While proposals to reform the regulatory process are almost as old as administrative regulation itself, what is meant by "reform" has changed along with prevailing ideas about how the government should be organized and what it should and should not be doing. At one time or another in the past regulatory reform has meant, among other things: improving the internal management of the regulatory commissions and agencies, making them more accountable to the President (or to the Congress, or to the courts), and making them less accountable to regulated businesses and other private groups. The current regulatory reform movement, which began in the early days of the Ford administration and has gathered steam under President Carter, has been concerned primarily with erecting new institutional constraints on the costs that regulatory decisions impose on the economy. Before considering the policies and policy proposals that have grown from this conception of regulatory reform, it will be useful to observe how closely the idea of constraining regulatory costs parallels today's broader political developments.

The late 1970s were a time of retrenchment in American politics. The "deadlock of democracy" which so worried political analysts in the 1950s and early 1960s—and then was forgotten during the post‑Kennedy years of heady legislative activism—appeared to reassert its grip. The present Congress is a "do nothing" Congress despite constant hectoring from President Carter on energy legislation and other matters. But this time around it would be difficult to argue that Congress is simply being unresponsive, since the public seems extremely wary of the government's doing anything at all. The situation is exasperating, both to those who would like to press forward the expansion of the federal government's domestic functions and to those who would like to roll back the advances of earlier political eras.

Explanations for the current paralysis range from the narrowly journalistic and ad hoc (President Carter's distaste for legislative politics, or post‑Watergate disillusionment with government) to the broadly economic (the decline in real personal income sapping enthusiasm for expensive new public ventures). An intermediate explanation draws upon political science at its customary middle ground between journalism and economics—concerned with both description and measurement, but interested primarily in the conditions of effective governance. This explanation is the recent decline of political institutions whose (usually implicit) function was to promote a broad public consensus over the appropriate purposes of government action—and, in so doing, to channel policy debate toward the resolution of narrower political conflicts. As James Q. Wilson has noted, the legislation boom years of the 1960s and early 1970s were accompanied by the collapse of many traditional institutions that had long stood against the expansion of government, especially the federal government (Commentary, February 1979). Among these were the congressional seniority system, the monopoly of decentralized political parties in the selection of public officials, and the general agreement on political and constitutional doctrines limiting the role of the federal government. For better or worse, constraining institutions such as these appear to have passed unmourned; there is as little prospect of resurrecting them as of repealing the domestic legislation that accompanied their demise.

But if the current era is not reactionary, it is plainly a time of searching after new and more formal institutions that will place some discipline on the political process and provide a new framework for debate over the strengths and weaknesses of particular government programs. Today, the policy proposals that have any momentum at all are those that are directed at restraining the size and scope of government.

Fiscal Limitation and Regulatory Reform

The most conspicuous attempts to establish new forms of political discipline have been those directed at "fiscal limitation." An early augury, in 1973‑74, was President Nixon's policy of selective executive "impoundment" of congressional appropriations, which provoked furious opposition but did spur Congress to establish budget committees and a budget office in order to exert greater legislative control over government spending. More recently, the success of Proposition 13 in California and similar programs in other states—along with the remarkable momentum that has gathered behind proposals for a constitutional amendment limiting federal expenditures—testify to the depth of public concern about government spending. Though both Democrats and Republicans have done well at the polls recently by portraying themselves as "fiscal conservatives," many voters seem to feel that something more basic than a change of legislative attitudes—or faces—is required.

The current regulatory reform movement is largely animated by the same quest for new mechanisms of governmental constraint. It is true that regulatory reform has been the rallying cry for a great variety of policies and proposals: abolition of traditional economic regulation of the airline and surface transportation industries; application of cost/benefit analysis to newer programs of health, safety, and environmental regulation; heightened judicial scrutiny of regulatory policies under freedom-of‑speech and right‑of‑privacy doctrines developed in noncommercial contexts; regulatory "sunshine" and "sunset" laws; even demands that federal regulations be written in plainer language. These and other so‑called reform initiatives have one thing in common, however: they all aim to reduce the impact of regulation on the economy through an appeal to a general principle—such as economic efficiency or freedom of speech—that transcends the pros and cons of individual regulatory schemes. Current proposals to increase regulation, such as those for federal corporate chartering and for a new "consumer advocacy" agency, have not gone forward under the banner of reform, although they might well have in earlier times.

It is not only the quest for systematic constraints that makes regulatory reform a close policy complement of "fiscal limitation." Regulation differs from other forms of government action in that it pursues the government's objectives not by spending public funds, but by causing private funds to be spent. Since the government's own administrative expenses of regulating are relatively small, a government that operates under some form of fiscal limitation (even if this is only a fixed tax schedule and an annual fiscal budget) will have an incentive to increase its reliance on regulation—so long as there is no parallel mechanism of regulatory limitation.

For this reason, further tightening of the government's fiscal constraints might have very little influence on the actual share of the economy controlled by government. With a little ingenuity one could probably establish a comprehensive national health insurance program entirely through regulation, with little effect on the federal budget. As a matter of fact, successive versions of the major health insurance proposals have become distinctly more "regulatory"—relying on compulsory private health insurance under federal standards rather than direct federal insurance—as concern over the size and chronic deficits of the federal budget has taken hold in Washington.

