Two new pay disclosure regimes have arrived in the past twelve weeks, alongside a UK regime nine years old, and not one of the three measures the gap inside the executive committee. That is precisely where an 8% gap survives every control the research can apply.
Two new pay disclosure regimes have arrived in the past twelve weeks, alongside a UK regime nine years old. If you sit on a board, you are about to have more remuneration data in front of you than you have ever had, and none of it will tell you what your executive committee pays men and women for comparable work.
The United Kingdom has required employers with 250 or more employees to publish a whole-workforce gender pay gap since the 2017 regulations. The European Union’s Pay Transparency Directive passed its transposition deadline on 7 June 2026, and its reporting obligations are whole-workforce as well, by category of worker. South Africa’s new remuneration report, which commenced on 22 May 2026, requires the ratio between the top 5% of earners and the bottom 5%, with no gender dimension at all.
Each of the three is a serious instrument and each was hard won. Not one looks inside the executive committee. That would be a technicality if the top of an organisation were where the problem thinned out. It isn’t, and the study that shows this most cleanly is the one most often cited to argue the opposite.
Every disclosure regime now in force measures the workforce, and none measures the executive committee
Start with what is actually in force, because much of what has been written in the past two years describes a state of affairs that has not arrived. Directive (EU) 2023/970 entered into force on 6 June 2023, and member states had until 7 June 2026 to transpose it. That deadline has passed, and only a small minority, fewer than a quarter, had completed transposition by it. The European Commission confirmed in December 2025 that the date would not move. No reporting falls due until 7 June 2027 for employers of 250 or more, and not until 7 June 2031 for employers of 100 to 149. The directive has produced no data. It cannot yet have moved anything. When it does bite, Article 9 reports at employer level across the whole workforce by category of worker.
South Africa’s reform is the one most likely to be misdescribed, because it arrived in the same conversation and does something different. Sections 30A and 30B of the Companies Act commenced by proclamation on 22 May 2026 with no transitional period, which Cliffe Dekker Hofmeyr noted caught companies mid-cycle. Section 30A puts the remuneration policy of public and state-owned companies to a shareholder vote. Section 30B requires an implementation report disclosing the pay of each director and prescribed officer, the highest and lowest paid employee, and the ratio between the top 5% of earners and the bottom 5%. Reject that report at two consecutive annual general meetings and the non-executive members of the remuneration committee face a two-year bar. Read the section closely and the gender dimension is simply absent: this is vertical dispersion and shareholder accountability, a different question from the one this article is about.
So the position a board is in this year is specific rather than general. You will soon be able to say what your organisation pays at the bottom relative to the top, and in the UK, eventually the EU, what it pays women relative to men across the whole headcount. If a shareholder asks what the gap is between the men and the women around your executive committee table, you’ll be answering from your own records or not at all.
About 70% of the executive gap is which job you hold, and an 8% residual gap is not
The best like-for-like evidence comes from Wolfgang Keller, Teresa Molina and William Olney, in NBER working paper 28216. It matches ExecuComp and Compustat data derived from filings to the US Securities and Exchange Commission, covering the top five highest-paid executives at roughly 3,500 firms, close to 240,000 observations, from 1992 to 2017. Women are 6% of that sample.
Unconditionally, women in it earn 26% less than men. That falls to 8% once executive characteristics, firm characteristics and, in particular, job title are accounted for. In the authors’ own words, female executives with similar experience and education, at similar firms, doing similar jobs, earn less than their male colleagues.
Both halves deserve equal weight and they rarely get it. Roughly 70% of the raw executive gap is explained by composition, principally by which title a person holds. If you have argued that the headline number overstates unequal pay for the same work, the better part of this finding is behind you, and that is worth conceding plainly. The rest of it is that a conditional gap of about 8% survived controls for title, firm, experience and education. It did not disappear.
Two disciplines travel with the number. It is a working paper rather than a peer-reviewed journal article, and the data ends in 2017, which makes it nine years old as you read this. The 8% may since have narrowed, widened or held, and no equivalent study using later data could be located. That absence is not an oversight by the research community: no regime in force anywhere requires executive-level gender pay data, so the raw material for a current study largely does not exist.
The same paper closes off the explanation that surfaces first in most boardrooms. Keller, Molina and Olney find female executives disproportionately represented at firms with more temporal flexibility and what the study classifies as female-friendly cultures, and state that this does not explain the pay gap. Their related finding is the more interesting one: the within-firm gap is significantly smaller at those firms, which locates the variation in employer behaviour rather than in the executives themselves. What actually produces the residual is unsettled, with performance-pay allocation, within-firm culture and differing executive entry and exit rates all present in the literature and none established. Any account that picks one has gone past the evidence.
