Incentive pay for bank regulators

Larry Ribstein —  9 September 2011

Now that regulating banker pay has been studied exhaustively, here’s something else worth studying:  bank regulator pay.  Fred Tung and Todd Henderson are on the case, in Pay for Regulator Performance.  Here’s the abstract:

Few doubt that executive compensation arrangements encouraged the excessive risk taking by banks that led to the recent Financial Crisis. Accordingly, academics and lawmakers have called for the reform of banker pay practices. In this Article, we argue that regulator pay is to blame as well, and that fixing it may be easier and more effective than reforming banker pay. Regulatory failures during the Financial Crisis resulted at least in part from a lack of sufficient incentives for examiners to act aggressively to prevent excessive risk. Bank regulators are rarely paid for performance, and in atypical cases involving performance bonus programs, the bonuses have been allocated in highly inefficient ways. We propose that regulators, specifically bank examiners, be compensated with a debt-heavy mix of phantom bank equity and debt, as well as a separate bonus linked to the timing of the decision to shut down a bank. Our pay-for-performance approach for regulators would help reduce the incidence of future regulatory failures

This is an interesting and well-executed idea which inspired a few thoughts.

First, if private-sector-type compensation for regulators is worthwhile, why not just rely on private sector rather than government regulators?  It would seem that the difference between government and private industry is really all about incentives.  Indeed, if we only semi-privatize by injecting market incentives into government actors, this could create tension in incentives analogous to imposing “corporate social responsibility” on corporate agents.

Second, might this proposal work for other types of gatekeepers, such as auditors?  Of course accounting regulation has moved in the opposite direction, toward making them more “independent.”  The Tung-Henderson analysis suggests that carefully structured performance pay may be a better way to go.

Third, if bank regulators, what about other public servants, including prosecutors? As I’ve written in my recently published Agents Prosecuting Agents (SSRN draft), they are agents too. Would incentive compensation better align their interests with those of the public?  My article discusses some problems with this (footnotes omitted):

Designing incentive compensation for prosecutors presents significant challenges.  Even in private law firms in which lawyers produce a clear financial output in the form of fees, there is controversy over whether billable hours or lockstep seniority-based compensation provides the best overall incentives.  The compensation design challenge is greater for prosecutors because there is no measure of the value of prosecutorial efforts.  Obviously a simple metric such as number of prosecutions would skew incentives, in that it may induce prosecutors to ignore the social costs of misguided prosecutions.  One author has proposed compensation based on convictions of charged crimes with deductions for findings of prosecutorial misconduct.  However, this could skew incentives toward, for example, undercharging defendants and over-caution.  On the other hand, tests that try to take more factors into account would be very costly to apply.

Tung and Henderson show how to overcome these metrics problems with bank regulators.  I’m skeptical analogous devices would work for prosecutors, even regarding purely corporate crime.  You’d have to, among other things, measure the deterrence effects of particular prosecutions. Not sure event studies could manage this.

There’s a lot more in this thought-provoking piece. Read the whole thing.

Larry Ribstein


Professor of Law, University of Illinois College of Law

One response to Incentive pay for bank regulators


    Thanks for posting this. I’ve been trying to find studies of where regulators went wrong with banks. This doesn’t give the answer, but it’s interesting anyway. My comments on the paper after hearing about it on this site, intended mainly for the authors but of interest to any reader of it, are at