Restrictions on taxing and spending unaccompanied by equivalent restrictions on regulation might also diminish the government's political accountability. Regulatory programs, however agreeable their formal purposes, frequently end up redistributing wealth in ways that are crude and surreptitious as compared to outright spending programs. Whether the issue is the rate structure for telephone or electricity service, minimum wages or maximum interest rates, crude oil "entitlements," or automobile fuel‑economy standards, it is usually unclear just who is subsidizing whom in what amount, but only too clear that none of the possible subsidy patterns accords with any widely accepted ideal of distributive justice. Certainly, regulatory subsidies can rarely be traced to a clear‑cut decision adopted on‑the‑record by an identifiable legislative majority.

Traditionally, of course, the formal purpose of regulation has been to improve economic efficiency by correcting for so‑called market failures such as "natural monopoly" and "external costs." Subsidies accomplished in the course of controlling pollution levels or utility prices acquired a stigma of illegitimacy in part because they seemed to go beyond the formal purpose of the regulatory enterprise. Increasingly, however, redistribution is not just the concomitant or even the furtive purpose of regulation, but the explicit public purpose, as in the cases of petroleum price controls and allocation schemes, state regulation of insurance rates, and some antidiscrimination regulations. Even in these cases, however, regulatory programs are more likely than outright spending programs to result in perverse redistributions that no legislature would embrace publicly, since regulation by its nature involves a greater delegation of discretion to administrative agencies and has economic consequences that are less easily apprehended. A dramatic case in

point is the current federal program of crude oil price controls, whose explicit redistributive purpose is to transfer scarcity rents from owners of American oil reserves to ultimate customers, but whose primary accomplishments are to subsidize owners of foreign (OPEC) oil reserves and domestic oil refineries.

Regulatory Reform by Regulatory Review

To date, systematic regulatory reform at the federal level has consisted of regulation‑review procedures in the Executive Office of the President. These have been designed to integrate regulatory decisions with broader government policies—in particular the minimization of inflation—by encouraging regulatory agencies to take greater account of the costs that their de­cisions impose upon the economy. The theory behind these review procedures appears to be (1) that regulatory agencies, with no budgetary constraints on the total social costs their deci­sions entail, take insufficient account of the costs of regulation and (2) that a degree of superordinate review by those who do worry about costs might help to redress the imbalance in regulators' incentives.

The first serious effort of this sort was the Inflation (or Economic) Impact Statement pro­gram instituted by President Ford early in his administration, which required the executive branch agencies to prepare evaluations of the expected impact of all major new regulations upon prices, productivity, and competition (see James C. Miller Ill, Regulation, July/August 1977). This program was supplanted in 1978 by President Carter's more ambitious Improv­ing Government Regulations program, which among other (mostly hortatory) things required the agencies to publish semiannual agendas of all upcoming regulations and to prepare "reg­ulatory analyses" of all regulations projected to have "an annual effect on the economy of $100 million or more." These analyses were to in­clude "a description of the major alternative ways of dealing with the problems that were considered by the agency; an analysis of the economic consequences of each of these alterna­tives; and a detailed explanation of the reasons for choosing one alternative over the others."

The task of implementing the regulation ­review programs has fallen upon a loose coali­tion of staff offices within the Executive Office of the President. Officially the programs have been the responsibility of the Office of Manage­ment and Budget (OMB), but OMB has dele­gated most of the day‑to‑day oversight to two other offices more directly concerned with in­flation in the private sector—the Council on Wage and Price Stability (CWPS), whose statu­tory function is to monitor government and private decisions far their effects upon prices, and the Council of Economic Advisers (CEA). And under the Carter program, a new informal group has been established, the Regulatory Analysis Review Group (RARG), whose mission is to select ten to twenty of the agencies' regula­tory analyses each year for independent review. While RARG is technically an interagency com­mittee including representatives of all the major departments and agencies of the executive branch, it is chaired by the chairman of the Council of Economic Advisers, staffed by indi­viduals from CWPS, and in practice operates as a joint venture of CEA, CWPS, and OMB. While the executive orders establishing these pro­grams do not officially reach the so‑called inde­pendent regulatory agencies, CWPS has been an active intervenor before these agencies un­der its broader statutory mandate, arguing against regulatory proposals that might be an­ticompetitive or unduly inflationary.

The formal purpose of both the Ford and Carter regulation‑review programs has been to reduce, not just to monitor, regulatory costs. President Carter's order states, for example, that regulations "shall achieve legislative goals effectively and efficiently" and shall "not im­pose unnecessary burdens on the economy"; it directs agency heads to issue regulations only after making sure that "regulation is needed," that "meaningful alternatives are considered and analyzed," and that "the least burdensome of the acceptable alternatives has been chosen." The programs, however, have lacked enforcement provisions that would give teeth to these policies. They have even lacked standards with which to judge the regulatory analyses that the agencies are required to perform. The executive orders have not, for example, required that regulations pass any sort of "cost/benefit" test, that the magnitude or distribution of costs be related in any less precise way to benefits, or that costs or effects on prices be held within any particular limits.

