Racial and ethnic earnings gaps in BLS data, and how they have changed

Want the official numbers on how pay differs by race and Hispanic ethnicity in the US, and whether the gaps are shrinking? This guide walks through the newest Bureau of Labor Statistics (BLS) figures, how the gaps have moved over about a decade, what changes when you compare people with the same education, and how to use it all when you check your own pay.

One quick definition before the numbers. Every figure here is a median: line up all full-time workers in a group from lowest paid to highest paid, and the median is the person in the middle. Half earn more, half earn less. Unlike an average, a handful of very high earners can't drag it up.

The latest figures: median weekly earnings by race and ethnicity

BLS publishes these numbers every quarter in its Usual Weekly Earnings release, using the Current Population Survey (CPS). The newest release, published October 21, 2026, covers the third quarter of 2026. Figures are for full-time wage and salary workers age 16 and over and are not seasonally adjusted.

Source: BLS, Usual Weekly Earnings, third quarter 2026. The percentage column is a Simply Salaries calculation from the BLS medians.

Put simply, in Q3 2026 the typical full-time Black worker earned about 80 cents for every dollar the typical full-time White worker earned, and the typical Hispanic worker earned about 78 cents. The typical Asian worker earned well above the White median. In dollars, the weekly gap between the White and Black medians was $249, and between the White and Hispanic medians it was $280.

The gaps look different for men and women

BLS also splits each group by sex, and it calculates some of the ratios itself. In Q3 2026:

• Black men's median ($1,040) was 74.6% of White men's ($1,394).
• Hispanic men's median ($1,028) was 73.7% of White men's.
• Black women's median ($994) was 87.4% of White women's ($1,137).
• Hispanic women's median ($919) was 80.8% of White women's.
• Asian men ($1,841) and Asian women ($1,506) had higher medians than White men and women.

So the race gap is noticeably wider among men than among women. (The gap between women and men has its own coverage on our pay trends page.)

How BLS defines the groups

The categories matter more than most people realize:

• Race is reported by the household respondent to the survey, using the five federal (OMB) race categories.
• Only three race groups appear in the quarterly release: White, Asian, and Black or African American. Native Hawaiian or Other Pacific Islander workers, American Indian or Alaska Native workers, and people reporting more than one race are left out because the samples are too small for reliable estimates.
• Hispanic or Latino is an ethnicity, not a race. Hispanic workers can be of any race, so they also appear inside the White, Black and Asian figures. In 2023, 89% of Hispanic people in the labor force were White, according to BLS's annual race and ethnicity report. That means the "White" median includes most Hispanic workers.
• "Full time" means usually working 35 or more hours a week at your main job.
• Self-employed people are excluded, whether incorporated or not.
• "Usual weekly earnings" are before taxes and include overtime, commissions and tips you usually receive.

How each gap has moved over the past decade

Quarterly numbers bounce around, so for trends the annual averages are the better guide. The table below uses BLS annual averages for full-time wage and salary workers.

Medians from BLS annual averages tables and the 2023 race and ethnicity report. Percentages are Simply Salaries calculations.

Hispanic or Latino workers: the clearest narrowing. The Hispanic median rose from 72.3% of the White median in 2015 to 77.3% in 2025. The same direction shows up in quarterly data: comparing third quarters, the ratio went from 74.0% in Q3 2016 to 77.8% in Q3 2026.

Black workers: some narrowing, but it depends on the years you pick. On annual figures, the Black median went from 76.8% of the White median in 2015 to 80.1% in 2025, after peaking at 81.5% in 2024. But compare the third quarter of 2016 ($685 against $854, or 80.2%) with Q3 2026 (80.3%), and the ratio barely moved. When two reasonable comparisons give different stories, the honest summary is: modest progress, not a clear trend.

Black men and Black women moved in opposite directions. In Q3 2016, Black men earned 78.5% of White men's median and Black women earned 84.1% of White women's. In Q3 2026 those figures were 74.6% and 87.4%. On 2025 annual averages, they were 76.7% and 85.0%.

