What Is The Gender Pay Gap? (Earnings Analysis)
The gender pay gap is the difference in earnings between women and men, measured for a defined group and period. In 2024, women working full time in the United States had median usual weekly earnings equal to 82.7% of men’s, according to the Bureau of Labor Statistics. That overall comparison describes a difference; by itself, it does not explain why it exists.
When you see a pay-gap figure, the key question is: what exactly is being compared? A statistic based on weekly earnings may differ from one based on hourly wages or yearly income. The group of workers, the year, and the calculation all matter.
I start by identifying those details before interpreting a number. That simple step helps prevent a common mistake: treating an overall earnings gap as proof that two people doing the same job receive different pay, or treating an adjusted estimate as proof that unequal treatment never occurs.
Defining the Gender Pay Gap and Choosing the Right Measure
The gender pay gap is a statistical comparison of earnings for women and men in a stated population. It can use hourly, weekly, or annual pay, and it can describe the overall difference or compare workers after accounting for selected factors. A useful figure always names its measure, population, location, and time period.
The gap is often reported as a ratio: women’s earnings divided by men’s earnings, then multiplied by 100. A ratio of 82.7% means that women’s median earnings were 82.7 cents for each dollar of men’s median earnings in that comparison. The related gap is 17.3% of men’s earnings.
These figures describe a group, not every worker. A particular woman may earn more, the same, or less than a particular man. The statistic does not match individual workers or jobs unless the analysis is specifically designed to do so.
What the 2024 U.S. benchmark measures
The Bureau of Labor Statistics (BLS) reported that women’s median usual weekly earnings were 82.7% of men’s in 2024. The comparison covers full-time wage-and-salary workers in the United States and uses median weekly earnings. It is an unadjusted, descriptive comparison: it does not control for job, hours beyond the full-time definition, experience, or other worker characteristics.
The BLS measure uses the Current Population Survey (CPS). “Median” means the middle value when workers’ earnings are arranged from lowest to highest. Half of the measured workers earn more than the median, and half earn less. A median is less affected by very high or very low earnings than an average.
The earnings are nominal dollars, meaning they are not adjusted for inflation. The 2024 ratio compares women and men within 2024, so inflation is not needed to calculate that ratio. Inflation matters when comparing the dollar amounts across different years.
How to calculate a pay-gap ratio
To calculate the ratio, divide the selected women’s earnings measure by the selected men’s measure, then multiply by 100. To calculate the gap as a percentage of men’s earnings, subtract the ratio from 100%. Always use the same earnings measure, population, location, and time period for both groups.
Using the BLS 2024 ratio:
- Earnings ratio: 82.7%
- Gap relative to men’s median earnings: 100% − 82.7% = 17.3%
The 17.3% is not a statement that every woman earns 17.3% less than a man in the same role. It is the difference between two group medians, expressed as a share of men’s median earnings.
Data, Methods, and Limits of Earnings Comparisons
A sound earnings analysis states where its data came from and how the comparison was made. The data source, worker sample, earnings measure, and period affect the result. If a study adjusts for worker or job characteristics, it must also list those factors and explain that they change the question being answered.
| Measure | What it compares | What it can show | What it cannot show alone |
|---|---|---|---|
| Hourly earnings | Pay per hour worked | A comparison less affected by differences in hours | Why hourly rates differ |
| Usual weekly earnings | Typical weekly pay | Differences in pay and, in part, work patterns | Equal pay for identical work |
| Annual earnings | Pay over a year | Total yearly earnings in a defined group | Whether differences stem from rates, hours, or time in work |
| Adjusted estimate | Earnings after selected factors are considered | Whether a difference remains under that model | A definitive cause or proof that discrimination is absent |
Unadjusted and adjusted estimates answer different questions
An unadjusted gap compares earnings as observed, without statistically accounting for other characteristics. It captures the combined result of many influences, such as occupation, industry, work hours, seniority, and access to higher-paid roles. It does not separate those influences or establish why the difference exists.
An adjusted gap uses a statistical method, often regression analysis, to compare earnings while holding selected factors constant. “Holding constant” means the model accounts for measured differences in those factors. For example, a study might include occupation, education, hours, experience, and location.
Adjustment changes the question from “How do earnings differ overall?” to “How do earnings differ among workers who are alike on the factors included in this model?” The answer depends on the model. If researchers leave out relevant factors, measure them poorly, or include factors that are part of the inequality being studied, the result may change.
For that reason, an adjusted estimate is not automatically the “true” gap. It is an estimate under stated assumptions and controls. It also does not, by itself, prove or disprove discrimination.
Why earnings measures and samples matter
Hourly pay, weekly pay, and annual earnings are not interchangeable. Someone who works fewer paid hours may have lower weekly or yearly earnings even if their hourly rate is the same. Annual earnings can also reflect periods when a person was not working or was working fewer hours.
The population matters, too. The BLS 2024 figure concerns full-time wage-and-salary workers, not all people, all forms of work, or every occupation. It should not be presented as the pay gap for part-time workers, self-employed people, a single industry, or a specific employer.
When comparing studies, check that they use matching definitions. A figure for hourly wages in one year cannot be directly compared with annual earnings from another year as if they measured the same thing. Also check whether a study reports a mean (average) or median; each summarizes the data differently.
A step-by-step check for any reported gap
Before using a figure in a paper, workplace review, or news discussion, record:
- Population: Which workers are included? Note location, age range, employment status, and any other limits.
- Earnings measure: Is the figure hourly, weekly, or annual? Is it a mean or median?
