Abstract

The study of government regulation has emerged as a distinct field of policy analysis in recent years, yet it remains an open question whether “regulation” describes a distinct set of government policies—policies with features that distinguish them clearly from other policies.  This chapter considers several different conceptions of regulation that have appeared in recent debate and scholarship and advances the view that regulations should be taken as the prescription by government of terms of private transactions.  The arguments for this view are that it is more comprehensive and exclusive than alternative views; that it describes regulation in a neutral, functional way, free of insinuations about the purposes or consequences of regulation; and that it defines the limit of the ability of regulation either to redistribute income or to improve the efficiency of markets.

The study of government regulation has emerged as a distinct field of policy analysis.  Beginning in the mid-1970s, leading universities established programs of regulatory studies, new professional journals devoted to regulation appeared, and several books on the politics and economics of regulation were published (Stigler, 1975; Owen & Braeutigam, 1978; MacAvoy, 1979; Weidenbaun, 1979; Mitnick, 1980; Wilson, 1980; Breyer, 1981).  Presidents Ford, Carter, and Reagan began a tradition of appointing experts on regulation to the Council of Economic Advisors and established several new agencies to evaluate federal regulatory policies.

Can the subject of all this study and debate by distinguished clearly from other endeavors of government?  The word “regulation” brings to mind the various federal “alphabet agencies” and independent commissions, such as the ICC, FTC, SEC, EPA, OSHA, and so on, and the similar state agencies, such as insurance and public utility commissions.  What these agencies primarily do is set prices, terms of service, and quality standards for particular firms and products.  But if these activities are regulation, then we cannot stop with the alphabet agencies and commissions.  We should also include the government’s sporadic efforts to establish wages and prices by statute rather than through administrative agencies, as in the federal minimum wage requirement and municipal rent controls.

There are many other things we sometimes mean by “regulation.”  All organizations, including government agencies, have “rules and regulations” regarding their own operations.  The agencies administering the government’s expenditure programs such as Medicare, Social Security, and housing and educational subsidies have thousands of regulations concerning eligibility, reimbursement, accounting procedures, and much else, which are published in the Federal Register alongside the regulations of the Environmental Protection Agency (EPA) and the Occupational Safety and Health Administration (OSHA).  The Internal Revenue Service (IRS) is also a prodigious author of regulations.  Taxing and spending programs themselves are often described as regulatory, meaning that they alter the allocation of resources in the economy from what it would be otherwise, and legal scholars increasingly characterize common law rules as regulatory with the same thought in mind.  If we add up all these usages, “regulation” means all of law viewed instrumentally.

While the meaning of the word is often clear enough in context, it is nevertheless a matter of practical interest to set more precise bounds on the idea of regulation.  The automobile executive who complains that his industry is a victim of “overregulation” and goes on to demand import quotas in the next breath confuses only the economists among his listeners.  But President Reagan could not be so casual once he had been elected.  The “regulatory freeze” he ordered early in his administration raised a host of questions concerning scope and application and required decisions whether to exempt such regulations as IRS Revenue Rulings, Security and Exchange Commission (SEC) enforcement actions, antitrust guidelines, notices of Federal Reserve Board actions, Department of the Interior rules for leasing federal lands, and amendments to guidelines for various grant-in-aid programs.  On what logical grounds could these rules be exempted from the freeze (as they were) while those of EPA were not?  That the exemptions were reportedly decided with little controversy or formality is evidence that there is a particular set of government activities sharing important features that distinguish them from the rest.

This chapter advances the idea that government regulation should be taken as the prescription by government of the terms of private transactions.  I attempt to distill this view from the most interesting conceptions of regulation explicit or implicit in recent scholarship.  In sections II and III, I discuss two narrower views of regulation—the “public utility” approach and the “regulatory reform” approach—and argue that they are both too narrow and, in certain respects, too vague to serve as durable definitions of the subject.  In section IV, I take up the question whether regulation can usefully be confined to something less than all of government action, and I draw on distinctions among types of government action suggested by several students of politics and economics.  In section V, I conclude with a number of affirmative arguments for the view I suggest.

The Public Utility Approach

The traditional “public utility” approach to regulation is well exemplified by Alfred E. Kahn’s The Economics of Regulation (1970), which is still widely used in colleges and graduate schools.  Kahn treats regulation as meaning government control of public utilities and common carriers—control of entry, prices, and exit in the major infrastructure industries of power, communications, and transportation.  He ignores the areas of health, safety, consumer protection, and environmental regulation, although regulation for such areas was conspicuous even in 1970.

Kahn acknowledges at the outset that government influences the operation of private markets in many ways other than utility regulation, e.g., by regulating the money supply, enforcing contracts and property rights, providing subsidies and tariffs, and imposing product and packaging standards.  He argues, however, that in these cases government is simply “maintaining the institutions within whose framework the free market can continue to function” (Kahn’s emphasis).  By contrast, in the “regulated sector” (i.e., the utility and common carrier sectors), government is supplanting the market itself by directing what should be produced, by whom, and at what price (Kahn 1970:1,2).

