Why non-binding say-on-pay votes still cut executive compensation
Dodd-Frank requires an advisory vote on executive pay, but boards respond anyway. Here's why the threat of failure shapes compensation, and when weak votes trigger change.
When shareholders cast a say-on-pay vote, they are casting a vote that carries no legal weight. The board is not required to follow the result, and the vote cannot force a company to change executive compensation. Yet companies spend considerable time and resources responding to negative votes, and research shows executives earn measurably less, on average, simply because the threat of a failed vote exists. The paradox reveals how non-binding votes can drive real changes: boards anticipate rejection and adjust pay accordingly, and when rejection actually occurs, the signal forces immediate action.
The Dodd-Frank Wall Street Reform Act, enacted in 2010, created this advisory mechanism through Section 951, which requires public companies to put executive compensation to a non-binding shareholder vote at least once every three years. Section 951 also requires companies to let shareholders vote every six years on whether to hold that vote annually, every two years, or every three years. The vote result carries no obligation, but boards have learned to treat it as a serious governance test.
How the rule works and what boards must disclose
The SEC's rule, finalized in 2011, defines say-on-pay simply: it is an advisory vote on the compensation of named executive officers as described in the company's proxy statement. Companies must include the say-on-pay vote in proxy materials and disclose the results. In the proxy's Compensation Discussion and Analysis section, boards must explain "whether, and if so how, companies have considered the results of the most recent say-on-pay vote." That disclosure requirement is the rule's lever: if a vote performs poorly, the company must publicly account for its response or explain why it chose to disregard shareholder opposition.
Companies can hold say-on-pay votes annually rather than every three years, depending on how shareholders voted in the frequency vote. A separate, related vote lets shareholders weigh in on golden parachute severance arrangements.
Why boards respond to a vote they can ignore
A company receiving strong opposition has several incentives to act despite the vote's advisory status. First, the vote is a public signal of shareholder unhappiness, visible to investors, media and analysts. A vote below 80 percent support typically triggers engagement. In the 2024–2025 annual meeting season, 25 Fortune 1000 companies received less than 80 percent support, and in preliminary 2024 data, three S&P 500 companies received less than 50 percent support outright.
Second, the board faces reputational and governance risk. Major shareholders often coordinate to vote against compensation if their concerns are ignored. If a company receives a weak vote, institutional investors and proxy advisers treat it as a warning signal. Some shareholders threaten votes against board members if compensation issues persist.
Third, and less visible, research shows boards adjust pay downward to avoid triggering a vote failure. A model of say-on-pay's effects on U.S. CEO compensation finds that providing shareholders this vote "lowers total CEO pay levels by about 6.6 percent" and "increases firm value by 2.4 percent, on average," compared with not having the vote at all. This effect operates through what researchers call the threat mechanism: boards factor in the cost of potential vote failure when setting packages, so the mechanism works even when failure is uncommon. Shareholder support for say-on-pay proposals has averaged around 93 percent, with roughly 7 percent of votes considered a failure, yet the mere possibility of that failure drives compensation decisions.
When votes actually fail and what companies do next
Failures are rare. In 2023, 13 of the S&P 500 companies that held say-on-pay votes received less than 50 percent support; in preliminary 2024 data, only three S&P 500 companies did, as the share of failed votes fell by more than half year over year. But when failure happens, boards respond with concrete changes. Among companies with failed votes, the average number of compensation adjustments reached 2.5 in the year following the rejection. The most common response, accounting for 66 percent of failing companies in 2023, was to adjust the metrics or weightings used in incentive plans—for example, changing the targets in a bonus formula or the performance metrics tied to incentive awards.
In some cases, the changes are structural. Pebblebrook Hotel Trust raised its target performance goal from the 50th to the 55th percentile among peers, and Norwegian Cruise Line Holdings adopted an adjusted EPS metric requiring the company to outperform the S&P 500 Index. Riot Platforms added detail to its proxy statement explaining the range of equity awards granted to executives. Most companies combine more than one such change: two changes in a single year was the most common outcome, made by about a quarter of the 77 companies Equilar studied.
“Boards adjust pay downward to avoid triggering a vote failure, with research showing the vote lowers total CEO pay levels by about 6.6 percent compared with not having the vote at all.”
How shareholder engagement influences votes between formal ballots
Board responsiveness is not limited to the year after a failed vote. Before a vote even takes place, companies facing low support in preliminary indicators engage intensively with shareholders. In one recent study of Fortune 1000 companies with adverse votes, boards conducted broad direct outreach, with independent directors participating in meetings. PENN Entertainment and UnitedHealth reached between 48 and 60 percent of outstanding shares through direct shareholder engagement. Goldman Sachs conducted more than 120 meetings with representatives of shareholders owning more than 45 percent of shares.
This engagement typically combines dialogue with concrete action. Companies address shareholder concerns about specific practices—off-cycle awards, severance multiples, or performance metric design—and make changes tied to feedback. The results show strong recovery. All 14 Fortune 1000 companies examined in one study that received adverse votes showed improved support in the following year. Improvements ranged from 3.9 percent to 66.1 percent. Companies achieving the largest gains, like Tutor Perini at 66.1 percent and Otis at 54.5 percent, combined broad shareholder engagement with meaningful compensation redesigns.
What the threshold for trouble looks like
Because say-on-pay votes pass so frequently, boards and compensation committees watch for softer warnings. Votes below 70 percent, while still passing, typically trigger the kind of shareholder engagement and compensation committee review that resembles a post-failure recovery.
The SEC's disclosure rule requires companies to address weak votes in their next proxy statement, explaining either the changes made or why shareholder concerns were deemed unfounded. This explanation is rarely credible when support has dropped dramatically. Boards that attempt to defend compensation after receiving minority support typically face pressure from governance advocates and sometimes face votes against board members the following year.
The non-binding nature of say-on-pay has thus created a system where formal rejection is rare but the anticipation of it is constant. Companies model compensation decisions partly around shareholder preferences, knowing that a vote failure would require immediate remediation and carry governance consequences. The mechanism is diffuse but effective: it channels shareholder pressure through a vote that has no legal force.
Related coverage: How CEO compensation packages combine cash, equity and perks; How boards tie executive pay to revenue, profit and stock returns; How Executive Compensation Is Actually Set.
