Does Competition Destroy Ethical Behavior?

By ANDREI SHLEIFER*

This paper shows that conduct described asunethical and blamed on greed is sometimes aconsequence of market competition. I illustratethis observation using examples of five cen-sured activities: employment of children, cor-ruption, excessive executive pay, corporateearnings manipulation, and involvement of uni-versities in commercial activities. In all theseexamples, I show how competitive pressureslead to the spread of the censured behavior. Mypoint is not to excuse or condemn any or all ofthese practices, but merely to pinpoint the cru-cial role of competition, as opposed to greed, intheir spread.I focus here on ethics, not efficiency. The

relationship between the two is complex. Inmany cases, ethical norms evolve to sustaincooperative behavior and thus to promote suc-cessful functioning of social institutions. Forexample, ethical condemnation of corruption isbased on the idea that officials should not self-ishly abuse public trust, and that a society func-tions better when its government works fairly.Likewise, the ethical norm against the employ-ment of children is driven in part by the moregeneral concern with abuse of the weak by thestrong. When ethics promote social coopera-tion, ethical behavior and efficient behavior typ-ically go together.In other instances, there is a mismatch be-

tween ethics and efficiency, either because eth-ical standards which might have had efficiencyjustifications long ago no longer do or becausethe behavior that is ethical in some idealizedsociety makes matters worse in the real world.For example, the ethical norm against debt orinterest, which might have been justifiable amillennium ago, is clearly no longer efficient.And while child labor might be a bad idea in aworld with good access to capital markets andeducational opportunities, for many families in

the developing world the alternative to childlabor is malnutrition and disease. These exam-ples of a mismatch suggest that behavior con-demned as unethical is not always inefficient.In still other instances, the conduct that is

advertised as ethical is the result of politicalindoctrination by parochial interests or of sim-ple confusion. For example, the ethical exhor-tations to Buy American! or to pay the livingwage are underwritten by labor unions servingtheir members, not the public. In various timesand places, tribalism, racism, anti-Semitism, thehatred of the rich, and other unsavory senti-ments reflected the dominant ethic. In theseinstances, what is considered ethical is very farfrom what economists would consider efficient.In some of the examples I discuss below, a

credible case can be made that the conductperceived to be unethical is also inefficient; inother examples, there is more ambiguity. Myinterest, however, is not to evaluate efficiency,but only to bring the crucial role of competitioninto the explanation of why activities morallysanctioned by the society spread.In four of the examples I discuss, censured

behavior reduces costs (in the last one, it raisesrevenues). I assume that the proprietor of thefirm values ethical behavior, but that such be-havior is a normal good. When sanctioned be-havior by competitors reduces their costs, it alsoreduces prices in the market, and as a result theproprietors income falls. When his incomefalls, so does his own demand for ethical be-havior, leading to the spread of censured prac-tices. The analysis I present is closely related tothe ideas in Gary Beckers (1957) classic studyof discrimination and reveals a broad range ofcircumstances where competition promotescensured conduct.

I. Child Labor

Under many plausible scenarios, the pres-sures of competition bring children into thelabor force. If hiring children is cheaper than hir-ing adults (even taking into account differences

*Department of Economics, Harvard University, Cam-bridge, MA 02138. I thank Olivier Blanchard, Mihir Desai,James Hines, Jesse Shapiro, and especially Daron Acemo-glu and Edward Glaeser for helpful comments.

414

in productivity) and a firm hires children, it canreduce prices. Its competitors must then hirechildren also, or be driven out of business (or, ina less extreme world, their willingness to payfor not hiring children declines when profitsfall). On the other side of the market, if parentsin one family can compel their children to workand thereby advantage themselves in competi-tion with other families (for food or for status),then competition among families forces morechildren into the labor force. Either of these twoforces of competition (on the demand side or onthe supply side) would bring children into thelabor force.Whether or not child labor is efficient de-

pends on a number of factors, such as whetherchildren and/or their parents correctly value ed-ucation, whether education is even available,whether capital markets are sufficiently devel-oped that families can borrow rather than sendchildren to work, what the family structure is,and so on. Presumably, the more efficient are allthese markets, the less likely we are to seechildren in the labor force.

