Print
E-MailThe following post comes to us from Steve Crawford of the Department of Accounting & Taxation at the University of Houston, and Karen Nelson and Brian Rountree, both of the Accounting Area at Rice University.
Will knowing how much the CEO makes relative to rank and file employees provide information to investors? We may soon find out as a result of a provision in the Dodd Frank Act that requires companies to report the ratio of the CEO’s compensation to that of the median employee. A number of different sources have developed industry-based estimates of the ratio using information about CEO pay from corporate disclosures and employee pay from the government’s Bureau of Labor Statistics. For instance, an article in Bloomberg BusinessWeek on May 2, 2013 found the ratio of CEO pay to the typical worker rose from about 20-to-1 in the 1950s to 120-to-1 in 2000, with the ratio reaching nearly 500-to-1 for the top 100 companies.
In our The CEO-Pay Ratio, which was recently made publicly available on SSRN, we take advantage of unique reporting rules for the banking sector, which requires disclosures concerning compensation to all employees, as well as the CEO. With this data, we calculate the ratio of CEO compensation to that of the average employee. Over the years 1995-2012, the ratio is relatively stable with an average of 16.58-to-1. In fact, it is only in the highest decile of CEO pay where we find ratios rising to the levels popularized in the financial press and policy debate. Thus, for the vast majority of corporations in the banking sector we find ratios that are well within the bounds espoused by management experts such as Peter Drucker.
A more important question is whether disclosure of the ratios will influence investor behavior. To provide some evidence on this issue, we investigate whether the ratios we calculate for the banking sector systematically relate to the way investors vote on Say on Pay (SOP) proposals. The Dodd-Frank Act also mandates that all corporations administer a non-binding shareholder vote on the compensation of executives reported in the firms’ annual proxy statements. This portion of the law is currently in effect, providing us with three years of data on the preferences of shareholders as revealed through their voting behavior. We find that voting dissent is greatest at both the lowest and highest levels of the ratio, consistent with information on pay disparity influencing voting behavior. Increased voting dissent at the highest levels of the ratio aligns with arguments that disclosure of the ratio may serve as a catalyst to reign in what investors believe to be excessive CEO compensation. However, it is interesting to note that dissent is also high for banks with the lowest levels of the pay ratio, which could be consistent with the view that some level of pay disparity is necessary to provide appropriate incentives for effort within organizations.
We further examine whether the ratios are predictive of future firm performance and risk to see if investors voting behavior is consistent with underlying firm outcomes. Our findings reveal a similar non-linear relationship where the highest and lowest pay ratios result in the lowest (highest) performance (risk). The economic magnitudes of these effects, however, are relatively small. Thus in the end, it appears that the pay ratio provides significant information concerning shareholder voting behavior, but only limited information about actual economic outcomes.
Overall, the results in our study help to inform the ongoing policy debate on the magnitude and consequences of pay disparity in public corporations. If the Securities and Exchange Commission issues its final pay ratio disclosure rule in 2015, investors may soon have this information to inform their voting decisions for a broad range of firms.
The full paper is available for download here.
