Without estimates of the costs of regulation there could be no regulatory budget. Yet the art of estimating regulatory costs is not fully developed. The existing detailed cost studies encompass only a handful of agencies, and the cost estimates in those studies apply only to a specific industry or group of industries. There is as yet no generally accepted convention regarding the cost elements to be included or excluded, nor is there a single methodology for cost calculation which has been generally applied.

Nevertheless, the budgeting of regulatory compliance costs would be possible within the present state of the art.  The history of the Federal fiscal budgeting process demonstrates that budgetary control can be achieved with far less sophisticated techniques than are being used today. The very existence of a regulatory budget would, of course, provide an incentive to develop better procedures for estimating the costs of regulation.

This chapter examines the measurement of regulatory compliance costs.[1]  The concept of compliance costs is defined, and the desirable criteria for evaluating measures of compliance costs are discussed.  Then the estimation of compliance costs for existing vs. proposed regulations and the assignment of compliance costs to appropriate fiscal years are explored. The chapter concludes with a survey of the current state of the art of estimating regulatory costs.  

A.        Definition of Compliance Costs

Compliance costs consist of expenditures made expressly to meet the requirements of Federal regulations. The expenditures may be made by the private sector, by state and local governments, or by other Federal agencies. In principle, only incremental costs due to regulation—that is, costs in excess of what would have been spent without regulation—should be counted in compliance costs. As noted below, in practice the precise isolation of incremental costs is difficult.

Both capital and operating costs may be incurred in complying with regulations.  Examples of capital costs of compliance include outlays for extra construction or new equipment. Examples of operating costs of compliance include expenditures on added research-and-development, extra variable inputs (such as labor and raw materials), additional supporting services, and further administration (such as paperwork).

Weidenbaum and De Fina have estimated that, in 1976, total compliance costs in U.S. industry were some twenty times greater than Federal administrative costs. For some regulations specific to certain industries, the estimated ratio was greater than 50:1.[2]  To date, no estimates have been made of the corresponding ratio of compliance to indirect costs.

Determining exactly what to include in the compliance costs to be charged against agencies' regulatory budgets would not be a straightforward matter. To illustrate the difficulty:

·        Capital outlays on girls' locker rooms under Title IX and the expense of seeing that truck drivers keep proper time logs for the Interstate Commerce Commission should obviously be counted as compliance costs.

 

·        Less clear would be the cost of employing extra workers as counterparts of Federal inspectors, as the oil companies claim they must do to comply with crude-oil price-control rules.  

 

·        Controllable expenses like lobbying in the Congress or maintaining an office in Washington to monitor the regulatory agencies most probably would not be allowed as compliance costs.

B.        Criteria for Evaluating Measures of Compliance Costs

A regulatory budget would play an important role in the policy making process. For that reason, the quality of the measures of compliance costs used in the budget would not be a matter of indifference.

What is acceptable quality is ultimately a political matter that cannot be resolved in this study. It is useful, however, to suggest certain criteria that would affect the quality of compliance-cost measures and that would therefore provide points of reference against which to evaluate particular measures and the methods used in obtaining them.

The following list of criteria also indicates possible directions for further work on the methodology of measuring compliance costs.

1.         Precision

A workable regulatory budget would have to employ point estimates of specific compliance costs. The closer those point estimates were to actual costs, the greater the success in husbanding the resources claimed by Federal regulation—the very purpose of a regulatory budget. In part, precision would be a function of time: more precision would be possible the longer a given regulation had been in effect. Perfect precision, of course, would be unattainable; put differently, perfect precision would be far too costly to strive for.  Note, however, that even the easily quantifiable dollar magnitudes in the fiscal budget, after years of evolution, are not precise: estimating errors of millions of dollars are common.

2.         Consistency

In a workable regulatory budget, individual compliance costs would have to aggregate into totals (by agency, sector, or total budget) that reflected the resources claimed by efforts to conform with regulatory requirements. In other words, a million dollars' worth of costs should represent the same claim on economic resources in one agency or sector as in the next.  If that were not the case, the incentives created by a regulatory budget would be distorted, and the rationale for such a budget would be weakened.

The methodology used to estimate compliance costs would have to be consistent across different regulations and across the various regulated sectors of the economy. Note that striving for greater consistency might lead to a loss of coverage incompliance-cost estimates if one element of costs could be measured in some areas but not in others.

3.         Transparency

An important quality of compliance-cost measures would be that anyone, using the same data and methodology, could duplicate the estimates used in a regulatory budget. To permit duplication, both the data and the methodology would have to be transparent to all parties concerned. The procedures and assumptions of the methodology would need to be visible and well documented.[3]  The data, once decided upon, would have to be readily available and subject to outside, independent audit. Note that the audit requirement could pose problems of privacy, particularly for business firms.

4.         The Cost of Making the Estimates Themselves

The very process of obtaining measures of compliance costs would itself entail costs. Data collection, processing, analysis, and dissemination all require the use of scarce resources. If the cost estimation itself was judged burdensome (a matter for political judgment), it would constitute a powerful argument against adopting a regulatory budget. Obtaining compliance cost measures would probably be most burdensome when a regulatory budget process was first introduced.

