using activity based costing to manage more effectively

G r a n t R e p o r t

January 2000

The PricewaterhouseCoopers Endowment for

The Business of Government

Using Activity-Based Costing

to Manage More Effectively

Michael H. GranofProfessor

David E. PlattAssistant Professor

Igor VaysmanAssistant Professor

Department of AccountingCollege of Business AdministrationUniversity of Texas at Austin

About The EndowmentThrough grants for Research, Thought Leadership Forums, andthe SES Leadership Program, The PricewaterhouseCoopersEndowment for The Business of Government stimulatesresearch and facilitates discussion on new approaches toimproving the effectiveness of government at the federal, state,local, and international levels.

Founded in 1998 by PricewaterhouseCoopers, The Endowmentis one of the ways that PricewaterhouseCoopers seeks toadvance knowledge on how to improve public sector effec-tiveness. The PricewaterhouseCoopers Endowment focuses on the future of the operation and management of the publicsector.



Using Activity-Based Costing to Manage More Effectively 1

Using Activity-Based Costing

to Manage More Effectively

Michael H. GranofProfessor

David E. PlattAssistant Professor

Igor VaysmanAssistant Professor

Department of AccountingCollege of Business Administration

University of Texas at Austin

January 2000

2 Using Activity-Based Costing to Manage More Effectively


TABLE OF CONTENTS

Foreword ......................................................................................5

Executive Summary ......................................................................6

Introduction ................................................................................7The Virtues of ABC ................................................................7Unique Characteristics of Universities ....................................8

Description of the Study ............................................................12Reason for Focusing on a Department Rather Than an Entire University ....................................................12Features of the Department ..................................................12Features of the ABC Model ..................................................14Examples of Problems and Issues Encountered ....................18

An ABC Model for a University Department ..............................21Finding One: Great Disparities Exist Among Various Programs ................................................................21Finding Two: Unused Capacity Is Costly ..............................22Finding Three: ABC Accounting Provides Useful Efficiency Information ..........................................................23Finding Four: Support Services Do Not Benefit Programs Uniformly ............................................................24Finding Five: Space Costs Are Significant ............................25

Lessons Learned ........................................................................26

Appendix ....................................................................................28

About the Authors ......................................................................30

Key Contact Information ............................................................31


Foreword

January 2000

On behalf of The PricewaterhouseCoopers Endowment for The Business of Government, we are pleased to present this report by Professors Michael Granof, David Platt, and Igor Vaysman, “Using Activity-BasedCosting to Manage More Effectively.”

In this report, Professors Granof, Platt, and Vaysman demonstrate how activity-based costing (ABC) can be used by executives to manage more effectively. The applicability of the ABC model presented has far-reaching implications for all organizations. This model can clearly be applied to government and nonprofitorganizations, as well as universities. Instead of measuring traditional “inputs” of salary and administrativecosts, ABC accounting provides a methodology to measure the costs of “outputs.”

This report arrives at an opportune time. Both the public and nonprofit sectors are working hard to quantifytheir “outputs.” The federal government is now engaged in dramatically improving its ability to document“outputs.” This increased emphasis on measurement can be linked to two recent events in the federal govern-ment. First, the Government Performance and Results Act of 1993 legislated that the executive branch docu-ment its performance and the cost of its services to the American public. Second, the rise in the number offranchise funds and reimbursable programs throughout the federal government has now made it necessary forfederal executives to develop new methodologies to understand and document the “true costs” of providingservices within their own organizations and to other units within government.

We trust that this report will be useful to executives at all levels of government who are now involved inquantifying their performance and financial costs.

Paul Lawrence Ian LittmanPartner, PricewaterhouseCoopers Partner, PricewaterhouseCoopersCo-Chair, Endowment Advisory Board Co-Chair, Endowment Advisory [email protected] [email protected]




Although universities and other governmental orga-nizations can boast of elaborate accounting sys-tems typically featuring thousands of accounts,their faculty and administrators often lack the mostrudimentary cost information on the programs thatthey manage and the activities in which theyengage. The systems are almost always based on aform of fund accounting and are intended to satisfylegal and donor stipulations rather than to provideinformation for administrative decisions.

In this project we show how activity-based costing(ABC) can be applied to a single department of amajor institution of higher education and therebyprovide more management-oriented informationthan the systems currently employed. ABC has anadvantage over conventional accounting systemsmainly in that it allocates “overhead” costs to pro-grams and activities in a way that is more reflectiveof the factors that influence them. Although ABChas occasionally been adopted in governmentaland not-for-profit settings, it is usually applied torepetitive processes rather than the more costlyintellectual activities.

We chose to apply ABC in a university settingbecause of the widespread concern about the risingcosts of higher education and the unique character-istics of universities that make for special chal-lenges. Universities are characterized, for example,by an absence of well-defined products or out-comes, unusual interrelationships among out-comes, and capacity constraints that are seeminglyelastic. We opted to focus on a single department,the accounting department of a large business

school, both to make the project manageable andbecause in universities key decisions, such aswhether to add new programs or courses, are typically made at the lower portion of the organiza-tion chart.

Although our project was a case study, we believethat we learned several lessons that are applicablenot only to universities, but also to other govern-mental and not-for-profit organizations. Theseinclude:

• The primary benefit of ABC may be not that itis an improvement upon an already adequateaccounting system, but that it provides thestructure for the establishment, for the firsttime, of a true management-oriented system.

• To be successful, a system must be flexible.Rigid allocation rules cannot readily beimposed upon organizations, like universities,characterized by decentralized managementsystems.

• ABC, by assigning costs to previously-unmea-sured factors in decisions and providing a mea-sure of the full cost of programs and activities,helps identify circumstances in which goalsand objectives are out of line with spendingdecisions.

Executive Summary


University faculty and administrators, it has beensaid, know the value of everything but the cost ofnothing. Indeed, although they may add newcourses, programs, and extracurricular activitiesafter having eloquently set forth the benefits, theyoften lack even rudimentary information about thecosts. Large universities may maintain intricateaccounting systems, often with thousands ofaccounts. However, the systems are almost alwaysbased on a form of fund accounting and areintended to satisfy legal and donor stipulationsrather than to provide information for administra-tive decisions.

In our study we show how activity-based costing(ABC) can be applied to institutions of higher edu-cation and, we believe, can result in improvedinformation of benefit to academic administrators,legislators, voters, and consumers. We conducted afield study of a large college of business that is partof a major state university. We compared the infor-mation that was available with that which webelieved to be decision-relevant. On the basis of our findings we developed and tested an ABCmodel to provide the decision-relevant data. Thepurpose of our study was not to develop a com-plete, working model, but rather to demonstrate thefeasibility and benefits of applying ABC in an acad-emic environment. We also sought to identify theobstacles that colleges would face in applying ABCand to discern the limitations of our approach.

Despite the unique features of business schools, webelieve that our model can be adapted readily toother colleges and institutions of higher education.More importantly, our findings can be generalizedto other government and not-for-profit organiza-tions in which personnel routinely engage in multi-ple activities or programs. These include hospitals(especially those associated with medical schools),public schools, police departments, fire depart-ments, health departments, social service agenciesand governmental agencies.

The Virtues of ABCActivity-based costing is now an accepted elementof the accounting and control systems of industrialand service firms, and it has been employed in bothgovernmental and not-for-profit organizations. ABCis a product of the technological era. Conventionalmanagerial information systems can trace their rootsto the industrial age, when labor was the dominantfactor of production. Within these systems, over-head cost is first allocated from service departmentsto production departments and then distributed,using an “overhead charging rate,” to specific prod-ucts. This method was developed to measure manu-facturing processes in which overhead was eitherimmaterial or was mainly a function of direct labor,which, in turn, was dependent upon productionvolume. Moreover, in a conventional industrial set-ting, production departments were mainly responsi-ble for the key manufacturing activities and wereclearly distinguishable from service departments,which provided only ancillary support.

