Problems With Recent Estimates Of The Costs And Benefits Of Regulations

See Reading Past The Headline In OMB's Report To Congress by Susan E. Dudley of Forbes.
Susan Dudley is Director of the George Washington University Regulatory Studies Center.

"On Friday, the Office of Management and Budget released a draft report to Congress on the benefits and costs of the federal regulations issued over the last decade. Reporters will likely seize on the report’s finding that the annual benefits of rules issued during that period range from $287 to $911 billion, while estimated costs are between $78 and $115 billion. These figures imply that regulatory benefits are at least 3-and-a-half times—and possibly as much as eight times—their costs.

Given such dramatic beneficial results from regulation, why is the Trump administration so bent on deregulation? Reading past the top line numbers offers a rationale.

Problems with Benefit and Cost Estimates

First, OMB’s figures are not independent estimates, but tallies of estimates agencies presented to justify individual rules at the time they were issued. Regulators naturally want to present their rules in the best light, and they rely on hypothetical models and assumptions to project benefits that exceed costs. The Environmental Protection Agency has been the most adept at calculating very large benefits from its rules. According to the Report, “EPA rules account for over 80% of the monetized benefits and over 70% of the monetized costs” of all federal regulations over the last decade.

The Report presents a list of important caveats regarding the EPA benefit estimates in particular, including large uncertainty regarding causation, projected exposures, health risk assumptions, and quantification of health effects. It emphasizes that the fact that it reports these results “does not imply an endorsement by the current Administration of all of the assumptions made and analyses conducted at the time these regulations were finalized.”

Second, OMB emphasizes that “none of these costs reflect retrospective evaluation of their impacts.”  The numbers summarized in the Report were developed before anyone had a chance to observe what compliance really entailed or what impact the rules actually had. As the Report notes, “prospective analyses can never fully capture all uncertainties and future changes in the regulated industries and communities.” Despite requirements to do so, agencies rarely look back to evaluate whether their regulations worked as intended.

Another reason to interpret the top line numbers with caution is the Report’s limited scope. First, because it covers fiscal years 2007 through 2016, it does not include the costs and benefits of regulatory activity in the current administration nor during the last four months of the Obama administration (a period that saw a higher than normal rate of regulations). We’ll have to wait until the 2018 report for those data.

Conversely, some of the rules issued in FY 2016 have since been overturned by the Congressional Review Act or are being reconsidered by the new administration, yet the Report includes them “in order to provide an estimated, but not necessarily realized, impact of the rules finalized during the fiscal year the Report covers.”

Also, as has been OMB’s practice, the Report only covers major regulations for which agencies have estimated both benefits and costs.  That means that although agencies issued 36,255 final rules over the 10-year period, the Report only includes benefit and cost information for 137 of them, or less than one-half of one percent. The Report covers only 5% of rules that were considered significant enough to require OMB review, and 22% of “major” rules (those likely to have impacts of at least $100 million in a year).

The reported estimates do not include benefits and costs of regulations issued by independent regulatory agencies, whose rules are not subject to OMB review. However, based on publicly-available information, OMB presents information for the 18 rules those independent agencies classified as “major” in FY 2016 and finds that, while most include some quantification of some costs, none attempt to quantify benefits."