analysis 2: 169-96.

Hays, Constance L. 1997. "Seeing Green in a Yellow Border: Quest for Profits Is Shak-ing a Quiet Realm," New York Times, August 3, §3, pp. 1, 12, 13.

Heilbrun, James, and Charles M. Gray. 1993. The Economics of Art and Culture: AnAmerican Perspective. Cambridge and New York: Cambridge University Press.

Henderson, Rebecca, Adam B. Jaffe, and Manuel Trajtenberg. 1995. "The Bayh-DoleAct and Trends in University Patenting, 1965-1988," Center for Economic Pol-icy Research, Paper no. 433, Stanford University.

324 References

Herman, Tom. 1995. "Tax Report," Wall Street Journal, February 8, p. Al.1997. "Tax Report," Wall Street Journal, June 18, p. Al.

Herzlinger, Regina E. 1996. "Can Public Trust in Nonprofits and Governments Be Re-stored?" Harvard Business Review 74(2) (March-April): 97-107.

Herzlinger, Regina E., and William S. Krasker. 1987. "Who Profits from Nonprofits?"Harvard Business Review 65(1) (January-February): 93-106.

Hicks, Diana M., and J. Sylvan Katz. 1996. "Where Is Science Going?" Science, Tech-nology, and Human Values 21(4): 379-406.

Hilgert, Cecilia, and Paul Arnsberger. 1992. "Charities and Other Tax-Exempt Organi-zations, 1988," Statistics of Income Bulletin 12(1) (Summer): 60-72.

Hirth, Richard A. 1997. "Consumer Information and Ownership in the Nursing HomeIndustry," Working Paper, School of Public Health, University of Michigan.

Hodgkinson, Virginia Ann, H. A. Gorski, S. M. Noga, E. B. Knauft. 1995. Giving andVolunteering in the United States. Washington, DC: Independent Sector.

Hodgkinson, Virginia Ann, and Murray S. Weitzman. 1996. Nonprofit Almanac 1996-1997. San Francisco: Jossey-Bass.

Hodgkinson, Virginia Ann, Murray S. Weitzman, Christopher M. Toppe, and StephenM. Noga. 1992. Nonprofit Almanac 1992-1993. San Francisco: Jossey-Bass.

Hoerger,Thomas J. 1991. "'Profit' Variability in For-Profit and Not-for-Profit Hospitals,"Journal of Health Economics 10(3) (October): 259-89.

1995. "Hospital Investment in the 1980s." Research Triangle Park, NC: Research Tri-angle Institute.

Hollis, Steven R. 1997. "Strategic and Economic Factors in the Hospital ConversionProcess," Health Affairs 16(1) (March-April): 131-43.

Holtmann, A. G. 1983. "A Theory of Nonprofit Firms," Economica 50: 439^9.Holtz-Eakin, Douglas, Whitney Newey, and Harvey S. Rosen. 1988. "Estimating Vec-

tor Autoregression with Panel Data," Econometrica 56(6): 1371-95.Honan, William H. 1997. "Scholars Attack Public School TV Program," New York

Times, January 22, p. A17.Ibrahim, Youssef M. 1996. "We Can't Do Business, the Dons Tell a Big Donor," New

York Times, November 26, p. A4.Iglehart, John K. 1984. "HMOs (For-Profit and Not-For-Profit) on the Move," New

England Journal of Medicine 310: 1203-8.1995. "Inside California's HMO Market: A Conversation with Leonard D. Schaeffer,"

Health Affairs 14(4): 131-42.Institute for Community Development and the Arts. 1997. Building America's Com-

munities II. Washington, DC: Americans for the Arts.Institute of Medicine. 1986. For-Profit Enterprise in Health Care. Washington, DC: Na-

tional Academy Press.Internal Revenue Service (IRS). 1994-5. Form 1023, "Application for Recognition of

Exemption Under Section 501(c)(3) of the Internal Revenue Code," applicationsfor individual organizations.

1997. The EO Tax Bulletin, no. 18, electronic version.Jaffe, Adam B., Manuel Trajtenberg, and Rebecca Henderson. 1993. "Geographic Local-

ization of Knowledge Spillovers as Evidenced by Patent Citations," QuarterlyJournal of Economics 79(5): 957-70.

Jaffe, Greg, and Monica Langley. 1996. "Fledgling Charities Get Billions from the Salesof Nonprofit Hospitals," Wall Street Journal, November 6, p. Al.

James, Estelle. 1978. "Product Mix and Cost Disaggregation: A Reinterpretation of theEconomics of Higher Education," Journal of Human Resources (Spring): 157—86.

References 325

1983. "How Nonprofits Grow: A Model," Journal of Policy Analysis and Manage-ment 2 (Spring): 350-65.

1986. "Comment," in Susan Rose-Ackerman (ed.), The Economics of Nonprofit In-stitutions: Studies in Structure and Policy. New York: Oxford University Press,pp.154-8.

1987. "The Nonprofit Sector in Comparative Perspective," in Walter W. Powell (ed.),The Nonprofit Sector: A Research Handbook. New Haven: Yale University Press,pp. 397^15.

Jarvik, Laurence. 1992. After Privatization: Public Television in the Cultural Market-place. Washington, DC: Heritage Foundation.

Jeavons,T. H. 1993. "The Role of Values: Management in Religious Organizations," inDennis R. Young, Robert M. Hollister, and Virginia A. Hodgkinson and Associ-ates (eds.), Governing, Leading, and Managing Nonprofit Organizations. SanFrancisco: Jossey-Bass, pp. 52-76.

Jensen, Elizabeth. 1994. "PBS May Ease Rules as Way to Increase Corporate Support."Wall Street Journal, June 1, p. A3.

Johnson, Lyndon. 1967. "Remarks on Signing the Public Broadcasting Act," Novem-ber 5. MS on file at LBJ Library, University of Texas, Austin, TX.

Kagen, Edward. 1995. Budgetary Trends in Jewish Community Centers & YM-YWHA's:1995. JCC Association, New York, July (and previous years).

Kaiser Family Foundation. 1995. "Kaiser Family Foundation Survey of Americans' Per-ceptions about For-Profit and Not-for-Profit Health Care." Menlo Park, CA:Kaiser Family Foundation.

Kane, Nancy M. 1997. "Some Guidelines for Managing Charitable Assets from Con-versions," Health Affairs 16(2): 229-37.

Kanter, R. M. 1995. World Class. New York: Simon & Schuster.Keeler, Emmett B., L. Rubinstein, K. Kahn, D. Draper, E. Harrison, M. McGinty, W.

Rogers, and R. Brook. 1992. "Hospital Characteristics and Quality of Care,"Journal of the American Medical Association 268(13) (October 7): 1709-14.

Kingma, Bruce R. 1995. "Do Profits Crowd Out Donations, or Vice Versa? The Impactof Revenue from Sales on Donations to Local Chapters of the American RedCross," Nonprofit Management and Leadership 6(1) (Fall): 21-38.

Kirchberg, Volker. 1996. "Entrance Fees as a Subjective Barrier to Visiting Museums:Results from a Representative Survey in Germany." Paper presented at the 9thInternational Conference of Cultural Economics, Boston, May.

Knapp, Martin. 1989. "Intersectoral Differences in Cost Effectiveness: Residential ChildCare in England and Wales," in Estelle James (ed.), The Nonprofit Sector in In-ternational Perspective. New York: Oxford University Press, pp. 193-216.

Kolata, Gina. 1997. "Safeguards Urged for Researchers: Aim Is to Keep Vested Inter-ests from Suppressing Discoveries," New York Times, April 17, p. A13.

Koput, Kenneth, Walter W. Powell, and Laurel Smith-Doerr. 1997. "IntergenerationalRelations and Elite Sponsorship: Mobilizing Resources in Biotechnology's De-velopment." Working Paper, Department of Scociology, University of Arizona.

Krashinsky, Michael. 1993. "Does Auspice Matter? The Case of Day Care for Childrenin Canada," Working Paper, Department of Economics, University of Toronto-Scarborough.

Krimsky, Sheldon. 1991. "Academic-Corporate Ties in Biotechnology: A QuantitativeStudy," Science, Technology, and Human Values 16: 275-87.

Krugman, Paul. 1991. "Increasing Returns and Economic Geography," Journal of Po-litical Economy 63(3): 483-99.