The cost of making the estimates would not, of course, be independent of the degree of precision or consistency required of measures of compliance costs.  For example, it might be necessary to sacrifice some precision in order to reduce the burden of making the cost estimates.

C.        Compliance Costs of Existing vs. Proposed Regulations

The logic of a regulatory budget would require measures of the compliance costs of both existing and proposed (new or revised) regulations. Only with both existing and proposed regulations covered would the discipline of the budget cover the total costs of complying with Federal regulation.  It is important to recognize that the decision to leave an existing regulation in place for another budget period has the same cost impact as the decision to introduce a new regulation with equal compliance costs. Also, the costs of complying with existing regulations would have to be known if agencies were to receive budget credits for regulations that they removed (as was suggested in Chapters 1 and 3).

It was noted earlier, however, that obtaining compliance-cost measures for all the regulations promulgated prior to the adoption of a regulatory budget would be a formidable task. It would be impossibly costly to achieve in the first few years of operation, at least with any degree of precision, consistency, and transparency. Thus, a gradual approach would be required—perhaps one that attempted estimates only for broad categories such as major existing programs, or even entire agencies, in the first few years after adoption. With time, however, more detailed measures of the compliance costs of existing regulations could be developed. Once acceptable base-cost measures had been obtained, annual updates would be relatively straightforward.[4]

Measuring the compliance costs of proposed (new or revised) regulations would pose much less of a problem than existing regulations, in terms of the volume of information and computation involved. Proposed regulations would, however, have estimation problems of their own. Unlike existing regulations, for which empirical track records could (given enough effort) be compiled, the costs of complying with proposed regulations would by definition have to be estimated on the basis of few hard data.[5] 

Preparing estimates of the compliance costs of proposed regulations would require two distinct steps. First, it would be necessary to forecast the probable methods of compliance. Second, constructive cost estimates would need to be prepared for those methods.  Of the two steps, the first would appear by far the more challenging and difficult. It would probably require some active involvement by the parties affected by the regulations. Such involvement could take the form of voluntary comments on the agencies' preliminary forecasts, or of more formal analysis or audit.

In the case of larger business firms and governmental units, there is precedent for such involvement under the present system. For small businesses and governments, and for consumers, however, there is little precedent. The exact nature of active involvement by fragmented, unorganized individuals is not easy to envision. Representation of consumers by public-interest groups would be one possibility; although it would be fraught with possible objections.  And as suggested in Chapter 3, small business could be represented by trade associations.

D.        Assigning Compliance Costs to Appropriate Fiscal Years

The costs of complying with a regulatory are requirement typically incurred over a number of years. Moreover, the time pattern in which compliance costs are incurred bears no simple relationship to the fiscal year in which the requirement is promulgated. Indeed, some costs (especially investments) may actually anticipate (and thus precede) promulgation of a new regulation.  A successful regulatory budget system would have to include all compliance costs within its authorization limits regardless of when they were incurred. In theory, this could be accomplished with a system which charged agencies with the compliance costs of all their regulations in force in the fiscal year in which the costs were incurred.

In practice, however, charging compliance costs as incurred could undermine the goal of using a regulatory budget system as a management tool. Politically-appointed heads of regulatory agencies—who can reasonably expect to be in their jobs for relatively short periods of time—would have little incentive to worry about the future budgetary claims from regulations that they take the credit for promulgating. This could lead to neglect of future compliance costs in decisions on new regulatory requirements.

An alternative to charging compliance costs only as incurred would be to charge an annualized compliance-cost value of each regulation that was in effect. At the time a regulation was promulgated, an agency would estimate the time-pattern of compliance costs, calculate its present discounted value, and convert that value to an annualized (i.e., annuity-equivalent) amount.[6]   In that manner the regulatory budget process would better take into account the total compliance costs over the expected lifetime of a regulatory requirement.

The annualized cost approach would have the disadvantage of masking the economic impact of extraordinarily large costs incurred in one or two years (for example, automobile-industry retooling required for fuel economy and emission standards). By the same token, the failure to use annualized compliance costs would tend to discriminate against regulations that require very large outlays in one or two years, but provide benefits over long periods of time with relatively low additional costs. 

A variant that would combine elements of both of the above approaches would be to require agencies to submit multi-year regulatory cost estimates along with their annual requests for compliance cost allocations. There is precedent for this variant in fiscal budgeting for costly programs extending over several years (e.g., weapons systems).  The multi-year cost estimates would at least raise the issue of future costs in the formulation of a given year's regulatory budget. It would, however, impose relatively weaker discipline on future costs than the annualized cost approach.

The present analysis suggests that the use of annualized compliance costs would be the best approach to capturing total costs over the lifetime of a regulatory requirement.[7] However, the issue merits further study, both of the defects of charging compliance costs as incurred and of the actual calculation of annualized costs.

E.        The State of the Art in Estimating Regulatory Costs

It is useful to conclude the discussion of measuring compliance costs with a survey of existing cost of regulation. Some 70 studies have been examined.  The methodologies used in those studies range from guesses to detailed, highly structured cost accounting.  The existing studies address primarily administrative costs and certain elements of compliance costs.  Although some of the studies comment on the existence of indirect costs, only a few quantitative estimates of indirect costs have been made.