Kahn seems to have been clearing the deck in these passages, describing the scope of his book rather than propounding a rigorous distinction.  His distinction between “influencing” and “supplanting” markets is, in any event, untenable.  Public utility programs certainly do not supplant markets entirely.  Even the ideal utility or common carrier commission described in textbooks merely limits the total revenues (or rates of return) regulated firms may earn and resolves controversies over particular rates, services, and investments.  Innumerable other decisions remain market decisions, left to the discretion of the firms’ managers, investors, suppliers, customers, and competitors.  Many such decisions are beyond the legal authority of even the most comprehensively endowed commission.  For example, except in the case of automobile liability insurance, consumers cannot be forced to purchase any particular goods or services under any regulatory regime in the United States.  And with or without legal authority, many important matters will always be arranged privately in a manner satisfactory to all concerned (or at least in a manner not unsatisfactory enough to inspire a complaint to a commission), and many will simply be immune to effective administrative control.  In practice, moreover, utility and common carrier commissions exercise only haphazard control even over total revenues (Joskow, 1974).

At the same time, many non-utility policies mentioned by Kahn, such as product quality standards and monetary controls, do supplant markets to some extent—they dictate certain aspects of market transactions or forbid certain kinds of transactions.  In many cases—drug regulation, for instance—they manifestly amount to deciding what should be produced and by whom.  While utility controls have in common the setting of prices, non-utility controls frequently set prices too, as in SEC regulation of brokerage commissions, the minimum wage requirement, and state regulation of agricultural goods and professional services.  Only some of the government policies mentioned by Kahn, such as taxes and subsidies and common law doctrines, can properly be characterized as merely influencing private markets, insofar as they stop short of prescribing the terms of private transactions; and these are precisely the kinds of policies lease likely to be described as “regulation.”

Kahn’s is the best of the traditional regulatory economics textbooks, and probably the last.  Just as it was being published, many of the framework regulatory programs began to grow rapidly in mission and political importance and soon were housed in several new federal agencies of their own, such as OSHA (1970), EPA (1971), and the Consumer Product Safety Commission (CPSC) (1972).

These programs were at first considered a distinct variety of policy—“social regulation” rather than “economic regulation.”  The distinction seemed rhetorically useful at first.  It was said that business favored economic regulation but despised social regulation, because the former restricted competition whereas the latter imposed costly social obligations, while, for the same reasons, consumer and “public interest” groups favored social regulation but despised economic regulation.  But no sooner was the distinction made in these terms than it began to melt away, for it became evident that consumer groups sometimes profited from economic regulation, as in the case of natural gas (Breyer and MacAvoy, 1973), and that producer groups sometimes profited from social regulation, as in the case of product standardization (Cornell, Noll & Weingast, 1976; Leland, 1979; Thomas, 1979).  In the mid-1970s, as price controls were extended to all levels of the petroleum industry—apparently at the expense of most sectors of the industry—and then became increasingly tangled in the administration of pollution and automobile safety controls, it was difficult even to remember what the differences between economic and social regulation had been.  Today it seems obvious that the distinction was a confusion from the start: railroad rates were as much a social issue as was meat quality a century ago, and the newest regulatory policies government medical care, such as requirements for “certificate of need” and price regulations, are reinventions of public utility controls.

The Regulatory Reform Approach

The regulatory reformers have a very different conception of regulation, one which treats regulation as a method of government administration rather than as a cluster of programs applied in certain markets or in response to certain problems.  By “regulatory reformers” I mean those scholars and government officials, mainly economists, who argue that public objectives should usually be pursued through economic incentives rather than through government command and control.  In the usage of most regulatory reformers, command-and-controls policies are regulation, and reform consists of replacing them with other policies that alter the economic incentives facing organizations and individuals in private markets.  Other reformers speak of economic-incentive policies as “regulatory alternatives,” but their meaning is identical: the command an control typifies current regulatory practice and ought to be minimized in favor of economic alternatives (United State Regulatory Council, 1980a, 1980c).

The policy recommendations of the regulatory reformers follow a rough division between economic regulation and social regulation; they usually recommend deregulation (abolition of regulatory programs) for utilities and common carriers, but only reform (the infusion of economic-incentive techniques) for programs of health, safety, and environmental regulation (Schultze, 1977; Joskow and Noll, 1978; Breyer, 1981).  The difference arises, however, not because they see any fundamental difference between social and economic regulation, but rather because their approach is to identify market failures justifying government intervention and then to recommend the most economically efficient corrective intervention.  Most programs of utility and common carrier regulation have been applied in cases where the market-failure argument—that monopoly is natural—is spurious (e.g., in the case of truck transportation), and where the appropriate reform is therefore to eliminate the government intervention itself.  On the other hand, in cases where a market failure appears real and serious—e.g., in plausible natural monopolies (electricity distribution) as well as in other sorts of failures—intervention is appropriate but should correct he failure as efficiently as possible.  In both cases the regulatory reformer prescribes maximum reliance on economic incentives and market forces.  But this reliance actually involves no intervention at all in the case of spurious natural monopolies, and only some intervention, of a less “regulatory” sort, in the case of true natural monopolies and other market failures. 