- Period: What year or date range does the data cover? Are dollar amounts inflation-adjusted?
- Comparison: Is the result unadjusted, or does it account for selected factors?
- Source and method: Which data set and statistical approach were used?
- Limits: What does the measure leave out, and what conclusion does it support?
This checklist makes it easier to compare like with like and to describe findings without claiming more than the data show.
Interpreting Results and Evaluating the Evidence
A pay-gap figure is strongest when read as a carefully defined finding, not a complete explanation. An overall ratio can show that group earnings differ. An adjusted estimate can show whether a difference remains after selected controls. Neither result alone identifies every cause or answers every legal question.
Worked examples: what a number means
The first example below uses the BLS 2024 ratio. The other rows are hypothetical calculations, included to show how different measures can lead to different descriptions. They are not reported findings about actual workers.
| Scenario | Women’s measure | Men’s measure | Calculation | Careful interpretation |
|---|---|---|---|---|
| BLS 2024 weekly earnings ratio | 82.7% of men’s median | 100% | Reported ratio | An overall comparison for the stated worker group |
| Hypothetical hourly pay | $24 | $30 | $24 ÷ $30 = 80% | Women’s hourly rate is 80% of men’s in this example |
| Hypothetical yearly pay | $48,000 | $60,000 | $48,000 ÷ $60,000 = 80% | Women’s annual earnings are 80% of men’s in this example |
Suppose an analysis finds an unadjusted difference, then a smaller difference after controlling for occupation and hours. A careful report would say that those controls account for part of the measured difference in that data and model. It should not say that occupation and hours explain every cause, or that the remaining difference has only one explanation.
What the numbers can and cannot establish
A descriptive earnings gap can help identify patterns that deserve closer study. Researchers may examine whether the gap differs by occupation, age group, race, education, or other characteristics, provided the data support those comparisons. Results can differ across these groups, so one national figure should not stand in for every worker.
An adjusted comparison can help test whether a measured difference remains among workers similar on specified characteristics. But observational data usually cannot establish cause on their own. A model may not capture every relevant difference, and a factor included as a control may itself reflect unequal access to jobs or opportunities.
A pay-gap statistic also is not, by itself, a finding about a particular employer or a legal decision. To assess a specific workplace or claim, more detailed evidence and the relevant rules are needed. Keep the conclusion close to the evidence: state the result, name the method, and explain the limits.
Career, Policy, and Employment Implications
Earnings-gap analysis can inform questions about pay practices, job patterns, and public policy, but the implications depend on the data and the purpose of the analysis. A group-level statistic can point to a pattern for further review. It cannot determine an individual’s pay, prove a cause, or replace a detailed workplace analysis.
For employers, a well-defined comparison may help identify areas to review, such as pay ranges, starting salaries, promotion patterns, or job classifications. The result should be examined with suitable data and consistent job comparisons. A broad national ratio is not a substitute for reviewing an organization’s own workforce.
For students and researchers, the main task is to make the claim match the evidence. A clear conclusion might say: “In the specified population and year, women’s median earnings were a stated share of men’s; this unadjusted comparison does not isolate causes.” If reporting an adjusted result, name the controls and sample.
When reviewing a policy claim, ask whether it is based on an overall earnings difference, a matched-job comparison, or an adjusted estimate. Those measures can all be useful, but they answer different questions. The next step is to verify the underlying data and avoid treating one measure as a complete account.
Frequently Asked Questions
These short answers clarify common terms and interpretation issues. For any figure, check the source’s population, earnings measure, time period, and method before comparing it with another statistic or applying it to a specific workplace.
What does a gender pay gap of 17.3% mean?
It means the earnings measure for women is 17.3% below men’s measure, using men’s earnings as the reference, in the stated comparison. It does not describe every individual worker.
What was the U.S. gender pay gap in 2024?
The BLS reported women’s median usual weekly earnings at 82.7% of men’s for full-time wage-and-salary workers. This is an unadjusted comparison for that population and year.
Does the gender pay gap mean women get paid less for the same job?
Not by itself. An overall gap compares group earnings and does not establish whether two people doing the same work receive different pay.
What is the difference between an adjusted and unadjusted pay gap?
An unadjusted gap compares earnings without accounting for other factors. An adjusted estimate accounts for selected factors, such as occupation or hours, and depends on the data and model used.
Is the pay gap based on hourly or yearly pay?
It depends on the source. The BLS 2024 benchmark above uses usual weekly earnings. Other studies may use hourly wages or annual earnings, so check the measure before comparing figures.
Why can the pay gap change from one study to another?
Studies may use different years, worker groups, earnings measures, data sources, and statistical controls. Those choices can produce different results without either study using the same definition.
Does an adjusted pay gap prove discrimination is absent?
No. An adjusted estimate accounts only for measured factors included in the model. It cannot, on its own, establish that discrimination is absent or identify every cause of a remaining difference.
Can I apply the national pay-gap figure to my workplace?
Not directly. A national statistic describes a broad population. A workplace analysis needs suitable employer data and a clear comparison of workers, jobs, earnings, and relevant factors.
What should I cite in an earnings analysis?
Name the data source, release period, population, earnings measure, and method. For the 2024 U.S. weekly earnings benchmark, cite the BLS Current Population Survey and its 2024 women’s earnings report.
Should I compare an average or a median?
State which one you use. A mean is the arithmetic average and can be pulled by very high earnings. A median is the middle value. They summarize distributions differently and should not be treated as interchangeable.
(This article was written by one of our staff writers, Nathan Sterling. Visit our Meet the Team page.)
