AI impact on new grads: Unemployment data defies expectations
▼ Summary
– A new working paper from CESifo researchers challenges previous Stanford findings by stating there is no evidence of significant hiring displacement for recent college graduates due to AI.
– Researchers Robert Fairlie and Jane Wu analyzed US Census microdata to track unemployment trends among recent graduates, finding that labor demand changes have not yet manifested as widespread hiring reductions.
– Despite the study’s conclusions, anecdotal evidence from venture capitalist Marc Andreessen and BlackRock CEO Larry Fink suggests fears are growing that AI capabilities may soon replace entry-level roles.
– The article notes a sharp increase in corporate AI spending and task automation, leading some experts to predict higher unemployment rates for the 2026 graduating class.
– The study focuses on recent graduates because firms are likely to reduce hiring for standardized entry-level tasks before laying off more experienced employees.
Recent labor market data challenges the prevailing narrative that artificial intelligence is causing widespread job losses for new graduates. While a Stanford University study previously suggested that entry-level roles in AI-impacted occupations are struggling compared to other fields, fresh research from economics scholars at Munich’s CESifo presents a contradictory view. The working paper asserts that “there is no evidence of any significant, widespread displacement or reduction in hiring of recent college graduates in absolute or relative levels.”
The study, titled “The Early Impacts of AI on Employment Among Recent College Graduates,” was authored by researchers Robert Fairlie and Jane Wu. They chose to concentrate on recent graduates because shifts in labor demand often manifest first through hiring freezes rather than layoffs. As AI systems become proficient at handling the standardized tasks typical of many entry-level office positions, companies may opt to stop recruiting for junior roles instead of terminating tenured staff. This strategic shift allows firms to reduce costs while retaining experienced workers who possess institutional knowledge and complex problem-solving skills that current AI models cannot fully replicate.
Despite these findings, there are indicators that the 2026 graduating class faces higher risks than their predecessors. CESifo researchers highlight a surge in corporate adoption of AI technologies, citing Census survey data showing a sharp rise in firms replacing a large number of employee tasks with AI. Additionally, metrics such as increased AI spending per employee and heightened usage of ChatGPT Enterprise tokens over the past year suggest accelerating integration into business operations.
High-profile industry leaders have echoed concerns about this rapid technological shift. Venture capitalist Marc Andreessen stated earlier this year that “AI literally until December [2025] was not actually good enough to do any of the jobs that they’re actually cutting.” Similarly, BlackRock CEO Larry Fink expressed caution in March, noting that “the speed at which AI is changing” raises fears that “when this year’s college graduates enter the workforce, we could see the highest unemployment rate among them in years,even without a recession.” These warnings underscore the tension between anecdotal industry anxiety and the broader statistical trends observed by academic researchers.
To evaluate whether these apprehensions are grounded in reality, the CESifo team analyzed granular microdata from the US Census’ Current Population Survey. Their focus was on identifying unemployment patterns among recent college graduates, defined specifically as Bachelor’s degree holders aged 22 to 25 who are not pursuing further education. Since unemployment figures naturally fluctuate as new entrants flood the job market during the summer months, the researchers examined both year-over-year and seasonal trends dating back to 2022. This starting point is significant as it marks the return of employment to pre-pandemic levels and coincides with the public release of ChatGPT, providing a baseline to measure the early effects of generative AI on the labor market.
(Source: Ars Technica)