Introduction*

* The authors gratefully acknowledge The Pricewaterhouse-Coopers Endowment for The Business of Government for their support of this study.


In the service economy (as well as in modern,computer-driven manufacturing facilities), directmanufacturing labor is no longer the overriding fac-tor of production and the distinction between pro-duction and service departments has becomedecidedly blurred. Overall product and servicecosts are more influenced by research, materialshandling, procurement, equipment maintenance,quality control, and customer service requirementsthan by direct labor.

To compensate for the deficiencies of the conven-tional information systems, ABC requires firms tocollect costs in specially constructed “activity pools”rather than service departments or overhead costcenters. Each of the pools corresponds to a group of similar business processes or activities that arehomogeneous in that all costs assigned to the poolare influenced or driven by a common factor. Theactivity pools can cut across departmental bound-aries and can include overhead costs incurred byproduction as well as service departments.

After collecting the costs in the activity-based costpools, the firm distributes them directly to its vari-ous products or services by means of a “cost driver.”A cost driver is similar to an overhead charging rate,but it should represent the factor that has the great-est influence on the behavior of the overhead costswithin a particular activity pool.

Common cost drivers include production-orienteddrivers such as cycle times, setups, number of pur-chase orders, number of machine hours and num-ber of inspections. Other cost drivers address thecost of providing service resources by measuringspecification changes, ordering characteristics, andother measures of clients’ needs for attention.

While direct labor is often a cost driver, it shouldbe used only when, in fact, the causal relationshipbetween labor and the costs in the activity pool isstronger than that between the pool and any otherpotential cost driver. The total costs in each poolare distributed to the products on the basis of eachproduct’s cost driver volume. Thus, if a particularproduct requires 60 percent of the quality controlinspections (a cost driver), then it is assigned 60percent of the quality control costs (the relatedactivity-based cost pool).1

The differences between ABC and traditional cost-ing are highlighted in Figure 1.

ABC has now been adopted by governmental, acad-emic and other types of not-for-profit organizations.However, it has been applied mostly to these institu-tions’ repetitive processes, rather than to their intel-lectual activities.2 For example, two recent articlesthat describe the application of ABC to universitiesconcentrate on the activities of the support depart-ments. These articles discuss ways to allocate thecosts of libraries3 and computing support4, as well asother services. In fact, many of the activities carriedout by these types of organizations are similar tothose carried by businesses. Clerical functions, forinstance, such as procurement and payroll process-ing, are common to all types of organizations. Thus,as noted in the books and articles promoting the useof ABC in the government and not-for-profit sectors,ABC is no less applicable to those sectors than tobusinesses. But these books and articles are directedmainly to the functions that governments and not-for-profits have in common with businesses, not tothe unique, often “intellectual,” activities in whichthey engage.

In many governmental and not-for-profit organiza-tions ABC systems did not merely replace conven-tional cost accounting systems that provided reasonable, if not necessarily optimal, measures of costs. These organizations previously had eitherno, or inadequate, costing systems. For these orga-nizations, ABC was the first real measurement system, and the primary benefit of ABC was inproviding the structure for needed accountingreforms.

Unique Characteristics ofUniversitiesWe chose to apply ABC to universities rather than to another government or not-for-profit organizationfor two primary reasons. First, the costs of higher

1 This description of ABC is adapted from M.H. Granof, P.W Belland B.R. Neumann, Accounting for Managers and Investors(Englewood N.J, Prentice-Hall, 1993).

2 See, for example, Brimson, J.A. and Antos, J. Activity-BasedManagement for Service Industries, Government Entities andNonprofit Organizations (New York: John Wiley & Sons,1994), and Kehoe, J., Dodson, W., Reeve, R. and Plato, G.,Activity-Based Management in Government (Washington,Coopers & Lybrand, L.L.P., 1995).

3 D.D. Acton, and W. D. J. Cannon. “Activity-based Costing in aUniversity Setting,” Journal of Cost Management (March1997): 32-38.

4 J.M. Trussel, and L. N. Bittner. “As easy as ABC,” NACUBOBusiness Officer (June 1996): 34-39.


ABC Costing

ABC systems accumulate costs intoactivity cost pools. These are designedto correspond to the major activitiesor business processes. By design, thecosts in each cost pool are largelycaused by a single factor—the costdriver.

ABC systems allocate costs to prod-ucts, services, and other cost objectsfrom the activity cost pools using allo-cation bases corresponding to costdrivers of activity costs.

Allows for non-linearity of costs within the organization by explicitlyrecognizing that some costs are notcaused by the number of unitsproduced.

Focuses on estimating the costs ofmany cost objects of interest: units,batches, product lines, businessprocesses, customers, and suppliers.

Because of the ability to align alloca-tion bases with cost drivers, providesmore accurate information to supportmanagerial decisions.

By providing summary costs of orga-nizational activities, ABC allows forprioritization of cost-managementefforts.

Relatively expensive to implementand maintain.

Traditional Costing

Traditional costing systems accumu-late costs into facility-wide or depart-mental cost pools. The costs in eachcost pool are heterogeneous—theyare costs of many major processesand generally are not caused by a single factor.

Traditional systems allocate costs toproducts using volume-based alloca-tion bases: units, direct labor input,machine hours, revenue dollars.

Generally estimates all of the costs ofan organization as being driven bythe volume of product or servicedelivered.

Focuses on estimating the cost of asingle cost object—unit of product or service.

Because of the inability to align allo-cation bases with cost drivers, leadsto overcosting and undercostingproblems.

Cost control is viewed as a depart-mental exercise rather than a cross-functional effort.

Inexpensive to implement and maintain.

Figure 1Difference Between ABC and Traditional Costing

Cost Pools

AllocationBases

Hierarchy ofCosts

Cost Objects

DecisionSupport

Cost Control

Cost


education are increasing far more rapidly than thatof other goods and services. In fact, public concernover escalating costs led Congress, in 1997, toestablish a special concern to investigate the rea-sons behind the increases and to make recommen-dations as to how the costs can be contained. TheAppendix to this report provides background on the commission and summarizes some of its con-clusions that are relevant to this report. Consistentwith the commission’s findings, we believe thatproper cost information and computation are essen-tial ingredients of cost control and containment.

Second, with respect to cost measurement andcontrol, universities seem to present the MountEverest of challenges. If adequate systems of mea-surement and control can be established in univer-sities, then they can be imposed upon virtually allother types of organizations. By nature, universitiesare not amenable to the constraints required ofsound management and cost control:

• Many members of university faculty are “freespirits” who have chosen academic careers so as to escape the authoritarian governancestructures that are typical in other types oforganizations. Viewing their scholarly efforts as having intrinsic, rather than monetary, value, they are often suspicious of practicesthat hold them accountable for measurableoutcomes. Most certainly they disdain any linkages between the costs and benefits of their activities.

• University administrators lack the authorityconventionally accorded managers. In contrastto most businesses, key decisions that affectcost are made by faculty and administrators at the lower, rather than the upper, echelons of the organization chart. Examples includedecisions relating to:

– course offerings

– number of sections of a course to be scheduled

– new programs, such as Ph.D. or Master’s programs, either within a specific disciplineor across disciplines

– faculty research projects, and

– teaching load reductions for administrativeand research assignments.

More tellingly, these decisions are often madewith, at best, only a passing nod to cost impli-cations. Although considerable attention maybe paid to how the new activities will affectfaculty and staff workloads, almost never is adollar cost assigned. Even if costs are takeninto account, the focus is invariably upon thosethat are short term and incremental. Oftenthese costs are determined to be minimal, asthe proposed activities seemingly necessitateno new faculty. All the same, the long-term,indirect costs may be considerable. For exam-ple, if existing faculty must direct their effortsto new courses, additional instructors may berequired to teach the courses currently taught.Further, the new courses or programs maynecessitate additional office and classroomspace and place additional demands on ancil-lary service departments, such as the dean’soffice, the counseling staffs, and the technologysupport groups.

