Shareholder Vote on Golden Parachutes: Determinants and Consequences

Since the 1980s, the federal government has repeatedly attempted to influence pay-setting for top managers at public companies. Most recently, Congress and the SEC have attempted to amplify the voice of public company shareholders on executive compensation by requiring advisory shareholder votes. These two interventions, known as “Say-on-Pay” and “Say-on-Golden-Parachute,” were promulgated under the Dodd-Frank Act and promised to focus and identify shareholder outrage over problematic pay practices. Say-on-Pay (“SOP”) asks shareholders to vote on the previous year’s executive pay practices in their entirety, while Say-on-Golden-Parachute (“SOGP”) asks shareholders to pass on merger-related severance payments that would become payable to executives when the change-in-control takes place.

SOP in the United States and its cousins around the world have received a good deal of attention from both practitioners and scholars. A series of recent studies found that SOP in fact influences corporate behavior. The studies also present a somewhat sanguine picture of proxy advisor recommendations, which are strongly correlated with subsequent shareholder voting, as well as shareholder voting itself. In a recent working paper, titled Shareholder Voting on Golden Parachutes: Determinants and Consequences, we look at the other “Say-on” provision in Dodd-Frank. Utilizing hand-collected data on golden parachutes, we analyze the first six years’ experience with SOGP votes to answer questions surrounding this expansion of shareholder power: how effective the law has been in influencing pay practices, how proxy advisors have responded to the new vote, and how shareholders behave when given this new power. Our central finding is that SOGP does not function like SOP and, in fact, may be more problematic across the board.

While both SOP and SOGP rely on advisory voting by the same groups of shareholders, important differences separate the two. As an advisory vote, SOP necessarily relies on indirect pressure on corporate directors. Most commonly, this entails an implicit or explicit threat to subsequently remove directors or discipline executives who fail to respond to shareholder and proxy advisor opinion. By contrast, the potential for future discipline of managers is more limited in the SOGP context. Directors will not be up for re-election by the same SOGP shareholders since the takeover triggering the golden parachute usually spells the end of the directors’ service at the firm or the advent of a vastly different shareholder base. Furthermore, executives who might be asked to renegotiate golden parachutes in the face of shareholder pressure may have little incentive to appease shareholders since their employment with the firm is usually ending as well. In the absence of any explicit or implicit second-stage disciplinary mechanism, directors and executives may have little or no incentive to eschew outsized change-in-control severance payments.

Theory thus suggests that the effectiveness of an SOGP regime may be compromised when compared to SOP. If so, we would expect to observe no effect on golden parachute incidence and levels after the onset of SOGP. We would expect few, if any, last-minute amendments ahead of a SOGP vote aimed at mollifying shareholders about to vote on golden parachutes. Also, we would expect to observe a number of other dissimilarities with the SOP experience. To the extent that their institutional shareholder clients care less about the issue, proxy advisor recommendations may be more likely to rely on simple, one-size-fits-all criteria so as to economize their resources. Relatedly, actual SOGP votes may be more highly correlated with proxy advisor recommendations if shareholders are less willing to expend their own resources to sort through the merits of SOGP votes. Finally, there may be less SOGP dissent than SOP dissent overall if shareholders perceive little chance of effecting a change in golden parachutes.

Our paper empirically assesses these hypotheses in order to shed light on the experience with SOGP. Specifically, we aim to assess the basic question of SOGP’s consequentiality for golden parachute pay practices. In doing so, the paper adds to the literature evaluating advisory votes at public companies and the behavior and influence of proxy advisory firms. Our empirical assessment can be roughly divided into three questions. First, we examine which factors are related to (or possibly influence) proxy advisor recommendations. We have collected data on recommendations from ISS since the inception of SOGP, and we find that several attributes are correlated with its SOGP recommendations. For instance, we find that the size of a golden parachute is significantly related to ISS’s Against recommendations, but only for parachutes in the top quartile of packages. ISS states that it looks for problematic provisions in golden parachutes—tax gross-ups, single triggers, and golden parachutes’ cash awards being three or more times larger than annual pay—when making their recommendation decisions. Although the data on whether components of the golden parachute are single or double trigger is not readily available, we are able to at least corroborate that the presence of a tax gross-up provision is significantly and positively related to ISS’s Against recommendations, consistent with ISS’s stated policy.

Second, we look to find the determinants of shareholder SOGP voting. Despite the lack of a strong disciplinary mechanism, we do find that shareholders, on occasion, vote against golden parachutes rather than simply rubber-stamping them. In terms of what determines such outcomes, we find that ISS Against recommendations explain by far the most significant amount of the variation in voting and more than has been found with respect to SOP voting, consistent with the hypothesis that shareholders do not take SOGP as seriously as they do SOP. At the same time, the firm’s prior performance, as measured by the return on assets (ROA), also seems to play an independent, albeit lesser, role in determining voting outcome. That is, the firms that perform better financially seem to attract lower dissent from their shareholders. The latter result is consistent with the existing scholarship on SOP that shows that firm performance is correlated with the voting outcome.

Third and finally, we examine the effect of SOGP on golden parachutes. First, looking at the broad trends in golden parachutes, we examine how they have evolved before and after the advent of SOGP under Dodd-Frank. Even though SOGP itself may lack any direct disciplinary force, one possibility may be that the SOGP could put some indirect pressure on the growth of golden parachutes over time. However, when we look at the absolute size of golden parachutes, we find that they grew at a faster rate after the adoption of SOGP than before, even after controlling for an extensive set of plausible variables. Second, we look at changes to golden parachutes during the period one year prior to the public announcement of a deal through its closing to see if firms amend existing golden parachutes in anticipation of SOGP votes. We find that firms that amend golden parachutes during the run-up to a deal are more likely to increase (rather than reduce) them over the previously-promised golden parachutes than are firms that do not amend their contracts. This suggests that firms do not act out of fear of negative shareholder votes, contrary to the experience with SOP.

Our findings suggest that SOGP may not be achieving its desired purpose of constraining golden parachutes. To the extent that one believes golden parachutes reflect arms-length bargaining by executives, boards, and, potentially, acquirers, that is of little consequence. If, on the other hand, one believes golden parachutes are often inefficient contractual terms requiring regulatory intervention, this poses a problem. Based on the empirical findings, we analyze a few possible policy changes to the existing regime. One possible answer to SOGP’s inconsequentiality is to make the vote binding somehow. Unlike SOP, SOGP votes are being taken before the executives are to receive the severance payments. By making the SOGP votes binding, we can provide more meaningful formal, disciplinary tool to the shareholders. Another option is to try to harness the implicit, market-based discipline on the directors and the executives through more robust disclosure. Two groups, in particular, come to attention: the target company directors who also serve (or expect to serve) as directors on other companies and the directors of the acquiring corporation. For instance, to the extent that the target company directors, who approve golden parachutes that receive strong negative votes from target shareholders, are serving as directors at other companies, by disclosing such fact to the shareholders of the other companies, one could attempt to impose some implicit discipline on the directors. Similarly, more robust disclosure to the acquiring corporation’s shareholders could generate indirect pressure on the target directors through the acquiring corporation.

The complete paper is available here.