In recent years, executive compensation in the U.S. has become a hotly debated issue. A central point of contention is peer benchmarking, an integral part of the pay-setting process in which firms compare their executives’ compensation with that of rivals in the labor market.  Proponents of this practice claim that compensation benchmarking is an efficient mechanism used to gauge market wages within a competitive labor market, while detractors allege that it inflates pay because firms may arbitrarily select peers with generously remunerated executives.

In a recent paper, I study the dynamics of compensation benchmarking using a comprehensive, hand-collected dataset of explicit peer group relationships disclosed in mandatory filings to the U.S. Securities and Exchange Commission by publicly traded companies. This novel dataset enables me to explore peer benchmarking by smaller public firms, where the potential effects are large, because the benchmark is often executive compensation at larger firms that can pay higher salaries.

I document that smaller firms do indeed select better paying peers, on average. Relative to larger companies, small firms select peers that pay their executives approximately 30 percent more than they pay their own executives. However, my results suggest that this apparent bias in peer selection is not a manifestation of agency conflicts. Instead, I find that compensation benchmarking is consistent with rational contracting in which boards strategically adjust pay to retain valuable managers. Additionally, the data imply that there are notable labor market differences between the general population of public firms and the largest firms in the U.S. economy.

I find that peer benchmarking reflects future performance among smaller firms. If the board anticipates that the firm will perform well or grow rapidly, the directors may select better paying peers and adjust next-year compensation upward to retain valuable managers. My tests support this conjecture: Smaller firms with better paying peers experience superior future performance as measured by both stock and accounting returns. Specifically, a one standard deviation increase in peer pay relative to firm pay is associated with a 6 percentage point higher stock return and 2 percentage point greater return on assets per annum for smaller firms. My results are weaker or nonexistent for larger firms.

I also find that compensation benchmarking plays a more prominent role in the pay-setting process at firms with executives that have highly transferable skills.  If boards use compensation benchmarking to retain managers, the extent to which an executive’s pay is influenced by peer compensation will relate to how readily the executive could employ her skills at another firm. Using a popular metric for skill transferability, I find that small firms with highly transferable managers adjust pay towards their target labor market group by over 30 percent more per year relative to baseline peer pay sensitivity levels. This relation is present, but greatly attenuated, in large firms.

What’s more, I directly test whether selecting better paying peers is associated with weak corporate governance. I examine several characteristics that have been shown to reflect poor corporate governance, and find no evidence that poor governance is related to upward compensation benchmarking. If anything, the difference between peer pay and firm pay is negatively correlated with poor governance.  This suggests that sufficiently powerful managers may simply dictate their pay without undergoing the complexities of peer group manipulation.

Finally, I study the role that compensation benchmarking plays in executive turnover. Under rational contracting, boards select peers in a manner that offsets executives’ tendencies to take outside positions, but simultaneously, boards do not wish to pay their executives more than it would cost to obtain equally able individuals. I find that peer pay does not have a detectable effect on turnover probabilities in tests that model turnover as a function of peer selection. This is consistent with an equilibrium in which firms select peers to offset their executives’ unobservable propensity to leave the firm.

Taken together, these results suggest that executive compensation benchmarking is an important characteristic of a well-functioning executive labor market. Although smaller firms tend to choose higher paying peers, this behavior is more in line with rational contracting rather than rent extraction. When boards anticipate good performance, they select aspirational peers that reflect their executives’ potentially growing set of outside options, and the rate at which pay adjusts towards peer levels is sensitive to the transferability of managerial capital. Overall, my study contributes to the debate on executive compensation by demonstrating that peer benchmarking is a practical mechanism used to set wages and retain valuable managerial talent.

This post comes to us from Thomas Ian Schneider, a Ph.D. candidate in finance at Boston College’s Carroll School of Management. It is based on his recent paper, “Executive Compensation and Aspirational Peer Benchmarking,” available here.