• Universities employ fund accounting systems.These are designed primarily for compliancerather than for providing the information need-ed for effective management. They are requiredbecause the funds of universities come frommultiple sources and may be restricted for spe-cific purposes. The salaries of individual facultymembers may be paid out of several differentaccounts — a standard state appropriationaccount, one or more endowment accountsand one or more research accounts. In fact,even when faculty entertain new recruits, thedinner tab may have to be divided among twoaccounts — one for food, another for alcoholicbeverages. In addition, some of these accountsmay be under the control of administrators atthe university level, whereas others may beunder the control of college deans or depart-ment heads. Thus, the answer to simple ques-tions, such as “What is the total compensationof faculty member X?” or “How much financialaid was given to Ph.D. candidate Y?” mayrequire hours of research.

• The budgets of universities are likely to mirrortheir fund accounting systems. They are not


typically tied to strategic plans or measurableoutcomes. Often allocations to colleges anddepartments are driven by set formulas or“across the board” percentage increases. Toaddress the special needs of individual units or to promote high-priority projects, the centraladministration may “hold back” or “tax” a por-tion of the amounts a college or departmentwould otherwise be given. The result is a bud-get process that is as complex as the account-ing system and equally deficient in promotingsound management and control.

• The accounting systems of large universitiesmay produce a plethora of periodic reports;most may be unintelligible to their intendedusers. In fact, the budgets and related financialdocuments are likely to be understood by rela-tively few university administrators — mostlikely budgeting and accounting officers butnot senior academic officials.

• Universities lack well-defined objectives ormeasurable outcomes. With respect to theirmain “products” — teaching, research and ser-vice — more is not necessarily better. At thesame time, quality is very much in the eye ofthe beholder. Thus, small classes are not neces-sarily less cost-efficient than large and may noteven be qualitatively superior.

• The outputs of faculty — the key universityemployees — are interrelated and not clearlyseparable from each other. There are no obvi-ous ways to distinguish between facultyresearch, teaching and professional service. Inconducting their research, for example, facultydirect and train graduate students. In reviewingjournal articles for academic publications (aform of professional service), they gain insightsthat will influence both their teaching andresearch.

• Just as the outputs of faculty are not readilyseparable, the distinctions between inputs andoutputs and producers and consumers may beequally blurred. Ph.D. students, for example,may be appointed as research assistants andteaching assistants — positions for which theyare compensated. In those positions they obvi-ously provide service to the university and addto its research and teaching costs. At the sametime, however, their teaching and research

assignments are essential elements of their edu-cation and may be degree requirements. Fromthat perspective, the research and teachingassistants are educational consumers ratherthan providers and their compensation may beinterpreted more as financial aid (like a schol-arship) than as a research or instructional cost.

• The costs and revenues of a university may beintegrally related and certain costs may not beincurred unless they were to be explicitly fund-ed from outside sources. For example, universi-ties carry out certain research projects only ifindependently funded. Their students receivescholarships from private parties. They receivedonations of computer equipment. It is notalways apparent whether the funds receivedfrom outside sources are a substitute forresources received from state appropriations orwhether they enable the university to carry outactivities that they otherwise would not.

• The capacity constraints of universities are notclearly discernable. Faculty, for example, donot work a standard number of hours per weekand hence can undertake additional assign-ments without any readily apparent reductionin the time devoted to other assignments.Similarly, new programs and activities can beadded without any obvious sacrifice in thequantity or quality of existing ones. At thesame time, capacity costs are proportionatelyhigh. Classrooms, laboratories and dormitoriesare subject to long planning, approval andconstruction processes and it often takes atleast two or more years before new facultypositions can be approved and filled.


This study applies the concepts of activity-basedcosting to the “intellectual” and business activitiesof a university setting. More specifically, we focuson the activities of one academic department aspart of (and benefiting from the resources of) a larg-er College of Business. Our goals, however, aretwofold: (1) to address some of the special prob-lems of managing the university enterprise as dis-cussed above; and (2) to demonstrate an improvedconceptual model for measuring the costs of ser-vices provided by knowledge workers in not-for-profit and governmental enterprises.

Reason for Focusing on aDepartment Rather Than an Entire UniversityTo test the feasibility and value of ABC in a univer-sity environment, we focused on the accountingdepartment of a large public research university. Thedepartment is one of five academic departments ina college of business administration — the othersbeing finance, marketing, management, and man-agement science information systems. In terms ofsize and range of programs and activities, thedepartment is comparable to a mid-sized college.

We opted to apply ABC at the departmental ratherthan the university level for several reasons:

• As previously discussed, university costs are asmuch affected by decisions made at the depart-mental and college level (e.g., those relating tocourse offerings, new programs, and researchprojects) as at the university level. Whereas the

allocation of university-wide costs, includingthose of services such as maintenance andpolice, may be necessary to determine the“full” cost of an academic program or a studentcredit hour, such cost would be relevant forfew routine operational decisions that universi-ty administrators are called upon to make.

• Our preliminary research indicated that theissues that we would have to address are moresalient at the lower than at the higher levels.For example, one of the least tractable (andcontroversial) questions that we faced was howto allocate the compensation of faculty mem-bers among the activities in which they engage.By contrast, issues of allocating the cost ofhigh-level administrative offices among the various colleges and other organizational unitswere not unlike those faced by a corporation in allocating its headquarters costs.

• The accounting system of a department wasmanageable and within the resource constraintsof this project, whereas that of an entire univer-sity was not. The modern university consists ofan extraordinarily wide range of organizationalunits — athletics, housing, food service, plantmaintenance and police — in addition to teach-ing and research.

Features of the DepartmentThe accounting department to which our study wasdirected consists of 25 tenure-track faculty, andapproximately 20 lecturers and 8 staff members. Ithas direct responsibility for four degree programs: a

Description of the Study


four-year undergraduate program (BBA) in account-ing; a five-year combined undergraduate and mas-ter’s in accounting program (PPA), consisting of twoyears of general undergraduate study followed bytwo years of specialized undergraduate and oneyear of graduate accounting studies; a stand-alonemasters in accounting program (MPA), a one-to-two-year program; and a Ph.D. program. In addi-tion, it offers both required and elective courses for university undergraduate students who do notmajor in accounting, for MBA students, and forseveral executive programs.

As is typical in research institutions, tenure-trackfaculty engage in three broad categories of activi-ties: teaching, research, and “service.” Promotionand salary increases are based mainly upon qualityof teaching (as assessed by student ratings andother means) and research (as evaluated primarilyby number and quality of publications). However,faculty are also called upon to fill various depart-mental administrative positions (e.g., directors of

the various programs) and to serve on department,college, and university committees. They are furtherexpected to serve as officers, editorial board mem-bers, and committee members of academic andprofessional organizations.

The department receives its resources from multiplesources. As summarized in Figure 2, its budgeted(state) funds are allocated from several accounts,the moneys in which are generally not interchange-able. These include accounts for salaries and wages(including separate sub-accounts for faculty, assis-tant instructors and teaching assistants, full-timeand part-time administrative staff) and for “mainte-nance, operations and equipment.” Faculty salariesare supplemented by income from endowments,such as professorships and fellowships, some ofwhich are within the direct control of the depart-ment while others are under the control of the college. Further, the department receives bothrestricted and unrestricted grants from alumni, CPA firms, and corporations.

Figure 2Departmental Expenditures by Source

Academic Year 1997-98

State Funds—Restricted Accounts

Faculty salaries 76%

Staff salaries and wages 6%

Teaching assistant and assistant instructor salaries 7%

Maintenance, operations and equipment 2%

Total expenditures from state funds 91%

Endowment Income and Unrestricted Gifts

Professorships and fellowships (faculty salary supplementsand research support) 6%

Unrestricted gifts used for travel, dues and subscriptions,and miscellaneous departmental activities 3%

Total expenditures from private sources 9%

Total expenditures from all sources 100%


Features of the ABC ModelWe developed our model in two stages. In the first,we focused exclusively on costs that were withinthe control of the department. The largest of thesewas faculty and staff compensation. Other costsincluded computers and other equipment, suppliesand faculty allowances for travel, books and sub-scriptions, and other costs. In the second stage, weadded college costs that benefited all academicdepartments. These included costs of the dean’soffice, computers and multi-media services, andcareer services. We also included a charge foroccupancy. Our purpose in adding the collegecosts was not to extend our analysis to academicdepartments other than accounting. Rather it wassimply to incorporate the college costs that benefit,in part, the accounting department.

