As part of a broader anti-poverty agenda, Rep. Paul Ryan (R-Wis.) has recently proposed steps to reform how regulations are written. His plan, which would require “distributional analysis” of agency regulatory proposals and congressional approval of those proposals found to have a regressive impact, has attracted some cautious support from Democrats, including President Obama’s former regulatory czar, Cass Sunstein.
{mosads}The idea that agencies should look at distributional impacts of their regulations is not a new one. Agencies are required under President Clinton’s Executive Order 12866, which governs the agency economic analysis of regulations, to consider “distributive impacts and equity.” Unfortunately, Harvard researchers Lisa Robinson, James Hammitt and Richard Zeckhauser have found that “agencies provide little information on distribution” despite being required to do so by the executive order. This raises the question of why the Ryan proposal would work when the executive order has failed.
A tempting answer is that putting a requirement for distributional analysis in law will be more effective than putting it in an executive order. However, many of the problems that bedevil distributional analysis won’t go away just because a law is written. Primary among these problems is that figuring out these impacts is very challenging. Thirty-four years into requiring analysis of all benefits and costs for certain regulations, we have made great progress but are still wrestling with the questions of how to do so. Now trying to figure out who (or what classes) are affected by a regulation adds another degree of difficulty.
Even if we could measure distributional impacts, it is not clear that we will learn anything new by doing so. Regulations that have been cited as disproportionately affecting the poor, such as the Department of Transportation’s requirement for rear-view cameras, or the Food and Drug Administration’s regulation prohibiting ozone-depleting chemicals in inhalers used by asthma patients, already fail a benefit-cost test without considering distributional impacts. In other words, adding distributional analysis only tells us that, as with the population as a whole, the costs of the regulation exceed the benefits for poorer people.
Also limiting the knowledge that could be gained from a distributional analysis, is the fact that the Ryan proposal (as noted by Sunstein) focuses primarily on the costs of regulations to the poor and downplays the benefits. Many regulations designed to protect public health are likely to provide disproportionate benefits to the least fortunate members of society. The harm caused by climate change, other environmental hazards and occupational risks all fall most acutely on the most vulnerable. Any true analysis of distributional impacts would need to focus on the benefits to the same degree it focuses on the costs.
Once benefits are included in a distributional analysis, and we recognize that most of the rules that would be deemed regressive also fail a more general benefit-cost test already in place, the number of regulations likely to be affected by Ryan’s proposal for distributional analysis becomes vanishingly small. However, Ryan’s proposal also raises the question of congressional review of regulations. Congressional review of all regulations has long been a goal of reformers of the regulatory process and is a centerpiece of many other regulatory reform bills currently being considered by Congress. It is a worthwhile topic for debate and raises complicated issues of separation of powers and the role of the legislature in overseeing the executive branch. These issues should be debated on their merits and not shoehorned into a bill that is ostensibly about reducing burdens on the poor but is unlikely to do so.
Shapiro is an associate professor and director of the Public Policy Program at Rutgers University and a member of the Scholars Strategy Network.