1994. The Age of Diminished Expectations. Cambridge, MA: MIT Press.

326 References

Kuttner, Robert. 1996a. "Columbia/HCA and the Resurgence of the For-Profit HospitalBusiness (Part I)," New England Journal of Medicine 335(5): 362-7.

1996b. "Columbia/HCA and the Resurgence of the For-Profit Hospital Business (PartII)," New England Journal of Medicine 335(6): 446-51.

LaMay, Craig L. 1994. "Thinking the Unthinkable: Is Advertising in Public TV's Fu-ture?" Chicago Tribune, June 8, p. A26.

Lambert, Wade. 1997. "Business Bulletin," Wall Street Journal, August 14, p. 1.Langley, Monica. 1996. "A Nonprofit Hospital Finds Its Executives Were Making the

Profit," Wall Street Journal, November 20, p. Al.1997. "Nonprofit Hospitals Sometimes Are That in Little but Name," Wall Street Jour-

nal, July 14,pp.Al,A6.Langley, Monica, and Anita Sharpe. 1996. "As Big Hospital Chains Take Over Non-

profits, a Backlash Is Growing," Wall Street Journal, October 18, p. Al .Lee, M. L. 1971. "A Conspicuous Production Theory of Hospital Behavior," Southern

Economic Journal 38 (July): 48-59.Lee, Yong S. 1994. "Technology Transfer and Public Policy in an Age of Global Eco-

nomic Competition," Policy Studies Journal 22(2): 260-6.Levit, K. R., Helen Lazenby, David Waldo, and Lawrence M. Davidoff. 1985. "National

Health Expenditures, 1984," Health Care Financing Review 7(1) (Fall): 1-35.Lincoln Library of Essential Information. 1966. 29th ed. Buffalo, NY: Frontier Press.Lohmann, Roger. 1980. Breaking Even: Financial Management in Human Service Or-

ganizations. Philadelphia: Temple University Press.Lutz, S. 1996. "Multiple-Partner Deals Let Players Hedge Their Bets," Modern Health

Care 26(26) (June 24), 146-50.Lyall, Sarah. 1997. "British Museum Is Trying to Mend Its Unmodern Ways,"New York

Times, January 7, pp. Bl, B7.Lynk, William J. 1995. "The Creation of Economic Efficiencies in Hospital Mergers,"

Journal of Health Economics 14(5): 507-30.McArthur, John H., and Francis D. Moore. 1997. "The Two Cultures and the Health Care

Revolution," Journal of the American Medical Association 277(12) (March 26):985-9.

McCready, Douglas. 1988. "Ramsey Pricing: A Method for Setting Fees in Social Ser-vice Organizations," American Journal of Economics and Sociology 47: 97-110.

Mclver, LaVonne. 1995a. "March of Dimes to Receive $100,000 for Health Message,"NonProfit Times, September, p. 5.

1995b. "Pitching to Donors, You Be the Jury," NonProfit Times, September, p. 33.Malaro, Marie C. 1994. Museum Governance: Mission, Ethics, Policy. Washington, DC,

and London: Smithsonian Institution PressMann, Joyce, G. Melnick, A. Bamezai, and J. Zwanziger. 1995. "Uncompensated Care:

Hospitals Response to Fiscal Pressures," Health Affairs 14(1) (Spring): 263-70.

Marshall, Eliot. 1994. "Congress Finds Little Bias in the System," Science 256(August12): 863.

Marshall, Eliot, and Andrew Lawler. 1996. "Congress: Biomedical Research Wins Big,"Science 274(October 4): 27-8.

Matkin, Gary W. 1990. Technology Transfer and the University. New York: Macmillan.1994. "Technology Transfer and Public Policy: Lessons from a Case Study," Policy

Studies Journal 22(2): 371-83.Matzke, Gerald E. 1997. "A Road Map from Nebraska," Health Affairs 16(2): 89-91.Mauser, Elizabeth. 1993. Comparative Institutional Behavior: The Case of Day Care,

Ph.D. diss., University of Wisconsin-Madison.

References 327

Menchik, Paul L., and Burton A. Weisbrod. 1987. "Volunteer Labor Supply," Journalof Public Economics 32 (March): 159-83.

Menke, William. 1996. "Letter to the Editor," Science News 149 (April 20): 243.Merges, Robert P., and Richard R. Nelson. 1990. "On the Complex Economics of Pat-

ent Scope," Columbia Law Review 90(4): 839-916.Meritt, Dennis A., Jr. 1980. "A Species Survival Program for AAZPA," AAZPA Annu-

al Proceedings, 69-75.Merton, Robert K. 1957. "Priorities in Scientific Discovery: A Chapter in the Sociolo-

gy of Science," American Sociological Review 22(6): 635-59.1968. "The Matthew Effect in Science," Science 159(3810) (January 5): 56-63.1988. "The Matthew Effect in Science, II: Cumulative Advantage and the Symbol-

ism of Intellectual Property," Isis 79(299): 606-23.Mifflin, Laurie. 1996. "Reprieve for PBS, but Hunt for Funds Continues," New York

Times, January 2, p. Bl.Mikesell, John (in press), "State Taxation of Nonprofit Organizations," in Joseph Cordes,

Robert Ebel, and Jane Gravelle (eds.), Encyclopedia of Taxation and Tax Poli-cy. Washington, DC: Urban Institute Press.

Milofsky, C , and S. Blades. 1991. "Issues of Accountability in Health Care Charities:A Case Study of Accountability Problems in Nonprofit Organizations," Nonprofitand Voluntary Sector Quarterly 20(4): 371-93.

Minnesota Taxpayers Association. 1996. 50 State Property Tax Comparison Study, St.Paul, MN:MTA, June.

Minow, Newton N. 1961. "The Vast Wasteland," address to the National Associationof Broadcasters, Washington, DC, May 9; see, e.g., Frank J. Kahn (ed.), Docu-ments of American Broadcasting, 4th ed. Englewood Cliffs, NJ: Prentice-Hall,1984,p.207.

Minow, Newton N., and LaMay, Craig L. 1996. "The Land Grant of the Airwaves," WallStreet Journal, January 24, p. A26.

Mintz, A. 1994. "That's Edutainment!" Museum News 73: 32-5."MIT Wins a New Trial in Price-Fixing Case." 1993. New York Times, September 18,

p.A6.Moran, R. Allen. 1985. "Queues: The Economic Theory of Screening and Human Ser-

vice Productivity," Southern Economic Journal 52(2): 492-8.Morrisey, Michael A. 1994. Cost-Shifting in Health Care: Separating Evidence from

Rhetoric. Washington, DC: AEI Press.Morrisey, Michael A., Gerald J. Wedig, and Mahmud Hassan. 1996. "Data from Califor-

nia Shed Light on Whether the Benefit from Nonprofit Hospitals Warrants TheirSpecial Tax Status," Health Affairs 15(4) (Winter): 132-44.

Moulin, Herve. 1988. Axioms of Cooperative Decisionmaking. Cambridge: CambridgeUniversity Press.

Mowery, David C , and Nathan Rosenberg. 1993. "The U.S. National Innovation Sys-tem," in Richard R. Nelson (ed.), National Innovation Systems: A ComparativeAnalysis. New York: Oxford University Press, pp. 29-76.

Mullen, William. 1997. "Museum, Zoo Shops Putting on Glitz, Raking in Profits," Chi-cago Tribune, June 11, pp. 1, 24.

Musgrave, Richard A. 1959. The Theory of Public Finance. New York: McGraw-Hill.National Academy of Sciences (NAS). 1992. The Government Role in Civilian Technol-

ogy: Building a New Alliance. Washington, DC: National Academy Press.National Academy of Sciences, National Academy of Engineering, Institute of Medicine

(NAS/NAE/IOM). 1992. Responsible Science: Insuring the Integrity of the Re-search Process, vol. 1. Washington, DC: National Academy Press.

328 References

National Assembly of State Arts Agencies (NASAA). 1992. The State of the State ArtsAgencies. Washington, DC: NASAA.