A prominent example of the regulatory reform approach is Charles L. Schultze’s The Public Use of Private Interest (1977), which has become well known for its advocacy of greater reliance on private incentives and of the price known for its advocacy of greater reliance on private incentives and of the price system in federal policymaking.  Schultze’s advocacy is encased in a larger and more descriptive argument, which is briefly this: (a) Our libertarian political tradition in the United States has created a “rebuttable presumption” against government intervention in private markets, and as a result we usually intervene only when there is wide agreement that shortcomings in certain markets need correction.  (b) Paradoxically, our libertarian tradition has also created the rule that government, when it does intervene, may “do no direct harm” to any identifiable individual or group, and as a result our government policies are less automatic and private-market oriented than they could and should be, since they rely too heavily on case-by-case decisionmaking.

The nature of the distinction between actual and desirable policies is suggested in the book’s opening passages (1977:5-6, 13):

There is a growing need for collective influence over individual and business behavior that once was the domain of purely private decisions.  But as a society we are going about the job in a systematically bad way…We usually tend to see only one way of intervening—namely, removing a set of decisions from the decentralized and incentive-oriented private market and transferring them to the command-and-control techniques of government bureaucracy…

 

…Once the decision to intervene has been taken, there remains a critical choice to be made: should intervention be carried out by grafting a specific command-and-control module—a regulatory apparatus—onto the system of incentive oriented private enterprise, or by modifying the informational flow, institutional structure, or incentive pattern of that private system.  Neither approach is appropriate to every situation.  But our political system almost always chooses the command-and-control response.

Here and throughout the book Schultze neglects to say just what command-and-control techniques are.  Perhaps he thought a definition would seem pedantic: the book abounds with examples, and the term “command and control,” borrowed from military parlance, is highly suggestive.  In particular it suggests centralized regimentation, the attempt to move society in a particular direction simply by ordering people to march that way—policy that is all stick and no carrot.  The difficulty with this imagery is that, where government is concerned, there is no difference between sticks and carrots.  The essence of government is command, the use of coercion to change behavior.  Virtually any act of state can be characterized in this way, and all such acts alter the incentive pattern of private markets—inevitably, and often deliberately. 

Consider Schultze’s (and other regulatory reformers’) favorite example of the superiority of economic-incentive techniques over regulatory command and control: the emission fee.  His argument, now widely familiar, is that charging a fee based on the level of pollutant emissions would permit firms to respond differentially according to their differing marginal costs of abatement.  By contrast, the uniform emission standard (the predominant approach in practice) requires all firms to abate in the same degree regardless of their abatement costs—some too much and some too little to minimize the total costs of a given amount of abatement.  Variable emission standards might set abatement requirements according to costs but would place enormous information-gathering burdens on regulatory agencies, burdens that can be avoided, or at least delegated to the market, with an emission fee.

The argument is persuasive as far as it goes, but why wouldn’t a program of emission fees actually be a “regulatory apparatus” grafted onto incentive-oriented private enterprise?  There is certainly no inherent logic in calling an emission fee an economic-incentive policy and an emission standard a command-and-control policy.  A standard affects incentives and a fee is a command made in order to control, and both require administration and enforcement at some level of government.  Indeed, an emission standard, enforced by a civil fine that varies with the degree of violation and not with the abatement costs of individual firms, is economically equivalent to an emission fee with a “deductible” in the amount of the standard.  A revealing irony is that one of the most touted achievements of regulatory reformers has been the calibration of pollution-control enforcement fines according to the abatement costs of individual firms—that is, according to firms’ differential savings from noncompliance with emission standards.  While this policy has been described as the substitution of economic incentives for command and control, it is rather conspicuously a move in the opposite directions: its purpose is to achieve uniform compliance with administratively determined abatement standards regardless of differential abatement costs.

Schultze elaborates his distinction with the argument that economic-incentive policies are “process-oriented, seeking to correct the faulty process,” while command-and-control policies are “output-oriented, seeking to bypass the process and determine outputs directly” (1977:29, 65-66, 74).  At times, the difference appears to be that “process-oriented” policies like taxes on cigarettes and gambling aim only to moderate the general tendency or degree of disapproved private behavior rather than to achieve “any desired set of water-quality standards” (1977: 35, 53-54).  At other times, however, he undermines the distinction by suggesting that economic incentive policies actually “correct” market failures by incorporating social considerations into the price system, and that “one of the major efficiency gains from use of a price system is precisely its ability to induce individuals and firms to balance costs against gains…The trick is to make sure that the costs and gains they confront also reflect, as far as possible, true social costs and gains” (Schultze’s emphasis; 1977:81).  An emission fee calibrated according to the marginal social benefit of pollution abatement might or might not be “stiff.”  It certainly would not be merely stiff, nor would it be adjustable to obtain “any desired” level of air or water quality.  A fee calculated in this manner most emphatically would be “output-oriented,” since its whole purpose would be to maximize total economic output, not simply to minimize the total costs of achieving some degree of abatement that might or might not be optimal.  It would also supersede the market with an administrative order: the fee would necessarily be set by administrative procedures rather than by the price system, since the absence of a market generating true supply-and-demand-based prices would be the very problem being addressed.

