AI displacement has not driven a broad rise in US graduate unemployment, study finds
A CESifo analysis found recent US graduates’ summer unemployment rate remained within its range from earlier years, but researchers say later classes could face greater exposure.
A CESifo working paper found no statistically significant evidence that artificial intelligence had caused widespread displacement or reduced hiring among recent US college graduates through summer 2026. The researchers cautioned that the findings are an early test, not evidence that AI has had no effect.
Robert Fairlie and Jane Wu analysed US Census Current Population Survey data for bachelor’s degree holders aged 22 to 25 who were not pursuing further study. They focused on recent graduates because any change in labour demand could first appear in reduced hiring.
The group’s summer 2026 unemployment rate was 7.3 per cent, within the 6.3 to 7.8 per cent range recorded in summers from 2022 to 2024. Comparisons with other age and education groups, and with jobs classified by potential AI exposure, generally found no statistically significant differences in employment trends from 2022 to 2026.
The findings differ from a recent Stanford study using ADP payroll data, which reported weaker entry-level employment in occupations considered exposed to AI. The studies measure different things: ADP tracks payroll employment in selected fields, while the CESifo paper examines unemployment using Census survey data.
Fairlie and Wu said later graduating classes could be more affected if workplace use of AI continues to grow. They said further years of data will be needed to determine whether effects emerge as adoption deepens.