National Endowment for the Arts (NEA). 1996. Annual Report. Washington, DC: NEA.National Public Radio (NPR). 1995. Morning Edition, August 16, Transcript. Washing-

ton, DC: NPR.1996a. Interview with I. Michael Heyman, Secretary of the Smithsonian Institution,

December 18, 8:30 A.M., CST, as broadcast on WBEZ-FM, Chicago.1996b. "Museum Sharing Brings Whitney Art to San Jose," All Things Considered,

October 22, Transcript no. 2374, Segment no. 9. Washington, DC: NPR.1997. "College Price Competition," Morning Edition, April 22, 7:45 A.M. CDT, as

broadcast on WBEZ-FM, Chicago.National Science Foundation (NSF), National Science Board. 1996. Science and Engi-

neering Indicators 1996. Washington, D.C.: USGPO.Needleman, Jack, Deborah Chollet, and Jo Ann Lamphere. 1997. "Hospital Conversion

Trends," Health Affairs 16(2) (March-April): 187-95.Nelson, Richard R. 1990. "U.S. Technological Leadership: Where Did It Come from and

Where Did It Go?" Research Policy 19: 119-32.Netzer, Dick. 1978. The Subsidized Muse: Public Support for the Arts in the United

States-A Twentieth Century Fund Study. New York: Cambridge University Press.Newhouse, Joseph P. 1970. "Toward a Theory of Nonprofit Institutions: An Economic

Model of a Hospital," American Economic Review 60(1) (March): 64-74.Nichols, Donald, Eugene Smolensky, and T. Nicolaus Tideman. 1971. "Discrimination

by Waiting Time in Merit Goods," American Economic Review 61(3): 312-23.Noble, Joseph V. 1970. "Museum Manifesto," Museum News 49: 17-20.Nonaka, I. 1994. "The Dynamic Theory of Organizational Knowledge Creation," Or-

ganization Science 14: 319-40.Norton, Edward C , and Douglas O. Staiger. 1994. "How Hospital Ownership Affects

Access to Care for the Uninsured," Rand Journal of Economics 25(1) (Spring):171-85.

Olson, Mancur. 1965. The Logic of Collective Action. Cambridge, MA: Harvard Uni-versity Press.

Oster, Sharon. 1995. Strategic Management for Nonprofit Organizations. New York:Oxford University Press.

Owen, Bruce M., and Wildman, Steven S. 1992. Video Economics. Cambridge, MA:Harvard University Press, chaps. 1-3.

Packer, Kathryn, and Andrew Webster. 1996. "Patenting Culture in Science: Reinvent-ing the Scientific Wheel of Credibility," Science, Technology, and Human Values21(4): 427-53.

Parker, Harry S., Thomas Krens, William H. Luers, and Neil Rudenstine. 1991. "Mu-seum Finances," in Martin Feldstein (ed.), The Economics of Art Museums.Chicago and London: University of Chicago Press, pp. 61-90.

Passell, Peter. 1997. "The New Economics of Higher Education," New York Times, April22,pp.Cl,C6.

Pattison, Robert V., and Hallie M. Katz. 1983. "Investor-Owned and Not-for-Profit Hos-pitals," New England Journal of Medicine 309(6) (August 11): 347-53.

Pauly, Mark V. 1987. "Nonprofit Firms in Medical Markets," American Economic Re-view 77(2) (October): 257-62.

Pauly, Mark V., and Michael Redisch. 1973. "The Not-for-Profit Hospital as a Physi-cians' Cooperative," American Economic Review 63(1) (March): 87-99.

Pearce, J. A., and R. B. Robinson, Jr. 1991. Strategic Management: Formulation, Im-plementation, and Control. Homewood, IL: Richard D. Irwin.

References 329

Pennisi, E. 1993. "Concern Grows over Expansion of Earmarking," Science News 144(8)(August 21): 119.

Peterson, Ivars. 1983. "Academic Questions: Campus and Company Partnerships," Sci-ence News 123 (January 29): 76-7.

Phelps, Charles E. 1986. "Cross-Subsidies and Charge-Shifting in American Hospitals,"in Frank Sloan, James Blumstein, and James Perrin (eds.), Uncompensated Hospi-tal Care: Rights and Responsibilities. Baltimore: Johns Hopkins University Press,pp. 108-25.

Pope, Gregory C. 1989. "Hospital Nonprice Competition and Medicare ReimbursementPolicy," Journal of Health Economics 8(2): 147-72.

Porter, M. E. 1979. "How Competitive Forces Shape Strategy," Harvard Business Re-view 57 (March-April): 137—45.

1980. Competitive Strategy: Techniques for Analyzing Industries and Competitors.New York: Free Press.

Powell, Walter W. 1990. "Neither Market nor Hierarchy: Network Forms of Organi-zation," in B. M. Staw and L. L. Cummings (eds.), Research in OrganizationalBehavior, vol. 12. Greenwich, CT: JAI Press, pp. 295-336.

1996. "Interorganizational Collaboration in the Biotechnology Industry," Journal ofInstitutional and Theoretical Economics 152(1): 197-216.

Powell, Walter W., Kenneth Koput, and Laurel Smith-Doerr. 1996. "InterorganizationalCollaboration and the Locus of Innovation: Networks of Learning in Biotech-nology," Administrative Science Quarterly 41(1): 116-45.

Powell, Walter W., and Laurel Smith-Doerr. 1994. "Networks and Economic Life," inR. Swedberg and N. Smelser (eds.), Handbook of Economic Sociology. Prince-ton, NJ: Princeton University Press, pp. 368-402.

Preston, Anne E. 1989. "The Nonprofit Worker in a For-Profit World," Journal of La-bor Economics 7: 438-63.

Prospective Payment Assessment Commission (ProPAC). 1995. Medicare and theAmerican Health Care System: Report to Congress, Commerce Clearinghouse,Medicare and Medicaid Guide no. 858, pt. 2, June 15. Washington, DC:USGPO.

1996. Medicare and the American Health Care System: Report to Congress. Wash-ington, DC: USGPO.

Puzzanghera, Jim. 1997. "How Federal Funds Fueled DNA Research," San Jose Mer-cury News, May 5, pp. Al , A12-13.

Quintanilla, Carl. 1996. "Work Week," Wall Street Journal, June 11, p. 1.Randall, Alan. 1986. "Human Preferences, Economics, and the Preservation of Species,"

in Bryan G. Norton (ed.), The Preservation of Species: The Value of BiologicalDiversity. Princeton: Princeton University Press, pp. 79-109.

Rentschler, Ruth. 1996. "Museum and Performing Art Marketing: A Climate of Change."Paper presented at the 9th International Conference on Cultural Economics, Bos-ton, May.

"Republican Effort to Eliminate Arts Agency Survives Challenge." 1997. New YorkTimes, June 18, p. A16.

Riley, Gresham, and Stephen Urice. 1996. "Forum: Art Museum Directors: A ShrinkingPool?" Museum News 75: 48-9, 64.

Riley, Margaret. 1995. "Exempt Organization Business Income Tax Returns, 1991,"Statistics of Income Bulletin 14(4) (Spring): 38-63.

1997. "Exempt Organization Business Income Tax Returns: Highlights and an Anal-ysis of Exempt and Nonexempt Finances, 1993," Statistics of Income Bulletin16(4) (Spring): 75-98.

330 References

Ritenour, Shawn. 1997. "How Subsidies Kill Symphonies," Wall Street Journal, April8,p.A22.

Ritter, John. 1996. "School Puts Its Principles First," USA Today, December 9, p. 3A.Robertson, Jim. 1992. "The Public Buys In." Current, November 2, p. 2.Robinson, James C. 1995. "Health Care Purchasing and Market Changes in California,"

Health Affairs 14(4) (December): 117-30.1996. "Decline in Hospital Utilization and Cost Inflation under Managed Care in Cali-

fornia," Journal of the American Medical Association 276(13): 1060^.Robinson, James C , and Harold S. Luft. 1987. "Competition and the Cost of Hospital

Care, 1972 to 1982," Journal of the American Medical Association 257(23)(June 19): 3241-5.

Roessner, J. David, and Anne Wise. 1994. "Public Policy and Emerging Sources of Tech-nology and Technical Information Available to Industry," Policy Studies Journal22(2): 349-58.

Romeo, Anthony A., J. Wagner, and R. Lee. 1984. "Prospective Reimbursement and theDiffusion of New Technologies in Hospitals." Journal of Health Economics 3(1):1-24.

Romer, Paul. 1986. "Increasing Returns and Long-Run Growth," Journal of PoliticalEconomy 94(5): 1002-37.

Roomkin, Myron, and Burton A. Weisbrod. 1998. "Managerial Compensation in For-Profit and Non-Profit Hospitals: Is There a Difference?" Working Paper, Depart-ment of Economics, Northwestern University.

