Toward a Pigouvian state.

• Author: Masur, Jonathan S.
• Position: Introduction through II. Agency Authority and the Absence of Pigouvian Taxes B. The Clean Water Act 1. Existing Point Sources, p. 93-122

Most economists believe that the government should impose Pigouvian taxes on firms that produce negative externalities like pollution, yet regulatory agencies hardly ever use their authority to create Pigouvian taxes. Instead, they issue command-and-control regulations. Our major point is that, contrary to the conventional wisdom, regulators typically have legal authority to create Pigouvian taxes--they just do not use it. While regulators may hesitate to impose Pigouvian taxes for a range of political and symbolic reasons, we argue that these reasons do not justify this massive failure of regulatory efficiency. It is time for the regulatory state to take a Pigouvian turn.

INTRODUCTION I. PIGOUVIAN TAXES IN THEORY AND PRACTICE A. The Theory of Pigouvian Taxes B. Existing Pigouvian Taxes II. AGENCY AUTHORITY AND THE ABSENCE OF PIGOUVIAN TAXES A. The Clean Air Act 1. National Ambient Air Quality Standards 2. Regulation of Stationary Sources 3. Hazardous Air Pollutants 4. Regulating Through Taxation B. The Clean Water Act 1. Existing Point Sources 2. New Sources 3. Toxic Pollutants C. Financial Regulation 1. Negative Externalities in Finance a. The Risk of a Panic b. Races to Information c. Speculation on Asset Prices 2. The Legal Regime a. Capital Regulation and Risks of Panic b. Regulation of High-Speed Trading and Information Races. c. Regulation of Financial Instruments in Order to Deter Speculation D. The Occupational Safety and Health Act III. OBJECTIONS AND OBSTACLES A. Administrative Problems B. Interest Group Explanations C. Negative Symbolism & Ideology D. Legal and Conceptual Hurdles E. Risk-Averse Regulators CONCLUSION INTRODUCTION

According to most economists, the optimal form of regulation of firms that produce negative externalities is a tax known as a Pigouvian tax, named after the English economist, Arthur Pigou. (1) A Pigouvian tax is a tax equal to the harm that the firm imposes on third parties. For example, if a manufacturer pollutes, and the pollution causes a harm of $100 per unit of pollution to people who live in the area, then the firm should pay a tax of $100 per unit of pollution. This ensures that the manufacturer pollutes only if the value of the pollution-generating activities exceeds the harm, such that the social value of those activities is positive.

Other forms of regulation are inferior to the Pigouvian tax. Consider command-and-control regulation, in which a regulator forces a firm to take a particular action, such as installing a pollution-reducing scrubber. (2) Under this form of regulation, the regulator may conduct a cost-benefit analysis to determine whether the benefit of alleviating the pollution for the firm's neighbors is greater or less than the cost to the firm of having to install the scrubbers or take other precautions. (3) If scrubbers pass a cost-benefit analysis, then the regulator orders the firm to install them. If they do not pass, the regulator allows the firm to continue its activity unabated. A perfectly conducted cost-benefit analysis should produce results as efficient as a Pigouvian tax, but in a world of administrative costs, command-and-control regulation will be inferior. The reason is that in order to determine the correct command-and-control rule, the regulator must know both the cost and benefit of the activities. (4) In contrast, the regulator only needs to know the cost of the activity to determine the correct Pigouvian tax. It is not necessary to know the benefit. Thus, as long as regulators make errors (as they unavoidably do), a Pigouvian tax is superior to command-and-control regulation.

It would be an understatement to say that economists endorse Pigouvian taxes over command-and-control regulation. Pigouvian taxes are constantly advocated by economists who seek to influence public policy. (5) Professor Greg Mankiw, a prominent Harvard economist and chairman of the Council of Economic Advisers under President George W. Bush, invited numerous public figures to join his "Pigou Club," which advocates a Pigouvian tax on gasoline. (6) Club members include several Nobel laureates with diverse views, including Gary Becker, Paul Krugman, and other prominent journalists, scientists, and politicians. (7)

