How NOT to prevent another Enron

Everyone I've talked to, from the left and right, agrees that Sarbanes-Oxley was bad law, politically motivated, hastily thrown together, and porly thought through. It also suffers from a fatal conceit that anticipatory regulation--i.e. treating executives as a criminal class--can actually work. An excerpt from "How Not to Prevent another Enron" by Richard Epstein.

It is instructive to remember that Sarbanes-Oxley was brought on by a half-dozen corporate bankruptcies--Enron, Tyco, and WorldCom among them. Each case was a distinctive breakdown, yet, in the concern over these high-visibility cases, it is easy to forget that 99 percent of public corporations have done nothing that merits reproach. Most have behaved properly over the long haul: They make good products that generate profits for shareholders and jobs for employees. But the commonplace becomes inconspicuous when a zealous Congress seeks to "prevent the next Enron" by introducing elaborate safeguards and sanctions. Similar refrains have been sounded after previous frauds, from Teapot Dome to the almost forgotten Salad Oil and Equity Funding scandals. Yet that naive optimism ignores the central truth of any large and diverse economy: Things will go wrong no matter how sound the legal regime.

The grand conceit is that we can "fix" such problems once and for all by anticipatory regulation. Even the most clever, adroit, and skillful legislature cannot achieve zero risk in human affairs. Any sensible legal system has to rely in part on sanctions brought to bear after people have stolen property or looted corporations. We do not possess a laser precise enough to target in advance only thieves. Any regulatory scheme sweeps into the net the good corporations along with the bad. The effort to slow down abuse in poorly run corporations often makes it more difficult for well-run corporations to remain that way. Compliance with external standards could easily undermine internal safeguards, formal and informal, against fraud and abuse.