Rose-Ackerman, Susan. 1982. "Unfair Competition and Corporate Income Taxation,"Stanford Law Review 34: 1017-36.

1990. "Competition between Non-Profits and For-Profits: Entry and Growth," Vol-untas 1(1): 13-25.

(ed.). 1986. The Economics of Nonprofit Institutions: Studies in Structure and Policy.New York: Oxford University Press.

Rosegrant, Susan, and David Lampe. 1992. Route 128: Lessons from Boston s High-TechCommunity. New York: Basic.

Rosenberg, Nathan, and Richard R. Nelson. 1994. "American Universities and Techni-cal Advance in Industry," Research Policy 23: 323^8.

Rosenbloom, Richard S., and William J. Spencer. 1996. "The Transformation of Indus-trial Research," Issues in Science and Technology 12(3): 68-74.

Rosenthal, Elisabeth. 1996. "Empire Is Emerging from a Cloud," New York Times, Sep-tember 26, p. D20.

Rowland, Willard D. 1993. "Public Service Broadcasting in the United States," in Rob-ert K. Avery (ed.), Public Service Broadcasting in a Multichannel Environment.White Plains, NY: Longman Publishing, pp. 173-92.

Salamon, Lester M. 1992. America's Nonprofit Sector. New York: Foundation Center.1995. Partners in Public Service. Baltimore: Johns Hopkins University Press.

Salamon, Lester M., and Helmut K. Anheier. 1994. The Emerging Sector. Baltimore:Johns Hopkins Institute for Policy Studies.

Salamon, Lester M., Helmut K. Anheier, Wojciech Sokolowski, and Associates. 1996.The Emerging Sector: A Statistical Supplement. Baltimore: Johns Hopkins In-stitute for Policy Studies.

Saxenian, Annalee. 1994. Regional Advantage. Cambridge, MA: Harvard UniversityPress.

Schaeffer, Leonard D. 1996. "Health Plan Conversions: The View from Blue Cross ofCalifornia," Health Affairs 15(4): 183-7.

References 331

Schatz, Robin D. 1996. "Public Radio Being Pressed to Turn Its Success into More In-dependence." New York Times, March 25, p. Cl.

Schiff, Jerald, and Burton A. Weisbrod. 1991. "Competition between For-Proflt andNonprofit Organizations in Commercial Markets," Annals of Public and Coop-erative Economics 62(4): 619-39.

Schlesinger, Mark. 1998. "Mismeasuring the Consequences of Ownership: External In-fluences and the Comparative Performance of Public, For-Proflt and PrivateNonprofit Organizations," in Walter W. Powell and Elisabeth Clemens (eds.),Private Action and the Public Good. New Haven: Yale University Press, pp. 85-113.

Schlesinger, Mark, and Robert Dorwart. 1984. "Ownership and Mental Health Services:A Reappraisal of the Shift Toward Privately Owned Facilities," New EnglandJournal of Medicine 311: 959-65.

Schriesheim, Alan. 1990-1. "Toward a Golden Age of Technology Transfer," Issues inScience and Technology (Winter): 52-8.

Schultz, Jerome. 1996. "Interactions between University and Industry," in F. B. Rudolphand L. W. Mclntire (eds.), Biotechnology. Washington, DC: Joseph Henry Press,pp.131-46.

Seal, Ulysses S., D. G. Makey, N. Flesness, D. Bridgwater, L. Simmons, and L. Murt-feldt. 1976. "A Computerized Record System for the Management of Wild An-imals in Captivity," AAZPA Annual Proceedings, 78-93.

Sharp, Phillip A. 1994. "The Biomedical Sciences in Context," in D. H. Guston and K.Keniston (eds.), The Fragile Contract: University Science and the Federal Gov-ernment, Cambridge, MA: MIT Press, pp. 146-52.

"Should Public Broadcasting Go More Commercial?" 1996. Current, May 12, p. 12.Shortell, Stephen M., R. Gillies, D. Anderson, K. Erickson, and J. Mitchell. 1996. Re-

making Health Care in America. San Francisco: Jossey-Bass.Shortell, Stephen, R. Gillies, and K. Devers. 1995. "Reinventing the American Hospi-

tal," Milbank Quarterly 73(2): 131-60.Shortell, Stephen, and Edward F. X. Hughes. 1988. "The Effects of Regulation, Com-

petition, and Ownership on Mortality Rates among Hospital Inpatients," NewEngland Journal of Medicine 318(17) (April 28): 1100-7.

Silvers, J. B., and Robert T. Kauer. 1986. "Returns on Equity for Not-for-Profit Hospi-tals: Some Comments," Health Services Research 21(1): 21-8.

Simon, John G. 1987. "The Tax Treatment of Nonprofit Organizations: A Review of Fed-eral and State Policies," in Walter W. Powell (ed.), The Nonprofit Sector: A Re-search Handbook. New Haven: Yale University Press, pp. 67-98.

Skloot, Edward (ed.). 1988. The Nonprofit Entrepreneur. New York: Foundation Center.Slaughter, Sheila, and Gary Rhoades. 1996. "The Emergence of a Competitiveness R&D

Policy Coalition and the Commercialization of Academic Science and Technol-ogy," Science, Technology, and Human Values 21(3): 303-39.

Sloan, Frank A. 1988. "Property Rights in the Hospital Industry," in H. E. Freeh III (ed.),Health Care in America: The Political Economy of Hospitals and Health Insur-ance. San Francisco: Pacific Research Institute for Public Policy, pp. 103^41.

Sloan, Frank A., T. Hoerger, M. Morrisey, and M. Hassan. 1990. "The Demise of Hos-pital Philanthropy," Economic Inquiry 28(4) (October): 725^3 .

Sloan, Frank A., and Bruce Steinwald. 1980. Insurance, Regulation, and Hospital Costs.Lexington, MA: D. C. Heath-Lexington Books.

Sloan, Frank A., J. Valvona, M. Hassan, and M. Morrisey. 1988. "Cost of Capital to theHospital Sector," Journal of Health Economics 7(1): 25^-6.

332 References

Sloan, Frank A., J. Valvona, and J. Perrin. 1986. "Diffusion of Surgical Technologies:An Exploratory Study," Journal of Health Economics 5(1): 31-61.

Sloan, Frank A., and Robert A. Vraciu. 1983. "Investor-Owned and Not-for-Profit Hos-pitals: Addressing Some Issues," Health Affairs 2(1): 25-37.

Sloan, Frank A., K. Whetten-Goldstein, and A. Wilson. 1997. "Hospital Pharmacy De-cisions, Cost Containment, and the Use of Cost-Effectiveness Analysis," SocialScience and Medicine 45(4) (August): 525-33.

Small Business Administration ( SB A). 1983. Unfair Competition by Nonprofit Organi-zations with Small Business: An Issue for the 1980s. Washington, DC: SB A.

Smilor, Raymond W., George Kozmetsky, and David Gibson, eds. 1988. Creating theTechnopolis: Linking Technology, Commercialization and Economic Develop-ment. Cambridge, MA: Ballinger.

Smith, Steven Rathgeb, and Michael Lipsky. 1993. Nonprofits for Hire: The WelfareState in the Age of Contracting. Cambridge, MA: Harvard University Press.

Sokolov, Jacque J. 1995. "Integrated Delivery Systems: The Future Is Now," in RalphSnyderman and Vicki Y. Saito (eds.), The Academic Health Center in the 21stCentury. Durham: Duke University Medical Center, pp. 95-110.

Sontag, Deborah.1997. "A Nature Walk through Halls of Time, at a Brisk Pace," NewYork Times, February 6, p. A17.

Spiegel, Menahem. 1995. "Charity without Altruism," Economic Inquiry 33(4): 625-40.

Starkweather, D. B. 1993. "Profit Making by Nonprofit Hospitals," in D. Hammack andDennis Young (eds.), Nonprofit Organizations in a Market Economy. San Fran-cisco: Jossey-Bass, pp. 105-37.

Stead, Deborah. 1997. "Corporations, Classrooms and Commercialism," New YorkTimes, January 5, §4A, 30-3ff.

Steinberg, Richard. 1986. "Should Donors Care About Fundraising?" in Susan Rose-Ackerman (ed.) The Non-Profit Sector: Economic Theory and Public Policy. Ox-ford: Oxford University Press, pp. 347-64.

