Document Type



Over the past few years the debate over the economic rationality of health, safety and environmental regulation has morphed into a sustained controversy over the tests and methods by which that rationality is judged. Critics have argued that the main regulatory scorecards which comprise much of the empirical foundation for the regulatory reform movement are fundamentally flawed because they: alter agency estimates of future costs and benefits; disregard most uncertainties; and misrepresent ex ante guesses as the costs and benefits of regulation. They also zero out whole categories of benefits that cannot be quantified and/or monetized even when the benefits clearly are substantial. Finally, most scorecards contain no test to detect untaken regulatory opportunities or regulations that are too lenient. They only identify regulations that appear too strict, an asymmetry that creates a built-in bias favoring the conclusion they purport to have proven. This Article responds to the latest defenses of scorecards offered by their proponents. It observes that scorecardists do not, for the most part, deny engaging in the practices described above, and cannot deny that they stray from widely-agreed principles of sound cost-benefit analysis in so doing. Rather, defenders of scorecards have asserted that such deviations are justified on policy grounds, or that they do not materially affect the results, or (most fundamentally) that there is no alternative. This Article argues that these errors occur, they are inherent to strictly numeric scorecards, they are not justified, and they cannot rationally be disregarded as harmless errors. Moreover, there is a clear alternative. Implementing that alternative, however, requires abandoning strictly numeric regulatory scorecards as a defunct mode of analysis. It also requires certain changes in the way regulatory analyses are prepared initially at the agency level, and in the way they are reviewed after the fact.