Professor Urska Velikonja at Emory has written an article that unpacks the enforcement statistics ("stats") annually issued by the Securities and Exchange Commission. The paper is here.
That a federal agency would have an incentive to develop a counting system that maximizes numbers is no surprise and I am guessing it is common. In truth, stats (and their constant increase) have little value except to reduce the inevitable criticism from SEC critics that would occur if the stats dropped.
Irrespective of how they are counted, the broader issue is whether the SEC should rely on stats as heavily as a measure of success. In truth, an effective approach to enforcement should involve a certain amount of investigation designed to “look around the corner” and find fraud or misbehavior before it becomes public. You want to find the Madoff’s before they are on the front page of the Wall Street Journal. This approach would, by definition, result in investigations that ended without cases being filed. Under the current system, they would not produce any stats.
The SEC is presumably heading in this direction. Nonetheless, the agency is still under pressure to maintain stats. Hopefully the article will not just encourage a conversation on the best method of calculation but will also encourage a conversation on the need to reduce reliance on this metric as a measure of success.