Various attempts have been made over the years to estimate the total societal costs of regulation. The significance of the costs of regulation was first documented by Weidenbaum and DeFina for 1976.1 Later estimates by Litan and Nordhaus for 1977 and Hahn and Hird for 1988 confirmed the Weidenbaum and DeFina results and raised their estimates.2 The Hahn and Hird study estimates gross societal costs of $327 billion to $401 billion, or 29 to 36 percent of 1988 Federal expenditures. When various regulatory paperwork costs not already included (about $90 billion) are added to the Hahn and Hird study as well as the costs of final regulations issued since 1987 (approximately $17 billion, from table 1 above), the total cost of Federal regulation for 1990 increases to Б between $434 and $508 billion, or as high as 40 percent of 1990 Federal expenditures. Though estimates of the cost of Federal regulation vary widely, these data show that the economic effects of regulation are substantial. What may be needed to supplement and complement the existing regulatory oversight process is a system that forces regulatory decisionmaking to take place within governmentwide constraints on the total costs imposed by Federal regulation, that is, within a regulatory budget. TOWARD A REGULATORY BUDGET Earlier in this century, the need for a centralized budget process to manage effectively the programs of the Federal Government led to the development of the fiscal budget process in the United States and the passage of the Budget and Accounting Act of 1921. Through the creation of the President's budget and the appropriation process that follows, Congress and the Executive establish overall budget and program constraints. In spite of arguments about the complexities of this process, it serves its general purpose of requiring the Government to decide how it is going to allocate its resources. As discussed above, only rough estimates have been made of the total costs imposed by regulation. And so far, neither Congress nor the Executive has felt compelled to create a budget-like process to force tradeoffs to be made among the societal outlays mandated by different regulatory programs. Although concerns regarding the costs of regulation have led to the establishment of the Executive Order Nos. 12291 and 12498 review processes, these concerns have not yet forced the Federal Government to create a process analogous to the budget process to manage societal expenditures required by Federal regulation. The effects of budgetary and regulatory outlays are analogous in two important ways. First, the expenditures required by both have many of the same overall economic effects on output, employment, prices, and growth. The Federal Government finances outlays by diverting resources from the private sector through taxation and borrowing. Business firms finance expenditures required by regulation (e.g., for pollution control) by borrowing, increas 'M. Weidenbaum and R. DeFina, The Cost of Federal Regulation of Economic Activity (American Enterprise Institute Reprint No. 88, 1978). See Robert Litan and William Nordhaus, Reforming Federal Regulation, New Haven: Yale University Press, 1983; and Robert Hahn and John Hird, "The Costs and Benefits of Regulation: Review and Synthesis," Yale Journal of Regulation, Vol. 8, No. 1, Winter 1991, pp. 233-278. Although these studies also provide estimates of the benefits of regulation, they are not discussed here because this section focuses on the analogy between the regulatory budget and the fiscal budget. Benefits are not accounted for in the fiscal budget. 'Hahn and Hird provide separate estimates for social costs (the value of opportunities lost to society) and transfers (the value of losses to some members of society offset by equal gains to others) and do not aggregate them as is done here. They are added together to be consistent with fiscal budget practice. 6 REGULATORY PROGRAM OF THE UNITED STATES GOVERNMENT ing prices, reducing other expenditures, and reducing dividends. These, of course, are the same ways in which firms finance taxes, and thus have the same broad effects on the economy. Regulation mandating expenditures may, however, be more analogous in its effects to user fees and excise taxes than to income taxes, since income tax liability is more directly tied to earnings, profits, and interest income. Second, both regulation and budget outlays divert private resources to public purposes. Furthermore, in many cases, expenditures required by regulation may be an alternative means of achieving the same public-policy objectives as budget outlays or other instruments of government policy such as taxes, tax expenditures, or loan guarantees. For example, firms may be required by regulation to treat their effluents before discharging them into the air or water. Alternatively, public wastewater treatment facilities can be constructed by direct expenditures by the Federal Government or by States and localities with Federal assistance. The basic allocative effects are similar, although economic efficiencies and income distribution implications may vary from one policy instrument to another. The similarities between fiscal and regulatory expenditures have led a number of observers