Every U.S. president in modern history has required that regulators conduct a cost–benefit analysis as part of the promulgation of “economically significant” regulations. Major new regulatory proposals by executive branch agencies must be supported by a Regulatory Impact Analysis (RIA), a requirement introduced in President Ronald Reagan’s Executive Order 12291. Each RIA’s content and methodology are reviewed by the Office of Management and Budget (OMB) to ensure quality. According to the OMB, the goals of an RIA are “(1) to establish whether federal regulation is necessary and justified to achieve a social goal and (2) to clarify how to design regulations in the most efficient, least burdensome, and most cost-effective manner.”

From both a theoretical and practical perspective, an RIA is a valuable tool because it helps to improve regulation. For example, the Environmental Protection Agency’s phase-out of lead in gasoline—a significant public health success story—was informed strongly by cost–benefit analysis.

However, not all regulatory agencies produce RIAs as part of major rulemakings. So-called “independent” agencies (e.g., the Federal Reserve, Federal Communications Commission, and Securities and Exchange Commission) are not subject to EO 12291. Those agencies are responsible for approximately 20% of all major rules. Even for agencies covered by the order, RIAs have only a limited effect on rulemaking. The analyses are rarely revisited to assess their accuracy after a regulation goes into effect. A handful of academic studies show that agencies can significantly under- or overestimate actual costs and benefits. The biggest problem with the RIA process, however, relates to something fundamental: objectivity. A regulatory agency is not unbiased. It has every incentive to develop RIAs that support its preferred regulatory approach. And because regulatory agencies are designed to regulate, RIAs seldom (if ever) conclude that federal regulation is not needed.

Fortunately, the OMB can—and often does—hold agencies accountable for deficient RIAs. But the OMB is not the optimal watchdog because it reports to the president and therefore is subject to political decisions that are not always consistent with objective analysis of regulatory effects. In addition, the OMB does not opine publicly on the quality of agency analysis.

Over the years, observers of this process have suggested many ways to improve the quality of RIAs. These suggestions fall into a few general categories: requiring greater or earlier OMB review of agency analysis, requiring a more searching judicial review of agency analysis, or establishing a new federal agency to conduct RIAs on behalf of regulatory agencies.

Each of these proposed solutions has its merits. But each also has its critics, and their arguments (e.g., delays in achieving public protections, higher cost to the government) have been sufficient to prevent reform. There is an additional argument that lies just below the surface: reform would change the balance of power among the three branches of government, and no branch will support a lessening of its influence.

A proposal / How, then, can we improve the objectivity of regulatory analysis? We can start by considering the source of the problem. Federal regulatory agencies have a monopoly position on the production of RIAs. There is no competition and—just like a market where a single producer controls supply—the result is an insufficient quantity of insufficient quality at too high a price.

To inject needed competition into the RIA market, the government could leverage external expertise on cost–benefit analysis. Specifically, consider the following proposal: During the public comment period on a proposed major rule, if an agency receives a public comment in the form of an RIA for the proposed rule, then it must submit that public comment to the OMB. The OMB would then determine whether the submitted RIA comports with its established guidelines on regulatory analysis (as outlined in OMB Circular A‑4, for instance) and, if not, provide an explanation for why the RIA falls short. The OMB must then send its determination back to the agency, which must include it in the rulemaking record as part of the agency’s final action.

Although this proposal can be described in just four steps (the public submits an RIA, the regulatory agency submits the public comment to the OMB, the OMB makes a determination, and the regulatory agency includes this determination in the rulemaking record), some points are important to stress:

  • The agency should not be allowed to evade its requirement to share such a public comment with the OMB. It must send the comment to the OMB within a prescribed period of time (say, within two weeks of receipt).
  • The OMB should not evade its responsibility to make a yes/​no determination or opine why a submitted RIA does not comport with its established guidelines. A terse and indecipherable explanation from the OMB would do little to foster improved quality.
  • The threat of judicial scrutiny is critical. If this policy were established via presidential executive order, it would not allow a private party to sue a recalcitrant agency for failing to comply. Therefore, the policy is best established through an act of Congress.

Such a policy would be considered a “nudge” in the sense used by Richard Thaler and Cass Sunstein. As Sunstein wrote in his 2013 book Simpler: “An analysis of costs and benefits is an important way to nudge regulators. Indeed, requiring such analysis is a way of creating a good choice architecture for those who make the rules.”

Getting the incentives right / One way to evaluate this proposal is to consider the incentives of such a requirement on each of the major actors in this drama: the regulatory agency, the OMB, the courts, and the public.

To inject needed competition into the RIA market, the government could leverage external expertise on cost–benefit analysis.