The composition problem is not a shortage of women in senior roles, it is which senior roles
If composition does most of the work, the useful question is what it looks like inside an executive committee. Altrata’s 2025 analysis of the S&P 500 gives the sharpest answer available: women hold 72% of chief human resources officer roles and 59% of chief marketing officer roles, against approximately 19% of chief financial officer roles and 10 to 12% of chief executive, chief operating officer and chair roles.
That is not a pipeline running dry. It is a pipeline running into two functions, and they are the functions that pay least at executive level and lead to the chief executive’s office least often. “We need more women in senior roles” describes something the S&P 500 has substantially achieved in human resources and marketing. “We have concentrated them in the seats from which chief executives are seldom appointed” describes something a nomination committee can act on this year, in its succession slate rather than in its diversity report.
Attribute this one carefully. It is Altrata’s analysis, single-source, and its 2025 board figure of 33.6 to 33.7% sits below Spencer Stuart’s 35%, drawn from a different population: 488 S&P 500 companies, from proxy statements filed between 1 May 2024 and 30 April 2025. Different measurements, not to be averaged.
Disclosure narrows the gap by slowing men’s pay, which matters most to people paid by negotiation
Disclosure works. The mechanism is not the one most boards assume, and it lands hardest on the population this article is about. Denmark required gender-disaggregated wage statistics from firms with 35 or more employees from 2007. Bennedsen, Simintzi, Tsoutsoura and Wolfenzon studied the change with Danish matched employer-employee administrative data and a difference-in-differences design, publishing in the Journal of Finance in 2022. The gap fell by about 1.9 percentage points, and it fell because male wage growth slowed by 1.7 percentage points while female wages were essentially unchanged. Firm productivity fell 2.7%.
Blundell, Duchini, Simion and Turrell examined UK gender pay gap reporting in the American Economic Journal: Economic Policy in 2025, exploiting variation in the mandate across firm size and time. Pay transparency closes 19% of the gap, measured across the whole workforce at UK employers of 250 or more, again by reducing men’s pay growth. Cullen and Pakzad-Hurson supply the theory in Econometrica in 2023: studying US state laws protecting the right to ask about colleagues’ salaries, they find wages declined by about 2% overall, because under transparency an employer will credibly refuse to pay a high wage to any one worker when doing so triggers costly renegotiation with everyone else. That paper measures wage levels rather than the gender gap and should not be read as a finding about the gap.
Three countries, three identification strategies, one mechanism. Disclosure narrows measured gaps by compressing pay at the top of the distribution rather than by lifting the bottom. Now consider who in your organisation is paid almost entirely through individually negotiated premia, and you have the implication that rarely reaches the board pack. The instrument works by suppressing precisely the kind of pay-setting that executive remuneration consists of. A remuneration committee treating a transparency regime as an equity project with no consequence for its own negotiating position has not read the evidence.
The strongest arguments against this reading, and what survives them
Three pieces of evidence cut against the argument above. A director will already half-know all three.
The first is the most direct. The Conference Board’s 2025 executive compensation release, produced with FW Cook and ESGAUGE, found female S&P 500 chief executives on median compensation of $16.5 million against $15.6 million for men, and female chief financial officers on $5.8 million against $5.35 million. A reversal, at exactly the level this article is about. It is also unadjusted, a median across a very small female group, with women holding 7.9% of S&P 500 chief executive positions, and it controls for neither company size nor sector nor tenure: very large companies pay more, and a handful of women lead very large companies. The Conference Board’s Matteo Tonello offers scarcity pricing as the explanation, that companies are likely offering more competitive packages to attract and retain female leadership, and offers it as a hypothesis rather than a finding. If he is right, the binding constraint at the very top is access to the role rather than the pay attached to it, which strengthens the composition half of this article and weakens the pay half.
The second is a problem for anyone treating board diversity as the operative lever. FTSE 350 boards are 42.7% women, per the FTSE Women Leaders Review 2026 report, with 35.9% of leadership roles held by women. FTSE 350 chief executives are 8.2% women. Boards have reached near-parity and the executive line has not moved with them. Galanakis and Gosling put a number on the ceiling in the Journal of Population Economics in 2026, using linked administrative data for 8,411 UK firms with 250 or more employees between 2017 and 2021: a one percentage point rise in the female director share reduces the reported gap by 0.043 percentage points, so full board parity would close about one sixth of the reported whole-workforce gap of 9.7%, which is not the same measurement as the executive gap this article is about. Board composition is a lever. It is not the lever, and anyone told otherwise has been oversold. The same paper locates a mechanism worth borrowing: female directors are associated with more equitable allocation of performance-related pay, and discretionary pay is where a same-role gap has room to live.
The third is broader. Equileap’s 2026 assessment of 3,430 public companies across 24 developed markets found the share publishing a gender pay gap rose from 15% in 2021 to 48% in 2026, while only 1.66% had closed theirs. Disclosure has spread roughly fivefold. Closure has not followed it.
