The Invisible Hand and Externalities
E. Maskin
Abstract:When economists contemplate the invisible hand at work, they generally think of competitive markets. But there are some circumstances in which markets are not supposed to operate well (i.e., in which the invisible hand is thought to falter). A leading cause of market failure, many argue, is the presence of significant externalities. With such externalities, the first welfare theorem does not apply, and so competitive equilibrium-if it exists at all-is not typically Pareto optimal. In the tradition of A. C. Pigou (1932), the typical response to this lack of optimality is for the government to step in and introduce corrective policy, usually in the form of taxes or subsidies. There is, of course, a strong antiPigouvian tradition, as well. Specifically, proponents of the Coase theorem (Ronald Coase, 1960) have contended that, despite externalities, unrestrained bargaining and contracting ought to be sufficient to generate an efficient outcome. (Indeed Coase's own celebrated example was a case of externalities.) Thus, even if formal markets themselves fail, the invisible hand nevertheless succeeds, and outside intervention or design is not required. Recently, Joseph Farrell (1987) argued that, even when free bargaining is permitted, the laissez-faire conclusion inherent in the Coase theorem may founder if agents have incomplete information about one another's relevant characteristics. I shall show, however, that the problem that Farrell identified is due only to monopoly power and is not peculiar to externalities. Indeed, in this paper, I shall take a modified Coasian stance. I shall attempt to show that, in spite of externalities and incomplete information, private contractual agreements suffice to achieve efficiency, as long as no agent is big enough to have significant market power. This conclusion must be qualified, however, with the proviso that, if the externality is "nonexcludable" (i.e., no one can be excluded from its effects), the government must intervene to prevent free-riding on the agreements. Intervention, in this case, amounts to establishing the right of an agent providing a positive external effect to collect a fee for increasing the effect from all who enjoy it, even if they are not parties to a contract with the provider. Symmetrically, providers of a negative externality can collect a fee for diminishing the effect from all their victims. In either case, however, the fee is set endogenously, that is, it is determined by the contractual arrangements rather than by the government. This result for nonexcludable externalities (which include pure public goods) provides support for a fairly laissez-faire stance toward externalities but turns on an important assumption, namely, that parties always write contracts so as to maximize their social surplus (subject to incentive and individual-rationality constraints). I will return to this assumption in the final section.
Economics