Table 1 gives a breakdown of the studies examined, by affiliation of authors and time period analyzed (historical or future).  A study that addresses both future and historical costs is counted in both columns.  Independent investigators have been responsible for more than half of the historical studies, but this group has attempted almost no future estimates. This is not surprising in view of the previously-mentioned need to forecast the method of compliance before estimates of future costs can be made. Business firms are better able than independent investigators to make forecasts of compliance methods, and the government can require such forecasts from industry.

                                                Table 1

                        Existing Regulatory Cost Studies

 

Affiliation                                            Historical        Future

of Author                                             Costs              Costs

 

Government (or Contractor)               17                     15

Industry (or Contractor)                        6                       5

Academic Institutions                          27                     2

There is a glaring lack of uniformity in the costs included in or excluded from these studies, as well as in the methods used to account for those costs. Hence it is difficult to draw comparisons. However, several pairs of studies conducted by the two opposing parties to a specific debate suggest the range of discrepancies that can occur.  Here are three examples:

1.         In 1978 the Consumer Product Safety Commission estimated the direct compliance cost of a proposed fabric flammability regulation for the furniture industry at $57 to 87 million a year; the market impacts would be a two to three percent increase in the wholesale price of furniture and added consumer costs of $144 million per year. The American Textile Manufacturers Institute, in contrast, estimated direct compliance cost at $1.3 billion per year.

 

2.         In 1978 the Environmental Protection Agency estimated the annual direct compliance costs of a proposed ambient air quality standard for ozone at $6.9 to 9.5 billion per year. The Council on Wage and Price Stability, using a different but equally logical methodology, estimated those direct compliance costs at $14.3 to 18.8 billion per year.

 

3.         In February 1979, the Environmental Protection Agency estimated the costs of a proposed 1981 diesel engine particulate standard at a negative $160 per ton of particulates removed. The Council on Wage and Price Stability estimated this cost at a positive $4740. For a proposed 1983 standard the estimates per ton were $3200 and $7650, respectively.

There are two explanations for such wide discrepancies in estimated compliance costs. First, there are currently no generally accepted conventions for choosing the specific costs to be included or for the methodology of computing total compliance costs. Second, because each of the above examples concerned the impact of a proposed regulation, it was necessary to make assumptions about the methods to be used for compliance. The assumptions of the opposing parties were quite different in each case. The existing studies of historical costs, where the methods of compliance were known, show much smaller discrepancies.  An important contribution to the literature on regulatory costs was made by the first issue of the Regulatory Calendar, which, appeared, in the Federal Register, February 28, 1979. The Calendar (which will be published semiannually by the Regulatory Council) listed 109 major rules being considered by twenty Federal departments; compliance cost estimates were included for about one third of the entries. 

In a statement at the time of publication, Douglas Costle, the chairman of the Regulatory Council, acknowledged that “agencies presently calculate costs in different and sometimes conflicting ways," and he cautioned against attempting to aggregate the costs published in the calendar.  He called for development of better cost-estimating methods and for dissemination of those methods among the agencies.  Nevertheless, the work currently being done by the Council may help lay the groundwork for making the cost estimates than would be needed to implement a regulatory budget system.

Significant progress toward resolving the problems of estimating regulatory compliance costs, particularly settling upon a uniform methodology, was made in a study commissioned and recently released by the Business Roundtable.[8]  That study, conducted by Arthur Andersen and Company, estimated the costs incurred in 1977 by 48 cooperating companies in complying with all the regulations of six agencies or statutes:

1. Environmental Protection Agency

2. Occupational Safety and Health Administration

3. Equal Employment Opportunity Commission

4. Department of Energy

5. Employee Retirement Income Security Act

6. Federal Trade Commission

The 48 participating companies represented 23 two-digit SIC industry groups; each group contained a minimum sample of three company divisions or other operating units.  While the numerical results of the Arthur Andersen study themselves are of interest, much the more interesting for present purposes is the methodology developed for the study. The procedures for consistently determining compliance costs across firms aid regulations, together with the supporting documentation from the companies, would provide the kinds of precise, consistent, and transparent cost estimates required for a workable regulatory budget. Of particular note is the assiduous care taken to determine the increment in company costs due solely to complying with regulations. The methodology has been criticized by some economists for failing to include indirect as well as compliance costs; for purposes of a regulatory budget, however, the criticism is beside the point, as argued in Chapter 1 of this study.

Arthur Andersen and Company estimated that the study itself cost 0.4 percent of the computed compliance costs.  Subsequent studies made by the same participants would probably reduce that percentage because of the experience gained in the initial study.

The studies surveyed above illustrate a point made at the very outset of this chapter—namely, that the art of estimating regulatory costs is still undergoing development.  At the same time, a handful of those studies—in particular, the Regulatory Calendar and the Arthur Andersen study for the Business Roundtable—also illustrate a second point made at the outset: a workable regulatory budget would be achievable within the present state of the art.