1987. "Nonprofit Organizations and the Market," in Walter W. Powell (ed.), The Non-profit Sector: A Research Handbook. New Haven: Yale University Press, pp. 118-38.

1990. "Profits and Incentive Compensation in Nonprofit Firms," Nonprofit Manage-ment and Leadership 1(2): 137-52.

1991. "Unfair Competition by Nonprofits and Tax Policy," National Tax Journal 44:351-63.

1993a. "Does Government Spending Crowd Out Donations?" in Avner Ben-Ner andBenedetto Gui (eds.), The Nonprofit Sector in the Mixed Economy. Ann Arbor:University of Michigan Press, pp. 99-125.

1993b. "How Should Antitrust Laws Apply to Nonprofit Organizations?" in D. R.Young, R. M. Hollister, Virginia A. Hodgkinson, and Associates (eds.), Gov-erning, Leading, and Managing Nonprofit Organizations. San Francisco: Jossey-Bass, pp. 279-305.

Steinberg, Richard, and Burton A. Weisbrod. 1997. "To Give or to Sell? That Is the Ques-tion: Or, . . . Price Discrimination by Nonprofit Organizations with DistributionalObjectives." Working Paper, Department of Economics, Indiana University / Pur-due University at Indianapolis, and Department of Economics, NorthwesternUniversity.

Steinwald, Bruce, and Laura A. Dummit. 1989. "A Hospital Case-Mix Change: SickerPatients or DRG Creep?" Health Affairs (Summer): 35^7 .

References 333

Stephan, Paula E. 1994. "Differences in the Post-entry Value of Biotech Firms: The Roleof Human Capital." Working Paper, Department of Economics, Georgia StateUniversity, Atlanta, GA.

1996. "The Economics of Science," Journal of Economic Literature 34 (September):1199-235.

Stigler, Stephen M. 1993. "Competition and the Research Universities," Daedalus122(4): 157-78.

Stone, Brad. 1997. "Rx: Thirty Minutes on the StairMaster Twice Weekly: HospitalsCourt the Sweaty Set with Health Clubs," Newsweek (March 17): 46.

"The Struggle Over Federal Aid, 1995-1996." 1996. Current, July 24, p. 1.Suhrke, Henry C. 1996. "More Scrutiny for Nonprofits Unrelated Business Income,"

Philanthropy Monthly 29(4) (May): 5-9.Swanson, Stevenson. 1997a. "Museum's Corporate Ties Rattle Some Bones," Chica-

go Tribune, October 12, pp. 1, 17.1997b. "The Wheeling and Dealing That Snared a Tyrannosaur," Chicago Tribune,

October 7, pp. 1,9."Symposium on Free versus Admission Zoos." 1965. Participants include Roland Linde-

mann, Robert Bean, Jr., A. LaMar Farnsworth, Don G. Davis, Philip Ogilvie, andWalter Kuenzli, AAZPA Annual Proceedings.

Tax Foundation. 1993/4. Facts and Figures on Government Finance. New York: TaxFoundation.

Temin, Peter. 1991. "An Economic History of American Art Museums," in Martin Feld-stein (ed.), The Economics of Art Museums. Chicago and London: University ofChicago Press, pp. 179-93.

Thorpe, Kenneth E., and Charles E. Phelps. 1991. "The Social Role of Not-for-ProfitOrganizations: Hospital Provision of Charity Care," Economic Inquiry 29(3)(July): 472-84.

Throsby, David. 1994. "The Production and Consumption of the Arts: A View of Cultur-al Economics," Journal of Economic Literature 32: 1-29.

"Time to Adjust Your Set." 1995. The Economist, December 23, p. 69.Titmuss, Richard. 1971. The Gift Relationship: From Human Blood to Social Policy.

London: Allyn & Unwin.Tober, James A. 1981. Who Owns the Wildlife? The Political Economy of Conservation

in Nineteenth-Century America. Contributions in Economics and Economic His-tory, no. 37. Westport, CT: Greenwood Press.

Tomsho, Robert. 1996. "Ohio Blue Deal Raises Question of Who Gets the Green," WallStreet Journal, June 1, p. B4.

Trescott, Jaqueline, 1994. "Santa, Patron of the Arts," Washington Post, December 11,p. G4.

Trias, Jos. 1996. "Conflict of Interest in Basic Biomedical Research," in F. B. Rudolphand L. V. Mclntire (eds.), Biotechnology. Washington, DC: Joseph Henry Press,pp.152-60.

Twentieth Century Fund. 1993. Quality Time? The Report of the Twentieth CenturyFund Task Force on Public Television. New York: Twentieth Century FundPress.

University of Arizona, Office of Vice President for Research. 1996. Profile '95: Summa-ry of Sponsored Project Activity. Tucson: University of Arizona.

Useem, M. 1987. "Corporate Philanthropy," in Walter W. Powell (ed.), The NonprofitSector: A Research Handbook. New Haven: Yale University Press, pp. 340-59.

334 References

U.S. Congress, House of Representatives. 1993. "Academic Earmarks - Part III," Hear-ings before the Committee on Science, Space, and Technology, September 21,22, and October 6.

U.S. Congress, Senate. 1994. Hearings before Subcommittee on Patents, Copyrights,and Trademarks, April 19.

VHA and Deloitte & Touche LLP. 1996. New Roles, New Responsibilities for Health:Responding to Imperatives for Change. Irving, TX, and Detroit: VHA and De-loitte & Touche LLP.

Von Hippel, Eric. 1988. Sources of Innovation. New York: Oxford University Press.Wade, Betsy. 1997. "A Free Museum Is Harder to Find," New York Times, December

28, §5, p. 4.Ward, Richard Somerset. 1993. "Public Television: The Ballpark's Changing," in Twen-

tieth Century Fund, Quality Time? The Report of the Twentieth Century TaskForce on Public Television. New York: Twentieth Century Fund Press.

Waters, Teresa M. 1992. Non-Price and Price Competition in Health Care Markets: AreConsumers Sensitive to Quality and Price Differences? Ph.D. diss., VanderbiltUniversity, Nashville, December.

Weber, Nathan, and Loren Renz. 1993. Arts Funding: A Report on Foundation and Cor-porate Grantmaking Trends. New York: Foundation Center.

Webster, Andrew. 1994. "University-Corporate Ties and the Construction of ResearchAgendas," Sociology 28(1): 123^2.

Wedig, Gerald J., M. Hassan, and F. Sloan. 1989. "Hospital Investment Decisions andthe Cost of Capital," Journal of Business 62(4): 517-37.

Wedig, Gerald J., F. Sloan, M. Hassan, and M. Morrisey. 1988. "Capital Structure,Ownership, and Capital Payment Policy: The Case of Hospitals," Journal of Fi-nance 43(1) (March): 21-40.

Weisbrod, Burton A. 1975. "Toward a Theory of the Voluntary Nonprofit Sector in aThree-Sector Economy," in Edmund S. Phelps (ed.), Altruism, Morality, and Eco-nomic Theory. New York: Russell Sage Foundation, pp. 171-95.

1977. The Voluntary Nonprofit Sector. Lexington, MA: D. C. Heath.1983. "Nonprofit and Proprietary Sector Behavior: Wage Differentials among Law-

yers," Journal of Labor Economics 1(3): 246-63.1988. The Nonprofit Economy. Cambridge, MA, and London: Harvard University

Press.1991. "The Health Care Quadrilemma: An Essay on Technological Change, Insur-

ance, Quality of Care, and Cost Containment," Journal of Economic Literature29(2): 523-52.

1995. "Comparative Institutional Behavior: For-Profit, Church-Related Nonprofit,and Other Nonprofit Organizations." Working Paper, Department of Economics,Northwestern University.

1996. "Nonprofit Organizations: What Is Their Role?" Working Paper, Departmentof Economics, Northwestern University.

1998. "Institutional Form and Organization Behavior," in Walter W. Powell and Elis-abeth Clemens (eds.), Private Action and the Public Good. New Haven: YaleUniversity Press, pp. 69-84.

Weisbrod, Burton A., and Nestor D. Dominguez. 1986. "Demand for Collective Goodsin Private Nonprofit Markets: Can Fundraising Expenditures Help OvercomeFree-Rider Behavior?" Journal of Public Economics 30 (June): 83-95.