II. Corruption

Robert Vishny and I (Sheifer and Vishny,1993) distinguish two kinds of corruption: withtheft and without theft. In the first kind, a gov-ernment official takes money in exchange forreducing the payments, such as taxes or tariffs,that the briber owes to the government. In thesecond kind of corruption, a government officialtakes additional money in exchange for givingthe briber goods, such as permits, to which he isentitled without the bribe. Relative to the re-gime with no corruption, corruption withouttheft raises costs, but corruption with theft re-duces costs. As a consequence, the latter kind ofcorruption spreads when markets are competi-tive. When a firms competitor can reduce histaxes through corruption, or can import by pay-ing lower bribes rather than higher tariffs, hecan pass on his savings to consumers. In acompetitive market, then, every firm must itselfpay bribes or go out of business. Even if theproprietor has some rents, his willingness to payfor ethical conduct declines as his profits do,leading him to bribe. The keener is the compe-tition, the higher is the pressure to reduce costs,and the more pervasive is corruption.

Corruption with theft has one additional com-petitive advantage: both the official and thebriber benefit, and neither has any incentive toreport the bribe to the police. In contrast, cor-ruption without theft raises costs, and hence thepotential briber has an incentive to complain.A further pressure comes from competition

for government jobs. In some countries, posi-tions allowing extensive bribe collection areauctioned off by senior officials. The prospec-tive officials who will collect the bribes mustpay for their jobs. This competition for jobscreates further pressure on corruption to spread:honest officials simply cannot afford the ap-pointments, and the officials who get the jobsare the ones who can collect the most bribes.This source of competitive pressure applies tocorruption both with and without theft, whichmight explain why corruption without theft isalso pervasive, even though it raises costs.

III. Executive Pay

Executive compensation has received a greatdeal of negative attention recently, with anemerging consensus that a large portion of whattop managers get is not incentive pay but rentsextracted by greedy CEOs from dysfunctionalboards (Marianne Bertrand and Sendhil Mul-lainathan, 2001; Lucian Bebchuk and JesseFried, 2003). There is undoubtedly some truthto this view, but competition for executives hasalso played a role, at least during the recentstock-market bubble.Suppose that managers can extract no rents,

so their pay is determined by their marginalproductivity or their contribution to the profit-ability of the firm. Shareholders care aboutmaximizing the market value of the firm, whichin an efficient market is the present value ofexpected profits. Assuming that a given man-ager does not contribute so much more to thefirm than the next best alternative, his level ofcompensation would be likewise moderate.Now suppose alternatively that markets are

not efficient, and expectations about futuregrowth, and consequently share values, can bemanipulated. Suppose some managers possess asuperior ability to manipulate investor expecta-tions, allowing them to drive up the price ofcompany stock, at least in the short run. Suchtemporary price bubbles in their stock can be

415VOL. 94 NO. 2 COMPETITION

very valuable to shareholders: they can sell intothe bubble, or their companies can acquire oth-ers using overvalued stock. In other words,managers who can create or sustain a bubble incompany stock reduce its cost of equity capital.If some managers are more able than others todo that, then their compensation is determinedby the competition among firms for this service.In such a market, the value of having an

executive who can reduce the cost of capitalthrough sustaining a bubble can be enormous,and so will be his competitive pay. The starriestexecutives would then be allocated to compa-nies, perhaps in technology or other glamouroussectors, where persuasion can generate a lot ofshareholder value. Companies will pay for theseservices, even when the executives do not con-tribute to profits or their growth at all.In this example, the market for executives is

perfectly competitive and efficient, yet execu-tive talent is defined by ability to raise valua-tions in an inefficient financial market. In recentyears, theory and evidence of stock-market in-efficiency have grown significantly (Shleifer,2000), especially with respect to the technologybubble (Eli Ofek and Matt Richardson, 2003).The very same period saw an explosion of ex-ecutive pay, especially in the high-valuationfirms. As the bubble collapsed, these episodesof extraordinary pay were attributed to the greedof the CEOs. Yet it seems equally plausible thatshareholders during the bubble were happy tocompensate these executives for promoting thebubbles in their shares.

IV. Earnings Manipulation

Parallel to the growth of executive pay duringthe technology bubble, many observers havepointed to an increase in a range of unsavorycorporate practices, such as manipulation ofearnings reported to shareholders. Companiesincreasingly reported pro forma earnings (earn-ings not computed according to generally ac-cepted accounting principles) on their web sites(Gene DAvolio et al., 2002). For high-technologyfirms in particular, these earnings tend to beconsiderably higher than those computed in thestandard fashion. Companies have also increas-ingly smoothed their earnings growth, usedaccruals to drive up earnings, and changed pen-sion rate-of-return assumptions to show higher

earnings (Daniel Bergstresser and Thomas Phil-ippon, 2003; Bergstresser et al., 2003). At leastsome companies pursued downright illegalpractices in reporting earnings.There is growing evidence that these account-

