A long time ago, I wrote about President Obama's nomination of Cass Sunstein to head OIRA and noted that some academics and commentators opposed it because of Sunstein's support for cost-benefit analysis. Sunstein is not yet confirmed many months later, not because of the opposition of critics on the left, but because some senators took fright at Sunstein's views about protecting animals from cruelty. This is hardly germane to the nomination to head OIRA, and one hopes that now that the political point has been made, confirmation will come in due course.
Meanwhile, the academic debate about cost-benefit analysis continues. Critics of cost-benefit analysis have argued that a better approach is to use "feasibility analysis." Feasibility analysis requires the regulatory agency to identify hazards and regulate the activities that cause them to the extent possible without causing widespread economic disruption—which is cashed out in terms of revenue or profit loss for the affected industry, bankruptcies, or plant closings. My colleague Jonathan Masur and I have written a new paper that argues that feasibility analysis is a conceptually confused and economically incoherent approach to regulation. It should appeal to neither pro- nor anti-regulatory forces. The abstract is below.
Feasibility analysis, a method of evaluating government regulations, has emerged as the major alternative to cost-benefit analysis. Although regulatory agencies have used feasibility analysis (in some contexts called 'technology-based' analysis) longer than cost-benefit analysis, feasibility analysis has received far less attention in the scholarly literature. In recent years, however, critics of cost-benefit analysis have offered feasibility analysis as a superior alternative. We advance the debate by uncovering the analytic structure of feasibility analysis and its normative premises, and then criticizing them. Our account builds on two examples of feasibility analysis, one conducted by OSHA and the other by EPA. We find that feasibility analysis leads to both under- and over-regulation, and we conclude that it lacks a normative justification and should have no place in government regulation.