First Stage In the first stage, based on discussions with thedepartment chair and other college administrators,we developed a set of multi-level cost objects. As illustrated in Figure 3, we began by allocatingfaculty compensation and related costs (such asmiscellaneous allowances for travel, books, andjournal submission fees) to four activity cost pools:

• teaching

• research

• service

• Ph.D. advising (distinguished from advising inother programs mainly because of the substan-tial time commitment required of several facul-ty members).

ResearchSupport

ResearchOutputs

ServiceOutputs

UnusedCapacity

CourseSections

Research

Service

Teaching

Ph.D.Advising

ProgramsBBAPPAMPAMBAPhDOp2

Other

GeneralAdmin.

TeachingSupport

Faculty Resource

CostsNon-Faculty

ResourceCosts

• Dept.• OH• Admin. Staff

= intermediate cost pool

= cost object

• Salary• Benefits• Support• TA/RA

Figure 3Design of ABC Cost Allocation System—Department Level


Owing to the magnitude of the dollars, this alloca-tion was the most critical in our model. We madethe allocation individual by individual, based ongeneral guidelines, tempered by our own judg-ments as to the activities engaged in by each indi-vidual faculty member. Typically we assigned 50percent of a tenure-track faculty member’s costs to teaching. The university counts a “full” teachingload as four courses per semester. That is, a facultymember, such as a lecturer, who has no research oradministrative responsibilities, is expected to teachthat many courses. However, in light of their otherresponsibilities, tenure track faculty typically areassigned only two courses — 50 percent of the“full” load. Therefore, we assigned 50 percent of a tenure-track faculty member’s costs to teaching.However, if a faculty member, whether tenure trackor not, taught more or fewer courses, we changedthis percentage accordingly.

The department’s policy is to shield newer facultyfrom administrative assignments. Therefore, weassigned the entire non-teaching portion of assis-tant professors’ compensation and related costs toresearch. By contrast we assigned smaller percent-ages of more senior (tenured) faculty’s costs toresearch and assigned the balance to either serviceor Ph.D. advising. In this pilot project, we basedthe allocation on our own knowledge of the facultymember’s activities. However, it is our expectationthat were the model to be fully implemented, thefaculty member himself would make the allocation,perhaps in consultation with the departmentalchairman. Faculty members would have littleincentive to “fudge” the percentages because, ifand when their contributions to the departmentwere to be compared against their costs, all costsand contributions would be taken into account.Gaming this allocation to make the cost-to-contri-bution ratios in one area look better would thencause those in another area to look worse.

Having allocated the faculty costs to the four activi-ty cost pools, we then assigned the teaching portionof each faculty member’s costs to the specific sec-tions that he or she taught. We generally assumedthat the faculty member devoted an equal amountof time to each section, but we made exceptions fornew preparations or other special circumstances.Correspondingly, we added the compensation ofteaching assistants to each section.

In the final allocation of faculty teaching costs, weassigned each section to a program. This allocationwas based on the proportion of students in the sec-tion that were enrolled in the particular program.

Taking a similar approach with regard to staff andother administrative costs, we apportioned theexpenditures among three pools of costs, compara-ble to those to which faculty costs were assigned:

• research support

• general administration

• teaching support.

Then, we assigned the teaching support costs to thespecific academic programs. Both allocations weremade on a case-by-case assessment of the particu-lar duties of individual staff members.

The sum of the faculty costs and the administrativecosts that were assigned to the academic programsconstituted the first-stage costs — those that werewithin the control of the department.

Second StageIn the second stage of the project we identified and assigned various college costs to the programsof the departments. As illustrated in Figure 4, theCollege comprises several administrative and sup-port departments. We directed attention to six specific cost centers:

• Career Services

• Media Services

• Technology (computer services)

• The College Dean’s Office

• The Graduate Business Dean’s Office

• The Undergraduate Business Dean’s Office.

The Business School has several other servicedepartments, including an office of internationalprograms, an executive education department, abureau of business research and a developmentoffice. We determined, however, that these depart-ments do not provide sufficient support to theaccounting department programs to warrant allo-cating their costs. Indeed, for some of these depart-


ments, such as executive education, the accountingdepartment provides support to the service depart-ments rather than the other way around.

The Career Services Office coordinates job recruit-ing and employment advising for all studentsenrolled in the Business School, with the exceptionof Ph.D. students. However, it does not directequal amounts of time per student to different pro-grams. For the most part, students avail themselvesof the office’s services mainly in the last two yearsof their programs. Thus, the office might provideservices to all students in the two-year MBA pro-gram, but to only half of those in the four-year BBAprogram. Similarly, most PPA and MPA students arecommitted to careers with major CPA firms. Therecruiting programs of these firms are highly struc-

tured, and the firms conduct interviews within aspecific window of time. Hence, the office pro-vides intensive services to these students but onlyfor a short period of time.

The Career Services Office is funded mostly by stu-dent fees. Moreover, the majority of its staff isassigned specifically to either graduate or under-graduate students and is paid from separate gradu-ate or undergraduate student fee accounts. As aresult, we were able to identify most of the office’scosts as being applicable to a graduate (MBA andMPA and PPA-Graduate Year) or undergraduate(BBA and PPA-Undergraduate Years) category ofprograms. However, once we made this determina-tion, we then had to assign them to a specific pro-gram within that category. Although we considered

ResearchOutputs

ServiceOutputs

ExcessCapacity

ProgramsBBAPPAMPAMBAPhDOp2

Other

College Cost CentersAcademic

Department Costs

Faculty R

esou

rce Co

sts

No

n-Facu

lty Reso

urce C

osts

= cost object

AdministrationResource Costs

Support DepartmentResource Costs

CollegeDean

GraduateDean

Undergrad.Dean

CareerServices

MediaServices

Technology

BusinessResearch

ExecutiveEducation

Development

International

= costs not allocated

Figure 4Design of ABC Cost Allocation System—College Level


doing this on the basis of student enrollment orsome other “hard” percentages, we opted to relyon the considered judgment of the program’s direc-tor as to how much time the office staff devoted toeach of the programs.

Media Services provides, maintains and repairsclassroom computers, projection equipment andrecording equipment. In addition it provides vari-ous media-related services, such as creating slidesand duplicating videotapes. As with most BusinessSchool units the department is funded from severaldifferent sources: student fees, user charges andoperating budget.

Media Services keeps records indicating the affilia-tions of the departments that use its equipment.Accordingly, it was able to provide us with thenumber of “uses” by each of the academic depart-ments (e.g., finance, management, accounting, etc.)as well as the name of the individual, usually a fac-ulty member, requesting the equipment. It doesnot, however, keep track of the programs in whichthe equipment is used. Thus, for example, equip-ment used by the finance department could beused in an undergraduate or an MBA course. Thatused by the accounting department could be usedin courses that served any one of its programs. Tofurther complicate the cost allocation process,when students check out equipment MediaServices does not note either the course or the pro-gram in which the student was enrolled. Further, itdoes not maintain records as to the number of ser-vice calls provided to the computer and otherequipment in classrooms that were specially desig-nated for MBA courses, each of which has a fullrange of multimedia equipment installed.

In general, we were able to base the allocation ofMedia Services costs on the number of “equipmentuses” by each academic department, with costsassigned to programs based on our knowledge ofwhich faculty member generally teaches in whichprogram. However, we made important exceptionsfor certain categories of costs. Because the expen-sive equipment permanently installed in the dedi-cated MBA classrooms was available for use at alltimes and no record of actual usage is kept, weassigned (based on our discussions with the depart-ment’s director) a disproportionately large equip-ment charge directly to the MBA program.

