The after-tax real wage of the average worker in the United States has fallen 13 percent in the last 20 years, while the average chief executive officer has received a pay raise of over 300 percent. This glaring contrast has sparked a flood of papers analyzing CEO compensation contracts. One of the main justifications for the extraordinary pay of top CEOs is that they receive contracts that link CEO compensation to the performance of the firm. The empirical literature, however, has found little evidence that CEO contracts provide such incentives. The compensation of CEOs appears to respond very little to the performance of their firms.
This article addresses three reasons why the previous literature may have been underestimating the response of compensation to firm performance. First, only firms where monitoring the CEOcis costly should have CEO compensation that is performance-sensitive. Restricting the sample to these firms yields a 67 percent increase in the performance sensitivity of compensation contracts. Second, the parameter that measures the performance sensitivity of CEO pay is negatively correlated to performance, causing it to be underestimated in standard regressions. Finally, econometricians do not observe exactly what compensation boards use as performance measures. Correcting this error shows that the elasticity of CEO pay with respect to firm performance is 10 times higher than previously believed.