Weisbrod, Burton A., and Mark Schlesinger. 1986. "Ownership and Regulation in Mar-kets with Asymmetric Information: Theory and Empirical Application to the

References 335

Nursing Home Industry," in Susan Rose-Ackerman (ed.), The Economics of Non-profit Institutions: Studies in Structure and Policy. New York: Oxford Universi-ty Press, pp. 133-51.

Williams, James D., and Ronald M. Nowak. 1993. "Vanishing Species in Our OwnBackyard: Extinct Fish and Wildlife of the United States and Canada," in LesKaufman and Kenneth Mallory (eds.), The Last Extinction, 2d ed. Cambridge,MA: MIT Press, pp. 115-18.

Wilson, Edward 0.1993. "Is Humanity Suicidal?" New York Times Magazine (May 30):25-9.

Winslow, Ron. 1996. "Perkin-Elmer Picks Duke Hospital CEO for Post, as IndustryTurns to Academics," Wall Street Journal, May 14, p. B6.

1997a. "Hope and Hype Follow Heart-Surgery Method That's Easy on Patients,"Wall Street Journal, April 22, pp. Al , 10.

1997b. "Scientists to Advocate Health-Care Tax," Wall Street Journal, January 13,p.B9.

Wong, Doris Sue. 1997. "Food Pantries See Sharp Increase in Demand," Boston Globe,November 29, pp. B1,B4.

Wu, Corinna. 1996. "Materials in the Magic Kingdom," Science News 150 (December21 & 28): 394-5.

Young, Dennis R. 1987. "Executive Leadership in Nonprofit Organizations," in WalterW. Powell (ed.), The Nonprofit Sector: A Research Handbook. New Haven: YaleUniversity Press, pp. 167-79.

Young, Dennis R., and Richard Steinberg. 1995. Economics for Nonprofit Managers.New York: Foundation Center.

Young, Gary J., K. Desai, and C. Lukas. 1997. "Does the Sale of Nonprofit HospitalsThreaten Health Care for the Poor?" Health Affairs 16(1) (January-February):137-41.

Young, H. Peyton. 1994. Equity in Theory and Practice. Princeton: Princeton Univer-sity Press.

Zolberg, Vera. 1986. "Tensions of Mission in American Art Museums," in Paul J.DiMaggio (ed.), Nonprofit Enterprise in the Arts: Studies in Mission and Con-straint. New York and Oxford: Oxford University Press, pp. 184-98.

Zolla-Pazner, Susan. 1994. "The Professor, the University, and Industry," ScientificAmerican 268(3) (September): 72-7.

Zucker, Lynne G., M. R. Darby, and M. B. Brewer. 1994. "Intellectual Capital and theBirth of U.S. Biotechnology Enterprises." National Bureau of Economic Re-search, Cambridge, MA, Working Paper no. 4653.

Zuger, Abigail. 1997. "After Bad Year for AMA, Doctors Debate Its Prognosis," NewYork Times, December 2, p. B11.

Zwanziger, Jack, and Glenn A. Melnick. 1988. "The Effects of Hospital Competitionand the Medicare PPS Program on Hospital Cost Behavior in California," Jour-nal of Health Economics 7: 301-20.

Index

Abbott, Jonathan, 266academic medical centers, see

hospitalsAdelphi University, 8admission fees, see user feesadvertisers, public broadcasting

and, 252-4, 256-8allocation mechanisms, see

distributional objectivesAmerican Association of Retired

Persons (AARP), commercialactivities of, 197, 198, 207-9,277,289

American Cancer Society (ACS),commercial activities of, 31,201-2

American Economic Association,62

American Heart Association, 30American Lung Association

(ALA), 200-1, 277American Medical Association

(AMA), 2, 62American Museum of Natural

History, 20American Zoo and Aquarium

Association (AAZA), 217-18Amgen, 178, 179ancillary activities, 11,48,61,63,

171,296aversion to, 53, 108mission and, 15-17,21,54-5,

84, 294-5see also specific nonprofit type

antitrust laws, 8, 142Apollo Group, Inc., 132aquariums, see zoos and aquariumsAramony, William, 8Archaeological Society of Central

Oregon, 49Arrow, Kenneth, 153Art Institute of Chicago, 239, 288art museums, 233-5, 237^44

admission fees, 234, 237-8ancillary outputs, 239"blockbuster" exhibitions, 238

commercialization pressures,235-7

cross-subsidization and, 211-8data analysis, 239^6government role in funding,

234,235-6,237,241-2income: activities to produce,

244-6; increases in, 240-4as labor intensive, 236management and, 235, 237mission, 234,238public policy and, 235restaurants and shops, 1,18,

234, 238-9see also specific museum

arts organizations, 15-16,27,38-9,95,97,99,118,119,125,126

asset distribution, conversion tofor-profit and, 146

AT&T, 42athletics, nonprofits and, 6aversion premium, 87, 88,90, 93,

124

bad debts, in hospital care, 161Bakker, Jim, 8Baptist Hospital (Memphis), 152,

288,292Baumol, William, 3,236Bayh-Dole Act, 171-2Baylor College of Medicine, 57Beatrice Foods, 135Behavioral Systems Southwest,

132Bennington College, 18, 136bidding, in nonprofit's price deter-

mination, 141-2biopharmaceutical research, 173biotechnology research, 173-85,

276"blockbuster economics," 238Blue Cross and Blue Shield, 130,

131, 137, 140, 145, 163;.seealso health insurance/insurers

Blumenthal, David, 183bonoficing, 52, 53

Boston University, 187-8Bowen, William, 3, 236Boyer, Herbert, 181Bristol-Myers Squibb, 57British Broadcasting Corporation

(BBC), 250British Museum, 58Brookings Institution, 34bundling, product, 69,73buyer influence, 31-2

cable TV, public broadcasting and,258-69,278,279

California, University of, 6-7, 8,15, 181, 182,190-1

Cambridge University, 57Camp Fire Boys and Girls, com-

mercial activities of, 204-5Canadian Broadcasting System,

251capital, weighted cost of, 65catalog businesses, 234CATO Institute, 34causality, tests of, 113, 124cause-related marketing, 203, 206-

7,209-10certificate of need (CON), 31, 152,

158Channel One, 56charge shifting, 159; see also cost

shiftingcharity care, 161, 162; see also

social service organizationsChild Welfare League of America

(CWLA), commercial activ-ities of, 197,198,205-7

Chiron, 179-80church-based organizations, see

religious nonprofitscoevolutionary arrangements, A2-ACohen, Stanley, 181collaboration, 5-7, 48-9, 52, 61

compensation and, 134among nonprofits, 4 2 ^price discrimination and, 116see also specific nonprofit type

336

Index 337

collective goods and services, seeoutputs

collectiveness index, 111colleges, see education nonprofitsColumbia/HCA, 37, 131, 141, 145commercialization of nonprofits,

1-5,16,26-7, 108-9,112,273-80, 304-5

aversion to, 53,92-3competition and, 36-7,44-5donations and, 56-9, 105-27,

198issues needing further study,

298-300public policy and, 44-5, 102-3,

166-8,198,235,280-4,300-4

removal of constraints and,273-9

tax issues and, 196, 292, 293,302

technology and, 276,278-9see also specific nonprofit type

commercial share, as commercial-activity measure, 91,96

compensation, 101, 134executive, 1,8

competitionbasis for, 26nonprofit-for-profit, 4-5,25-7,

37-40,155,162,275,288,291

nonprofit-nonprofit, 26-36:collaborative ventures and,42—1; commercialization and,36-7,44-5; exclusively non-profit marketplaces, 27-34

organizational form and, 40-2,44-5

public policy and, 44-5unfair, 4, 87, 264, 291

competitive bidding, in nonprofit'sprice determination, 141-2

competitive medical plans (CMPs),155-6

concessions, zoo and aquarium,229-31

conflicts of interest, 144-5,189—91,273; see also nonprofitmission, commercializationand

constraints, 4-5, 282-3, 287budgetary, 299-300on conversion to for-profit,

297-8on costs, 2 9 3 ^market-imposed, 134removal of, 273-9on revenue, 14-16,293—1tax-based, 296-7see also specific nonprofit type

consumers, influence of, 31-2,227; see also target popula-tions

Consumers Union, 30, 37Contra Costa County Earth Day,

49contract failure, 73contributions, gifts, and grants

(CGG),9, 112in basic nonprofit model, 55-9corporate sponsorships, 12commercial activities: effect on,