One reason for the lack of standards and enforcement provisions is that the regulatory statutes vary in the extent to which they per­mit cost considerations to affect agency deci­sions. For instance, the Toxic Substances Con­trol Act requires the Environmental Protection Agency (EPA) to consider cost‑effectiveness in regulating use of toxic substances, as does the Consumer Product Safety Act in the case of the Consumer Product Safety Commission's (CPSC's) product standards; the Clean Air Act permits the EPA to consider costs for some pol­lution‑control purposes but not for others; and the Delaney Amendment to the Food, Drug, and Cosmetic Act, which requires the Food and Drug Administration (FDA) to prohibit certain food additives shown to cause cancer in ani­mals, implicitly forbids consideration of the costs of such prohibitions. An executive order cannot treat these statutory policies in the same way: while the President has general constitu­tional authority to supervise the conduct of of­ficers of the executive branch, he may not direct his subordinates to implement his policies in ways that contradict the directives of statutes. Nevertheless, many regulatory statutes are si­lent on the matter of cost‑effectiveness, simply admonishing the agencies to be "reasonable" in pursuing statutory goals, and there is judicial authority (as we shall see) that such statutes do oblige agencies to take account of costs in framing regulations. So the hortatory nature of the regulation‑review orders probably reflects doubt that, purely as a political matter, the President and his aides could force the regula­tory agencies to live up to an explicit cost standard.

Whether from legal or political timidity, the presidential orders establishing the regula­tion‑review programs have imposed no actual constraint upon the substance of regulatory de­cisions; their only important requirement has been the procedural one that agencies commis­sion studies of the costs of prospective regula­tions and "meaningful alternatives." Apparent­ly, the hope behind the cost‑analysis requirement—like that behind the earlier environmen­tal‑impact statement requirement established by statute in 1970—is that the mere exercise of anticipating particular consequences from gov­ernment decisions will lead those making the decisions to pay greater heed to the consequen­ces. Also, of course, forcing an official to ac­knowledge publicly the existence and extent of some negative consequence of a policy makes it easier for those who oppose the policy to mo­bilize their opposition; the environmental im­pact program is also precedent for this surrep­titious function.

What has happened in practice, in any event, is that enforcement of the regulation‑re­view programs has consisted of political skir­mishing within the executive branch, with the staff agencies of the Executive Office of the Pres­ident (CWPS, CEA, and RARG) conducting economic search‑and‑destroy missions against the line regulatory agencies, stalking through the Federal Register and the accumulating reg­ulatory analyses after proposals that appear particularly costly, ill‑advised, or politically vulnerable. (Within the regulatory agencies, to have one's regulations selected for higher‑level review is to "get RARGed.") In late 1978 the heads of thirty‑five regulatory agencies, includ­ing a number of the independent agencies, re­sponded by forming a sort of joint chiefs of staff for the defense, called the Regulatory Council and chaired by the administrator of the Environmental Protection Agency. The Reg­ulatory Council has taken over and expanded some of the activities required under President Carter's regulation‑review program, such as the publication of a semiannual Calendar of Feder­al Regulations (see "The Regulatory Calendar," page 3, this issue), and has commissioned stud­ies of better methods of assessing the benefits of regulation.

It is difficult to determine whether all of the jostling and proliferation of organizations has been worthwhile. While it has clearly brought about improvements in the substance of a few regulations, it has also brought about a great deal of confusion, intrigue, and bad feel­ing within the executive branch. It has necessi­tated occasional (and reluctant) presidential intercession to arbitrate disputes over particular regulatory proposals, and has provoked growing criticism from Congress and the lead­ers of important political factions—who believe they ought to be the principal arbiters of policy disputes and resent meddling from the Office of the President.

The Cotton Dust Case

A well‑publicized episode illustrating all of these effects is the Occupational Safety and Health Administration's adoption of a regula­tion limiting workplace exposure to cotton dust, which causes a respiratory disease known as byssinosis. In late 1976, OSHA proposed to re­quire a substantial reduction in permissible levels of cotton dust in workplaces in all seg­ments of the cotton industry—ginning, milling, yarn and fabric manufacturing, and waste proc­essing—through the installation of ventilation equipment and controls. Before proposing the regulation, the agency commissioned a consult­ing firm to prepare the economic impact state­ment required under the Ford administration's program. The impact statement estimated, among other things, that the costs of complying with the regulation would be very high—$808 million on an annualized basis (including cap­italized initial costs plus annual operating costs).

During 1977, OSHA officials heard oral tes­timony and received voluminous written com­ments on the cotton dust problem and various ways of dealing with it, and by the end of the year they had decided to pursue a more flexible approach. In particular, they determined to permit different exposure limits (and different means of compliance) for the various stages of cotton production and manufacturing, based upon differences in costs of compliance and de­grees of health risk at each stage. OSHA's change of approach was the result of a variety of factors, including the more exacting infor­mation on the nature of the problem gained from oral and written testimony, and the political pressures brought to bear by management and labor from the different sectors of the cot­ton industry. But clearly an important factor was the enormity of the costs projected by the economic impact statement and the public criti­cism the projection had evoked. In early 1978 OSHA conducted a further cost analysis, based upon the data in the impact .statement, which concluded that its modified, variable‑exposure ­limit regulation would be only about one‑quar­ter as costly as its original proposal.