to look at the fiscal budget process as a source to design a similar oversight process to manage regulatory expenditures. Members of Congress and the last three Administrations have considered developing an accounting framework to track the expenditures that are directly required by regulation. A full accounting framework, however, is still in the developmental stage, and more work needs to be done to solve some practical accounting problems inherent in measuring the private expenditures required by Federal regulation. One practical accounting problem in establishing a regulatory budget involves the extensive records private firms and individuals would have to keep to validate projected budget estimates. These would not necessarily be accurate and could create an expensive compliance burden. A second difficulty involves subjectivity in differentiating between expenditures made because of a regulation and those which would have occurred in the absence of regulation. For example, in the absence of regulations for automobile safety, some level of safety would still be built into vehicles, because safety is an attribute consumers desire and value when they consider alternative purchase decisions. A third problem involves the difficulty of estimating the indirect costs of regulation. These include, for example, the loss to a consumer of opportunities to purchase goods because of higher prices, less desirable products, or outright bans resulting from regulation. Such indirect costs are relatively more important for regulatory accounting than they are for fiscal budget accounting. Since indirect costs are not directly measurable, and can only be estimated by complicated statistical models, it might be problematic to combine estimates of these indirect costs with measures of the direct costs of regulation. Yet measuring only the direct expenditure costs of regulation could create a bias toward banning substances and products rather than controlling them, since bans primarily give rise to indirect costs. Although these practical problems make regulatory budgeting significantly more difficult than fiscal budgeting, they should not be insurmountable. For example, the spending forecasts for fiscal budgets do not have to be perfectly precise for the fiscal budget process to be effective in controlling spending. Likewise the spending forecasts for regulatory budgets do not necessarily have to be absolutely accurate for the regulatory budget process to act as a constraining device for regulatory spending. The second measurement problem concerning the proper baseline to use diminishes if an incremental budget approach is used. For example, since the amount of safety that a firm builds into a car is not likely to change significantly from one year to the next in the absence of new regulations, the baseline changes little. Finally, the problem that a regulatory budget would create incentives for agencies to ban products and processes to avoid budget constraints on direct spending should be less a problem than the current tendency to impose both direct and indirect costs on the public, with no regulatory budget constraint. Furthermore, simple rules could be agreed upon to create proxies for indirect costs that would mitigate the incentive for bans. One way to get started is to begin implementation of an accounting system that makes use of what information is available. The fiscal budgetary process evolved from the Treasury Act of 1789 through continual refinement over the years. It was not until the Budget and Accounting Act of 1921 that a comprehensive Federal budget system was established. This Act established the institutional framework for the President to prepare a budget for the United States Covernment as a whole. The new framework included the Bureau of the Budget, to assist the President in the preparation of the budget, and the General Accounting Office, to assist the Congress in carrying out its legislative and oversight responsibilities. Since 1921, the accounting principles, standards, and forecasting methodologies for the OVERVIEW budget have continued to evolve and improve as a result of both executive and legislative action. In a similar manner, though on a much smaller scale, the information-collection budget, a mechanism to manage the Federal collection of information from the public, has evolved over time, with refinements to its accounting and estimation procedures. The Federal Reports Act of 1942 first established the requirement for agencies to measure and control their paperwork burdens. Executive Order No. 12174, "Paperwork," issued November 30, 1979, required agencies to plan and budget total paperwork burden in a manner analogous to fiscal resources. The Paperwork Reduction Act of 1980 directed the Office of Management and Budget to establish general policies and procedures for controlling information collections and to report to Congress each year the estimated "burden hours" imposed by each Federal agency. That Act and its 1986 amendments set paperwork-burden reduction goals. Over the last 11 years of administering the information-collection budget, the paperwork-burden measuring and accounting systems have evolved and improved. REGULATORY COST CEILINGS One approach to