Utility Regulation
Prediction markets sometimes might be useful not only in conjunction with permit trading systems but also in conjunction with other market-oriented devices such as auctions. Consider, for example, Harold Demsetz’s proposal to use an auction as an alternative to traditional regulation of utilities such as electricity or cable television providers. Traditionally, utilities have been regulated on the theory that they are natural monopolies, that is, that economies of scale in the industry are so great that inevitably there will be only a single producer. Because prices are higher and quantity is lower in a monopoly than with various forms of competition, the government regulates the prices that natural monopolists can charge and the services that they provide.33 Demsetz suggested that instead the government might auction the right to serve as the natural monopolist for some period of time.34 The winner of the auction would be the bidder committing to the best overall deal to society. For example, firms might compete for the right to offer service at the lowest price, with the lowest bid winning the auction. Demsetzian auctions have not become widespread, and one possible problem is the difficulty in assessing bids.35 If price were the only variable in bids, then the winning bidder would charge a very low price but offer very low quality. Thus, for the mechanism to work well, there must be some means of assessing the quality of services that bidders will provide. One approach, of course, is simply to rely on the government to assess the proposals. The purpose of the Demsetzian auction, however, is to limit the government’s role. One might worry that the government would not do an adequate job of predicting quality and that competing bidders would have incentives to spend a great deal of money to persuade the government that they will offer high-quality service (for example, by wining and dining lawmakers), without necessarily actually providing such service. An alternative approach is to use a conditional prediction market to gauge how satisfied consumers will be with different proposed service providers. For example, it might be used to predict consumer surplus in future years, and those numbers could then be discounted to present values. The consumer surplus from a transaction is the difference between the maximum that a consumer would pay for the service and the amount that the consumer actually pays. For the firm selected, this might be established ex post by conducting surveys of consumers. Surveys might not produce perfect predictions because consumers might not answer correctly. A consumer might really prefer to pay ten thousand dollars a year rather than not have electricity yet give an answer of two thousand dollars, biased in the direction of the amount that consumers actually pay. Nonetheless, consumers should generally report higher valuations for utilities that succeed in offering higher quality services, and so the relative forecasts from one prediction market to another might be meaningful. Alternatively, a governmental official might make the evaluation independent of consumer surveys. Because this evaluation would affect only the prediction market participants, the franchisee would have little incentive to try to influence the official making the ex post decision. This is a critical distinction between the predictive decision making proposal and existing approaches to rate regulation. In existing approaches, the regulated entity has every incentive to seek to capture the governmental agency, and short of that, to spend a great deal of money trying to persuade the agency to allow it to charge more to consumers. In the predictive proposal, the eventual decision by the agency would not affect the regulated entity, and so the only parties with a direct incentive to seek to influence the decision would be prediction market participants. At least these participants would have varying interests, however, and the overall stakes would be so much lower that the amount spent would be much less than in the ratemaking context. It might appear that Demsetz’s auction proposal in general and this refinement in particular fail to recognize that a prediction of performance cannot substitute for active monitoring. The government’s role, the argument goes, is not merely to select as natural monopolist the company with the best proposal but to ensure that the natural monopolist meets its commitments. Otherwise, the auction winner inevitably would shirk on its promises, and no party would be in a position to enforce them. Yet proposals from prospective utility regulators could include enforcement devices that establish precommitment. For example, one prospective utility operator might authorize consumers to sue to enforce rights that the proposal would specify. Another might promise to pay a third party some set amount of money if quality, as established by independent surveys, falls below a certain level.36 Yet another might agree to government-conducted (or perhaps third-party-conducted) price regulation. Prediction market participants would have an incentive to anticipate the degree to which such commitments would succeed in improving quality, and prospective utility operators would need to identify the commitment approaches that can achieve quality at the lowest possible price.
Leave a Reply