Yet, turning from the scholarly literature to government practice, one discovers that Pigouvian taxes are used rarely by Congress and almost never by regulators, at least in a self-conscious way. (8) There is no political support for a Pigouvian tax on gasoline. (9) And while gasoline taxes do exist, they do not appear to be based on Pigouvian theory; (10) they are not calculated on the basis of an assessment of the social costs of gasoline-powered driving, and they are much too low. (11) As far as we have been able to discover, the Environmental Protection Agency (EPA) has never ordered a Pigouvian tax. Nor has any other agency. We have been able to find only a few isolated examples of a pure Pigouvian tax in U.S. law. (12) Indeed, we located only twenty references to Arthur Pigou or Pigouvian taxes in the entire history of the Congressional Record. (13) All but one of these references were general comments on Pigou's work or theories. The lone Pigouvian reference directly involving pending legislation related to a bill that eventually became the Noise Control Act of 1972. (14) That law provides EPA with authority to impose--wait for it--command-and-control regulations governing aircraft noise pollution.

There are laws and regulations that could be rationalized on Pigouvian grounds, and that probably reflect some of the economic theory that underlies the Pigouvian tax. As we noted, gasoline taxes do exist, and while they are not Pigouvian in the sense of being equated to the social cost of driving, they obviously do deter excessive driving on the margins. One can also find examples of usage fees, congestion pricing, and the like, but these are found in contractual settings that are distinguishable from the problem of negative externalities that Pigouvian taxes seek to correct. (15) Tradable-permit regimes also reflect a kind of Pigouvian thinking, and are superior to command-and-control regulation, but as we discuss in Part I, they are inferior to Pigouvian taxation in the pure sense. Moreover, essentially all of the examples we have found of rules that resemble Pigouvian taxes derive from legislation, rather than administrative regulation. So the existence of some laws and regulations that bear a family resemblance to Pigouvian taxation, and reflect some of the economic thinking that motivates Pigouvian taxation, reinforces rather than solves the puzzle of why regulators never, or very rarely, use Pigouvian taxes.

In this Article, we attempt to solve the puzzle. We suggest that the principal reason regulators do not employ Pigouvian taxes is that they do not believe they have the authority to do so under existing law. We then demonstrate that regulators' pessimism is misplaced. Across a variety of regulatory areas covering a vast swath of economic activity, existing regulatory statutes provide regulators with at least plausible authority to use Pigouvian taxes in regulation. If we are correct, this would not be the first time that regulators have discovered such authority where conventional wisdom held that none existed. The Obama Administration's 2014 proposal that states regulate greenhouse gases using cap-and-trade was spurred in part by a Natural Resources Defense Council argument that EPA possessed such authority under section 111(d) of the Clean Air Act. (16) One objective of this Article is to map out similar arguments for Pigouvian taxes across a range of regulatory areas and statutes.

Having concluded that administrative agencies likely have authority to regulate using Pigouvian taxes, we next consider whether there is any compelling reason why they should not or will not be capable of doing so. We canvas five potential objections or hurdles. First, Pigouvian taxes do not solve a significant information problem, which is how the regulator values the harm caused by economic activity. This problem is compounded by the second-best nature of regulation: gasoline taxes, for example, may be inefficient if cars are already over- or adequately regulated.

Second, Pigouvian taxes may lack political support because they do not serve the interests of those with political power. Suppose that laws and regulations typically reflect interest-group compromises, and not the general interest of the public because it is too costly for the public to organize. If so, one would expect regulations that reflect the interests of those groups--industry, unions, and so on. Some types of regulation produce natural interest-group constituencies: the firms that will produce the scrubbers or other technologies that the regulation will mandate. Command-and-control regulation will also sometimes serve the interests of regulated parties by serving as a high barrier to new market entrants. Pigouvian taxes possess neither of these features, at least not to the same degree.

Third, Pigouvian taxes have negative symbolic resonance. For the right, they are unattractive because they are "taxes," which people on the right oppose. For the left, they are unattractive because they seem to put a price on intrinsically valuable goods like human life and the environment, and because they seem to permit a firm to commit ongoing harm so long as it is willing to pay a fee. In contrast, tradable permits are more attractive to the right because they seem to create markets, while command-and-control regulation is more attractive to the left because it seems to avoid pricing intrinsically valuable goods and bans harmful activity outright.

Fourth, Pigouvian taxes breach the divide between taxation and regulation, which is firmly entrenched in the institutional organization of the U.S. government. We suspect that in the minds of many government officials, only Congress can...