Should we pay regulators according to their performance? In a new paper, Jason Chen, Jakub Hajda, and Joseph Kalmenovitz show that a pay-for-performance system has a surprising effect: it increases regulatory effort but also motivates regulators, especially the productive ones, to quit and join the private sector.
In the wake of the recent regional banking crisis, critics argued that bank regulators are insufficiently motivated to perform their duties. In response, commentators have suggested that federal agencies should consider adopting a pay-for-performance scheme. This solution may work in the private sector, and many studies conducted on C-Suite executives demonstrate the effectiveness of pay-for-performance. However, there is very little understanding on how pay-for-performance may affect the behavior of public sector employees.
In our paper, we directly study this fundamental question: should we pay regulators based on performance? We utilize a new dataset that tracks the careers of all senior federal regulators. We show that in 2002, when a select group of senior regulators transitioned to a new pay-for-performance system (which is still in place), they responded by leaving the government for the private sector. To understand the reasons behind this phenomenon, we build an economic model that highlights the unique aspect of public sector salaries: they are capped from above. Thus, combining pay-for-performance with a pay cap yields a complicated response. The performance pay increases effort, that is, the time and dedication regulators put into their job. But it also causes productive regulators to reach the pay cap more quickly, prompting them to transition to the private sector where pay is unlimited.
Our analysis focuses on federal executives. This is a group of career bureaucrats, roughly 0.5% of the federal workforce, who hold leadership positions in their respective agencies. For instance, the Financial Crimes Enforcement Network is run by seven federal executives. Rising through the ranks of the Treasury Department over many decades, they now oversee the agency’s activities to safeguard the financial system from illicit use, such as money laundering and terrorism.
By the late 1990s, the pay system for federal executives faced criticism for two main reasons. First, they were not properly rewarded according to their performance, and largely received automatic annual pay raises. Second, there was a severe compression in executive pay, with the pay floor rising faster than the pay ceiling. This compression meant that, even if a meaningful pay-for-performance system was implemented, its efficacy will be limited. To address these issues, a major pay reform was rolled out between 2002 and 2004. The goal was to retain executives and stimulate better performance. To achieve that, the reform simultaneously raised the pay ceiling for federal executives and linked their pay to performance using an enhanced pay-for-performance evaluation process.
A key aspect of the reform was that it only affected executives working for certain agencies, and did not affect executives working in agencies with independent payroll systems. In the first part of our paper, we leverage this feature to study the causal impact of pay-for-performance on the revolving door: using a difference-in-differences methodology, we compare the response of treated executives (“treatment”) to the response of non-treated executives (“control”), before and after the reform. We find that the pay-for-performance reform increased the exit likelihood among treated executives by 2.5%, which is nearly half the sample average. In other words, the adoption of a robust pay-for-performance system significantly accelerated the revolving door among federal executives.
At first glance, this finding is surprising: the reform was intended to retain talent within the public sector, but it achieved the exact opposite. To understand the economic forces driving this result, we develop an economic model that links executive pay to exit and effort. One advantage of the model is that it allows us to study changes in effort, even though we cannot measure effort directly (for example, we cannot measure “dedication to the job”). The model provides several important insights. First, an increase in pay-for-performance and a higher pay ceiling both induce more effort. Performance pay increases the reward for hard work, and a higher pay ceiling allows for greater pay raises, stimulating more effort. Second, pay-for-performance and pay ceiling have contradictory effects on exits. A higher pay cap increases the salary potential in the government, which in turn reduces the incentive to leave. In contrast, performance pay induces exit: productive executives are hitting the pay ceiling faster, and they are therefore more likely to take their talents to the private sector where pay is uncapped.
To draw further insights from our model, we combine it with our rich executive-level dataset and use an estimation technique known as Simulated Method of Moments. This exercise yielded several interesting results. First, performance pay constitutes a relatively small fraction (21%) of federal executives’ salary. This is in stark contrast with the pay structure of the private sector, where incentive pay accounts for 75-80% of total CEO pay. Second, despite the relatively small incentive, performance pay has a large impact on federal executives. Concretely, we estimate that a 1% increase in pay-for-performance will, on average, increase effort by 0.04% and exits by 7.24%. This highlights the trade-offs associated with performance pay in the public sector; any increase in performance pay will lead to a modest boost in effort but will also trigger a wave of resignations. Third, motivated by this trade-off we design a set of zero-exit pay packages that induce greater effort without increasing exits. For example, a combination of 10% increase in pay-for-performance and 9% increase in the pay cap. Both changes will induce greater effort, and they will have a contradicting effect on exits that will perfectly offset each other, resulting in net zero change in the revolving door.
Overall, our research provides insights into the consequences of implementing pay-for-performance in the public sector. While intended to retain and motivate talent, such systems may inadvertently motivate successful regulators to quit and join the private sector. Policymakers may want to consider these dynamics when designing compensation structures, to ensure that they align with the goal of maintaining an effective regulatory workforce.
Authors’ Disclosures: the authors reports no conflicts of interest. You can read our disclosure policy here.
Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.
