Impression Management and the Biasing of Executive Pay Benchmarks: A Dynamic Analysis

Abstract

We study the selection of peers into compensation peer groups reported by U.S. corporations. Securities and Exchange Commission (SEC) regulation requires firms to report these peer groups which are used by investors and shareholders to benchmark the compensation of CEOs. Building on a novel, dynamic analysis of more than 1,400 compensation peer groups since 2006, this paper presents new evidence that compensation peer groups are biased upwards relative to neutrally chosen “natural” peer groups. This upward bias is masked by several factors including normative selection on other criteria than compensation and the strong negative relationship between bias and the percentile at the named peer group at which the CEO is compensated. We find that adjustments to compensation peer groups are often better explained as bias-maintaining impression management than as structural adjustments to align peer groups more closely with the normative principles for peer group selection. We also find that peer group bias cannot generally be explained away as a reward for CEO talent. Our research suggests that although the SEC regulation was intended to minimize rent extraction, it has given firms a tool to manage impressions of shareholders and justify excessive pay.