What survives is narrower than the argument this article opened with, and more defensible for the narrowing. Not that the executive gap is large and growing. That it is the one gap nobody is required to measure, that the best evidence available finds a residual after every control it can apply, that the evidence is nine years old, and that the instruments now arriving will not refresh it.
The Executive Committee Pay Test: four things a board can measure that no regime requires it to publish
Four things a board can measure about its own executive committee, none of which any regime requires it to publish. Deliberately few enough to hold in your head during a remuneration committee meeting.
- Composition. Which functions do the women on your executive committee hold, and are they the functions from which your last three chief executives were appointed?
- Conditional gap. What is the pay difference between comparable roles once title, tenure and business scale are held constant? Not the whole-workforce number. The one inside the room.
- Discretion. How is performance-related and discretionary pay allocated across the committee, and would the pattern hold up if it were read aloud?
- Flow. What are appointment and exit rates by gender at this level over five years, rather than the standing headcount on the day of the report?
The first is a nomination committee question, the second and third belong to remuneration, and the fourth usually belongs to nobody, which is why it goes unanswered. A board that can answer all four is ahead of every regime discussed here, and the answers are already in its own records.
What this means for a board, and for anyone whose own pay is the subject
Five questions worth being able to answer, whichever side of the table you sit on.
- Do you know your own executive committee’s gender pay gap, conditional on role, and if not, what is stopping you from calculating it before someone asks?
- Which regime will first require your organisation to publish something, on what date, and is anyone accountable for that date?
- When your remuneration committee allocates discretionary and performance-related pay, does it review the pattern across the committee, or only each case on its own?
- If transparency compresses individually negotiated premia, as three independent studies suggest, what does that mean for how your organisation attracts and holds executive talent?
- And if your own pay is the subject rather than the item on the agenda: can you evidence the scale of what you have run, in the terms a remuneration committee uses, before the conversation reaches a benchmark?
That last question is where an individual has the most influence and usually the least preparation. The composition finding says the roles you have held shape your pay more than the negotiation does, which puts the weight on the trajectory rather than the conversation at the end of it. Our guidance on the route into a C-level role and the portfolio career at executive level covers the trajectory itself, and career transition for women in senior roles is the practitioner companion to this analysis. On how that record is then read and tested, see how executives are found and approached, why executive visibility is a governance asset and what happens when a claimed record does not hold up.
Where the work is making the scope of what you have run legible before anyone reaches for a benchmark, that is the substance of our Executive Positioning Strategy engagement. The executive services and engagement options pages set out how the firm works, and confidential enquiries go to [email protected].
Sources and further reading
- Directive (EU) 2023/970 on pay transparency, adopted 10 May 2023, in force 6 June 2023. EUR-Lex
- L&E Global, EU pay transparency transposition tracker, 28 July 2026. L&E Global
- Ogletree Deakins on the European Commission’s confirmation of the 7 June 2026 deadline, December 2025. Ogletree Deakins
- Companies Amendment Act 16 of 2024 (South Africa), sections 30A and 30B, commenced 22 May 2026. gov.za
- Cliffe Dekker Hofmeyr, corporate and commercial alert on the new remuneration provisions, 10 June 2026. Cliffe Dekker Hofmeyr
- Baker McKenzie, South Africa: New Companies Act Pay Rules, June 2026. Baker McKenzie
- GOV.UK, gender pay gap reporting guidance for employers, current at 13 August 2026. GOV.UK
- Keller, Molina and Olney, “The Gender Gap Among Top Business Executives”, NBER Working Paper 28216, December 2020, revised December 2022. NBER
- Bennedsen, Simintzi, Tsoutsoura and Wolfenzon, “Do Firms Respond to Gender Pay Gap Transparency?”, Journal of Finance, 2022. NBER working paper 25435
- Blundell, Duchini, Simion and Turrell, “Pay Transparency and Gender Equality”, American Economic Journal: Economic Policy 17(2), 2025. American Economic Association
- Cullen and Pakzad-Hurson, “Equilibrium Effects of Pay Transparency”, Econometrica 91(3), 2023. NBER
- Galanakis and Gosling, “Mind the (gender pay) gap: the role of board gender composition”, Journal of Population Economics 39(16), 2026. Springer
- Altrata, Gender Diversity in Corporate America 2025, published 21 August 2025, data through Q2 2025. Altrata
- Spencer Stuart, 2025 US Board Index, October 2025, proxy statements filed 1 May 2024 to 30 April 2025. Spencer Stuart
- FTSE Women Leaders Review, 2026 report, February 2026. FTSE Women Leaders Review
- Equileap, The Gender Pay Gap in 2026, 23 April 2026, 3,430 companies across 24 developed markets. Equileap
- The Conference Board with FW Cook and ESGAUGE, Executive Compensation Release 2025, reported via ESG Dive