93, 196; crowding out/in byCGG, 60, 106-7

endogenous, 55,113program services and, 106-26,

296public policy and, 302-3as revenue source, 13-14, 272;

constraints of, 14-16see also specific nonprofit type;

revenueconversion to for-profit, 129,

132-3abuse of, 147asset distribution and, 146constraints on, 297-8effect on society, 2 8 3 ^ , 285managerial control, transfer of,

132mission and, 19, 135,297organizational form and, 133-4partial vs. complete, 139public policy and, 135,138-18,

152, 283-*, 303reasons for, 18-19,133-8, 146tax issues and, 134-5, 140valuation, social vs. market,

139^5see also specific nonprofit type

cooperation, nonprofit-for-profit,5-7

Cooperative Research Act, 172cooperative research and develop-

ment agreements (CRADAs),172

corporate sponsorships, 12, 252-4Corporation for Public Broadcast-

ing (CPB), 256-8, 261, 263;see also public broadcasting

corporations, see for-profitsCorrections Corporation of Amer-

ica, 132cost disease, 3,236cost shifting, 91-2, 98-9, 159-60costs

of capital, weighted, 65complementarities, 88, 90,92-3constraints on,

commercialization and, 293—4interdependence among, 17,

60-1,113-14see also joint costs

cream skimming, 79-80cross-subsidization, 273, 279crowding out/in, 60, 106-7, 120-

1,198,216,277,285

Daughters of the American Revo-lution (DAR), 39

day-care centers, 27, 134Diagnosis Related Groups (DRGs),

72-3,154Diamondopoulos, Peter, 8Discovery Network, 255distributional objectives and mech-

anisms, 66-74, 77, 79, 108,220, 238; see also nonprofitmission; target populations

Dole, Bob, 262donations, see contributions, gifts,

and grantsDow Chemical, 6DuBridge, Lee, 257-8Duke University, 57

"earmarking," 186-7Economic Recovery Tax Act

(ERTA) of 1981,102economies of scale, 38-9economies of scope, 107Edison Project, 132Educational Testing Service, 30, 37education nonprofits, 27,51, 132

commercial activity in, 1,4,6,8, 10; and donations, 14-15,56-7,107,276-7

tax issues and, 18,95,97,99tuition, 68, 107see also intellectual property,

universities andeligibility requirements, raising

revenue via, 69,71-3endangered species, zoos and

aquariums and, 217-19endorsements by nonprofits, 2,

202,209exogenous donations, 55, 109-10,

111

fairness, 76; see also competition,unfair

fees, see user feesfederally funded research and de-

velopment centers (FFRDCs),172

Field Museum of Natural History,2,288

financing, see revenuefixed-effect model, 112, 121-2food banks, 135Ford Foundation, 30for-profits, and nonprofits, 4, 273

affected by, 291-2, 296competition, 4-5, 25-7, 37—W),

155, 162,275,288,291cooperation/collaboration, 5-7,

238-9as subsidiaries, 1, 263see also conversion to for-profit

"for-profits in disguise," 11-12,48,65,66,273,281

338 Index

Fowler, Mark, 262funding, see revenuefund-raising, 28,55-6, 86,109-10,

197; see also specific non-profit type; contributions,gifts, and grants; revenue

genetic engineering, 180-2genomics, 178Genotech, 178Giamatti,A. Bartlett,6gifts, see contributions, gifts, and

grantsgift shops, 229-30, 231Girl Scouts of the USA, commer-

cial activities of, 197, 202-4,289

Glen Oak Zoo, 220, 223, 224, 229,230

goals, nonprofit, see nonprofitmission

governmentinteraction with nonprofits, 3,

7-8,73research support by, 172, 182,

186-8,283university-industry collabora-

tions and,171—3see also specific nonprofit type;

contributions, gifts, andgrants; public policy; taxissues

Grainger causality, 124grants, see contributions, gifts, and

grantsGreater Richmond Partnership, 37Grove City College, 14

Habitat for Humanity, 70Hammond Maps, 291Harvard University, 6, 57Health Care Cost Containment

Board, Florida, 158health-care nonprofits, 1,4, 11-12

charities, 200-2, 210competition and, 37-8,40, 155-

6conversion to for-profit, 130-2,

136-8, 152perceived quality of care and,

167taxation and, 95, 97,99see also health insurance/

insurers; hospitals; Medicaid;Medicare

health insurance/insurers, 156,208

cross-subsidization and, 274-5for-profit conversion of, 130-2,

136-8,283,297organizational form and, 40,42,

156uninsured patients, 161-2, 164see also health-care nonprofits

health-maintenance organizations(HMOs), see health insur-ance/insurers

Health Net, 130Heller, Craig, 189Heritage Foundation, 34Heyman, I. Michael, 12Hill-Burton Act, 137HoechstA.G.,6holding companies, 40-1hospitals, 27, 158, 166-8

capital funds, 163-5coevolutionary arrangements

and, 4 2 ^commercialism in, 151-3, 167—

8; care and, 154-6, 161-2,163; cross-subsidization and,274-5,284; public policyand, 166-8

conversion to for-profit, 131,132,152

investments, 165-6nonprofit status, reasons for,

organizational form and, 40-2pricing patterns in, 159-60, 162profitability of, 156-9and research, 59-60, 153, 166see also health-care nonprofits

Human Genome Project, 182human-services nonprofits, 95,97,

99hurdle rate, 86

income, see revenueincome tax, see tax issuesIndependent Sector, 117Indiana University, 10industry-university collaborations,

see intellectual property,universities and

Institute for Museum Services,236

intellectual property, universitiesand, 6,56,169-71, 174-5,182-91

business strategy, 173—4conflicts of interest, 189-91constraints on, 282-3cross-subsidization and, 275-6funding, politics of, 186-8licensing and corporate support

ofR&D, 176-80, 183in life sciences, 175-82organizational adaptations and,

188-9patenting, 176, 190-1, 282-3public policy, 171-3at Stanford and Johns Hopkins,

180-2Internal Revenue Service (IRS),

117Code §501(c)(3)/(4), 4, 30,49,

91,207

determining taxability, 61-2,105, 107-8,116,203^,205,264,288,290,292,295

as regulatory agency, 50-1, 63,123, 136, 144,264,288-9,293,304

Form 990,44,55,91,110,117,120,233,239^0,274,280,296,301-2,307-12

Form 990-T, 99, 100-1, 296-7,302,313-16

Form 1023,49

Statistics of Income (SOI) Divi-sion, 117,240

see also tax issuesintertemporal price discrimination,

67,70investments

active, 41of nonprofits, 53, 85-90, 92-

102,165-6,208passive, 40-1, 53, 88, 89

Jewish Community Centers Asso-ciation (JCCs), commercialactivities of, 198,210-15

Johns Hopkins University, 180-2Johnson, Lyndon, 256joint costs, 90, 99, 102, 296-7Joint Council on Economic

Education, 30joint ventures, 41-2, 132, 141,

264, 265, 305; .«?<? a/sospecific nonprofit type

Junior Achievement, 30

Kaiser Family Foundation, 167Kellogg's, 6Kraft, 135

laborexpense of, 10-11,101,236as nonmonetary fee, 70volunteer, 15,29,55

leased-space ventures, 42legislation, see government; public

policylicensing, 176-80, 238-9life sciences, university research,

175-82Lincoln Park Zoo, 220, 223, 224,

229,230

mail-order businesses, 234management

behavior, affecting nonprofits,235,237,272,273

competition and, 28, 33, 34conversion and, 131-2, 134,

145-6revenue sources determining,

291-2March of Dimes, 6market-imposed constraints, 134

Index 339

marketing, cause-related, 203,206-7,209-10

Massachusetts Institute of Technol-ogy (MIT), 176

McDonald's, 2medical centers, see hospitalsmedical insurance, see health

insurance/insurersMedicaid, 161Medicare, 72-3, 154-5

for-profit vs. nonprofits and,161,163

membership, categories of, 68,70;see also specific nonprofittype

mental-health services, 32, 34; seealso health-care nonprofits

mergers, 151, 152,155,279merchandising, by nonprofits, 239,

254-5,263Merck & Co., 57merit goods, 219Metropolitan Museum of Art, 234,

239Michigan, University of, 6Millennium, 180Minnow, Newton, 257mission, see nonprofit missionMonsanto Chemical Company, 6Mullis,Kary, 180multiproduct model, 48, 111, 114—