A more complete example of the use of the price system to avoid administrative command and control is the institution of private, marketable property rights to some naturally occurring resource.  This would have happened in the 1920s if the government had established a system of property rights in the electromagnetic spectrum rather than an administrative allocation of broadcast frequencies (Coase, 1959).  Such as system might have developed spontaneously, as in the case of land, without any government action other than provision for a public filing system and for judicial resolution of conflicting claims.  Or the government might have coaxed things along by designing and running an initial auction.  But even this is a false example of an economic-incentive alternative to regulation, since the solution (an effective market) is simply the absence of the problem (a failing or nonexistent market).  Schultze recognizes this in noting that the economic case for government intervention, and the dilemma of choosing between command-and-control and economic-incentive intervention, arises only when the institution or improvement of property rights to resolve a market failure is for some reason infeasible (1977:29-32).

It seems impossible to erect any unique conception of regulation upon generalizations about command and control or output orientation or the efficiency of emission fees.  Indeed, in the context of Schultze’s larger argument about the policy consequences of the American political tradition—that our distrust of government has led us to hobble public policies with excessive legal protections of individual interests—the contrast between regulation and economic incentives pales to a contrast between the actual and the (economic) ideal.  At the end of his book, he concludes that what we need is not economic-incentive techniques themselves, but rather the political maturity and economic sophistication to let them work their magic (1977:76-90).  In the meantime, presumably, even the most elegantly conceived programs for harnessing private incentives to public purposes will be disfigured by judicial doctrines and political pressures until they become mere regulatory programs, marred by absolute-sounding standards, detailed exemptions, special conditions, deadlines, precedent, and waste.  Notably, when Schultze himself was chairman of the Council of Economic Advisors and architect of President Carter’s regulatory reform program, he devoted himself to applying economic analysis within the existing structure of regulatory policy—attempting to conform regulatory decisionmaking to cost-benefit analysis—rather than to replacing existing policies with economic incentives (DeMuth, 1980).

Schultze is no doubt correct that a policy proposal aimed at adjusting marginal economic incentives to that “costs and gains reflect true social costs and gains” would be molded by the political process until it departed substantially from its original conception.  The interesting question is whether the results would significantly differ from existing regulatory policies.  Policies in the form of taxes and expenditures involve different political institutions than policies in the form of standards.  An emission fee in the form of a flat national tax would have to be initiated by the House of Representatives and would have no success without the support of the Treasury Department.  Such a tax would fall under the jurisdiction of several congressional committees, composed of representatives with interests and constituencies different from those which influenced the writing and administration of the present environmental laws.  Because the tax would be a statute rather than an administrative order, it could not be thrown out by a court for lack of evidence or an incomplete hearing record.  On the other hand, an emission fee scheme that required greater administrative supervision and held little potential as a revenue raiser, such as one whose fee schedules varied according to different marginal damages in different locales, might evolve out of the current EPA program of administrative fines and might require only minor tempering to gain the approval of congressional finance committees and the courts.  Political contingencies are surely different in different cases, and if they are systematically different, approaching problems in certain ways may produce certain distinctive results.

Roger G. Noll, in a recent paper whose title I have borrowed (1980), argues explicitly that institutional forms make a difference.  He begins where Schultze leaves off, by noting that the need to define regulation as a distinct form of policy arose when reform-minded economists began to analyze the effects of “different institutional approaches to the same policy objective” in order to recommend the most economic approach.  In an earlier paper he wrote with Alain Enthoven on the problem of rising medical care expenditures, Noll argued that altering “the basic financial incentives facing [medical care] providers” was preferable to “imposing economic and technical regulation on providers” in an attempt to defeat their (perverse) financial incentives.  To make his case Noll had to explain how “economic and technical regulation” differed from what he was proposing (Enthoven & Noll, 1979:215):

Regulation refers to a type of social control of transactions that is characterized by its procedures as well as by the substantive purpose of the regulation.  The two key characteristics of regulation are as follows.  First, the regulatory authority is not a party of the transactions it regulates.  Instead, it acts as the referee of transactions between other parties.  By contrast, eligibility requirements and cost reimbursement formulas for Medicare and Medicaid recipients are not, in this sense, regulations because they are written by the purchaser of the service.  These controls are more properly regarded as terms of a contract between a purchaser and a vendor…Second, regulation is operated according to procedural rules that were developed from case law and formalized after the fact in the Administrative Procedure Act of 1946.  The most important features of these rules are that decisions must by based on evidence that is presented in formal proceedings, that substantial evidence must be submitted in support of each decision, and that the courts may review a decision if it is appealed by a participant in the regulatory proceeding.