OSHA's change of heart did not result from any regulatory review or pressure from the Council on Wage and Price Stability. The coun­cil was occupied with other matters throughout 1977 and did nothing beyond submitting a brief critique of the original proposal during the offi­cial comment period. Only at the very last stage, when OSHA was about to issue its highly modi­fied proposal, did CWPS become involved. After a briefing from OSHA on the modified regula­tion, CWPS prepared an informal memoran­dum listing numerous objections and request­ing that final action be delayed for further con­sideration.

In particular, CWPS objected that the reg­ulation would rely on prescribed "engineering controls" (ventilation equipment) rather than on "performance standards" (such as byssino­sis‑reduction quotas) that could be met in the most cost‑effective manner available (for ex­ample, through the use of personal respirators, rotating work assignments, or increased medi­cal surveillance). OSHA officials who had been working on the cotton dust project felt the CWPS criticisms were naive, amounting to lit­tle more than the economist's reflexive prefer­ence for economic incentives over command-­and‑control regulation regardless of the prac­tical difficulties of enforcing performance standards. They noted, for example, that OSHA inspectors could easily police the installation of ventilation equipment, but could not rely on mill workers to wear respirators or report res­piratory problems to their supervisors.

A furious interagency dispute ensued, and in the end not even the top officials of the agen­cies could resolve it. Charles Schultze, chair­man of the Council of Economic Advisers (and also of RARG), eventually made a direct appeal to Secretary of Labor Ray Marshall, reiterating the CWPS objections and request for postpone­ment. Secretary Marshall responded that while he greatly valued the advice of Dr. Schultze on all matters of national concern, Congress had entrusted the Department of Labor, not the CEA, with the responsibility for ensuring a healthy workplace to all American workers, and the department intended to press ahead with its responsibilities. Then Dr. Schultze obtained a private meeting with President Carter and thought he received a commitment to relax the proposed regulation along the lines suggested by CWPS. But when Secretary Marshall heard of the decision he demanded a meeting of his own with President Carter and Dr. Schultze, and in the course of the second meeting the President apparently changed his mind. In the end, President Carter permitted the regulation to go forward substantially as OSHA had planned, though he did agree (at Secretary Marshall's suggestion) to a leisurely four‑year "lead period" to postpone the economic effects of the regulation.

This was not, of course, the end of the mat­ter. As soon as the final regulation was pub­lished it was challenged in court, both by labor—on grounds that it was too lenient and that the last minute involvement of the President and his economic advisers was contrary to pro­cedural requirements of administrative law­, and by management—on grounds that it was too stringent and took insufficient account of the costs projected in the economic impact statement. In addition, the regulation proved unpopular in Congress due to its projected costs, and the House‑Senate Conference Com­mittee Report on the Department of Labor's next appropriations bill directed OSHA to conduct a study that reviews the technical standards relating to the occupational ex­posure of cotton dust in an effort to devel­op viable alternatives which are less costly and more technologically feasible. The study should also review the inflationary impact of these standards and the effect of the standard on the ability of the U.S. cot­ton and textile industry to compete with foreign industry.

OSHA's report, sent to Congress in May 1979, reworked the data from the original im­pact statement to estimate compliance costs for the agency's final regulation and three other, industry‑wide exposure standards of varying stringency (not, however, including the CWPS alternative of permitting use of personal respirators and medical surveillance in place of en­gineering controls). The report also employed medical data to estimate the number of cases of byssinosis avoided under each of the four exposure limits and concluded, to no one's great surprise, that the very regulation OSHA had issued the year before was the most cost­ effective of the alternative approaches. (More lenient exposure limits with lower compliance costs were estimated to be more costly on a "per case avoided" basis because they were projected to prevent so many fewer cases in all.)

An Assessment of the Regulation‑Review Program

The cotton dust episode suggests that any prac­tical impact of the regulation‑review program arises simply from its requirement that agen­cies issue formal estimates of regulatory costs. The review process itself, aside from its inher­ently haphazard character, is unlikely to have much influence on the substance of regulations, since the people doing the reviewing have neith­er the legal authority nor the political responsi­bility for issuing final regulations. In the end the regulators can spurn the counsel of the reg­ulation reviewers, as Secretary Marshall spurned the counsel of Dr. Schultze—and with some justification in the particulars of the case, since the reviewers are necessarily removed from the daily give‑and‑take of policy develop­ment. While in theory the reviewers can appeal to the President on behalf of their own concep­tions of regulatory policy, in practice they will do so only as a final recourse—which means at the last moment, after the regulatory agency and influential political groups have acquired organizational stakes in proceeding in a certain manner, and the President's practical options are therefore narrowly constrained. While one can only guess what transpired between Presi­dent Carter, Secretary Marshall, and Dr. Schultze in the cotton dust case, it is reason­able to suppose that Secretary Marshall was much better informed of the technical and po­litical details of the policy under discussion, and was able to argue persuasively that any last‑minute revision of OSHA's proposal would be both substantively unwise and politically harmful.

