Escaping Pay-for-Performance

When Silicon Valley Bank collapsed in March 2023, the blame quickly spread from the bankers to their supervisors. The Federal Reserve's own post-mortem concluded that examiners had spotted the bank's vulnerabilities but did not push hard enough to get them fixed. An old idea promptly returned to the policy debate: if regulators lack incentives to do their jobs, pay them for performance. The proposal has a long pedigree. After the 2008 crisis, legal scholars argued that paying bank examiners like bankers would help prevent the next disaster, and commentators and policymakers have revived versions of the idea ever since.

The logic is seductive. Performance pay is everywhere in the private sector, and a large literature shows that it raises effort and output. Why should government be different? In our working paper, we show why: performance pay makes regulators work harder, and then it helps them leave. Any evaluation of these reforms that looks only at productivity misses half the story, and arguably the more important half.

Our laboratory is a set of nine reforms adopted by U.S. financial regulatory agencies between 1981 and 2006, including the OCC, the FDIC, the SEC, and the CFTC. The reforms replaced the largely automatic step increases of the federal pay scale with merit-based raises and wider pay bands, with the explicit goal of motivating effort and retaining regulatory talent. Because agencies adopted the new schemes at different times, we can compare adopters with otherwise similar agencies that had not yet adopted, using a stacked difference-in-differences design. To do so, we assemble payroll records covering 31,784 regulators from 1973 to 2013 and link them to two new data sources: rulemaking records from the Federal Register, which reveal which regulators contribute to which rules, and career histories from Revelio Labs, which trace regulators into their post-government jobs.

The first finding gives the paper its title. Once agencies adopted performance pay, voluntary exits from government jumped by 43% to 57% relative to the sample mean. And this is not a quirk of financial regulators. When performance-based pay reached the Senior Executive Service in 2004, covering 23,763 top officials across 397 agencies, we find the same response.

The second finding is that, on its own terms, the reform worked. Among regulators who stayed, the likelihood of contributing to a new regulatory document rose by 62% to 68% relative to the sample mean. Agencies got more observable output from the workers who remained.

The third finding ties the first two together. Regulators who left after the reform earned 53% higher starting salaries, and 69% higher total compensation, in their first private-sector jobs than comparable leavers from control agencies. We read this as the central mechanism. Performance pay pushes regulators to exert effort and build skills that private employers value, and it makes individual ability easier to observe from outside the agency. The same incentives that raise effort therefore raise the value of the outside option. The revolving door does not merely keep spinning; it spins faster.

Who leaves matters as much as how many. Exits were concentrated where performance evaluations were noisier or less trusted, and among regulators with weaker attachment to public service: those hired during recessions, or those who had sacrificed little pay to join government in the first place. In other words, performance pay did not just change behavior. It reshaped the composition of the regulatory workforce, tilting it away from employees with strong outside options and toward those who value public service the most.

To think through alternatives, we also estimate a dynamic model of regulator effort and career choice. The estimates sharpen the tradeoff. Stronger performance incentives increase effort, but they also increase exits. A higher level of government pay, by contrast, reduces turnover without the same side effect. Implementation matters too: the exit response is largest precisely where the evaluation process is least transparent and least credible, which tells us that how performance is measured matters as much as how it is paid.

So, should regulators be paid for performance? Our answer is a qualified yes, with eyes open. Performance pay delivers what its advocates promise: more effort and more regulatory output. But it also accelerates departures and strengthens the very revolving door that critics of regulatory agencies worry about. For agencies whose employees hold specialized expertise and enjoy strong private-sector demand, which is precisely the profile of a bank examiner or an SEC attorney, the retention cost can be first order. If the policy goal is a capable and experienced regulatory workforce, raising pay caps and improving the credibility and transparency of performance evaluations will likely achieve more than simply ratcheting up incentive intensity. The broader lesson is that compensation reform in government should be judged on two margins, productivity and retention, because the second margin is where the bill comes due.

Jason Chen is an Assistant Professor of Finance at the Harbert College of Business, Auburn University.

Jakub Hajda is an Assistant Professor of Finance at HEC Montréal.

Joseph Kalmenovitz is an Assistant Professor of Finance at the Simon Business School, University of Rochester. 

This blog is based on a paper presented at the twelfth annual Global Corporate Governance Colloquia (GCGC), hosted by the National University of Singapore and ECGI. Visit the event page to explore more conference-related blogs.

The ECGI does not, consistent with its constitutional purpose, have a view or opinion. If you wish to respond to this article, you can submit a blog article or 'letter to the editor' by clicking here.