Shareholder Votes and Proxy Advisors: Evidence from Say on Pay

In our paper, Shareholder Votes and Proxy Advisors: Evidence from Say on Pay, which was recently accepted for publication at the Journal of Accounting Research, my co-authors (Yonca Ertimur of the University of Colorado at Boulder and David Oesch of the University of St. Gallen) and I examine the economic role of proxy advisors. As non-binding shareholder votes have come to increasingly affect firms’ governance practices, there has been growing interest in understanding the value of proxy advisors’ recommendations, a key driver of shareholder votes.

To shed light on this question, we follow the entire process surrounding proxy advisor activities and examine the analyses underlying proxy advisor recommendations, how firms, stock prices and voting shareholders respond to the release of these recommendations, firms’ reactions to the votes triggered by them and, ultimately, whether they have an impact on firm value. The setting we use for our examination is based on the analyses provided by the two most influential proxy advisors, Institutional Shareholder Services (ISS) and Glass Lewis & Co. (GL) to arrive at voting recommendations for the non-binding shareholder vote on executive pay mandated by the Dodd-Frank Act in 2010, commonly known as say-on-pay (SOP).

We start with an analysis of the SOP-related portion of ISS and GL proxy reports for S&P 1500 firms in 2011. ISS issued recommendations to vote against compensation plans at 11.3% of sample firms while GL issued negative recommendations at 21.7% of sample firms. In both cases, a perceived disconnect between past pay and past performance is the most frequent reason for the negative recommendation. Firms that receive an Against recommendation from GL are not simply a superset of those receiving an Against recommendation from ISS, suggesting that GL not only has a more aggressive stance than ISS but likely also uses a different methodology to arrive at voting recommendations. In contrast to concerns voiced by critics of proxy advisors, we only find limited evidence of a one-size-fits-all approach. In most cases, the simple existence of certain provisions in compensation plans does not mechanically translate into a negative recommendation. Instead, we find that firms with similar potentially controversial provisions end up receiving different recommendations because proxy advisors take into account firm-specific circumstances, the severity of the issue, and the overall compensation plan quality.

We next examine the market reaction to the release of proxy advisors’ SOP recommendations. Controlling for other information and contemporaneous firm-specific news, we find small but significantly negative mean abnormal returns (-0.5% to -0.7%) around the release of ISS reports with a SOP-related Against recommendation, particularly when they are less expected. This result does not appear driven by expected costs potentially associated with negative recommendations (e.g. litigation, sub-optimal changes to compensation plans) and, thus, is more consistent with unexpected ISS Against recommendations revealing information about compensation plan quality.

We also investigate the association between shareholder votes and the SOP-related part of proxy advisor reports. Compensation plans are voted down at 2% of the sample firms only, but votes against the plan exceed 20%, a threshold viewed as an indication of considerable shareholder dissatisfaction, at more than 15% of the firms. We show that proxy advisor recommendations are a key determinant of voting outcomes: a negative recommendation from ISS (GL) is associated with an increase of votes against the compensation plan by 24.7% (12.9%). Further tests reveal that the association between proxy advisor recommendations and votes is not higher for SOP than for other items, contrary to the concern that mandatory SOP votes would lead more investors to rely on proxy advisors to avoid costly analyses of thousands of compensation plans. Finally, we show that the association between proxy advisors’ Against recommendations and shareholder votes depends on the institutional ownership structure, the specific rationale behind the recommendation, and certain firm characteristics. This evidence suggests that at least some investors do not follow the recommendations blindly but take into account the underlying rationale and other firm-specific factors.

Our next set of tests investigates how firms respond to SOP votes triggered by negative recommendations. We find that 55% of sample firms with a 2011 negative ISS recommendation report compensation changes explicitly made in response to the 2011 SOP vote, a remarkably high number given firms’ documented reluctance to respond to non-binding votes. Firms’ responsiveness is proportional to the extent of dissent and exhibits a sharp discontinuity around the 30% dissent threshold. This further manifests ISS’ significant influence as ISS had stated after the 2011 proxy season that firms failing to respond to dissent above 30% would receive negative recommendations on SOP proposals and the election of compensation committee members in 2012.

Our study concludes with an attempt to address the question of whether proxy advisor recommendations are beneficial or detrimental for shareholder value. This question is difficult because we do not have direct ex post measures of the quality of compensation plans that would allow us to validate the quality of proxy advisors’ SOP recommendations. To shed some light on the issue, we investigate the stock price reaction to announcements of compensation changes triggered by negative proxy advisor recommendations and the following SOP vote. We do not find a statistically significant positive or negative share price reaction, even for the subset of the most substantial compensation changes, suggesting that proxy advisors’ key economic role is to process and organize a substantial amount of compensation information for institutional investors rather than identifying and promoting superior compensation practices.

Our study contributes to the literature on proxy advisors and shareholder voting. We are the first to examine the analyses underlying proxy advisor recommendations, and we are able to do so in a setting where such analysis is especially complex. We thereby provide answers to a number of questions that are of interest to academics, policy-makers, and practitioners, such as how recommendations and underlying analyses differ across proxy advisors or to what extent they reflect a “one-size-fits-all” approach. Finally, our paper is the first to provide evidence on compensation changes made by U.S. firms in response to SOP votes and the market’s response to such changes.

The full paper is available for download here.