US Pay Gap Calculator

Median, not average: why the midpoint matters

Every serious pay-gap number starts with a choice: measure the typical worker or measure all dollars. The gap you see quoted in headlines is built on the median, the worker standing exactly halfway through the earnings lineup, half earning more and half earning less. The median resists distortion, because a handful of nine-figure chief executives cannot move it no matter how far above the line they sit.

The average behaves differently. One outsized salary drags the whole mean upward, and since the top of the earnings distribution is disproportionately male, an average-based gap comes out wider and means something different: it partly measures concentration at the top, not the experience of the typical worker. Neither number is wrong, but they answer different questions, and mixing them up is the fastest way to argue past someone.

This page sticks to medians from the Bureau of Labor Statistics, the standard source for US earnings data, and shows the arithmetic behind every figure so you can reproduce it with any two numbers you choose.

The 2024 numbers: $1,043 against $1,261 a week

The BLS annual report on women's earnings in 2024 gives the baseline. Among full-time wage and salary workers, women's median usual weekly earnings were $1,043 and men's were $1,261, a women's-to-men's ratio of 83%. That ratio was 62% in 1979, the first year of comparable data, so the long arc is real even though recent progress has been slow.

Converting weekly medians to a year is one multiplication:

• Women: 1,043 x 52 = $54,236 a year
• Men: 1,261 x 52 = $65,572 a year
• Gap: 65,572 - 54,236 = $11,336 a year, or about $218 a week

Two cautions keep the number honest. First, it covers full-time workers, usually 35 hours or more a week, so it does not capture the part-time story at all. Second, it compares all industries and ages together, so it is a snapshot of the whole workforce rather than a like-for-like comparison of two people in the same chair. The table below starts to pull those threads apart.

Worked example: comparing a $58,000 salary

Individual comparisons are where the medians become useful. Say you earn $58,000 a year and want to know where you stand against the 2024 medians.

• Against the women's median: 58,000 / 54,236 = 1.069, so about 106.9% of it, a positive gap of 58,000 - 54,236 = $3,764
• Against the men's median: 58,000 / 65,572 = 0.885, so about 88.5% of it, a shortfall of 65,572 - 58,000 = $7,572
• Weekly equivalent: 58,000 / 52 = about $1,115 a week

The same salary sits above one median and below the other, which is the entire point: "above average pay" depends on which line you are measured against. Run the calculator with your own number and both comparisons appear together, in dollars rather than percentages, because dollars are what show up in a lease application.

The gap by race and ethnicity

The headline 83% averages across groups that fare differently. The BLS quarterly earnings release for the third quarter of 2024 reports the ratio separately for each group, always comparing women with men of the same group.

The pattern looks paradoxical until you read the columns separately. Black women earn close to 95% of Black men's median, a tight ratio inside a group whose overall pay levels are lower. Asian women earn the highest weekly figure in the table yet the widest gap to their male counterparts. A ratio and a level measure different things, and equal-opportunity conversations need both, which is why demographers at the Census Bureau's income program publish levels and ratios side by side in their annual series.

The gap by age and career stage

The gap is not one number over a career, it starts small and compounds. Late 2024 BLS quarterly data show the ratio moving sharply by age band, and the shape of that curve tells the story better than the average does.

Early careers start nearly level, which is what you would expect when pay scales are standardized and tenure is short. The spread then widens through the prime career decades, the years when caregiving interruptions, occupational sorting and seniority gaps stack up, and it is widest for workers 55 and over. The same dollars compound: a gap that is roughly $113 a week at 24 is $280 a week at 60, which is also why later-career retirement balances show the gap even more starkly than paychecks do.

What drives the gap

The raw 83% ratio is not a measure of unequal pay for identical work. Most of it reflects three measurable forces. Occupational sorting is the biggest: women remain overrepresented in lower-paying fields and underrepresented in the highest-paying ones, and BLS's occupational tables show wide spread between, say, health support roles and software development. Hours within full-time matter next, since a 50 hour week in a salaried job is still one full-time job, but pays materially more than a 38 hour one.

Career interruptions form the third force, and they act like a tax on seniority: time out of the workforce slows raises and promotions in ways that persist after return. Controlled studies that match workers on industry, occupation, hours and experience shrink the gap substantially but rarely to zero, and the unexplained remainder is where discrimination questions live. economists disagree about its size, not its existence.

Knowing the drivers changes what the calculator tells you. A gap driven by sorting suggests questions about which field and which level; a gap that survives matching inside one workplace suggests questions about the pay-setting process itself, and several states now require employers to disclose pay scales on request, which turns that conversation from vibes into numbers.

The gap in dollars at different salaries

Percentages hide what people buy and rent with. Applying the 2024 national ratio of 83% to round salaries shows the annual cost of the gap as pay rises, which is why senior roles have the most money on the table.

These rows are illustrative math at the national ratio, not measured gaps for those jobs; real gaps vary enormously by occupation, state and employer. But the scaling is the takeaway: the same percentage gap costs a $40,000 worker $6,800 a year and a $150,000 worker $25,500, which compounds into retirement accounts, mortgages and negotiation room. Roughly 28 months of the national median rent sits inside the ten-year version of the $100,000 row.

How to use these numbers well

Three habits make pay-gap statistics useful instead of inflammatory. First, always state the population: full-time or all workers, all ages or a band, weekly or annual. The 83% figure is full-time weekly medians for 2024; quoting it as something else invites a fight over a claim nobody made. Second, compare like with like when the question is fairness inside a workplace: same role, same level, same geography. The national ratio was never designed to settle whether two analysts are paid equally.

Third, use the dollars for negotiation. The BLS medians give you an anchor for what typical full-time pay looks like nationally, and state-level Census data narrow it further. Pair that with job-posting ranges, which pay-transparency states now require, and the question "what should this role pay" becomes a bounded estimate instead of a guess. The Equal Employment Opportunity Commission's guidance covers the legal side, where equal pay for substantially equal work has been federal law since the Equal Pay Act of 1963.

One last framing note: the ratio has moved from 62% in 1979 to 83% in 2024, roughly 21 points in 45 years. At that pace, full convergence is not a near-term event, which is precisely why the numbers, their sources and their definitions are worth knowing precisely.

Figures are BLS medians for full-time wage and salary workers and are for orientation, not legal conclusions about any specific workplace.