Equal Pay Day: The US Gender Pay Gap in 2024

Published for Equal Pay Day in March 2025, this Economic Policy Institute article examines United States wage data for 2024. Katherine deCourcy and Elise Gould report an average hourly gender gap of 18.0% after controlling for race and ethnicity, education, age, marital status and state. The date of the article and the year of its data answer different questions: this is a 2025 analysis of 2024 wages.

What does a pay-gap measure compare?

A comparison of hourly wages is different from annual earnings, which also reflect hours and periods outside paid employment. An adjusted estimate is different again from an unadjusted comparison or a case of unequal pay for the same job. These distinctions determine what a number can support. None should be silently substituted for another, or carried from one country into another.

EPI examines the distribution of wages, differences by educational attainment, and racial and ethnic disparities. This makes it possible to ask whether a headline improvement reaches workers in different positions. An aggregate gap can shrink while substantial inequalities remain within the labour market. Likewise, greater educational attainment does not establish that the institutions allocating pay have become fair.

What the average leaves open

For feminist analysis, the useful question is how a statistic connects to the organization of work. Occupational segregation, bargaining power, care responsibilities and the valuation of different jobs are possible routes for further inquiry. A statistical control does not make the social process behind a variable disappear. Education, for example, is itself shaped by access to resources and opportunity.

The authors connect their findings to pay transparency, enforcement and collective bargaining. These proposals operate at different points: information can expose a disparity, enforcement can create consequences, and collective organization can affect workers’ ability to negotiate. Evaluating a reform requires asking which point it reaches and who can actually use it.

The article is valuable as a clearly bounded US data resource. Readers comparing it with another study should first check the population, time period, wage definition and reference group. That discipline makes international comparison more informative and prevents a memorable percentage from becoming a claim about every woman’s experience.