In the paper, Executive Superstars, Peer Groups and Over-Compensation — Cause, Effect and Solution, which was recently made publicly available on SSRN, we develop a pragmatic approach to understanding the run-up in CEO compensation over the past several decades. Rather than looking to markets or captured boards for the explanation, we argue that the actual mechanical process of peer benchmarking by which pay is set is the cause of the present controversy. From this perspective, we present what we believe will be an effective solution; additionally and collaterally, some interesting lessons about executive recruitment, particularly the CEO “superstar” culture, may be gleaned from our findings. We thank the Investor Responsibility Research Center Institute, which has long funded compensation research, for their financial support and helpful assistance in the development of this paper.
The piece makes a contribution to the executive compensation literature as it offers a novel explanation for the perpetual rise in CEO pay and suggests a significantly different solution to the compensation controversy. As boards have typically looked outside the organization to set CEO pay, we argue that this approach, known as “peer grouping,” is seriously flawed as it relies on the notion of an easy transferability of executive talent which empirically, is incorrect. Therefore, boards should look within the organization itself rather than to external comparators to create an appropriate CEO pay structure. We suggest that this approach should begin to resolve the CEO compensation problem.
Boards will almost always pay their CEO an amount based upon a comparison with what other similar companies, referred to as their peer group, are paying their CEOs. For various reasons referenced in the paper, the value of the total compensation package will be targeted to the 50th, the 75th or the 90th percentile. After the compensation committee selects the appropriate percentile level, a package is then designed by the consultant to meet the specified numerical goal.
Despite the prominent position of the peer group process within actual compensation negotiations, it is seldom given similar attention in the academic executive compensation literature. As part of the broader management-capture hypothesis, it has been suggested that these groups may be manipulated so as to create higher levels of pay. While several finance studies have found evidence of bias or manipulation in the selection of larger or higher paying companies as peers, a few other studies have not. Today, the proxy advisory firms will analyze the composition of disclosed peer groups in an attempt to ferret out this manipulation. On the other hand, market theorists view peer groups to be irrelevant, and, if anything, simply the expression of market prices. To them, the process is none other than an effective gauge of “market wages” which are necessary for the retention of an executive. Both theories view the peer group process as analytically secondary. It is simply a mechanism by which either market forces or management power is expressed.
We, however, position the process centrally within the dynamics of chief executive compensation theory. We argue that because CEO skills are not transferable from company to company, there is no distinct market price for an executive’s abilities. With firm-specific talent, a bilateral monopoly set-up, where there is a range of possible compensation amounts between both the manager and firm’s best outside opportunities, is more appropriate for analysis. Given the ambiguous nature of determining executive compensation, the peer benchmarking process then has a significant effect on pay decisions by providing the simple decision heuristic, whereby appropriate pay is determined by both boards who are well-aligned with shareholders, and likewise, those who are not.
We support our argument by surveying the empirical evidence of CEO turnover. Studies demonstrate that instances where CEOs jump from one company to another are exceedingly rare. Moreover, the ones who do are atypical, appearing to be primarily M&A experts who are skilled at preparing companies for a sale. Studies also show that, where CEOs and other executives separate from their original company, they experience a sharp decline in position and responsibility. They typically take jobs at much smaller enterprises.
Having concluded that the use of peer benchmarking by boards is unnecessary from a retention standpoint, we assess whether its use is otherwise problematic. It has been observed in both the academic and professional communities that the practice of targeting the pay of executives to median or higher levels of the competitive benchmark will naturally create an upward bias and movement in total compensation amounts. Whether this escalation has been dramatic or merely incremental, the compounded effect has been to create a significant disparity between the pay of CEOs and what is appropriate to the companies they run. This is not surprising. By basing pay on primarily external comparisons, a separate regime that was untethered from the actual wage structures of the rest of the organization was established. Internal consistency, a natural inhibitor to CEO pay growth, was abandoned. Over time, these disconnected systems were bound to diverge.
The pay of a chief executive officer, however, has a profound effect on the incentive structure throughout the corporate hierarchy. High pay thus has costs far greater than the amount actually transferred to the CEOs themselves. It is for this reason that we find peer benchmarking problematic. To mitigate this, we suggest that boards must set pay in a manner that is more consistent with the internal corporate wage structure. An important step in that direction is to diminish the needless focus on external benchmarking. Instead, independent and shareholder-conscious compensation committees must develop internally created standards of pay based on the individual nature of the organization concerned, its particular competitive environment and its internal dynamics.
The paper is available for download at here.