The SEC, Enforcement, and the Problem of Stats

James Kidney, an attorney in the trial division, recently retired. In an unusual step, his departing speech at the Commission has been posted. The speech is here.  

The remarks raised a number of issues with respect to the Commission. One in particular concerned the use of stats (number of cases brought) as a means of assessing productivity. This, he asserted, was a "cancer." 

• The only other item I want to be serious about, besides some personal observations in a minute, is the metric of the division of enforcement: number of cases brought. It is a cancer. It should be changed. . . . Please don’t tell me we account for other factors in our management of cases. We think about them, of course, but we all see cases frequently to which we offer a head scratching response. Really? The SEC spent time and money on that? These cases have no significant impact and the conduct is of minimal or no harm to the investing public. But the investigation has been intense and expensive. Could no one in management exercise judgment and call the investigation to a halt? Of course not! Bringing the case is a stat! The metric we have now is built into the soul of the Division. It has to be removed root and branch. 

We have discussed the issue of stats on this Blog before, including the resulting emphasis on quantity rather than quality. But we have also examined the issue from a different perspective. An emphasis on finished cases essentially penalizes those who spend time on "unfinished" cases. The easiest way to be sure a case will be finished is to investigate instances of fraud (or other violations) that have already become apparent. In other words, this encourages a post-hoc strategy of enforcement.

Yet there is plenty of room to argue that the Commission should be investigating in an anticipatory fashion, "looking around the corner" so to speak. DERA's accounting quality model is designed to spot problems before they become public. A consequence of looking around the corner, however, is that sometimes nothing will be there. In other words, finding violations that have not yet become public will result in more unfinished investigations. And if unfinished investigations count against the enforcement attorneys involved and weigh against the producivity of the Division, there will be far less incentive to implement this approach in any meaningful sense.  

The emphasis on stats also has the potential to encourage the staff to bring smaller cases. Actions against a large company or firm with top counsel can involve millions of documents and thousands of attorney hours. The complaint filed in the Goldman case (the one that settled for $550 million) included 10 attorneys. Goldman presumably counted as one stat. Had the same attorneys instead worked on smaller cases, they could have generated additional stats.  

The SEC has now been in place for 80 years (it was created in the Exchange Act, not the Securities Act of 1933). The Division of Enforcement was created in 1972.  As bureaucracies age, they can sometimes become excessively beholden to the way "things have always been done." The efforts by the Division to create speciality divisions and eliminate branch chiefs represented an effort to overcome this inclination. Perhaps the use of stats as a principle measure of productivity should likewise be reexamined.