The Say-on-Pay Vote - Does It Change Anything?
The say-on-pay vote is an annual advisory ballot that gives shareholders the opportunity to approve or disapprove of a company's executive compensation program. Mandated by the Dodd-Frank Act in 2010, the vote does not bind the board to any specific action. Even a company that receives less than 50% approval is free to maintain its existing compensation structure. This advisory nature leads many investors to dismiss say-on-pay as toothless. The data, however, tells a more complicated story.
Since 2011, when say-on-pay votes became mandatory for large public companies, the vast majority of votes have passed comfortably. The average approval rate across the Russell 3000 is approximately 90%. Fewer than 3% of companies receive less than 50% support in any given year, and fewer than 10% receive less than 70% support. These numbers suggest that the mechanism has little practical impact. But the averages mask the behavioral changes that occur at the margins, where companies facing low vote results or the prospect of a failed vote make substantive compensation changes to avoid shareholder rejection.
How Say-on-Pay Works
Section 951 of Dodd-Frank requires that public companies hold a non-binding shareholder vote on executive compensation at least once every three years. Most companies hold the vote annually. The vote covers the compensation disclosed in the proxy statement, including the Summary Compensation Table, the CD&A, and the related compensation tables. Shareholders vote to approve or disapprove the overall program, not individual compensation elements.
The proxy statement frames the vote with a resolution along the lines of: "RESOLVED, that the shareholders approve, on an advisory basis, the compensation paid to the Company's named executive officers as disclosed in this proxy statement." The board recommends a vote "for," and the proxy advisory firms, ISS and Glass Lewis, issue their own recommendations based on their evaluation of the pay-performance relationship.
The ISS recommendation is particularly influential. ISS evaluates say-on-pay proposals using a quantitative pay-for-performance model that compares the CEO's total compensation to the company's TSR relative to peers. If the model identifies a significant pay-performance disconnect, ISS will recommend a vote against say-on-pay. Research estimates that an ISS "against" recommendation reduces the approval rate by approximately 20-25 percentage points, making it the single most important factor in say-on-pay vote outcomes.
What Happens When the Vote Fails
Although rare, say-on-pay failures have concrete consequences. When a company receives less than 50% support, the board faces intense pressure from shareholders, proxy advisors, and the media to make changes. Research by Ertimur, Ferri, and Oesch documented that roughly 55% of companies that failed say-on-pay in one year made substantive compensation changes before the next proxy season. These changes commonly included:
Reduction in total CEO compensation. Increased weighting of performance-based pay relative to fixed or time-based components. Changes to performance metrics to better align with shareholder value creation. Elimination of controversial provisions such as excise tax gross-ups, repriced options, or discretionary bonuses. Enhanced disclosure in the CD&A to explain compensation decisions more transparently.
Companies that made these changes typically saw their say-on-pay approval recover to normal levels in the following year. Companies that did not make changes often faced continued low approval and, in some cases, director election campaigns or activist interventions.
The failed vote at Citigroup in 2012 was a landmark event. Shareholders voted down CEO Vikram Pandit's compensation package, which included a $15 million retention award on top of his $1 salary and existing equity compensation. The company subsequently redesigned its compensation structure, and Pandit departed later that year. While the failed vote did not directly cause the CEO change, it contributed to the broader accountability dynamic that made the board's position untenable.
The Discipline Effect
The most significant impact of say-on-pay may not be the failed votes but rather the compensation changes companies make to prevent failed votes from occurring. This discipline effect is harder to measure but likely more consequential than the small number of outright failures.
Before Dodd-Frank, compensation committees operated with minimal external feedback on their decisions. The annual meeting might include questions about pay, and shareholder proposals could address specific compensation practices, but there was no systematic, company-wide vote on the compensation program as a whole. Say-on-pay created that mechanism, and the prospect of a negative vote, with its attendant media coverage and proxy advisor scrutiny, has changed behavior at the margins.
Compensation consultants report that their clients now explicitly consider how ISS and Glass Lewis will evaluate proposed pay packages before finalizing them. The proxy advisory firms' quantitative models, which compare CEO pay to peer-group pay and company TSR to peer-group TSR, have become de facto constraints on compensation design. This is not because the advisory firms' models are necessarily correct. It is because a negative recommendation materially reduces the approval vote, and boards prefer to avoid that outcome.