The first prong of this definition may seem a little odd at first, until one realizes that may of the things we customarily call regulations are really only terms of voluntary transactions.  Regulations concerning military dress, teaching loads in universities, and vacations and promotions in private firms and government agencies are all terms of contracts between organizations and employees, just as Medicare reimbursement regulations are terms of contracts between government and medical providers, and just as procurement regulations are terms of contracts between government and suppliers.  Regulations such as these pertain sometimes to an organization’s internal management and sometimes to its dealings with the outside worlds, but as a class their function is to define an organization’s own purposes and manner of operation.  They might, therefore, be called “organizational regulations.”  Distinguishing them from “government regulation” was important in the Enthoven-Noll paper because their alternative to “regulating” medical care costs was to revise the terms of government payments under Medicare, Medicaid, and other programs so as to strengthen the incentives of doctors and hospital administrators to minimize their costs (1979:221-223).  The concept of “command and control” would not have helped to distinguish regulation from what they were proposing, but the distinction between a referee and a contracting party did.

A more direct approach would be to say that the medical care reimbursement provisions are aspects of public finance, which taxes some activities and subsidizes others.  Public finance also covers the other nonregulatory proposal in the Enthoven-Noll paper—to revise the tax deductions for medical care and medical insurance expenditures to give consumers greater incentives to minimize medical costs—and it puts both in the same category as other economic-incentive proposals such as emission fees.  Of course, all tax and subsidy programs necessarily involve terms and conditions: even farm subsidies are not just showered over the fields of Kansas.  But farm subsidies are surely different in kind from the terms of the government’s “organizational regulations”—from the terms of a contract between the government and its clients when the government is purchasing or selling goods and services to operate weather stations, naval bases, or medical insurance programs.

Most people would consider it euphemistic to call the government a “contractor” when it is taxing and subsidizing, or to call those who are taxed and subsidized “clients.” Even those who assume that all government activities are determined by the pressures of private economic interests would almost always distinguish activities in political markets from those in conventional economic markets: the provision of additional government jobs, for example, has probably been an insignificant motivating factor in the growth of Medicare and Medicaid.  Moreover, if we wish (with Enthoven and Noll) to distinguish public policies according to institutions and procedures, then it is surely important to recognize that taxing and spending programs are largely the product of legislative action, while the government’s organization regulations, such as civil service rules, are largely the product of administrative action.

If we put organizational regulations aside and distinguish other government actions as either “regulation” or “taxing and spending” (public finance), we can then dispense with the second prong of the Enthoven-Noll definition, the application of procedures of administrative law.  While there are exceptions, the Administrative Procedure Act (APA) generally does not apply to taxing and spending programs.  Indeed, administrative law may be described as a set of restraints that has grown up because government sometimes acts outside the restraints of public finance and yet acts purposively—not as a referee in the manner of a court.  Moreover, administrative law does not apply at all to policies involving neither taxing and spending nor administrative action other than prosecutorial discretion.  Especially at the state and local level, statutes themselves often prescribe the terms of specified private transaction; examples are building codes, rent control, drug substitution laws, and California’s law entitling artists to a 5 percent commission on resales of their works.  Even federal statutes, such as the Fair Labor Standards Act and the Clean Air Act, may stipulate ultimate regulatory standards, leaving little to the discretion of administrative agencies (Ackerman & Hassler, 1981).  To say that regulation is policy developed according to the APA is incomplete and obscures the interesting question why legislatures sometimes dictate regulatory standards and sometimes delegate the task to administrative agencies. 

Noll’s paper (1980) suggests that administrative procedures may determine the content of policy in a predictable fashion; his examples, however, are very brief and tentative.  He notes that the regulations of the Consumer Product Safety Commission have occasionally been baldly anticompetitive, presumably because of the influence of trade organizations in the administrative process, and that this is also a well-known effect of the work of traditional “producer protection” regulatory commissions such as the Interstate Commerce Commission (ICC).  Yet nonadministrative safety regulation, such as building codes, is often anticompetitive too.  Noll also suggests that administrative procedure, with its eternal delays and preoccupation with “equity” to protect interests in the status quo, strengthens the hypothesis that the dominant purpose of regulation is to reduce economic risk by dampening the rate of economic and technological change (Owen & Braeutigam, 1978:1-42).  But counter-examples abound here in both directions.  Building codes and rent controls (nonadministrative) retard economic adjustments and preserve the status quo, while environmental and automobile safety regulations (administrative) are often explicitly “technology forcing.”  The very flexibility, and hence the uncertainty of administrative policymaking appears recently to have increased economic risk in several sectors of the economy, such as electrical power generation.

Is Regulation All of Law?

