![Sample our Area Studies journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5922/TOC-Subject-Sample-2021-Area-Studies.jpg"})
ABSTRACT
In this study, we draw on the reputation and product quality literature to examine whether compensation consultants with higher reputation have incentives to provide higher-quality services in terms of CEO pay level and pay performance sensitivity (PPS). Using 1,827 US firm-year observations for 2009 and 2010, we find that consulting firms with a higher reputation, proxied by consultant size, are associated with lower compensation and higher PPS. We also find that the PPS for consultants that also provide non-compensation-related services (NCS) is significantly positive for consultants with a higher reputation, consistent with the reputation protection explanation, and significantly negative for the other consultants with a lower reputation consistent with the conflict of interest argument.
Disclosure statement
No potential conflict of interest was reported by the author.
Notes
1. See Murphy (Citation1999) for more details about the use of compensation consultants.
2. Another manifestation of these concerns is that, as part of its rulemaking under the Dodd–Frank Wall Street Reform and Consumer Protection Act of 2010 (Dodd–Frank Act), on 20 June 2012, the Securities and Exchange Commission adopted new Rule 10C-1 to expand disclosures regarding compensation consultants.
3. See Cadman, Carter, and Hillegeist (Citation2010) p. 269 for details of the three proxies to capture the potential conflicts of interest between compensation consultants and compensation committees.
4. In the context of the IPO industry, Megginson and Weiss (Citation1991) measure the market share using the dollar size of an IPO contract . Similarly, we use the market share using the size of the compensation contract, as measured by the CEO compensation.
5. This cutoff is arbitrary and based on trials with different cutoffs. Since the results disappear with cutoffs at 4 and above we select this cutoff to represent reputable and less reputable consultants.
6. See Section 2.2 for the three indirect proxies for conflicts of interests.
7. The six factors are: (a) any stock of the listed company owned by the consultant, (b) any business or personal relation of the adviser with a member of the compensation committee, (c) any business or personal relationship between the executive officers of the listed firm and the consultant, (d) the provision of other services to the listed firm, (e) the amount of consultant fees, (f) the policies of the firm employing the consultant.
8. Other terms used in the economics literature include externalities, external effects, or external economies (see Mishan Citation1976). These spillovers arise from interdependencies between production functions for both the services. See also Simunic (Citation1984).
9. Instead of measuring performance using ROA, we also use the annual buy-and-hold return on the firm’s stock, industry adjusted ROA, industry-adjusted returns to measure performance. The results are similar.
10. To exclude the effect of the client itself from REPU1, we repeat all the tests reported in the study by deducting 1 from the number of clients of each consultant. Similarly, we repeat all the tests reported in the study by deducting the CEO compensation of the client itself from REPU2. We find that all empirical results do not change qualitatively.
11. The results are similar if we calculate the ratio using the contemporaneous compensation for CEOs.
12. We also compute the clustered standard errors based on the consultants hired. The results are similar.
13. In additional tests, we also measure NCS_RATIO using the fees paid to NCS divided by total fees (the sum of NCS fees and compensation-related fees). The results are the same.
14. We follow Murphy and Sandino (Citation2010) to focus our analyses on the firms that hire a consultant. The results are the same if we use the full sample and introduce a dummy variable to capture the use of a consultant.
15. When multiple consultants are used, we focus on the primary consultant that provide advises to compensation committees. The results are the same if we remove the firms that hire multiple consultants.
16. We calculate the variance inflation factor (VIF) to examine if the multicollinearity significantly influences our empirical results for all the regression analyses we perform in the study. We do not find any case where the VIF value is greater than 10, suggesting that multicollinearity is not a serious problem. For brevity, we do not report the results.
17. Our model is in log-level form, and the coefficient *100 can be interpreted as the percentage change in compensation for a unit increase in ROA. Thus, 0.107 (a one standard deviation increase) * 1.529 (the coefficient on ROA*BIG3)*100 = 16.3%.
18. In our sample, there are 15 (28) firms voluntarily disclosing NCS fees in 2009 (2010) because their NCS fees are below $120,000. To avoid the bias from the factors that cause the voluntary disclosure, we also exclude these firms that voluntarily disclose. This does not qualitatively change the results.
19. Our results are robust to using a logit model instead of a probit model to calculate propensity scores.
20. We match without replacement. If we impose the constraint that the matching firm selected be within a distance of 0.01 of the treatment firm’s propensity score to guarantee the similarity of the observable variables between the treatment and control sample, the results are the same.
21. While we control for various CEO characteristics in our regression analysis, the CEO power index suggested by Morese, Nanda, and Seru (Citation2011) is more comprehensive and takes into account several dimensions that have been considered in the literature (see p. 1792).