The Technology (computer services) Departmenthad a relatively large budget, yet presented espe-cially intractable allocation decisions. Its recordsprovided comprehensive information as to expendi-tures by each of its divisions (e.g., training, network-ing, programming, etc.), but revealed little as to thebeneficiaries of its services. Accordingly we wereforced to allocate the costs to programs in propor-tion to number of students, after making assump-tions about computer usage by program and theirrelative consumption of this department’s resources.

The assumptions we were forced to make are illus-trative of the challenges inherent in allocating costsin a way that is decision-relevant for managers.Departmental cost reports gave adequate informa-tion as to the costs of resources provided by thedepartment. However, they were insufficient tomeasure how the resources were consumed and,hence, how to allocate them to particular groups ofusers. Thus, for this department there is no way toanswer such pressing managerial questions as:

• What is the cost of supporting the present poli-cy requiring all MBA students to purchase lap-top computers, and what is the correspondingsaving (if any) to the school when comparedwith providing computer labs?

• Which programs or groups of users are drivingthe increasing demand for computer supportservices?

• What is the cost saving associated with requir-ing all faculty and students to standardize onIntel-based Windows NT computers?

• What is the cost of supporting faculty computerusers, and is it significantly different betweendepartments?

• Should an internal charge for cost of supportbe imposed, and in light of such a marketmechanism, should external sourcing beallowed?

The lack of useful data in this department also illus-trates the importance of capturing operational data,in addition to traditional cost data, in an effectiveaccounting system.

The costs of the three administrative Deans wereallocated among their constituent programs based


upon the number of students in each program. Thisallocation is contrary to the spirit of ABC in that itdoes not reflect the consumption of resources bythese programs. It was necessitated, however, bylimitations on data availability, and, while a hugeimprovement over the existing measures of cost, it represents an approximation at best — one thatrests on the inherently flawed assumption that allstudents are homogenous in their consumption ofresources.

Our model, it must be emphasized, is exceedinglyflexible. All “arbitrary” allocation rules can be easi-ly overridden for any individual faculty member orcost item. Moreover, at least for the department towhich it was applied in this study, it provided simi-lar results, irrespective of subjective allocationdecisions.

Added Feature — A Course Capacity ModuleWe added to our ABC system a “course capacity”module — a feature that proved extremely useful inhighlighting potential scheduling inefficiencies. Thismodule enabled us to calculate a measure akin toa cost of underutilized capacity — the cost ofassigning fewer students to a course section thanthe section was intended to accommodate.

To develop this module, we categorized eachcourse by type and assigned a capacity to it. Thecapacity was based mainly on the “close limits”that the department itself imposed on the coursesections. Thus:

• Large lecture course 150 students

• Case course 60 students

• Standard discussion course 40 students

• Seminar 10 students

For each section, we calculated the ratio of stu-dents enrolled in the class to total student capacityand classified that proportion of the section’s costsas “used capacity.” The remainder of the section’scost was ascribed to “unused capacity” — a mea-sure of the “opportunity costs” of under-enrollment.

Unused capacity may be the consequence of ques-tionable policies, such as permitting faculty toteach two sections of a course when only one is

needed. Our approach allows universities to quan-tify the costs of these policies. On the other hand,unused capacity cannot automatically be equatedwith inefficiencies — no more than can emptyhighways at midnight. Facing a need to staff certaincourses at peak times (e.g., a fall or spring semes-ter), a department may have to tolerate unusedcapacity in sections in off-peak times. Or, it mayhave to schedule sections at odd times (e.g., earlymorning or late afternoon) to accommodate stu-dents who require a particular course but have timeconflicts with offerings at more popular times. And,of course, smaller classes may be qualitativelysuperior to larger courses.

Universities, of course, are not the only organiza-tions to incur costs of unused capacity. Virtually allgovernment and not-for-profit entities have “slack”in their systems — sometimes for legitimate rea-sons, sometimes as the consequence of carelessscheduling or, as we found at the university, as ameans of providing hidden “perks” to selectedemployees. To be sure, administrators can oftenidentify unused capacity simply by comparingactual and potential outputs (e.g., as we did, bycomparing actual enrollments with assigned classlimits). They do not need cost data to point topotential inefficiencies. However, cost data addpoignancy to the message and highlight the conse-quences of resource misallocation.

Examples of Problems and IssuesEncounteredIn applying our model we encountered severalissues and problems. Although some may be uniqueto the college studied, we believe that most arecommon to other university departments and aretypical of those that would be faced in applying themodel to a wide range of governmental agencies.

• The outcomes of a university cannot readily be assessed in terms of quality. In our model a course taught by a senior professor would be allocated a far greater cost than one taughtby a teaching assistant. Therefore, the seniorprofessor may appear to be less efficient thanthe teaching assistant. But by more effectivelycapturing and presenting the cost data, weallow administrators to make an informed judg-ment (using student evaluations, reviews of


exams and course materials, and any othermeans) whether that is, in fact, the case.

• As in any organization, some employees aremore productive than others. In our model, we allocated the cost of each faculty member’scompensation-related costs among the variousactivities in which he engages. The compensa-tion of a “slacker” tenured faculty memberwho devoted 100 percent of his time to teach-ing (and performed none of the research or ser-vice that is part of his job and in exchange forwhich he receives a reduced teaching load)would be assigned entirely to the class sectionshe taught. By contrast, only a small portion ofthe compensation of another faculty member— one who devoted only a small percentage ofhis time to teaching, but taught the same num-ber of sections and worked far more hours —would be allocated to teaching. Thus, the sec-tions of the slacker would be reported as morecostly than those of the more productive facul-ty member. More properly though with obviousproblems of measurement, a portion of theslacker faculty member’s compensation shouldbe assigned to “waste and inefficiency.” Wedecided not to make that judgment.

• The department receives support from numer-ous sources, some internal and some external.Our general approach was to focus on depart-mental costs, irrespective of how the costs werefunded. Thus, for example, we did not distin-guish between faculty salaries paid from stateappropriations and those paid from endow-ments that were under the control of the univer-sity. At the same time, however, we did not takeinto account the cost of activities that likelywould not have been incurred had they notbeen independently paid for. Therefore, if a fac-ulty member received an outside research grantthat covered his salary for a semester, we didnot include the salary as a departmental cost.Similarly, we excluded all financial aid to stu-dents from our calculations inasmuch as mostof the financial aid, although within the controlof the department, is from private donations.

• Ph.D. students are required to serve as teachingand research assistants in various stages of theirprogram. In these roles they not only make sig-

nificant contributions to the department, butreceive substantial educational benefits as well.Nevertheless, we accounted for their compen-sation as if they were employees, assigningtheir salaries and benefits to the teaching andresearch pools.

• The accounting and information system of nei-ther the department nor the college (which arepart of the larger university-wide accountingsystem) was sufficient to support ABC andwould require a major overhaul before itwould be able to do so. Like the systems ofmany other government and not-for-profit orga-nizations, the systems of universities are nottypically designed to provide management-rel-evant information — and certainly not to “costout” services, programs or activities. For exam-ple, the accounting system of the departmentdid not automatically generate data on the totalcompensation of individual faculty members.We did our best to obtain reliable data, butacknowledge that they are not as accurate asrequired before they can be relied upon fordecisions and publicly defended.

• The College’s Executive Education Departmentsponsors numerous executive developmentprograms and MBA programs targeted to spe-cial audiences. Faculty devote considerabletime to these programs and are compensatedapart from their normal salaries. Since facultydo not work a standard number of hours eachweek, it is almost certain that some of the timedevoted to executive education takes awayfrom time devoted to their other activities.Nevertheless, because faculty are compensatedseparately, we elected not to assign any portionof faculty time to executive education.

