Say-on-Pay Passed More Easily in 2026. The Special Award Is Still the Trigger.

Say-on-pay got easier to pass in 2026, and the remaining failures got easier to explain. FW Cook's review of S&P 500 results, published on the Harvard Law School Forum on Corporate Governance, found 74.8% of companies clearing 90% support, up from 70.0% a year earlier, while the share below 70% slipped to 5.0% from 5.9% ^[1][2][3]. Five companies failed outright, the same count as 2025, and every one of them had granted an outsized equity award ^[4][5].

The low-support cohort is smaller and more uniform in cause. Twenty-two S&P 500 companies landed below 70%, and 21 of those carried an adverse ISS recommendation ^[6][7]. Large special awards were cited in 11 of the 22, which is exactly half, tying with poor pay-for-performance alignment as the most common concern ^[8][9][10]. Incentive design, discretion, or weak goal rigor appeared in eight, and the issues frequently overlapped ^[11][12].

Ownership structure remains the strongest single predictor of surviving an ISS "against." ISS opposed 36 S&P 500 proposals; companies with controlled or concentrated ownership made up a third of that group and averaged roughly 81% support, lifting the overall average to 64.9% ^[13][14][15]. Strip them out and the 24 widely held companies averaged 56.9%, lower than any pre-pandemic season in the decade FW Cook reviewed, with support topping out in the mid-70s ^[16][17][18]. That is a gap of about 24 points between the controlled and widely held sets ^[19]. An adverse recommendation is not fatal, but for a widely held issuer it functions as a ceiling.

FW Cook also looked at award size directly, screening non-new-hire special awards of at least $15 million and excluding new-hire grants because those often replace forfeited pay ^[20][21]. Below $50 million, outcomes scattered. All five widely held companies that cleared 80% support had awards of roughly $30 million or less plus a favorable ISS recommendation ^[22]. But two companies with awards of roughly $29 million and $40 million drew support only in the high 50s: the smaller was entirely time-based and granted as the executive moved into a reduced role, the larger a promotional performance award with undisclosed goals layered on top of continued annual grants ^[23][24][25].

Above $50 million the pattern tightened. Among widely held companies granting at least that much, support peaked at roughly 70%, and all four failed votes in the special-award sample involved awards of at least that size, including two fully performance-based grants of approximately $50 million and $60 million ^[26][27][28]. Performance conditions did not rescue them. FW Cook is explicit that the $50 million line describes a small sample rather than a market threshold, but the direction is consistent: support generally declined as award value rose, and vesting terms and other design features provided less protection at higher values ^[29][30]. Magnitude, prior special-grant history, and the performance backdrop outweighed the design defenses committees usually rely on ^[31].

One arithmetic note worth carrying into next season: four of the five failures sit inside the $15 million-plus non-new-hire sample, meaning one did not, so the trigger is the size and story of the grant rather than that specific screen ^[32].