It appears, moreover, that the outcome of the cotton dust controversy dampened the re­solve of RARG officials to seek presidential re­view in subsequent cases. In June 1978, during the week the final cotton dust regulation was issued, the Environmental Protection Agency announced a proposal to establish a new na­tional ambient air quality standard covering ozone. The standard, used as a "target" for the design and enforcement of state and federal programs limiting emissions of two major air pollutants (hydrocarbons and nitrogen oxides), was to be relaxed somewhat from the existing standard. In the following months, RARG/CWPS prepared what was generally regarded as a strong economic case for a substantial further relaxation of the ozone standard. (RARG's anal­ysis was, in any event, considerably more thorough and persuasive than the hasty mem­orandum it had prepared in the cotton dust case.) But in spite of the RARG critique and an intense session between Dr. Schultze and EPA Administrator Douglas Costle, Mr. Costle re­fused to revise the proposed standard to the ex­tent recommended by RARG, arguing that the statute in question (the Clean Air Act of 1977) excluded consideration of costs in the establish­ment of "target" air quality standards. At this point Dr. Schultze dropped the matter rather than take it up to the President, and the new standard was issued without substantial modi­fication.

There were, however, newspaper reports that Mr. Costle had yielded to White House pressure in relaxing the air pollution standard, and soon thereafter President Carter was asked at a press conference about his personal in­volvement in regulatory decisions. He respond­ed with these words: 

The regulators, Doug Costle and the others, know that they have authority to consider . . . economic considerations and they're to make their judgments accordingly. I have not interfered in that process. I have a statutory responsibility and a right to do so, but I think it would be a very rare oc­casion whenever I would want to do so.

The President's remark was interpreted as a clear signal that he wished to leave final regu­latory judgments to his regulatory officials themselves—and that his staff regulatory re­viewers were to function primarily as kibitzers rather than as super‑regulators threatening presidential intercession to force compliance with their own policy ideas. While the matter is by no means free from speculation, the gen­eral opinion seems to be that since the Presi­dent's remarks, neither RARG nor CWPS nor CEA has asked him to countermand a regula­tory proposal because of its projected cost. In any event, since the Carter review program was established in early 1978, only nine regulatory proposals have gotten RARGed out of a total of forty or so proposals with projected costs of $100 million or more, which is substantially short of the envisioned ten to twenty RARG re­views each year. (The estimate of about forty regulations costing $100 million or more prob­ably understates the number of regulations that will actually cost this much, since under the Carter program a suspiciously large number of regulations have been projected to cost $90‑95 million.)

The RARG analyses, and the less weighty CWPS staff analyses submitted in a number of additional regulatory proceedings, have gener­ally been of high quality, and have uncovered numerous defects and omissions in the agen­cies' projections of costs and benefits. They have undoubtedly been read with great interest, even by regulatory officials who most resent be­ing second‑guessed by economists and account­ants. One can even discern a few major revi­sions in final regulations consistent with the RARG analyses. For instance, the Department of Transportation's 1978 regulation requiring "non‑discrimination on the basis of handicap" omitted a proposed requirement, which had been sharply criticized by RARG, that the five major urban mass transit systems install ele­vators at all subway stops (it required them only at some stops).

But it is impossible, even in these cases, to ascribe much influence to RARG or CWPS. For one thing, regulatory agencies customarily pro­pose extremely stringent regulations in their in­itial notices, so as to leave themselves ample negotiating room and the appearance of fair-mindedness when their final rules turn out to be more moderate. For another, private politi­cal groups (such as associations of handi­capped people and of mass transportation au­thorities) are far more important to the orga­nizational interests of the regulatory agencies than sister agencies within the executive branch, and at least one private group is almost certain to take a position similar to RARG's or CWPS's (advocating cost minimization) on every regulatory proposal. Indeed, CWPS offi­cials themselves do not claim any independent influence on the outcome of any particular reg­ulatory proceeding. While they have a strategic interest in being modest about their activities, especially while the legality of those activities is being challenged in court (as in the cotton dust suit), there is no reason to doubt they are correct. Surely they have had negligible influ­ence on the general course of regulatory policy and the total costs of regulation, considering their institutional inability to review more than a few regulations in any depth and their inher­ent inability, as mere reviewers, to affect the rate at which regulations are generated. The major impact of the review process has prob­ably been to single out for debunking a few un­usually preposterous regulatory proposals, like the one to build an elevator at every subway stop. But these, of course, are the proposals that would have been singled out anyway.

The preparation of the regulatory cost esti­mate, on the other hand, did have a substantial effect on the final cotton dust regulation inde­pendently of RARG or CWPS review. It had a similar independent influence on the final handicap‑nondiscrimination regulation as well. The Department of Transportation's own cost esti­mate for the subway elevators yielded a quo­tient of about $3 added cost for each ride by a handicapped person, and the mass transit asso­ciations' estimate, which RARG found more plausible, was two to three times higher; the estimates received a good deal of publicity, and the point was often made that it would be cheaper just to give handicapped people free taxi rides. Publication of cost analyses can be expected to influence final regulations so long as cost‑effectiveness is considered a desidera­tum of regulatory policy and some important group (the cotton and textile industry trade groups in the cotton dust case, and the mass transit authorities in the subway elevator case) has an interest in advocating the point.

