Gap Trends 2015-2023
Year-by-year gender earnings gaps across ACS and CPS. Raw and sequentially adjusted OLS series are the headline trend estimates; Oaxaca appears below as a supplemental decomposition check.
1. ACS Hourly Gap Trend
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.
| Year | Raw % | Adjusted % | N (approx.) |
|---|---|---|---|
| 2015 | 18.65 | 14.20 | ~900K |
| 2016 | 18.43 | 13.65 | ~900K |
| 2017 | 17.83 | 13.65 | ~900K |
| 2018 | 17.81 | 13.59 | ~900K |
| 2019 | 16.95 | 13.86 | ~900K |
| 2021 | 15.94 | 14.21 | ~900K |
| 2022 | 17.10 | 13.76 | ~900K |
| 2023 | 16.81 | 13.23 | ~900K |
2. CPS ASEC Trend
CPS ASEC Raw vs Adjusted Gap (2015-2023)
CPS ASEC includes 2020. Adjusted model includes all available CPS controls.
3. Cross-Dataset Comparison (2023)
2023 Gap Comparison Across Datasets
4. Oaxaca Decomposition Trend
Oaxaca: Explained vs Unexplained Share (ACS)
After 2019, the unexplained share rises sharply (74% → 88-94%) while the total gap stays stable. This reflects a collapse in the explained component under a stable total gap, not an explosion in the gap itself, which is why Oaxaca is treated as supplemental to the raw and adjusted trend lines above.
5. ACS Specification Ladder (2023)
Sequential Control Blocks (ACS 2023)
Adding occupation and industry controls drives the largest gap reduction: ~8 percentage points. Family controls add only 0.06 pp.