McKinsey's Diversity Wins: What the Methodology Actually Shows

The McKinsey "Diversity Wins" series is the most-cited evidence in corporate DEI business-case arguments. The three reports -- 2015, 2018, 2020 -- claimed that companies in the top quartile for diversity in executive teams are 15%, 21-33%, and 25-36% more likely to have above-median financial performance. Those numbers appear in thousands of corporate ESG reports, DEI policy documents, and board presentations.

The methodology underlying those numbers has significant problems. This is not a claim that diversity does not improve organizational performance. It is a claim that the McKinsey reports do not actually prove what they are being cited to prove.

Key findings

• McKinsey's three Diversity Wins reports (2015, 2018, 2020) used a methodology that does not establish causation.
• The studies correlate diversity quartile with above-median EBIT performance but do not control for major confounders including industry, company size, founding date, or country legal environment.
• A 2021 SSRN replication study found the correlation is not strong when controlling for country and industry.
• The direction of causality is unestablished: profitable companies may be more able to afford DEI programs, producing a correlation that runs opposite to the implied direction.
• McKinsey's consulting practice generates revenue from DEI advisory work, creating a conflict of interest in its research publications.
• Scott Page's peer-reviewed research on the diversity bonus (2017) provides a stronger methodological basis for diversity's impact on complex problem-solving, though with a narrower scope than McKinsey claimed.

The correlation vs. causation problem

McKinsey's headline finding is: companies in the top diversity quartile are more likely to have above-median financial performance. This is a correlation. It does not tell you which variable causes the other.

Three alternative causal stories are consistent with the same data:

1. Diversity causes better financial performance (McKinsey's implied story).
2. Financial outperformance causes more diversity (profitable companies can afford more expensive DEI programs, have resources to recruit broadly, and attract a wider applicant pool because of their brand).
3. A third variable causes both (companies in high-growth industries are both more profitable and more likely to be in urban centers where diverse talent concentrates; the correlation is spurious).

McKinsey's reports do not test these alternatives. They identify the correlation and present it, with the causal arrow implied by the framing.

The confounder problem

The McKinsey methodology compares the top diversity quartile against the bottom. High-diversity executive teams and low-diversity executive teams differ on many dimensions beyond diversity:

• Industry: Technology and financial services companies are both more profitable on average and more geographically concentrated in areas where diverse executive talent is available. Oil, gas, manufacturing, and agriculture companies are less so.
• Geography: Companies based in global cities (New York, London, San Francisco) are both more diverse (diverse talent concentrates there) and more profitable (higher-value industries concentrate there).
• Company maturity: Newer companies are more likely to have built diverse executive teams (norms have shifted) and are also more likely to be in growth industries.
• CEO characteristics: CEOs who prioritize diversity tend to cluster in certain industry types, and those same industry types have specific performance profiles.

The 2021 Baert et al. replication study tested whether the McKinsey finding held after controlling for country and industry. It did not. The correlation between diversity quartile and financial performance was not strong to those controls. McKinsey has not publicly addressed this finding.

The conflict of interest

McKinsey is a consulting firm. It generates significant revenue from DEI advisory engagements: workforce diversity audits, inclusion strategy consulting, DEI program design, and leadership development. The Diversity Wins series is free-to-download research that positions McKinsey as an authority on the business case for DEI.

Management consulting research that validates the services the publisher sells is a known category of problem. This doesn't mean the research is wrong -- it means the research has a funding and incentive structure that creates pressure toward positive findings, and that consumers of the research should weight the findings accordingly relative to independent peer-reviewed work.

Does diverse leadership actually cause higher financial performance?

The critique of McKinsey's methodology is not an argument that diversity has no organizational benefits. Scott Page's The Diversity Bonus (2017) provides a peer-reviewed theoretical and empirical argument for why diverse teams outperform homogeneous teams on specific complex problem-solving tasks. Page's mechanism: diverse cognitive profiles (shaped by different backgrounds, experiences, and training) apply different models to the same problem, and the interaction of different models tends to identify solutions no single model finds.

Page's framework is narrower than McKinsey's: it applies most strongly to complex, nonroutine tasks with multiple valid solutions (like strategic planning, product design, or scientific research). It applies less to routine operational tasks. And it's about cognitive diversity, which correlates with demographic diversity but is not the same thing.

The distinction matters: the evidence for diversity's benefit is strongest for specific task types, with a specific mechanism. The McKinsey reports imply a broader, economy-wide effect that is methodologically unsupported.

The WokeCorp assessment

The research. McKinsey's Diversity Wins series is correlation research with significant uncontrolled confounders, a conflict of interest in the publisher, and a replication failure on key controls. Citing it as proof that diversity programs cause financial outperformance overstates what the data shows.

The underlying question. Whether diversity produces organizational benefits is a legitimate empirical question with a real body of evidence. The answer from better-controlled research is: sometimes, for specific task types, with a specific mechanism. That is a genuine finding. It just isn't the "companies 36% more likely to outperform" headline.

The policy implication. DEI programs can be justified on legal compliance, fairness, and talent-pipeline grounds without requiring a 36%-outperformance claim. The business case doesn't need to be overblown to be real. Citing flawed research as the foundation of DEI policy creates fragility: when the research is criticized, the policy looks like it has no foundation, even if there are better arguments available.

Related reading

• The Great DEI Retreat: Corporate Rollbacks in 2024-2025 -- the rollback wave that the "business case" framing was supposed to prevent
• Accenture's DEI Rollback: The Consulting Firm Reverses Course -- consulting firms that cited this research while rolling back the programs
• Google's DEI Rollback: Aspirational Hiring Targets Dropped -- the tech company that published its own diversity data for ten years

Sources

• McKinsey, "Why Diversity Matters," January 2015. Verified June 2026.
• McKinsey, "Delivering Through Diversity," January 2018. Verified June 2026.
• McKinsey, "Diversity Wins: How Inclusion Matters," May 2020. Verified June 2026.
• Baert, Stijn et al., "Diversity Wins? A replication study of the McKinsey diversity reports," SSRN Working Paper 3910410, 2021. Verified June 2026.