The influence of the cost estimates may even be a matter of legal obligation, in spite of the absence of standards and enforcement pro­visions in the regulation‑review program itself. This is so because the estimates, once generated and placed into the record of regulatory pro­ceedings, may limit the discretion of regulatory officials in interpreting their own statutory mandates. This is what happened in another recent case involving occupational health, in which OSHA determined to require a reduction in workplace exposure to airborne benzene from the then‑existing level of 10 parts per mil­lion down to 1 part per million (ppm). The agency had received epidemiological data show­ing an increased incidence of leukemia among workers exposed to benzene over long periods of time; the data, however, involved benzene exposure at levels over 100 ppm, and there was no evidence that levels of 10 ppm or less were associated with any increase in the incidence of leukemia. At the same time, OSHA's economic impact statement estimated that the cost of compliance with the new standards would be several hundred million dollars.

In issuing the new benzene standard, OSHA argued that it had no obligation to take account of regulatory costs: the Occupational Safety and Health Act, it said, directed it to protect the health of workers but made no mention of the costs of doing so. When its decision was chal­lenged in court, however, the Fifth Circuit Court of Appeals disagreed with OSHA's rea­soning (American Petroleum Institute v. OSHA, 1978). Noting that OSHA's statute authorized the agency to issue regulations "reasonably necessary" to address specific problems of oc­cupational health and safety, the court held that "although the agency does not have to con­duct an elaborate cost‑benefit analysis, it does have to determine whether the benefits expect­ed from the standard bear a reasonable rela­tionship to the costs imposed by the standard." The court therefore struck down the regulation, on grounds that OSHA had no substantial evi­dence of health benefits from its new benzene exposure standard commensurate with the standard's costs. (The Supreme Court has since granted OSHA's petition to review the court of appeals' decision, and its decision is due during the current term.)

The preparation of cost analyses has influ­enced regulatory policy, at least in a few in­stances, by introducing new information about the consequences of regulations in a form that can be readily assimilated by those (such as judges, journalists, and "average citizens") out­side the narrow circle of parties directly inter­ested in individual regulations. As a result, reg­ulatory officials have had to be more careful about the formal logic underlying their policies—and therefore more careful about the sub­stance of their policies. It is less clear, however, that the regulation‑review programs have been the essential cause of the developments, and less clear still that the cost‑analysis require­ment need be continued in its present form. The process of policy formulation has, after all, been growing steadily more formal and analyti­cal in recent years, and the calculation of regu­latory costs has been receiving increasing atten­tion not just from government officials but from academic economists and private groups ranging from the Business Roundtable to the Ralph Nader organizations. Surely the regula­tion‑review programs have been but one more expression of this attention—the vehicle for, ra­ther than the cause of, debate over regulatory costs within the federal establishment. If the programs had not existed at all, regulatory con­troversies such as those described above might have arisen and concluded substantially as they did, based instead upon cost studies prepared by private parties involved in the disputes.

Cost estimates prepared by parties with an interest in a particular regulatory proceeding might, of course, be considered biased or less credible than an estimate prepared by the reg­ulatory agency itself. But a sophisticated politi­cal group such as the Environmental Defense Fund is at least as capable of dissecting an in­dustry cost estimate as RARG is of dissecting an agency cost estimate. Nor would such a group be without knowledgeable allies on the issue of regulatory costs. In a proceeding to es­tablish an air pollution standard, both the industries supplying pollution control equipment and those who might be required to purchase the equipment possess expert cost information.

It is, of course, a well‑known and impor­tant fact of political life that some groups are more energetic than others in pressing their views upon regulatory (and other government) agencies, because their demands for govern­ment action are greater or because their costs of political organization are smaller. And there is no reason to think that the interests of the most politically effective groups should be identical to those of less effective groups or the public as a whole. A regulator who is "cost con­scious" only to the extent of minimizing those costs of concern to the parties appearing be­fore him might very well end up imposing greater costs on society at large—the aggrega­tion of costs borne imperceptibly by large numbers of unorganized individuals who never appear before him, or costs that cannot be estimated concretely because they take the form of "forgone opportunity" or "deterred innovation."

This does not, however, mean that the cost information assembled independently by the government is likely to be more compre­hensive or to yield different policy implications than the information generated by the clash of contending private interests in a regulatory proceeding. For the problem remains that the kinds of regulatory costs that can be meas­ured with the least technical difficulty and controversy—and which, therefore, will be measured and relied upon most readily even by an utterly disinterested regulatory agency­ tend strongly to be the same as those which private groups will advance in their own inter­ests. It is fairly straightforward to measure the discrete expenditures that discrete organiza­tions must make to comply with a regulation (for example, the costs of building lawnmowers to comply with the precise safety standard spec­ified in a recent regulation of the Consumer Product Safety Commission), and the organiza­tions themselves will ordinarily have strong in­centives to measure these costs and advocate their minimization. It is far more difficult to quantify the lost opportunities of a large num­ber of unidentifiable individuals resulting from the same regulation (for example, the costs of lessened competition among lawnmower manu­facturers and of reduced utility among consumers who would prefer less safe but less costly lawnmowers), and no particular organization will ordinarily be interested in bringing these costs to the attention of the agency. Agency cost estimates prepared under the regulation‑review programs have, in fact, been limited almost exclusively to direct compliance costs for par­ticular industries and firms; the method of cost estimation has been to interview engineers and other experts at a few firms in an affected in­dustry and to extrapolate their estimates to the entire industry. While the RARG analyses have made occasional reference to the agencies' fail­ure to estimate indirect or opportunity costs, RARG itself has not attempted to make such estimates.

