The Role of Retrospective Analysis and Review in Regulatory Policy

There are many obstacles to better regulatory measurement, analysis and review. But major setbacks include the agencies’ lack of impartiality in reviewing their existing regulations, an inappropriately narrow focus, and a failure to promote steps to better measure the actual benefits vs. the costs of regulations.


In 2011 President Obama issued executive orders regarding analysis and review of existing federal regulations. More than a year later, agencies appear unable to perform meaningful retrospective analysis of their own rules. Past administrations have attempted similar reforms to the regulatory analysis and review process with little success. In fact, very few retrospective analyses of current regulations provide sufficient information to evaluate whether benefits exceed costs.

There are several reasons why we should analyze and review existing rules. One is to evaluate the soundness of the original economic analysis. It is also important to evaluate if the market failure that motivated the rule has been overcome by changes in market conditions or in technology and if the rule effectively addresses the underlying problem. 

The Code of Federal Regulations had 165,494 pages in 2010, and since 1970, the total pages have tripled, growing at an average of about 2.8% annually. Such a long and complicated set of regulations exceeds comprehension and taxes the resources of entrepreneurs.

An analysis of four regulatory agencies and their retrospective reviews indicates that regulatory processes resemble “business as usual’ with little discernible new work on the measurement called for in President Obama’s executive order. The agencies are the Environmental Protection Agency, the Food and Drug Administration, the National Highway Safety Transportation Administration, and the Securities and Exchange Commission. 

The most prominent practitioner of retrospective analysis is apparently the National Highway Traffic Safety Administration, which has completed 92 evaluations of the costs and/ or the effectiveness of various facets of its regulatory program since 1973.

Major obstacles to better regulatory measurement, analysis and review include the agencies’ lack of impartiality in reviewing their existing regulations, an inappropriately narrow focus, and a failure to promote steps to better measure the actual benefits and costs of regulations. 

Three measures can be taken to address these roadblocks: 

  1. Promote impartiality in retrospective analysis. The NHTSA has done a notable job in this respect, so one approach would be to pursue changes to make other agencies more like NHTSA. To create a data-driven environment for other regulatory agencies Congress could allow or encourage pilot programs. Such projects could help ensure that new rules have an empirical basis adequate to show that their benefits likely exceed costs. In addition, Congress could create a nonpartisan office charged with conducting and reviewing both retrospective and prospective economic analyses of regulations and regulatory programs. 
  1. Reconsider the scope of retrospective analysis. Evaluating regulatory programs rather than individual regulations may provide better information about benefits and costs of federal actions. With approximately 165,000 pages of extant rules, a wholesale approach focused on regulatory programs may be more likely to cover the wide scope of federal rules than a retail approach focused on individual regulations.
  1. Promote regulations that measure effects. Review plans of key regulatory agencies do not specify new steps for improving the measurement of federal regulations’ actual effects. Such improvements could reduce the controversy associated with use of benefit-cost analysis. A feasible and effective approach would be to design regulations to allow or ensure better measurement of their results. Agencies should promote market-based mechanisms like permit trading as well as greater use of surveys of the effectiveness of information disclosure requirements.