ing tricks in fact represent a conscious attemptto manipulate shareholder beliefs. Companiestend to use these strategies to show particularlygood earnings numbers right before they makeacquisitions, and right before their executivesexercise stock options (Bergstresser et al.,2003). As this evidence has accumulated, com-mentators have denounced the degradation ofcorporate standards of ethics.The discussion of earnings manipulation has

largely ignored the importance of competitionfor funds. During the bubble, firms faced pow-erful incentives to drive up their share prices,largely by reporting higher earnings growth. Ifearnings manipulation helps sustain a high val-uation, it reduces the cost of capital, enablingcompanies to make acquisitions for stock, toattract better executives and workers with stockoptions, and even to issue new shares. For somecompanies, this was an issue of survival as well.Many young high-technology firms, such asAmazon.com, financed their ongoing operationsby issuing equity. Without creative accounting,their cost of capital might have been too highfor them to survive. Likewise, for many maturefirms, a high equity valuation was the matter ofsurvival as an independent company, rather thanbeing acquired for stock by a company with ahigher valuation. Vishny and I have shown that,in the takeover market, the choice is to acquireor be acquired (Shleifer and Vishny, 2003), andto execute the former strategy a firm needs ahigh valuation. In these ways, competitive pres-sures contributed to the rise of aggressive cor-porate accounting practices.

V. Commercial Activities by Universities

In a provocative recent book, Harvardsformer president Derek Bok (2003) expressesconcern with commercialization of universities.According to Bok, universities are increasinglyinvolved in highly commercial athletic pro-grams, which generate profits but may under-mine academic standards in admissions. Theyincreasingly pursue purely commercial researchefforts, which can interfere with faculty time,

416 AEA PAPERS AND PROCEEDINGS MAY 2004

the selection of research priorities, and even theimperative of openness and full disclosure ofresults. In recent years, universities have alsoincreasingly pursued commercial educationalefforts through online education, which again,in Boks opinion, may undermine core missionsof research and teaching their own students.While Bok regrets these developments, he

recognizes the crucial role played by competi-tion. Universities need resources for their coremissions, such as attracting the best faculty andstudents, getting these faculty to teach thosestudents, building housing and research facili-ties, and so on. Universities compete fiercely forfaculty and students, and to the extent that prof-its from commercial activities give them anadvantage in this competition, they embracethem. As in the other examples I have given,competition assures the spread of commercialpractices among universities.

VI. What Is To Be Done?

I have presented five examples of activitieswhich many people find ethically objectionable:child labor, corruption, high executive pay,earnings manipulation, and commercializationof education. These examples show how com-petition can encourage the spread of censuredconduct.What can a society do to discourage these

activities? As noted earlier, it is sometimes farfrom obvious that discouraging them is efficient.Corruption may enable small businesses to getaround unreasonable regulations, and actuallyencourage economic development. Child labormay improve the economic circumstances ofboth children and their families in places wherethe feasible alternatives are hunger and malnu-trition. High levels of executive pay may allo-cate the most talented managers to growingyoung firms. Some technology entrepreneursargued that conventional accounting harmedtheir companies as investors failed to under-stand the temporary nature of their losses, andthat therefore some freedom of accountingchoice was essential for accuracy. And com-mercialization of universities might enablethem to better discharge their core duties.I can think of (at least) three strategies for

curtailing unethical conduct, involving increasingamounts of coercion: long-run market pressure,

moral suasion, and government regulation.Some economists express an optimistic viewthat, in the long run, markets cure the problemsI discussed. If public opinion really turnsagainst child labor, firms that do not hire chil-dren will be able to charge higher prices. Firmsthat pay executives to pump up their stock, ormanipulate their earnings, will face eventualinvestor disappointment and a higher, notlower, cost of capital. And universities thatovercommercialize their activities will lose se-rious faculty and students.These arguments about long-run competition

are not compelling. While public opinion mayexert pressure in the long run, in the short runpeople want cheaper shoes, and most do notcare who makes them. Emerging-market sub-contractors refusing to hire children, countingon the tide in consumer sentiment for adult-made shoes, surely cannot survive. Likewise,firms that do not manipulate their earnings orcompete for glamourous executives might not sur-vive as independent entities long enough for real-ity to intervene. Finally, universities that eschewcommercialism may find themselves too far be-hind academically to catch up. Competitionmay take too long to work, and even in the longrun it need not work to promote ethical values.A more aggressive strategy is moral suasion.

