Corporate governance data and measures revisited

Researchers in accounting, corporate finance, economics, and law regularly evaluate the impact of corporate governance provisions on firm performance and managerial actions. Many of these studies rely on publicly available governance summaries developed by the Investor Responsibility Research Center (IRRC). These summaries have been coded and used in a variety of influential studies, including analyses that assess whether corporate governance practices impact shareholder value. This paper reexamines the coding of the IRRC summaries for the 1990s and early 2000s. We find that the IRRC coding and one commercial source of corporate governance data disagree significantly for several important governance indicators. Our own review of source documents uncovered measurement errors in the IRRC summaries, especially for golden parachutes and supermajority voting provisions. We also find that the IRRC's definitions do not always clearly represent managerial entrenchment or the difficulty of a (hostile) takeover, as many researchers have assumed. To understand whether these issues matter for corporate governance research, we investigate the striking conclusion of prior work that finds that a hedge portfolio composed of a long (short) position in firms with few (many) anti-takeover provisions produces annual excess returns of six to twelve percent per annum from 1990 to 1999. We find that governance data corrected for errors and definitional concerns do not predict excess returns. We conclude researchers should exercise caution using IRRC data from this period, and that much more attention needs to be devoted to measuring corporate governance practices