Journal of Public Economics 76 (2000) 309–335www.elsevier.nl/locate/econbaseDoes voluntary participation undermine the CoaseTheorem?†*Avinash Dixit , Mancur OlsonDepartment of Economics,Princeton University,Princeton,NJ08544-1021,USAAbstractThe Coase Theorem states that costless enforcement of voluntary agreements yieldsefficient outcomes. We argue that previous treatments fail to recognize the full meaning of‘voluntary’. It requires a two-stage game: a non-cooperative participation decision, followedby Coaseian bargaining only among those who choose to participate. We illustrate this in asimple public-goods model, and find outcomes ranging from extremely inefficient to fullyefficient. However, the efficient equilibrium is not robust to even very small transactioncosts. Thus, we cast doubt on Coaseian claims of universal efficiency. Finally, we outline akind of coercion that restores efficiency. 2000 Elsevier Science S.A. All rightsreserved.Keywords:Public goods, Coase TheoremJEL classification:H11; H41; D711. IntroductionIn his article ‘The Problem of Social Cost,’ Ronald Coase introduced a verypowerful idea of great importance. Coase’s article has been arguably the single*Corresponding author.E-mail address:[email protected] (A. Dixit)†Footnote by Avinash Dixit: Mancur Olson was Distinguished Professor of Economics and Directorof the Center for Institutional Reform and the Informal Sector (IRIS) at the University of Maryland. Hedied very suddenly on February 19, 1998, in the midst of numerous projects including a revision of thispaper. I have had to prepare the final version without the benefit of his insight, scholarship, energy, andenthusiasm.0047-2727/00/$ – see front matter 2000 Elsevier Science S.A. All rights reserved.PII: S0047-2727(99)00089-4310 A.Dixit,M.Olson / Journal of Public Economics76 (2000) 309–335largest influence on thinking about economic policy for the last three decades. It isone of the most — if not the most — widely cited economics article in recent1times.Coase argued that, given a precise allocation of property rights and the absenceof any costs of information or negotiation, two parties would arrive at a bargainthat would internalize any externalities between them. Though Coase took forgranted a government that allocated the property rights between the parties and acourt that enforced their agreed bargain, he emphasized that an efficient outcome2would occur whatever the initial allocation of legal rights. Coase extended hisanalysis beyond two-party externalities to larger groups and even to ‘amorphous’externalities or public bads like air pollution (see the Coase, 1988 book, in whichthe Coase, 1960 article is reprinted, pp. 24–25, 170–177, 180–182). While it isadmitted that transaction costs will increase when the number of people in thegroup impacted by the externalities or served by the public good are large, theargument is that in Coase’s idealized world of zero transaction costs, efficientoutcomes can be achieved no matter how large the numbers. Thus, Coase’sargument applies to public goods for large numbers as well as to local exter-nalities. In addition, using his own earlier theory of the firm, Coase argued inSection VI of his article (Coase, 1960) that economic activity will be carried outby whatever means, market or non-market, that minimizes total costs: that is,production plus transaction costs. In short, the Pigouvian argument that govern-ment is needed to use taxes and subsidies to internalize externalities wasfundamentally unsatisfactory; even in the presence of externalities and publicgoods, the rational bargaining of the parties in the economy would bring efficiencywithout any governmental intervention.1.1.The narrow and the broad theoremCoase did not claim he had offered a theorem, but George Stigler and legions ofother economists have asserted that he had. Therefore, they attribute to him adeductive result that is, within its domain of application, necessarily anduniversally true. Though in some formulations of the theorem there are also otherclaims, the most basic claim of what has come to be called the Coase Theorem isthat only transaction (or bargaining) costs can prevent voluntary bargaining fromattaining Pareto-efficient outcomes. The theorem can be fairly stated as follows: ‘If1According to the Social Science Citation Index volumes since 1972, even Milton Friedman and PaulSamuelson do not have a single publication that has been cited even half as often as ‘The Problem ofSocial Cost.’ This article was also, by a huge margin, the most widely ordered article in theBobbs–Merrill Reprint Series in Economics.2The initial distribution of legal rights affects the distribution of income, and, thus, the outcomes maybe different because of income effects.A.Dixit,M.Olson / Journal of Public Economics76 (2000) 309–335311transaction costs are zero, rational parties will necessarily achieve a Pareto-efficient allocation through voluntary transactions or bargaining.’Different economists define transaction costs differently, but all agree that theresources devoted to transactions have alternative uses and, thus, an opportunitycost. Therefore, transaction costs must be taken into account in defining the Paretofrontier. When this point is used along with a comprehensive definition oftransaction costs, the Coase Theorem can easily be transformed into an evengrander proposition. If the familiar Coase theorem is true, it must also be true thatrational parties in an economy will make all those trades in private goods, and allthose bargains to internalize externalities, provide public goods, and deal with anyother potential market failures, that bring positive net gains — that is, gainsgreater than the transaction costs needed to realize them. They will not make thosedeals that cost more to make than they are worth, and obviously Pareto efficiencyrequires that such deals should not be made. Thus, if the Coase Theorem is true, sois a ‘super Coase Theorem,’ namely that ‘rational parties will necessarily achieve aPareto-efficient allocation through voluntary transactions or bargaining, no matterhow high transaction costs might be.’When transaction costs are important, so is the transaction technology. There areobvious incentives to come up with innovations that reduce transaction costs. Theabove argument then extends to say that the most cost-effective methods ofreducing transaction
View Full Document
Unlocking...