One can try to convince companies (throughpersuasion and perhaps even boycotts) not to hirechildren, not to manipulate earnings, and not topay too much to executives. Derek Bok clearlyfollows this approach in urging university ad-ministrators to slow down commercialization.Moral suasion is likely to work better when

competition is less keen. Protests can discour-age Nike (with its rents from its name) fromhiring children, either on its own or throughsubcontractors. Moral suasion can slow downcommercialization among the wealthiest uni-versities, which have significant endowments.The richer the competitor, the more it will payfor being ethical. But the trouble with moralsuasion is that it can rarely stand up to the forcesof competition when competitors have no rents.This is precisely the reason why the phenomenaI describe are so pervasive. Even if managers,boards, or shareholders of companies are con-vinced that it is bad to manipulate earnings,pay a lot to executives, or hire children in de-veloping countries, their discretion is severely

417VOL. 94 NO. 2 COMPETITION

circumscribed by the imperative of commercialsurvival. Even in the richest private universities,competition for faculty and students is sufficientlykeen to at best delay the advance of commer-cialization. In most cases, then, moral suasionwill not stop the pressures of competition.What about government regulation? Govern-

ments often restrict entry, or ban cost-reducingpractices that are deemed unethical. Govern-ments prohibit many kinds of corruption andimpose criminal penalties on those who arecaught. In developed countries, governmentsban child labor. In most countries, governmentsregulate earnings reporting and are beginning toregulate executive pay. Finally, even in theUnited States, where universities are private,their not-for-profit status severely limits theirpursuit of commercial activities. In all the casesI discussed, the government is involved, al-though the nature of its involvement varies.Here again, however, one should not expect

too much success from regulation, especiallywhere the states enforcement powers are lim-ited. Even if the government pursues benevolentpolicies (which is far from universal), when itbattles against cost-reducing competition, it islikely to lose. This is partly why corruption,child labor, and unsavory accounting practicesremain pervasive.If this analysis is complete, it leads to a fairly

depressing conclusion. Globalization and the vic-tory of laissez-faire economics has made compe-tition keener in many countries and marketsaround the world. Does this imply that unethicalconduct is also becoming more pervasive?The answer, I believe, is no, for two separate

reasons. First, competition is the fundamentalsource of technological progress and wealth cre-ation around the world. The very same marketforces that might encourage unethical conductalso motivate firms to innovate and create newproducts, leading to economic growth. As soci-eties grow richer, their willingness to pay forethical behavior, through both government en-forcement and private choice, increases as well.As a consequence, both moral and regulatorysanctions work better in the richer countries,leading to more ethical behavior. Second, asBenjamin Friedman (2004) demonstrates, as so-cieties grow richer their views of what is ethicalchange as well. More universalist ethics thatemphasize cooperation and inclusion replace

the more tribal and parochial beliefs. As I haveargued in the Introduction, the ethics of coop-eration are much more likely to coincide withobjective notions of efficiency. For both ofthese reasons, the increased willingness to payfor ethical behavior and the improving matchbetween ethics and efficiency, competition islikely to promote ethical behavior in the long run.

REFERENCES

Bebchuk, Lucian and Fried, Jesse. ExecutiveCompensation as an Agency Problem. Jour-nal of Economic Perspectives, Summer 2003,17(3), pp. 7192.

Becker, Gary. The economics of discrimination.Chicago: University of Chicago Press, 1957.

Bergstresser, Daniel; Desai, Mihir and Rauh,Joshua. The Gerstner Effect. Mimeo, Har-vard Business School, 2003.

Bergstresser, Daniel and Philippon, Thomas.CEO Incentives and Earnings Management.Mimeo, Harvard Business School, 2003.

Bertrand, Marianne and Mullainathan, Sendhil.Are CEOs Rewarded for Luck? The Oneswithout Principals Are. Quarterly Journal ofEconomics, August 2001, 116(3), pp. 90132.

Bok, Derek. Universities in the marketplace:The commercialization of higher education.Princeton, NJ: Princeton University Press,2003.

DAvolio, Gene; Gildor, Efi and Shleifer, Andrei.Technology, Information Production, andMarket Efficiency, in Economic policy forthe information economy. Kansas City, MO:Federal Reserve Bank of Kansas City, 2002,pp. 12560.

Friedman, Benjamin. The moral consequencesof economic growth. New York: Knopf, 2004(forthcoming).

Ofek, Eli and Richardson, Matt. DotCom Ma-nia: The Rise and Fall of Internet Stocks.Journal of Finance, June 2003, 58(3), pp.111338.

Shleifer, Andrei. Inefficient markets. Oxford,U.K.: Oxford University Press, 2000.

Shleifer, Andrei and Vishny, Robert. Corrup-tion. Quarterly Journal of Economics, Au-gust 1993, 108(3), pp. 599617.

. Stock Market Driven Acquisitions.Journal of Financial Economics, December2003, 70(3), pp. 295311.

418 AEA PAPERS AND PROCEEDINGS MAY 2004