Pay Gap Statistics (2026)

Median usual weekly earnings for the 120.9 million US full-time wage and salary workers were $1,251 in the second quarter of 2026.

That was 4.6% higher than a year earlier, against a 3.9% rise in the Consumer Price Index for All Urban Consumers over the same period, so real earnings rose slightly.

On a seasonally adjusted basis the figure was $1,258, up from $1,233 in the previous quarter.

Women's median weekly earnings were $1,131, or 82.0% of the $1,380 median for men.

The ratio varies considerably by race and ethnicity: 91.0% for Black women against Black men, 83.5% for Hispanic, 81.6% for White and 78.3% for Asian.

A high within-group ratio does not mean high pay. Black women's median of $991 is the second lowest in the table even though their ratio to Black men is the highest, because Black men's median of $1,089 is itself well below the $1,400 median for White men.

Women aged 16 to 24 earned 91.1% of what men the same age earned. For workers aged 55 and over, the ratio was 78.5%.

Both sexes peak in the same band. Men's median was $1,596 at 35 to 44 and $1,571 at 45 to 54; women's was $1,249 and $1,231 in the same bands.

The widening pattern is consistent with a gap that opens over a career rather than at hiring, which is what makes the age breakdown more informative than the headline number.

Full-time workers in management, professional and related occupations had the highest median weekly earnings, $1,928 for men and $1,476 for women.

Service occupations had the lowest, $920 for men and $730 for women.

The gap between those two occupational groups is more than $1,000 a week for men, which is far larger than the $249 gap between men and women overall.

Among full-time workers aged 25 and over, median weekly earnings were $803 for those without a high school diploma, $994 for high school graduates with no college, and $1,768 for those with at least a bachelor's degree.

A degree is therefore associated with $774 more a week than a high school diploma, roughly $40,000 a year before tax.

That is by a wide margin the largest single differentiator in this release, larger than sex, race or occupation group taken individually.

Among college graduates with advanced degrees, the highest-earning 10% of men made $5,363 or more per week and the equivalent women $3,676 or more.

In annualised terms that is roughly $279,000 against $191,000, a difference of about $88,000 at the same educational level.

Medians compress this. Looking only at the median understates how much of the aggregate gap sits in the upper tail of high-earning professional work.

The 82.0% figure compares two medians. It controls for nothing: not occupation, not hours beyond the full-time threshold, not experience, not industry, not seniority.

So it measures what men and women are actually paid, which is a real and useful fact, and it does not measure whether two people doing the same job are paid differently.

Studies that control for those factors produce a smaller gap. Both numbers are correct and they answer different questions, and quoting one as though it were the other is the standard misuse of this series.

A gap that is narrow at entry and wide by mid-career is a different problem from one that is constant across ages.

The first points at what happens during a career: promotion timing, hours, career breaks and the occupations people move into. The second would point at starting pay.

The data here shows the first shape, which is why interventions aimed only at starting salaries would address a smaller share of the gap than they appear to.

A $249 weekly difference is about $12,950 a year, which is the number most people stop at.

The larger effect is on saving. A retirement contribution is usually a percentage of pay, and an employer match is a percentage of that, so a pay gap becomes a contribution gap and then a compounding gap.

Over a thirty-year career the difference in accumulated retirement assets is a multiple of the difference in salary, which is how a pay gap turns into a wealth gap.

The single largest lever in this dataset is not negotiation, it is occupation and education, which move pay by more than a thousand dollars a week between groups.

The second largest, for anyone already in a job, is the contribution rate rather than the salary, because it is the part of compensation an individual fully controls.

Capturing an employer match in full is the specific version of that: it is the only guaranteed return available in a pay package, and it is forfeited quietly rather than refused.

All figures are median usual weekly earnings for full-time wage and salary workers, not annual pay and not all workers. Part-time workers and the self-employed are excluded.

Data comes from the Current Population Survey, in which respondents are asked how much each wage and salary worker usually earns, so it is self-reported.

Figures are not seasonally adjusted unless stated, which is why the release quotes both $1,251 unadjusted and $1,258 seasonally adjusted for the same quarter.

Every figure is from the BLS news release Usual Weekly Earnings of Wage and Salary Workers, Second Quarter 2026, published 21 July 2026 as USDL-26-1257.

Earnings by sex, race and ethnicity are table 2, by age table 3, by occupation table 4, and by education and decile table 5.

The inflation comparison uses the CPI-U over the same period, as quoted in the release itself.