Cornell Law Review Volume 101 Issue 4

Reporting Agency Performance: Behind the SEC’s Enforcement Statistics

Every October, after the end of its fiscal year, the Securities and Exchange Commission releases its annual enforcement report, detailing its activity for the year. The report boasts record enforcement activity, often showing significant in- creases over the prior fiscal year in the number of enforcement actions brought and monetary penalties ordered. The numbers suggest that the SEC is ever tougher on securities violators. The SEC includes these statistics in its budget requests; the figures are repeated in congressional testimony, scholar- ship, policy proposals, and the business press.

Yet the SEC’s metrics are deeply flawed. This Article, a pilot study, reviews fifteen years of enforcement actions and demonstrates that the widely-circulated statistics are invalid because they do not measure what they purport to measure, and unreliable because they are inconsistent and can be manipulated all too easily. The SEC double and triple counts many of the enforcement actions it brings and overstates the fines it orders. Once these measures are adjusted, they reveal that enforcement remained steady between 2002 and 2014, and shifted towards easier-to-prosecute strict-liability violations.

The SEC is not alone in using statistics that have a propensity to mislead to report its output. Multiple reporting statutes authorize Congress to cut agencies’ budgets for failing to meet performance targets. In response, agencies report flawed statistics to protect their ability to continue enforcing the law. This Article suggests that Congress should not threaten to reduce an agency’s budget because of year-to-year fluctuations in enforcement. In addition, to make reported numbers more reliable, nonfinancial performance measures should not be developed by the agency. Instead, the selection and development of performance indicators should be out- sourced and possibly standardized across agencies, much like financial reporting has already been standardized. Doing so would depoliticize reporting, as well as facilitate comparisons among agencies, both domestically and internationally.

:: View PDF