15Museum of Fine Arts, Boston, 239museums, 1, 5,27, 54,62; see also

art museums

National Competitiveness Technol-ogy Act, 172

National Endowment for the Arts(NEA), 15-16,236,237

National Geographic Society, 291,294

National Institutes of Health(NIH), 170, 276

National Rifle Association, 31National Science Foundation

(NSF), 169, 276National Taxonomy of Exempt

Entities (NTEE), 93, 117, 118New England Medical Center

Hospital, 123Nike,6nonexcludable, nonrival goods, 28nonmonetary payments, 70nonpreferred goods and services,

see outputs, preferred vs. non-preferred

nonprofit mission, 9, 16,48-55, 73ancillary activities and, 15-17,

21,42,54-5,84,238-9,254,294-5

commercialization and, 196-9,288-93,294-5,305

displacement of, 54, 57

for-profit conversion and, 135,297

and investment, 85-90and outputs, 50, 292-3public policy and, 19-20revenue and, 12-19,52vagueness of, 15, 51, 234, 253,

256-60,289-91,293see also specific nonprofit type

nonprofits, 2—4, 27behavior: differences among,

295-6; and differentialtaxation, 85-8; modeling,48-50,59-61

constraints on, 4-5, 282-3data improvement, 301-2diversity in, 112, 119,125,

3 0 3 ^"pure," 13, 110see also specific nonprofit type

or organization; commercial-ization of nonprofits; com-petition; conversion to for-profit; for-profits, andnonprofits; "for-profits in dis-guise"; government; invest-ments, of nonprofits; non-profit mission; outputs; targetpopulations

Northwestern University, 6Notre Dame, 10nursing homes, 27, 31, 134, 297-8

objectives, distributional, seedistributional objectives andmechanisms

odds ratio, 95organizational form, nonprofit

and competition, 40-2,44—5and conversion to for-profit,

133—4output mission, see nonprofit

missionoutputs (goods and services)

collective/public vs. private, 19,28,49,52,61,63,219,237,250, 272,280

disfavored vs. favored vs. neu-tral, 196-7

measuring, 301-2mission and, 50,292-3preferred vs. nonpreferred, 48-

9,52-3,62,85,86,106-16,196-7,219-20

target populations and, 61, 299see also specific nonprofit type;

quality of service or outputOxford University, 114

Patent and Trademark Amend-ments of 1980, 171,282

patenting, 176, 190-1,282-3Perkin-Elmer Corp., 57Philadelphia Museum of Art, 238

philanthropy, see contributions,gifts, and grants

preferred goods and services, seeoutputs, preferred vs. non-preferred

preferred provider organizations(PPOs), 155; see also healthinsurance/insurers

pricingdiscrimination in, 8, 67, 68,

69-71,115-16, 159fairness and, 76patterns, in hospitals, 159-60Ramsey, 75-6screening and, 78-80uniform, 67-8

Princeton University, 10prisons, conversion to for-profit,

132private firms, see for-profitsprivate goods and services, see

outputsprivatization, see conversion to for-

profitproduct bundling, 69, 73"profit," use in nonprofits, 66, 67,

295, 296; see also "for-profitsin disguise"

program service revenues (PSRs),16,91, 105, 110-11, 120,242,294,296

property-tax exemptions, 7, 88-9,94, 134, 2%2\ see also to*issues

Prospective Payment System(PPS), 154-5

PTL Ministry, 8public broadcasting, 27, 249-50

ancillary goods and services,254,264

corporate underwriting andadvertising, 249, 265-7

commercial activity, 53, 254-6conversion to for-profit, 261-2cross-subsidization and, 278-9and for-profit stations, similari-

ties, 255-6, 262mission, 252, 2 5 3 ^ , 256-8,

266output quality in, 249-50, 255,

261revenue sources, 250^, 256,

259; government funding,260-5; memberships, 249,254-5, 258; user fees and,250-1,252^

public goods and services, seeoutputs

public policyon contributions, gifts, and

grants, 302-3on commercialization, 44-5,

61-4,102-3,166-8, 198,235,280-4,300^

340 Index

public policy (cont.)competition and, 44—5on conversion to for-profit, 135,

138-48,152,283-4,303on intellectual property, 171-3nonprofit mission and, 19-20see also specific nonprofit type

quality of service or output, 26,32,38,69,73; see alsospecific nonprofit type

Ramsey pricing, 75-6rationing, 116Red Cross, 31regulation, see conversion to for-

profit; government; publicpolicy

Reimers, Neils, 181-2religious nonprofits, 27 ,73^ , 280

competition among, 35-6reputation, commercialization and,

196,197-8research

applied vs. basic, 52, 169-71,276

government support, 172, 182,186-8,283

see also hospitals, and research;intellectual property, univer-sities and

research and development (R&D),172,176-80, 183

restaurants/cafeterias, in artmuseums, 234, 239

retailing, by nonprofits, 1, 18, 234,238-9,254-5,263

revenue, 4, 6-7, 9-12,49, 63-4constraints: budgetary, 299-

300; commercialism and, 52,293-4

interdependencies, 59-60, 113,293^ , 300-1

nonprofit's mission vs., 12-19,294-5

see also contributions, gifts, andgrants; fund-raising; programservice revenues; user fees

risk, 77-8, 206Rockefeller, Sharon Percy, 266

schools, see education nonprofitsscience, nonprofit-private cooper-

ation in, see intellectualproperty, universities and

screening, as function of price,78-80

Sequana Therapeutics, 180service quality, see quality of

service or outputSharp, Philip, 186

Shedd Aquarium, 288sliding-scale fee schedules, 68-70SmithKline Beecham, 57, 202social mission, see nonprofit

missionsocial service organizations, 195—

9, 209-10memberships and user fees,

198-9,205,212-13see also specific nonprofit type

or organizationspecies preservation, in zoos and

aquariums, 217-19spectrum fee (broadcasting), 262Stanford University, 180-2Stevenson-Wydler Act, 172subsidies, see tax issues, subsidiesSuicide Anonymous Treatment, 49Sunbeam Corporation, 2sweat equity, 70

target populations, output and, 61,299

tax issuesarbitrage, 292commercialization and, 196,

292,293,302constraints imposed by, 289,

296-7conversion to for-profit and,

134-5, 140cost allocations and, 63; see also

joint costsneutral income tax, 88nonprofit vs. for-profit, 84-90organizational form and, 133—4premiums, 9 3 ^subsidies, 4, 7, 18-19,48, 88-9,

224,272,273,280-2see also specific nonprofit type;

Internal Revenue Service;Unrelated Business IncomeTax

Tax Reform Act (TRA) of 1986,102-3

technology, 158ancillary activities and, 60-1,

99,102,296commercialization and, 276,

278-9see also intellectual property,

universities andTelecommunications Act, 262television, public, see public

broadcastingTemporary Commission on Addi-

tional Financing, 265-6think tanks, 34

uncompensated hospital care, 161-2,275

unions, 11United Way, 8, 29, 210universities, see specific institu-

tion; education nonprofits;intellectual property, univer-sities and

university-industry researchcenters (UIRCs), 172

unrelated business activity andincome, 48, 63, 67, 89-90,108,116-17,246,264,281,296-7; see also specificnonprofit type

Unrelated Business Income Tax(UBIT), 53, 84-5, 89-90,91-2,95-102, 108

Urban Institute, 34U.S. Congress, see government;

public policyU.S. Health Care, 130user fees, 10,49,54,63

admission fees, 107, 195, 197,212,224-9,234,237-8,258,277

fee schedules, 68-70labor as, 70mission and, 79-81, 294-5

valuation, conversion to for-profitand, 141-5

Virginia Opera, 39visitor surveys, 225volunteer labor, 15, 29, 55, 255

wages, see compensationwaiting lists, 69, 71,78Walt Disney Company, 2, 56Washington, University of, 7Washington University, 6Wellness Foundation, 130WellPoint Health Networks, 130,

144work force, see labor

Yale University, 14YMCA/YMHA/YWCA, 27, 210youth services, 202-7

zoos and aquariums, 217-22, 227attendance, 225-6collaboration, 219-20commercialization and, 221-2conversion to for-profit, 223cross-subsidization and, 278revenue sources: admission fees,

224-9; concessions, 229-31;donations, 223-̂ 4-; member-ships, 229

species preservation in, andmission, 217-19

tax subsidies, 224

to profit or not to profit: the commercial transformation of the nonprofit sector

Documents