The most expansive conception of regulation is that of the theorists of the Chicago School, who in their efforts to understand regulation as an arm of the government’s redistributive apparatus have tended increasingly to treat “regulation” and redistribution” interchangeably.  Their work recognizes no distinction between economic and social regulation, nor between policies that do and do not apply to business or even to explicit markets, nor between policies that operate through rules and those that operate through taxing and spending.  Thus, in three influential articles during the 1970s, Richard Posner (1971) argued that regulation could be considered a branch of public finance, George Stigler (1971) incorporated tariffs and other differential taxes into the regulatory fold, and Sam Peltzman (1976) built a formal model of redistribution-by-regulation whose most obvious application was to direct fiscal redistribution. 

The Chicagoans’ usage departs boldly from the conventional understanding of regulation, for they have lost interest in regulation as it is conventionally understood.  They propose to explain the patter of government policies in strict economic terms, specifically in terms of the costs and gains of political action by differently situated groups.  Their interest is thus in the economics of politics, they have called the supply side of the political market “regulation” rather than “legislation” or “state action.”  Eventually it may turn out that “regulation” in some narrower sense is systematically more or less costly in supplying the state’s favors than other kinds of policies, but only then will a redefinition of terms be necessary.

The idea of regulation has tended to expand in the hands of political scientists as well as economists.  James Q. Wilson’s The Politics of Regulation (1980), a collection of case studies by political scientists and an explicit challenge to the Chicago economists, exhibits a similarly broad conception of what regulation is.  His book contains, for example, a chapter on the Antitrust Division, which enforces laws that have conventionally been considered as alternatives to regulation, and a chapter on the Office of Civil Rights, which polices racial and other discrimination in organizations receiving federal grants (school and other institutions of local government).  Noll would have omitted both chapters, since the first does not involve the APA and the second involves terms of government contracts.  Wilson differs from the members of the Chicago School in thinking that political and institutional detail is more important than “cosmic generalization” for understanding regulation, but nowhere does he suggest that the details are fundamentally different between regulation and other sorts of policies.  He concludes his book with an essay setting forth a typology of “regulatory politics” when regulatory benefits are dispersed and costs are concentrated,” client politics” when benefits are concentrated and costs are dispersed—a typology that might also be applied to “tax politics” or “welfare politics.”  It is noteworthy that Stigler, in a scathing review of the Politics of Regulation, described this typology as “simply a primitive version of the economic theory of regulation” (1980:12).

That scholars as diverse as Wilson and Stigler have found it unnecessary to distinguish sharply between regulatory politics and other kinds of politics obviously bodes ill for the hopes of regulatory reformers (such as Noll) that political institutions may have a determining effect upon policy.  There are, however, two glimmers of hope in the recent work of Richard Posner (1981) and Friedrich Hayek (1973, 1976, 1979).  Posner argues that while legislative law may typically promote redistribution of wealth and income, judge-made common law typically promotes allocative efficiency in the sense of maximizing economic product or wealth (1981:88-115).  He argues that legislative policies can have distributive effects that are “substantial and nonrandom” with respect to identifiable groups, so that individuals with similar interests have incentives to organize and exert political pressure to obtain economic benefits through legislation.  Their success typically brings them benefits smaller than the consequent losses to those outside the successful coalitions, and thus diminishes the wealth of the whole society.  In contrast, common law adjudication of discrete, individually small disputes cannot confer substantial and nonrandom benefits on identifiable groups, so individuals have little incentive to organize and exert pressure on the process.  The best anyone can do is to favor the efficient (wealth-maximizing) resolution of all common law disputes, since this confers the greatest benefits on each individual group.

Posner’s argument leaves room (implicitly) for treating regulation as an intermediate case.  Consider his example of the redistributive potential of legislation, progressive versus proportionate taxation (1981:101-102).  If tax revenues were held constant, lower income taxpayers could realize higher after-tax incomes under progressive taxation than under proportionate taxation, even though higher income taxpayers would do better—and aggregate income would be higher—under proportionate taxation.  This could happen even if, as today’s “supply side” tax-reduction advocates maintain, higher income taxpayers earn less taxable income under progressive taxation.  For they still might earn enough to generate greater tax revenues under higher progressive tax rates, thus permitting lower income taxpayers to enjoy lower tax rates and higher after-tax incomes.  Posner’s contrasting example of the redistributive deficiency of common law adjudication is landlord-tenant law.  Deciding all landlord-tenant disputes in favor of tenants, since landlords would simply react to the increased risk (and other costs) of doing business by raising their prices, or if this were ineffective or legally prohibited, by withdrawing their investments.

The crucial difference in these examples lies in the elasticity of response to the two policies.  Faced with the respective common law and legislative policies, landlords have greater alternative opportunities than higher income taxpayers.  But high response elasticity is also a characteristic of regulatory policies.  Indeed, rent control is said to be an ineffective means of aiding low-income apartment dwellers for precisely the reason given by Posner in his example of landlord-tenant law.  The principal normative argument against using regulation as a device for redistribution is that it is bound to produce greater allocative inefficiency than broad taxes on earnings or consumption; there is less elasticity of response to raising the price of all earnings, or the price of all or most of the things earnings can buy, than to raising the price of just one or a few of the things earnings can buy.  But this argument, like Posner’s is at the same time a positive argument.  Redistribution by regulation is ineffective—and hence an unpromising source of governmental benefits around which to organize—to the degree it is inefficient.