One is struck, finally, by a depressing irony in the Ford and Carter regulation‑review pro­grams. They have attempted to alleviate the problems of regulation by adding another layer of regulation that replicates the original prob­lems. The review programs have made no at­tempt to alter the natural incentives of regula­tory officials. They have established for the regulatory agencies an "engineering stand­ard"—the preparation of cost analyses—rather than a "performance standard." They have embraced the hope that adoption of a particular technology might improve matters, but they have been without the resources or authority to make the improvement, and they have paid little attention to whether matters have actually improved. They have been long on rhetoric, purpose, and moral suasion and short on specific results.

The review programs have undoubtedly played some role in making regulatory officials and the general public more sensitive to the costs of regulatory decisions. But it is not clear that they have done any more than this, or that the role is any longer necessary. If the Supreme Court's decision in the OSHA benzene case ap­proves the fifth circuit court's opinion that "reasonable" regulations are those whose ex­pected costs bear some "reasonable relation­ship" to expected benefits, the doctrine will henceforth affect a substantial portion of all federal regulatory decisions. The obligation to show a reasonable relationship between costs and benefits, however imprecise, is bound to have a greater impact on regulatory decision making than the similarly imprecise but unen­forced language of a presidential order and the interagency criticism of the RARG reviews.

Legal and Political Problems

These criticisms of the regulation‑review pro­grams are not heresies, even among enthusiastic regulatory reformers. They are, indeed, shared in some degree by many who had a hand in establishing the programs and many who are administering the current one. In addition, the Carter program is presently threatened by legal and political difficulties that could compromise both the formal cost analysis requirement and the informal negotiations between White House and regulatory officials. The program is under attack not only in court (as in the cotton dust case) but in Congress, by legislators such as Senator Edmund Muskie who are accustomed to their own close oversight of the activi­ties of particular agencies. (Senator Muskie has charged that the review program consists of "potshot, one‑man reviews that carry politi­cal clout" and that have a "chilling effect" on EPA's efforts to protect the environment.)

It is fair to say that the political threats are more serious than the legal ones. There is little doubt of the President's constitutional authority to supervise the general conduct of regulatory officials (as in directing them to analyze and consider the expected costs of their decisions). While presidential authority to in­tervene in particular cases is less clear, it is not, as a practical matter, very amenable to judicial control. The restraint that presidents have tra­ditionally shown toward the regulatory agen­cies is as much a matter of political custom as legal doctrine, but it is a custom unlikely to fall into desuetude. Congress may be no less con­cerned than the President over regulatory costs—in the cotton dust case Congress was plainly more concerned—but it has regarded regula­tion as a quasi‑legislative activity ever since it established the Interstate Commerce Commis­sion in the nineteenth century.

In the face of pending litigation and grow­ing congressional hostility, the regulatory re­view process has, in fact, slowed down dra­matically. Only one RARG analysis of a new regulation has been published since March 1979 and only three are currently under way. This is only partially because the regulatory agencies themselves have postponed several proposals until after the Supreme Court de­cides OSHA's petition in the benzene case. How­ever the benzene case is finally decided, it is unlikely that the review process will regain its earlier momentum until the legal and political threats are overcome.

Some Proposed Alternatives

In these circumstances, numerous proposals have been advanced for supplementing or re­placing the Carter regulation‑review program. One proposal is simply to codify that program in law and, in so doing, to extend it to the "in­dependent" regulatory commissions. President Carter has proposed this himself, and the idea has been championed by a number of legisla­tors. (For example, see separate articles by Abe Ribicoff and Clarence J. Brown, Regulation, May/June 1979.) Introducing a measure of cost consciousness to the work of the independent commissions would surely be desirable, since the commissions' decisions are at least as costly to the economy as those of the executive agen­cies. But although a statutory review program would presumably settle the current political dispute between the President and Congress, it would probably not do so to the satisfaction of the President himself. While President Carter's bill provides for continued oversight of agency cost analyses by OMB, Senator Ribicoff's bill would transfer oversight to the Congressional Budget Office. But regardless of how the political details are settled, simply enacting the cur­rent review program into law would not rem­edy any of its most serious substantive deficien­cies. Regulatory review would still be highly selective (or "potshot") and the agency cost analyses would still largely duplicate the sub­missions of private parties; the regulators would continue to have a superior grasp of the technical and political underpinnings of their proposals; and the overall rate at which regula­tions are generated would remain beyond the reach of the review process.

A more sweeping proposal is to give the President or Congress a specific role in indi­vidual regulatory decisions in order to con­form regulations to anti‑inflation strategies and other broad national polices. The best known version of this proposal is that of Lloyd N. Cut­ler and David R. Johnson, first published in 1975 in the Yale Law Journal (vol. 87, no. 7) and later embraced in a draft report of the American Bar Association's Commission on Law and the Economy ("Federal Regulation: Roads to Reform," August 5, 1978). Under the Cutler‑Johnson‑ABA proposal, the President would be given statutory authority to direct regulatory agencies and commissions to "take up, decide, or reconsider critical regulatory is­sues within a specified period of time, and thereafter to modify or reverse certain agency actions relating to such issues." "Critical" issues would be those the President "finds to be of major significance both to the national inter­est and to the achievement of statutory goals other than the goal primarily entrusted to the regulatory agency in question." Congress would be authorized to reverse any presidential order by a majority vote of either house within a specified period of time, and (at least in the original Cutler‑Johnson article) the courts would be authorized to review presidential orders to see that they had a "rational basis."

