Over the past decade, they have been successful, and the shift to majority voting has been one of the most popular and successful governance reforms.

Yet critics are sceptical as to whether majority voting improves board accountability. Tellingly, directors of companies with majority voting rarely fail to receive majority approval – even more rarely than directors of companies with plurality voting. Even when such directors fail to receive majority approval, they are unlikely to be forced to leave the board. This poses a puzzle: why do firms switch to majority voting and what effect does the switch have, if any, on director behavior?

We empirically examine the adoption and impact of a majority voting rule using a sample of uncontested director elections from 2007 to 2013. We test and find partial support for four hypotheses that could explain why directors of majority voting firms so rarely fail to receive majority support: selection; deterrence/accountability; electioneering by firms; and restraint by shareholders.

Our most dramatic finding is a substantial difference between early and later adopters of majority voting. The early adopters of majority voting appear to be more shareholder-responsive than other firms. These firms seem to have adopted majority voting voluntarily, and the adoption of majority voting has made little difference in shareholder-responsiveness going forward. By contrast, later adopters, as a group, seem to have adopted majority voting only semi-voluntarily. Among this group, majority voting seems to have led to more shareholder-responsive behavior.

These differences between early and late adopters have important implications for understanding the spread of corporate governance reforms and evaluating their effects on firms. Reform advocates, rather than targeting the firms that, by their measures, are most in need of reform, instead seem to have targeted the firms that are already most responsive. They then seem to use the widespread adoption of majority voting to create pressure on the non-adopting firms. Empirical studies of the effects of governance changes thus need to be sensitive to the possibility that early adopters and late adopters of reforms differ from each other and that the reforms may have different effects on these two groups of firms.