Of course, taxes may be narrowly targeted and regulatory programs may be more comprehensive than the setting of apartment rents.  But the most pervasive regulatory programs, such as the public-utility programs, which cover not only price buy quality of service and service initiation and discontinuance, apply at most to one or a few markets.  And the regulatory programs applying to many markets, such as the safety and environmental programs, cover at most one or a few terms of transactions in any one market.  While there is a powerful tendency for regulatory controls to expand in an attempt to defeat compensating reactions to earlier policies (McKie, 1970), the consistently negative conclusions of studies of regulation in many different fields suggest that the regulative effort is usually futile—that human behavior has many more, and more subtle and tenacious, facets than can be controlled, monitored, or even imagined by the regulator.

Friedrich Hayek’s Law, Legislation and Liberty (1976, 1976, 1979) introduces a different pair of categories.  Hayek divides all social rules (not just those established by governments) into two types: (a) general rules, such as customs, manners, and commercial and social standards of all sorts, which emerge spontaneously through social evolution and may or may not be adopted in legislative or common law, but “have no purpose” conceived or desired by any particular individual or group; and (b) purposive rules, which are adopted by private or public organizations in an effort to achieve some purpose.  His argument, very briefly, is that the only proper function of purposive rules is to direct the activities of individuals within organizations—within the context, that is, of voluntary relations.  When enacted by governments in pursuit of social goals, in an attempt to move society as a whole in a particular direction, purposive rules are based on the dangerous illusion that society can be managed as if it were a single organization, or that it can have any singular or consistent purpose at all.

Hayek’s formulation encounters serious difficulties when one gets down to identifying social rules that evolved purely through chance and competition and hence “have no purpose.”  Individuals formulate their interests consciously, use persuasion as well as outright coercion to engage others in cooperative ventures, and observe and form opinions about their fellows.  As a result, we can speak of any of our social rules as “having a purpose” without being teleological, whether we are speaking of rules that emerged out of the mists or the past or rules enacted by identifiable legislative coalitions.  (That we may disagree endlessly over the “true” purposes of laws and other rules is a separate matter.)

For example, Posner’s essay discussed above is an explanation of the content of common law rules in terms of purposive individual action, and it evidently grew from his and others’ dissatisfaction with attempts to explain common law rules in strictly evolutionary terms (Landes & Posner, 1979; Rubin, 1977; Priest, 1977).  Even seemingly neutral rules of social coordination, such as those of measurement and timekeeping, invariably involve divergent interests.  There is a distinct politics of daylight-saving time: people in different geographic and occupational situations wish to have it observed for longer or shorter periods, beginning earlier or later in the year, or not at all (Bartky & Harrison, 1970).  And every machinist or manufacturer knows the large economic stakes involved in the choice between metric and American-British measurement.  Presumably, so long as England and the United State remain major industrial powers, the American-British system will persist, however superior the metric system may be in the abstract.

The question of purposiveness can be avoided by the similar but less problematic distinction between general and specific rules, which often seems to be what Hayek has in mind.  Common law, and informal social rules such as customs and oral conventions, consist almost entirely of general rules.  They are standards rather than stipulations of conduct.  They do not prescribe the exact terms of either deliberate or circumstantial relations among individuals, but rather establish boundaries for acceptable conduct within which individuals are free to gauge their actions or negotiate joint actions with others.  Whether the boundaries should be declared violated is left for ex post determination in the event of a dispute.  This conception is a departure from Posner’s argument that common law is typified by its very specificity.  A court decision in a landlord-tenant case, he suggests, alters only one term of a contract, leaving contracting parties to alter all of the other terms in the future.  While common law can do no more than this, it often does less.  A single court decision is both highly specific and limited to the case at hand; the decision rests upon an interpretation of the behavior of the disputants under particular circumstances and is important for the future primarily as a guide to behavior under arguably analogous rather than identical circumstances.  In other words, common law consists not so much of specific rules as of general guidelines implied by a stream of specific decisions.  There are arguable exceptions.  The common law of contract occasionally adds fairly specific terms, such as implied warranties, to certain kinds of contracts.  But even the “implied warranty of habitability” in landlord-tenant law is much more general, and more open to interpretation in a particular case, than the terms of a rent-control ordinance.  Of course, this way of looking at the matter only strengthens Posner’s argument about the difficulty of effecting redistribution through common law.