Asian workers: the gap above the White median widened. The Asian median went from 118.9% of the White median in 2015 to 127.2% in 2025.

Dollar gaps grew even where ratios narrowed. The weekly gap between the White and Black medians was $194 in 2015 and $245 in 2025. Those are current dollars, not adjusted for inflation, and when all pay rises, a fixed percentage gap becomes more dollars. That's why ratios are the fairer way to track change over time.

The most recent year, adjusted for inflation

From Q3 2025 to Q3 2026, the overall median rose 1.5%, while consumer prices (CPI-U) rose 3.5%, so the typical full-time worker lost ground in real terms. By group, before inflation:

Percentage changes are Simply Salaries calculations from BLS current-dollar medians; constant-dollar figures are from BLS.

Black and Hispanic medians grew faster than White and Asian medians over the year, which is why both gaps narrowed between those two quarters. In inflation-adjusted terms, only the Black median rose, and the Hispanic median held steady.

Two data quirks to know about

• There is no fourth-quarter 2025 figure. The October 2025 survey wasn't collected because of the federal government shutdown, so BLS couldn't produce Q4 2025 estimates.
• The 2025 annual averages cover 11 months, leaving out October. BLS says they are not strictly comparable with other years. Treat the small dip in the Black-to-White ratio between 2024 and 2025 with some caution.

Gaps within the same education level

Do the gaps disappear when you compare people with the same schooling? Partly, for some groups. Not for others.

Start with who has which degree. In 2023, among people 25 and older in the labor force, 68% of Asian workers had at least a bachelor's degree, compared with 44% of White, 36% of Black and 25% of Hispanic workers. Twenty percent of Hispanic workers had not finished high school. BLS notes that median earnings rise with education in every major group, and it lists education among the factors associated with differences between groups.

The table below crosses education with race using 2014 annual averages for full-time workers 25 and older, published by BLS in The Economics Daily. It's older data, so read it for the pattern, not the dollar amounts.

Source: BLS, The Economics Daily, 2014 annual averages. Percentages are Simply Salaries calculations.

Three patterns stand out:

1. The Hispanic gap shrinks a lot within education levels. Compared with the overall ratio for that year, Hispanic workers with the same schooling as White workers were much closer to the White median at every level.
2. The Black gap stays at around 80% at most levels, including among college graduates, so in this table it narrows much less when education is held equal.
3. The Asian advantage shows up mainly among college graduates. Asian workers with a high school diploma had a lower median than White workers with the same education. The advantage appears at the bachelor's and advanced degree levels.

BLS updates this kind of cross-tabulation in its annual report, Labor force characteristics by race and ethnicity (Table 17 in the 2023 edition). Check the newest edition before quoting current dollar amounts.

Same job group, different medians

Occupation tells a similar story. In 2023, among men working full time in management, professional and related jobs, the highest-paid group, medians were $2,103 for Asian men, $1,782 for White men, $1,475 for Hispanic men and $1,447 for Black men. In service jobs, the spread was much tighter: $820 for White men, $797 for Asian men, $749 for Black men and $719 for Hispanic men.

Broad groups still mix very different jobs, so some of that spread reflects which jobs people hold inside each group.

Age matters too

In Q3 2026, among workers ages 16 to 24, the Black median ($708) was 90.5% of the White median ($782), and the Hispanic median ($762) was 97.4%. Among workers 25 to 54, those ratios were 79.1% and 77.0%. Gaps are much smaller at the start of working life and wider in the prime earning years.

What a raw gap measures, and what it doesn't

A raw gap compares the middle earner in one group with the middle earner in another. That's useful, but it's easy to over-read.

It tells you how the typical full-time paycheck differs between groups nationally in a given period, and whether that difference is growing or shrinking. What it doesn't tell you:

• Whether two people doing the same job are paid differently. The comparison doesn't hold job, employer, experience or location constant.
• Why the gap exists. BLS says long-standing differences between groups are associated with many factors, not all measurable. It lists age, education, occupation and industry, where people live (including urban versus rural) and the degree of discrimination encountered at work. BLS does not assign a share of the gap to any one factor, and neither should you on the basis of these tables.
• Anything about part-time or self-employed workers, who are outside these figures.
• Smaller groups. Several smaller race groups and multiracial workers aren't in the quarterly figures, and the broad categories hide big differences inside each group.