The proposal seems exceedingly unwise. It would, of course, settle explicitly the scope of presidential authority over the regulatory proc­ess. But it is doubtful that this issue ought to be settled very explicitly. There are bound to be occasional regulatory decisions of unusual national importance, such as the postponement of a major automobile safety or fuel economy requirement, where no regulatory official would proceed without consulting the President—and where no court would forbid such consultation. Giving the President a formal statutory role in all major regulatory proceedings, however, would inevitably expose him to intense political pressures concerning a multitude of decisions to which, as a practical matter, he cannot de­vote much attention. In any event, by the time a regulation found its way to the President for review it would have been molded by months or years of political pulling and hauling: his realistic options would be narrowly constrained, although his ability to make one group or another very angry, through action or inac­tion, would be great. Even if—as would surely happen—the President declined to become in­volved in most regulatory matters, in the pub­lic mind he could not escape personal responsi­bility for having made the final decision wheth­er drug X should or should not be allowed on the market, how bicycle reflectors should be mounted, and what ought to be written in war­ranties for used cars. Where the President did intervene in a truly "critical" case, say to post­pone an automobile safety standard for eco­nomic reasons, his decision would have to be made according to all the administrative para­phernalia of formal publication, distribution to all parties of record, and so forth. And after all that, his lawyers would still have to make his decision stick in court. A judge would have the final say whether the President acted "ration­ally" in permitting unsafe cars on the road in order to reduce the inflation rate a fraction of a point.

The Cutler‑Johnson‑ABA proposal would demean the presidency by weighting the office with procedures appropriate to appointed offi­cials and career administrators. It is safe to say that no President would want to be cast as a super‑regulator and that no Congress would force the role upon him against his veto. Pres­ident Carter's general remark, that he has the authority to influence regulatory decisions but would rarely wish to exercise it, is probably the most any President would want, or ought, to say on the subject. It need only be added that giving Congress responsibility to review indi­vidual regulatory decisions would be equally unwise. It is because Congress is institutionally incapable of making such case‑by‑case judg­ments that it establishes regulatory agencies in the first place. And where the occasional regu­latory decision runs strongly contrary to con­gressional sentiment, Congress can and does correct matters by direct legislative action, as it did in 1977 when it promptly reversed the Food and Drug Administration's ban on saccha­rin.

Conclusion

The regulation‑review program, the proposal to make it statutory, and the proposal to give the President direct authority to revise and reissue regulations are three increasingly strong appli­cations of the same basic idea—that regulatory decisions will be improved by centralizing con­trol over them within the executive branch. The idea of centralization, it is worth noting, is at the root of most regulatory policies them­selves; whether the problem at hand is mo­nopoly pricing, unfair commercial practices, pollution, or hazardous products or work­places, the essence of the regulatory response is to centralize control over one or more as­pects of the problem in a government agency. Whether centralization actually leaves a prob­lem better, worse, or unchanged is, of course, always an empirical question, although the dis­appointing results of so many regulatory pro­grams cast some doubt on the basic idea itself, even as applied to the further problem of un­constrained regulatory costs.

In any event, in the case of the Ford and Carter regulation‑review programs, the evi­dence to date is that a modest degree of fur­ther centralization, from the regulatory agen­cies to the White House, has failed to bring about much constraint on the private costs of regulatory decisions. Certainly the programs have failed to duplicate the panoply of institu­tional constraints that affect the government's expenditure programs, so as to eliminate the inappropriate (and increasing) incentives of government officials to pursue public goals through regulation rather than outright taxing and spending. While it is conceivable that further increases in White House control over in­dividual regulatory decisions would constrain regulatory costs more effectively, the most forthright proposal for doing so—the Cutler­Johnson‑ABA proposal—raises such serious in­stitutional problems of its own as to give pause about the entire approach.

A very different proposal for imposing systematic constraints on regulatory costs, one that has been receiving increasing attention within the federal establishment during the past year, is the "regulatory budget." In brief, the proposal is for the President and Congress to establish prior aggregate limits on the costs that the individual regulatory agencies may impose upon the economy, in a manner similar to the current process of fiscal budgeting; the agencies would then be obliged to live within their regulatory budgets just as they now must live within their fiscal budgets. In contrast to the regulation‑review program and its variants, the regulatory budget would be a decentralized method of constraining regulatory costs. The President and Congress would be assigned their appropriate role of setting general policy (in the sense of setting boundaries on the share of the nation's economy to be devoted to partic­ular endeavors), and the regulatory agencies would be left to make individual regulatory decisions within a budgetary framework that encouraged cost‑effectiveness in particular cases. The institutional constraints upon regu­latory costs would approximate those upon spending programs.

The regulatory budget idea has consider­able initial appeal, for these reasons and others. But the very rigor of the envisioned budgeting process raises acute problems concerning the nature and measurement of regulatory costs—problems that the more casual regulatory‑re­view programs have never had to face directly, and that have been ignored altogether in this article. The second article in this series will in­vestigate both the proposal and the problems.