developing a regulatory budget would be to follow the models of the fiscal budget and the information-collection budget and require a "regulatory cost ceiling" in any new legislation that imposes private-sector regulatory costs. Under this approach, each new statute would include a ceiling on the total private-sector, State, and local costs that agencies could impose in implementing the statute through regulation. Agencies would then track the estimated costs imposed by the regulations. Once the statutory ceiling was reached, imposing further regulatory costs would require either additional legislation to raise the ceiling or offsetting changes in other regulations that would keep total private-sector regulatory costs within the ceiling. Establishing regulatory cost ceilings would give Congress and the agencies more incentive to make accurate estimates of the likely costs of regulation than simply requiring economic impact estimates for proposed rules. Regulatory cost ceilings that were excessively low would frustrate the purpose of the statute, because agencies could not issue implementing regulations. Although Congress might be tempted 7 to authorize excessively generous amounts, it would have to declare itself willing to impose a specific level of costs on the public. Moreover, estimates developed during the course of congressional debate on legislation would have a real effect on agency decisionmaking, and would give agencies strong incentives to choose regulatory approaches that would produce benefits at the least possible cost. Throughout the course of developing and debating proposed revisions to the Clean Air Act, the Administration identified the costs of its proposals, and indicated that it would not approve requirements that significantly exceeded these costs. The Administration is now committed to tracking the long-term costs of the 1990 Amendments to ensure that the implementation of the law conforms with these expectations. In addition, a pilot regulatory budget for several industrial sectors covered by the new Clean Air Act is being tested by the Administration. The Environmental Protection Agency (EPA), the Council of Economic Advisors (CEA), and OMB are working cooperatively to establish nonbinding cost caps for certain industrial categories. EPA, consulting closely with OMB, will attempt to keep the projected costs of proposed regulatory options below the adopted "budget." EPA and OMB will work together to resolve the complex issues of identifying and estimating baseline costs. If this experiment proves successful, it can be expanded, with stronger constraints, to other regulatory areas, or perhaps extended to the legislative arena. A SUPERBUDGET If a regulatory budget process is adopted, it could eventually evolve toward structural similarity with the fiscal budget apparatus: the President proposing it; the Congress enacting it into law; and the Executive Branch implementing it and operating within its constraints. Institution of a regulatory budget could then lead to yet another evolutionary stage-integration of the two types of budgets into one "superbudget" that rationalizes and controls both direct government spending and private-sector regulatory spending mandated by the Government. Since the economic tradeoffs between the two types of government spending have similarities, such a superbudget could lead to a more efficient and fair public use of private resources. 8 REGULATORY PROGRAM OF THE UNITED STATES GOVERNMENT Regulating Risk: The Cost-Effectiveness of Federal Efforts To Reduce Health and Safety Risks Protecting and enhancing human health and welfare has long been an essential purpose of government. The Federal Government has, however, since World War II taken on an ever-expanding list of responsibilities focused on improving public health and reducing risks of death and injury. As one means of accomplishing these goals, the Federal Government promulgates regulations that compel private parties, including State and local governments, to dedicate resources to the protection of health and safety. It regulates the discharge of pollution that may harm human health and the environment. It administers comprehensive regulatory programs to assure the safety of the food people eat and the pharmaceuticals that they rely on to make them well. It establishes extensive safety standards across the full range of transportation technologies, from automobiles and aircraft to roads and railways. From the shop floors where people work to the consumer products they use, the Federal Government has a visible and authoritative presence that is grounded on the conviction that life and health are highly valued resources. This rapid expansion of Federal involvement in protecting public health and safety has not been achieved without cost. Indeed, the American people bear a burden that totals billions of dollars each year to obtain these benefits. Although attempts to measure the total costs and benefits of health and safety regulation are necessarily fraught with difficulty, the available estimates are instructive and sobering. The recent study by Hahn and Hird, mentioned above, which attempted to pull together many individual benefit and cost estimates, places the costs of