Watch out for small changes. The CPS surveys about 60,000 households, and earnings questions go to only a quarter of the monthly sample. BLS also estimates medians by grouping answers into $50-wide bands, and people tend to report round numbers. A one-quarter move of a point or two in a ratio may be noise. That's why the trend table above leans on annual averages.

How pay transparency and salary history bans come into it

Two kinds of laws are often discussed alongside these gaps:

• Salary history bans limit when an employer can ask what you earned before, or bar it from basing an offer on that figure. The idea is that a past pay gap shouldn't follow you from job to job.
• Pay transparency laws require some employers to share a pay range, in job ads, on request, or to current employees.

The BLS data in this article can't show whether these laws have changed the gaps. The figures are national, and the reports don't evaluate the laws. What the laws can do for you is practical: they give you a published range to measure an offer against, and in places with a history ban, you can anchor the conversation on the job instead of your last paycheck. Rules differ by state and city, and some apply only to employers above a certain size. Check your state labor department's website for what applies where you work, and see our negotiation and pay trends pages for more.

If you think your pay reflects discrimination

Under Title VII, along with the ADA and ADEA, employers may not discriminate in pay based on race, color, national origin and other protected traits. Unlike the Equal Pay Act, Title VII doesn't require that the jobs being compared be substantially equal. Title VII generally applies to employers with 15 or more employees.

Time limits are short. The EEOC says you generally have 180 days to file a charge under Title VII, a period that may be extended by state law, and federal employees have 45 days to contact an EEO counselor. The EEOC's equal pay and compensation page explains the rules and how to file. A national median gap is not evidence about your own employer, so if you're considering a complaint, talk to the EEOC, your state fair employment agency, your union or an employment lawyer about your specific situation.

Checking your own pay against a benchmark

The national race-by-race medians are a backdrop, not a benchmark for you. To find out if you're underpaid, compare yourself with people doing your kind of work.

1. Convert your pay to the same basis. BLS figures are weekly, before taxes, including usual overtime, tips and commissions. Divide an annual salary by 52, or multiply an hourly rate by your usual weekly hours.
2. Find the right comparison group. The quarterly release has medians by broad occupation group. For specific job titles and metro areas, use BLS occupational wage data, and see our salaries by role and salaries by city pages.
3. Look at the range, not just the middle. BLS publishes quartiles and deciles too. The 25th percentile is the point a quarter of workers earn less than, and the 75th percentile is where three quarters earn less. Where you sit in that range should match your experience and responsibilities.
4. Check the employer's own range. Where a transparency law applies, the posted or disclosed range for your role is the most relevant number you'll find.
5. Write it down. Keep the figure, the source and the date. Pay data goes stale fast.

Example (hypothetical): You work full time in production and earn $950 a week. In Q3 2026, BLS put the median for full-time workers in production occupations at $1,027 a week. You're below the middle of a very broad group. That's not proof of anything on its own, but it's a reason to look up your specific job title in your metro area and check where you land.

If the numbers support it, a short, factual script works well:

"I've been comparing my pay with published data for this role. The median for full-time workers in comparable jobs is about $1,027 a week, and I'm at $950. Given my results this year, I'd like to talk about bringing my pay closer to the middle of the range. What would that take?"

Or, if you're job hunting:

"Before we talk numbers, could you share the pay range budgeted for this role?"

What to do next

• Bookmark the BLS Usual Weekly Earnings release. It comes out quarterly; the next one, covering the fourth quarter of 2026, will also include annual revisions to seasonally adjusted data.
• Use ratios, not dollar gaps, when you compare years, and compare annual averages rather than single quarters when you can.
• Benchmark against your job, not your demographic group. Then use those numbers in your next review or offer conversation.
• If you suspect pay discrimination, move promptly. The filing windows are short, and the EEOC, your state agency, a union or a lawyer can tell you where you stand.