The empirical evidence supports the view that say-on-pay has moderated compensation growth. Research by Correa and Lel found that CEO pay sensitivity to poor performance increased significantly after the adoption of say-on-pay. CEOs at companies with low say-on-pay support experienced larger pay cuts in subsequent years than CEOs at companies with high support, controlling for performance. This suggests that the vote creates a feedback mechanism that makes compensation committees more responsive to the pay-performance relationship.
The Limitations
Say-on-pay's limitations are substantial and should temper expectations about what the mechanism can achieve.
Advisory status means the vote has no legal teeth. A determined board can ignore even a majority rejection, though the reputational and practical costs of doing so are significant.
The all-or-nothing structure forces shareholders to vote on the entire compensation program rather than specific elements. A shareholder who approves of the base pay structure but objects to a specific retention award must vote against the entire package or approve everything. This blunt instrument provides less precise feedback than element-specific votes would.
Retrospective rather than prospective. The vote covers compensation that has already been earned and, in many cases, already paid. By the time shareholders vote, the compensation decisions are largely irreversible. The vote can influence future compensation design, but it cannot undo past awards.
Institutional voting patterns create a high baseline approval rate that makes failures rare. Many institutional investors follow internal policies or proxy advisor recommendations that result in supporting say-on-pay unless specific quantitative triggers are met. This means that only the most egregious pay-performance disconnects result in failed votes. Moderate disconnects fly under the radar.
Low individual investor participation. Retail shareholders participate in proxy voting at rates well below institutional shareholders. This means that say-on-pay outcomes are overwhelmingly determined by institutional investors and proxy advisory firms, whose interests and incentive structures may not perfectly align with those of individual shareholders.
Gaming the metrics. Sophisticated compensation consultants help companies design pay packages that satisfy the proxy advisory firms' quantitative models while preserving the compensation levels management desires. If the ISS model flags a CEO's pay as excessive relative to peers, the solution might be to adjust the peer group, restructure the timing of equity grants, or shift compensation from one category to another, rather than to actually reduce pay.
When Low Votes Are Informative
Even when the say-on-pay vote passes, the level of support provides useful information. A company that receives 95% approval has a compensation program that shareholders broadly endorse. A company that receives 72% approval has a compensation program that a significant minority of shareholders find problematic.
Trends over time are more informative than any single year's result. A company whose say-on-pay support has declined from 92% to 75% over three years is experiencing growing shareholder concern about compensation, even though the vote has passed comfortably each year. A company whose support has declined to 65% is approaching the threshold where proxy advisory firm and institutional investor attention intensifies.
ISS recommends that companies receiving less than 70% support engage with shareholders to understand their concerns and disclose the results of that engagement in the following year's proxy statement. Companies that take this engagement seriously and make responsive changes tend to recover their approval levels. Companies that dismiss low votes and make no changes tend to see continued erosion.
The Broader Accountability Ecosystem
Say-on-pay does not operate in isolation. It is one component of a broader accountability ecosystem that includes proxy advisory firm recommendations, institutional investor engagement, shareholder proposals on specific compensation practices, director election results, and activist campaigns. The interaction among these mechanisms creates more accountability than any single mechanism provides alone.
A company that receives low say-on-pay support may also face shareholder proposals requesting specific compensation changes, withhold campaigns against compensation committee members, and private engagement from large institutional investors demanding reform. The cumulative pressure from these multiple channels is considerably stronger than any individual mechanism.
The compensation committee members themselves face personal accountability through director elections. At companies where say-on-pay support is low, proxy advisory firms may recommend against the reelection of compensation committee members. This personal risk creates an additional incentive for committee members to ensure that compensation programs are defensible.
Say-on-pay has not transformed executive compensation. CEO pay continues to grow, and the most extreme packages continue to draw criticism. But the mechanism has created a structured, regular, and visible feedback channel between shareholders and compensation committees that did not exist before. That channel has moderated the worst excesses and created pressure for better pay-performance alignment, even if it has not eliminated the fundamental challenges of executive compensation governance.
Put these principles into practice. Track fundamentals, build portfolios, and analyze stocks with AI-powered insights.
Start Free on GridOasis →