Cultural and Institutional Barriers to Smarter Regulation

Good morning, Chairman Graves, Ranking Member Napolitano, and members of the subcommittee. Thank you for inviting me here to testify today.

I am an economist and senior research fellow at the Mercatus Center, a 501(c)(3) research, educational, and outreach organization affiliated with George Mason University in Arlington and Fairfax, Virginia. I’ve previously served as a senior economist at the Joint Economic Committee and as deputy director of the Office of Policy Planning at the Federal Trade Commission. My principal research for the last 25 years has focused on the regulatory process, government performance, and the effects of government regulation. For these reasons, I’m delighted to testify on today’s topic.

I’d like to discuss three broad problems with the US regulatory process, explain how those problems cause undesirable results, and suggest some solutions that focus on altering the underlying culture and incentives that gave rise to the problems. The three problems are: (1) a regulatory focus on activities and outputs, rather than results; (2) significant deficiencies in the underlying analysis that is supposed to inform regulatory decisions, and (3) “ready-fire-aim” rulemaking. This written testimony summarizes these points. Further explanation, along with supporting empirical evidence, can be found in the research papers cited in this testimony.

1. The Regulatory Process Tends to Focus on Activities and Outputs, Rather Than Results

Prospective regulations should be judged based on whether they are likely to produce significant benefits that improve Americans’ quality of life. Existing regulations should be judged based on whether they actually produce those benefits. But, all too often, the federal government focuses on regulatory activities and outputs, rather than regulatory outcomes.

This malady manifests itself in several ways. 

First, regulatory agencies often act as if enforcement is more important than compliance or achievement of regulatory outcomes. For example, a colleague of mine who was advising the enforcement division of a regulatory agency on performance management in the early 2000s found that the enforcement officials objected strenuously to being held accountable for anything other than the level of enforcement activity and their win/loss record in enforcement cases. Effective performance management holds employees accountable for outcomes that benefit the public, not levels of activity.

Second, regulatory agencies often act as if their job is to produce regulations, rather than to produce outcomes. As one agency economist noted, “Success is putting out 10 regulations a year, and bigger regulations are bigger successes. They don’t say, ‘We examined 10 [situations], and we decided that 8 did not warrant regulation.’” Pay, bonuses, career advancement, and recognition go to staff who successfully complete regulatory proceedings. Scholarly research on implementation of the Government Performance and Results Act has found that regulatory agencies are less likely than other agencies to have outcome-oriented goals.

Third, there is little systematic effort to evaluate the actual results (benefits and costs) of regulations after they are implemented and reassess whether a regulation should be eliminated or modified. As a result, new regulations accumulate on top of old ones, even if some of the old ones no longer achieve their intended purposes or do so only at high cost. Patrick McLaughlin, director of the Mercatus Center’s Program for Economic Research on Regulation, and coauthors have estimated that the increase in the US regulatory burden since 1980 may have reduced gross domestic product by as much as $4 trillion in 2012, or about $13,000 per person. Presidents customarily ask regulatory agencies to find regulations that could be eliminated or updated, and these efforts produce some useful results. But such reviews rarely assess whether existing regulations are producing the intended results and at what cost. A study commissioned by the Administrative Conference of the United States indicates that the results of the most recent round of retrospective reviews in the Obama administration are typical in this regard:

The vast majority of status updates on agencies’ retrospective review programs do not include evidence of formal retrospective analysis, such as ex post estimates of benefits, costs, or efficacy. . . . Most of the analyses, such as estimated cost savings from removing regulatory burdens, in agency reviews focus on what can be achieved through reducing paperwork and reporting obligations, or transforming some of these obligations to electronic reporting. . . .

Streamlining the way the government collects information on the actions of regulated firms is fundamentally different than an assessment of whether an economically important rule is delivering on societal objectives identified in authorizing legislation and doing so in a cost-effective and/or efficient manner. 

Several types of reforms would help refocus the regulatory process on achievement of beneficial outcomes for the public, instead of activities or outputs:

• Clearly indicate the desired outcomes Congress seeks to achieve in legislation that authorizes regulations.
• Require regulatory agencies to develop and seek public comment on a plan for retrospective analysis of major regulations at the time these regulations are proposed. 
• Require regulatory agencies to disclose in the Unified Agenda of Federal Regulatory and Deregulatory Actions when they have decided not to issue new regulations, so they can receive credit and be accountable for these decisions.
• Focus congressional oversight on regulatory outcomes. Challenge agencies to provide credible evaluations showing whether regulations have achieved the intended outcomes, and at what cost. Oversight that focuses on levels of enforcement activity, production of regulatory outputs, or anecdotes provides little or no information about whether regulations are achieving the desired outcomes.
• Base budgeting for regulatory agencies on evidence that regulations are achieving their intended outcomes, not measures of activity or outputs.
• To the extent possible, require agencies to reward managers and employees based on regulatory results, rather than activities or outputs. At a minimum, prohibit regulatory agencies from rewarding personnel based on the number or size of regulations they produce.
• Promote independent retrospective assessment and revision of regulations by creating one or more expert commissions modeled on the Base Realignment and Closure Commission. 
• To inform congressional decision-making about authorization, oversight, reauthorization, and budgeting, establish a unit within the Congressional Budget Office or Government Accountability Office whose job is to assess the benefits and costs of regulatory legislation and agency regulations.

2. The Underlying Analysis That Is Supposed to Inform Regulatory Decisions Often Suffers Serious Deficiencies

President Clinton’s Executive Order 12866 requires that before issuing an “economically significant” regulation, executive branch agencies must understand the nature and cause of the problem they are trying to solve, develop alternative solutions, and assess the benefits and costs of each alternative. Independent agencies sometimes face similar statutory requirements. 

If the agency does not understand the problem and does not examine alternative solutions, the resulting regulations are likely to be less effective than they could be, excessively costly, or perhaps not needed at all. If the agency does not understand what caused the problem, then its estimate of the benefits of the regulation is suspect. How can the agency know that the regulation will produce benefits if it does not know whether the regulation will solve a problem? 

Independent analysis consistently shows that agencies’ assessments of the problem, alternatives, benefits, and costs are often seriously incomplete. For example, out of the 130 economically significant, prescriptive regulations proposed between 2008 and 2013, 48 percent were accompanied by no significant evidence demonstrating the existence, size, or cause of the problem to be solved. For only 19 percent of the regulations did the agency consider a wide range of regulatory approaches or levels of stringency. Only 22 percent of the regulations were accompanied by reasonably thorough evidence that the regulation would achieve the intended benefits or other desired outcomes. 

About 56 percent were accompanied by reasonably thorough estimates of compliance expenditures, but compliance expenditures are not the only cost of regulation. Agencies considered the effects of the regulation on prices for only 33 percent of the regulations, and they identified costs that stem from changes in behavior for only 12 percent of the regulations. Thus, costs are likely underestimated for the majority of regulations.

These deficiencies in analysis usually occur regardless of whether Congress has given the agency broad or narrow authority to decide whether to regulate, what form the regulation should take, how stringent the regulation should be, or who must comply. Several changes in the regulatory process would help remedy these deficiencies:

• Statutorily require all agencies to conduct regulatory impact analysis for regulations with economic effects exceeding a certain threshold, such as the $100 million per year threshold used in Executive Order 12866. Indicate in the statute the topics the analysis must cover: assessment of the existence, extent, and cause of the problem; development of alternatives; and assessment of the benefits and costs of alternatives.
• Expand the resources and personnel of the Office of Information and Regulatory Affairs so that it can conduct more thorough reviews of regulations and the accompanying analysis from independent and executive branch agencies.
• Allow courts to review an agency’s analysis to ensure that it covers the required topics and employs the best available evidence in the record. The court could set aside the regulation only if an error or omission in the analysis materially affected a decision about the regulation.

3. Agencies Engage in “Ready-Fire-Aim” Rulemaking

Even when regulatory impact analysis (or other economic analysis) is required by executive order or by law, regulatory agencies often make decisions first, then craft the analysis to support decisions that were already made for other reasons. This “ready-fire-aim” approach to rulemaking puts agencies in the position of selecting a regulatory option before they know whether there is a problem that regulation could solve, what caused the problem, or which solution might be most effective and efficient. 

Several reforms could help raise the odds that agencies conduct a more complete and objective analysis before they make major regulatory decisions:

• Require agencies to consult with stakeholders before writing major regulations.
• Require agencies to publish for public comment their preliminary analysis of the problem, and the benefits and costs of each alternative they are considering, before they select a preferred approach and write a regulation.
• Within agencies, free economists and other analysts to conduct objective analysis by locating them in a unit other than the program office that writes regulations. Have analysts report to and be managed by other analysts, with clear criteria for career advancement based on the quality and objectivity of their analysis.

Far too much of the regulatory debate is an unproductive screaming match about intentions, rather than a reasoned discussion of results. Experience has shown that intentions do not automatically produce results. For this reason, significant institutional and cultural changes are needed to refocus regulatory decision-making on regulatory outcomes, improve the quality of analysis that informs regulatory decisions, and ensure that agencies get the facts before they write regulations.