The legislative and administrative agencies of the modern state are more versatile than the courts, and their actions tend to be more specific.  Certainly they are highly specific in defining the terms of rights and obligations between the government itself and others, whether the others are citizens, in the case of programs of taxation and expenditure, or employees and other contractors, in the case of the government’s organizational regulations.  When, on the other hand, legislatures adopt rules for behavior among citizens, they often do so in fairly general terms, as in most traditional criminal law, which incorporates standards of conduct from the common law.  But the legislative and administrative branches also, and apparently increasingly, spell out rules for behavior among citizens that are highly specific, to the point of prescribing the exact terms of private transactions.  This seems to come closest to what we mean by regulation, and to set regulation apart in an unambiguous and useful way from other government actions.

Conclusion

The view of regulation suggested here defines a clear subset of government actions—those that prescribe the terms of specified private transactions.  Only some statutes and agencies do this, and those that do usually do little else.  The terms, of course, may be price terms, quantity terms, or quality terms.  Minimum wage laws and direct wage controls prescribe price terms in labor markets; maximum hour laws prescribe quantity terms in labor markets; and OSHA, pension fund, and equal employment regulations prescribe quality terms in certain information markets.  Economic regulation and social regulation are thus appropriately consolidated.

The suggested view is not entirely comprehensive.  Regulatory agencies may do less, as in denying a proposed price reduction without specifying what price is “just.”  And the prescription may be implicit, as when an agency awards compensation based on an existing term (a price or business practice) it finds to by discriminatory.  But prescribing certain terms of certain transactions is the most any regulatory authority can do.  Even prohibitions, as of child labor, the manufacture of certain chemicals and other products, or the making of certain advertising claims, are simply the limiting case of banning certain transactions altogether.  Thus, this approach captures the essence of regulation more completely than general references to “a type of social control of transactions” (Enthoven-Noll), because it specifies the extent of any regulator’s powers of social control.

The suggested view is exclusive, distinguishing regulation from other forms of government action discussed in this chapter: taxation and expenditure, the government’s own organizational regulations, general rules of common law, and specific rules of coordination such as those of measurement and timekeeping.  All these actions influence private transactions—sometimes massively, sometimes trivially—but none stipulates their terms.  If regulation is indeed a distinctive form of policy for any of the reasons mentioned by the leading authors in the field, the distinction is likely to rest upon the circumstance of setting on or a few terms of a transaction to which one is not a party, leaving a multitude of other terms to readjust according to the interests of those who are parties.  As an approximation, regulation is less effective than taxing and spending but more effective than common law in redistributing income.  The normative side of the same proposition, and the essential argument of the regulatory reformers, is that regulation is also less effective than taxing and spending in correcting market failure.

Another virtue of the approach suggested here is that it reduces the idea of regulation to pure function, stripping away accretions that are not descriptions at all but rather arguments over purposes and consequences.  We may debate whether the ICC or the EPA is correcting market failures, balancing the interests of producer and consumer groups, promoting safety or national integration or regional development, providing political vent to the power of certain factions, redistributing income in  worthy or unworthy manner, or reducing market uncertainty and the rate of economic change.  But surely it is best not to identify regulation itself with the pursuit of any of these goals.  The purposes of regulatory policies are deeply ambiguous.  For example, competing firms that agree to standardize their products in some respect (Through regulation or otherwise) may know it is in their interest to do so, but neither know nor care whether this is because standardization will “widen the market or narrow the competition” (in Adam Smith’s phrase).  Moreover, even if their intention is exclusively to narrow the competition, they may very well end up promoting a wider interest that is no part of their intention (as Smith again puts it).  Even the original Act to Regulate Commerce of 1887, long regarded as an imperfect effort to shore up a disintegrating cartel, is now regarded by at least one careful scholar (a Chicagoan) as having been plausibly related to the promotion of general economic welfare (Kitch, 1979).

A degree of reductionism would be a healthy thing for political debate as well.  Consider, as a small example, the perennial debates over the inclusion of certain expenses in the rate bases of utilities, such as expenses for political advertising, charitable contributions, and construction work in progress on generating plants not yet producing power.  These debates are almost always cast in terms of whether consumers or owners should pay for the activity in question, or whether it is fair to charge today’s customers for tomorrow’s power.  Yet the utility commissions cannot decide these questions; all they can decide is what rates the utilities may charge.  Depending on how financial and consumer markets react, the ultimate results may be just the opposite of those envisioned in the debates, and recent research suggests that they often are just the opposite (Lehn, Benham & Benham, 1980; Navarro, 1980; Comptroller General of the United States, 1980).

Finally, the suggested formulation is compatible with current economic thinking about market failure, which consolidates all of the various “failures” into the same underlying phenomenon—the costs of market transactions.  Emphasizing that regulation is in fact no more than the prescription of terms of transactions brings out the important point that, in any regulatory controversy, the issue is not whether a failure should be corrected, but whether a government agency can formulate appropriate terms of transactions at lower cost than the market can.  Monopoly, externality, and poor information are not problems unique to markets or amenable to technical administrative solutions, even by thoroughly disinterested government agencies.  Rather, they are problems that may be transferred from market settings to administrative settings, with attendant advantages and disadvantages.  Every contention for government regulation based upon market failure is more precisely a contention that economies of larger scale decisionmaking remain somehow to be exploited.