July 7, 2016

Regulatory Budgeting as a Solution to the Accumulation of Regulatory Errors

Testimony Before the House Committee on the Budget

Chairman Price, Ranking Member Van Hollen, and members of the committee: thank you for inviting me. As an economist and senior research fellow at the Mercatus Center at George Mason University, my primary research focuses on the regulatory process—its strengths, its weaknesses, and proposals for improving it. One such proposal is the creation of a budget for regulations.

My testimony contains two main themes:

  1. The regulatory process is not error free, but a regulatory budget could reduce the error rate.
  2. Regulatory budgeting can be implemented in a variety of ways. I will discuss the mechanics of a regulatory budget, including
    1. the unit of measurement,
    2. how budgets could be set, and
    3. who performs the measurement and enforcement.

1. Error Accumulation and an Error-Correction Process

The concept of regulatory budgeting has a long and bipartisan history, with the earliest versions attributed to Robert Crandall of the Brookings Institution and former Texas Senator Lloyd Bentsen in the 1970s. The 40-year history of the idea parallels the consistent recognition by every presidential administration since the mid-1970s that the body of federal regulations is lengthy and complex. 

While the efforts of presidents past and present implicitly recognized that an ever-growing volume of regulations is undesirable, recent theoretical and empirical advances offer more specific evidence as to why. Regulatory accumulation—that is, the buildup of rules over time—has been demonstrated to distort business investment choices, deter innovation, hinder productivity growth, and slow economic growth. The degree to which regulatory accumulation can directly harm the most vulnerable—including low-income households and would-be entrepreneurs with few resources—is even more troubling.

These negative effects of regulatory accumulation might be acceptable if we were confident that every rule in the 178,000-plus pages of federal regulations was effective and efficient, delivering the outcomes intended by Congress’s authorizing statutes at a reasonable cost. But the regulatory process, like any process involving human involvement, is subject to error. Not even the most optimistic among us can believe that our federal agencies have achieved an error rate approaching zero.

Even though the existing regulatory process is not error free, there is no formal mechanism or systematic procedure for the identification and correction of errors. The absence from the regulatory process of required and specific provisions for retrospective analysis and review of regulations means that the errors that do occur become ensconced in the regulatory code and often, in turn, serve as the foundation for subsequent regulations. Regulatory budgeting introduces dynamic elements into the regulatory process that would provide regulatory agencies incentives to identify and correct any existing errors.

Regulatory budgets, like other types of budgets, only work if they force the spender to identify and prioritize the most valuable options. The behavior of an agency with a budget differs from that of an agency without a budget. In today’s no-budget world, an agency’s objective is to fulfill its mission with the promulgation of rules. The effectiveness and efficiency of those rules are not evaluated in hindsight, and prospective evaluation of effectiveness and efficiency only occurs for less than one percent of all new rules. In contrast, an agency with a regulatory budget would act differently. First, the agency would avoid new regulations that would not achieve high benefits relative to their budgetary cost. Second, the agency would have incentive to eliminate old regulations that are found to be ineffective or intolerably inefficient. In other words, a regulatory budget process would resemble an error-correction process: it would lead to fewer new errors as well as aid in the identification and correction of existing ones.

2. The Mechanics of a Regulatory Budget

a. What is measured? Regulatory budgets are designed to account for regulatory activity undertaken by a government. The unit of measurement can vary. Some options include

  • regulatory restrictions,
  • administrative burden,
  • business costs, and
  • social costs.

The simplest unit of measurement is regulatory restrictions. Regulatory restrictions are obligations or prohibitions created by regulations. Measurement of regulatory restrictions depends simply on the text of regulations rather than on any estimation of cost associated with regulations. This method was and continues to be used by the government of the province of British Columbia in Canada. Using the regulatory restrictions method, budgets would set targets for the number of restrictions contained in regulatory code. Each agency could have a set budget. Alternatively, a budget could apply to a portfolio of regulatory programs related to one subject, such as education or railroad transportation. 

A second possible unit of measurement is administrative burden, which is similar to the idea of paperwork burden used in the Paperwork Reduction Act of 1980. Administrative burden typically refers to paperwork and other information-related activities required by regulations, such as the completion and submission of forms or the retention of data. Analyses of regulatory paperwork burden typically assess the number of paperwork hours associated with a rule—a figure that can be monetized using industry-specific wage and overhead data. But neither the estimation of hours nor its monetization is required. It is possible to simply count the number of regulatory requirements that impose any amount of administrative burden, which is the method used in Canada’s ongoing federal Red Tape Reduction Action Plan. 

Another unit of measurement, business costs, focuses on direct compliance costs imposed on businesses. As such, it is a more comprehensive measure of costs than paperwork costs, because it would include other direct costs such as those incurred by purchasing equipment or allocating labor to compliance activities. Business costs are the unit of measurement used by the United Kingdom’s one-in, two-out form of regulatory budgeting. 

Finally, the most comprehensive unit of measurement is social costs, which include those elements of regulations that apply to businesses, such as paperwork burden or compliance costs. But social costs also include the hidden costs that accompany government intervention and that apply to other segments of society rather than to business owners only. Some examples include negative effects on innovation and on entrepreneurship—both critical drivers of economic growth. Regulations, and regulatory accumulation in particular, tend to have deleterious effects on innovation and, in some cases, on competition. Along with Bentley Coffey and Pietro Peretto—both of Duke University—I recently demonstrated a methodology for estimating the effects of regulatory accumulation on innovation and, subsequently, economic growth, which could serve as a model for the development of a regulatory budget based on social costs.

b. How is the limit determined? A regulatory budget must actually constrain agencies in order to be effective. But how constrained should the agencies be? For explanatory purposes, it is useful to consider the methods of determining this limit in two categories: agency-centric and legislation-centric. Agency-centric methods focus on actions of the agencies themselves as if they are independent of other branches of government. Legislation-centric methods focus on authorizing statutes passed by Congress as the inherent wellspring of agency regulation.