The question of how to account for executiveeducation is similar to that of how to accountfor a wide range of outside activities. For someof these activities, such as independent con-sulting, faculty may get monetary compensa-tion; for others, such as scholarly lectures atother universities they receive only intangiblerewards, such as enhanced prestige. Mostactivities involve costs to the university in thatthey detract from time that can be devoted toother pursuits. Yet faculty are encouraged to


engage in many of these activities because theyprovide benefits to both the university and theindividual faculty member.

• The university does not charge colleges orother organizational units for the space thatthey occupy. From the perspective of theseunits, space, along with attendant utility andmaintenance costs, are considered free goods.Space, however, is a major university con-straint, always in short supply, and a relevantfactor in decisions relating to courses, pro-grams, and other activities. Accordingly, there-fore, we obtained data on the number ofsquare feet occupied by the various units anddepartments within the college and added a“shadow” cost to their budgets. We based thiscost on the prevailing market price of officespace in the area surrounding the university —an admittedly crude but nevertheless conve-nient measure.


Our specific results cannot be generalized beyondthe department and college to which we appliedour model. Nevertheless, they are of interestbecause they point to the types of findings thatapplication of the model can produce and to class-es of problems that are common in other academ-ic, governmental, and not-for-profit organizations.

Finding One: Great Disparities ExistAmong Various ProgramsWe found great disparities in the cost spent per student in the various academic programs. Asindicated in Figure 5, the average cost per studentin the Ph.D. program was $34,244. By contrast, thecost in the PPA program was only $9,346. Thesedifferences, however, greatly understate the gap

An ABC Model for aUniversity Department

Figure 5Annual Total Cost per Student in Accounting Programs, as Estimated by

ABC Costing Accounting Instruction and Support Costs Only*(Academic Year 1997-98)

PPA MPA Ph.D.Accounting Dept. Costs

Allocated Instruction $ 1,727 $ 2,935 $ 17,529

Space 1,311 2,227 13,300

Support Dept. CostsAllocated Support 3,729 6,821 2,355

Space 2,579 4,948 1,060

Total Cost $ 9,346 $ 16,931 $ 34,244

* Includes all support costs but excludes cost of instruction in departments other than accounting. Also excludes accountingdepartment unused capacity costs.


between the two programs because the measuresinclude only teaching and program-related admin-istrative costs. They exclude financial aid (because,as previously noted, much of this comes from out-side sources). Relatively few PPA students receivefinancial aid, and if they do it is generally in thehundreds of dollars. Ph.D. students, however, aretypically supported with grants of over $10,000.

These huge disparities can be attributed mainly to the small size of the Ph.D. classes (usually in single digits), the higher rank of faculty who teachin the Ph.D. program (exclusively tenure-track fac-ulty as opposed to lecturers and Ph.D. candidates),and special demands upon and rewards (such asreduced teaching loads) given to faculty who teachthe Ph.D. courses.

In businesses, product costs can readily be com-pared with product prices and “out of line” expendi-

tures can readily be identified. In governments andnot-for-profits, however, expenditures must be linkedto specific institutional objectives and the level ofexpenditures must be assessed in relation to the pri-ority assigned to those objectives. We can only ques-tion whether the college and department faculty andadministrators would opt for the same allocation ofresources among programs and courses if they con-sidered explicitly their costs and benefits. We arecertain, however, that administrators of other govern-ment and not-for-profit organizations are similarlylacking the cost data necessary to assure consistencybetween expenditures and objectives.

Finding Two: Unused Capacity IsCostlyWe became aware of the strikingly large costs ofunused capacity. As reported in Figure 6, this costexceeded the teaching costs that we had assigned

Figure 6Annual Accounting Department Costs, as Estimated by ABC Costing

(Academic Year 1997-98)

Panel A: Accounting department costs, by output

Cost %Research $ 1,621,560 31%Service 496,209 9%Teaching 3,176,368 60%

Total $ 5,294,137 100%

Panel B: Accounting department teaching costs, by program

Cost %PPA (A) $ 932,840 29%BBA* (C) 468,823 15%MBA (C) 370,706 12%MPA (A) 352,163 11%PhD (A) 210,351 7%Other (C) 78,251 2%

Subtotal 2,413,134 76%Unused Capacity 763,234 24%

Total $ 3,176,368 100%

(A) Costs incurred to provide courses in accounting degree programs (C) Costs incurred to provide courses in support of college-wide programs

* A small portion of these costs is incurred to educate students graduating with four-year undergraduate accounting degrees. Most are for supporting accounting classes provided to all undergraduate students.


to all of the academic programs except one.Examination of the data revealed situations such as the following:

• several sections of a multi-section coursewere oversubscribed while others were undersubscribed;

• two sections of a course were offered, so as to accommodate a single preparation for afaculty member, when one course would have sufficed;

• some elective courses were offered two timesper year and were substantially undersub-scribed in each.

As emphasized previously, this measure of unusedcapacity has to be interpreted guardedly. It isintended mainly to focus attention on potentialinefficiencies; by itself it is not necessarily indica-tive of wasteful practices.

Whereas the cost structures of businesses are beingconstantly tested in the competitive marketplace,governments and not-for-profits have no compara-ble mechanism to detect and root out inefficiencies.Our results highlighted the sort of staffing inefficien-cies that certainly are found in other academic, government and not-for-profit organizations.

Finding Three: ABC AccountingProvides Useful EfficiencyInformation As shown in Figure 7, cost information presented in a traditional, account-oriented fashion providesinformation about the costs of resources being sup-plied (in this case, by the accounting department)to make it possible to carry out the college’s mis-sion. However, an ABC perspective on the samedata provides (previously unavailable) informationabout what the college’s constituents are getting fortheir money. It also provides managerial informa-

Figure 7Annual Accounting Department Costs

Comparison of Traditional and ABC Cost Reporting(Academic Year 1997-98)

Traditional Reporting ABC Reporting

Faculty salary and benefits $ 4,253,309 Research $ 1,621,560

Other than faculty salary: Service 496,209

Research and teaching asst. 490,861 Teaching by program

Admin. salary and benefits 309,581 PPA 932,840

Admin. expenses 109,986 BBA 468,823

Furnishings and equipment 43,286 MBA 370,706

Travel 39,900 MPA 352,163

Events 18,757 PhD 210,351

Miscellaneous 28,457 Other 78,251

1,040,828 Unused capacity 763,234

Total expenditures $ 5,294,137 Total expenditures $ 5,294,137


tion that may be useful in assessing the school’sefficiency in delivering its various educational andresearch outputs.

For example, the accounting department makes asubstantial contribution to the MBA program,which is administered by the college rather thanthe department. Although obviously the depart-ment was aware of how many courses taught by its faculty were directed entirely or partially toMBA students, our model captured the financialvalue of its contribution. Unlike “number of MBAcourses taught” or even “numbers of MBA studenthours taught,” the dollar value of the contributiontakes into account the salary level of the facultywho teach in the MBA program (relatively high)and the extensive amount of teaching assistant and related support given to the MBA courses.

The problem of measuring the contributions ofdepartments to programs and activities other thanthose for which they are directly responsible isgeneric to a wide rage of academic and nonacade-mic organizations. The information that we pro-duced in our study — and which ABC is capable ofproviding — would unquestionably be of the typethat administrators need to take into account inmaking budget allocations among departments,programs, and activities.

Finding Four: Support Services DoNot Benefit Programs UniformlyAs made clear in Figure 8, support services do notbenefit programs uniformly. Whereas support costs(excluding space costs) were only $2,355 per Ph.D.student, they were $7,326 per MBA candidate.

Figure 8Annual Support Department Costs, as Estimated by ABC Costing

(Academic Year 1997-98)

Panel A: Support department cost per student by program, excluding space costs

Program #of Students Cost per Student

Ph.D. 50 $2,355

MBA 800 $7,326

MPA 120 $6,821

PPA 540 $3,729

BBA* 1,640 $2,955

Panel B: Support department cost per student by program, including opportunity cost of space**

Program #of Students Cost per Student

Ph.D. 50 $3,415

MBA 800 $13,476

MPA 120 $11,769

PPA 540 $6,308

BBA* 1,640 $5,314

* Undergraduate students (BBAs) are counted only for their last two years, when most of their classes are in the College of Business.