Federal health and safety regulation at between $78 billion and $107 billion as of 1988. Because signifi cant social costs were not counted in this estimate, it is primarily useful as a lower bound." Figure 1 summarizes the three major categories of aggregate costs and benefits assembled by Hahn and Hird. Logarithmic scales have been used on both the cost (horizontal) and benefit (vertical) axes. Thus, doubling the distance from the origin implies a tenfold increase in benefit or cost. The width of each rectangle represents the range in cost estimates; the height of each rectangle captures the range in benefit estimates. The elongation of each rectangle indicates the relative magnitude of uncertainty in estimating aggregate benefits and costs. As can be seen from the height of the rectangles along the vertical or benefits axis in figure 1, this uncertainty is particularly great in estimating benefits. More accurate estimation requires the establishment of common units of measurement that enable risk-reduction benefits and costs to be compared across a range of options. Figure 1 shows that highway safety regulation has, on average, provided substantially more benefits than costs. Hahn and Hird estimate benefits of $25 billion to $46 billion per year and costs of $6 billion to $9 billion per year. In contrast, Hahn and Hird estimate that regulations aimed at reducing occupational safety and health risks have imposed costs of about $9 billion per year, but offered negligible risk-reduction benefits. Hahn and Hird's results are mixed for environmental regulations. They estimate costs of $55 billion to $78 billion per year and annual benefits of $16 billion to $136 billion. The horizontal line in figure 1 represents Hahn and Hird's "best estimate” for annual benefits of $58.4 billion. As the demands mount for ever-increasing levels of safety and new or expanded Federal program Robert W. Hahn and John A. Hird, "The Costs and Benefits of Regulation: Review and Synthesis," Yale Journal on Regulation, Vol. 8, No. 1 (Winter 1991), pp. 233-278. As noted above, Hahn and Hird estimate total costs of Federal regulation at between $327 billion and $401 billion. The costs of Federal health and safety regulation are, of course, a subset of these costs. 8 Hahn and Hird offer several important caveats to guide the interpretation of these data. First, their cost estimates do not include the indirect effects of regulation on innovation, particularly where regulation mandates specified technologies instead of specified performance standards. Second, they recognize that the trend is toward increasing use of regulation to achieve health and safety objectives. Third, aggregation tends to conceal many regulations which do not appear to be cost-effective. See Hahn and Hird, op. cit., p. 259. In addition, the analysis by Hahn and Hird does not include certain important areas of safety regulation (e.g., airline travel) and health regulation (e.g., food additives). Hahn and Hird also acknowledge that their estimates for environmental regulation do not include hazardous waste site cleanup, nor do they include other major regulations under development (e.g., municipal solid waste landfill standards). See Hahn and Hird, op. cit., p. 254, especially footnote 79. Finally, their estimates necessarily exclude regulations expected under significant new statutory enactments (e.g., the Clean Air Act Amendments of 1990, the Americans With Disabilities Act). Hahn and Hird's estimates for environmental regulation differ considerably from other recent figures. For example, the Environmental Protection Agency recently estimated that the annual regulatory costs in the environmental sector alone amounted to $98 billion in 1987 (1990 dollars). EPA forecast annual costs to grow to $120 billion per year by 1990, and as much as $179 billion per year by 2000. See EPA, Environmental Investments: The Cost of a Clean Environment (July 1990 draft), p. ES-vi. EPA's cost estimate for 1987 exceeds Hahn and Hird's upper-bound cost estimate by $20 billion (26%). commitments, policymakers and the public alike have begun to ask whether the Nation's resources are being invested wisely. Responsible stewardship demands that the Government carefully examine its regulations to ensure that the American people obtain the best possible return on their investment. This means taking a hard look at the objectives of Federal regulations to verify that they are appropriate, and to strive to achieve these objectives in the most cost-effective manner. This section analyzes the cost-effectiveness of a substantial number of Federal regulatory decisions. In this context, cost-effectiveness refers to the aver age amount of societal resources expended to obtain a fixed amount of societal benefit-in this case, the prevention of injuries and premature deaths. This review suggests that the cost-effectiveness of Federal regulation aimed at reducing human health risks varies enormously. These variations can be seen both across and within Federal agencies according to the nature of the risk regulated, and over time. REASONS FOR GOVERNMENT INTERVENTION TO REDUCE RISK Risk is an essential part of life. The rewards and penalties that flow from risk-taking are at the heart |