Interested stakeholders may have a role in quantifying deregulatory impacts for agencies.
Early in his presidency, President Donald J. Trump issued Executive Order 13,771, which requires agencies to identify candidate rules for deregulation. Because agencies have dozens, if not hundreds, of deregulatory options available to them, their selection and justification of specific rules to modify or repeal may receive closer scrutiny from the public and the courts relative to initial regulatory actions. To deregulate successfully, agencies need to find ways to address uncertainty, particularly related to the forgone benefits of deregulation.
Agencies often face situations where they cannot quantify the incremental benefits for significant categories of effects of a rule. For example, of the 58 major rules reviewed by the U.S. Office of Management and Budget (OMB) in fiscal year 2016, more than half (31) had unquantifiable benefits. Given this level of uncertainty, one might wonder why agencies proceed with these rulemakings. In many cases, Congress may compel the agency to act through a statutory mandate, and courts may impose deadlines for mandatory regulatory action through judicial orders and settlement agreements. Such requirements motivate agencies to make decisions, even when the outcomes of their actions are uncertain.
Deregulatory actions, however, particularly actions that relax the stringency of existing regulations without completely repealing those rules, typically lack these same initial motivating forces. How will regulators choose among hundreds of rules and rule provisions that could be potential deregulatory candidates? This choice is made particularly difficult when, as is true in many cases, the adverse effects of these possible deregulatory actions—that is, the forgone benefits that existing rules currently provide—are uncertain. The risk assessment process is no simpler for deregulatory actions; all of the same data and resource constraints exist. Furthermore, some of the tools analysts traditionally use to evaluate the potential positive net benefits in prospective analyses are less useful in a deregulatory context.
Imagine, for example, a hypothetical regulation intended to improve safety in the United States by requiring that workers in a particular industry receive additional safety training. When promulgating the regulation in the first place, analysts could readily estimate industry compliance costs by collecting publicly available data and constructing models. Quantifying the benefits from additional training in terms of incremental risk reduction, however, may be difficult due to incomplete information about the efficacy of training programs. When significant categories of benefits cannot be quantified, OMB suggests that agencies describe the potential benefits qualitatively and use tools such as break-even analysis to answer the question, “How small could the value of the non-quantified benefits be…before the rule would yield zero net benefits?”
In the example above, analysts might use information about the value of avoided fatalities to estimate the smallest number of such deaths that, if avoided each year, would result in zero net benefits. Decision-makers can then evaluate whether the regulation, if promulgated, would likely deliver risk reduction that exceeds such a break-even threshold, based on available information about baseline fatality rates associated with the regulated activity. For example, if the activity of concern has historically resulted in approximately 150 fatalities each year, and the break-even threshold for a proposed regulation is ten avoided fatalities, comparing this threshold level to the baseline fatality rate suggests the break-even risk reduction is achievable. However, if the threshold level is 200 avoided fatalities, regulators will know that such a risk reduction is impossible, because it exceeds the baseline fatality rate. Thus, they can be reasonably certain that the required regulatory benefits will not be achieved, and they will need to think about other regulatory solutions. In other words, the baseline fatality rate provides an upper bound on the magnitude of benefits that can be achieved by the regulation because it quantifies the current level of harm the regulation seeks to alleviate. Even though the actual number of avoided fatalities likely to result from the regulation is unknown, at the extremes break-even analysis can highlight regulations that are very likely to be successful or unsuccessful.
In the context of a deregulatory action, break-even analysis is less useful for two reasons. First, the potential change in risk no longer has a clearly-defined upper bound. Analysts are unable to use the historical record, especially if the regulation has been in place for many years, to bound the potential forgone benefits because risks will presumably increase, rather than decrease, as a result of the regulatory change.
Continuing with the safety training example, assume that the safety training requirement has been in place for many years, and, as a result, the current fatality rate associated with the regulated activity is 50 fatalities per year. A deregulatory action eliminating the requirement might result in an increase in fatalities. However, how much this rate will increase is more difficult to bound. Fatalities could increase, in theory, by an amount equal to all the individuals undertaking the activity in question. Such an outcome is probably unreasonable; however, identifying a reasonable upper bound is difficult. And, without an upper bound to compare to the threshold amount, it is more difficult to gain helpful insights from break-even analysis.
