Corporate Governance Responses to Director Rule Changes

Much of the corporate governance literature has been plagued by endogeneity problems. In our paper, Corporate Governance Responses to Director Rule Changes, which was recently made publicly available on SSRN, we use a law change as a natural experiment to test how firms adjust the choice and magnitude of governance tools given a floor level of monitoring from independent directors. Through this analysis, we can recover the structural relationship between inputs in the governance production function. We study these relationships with a new board of director dataset with a much larger range of firm size.

In 2002, U.S. stock exchanges and the Sarbanes-Oxley Act established minimum standards for director independence. These director rules altered firm choice of other tools for mitigating agency problems. On average, treated firms do not increase the size of their board, instead inside directors are replaced with outside directors.

Our primary results test for changes in governance and provide a unique insight on the production function for governance. We find three economically significant changes.

• First, interlocking directorships increased at treated firms, by two interlocking directorships. This is consistent with Jiraporn, Singh, and Lee (2009) who document an increase in multiple board seats post-SOX across all firms. Because treated firms in our sample did not increase their board size on average, the interlock result suggests that these firms replaced inside directors with outside directors that serve on boards of other companies.
• Second, CEO ownership decreased 0.8 percentage points at treated firms relative to untreated firms which is equivalent to a 6 percent decrease. Given that tests on turnover rates show that treated firms were no more likely to change management, the decrease in CEO ownership was the result of continuing managerial divestment of shares.
• Third, we find the share of CEO compensation that are equity shares decreased 1–2 percentage points at treated firms, which is about one-third to one-half of the average share. Overall, these results suggest that CEO incentive pay moved away from long-term incentives based on ownership. This is consistent with those of Chang, Choy, and Wan (2012). They find an overall decrease in CEO ownership and pay-performance sensitivity after SOX among the S&P 1500 firms without controlling for the extent to which individual firms were treated by SOX. They interpret the reduction in CEO ownership as a reaction to the more stringent regulatory environment. Our results suggest that this adaptation is more specifically tied to the Director Rules.

Because treated firms do not outperform the market, these results are more consistent with governance reoptimization than with managerial entrenchment or governance improvement explanations.

Overall, these results are likely disappointing for advocates of director reform. The combination of substitution effects among governance strategies and no performance improvement of treated firms suggests that these firms had good reasons for their initial governance choices. The resulting homogenization of firm specific governance choices are notable unintended consequences of the reforms. A major policy implication of this work is that regulators must take into account countervailing corporate action when trying to improve specific areas of governance. Failing to do so is likely to raise costs (pecuniary and otherwise) without improving governance.

The full paper is available for download here.