36% More Likely to Outperform: What McKinsey's Study Actually Measured
November 2, 2026 · Frisian News
McKinsey's 2020 Diversity Wins report claimed companies with ethnically diverse executive teams were 36% more likely to outperform financially. A March 2024 paper in Econ Journal Watch found no statistically significant relationship between executive racial diversity and any of six financial performance measures across a comparable S&P 500 dataset.
Thirty-six percent. That single figure from McKinsey's 2020 report Diversity Wins has circulated in boardrooms and HR decks for six years as proof that hiring diverse executive teams improves financial performance. The source is McKinsey and Company. The mechanism, the report says, is outperformance probability. The finding has since been treated as settled science.
It is not settled. Two accounting professors, Jeremiah Green of the Mays School of Business and John R. M. Hand of the University of North Carolina's Kenan-Flagler Business School, published a replication attempt in Econ Journal Watch. They took the S&P 500 as it stood on 31 December 2019 and tested whether executive racial and ethnic diversity at mid-2020 predicted financial results over the prior five years, 2015 to 2019.
Across six separate financial metrics, Green and Hand found no statistically significant relationship. Their tests covered industry-adjusted earnings before interest and taxes margin, sales growth, gross margin, return on assets, return on equity, and total shareholder return. Not one showed a reliable link to the diversity measure McKinsey uses, the inverse normalized Herfindahl-Hirschman index applied to executive racial and ethnic composition.
McKinsey's own methodology compounds the problem. Green and Hand point out that the consultancy measured company financial performance before the diversity snapshot, not after it. If performance in 2019 is used to predict diversity in 2020, the causal arrow runs the wrong way: well-run, profitable companies may be the ones that can afford to pursue diversity goals, not companies that became profitable because they pursued them. McKinsey's studies, as Green and Hand describe them in their paper, are structured in a way that reverses cause and effect.
Methodology problems extend to transparency. McKinsey self-published all four reports in its diversity series, covering 2015, 2018, 2020 and 2023. Green and Hand had to construct a comparable sample from publicly available data, specifically the S&P 500, and work with what they could observe rather than what McKinsey chose.
One honest limit: Green and Hand did not find that diverse executive teams perform worse. Their result is a null result. The replication found no statistically significant relationship in either direction. Companies with diverse leadership may well be well-managed for other reasons. The only thing the Econ Journal Watch paper removes from the table is the specific 36 percent figure as evidence that US publicly traded firms can expect improved financial performance by increasing executive racial and ethnic diversity. That claim, in the words of Green and Hand, should not be relied on.
The figure has already done its work. It filled six years of consultant pitch decks before anyone with access to a comparable dataset ran the test again. The test has now been run. The number did not hold.
Published November 2, 2026 · Frisian News