In Business Roundtable v. SEC, 647 F.3d 1144 (D.C. Cir. 2011), the D.C. Circuit held that the SEC must justify all its regulations promulgated under the National Securities Market Improvement Act through cost benefit analysis. Cost-benefit analysis makes sense for many SEC regulations because they focus on easily quantified matters. But they make less sense for regulations targeting social issues. That’s in part because social issues are difficult to quantify—how do we value preserving pretty open spaces or preventing people from suffering physical pain?—and in part because it is difficult to predict how many people will be affected by social regulation and how they will respond to it.

So it is no surprise that the SEC’s cost-benefit analysis underlying its regulation relating to conflict minerals has come under attack. In a recent paper forthcoming in the Administrative Law Review (http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2714620), Jeff Schwartz and Alexandrea Nelson argue that the SEC’s analysis is completely unjustified

Under the SEC rule, which was required by the Dodd-Frank Act, public companies must investigate to determine whether particular minerals (like gold) in their products come from militarized mines in the Congo and publicly report those findings. Conflict Minerals, 77 Fed. Reg. 56,274 (Sept. 12, 2012). The rule also requires companies to develop plans and implement procedures to guide their investigations.

The SEC estimated that the rule would impose costs of $3-4 billion. It came to this conclusion largely by estimating the number of companies subject to the rule, and all the various costs the rule would impose on each of those companies.

Schwartz and Nelson assess every aspect of this conclusion and make a good case for being highly skeptical of the SEC’s entire analysis. Critics typically focus on the difficulty of quantifying the benefits of rules, but Schwartz and Nelson focus on demonstrating how the cost estimates are unreliable. To go through each criticism would be tedious, but a particularly nice illustration of the extent of the inaccuracy of the SEC’s analysis is that the SEC vastly overestimated the number of companies that would file disclosures under the rule. The SEC estimated 6000 companies; in reality, only 1300 companies filed disclosures the first year of the rule. That suggests a misestimate of 350%.

More glaring is that some costs in the analysis were not even imposed by the rule. For example, the analysis assumed that mineral suppliers would incur $1.2 billion to assist companies with their investigation—even though the rule does not require suppliers to do anything.

Although the shortcomings of cost-benefit analysis are well known, the extent of their impact on any given regulation is often unclear. Schwartz and Nelson provide a nice case study that’s well worth reading.