** Computed at the approximate retail lease price of $15 per sq. ft. per month.


Program costs would therefore be seriously distort-ed if support costs were allocated based on headcounts.

Finding Five: Space Costs AreSignificant Space costs, as discussed previously, although aprecious commodity, is a free good at the university(as well as in many government and not-for-profitorganizations). Yet, as is evident in both Figures 5and 8, they are of huge consequence, rivaling insignificance almost all other costs combined. Ourcomputations, however, seriously understate thefull cost of occupancy, in that the assigned cost isbased on prevailing rental costs. It excludes relatedcosts such as those for utilities and maintenance.


We believe that our project demonstrated theapplicability of ABC to universities and other typesof government and not-for-profit organizations inwhich personnel routinely engage in multiple activ-ities and programs. We drew several lessons fromour study:

• The stronger the existing managementaccounting and information system of an orga-nization, the easier it is to apply ABC. Yet theweaker the system, the greater the contribu-tion of ABC. Unfortunately, most governmentsand not-for-profits employ fund accounting sys-tems that are designed primarily to insure legalcompliance rather than to provide decision-useful information. Thus, we found it difficultto determine the compensation of individualfaculty members, let alone the costs applicableto programs and activities. The conventionalvirtue of ABC is that it results in cost determi-nations that better capture the full measure of aproduct’s cost. This is achieved mainly by tyingoverhead allocations to the factors with whichthe overhead most closely varies. In govern-ments and not-for-profits, by contrast, the maincontribution of ABC may be in encouraging theentity to establish the rudiments of a manage-ment-oriented accounting system.

• There are currently no comprehensive manu-als — and likely never will be — to provideoff-the-shelf instructions on how to install anABC system in government and not-for-profitorganizations. Each set of programs and activi-ties, as well as each type of cost, presents dif-

ferent issues and problems. We correctly antici-pated that many of the allocation issues facedby universities would be similar to those facedby industry. We were surprised, however, at thenumber and significance of the issues that wereunique to universities. For example, questionsof whether personnel (e.g., teaching assistants)are employees or “customers” and the issues ofscholarship assistance and research supportprovided by outside parties have no obviouscounterpart in industry. Yet they involve sumsof considerable magnitude.

• Flexibility is of the essence. Rigid allocationrules cannot readily be imposed upon organi-zations, like universities, characterized bydecentralized management systems. For exam-ple, we developed general guidelines for allo-cating faculty compensation costs among thethree main activities in which faculty engage— teaching, research and service. However,we allowed for exceptions when the guidelineswere inappropriate for individual faculty. Bydoing so, we were able to counter criticismsthat our allocations were unrealistic. At thesame time, we were able to demonstrate thatchanges from our allocation percentages towhat others could reasonably propose tendedto have little impact on the overall results.

• University faculty and, no doubt, many non-accounting employees of governments and not-for-profits are not only skeptical of but are threatened by, attempts to quantify thecost of the activities in which they engage.

Lessons Learned


Many faculty, for example, are understandablyfearful of how data on the cost of their researchmight be perceived. Others are concerned thatif the programs in which they teach have a highcost per student (perhaps because they requiresmall classes or individual instruction, such asprograms in music) they will be viewed asunnecessary drains upon university resources.Accordingly, if accountants are to gain the con-fidence and cooperation of members of theorganization, they must be sensitive to theirconcerns. More importantly, they must accom-pany their reports with warnings that seeminglyhigh costs or unfavorable variances (such as ourmeasure of excess capacity) are not necessarilyindicative of inefficiencies. Further, they musttake all reasonable steps to assure that theirdata will not, in fact, be used to draw unwar-ranted conclusions.

• ABC should be used to provide decision-usefulinformation, not to develop a conceptuallypure measure of costs. In conducting our pro-ject we quickly realized that the full cost ofcourses and programs must include the myriadof university-wide costs, such as those of thePresident’s office, admissions and student ser-vices. However, our focus was mainly on theaccounting department and the businessschool, and the addition of the university-widecosts would likely have little effect on the deci-sions made at those levels. In light of the limi-tations of existing systems, the benefits ofincluding these costs did not seem to justify the effort.

• ABC can provide interesting insights into thecosts of programs and activities. Our study, forexample, highlighted the cost of schedulinginefficiencies and pointed to the dramatic dif-ferences between the resources directed to thePh.D. students and undergraduate students. Nodoubt, many of our findings could have beendiscerned without the benefit of our study or of ABC. The reality is, however, that they werenot. Insofar as there is truth to the popularaphorism “what you measure is what you man-age,” then by measuring certain critical costs,ABC may force administrators to manage them.

• ABC may highlight changes in circumstancesthat have taken place gradually over time andof which administrators might not yet be cog-nizant. For example, in the Business Schooluntil a few years ago the computer supportdepartment consisted of a small staff of techni-cians. Their main job was to manage theschool’s mainframe computers and to installand support personal computers in the studentcomputer labs and in faculty offices. Today,however, the influence of computers is perva-sive and the technology group has a significantimpact on virtually all programs and activitiesin which the school is engaged. Yet, we wereunable to quantify meaningfully the group’scontributions to any of these programs andactivities. We raised questions, therefore, as towhether providing better cost data to thedepartment could assist them with managerialas well as accounting reforms.

• The rationale for using ABC in industry is toallocate indirect costs to goods or servicesbased, not simply on what is convenient, suchas direct labor, but on the factors by whichthey are most influenced. We found that thisrationale is no less compelling in universitiesthan in industry. We allocated the costs of sup-port services on the basis of the factors thatmost directly affect their magnitude rather thanthe conventional basis of, “number of students.”The result was cost information that we believeis more useful for virtually all decisions that fac-ulty and administrators are likely to make.

• Despite their unique features, universitieshave much in common with other types ofgovernment organizations. As public demandfor increased accountability becomes moreintense, governments must demonstrate thatthe benefits of the programs and activities inwhich they engage are commensurate withtheir costs. Accordingly they need accountingsystems that properly measure and report thesecosts. In governments, no less than universities,managers and other employees may contributeto more than one program and activity. Hencecosts must be allocated and ABC is therefore aslikely to be of benefit to other types of govern-ment entities as to universities.


The Need for Improved CostControls in Higher Education: The Conclusions of the NationalCommission on the Cost of HigherEducationOwing to concerns over rising costs of higher edu-cation, Congress established in 1997 the NationalCommission on the Cost of Higher Education.Constituted as an independent advisory body, theobjective of the Commission was to review compre-hensively the reasons for increases in college costsand prices and to make appropriate recommenda-tions as to how they might be contained. TheCommission issued its report in January 1998.1

The Commission acknowledged that “the UnitedStates has a world-class system of higher education,”and that “American higher education remains anextraordinary value.” Nevertheless, the Commissionreported that “…rising college tuitions are real. Inthe 20 years between 1976 and 1996, the averagetuition at public universities increased from $642 to$3,151 and the average tuition at private universitiesincreased from $2,881 to $15,581.”

Lest there be any doubt about the associationbetween costs and tuition, the Commission pointedout that during a portion of this period of rising

tuition, the instructional cost per student in publiccolleges universities increased from an average of$7,922 to $12,416 between 1987 and 1996, anincrease of 57 percent, and in their private sectorcounterparts the increase averaged 69 percent,from $10,011 to $18,387. Warning that collegesand universities had best pay heed to their risingcosts, the Commission noted that “Public anxietyabout college prices has risen along with increasesin tuition.”

The Commission concluded that “Financing a col-lege education is a serious and troublesome matterto the American people,” and now is “on the orderof anxiety about how to pay for health care orhousing.” Among the root causes of this problem:that academic institutions “have not seriously con-fronted the basic issues involved with reducingtheir costs and that most of them have also permit-ted a veil of obscurity to settle over their basicfinancial institutions.” It advised that unless UnitedStates colleges and universities get their fiscal hous-es in order, “policy makers at both the state andfederal levels could impose unilateral solutions thatare likely to be heavy-handed and regulatory.”

