U.S. Gender Earnings Gap

Multi-dataset analysis of the hourly gender pay gap using ACS, CPS ASEC, SIPP, ATUS, and SCE public data. Prime-age (25-54) wage/salary workers, 2015-2023.

16.8% Raw Hourly Gap (2023)
13.2% Adjusted Gap (ACS)
7M+ ACS Observations
13-14% Adjusted Gap Band

Headline figures come from the year-by-year sequential OLS trend. Gelbach and Oaxaca are reported as supporting decomposition tools in Methods.

Key Findings

Persistent adjusted gap

After controlling for age, education, race, occupation, industry, hours, and family status, a 13.2% hourly gap remains in 2023 ACS data. CPS shows 17.0% adjusted. The gap has been stable at 13-14% across 8 years of ACS data.

Trends →

Job sorting drives largest reduction

Adding occupation and industry controls reduces the raw gap by ~8 percentage points. Gelbach confirms job sorting is the largest observable channel, with reproductive burden the next-largest order-invariant block.

Mechanisms →

Gap varies sharply by subgroup

Healthcare 30.6%, Black 5.6%, Military 5.5%. The gap is widest among higher-educated workers (Masters 27.4%) and in sales/office occupations (28.6%).

Heterogeneity →

Reproductive burden explains ~3 pp

Adding reproductive stage and couple type reduces the adjusted gap from 13.7% to 10.8%. Job rigidity and motherhood interactions absorb most of the remainder.

Reproductive →

Male earnings more dispersed

Men’s raw hourly variance is about 10% higher, but after controls the residual ratio is only about 1.04; the bigger difference is concentration in the upper tail.

Variance →

At a Glance

ACS Raw vs Adjusted Hourly Gap (2015-2023)

Prime-age wage/salary workers. Adjusted series controls for age, race, education, state, occupation, industry, hours, work-from-home, commute, marital status, children. No 2020 ACS due to COVID collection issues.

Explore

Trends

Year-by-year gaps across ACS and CPS, 2015-2023.

Heterogeneity

Gaps by race, education, occupation, and industry.

Mechanisms

Specification ladders, Gelbach attribution, and mechanism evidence.

Reproductive

Reproductive-burden extension: fertility risk, same-sex placebo, interaction channels.

Variance

Distributional analysis: dispersion, tails, reproductive & job-context heterogeneity.

Occupations

Harmonized occupation-level variance, leaderboards, and pre/post-2020 shifts.

Methods

Sample construction, variable definitions, models.

FAQ

Common questions on measurement and interpretation.