The Curve

A natural pipeline produces an S‑shape over time. The dataset tracked here does not look like that.

What was expected

When a class of candidates that was previously absent from a category of senior positions begins to enter that category, the standard sociological expectation is an S‑curve. The rate of firsts begins slowly as a small number of pioneers establish that the role is possible to occupy. It accelerates as cohorts of candidates with relevant qualifications mature through middle-career positions and become available for promotion. It then plateaus as the pool of candidates and the pool of available positions reach equilibrium.

For first-woman appointments to senior Western institutional roles, the expectation under this model would have been: scattered firsts from roughly 1960 onward, accelerating through the 1990s and 2000s, with the rate visibly tapering by the 2020s as the most senior categories filled.

What is observed

Appointments per decade across all 17 domains. The 1960s through the 1980s are not the rising middle of an S‑curve. They are a flat baseline of five to ten events per decade across all senior institutional categories globally. The 2010s and 2020s are not the plateau. They jump by roughly fivefold over the 1980s.

The flat baseline runs from 1960 through the late 1980s — five to ten first-woman events per decade across all senior categories worldwide combined. The 1990s and 2000s rise to roughly twenty events per decade, distributed across continents and categories. The 2010s register a step change to forty-plus events. The 2020s, only partially elapsed at the time of writing, are on a trajectory comparable to or exceeding the 2010s.

The shape is a J, not an S. That is the central observation this site exists to document.

A geographic asymmetry in the baseline

At the level of national heads of state and heads of government, the dataset's pre-1990 record is predominantly non-Western. The world's first elected woman head of government was Sirimavo Bandaranaike in Sri Lanka in 1960, followed by Indira Gandhi in India (1966), Golda Meir in Israel (1969), and the early entries in Argentina, the Central African Republic, and Bolivia in the 1970s. Margaret Thatcher (1979) is the dataset's earliest Western European national-level entry, nearly two decades after Bandaranaike.

The county-sheriff pre-1990 record runs in parallel, but distinct: at the US local-government level, the dataset's earliest competitive first-woman events are Kathy Crumbley (Belmont County, Ohio, Nov 1976) and Ruth Carpenter (Hunterdon County, New Jersey, 1977), with Virginia Donnelly (Howard County, Maryland, 1982) following. These three are the only pre-1990 first-woman events in the dataset at the US sub-national level under the law-enforcement domain's competitive-election rule. They do not change the national-level geographic asymmetry; they sit beside it as a separate institutional-tier record.

The post-2018 cluster, in contrast, is concentrated in Western institutions: the European Union's senior offices, the United States executive branch and intelligence services, Western European member-state governments, major US art institutions, US universities, and US Fortune 500 firms. Latin American and African post-2010 first-women heads of state are real and documented but are individually distributed rather than clustered. The contrast — non-Western pre-1990 national record, Western post-2018 cluster — is one of the dataset's principal structural features.

The pipeline pre-cluster: universities

One layer of the dataset's record is upstream of the rest: first-woman presidencies of major research universities. The candidate pool for senior intelligence, judicial, financial, diplomatic, and corporate roles is drawn substantially from the faculty and graduate alumni of these institutions. The dataset records: Hanna Holborn Gray at the University of Chicago (1978), Judith Rodin at the University of Pennsylvania (1994 — first permanent woman president of an Ivy League institution), Tilghman at Princeton (2001), Simmons at Brown (2001), Hockfield at MIT (2004), Faust at Harvard (2007), Richardson at Oxford (2016), Prentice at Cambridge (2023).

The Ivy and equivalent firsts cluster in 1994–2007 — thirteen years — about a decade ahead of the broader institutional cluster. If the candidate pool for senior institutional roles were being substantially shaped by upstream university leadership and faculty composition, the timing offset between the academic cluster and the broader cluster is approximately what one would expect. This is the dataset's pipeline-institution sub-thesis. It is consistent with the data; it does not adjudicate between alternative interpretations of the same evidence.

The 2018–2026 cluster

Within the broader inflection, a tighter cluster is visible from approximately 2018 onward. The cluster spans:

The European Central Bank (Christine Lagarde, 2019).
The European Commission (Ursula von der Leyen, 2019).
The European Parliament (Roberta Metsola, 2022).
The EU’s foreign-policy chief role (Kaja Kallas, 2024).
The first woman as Italy’s prime minister (Giorgia Meloni, 2022).
The United States Treasury (Janet Yellen, 2021).
The United States Centers for Disease Control (Rochelle Walensky, 2021; Mandy Cohen, 2023).
The United States National Institutes of Health (Monica Bertagnolli, 2023).
The World Trade Organization (Ngozi Okonjo-Iweala, 2021).
The United States Coast Guard, the first uniformed service branch (Linda Fagan, 2022).
The Louvre (Laurence des Cars, 2021), the National Gallery of Art in Washington (Kaywin Feldman, 2019), and other major museums.

A number of those organisations had no female senior leadership in their preceding history of fifty to two hundred years. Several registered firsts within the same year as one another.

The cluster also extends into the Fortune 500 corporate leadership domain. A first wave (Rometty IBM 2012, Mayer Yahoo 2012, Hewson Lockheed 2013, Novakovic General Dynamics 2013, Barra GM 2014, Su AMD 2014, Wojcicki YouTube 2014) preceded the principal cluster by approximately five years. A second wave (Tómé UPS 2020, Lynch CVS 2021, Fraser Citigroup 2021) is contemporaneous with it. The first wave preceded the documented institutional-investor diversity policies; the second is contemporaneous with them.

What the shape does and does not establish

A J‑curve is consistent with several distinct underlying causes, and the curve alone does not adjudicate between them.

It is consistent with a long-deferred pipeline finally arriving. It is also consistent with the introduction of policy mechanisms in a narrow time window that affect promotion outcomes simultaneously across institutions. The two are not exclusive: a maturing pipeline could be amplified by simultaneous policy mechanisms, producing a steeper rise than either would produce alone.

What the shape rules out is the strong form of the natural-pipeline thesis: that the rate observed since 2018 is the unaltered output of a thirty-year demographic trend. The flat baseline contradicts that directly. Whatever else is going on, something different began to happen in the late 2010s.

What follows from this

The reasonable next questions are empirical. They are: which specific mechanisms changed in the late 2010s; whose promotion-eligibility pools they affected; and whether the cluster is concentrated in institutions that share governance features (institutional-investor pressure, EU policy frameworks, federal procurement rules) or is diffused across institution types.

The documented mechanisms page lists the policies and rules with public records that bear on the first question. The counterexamples page lists the institutions where the trend is absent and helps constrain the second.

Caveats

The dataset is currently weighted toward Western institutions and toward institutions with strong public sourcing. Coverage gaps could in principle obscure or exaggerate the curve.
Some of the late-cluster appointments remain in office at the time of writing. The full duration and consequence of these appointments is not yet observable.
The choice of “senior” positions is judgment-bound. The methodology page describes the inclusion criteria.

The dataset will continue to grow. The shape may shift as additional historical firsts are added — in particular, the inclusion of additional non-Western institutions could change the baseline. The site will document those changes openly.