These approaches would focus on different governmental actions. The agency-centric approach would focus on the effects of individual regulations emanating from an agency or, in the case of a portfolio approach, on groups of regulations, possibly from multiple agencies, that address the same subject. The legislation-centric approach would focus on congressional bills before their passage, in a process similar to the budget scoring process used by the Congressional Budget Office.

Once the focus—agency or legislation—is determined, there remains the question of determining the limit associated with regulations originating in either agencies or legislation. The regulatory budget processes that have been applied in Canada and the United Kingdom relied on a rule-based approach, such as one-in, one-out or one-in, two-out. The logic is simple: regardless of the unit of measurement, for each new cost added to a regulatory budget, some quantity of costs must be eliminated. The rule-based approach has the merits of simplicity and relevance, because any new intervention (regulatory or legislative) will require some amount of reprioritization. On the other hand, the rule-based approach’s simplicity is accompanied by some degree of inflexibility. A design element to consider with any rule-based approach would be a means for dealing with cases of emergency.

An alternative approach, which I call flexible regulatory budgeting, entails the implementation of a process that is similar to the budget process for outlays. In the agency-centric version of flexible regulatory budgeting, Congress would set regulatory budgets for agencies during the budgeting process, and the regulatory budget would exist as a parallel cost ledger. The regulatory budget could change from year to year, depending on Congress’s perception of agency needs, effectiveness, importance, or other factors. In the legislation-centric version, a bill would be scored by an independent body—perhaps an office of the CBO or a similar entity—to estimate the bill’s anticipated regulatory costs. Upon passage of the bill, these anticipated regulatory costs would be allocated to the agencies mandated to create new regulations under the new law. Subsequent rounds of budgeting would be informed by both prospective and retrospective analysis of agency regulations, allowing Congress to react to the effectiveness of regulations by allocating a larger or smaller regulatory budget associated with the regulations’ authorizing legislation.

c. Who does the measuring? Several studies warn of the conflict of interest that arises when regulatory analysis is performed by an entity that has incentive to create a regulation. The same logic applies in the case of regulatory budgeting: The estimation of the effects of regulations can turn into an exercise in advocacy if there is a conflict of interest. The simple recognition of this dynamic should inform the design of a regulatory budgeting scheme.

Two alternative approaches are generally used to overcome potential conflicts of interest that arise surrounding the performance of regulatory analysis. The first is to assign the task of regulatory analysis to a body that has no interest—political, financial, or budgetary—in seeing a regulation pass or fail. The second is to create a process for oversight of the analyses produced by regulatory agencies. If either of these approaches is applied to regulatory budgeting, a second layer of oversight—from the public at large—could be facilitated by the specification of a transparent and replicable methodology that the analyses must follow. 


A fundamental problem with the current regulatory process is that its design leads regulators to focus on one regulation at a time, rather than the entire body of regulations affecting the same segment of economic activity. Stanford University economist John Cochrane illustrated this problem with a focus on financial regulations:

The thicket of existing and proposed regulation deeply challenges cost-benefit analysis, which is designed to think about one regulation at a time and to cure market failures, not regulatory failures. Financial regulation often works at cross-purposes. One agency’s predatory lending, by which financial companies are accused of forcing borrowing on unsuspecting customers at punitive rates, is another agency’s opening of credit markets to underserved income-based, geographical, or racially defined groups. One agency wants lower loan-to-value ratios in the name of financial stability. Another wants higher loan-to-value ratios in the interest of community redevelopment. . . .

This sort of regulatory contradiction is pervasive. Our government subsidizes and requires the use of corn ethanol to reduce emissions, yet bans the import of sugarcane ethanol, which might actually have that effect. Our government heavily subsidizes solar cell production to lower prices, and then imposes tariffs against cheap Chinese solar cells to raise prices.

Regulatory budgeting could improve, if not resolve, this situation if it were applied to regulatory portfolios, or groups of regulations that touch on the same subject (e.g., financial regulations). Even if regulatory budgets were not applied to portfolios that extended across multiple agencies, the budgeting exercise would force individual agencies to weigh regulatory options against each other before promulgation and to retrospectively assess their effects. Both of these changes would represent vast improvements over the current process.

Regulators and legislators alike are not perfect. Regulations are perhaps unique in the sense that they are undeniably important to all actions in the economy, but are not subject to a process for error correction. These errors—most of which are probably undiagnosed owing to the lack of retrospective analysis—are far from benign. They contribute to regulatory accumulation, a force that disproportionately harms low-income households, deters innovation, and slows economic growth, without delivering offsetting benefits. The reduction of the error rate requires a process that ensures the development and application of high-quality information, both before and after the effects of regulations have been observed. Regulatory budgeting represents one option to achieve just that.

Regulatory budgeting would lead to the creation of better information about the effects of regulations. Simultaneously, it would create incentives for regulators to act upon that information, promulgating those regulations that offer the greatest benefit relative to costs and eliminating regulations that impose an undue burden on the American people.