Consider a regulatory agency that is not required to conduct an RIA for a proposed major rule. The threat of a publicly submitted RIA is very concerning. Perhaps the proposed rule, which the agency has taken years to develop, does not fare well when subject to an objective RIA. In such a case, the agency risks losing in court because its failure to undertake a cost–benefit analysis is more likely to be seen as arbitrary and capricious by a judge who examines the rulemaking record, including the positive OMB determination on a publicly submitted RIA. This possibility—that a court may overturn a rule not supported by objective analysis—is likely to increase the odds that a non-covered regulatory agency will conduct its own high-quality RIA.

An agency covered by the RIA requirement already must develop its own analysis, but it will also be concerned. Should the OMB make a negative determination on a publicly submitted RIA that mirrors closely the agency RIA, then the agency will have a strong incentive to address the deficiency in its own RIA before it issues a final rule. Alternatively, should the OMB issue a positive determination on a submitted RIA that calls into question the proposed rule, the agency will have a strong incentive to alter its proposed rule as it crafts a final rule.

Now consider the role of the OMB. Its economists have been evaluating the quality of agency RIAs for more than 35 years. They know how to do this and they can quickly determine whether a publicly submitted RIA comports with established guidelines. (An experienced OMB economist or desk officer can spot a poor-quality RIA very quickly.) But let’s consider a situation where the public submits an RIA that mirrors an agency RIA except in just one aspect, and that aspect represents a vast improvement in quality over the agency RIA. No matter the OMB determination (positive or negative) on the quality of the original RIA, the OMB will be signaling to the agency to improve its own RIA. An agency would be foolish to ignore this signal as it works on a final rule. So even though the OMB will be given a new task, it will also be given additional power—not a bad tradeoff from an OMB perspective.

The courts should also benefit. Judges are seldom trained as economists versed in the intricacies of cost–benefit analysis. They are ill-equipped to review an RIA in the rulemaking record to determine if an agency acted in an “arbitrary or capricious” manner. But judges are very good at determining whether an agency followed procedures required by statute. A judge is likely to give weight to an OMB determination, especially if it calls into question the merits of an agency’s final rule.

Last but not least, consider the public. It is not unusual for public comments to criticize an agency RIA. With the proposed law in place, the public can make this point with greater emphasis because it might have concurrence from the OMB. But the public might also receive a rebuke from the OMB. This possibility, along with the cost of developing an RIA, will temper would-be public commenters such that the number of submitted RIAs is likely to be relatively low.

Consider the analogous case of the Information Quality Act of 2001. That legislation allows the public to submit a correction request to an agency for disseminating information that does not meet certain quality standards. Critics thought agencies would be flooded with requests for correction; in fact, agencies have received very few such requests.

By examining the incentives the proposed law creates, one concludes that the aggregate effect is to improve upon the quality and objectivity of the RIA. The proposal also avoids problems that plague other proposals to improve cost–benefit analysis. For instance, there is no significant cost imposed on the federal government; the public will bear the cost of conducting an RIA and will do so only if it is in the interest of the commenter. The law will not delay the time it takes for an agency to issue a major rule; the same timeframe will hold. The OMB is not being required to review the rules or RIAs from non-covered agencies, which were created to have some degree of independence from the president. Nor is the judiciary asked to delve into the minutia of a cost–benefit analysis. The balance of power between the branches of government will not shift.

It is possible, of course, that this proposal creates unique concerns. For example, estimates of costs and benefits for a major rule are often highly uncertain. In such cases, the agency RIA and a publicly submitted RIA may differ in their conclusions even though both conform to established quality guidelines. But is this really a major concern? Such a situation makes explicit the uncertainty in an analysis that would otherwise go unnoticed. Uncertainty can be frustrating to regulators seeking a clear choice among regulatory options, but regulators ought to know the risk of making a bad choice. The other actors in our drama (Congress, judges, the public) should also know the certainty under which regulatory decisions are made.

Although this proposal is rather modest in terms of cost, it should have a relatively positive effect on the behavior of regulatory agencies responsible for considering the expected consequences of major rules. In other words, this proposal would pass a cost–benefit test.

Readings

  • Economic Analyses at EPA: Assessing Regulatory Impact, edited by Richard D. Morgenstern. Resources for the Future, 1997.
  • Nudge: Improving Decisions about Health, Wealth, and Happiness, by Richard H. Thaler and Cass R. Sunstein. Penguin Books, 2008.
  • “Regulatory Impact Analysis: A Primer,” published by the Office of Management and Budget. 2003.
  • Simpler: The Future of Government, by Cass R. Sunstein. Simon & Schuster, 2013.