Economic Characteristics, Corporate Governance, and the Influence of Compensation Consultants on Executive Pay Levels

This post is from David F. Larcker of the Stanford Graduate School of Business. Posts by Brian Cadman and Tatiana Sandino, available here and here also analyzed the role of compensation consultants in setting pay.

In a recent working paper, Christopher Armstrong, Christopher Ittner and I investigate the relation between the use of compensation consultants and CEO pay levels. We conduct an analysis using proxy disclosures by a diverse sample of 2,116 companies. Consistent with claims that executive pay levels in clients of compensation consultants are higher than justified by economic characteristics, ordinary least squares (OLS) regressions that control for a wide variety of economic determinants of compensation indicate that total pay is higher for clients of most (but not all) of the consulting firms relative to companies without consultants. The OLS results also suggest that pay levels of clients of the larger, most frequently used compensation consultants are higher than those of firms using other consulting firms (most of which are smaller, boutique compensation consultants) in some model specifications. However, when more sophisticated propensity score matched pair analyses are used to relax the stringent functional form assumptions imposed by OLS models and to assess correlated omitted variables problems, most differences between the individual consulting firms disappear, though the statistically higher levels of total pay at companies using compensation consultants persist.

When we add governance variables, we continue to find higher pay in clients of most consulting firms in OLS regressions. In contrast, we find no significant differences in total pay levels between users and non-users of consultants or among the various consulting firms when propensity score matched pair analyses are used. This evidence indicates that once companies with similar economic and governance characteristics are compared and OLS’s strict functional form is relaxed, pay levels are not significantly different, suggesting that governance differences account for much of the unexplained pay differences between consultant users and non-users.

Further analysis indicates that these results are due (at least partially) to pay levels for clients of individual consulting firms varying with governance strength, with weaker governance within clients of a given consultant associated with higher total pay. These results suggest that the higher pay found in consulting clients is at least partially explained by the link between weaker governance and higher pay in companies using consultants. This is consistent with the rent extraction view of the association between compensation consultant use and CEO pay. Finally, we find no support for claims that CEO pay is higher for clients of potentially “conflicted” consultants that offer a broad range of advisory services relative to clients of specialized, “non-conflicted” compensation consulting firms.

The full paper is available for download here.

Both comments and trackbacks are currently closed.