Nevertheless, the Commission emphasized thatthere is no clear cause and effect relationshipbetween rising costs and rising prices. “Linkingspecific cost increases to price increases is a trickymatter,” it noted. “Quite simply, the available dataon higher education expenditures and revenuesmake it difficult to ascertain direct relationshipsamong cost drivers and increases in the price ofhigher education.”

Appendix

1 National Commission on the Cost of Higher Education(NCCHE). Straight Talk about College Costs and Prices:Report of the National Commission on the Cost of HigherEducation, 1998.


A primary reason why rising tuition cannot be tiedto rising prices, the Commission contended, is theabsence of reliable data on college costs. “Institu-tions of higher education, even to most people inthe academy, are financially opaque,” it observed.“Academic institutions have made little effort,either on campus or off, to make themselves moretransparent, to explain their finances. As a result,there is no readily available information about college costs and prices nor is there a commonnational reporting standard for either.” Colleges, it said, “report on financial standards using onemethodology; report expenditures using another;and conform to government cost-recovery princi-ples with yet a third.” It further points out that sta-tistical data on college and university costs fail todistinguish between institutions that have markedlydifferent missions and therefore have very little incommon. Thus, the data may combine informationfrom major research universities and universitieswith medical schools and other costly professionalschools with that from small liberal arts institutions.

Our study confirms the Commission’s assertion thatthe data on college costs cannot be relied upon.However, whereas the Commission emphasizesthat the data are undependable owing to a lack ofcomparability, our research suggests that the prob-lem may be far more basic — perhaps colleges anduniversities lack the accounting systems to produceproper data.

Our study was premised on the notion that costinformation is an essential ingredient of cost con-trol — one that appears consistent with the conclu-sions of the Commission. Both the Commission andother researchers have suggested several possiblereasons for the increases in college and universitycosts. To cite but a few:

• Increase in faculty salaries and other laborcosts

• Increase in the number of nontraditional stu-dents who require remedial and other specialprograms

• Student demand for high quality equipmentand facilities

• Increases in numbers of programs and courses

• Increases in numbers of administrators

• Growth in the number of accreditation agen-cies and the demands that they place uponinstitutions.

Yet the Commission was unable to confirm any ofthese causes. “This Commission,” it said, “findsitself in the discomfiting position of acknowledgingthat the nation’s academic institutions, justlyrenowned for their ability to analyze practicallyevery other major economic activity in the UnitedStates, have not devoted similar analytic attentionto their own internal financial structures. Blessed,until recently, with sufficient resources that allowedquestions about costs or internal cross-subsidies tobe avoided, academic institutions now find them-selves confronting hard questions about whethertheir spending patterns match their priorities andabout how to communicate the choices they havemade to the public.” In other words, theCommission implied, colleges and universities donot know the cost of their courses, programs andother activities in which they engage.


Michael H. Granof is Ernst & Young Distinguished Centennial Professor in Accounting, Department ofAccounting, College of Business Administration, University of Texas at Austin. Professor Granof received his A.B. from Hamilton College, his M.B.A. from Columbia University, and his Ph.D. from the University of Michigan. His research and teaching interests include accounting in governmental and other nonprofitorganizations and financial accounting.

David E. Platt is Assistant Professor, Department of Accounting, College of Business Administration,University of Texas at Austin. Professor Platt received his B.S. from the Wharton School at the University of Pennsylvania, his M.B.A. from Syracuse University, and his Ph.D. from Cornell University. He is a CPAand has worked in manufacturing and public accounting. His teaching and research interests include man-agerial and cost accounting, non-financial measures of performance, and management control systems.

Igor Vaysman is Assistant Professor, Department of Accounting, College of Business Administration,University of Texas at Austin. Professor Vaysman received his B.S. from Trinity University and his Ph.D. from Stanford University. His teaching and research interests include managerial and organizational design, costing systems, and decentralization.

About the Authors

Igor Vaysman (left), Micahel Granoff (middle), and Dave Platt (right)


To contact the authors:Michael H. Granof Professor of Accounting Department of Accounting College of Business Administration The University of Texas at Austin CBA 4M.202 Austin, TX 78712-1172 (512) 471-5215

e-mail: [email protected]

David E. Platt Assistant Professor of Accounting Department of Accounting College of Business Administration The University of Texas at Austin CBA 4M.202 Austin, TX 78712-1172 (512) 471-5215


Igor Vaysman Assistant ProfessorDepartment of Accounting College of Business Administration The University of Texas at Austin CBA 4M.202 Austin, TX 78712-1172 (512) 471-5215


Key Contact Information

To download or order a copy of these reports, visit the Endowment website at: endowment.pwcglobal.com

ENDOWMENT REPORTS AVAILABLE

Managing Workfare: The Case of the Work ExperienceProgram in the New York City Parks Department (June1999)

Steven CohenColumbia University

Results of the Government Leadership Survey:A 1999 Survey of Federal Executives (June 1999)

Mark A. AbramsonSteven A. ClyburnElizabeth MercierPricewaterhouseCoopers

Credit Scoring and Loan Scoring: Tools for ImprovedManagement of Federal Credit Programs (July 1999)

Thomas H. StantonJohns Hopkins University

The Importance of Leadership: The Role of SchoolPrincipals (September 1999)

Paul TeskeMark Schneider State University of New York at Stony Brook

Leadership for Change: Case Studies in AmericanLocal Government (September 1999)

Robert B. DenhardtJanet Vinzant DenhardtArizona State University

Managing Decentralized Departments: The Case ofthe U.S. Department of Health and Human Services(October 1999)

Beryl A. RadinState University of New York at Albany

New Tools for Improving Government Regulation: AnAssessment of Emissions Trading and Other Market-Based Regulatory Tools (October 1999)

Gary C. BrynerUniversity of Colorado School of Law

Determining a Level Playing Field for Public-PrivateCompetition (November 1999)

Lawrence L. MartinColumbia University

An Assessment of Brownfield RedevelopmentPolicies: The Michigan Experience (November 1999)

Richard C. HulaMichigan State University

Religious Organizations, Anti-Poverty Relief, andCharitable Choice: A Feasibility Study of Faith-BasedWelfare Reform in Mississippi (November 1999)

John P. BartkowskiMississippi State UniversityHelen A. RegisLouisiana State University

Business Improvement Districts and InnovativeService Delivery (November 1999)

Jerry MitchellBaruch College, The City University of New York

Profiles in Excellence: Conversations with the Best ofAmerica’s Career Executive Service (November 1999)

Mark W. HuddlestonUniversity of Delaware

San Diego County’s Innovation Program: UsingCompetition and a Whole Lot More to Improve PublicServices (January 2000)

William B. EimickeColumbia University

Using Activity-Based Costing to Manage MoreEffectively (January 2000)

Michael H. GranofDavid E. PlattIgor VaysmanUniversity of Texas at Austin

For additional information, contact:Mark A. AbramsonExecutive DirectorThe PricewaterhouseCoopers Endowment for

The Business of Government1616 North Fort Myer DriveArlington, VA 22209

(703) 741-1077fax: (703) 741-1076e-mail: [email protected]: endowment.pwcglobal.com

1616 North Fort Myer DriveArlington, VA 22209-3195


The Business of GovernmentBulk Rate

US Postage

P A I DPermit 1112

Merrifield, VA

About PricewaterhouseCoopersPricewaterhouseCoopers (www.pwcglobal.com) is the world’s leading professional services organization.Drawing on the knowledge and skills of 150,000people in 150 countries, we help our clients solvecomplex business problems and measurably enhancetheir ability to build value, manage risk and improveperformance. PricewaterhouseCoopers refers to theUS firm of PricewaterhouseCoopers LLP and othermembers of the worldwide PricewaterhouseCoopersorganization.

using activity based costing to manage more effectively

Documents