Second, communicating the results of the break-even calculation is also more challenging. Going back again to the safety training example, instead of describing the number of fatalities an agency hopes would be avoided with additional training, analysts would have to describe how many additional future fatalities would be “acceptable” from an economic viewpoint if training requirements, and associated costs, are reduced. In other words, regulators would need to describe how many additional fatalities society would be willing to accept as a tradeoff for avoided compliance costs. It is difficult to envision how agencies would successfully communicate such information without creating considerable public concern and opposition to the proposed deregulatory action.
Similarly, when issuing new rules in the first instance, agencies sometimes use pilot programs to test regulatory interventions prior to promulgating a regulation. But to provide a basis for informing deregulatory actions, pilot programs are less likely to be viable. Pilot projects work in the context of new regulations because pilot participants agree to additional requirements on a voluntary basis. In a deregulatory setting, the opposite would be true; participants in a pilot program would receive regulatory relief. Providing such relief to a subset of the regulated community may not be possible from a legal or ethical standpoint. Competitors who do not receive the same relief may justifiably complain that they are being treated inequitably. Moreover, with respect to regulations that aim to protect workers or the public from risks, selectively waiving these requirements, even on a pilot basis, is tantamount to knowingly exposing individuals to potential harm.
The scrutiny likely to be applied by the courts to deregulatory decisions, particularly those that contravene earlier agency regulatory decisions, combined with more limited options for addressing uncertainty, suggests that agencies need new primary research on the risk changes resulting for deregulatory candidates. It is difficult to imagine, however, that agencies will have the resources to undertake this work, given their current workload and existing and anticipated resource constraints. To address the challenge of limited resources and overloaded agencies, some observers suggest that, as a means of getting preferred regulations into the queue for deregulation, interested stakeholders should consider funding relevant economic and legal analyses. Such an approach has several limitations but also some advantages.
During the development of the analysis conducted before an agency takes a regulatory or deregulatory action, the questions asked when constructing cost and benefit models often help agency decision-makers refine their vision of how to implement the regulation. If stakeholders were allowed to fund an agency’s analysis, agency leaders would need to ensure that this type of collaborative interaction between the regulated entities and the agency would adhere to legal requirements under the Administrative Procedure Act.
Without careful oversight by the agency, interested parties might make incorrect assumptions about how the agency plans to implement a deregulatory action, resulting in a study of limited usefulness. Furthermore, agencies would need to be cautious about potential adverse effects from conflicts of interest. Without adequate measures in place, interested stakeholders could skew analytic methods and assumptions toward a pre-determined finding or conclusion.
Research funded by interested stakeholders does have some advantages, however. Stakeholders would bring resources to overloaded regulatory agencies operating on limited budgets. Industry-sponsored research projects would also not be subject to the constraints of the Paperwork Reduction Act, which would mean researchers could survey industry members to collect information about costs or risks. Finally, stakeholders could provide important insights about technological or other innovations that are not obvious to agencies.
There is some precedent for allowing private entities to fund analysis in support of federal decision-making. Under the National Environmental Policy Act, an agency can use a third-party contractor, paid for by the private project proponent, to prepare an environmental impact statement (EIS). The agency is responsible for independently evaluating the information and ensuring its accuracy. The contractor is selected by the agency, and the contractor is responsible to the agency for preparing an EIS that complies with the law. Perhaps with this type of firewall, private entities could fund research to support deregulatory actions while also allowing for appropriate collaborative interaction between the agency and the analysts.
Under Executive Order 13,771, federal agencies must consider deregulatory actions. But to ensure that such actions result in positive outcomes for society, agency officials must also think about how to provide quantified estimates of potential forgone benefits. These estimates are likely necessary if these actions are to withstand legal challenge, given that the agency is not compelled by statute to undertake specific deregulatory action and given that some of the traditional tools used by agencies to address uncertainty are difficult to apply in a deregulatory context. Perhaps especially if interested stakeholders are willing to invest in suitably structured information-gathering and analysis to support deregulatory actions, agencies may be able to prioritize the primary research needed to measure risk changes and demonstrate that risks will not increase if some existing regulations are to be eliminated.
This essay is part of an eight-part series, entitled New Developments in Regulatory Benefit-Cost Analysis.