A new working paper from economics researchers at Munich CESifo institute argues point blank that there is no evidence of any significant, widespread displacement or reduction in hiring of recent college graduates in absolute or relative levels. The findings, published in The Early Impacts of AI on Employment, directly contradict a Stanford University study released last month that found entry-level employment in AI-impacted occupations lagging behind other fields.
The CESifo researchers analyzed labor-market data across multiple OECD countries, controlling for sector, education level, and regional variation. Their conclusion: while AI is reshaping task composition within jobs, the aggregate employment numbers for recent graduates in AI-exposed fields show no statistically significant decline.
What we are seeing is task substitution, not worker substitution, said one of the paper authors. AI is changing what people do within their roles, but it is not eliminating the roles themselves — at least not yet, and not at the scale the panic narratives suggest.
The Stanford study, by contrast, focused specifically on US labor markets and used a different methodology that weighted job-posting data more heavily. The discrepancy highlights how sensitive AI-employment findings are to methodology, geography, and time horizon.
Both research teams agree on one point: the data covers only the early phase of AI deployment. Whether mass displacement materializes depends on the pace of capability advances and the speed of organizational adoption — neither of which has a clear trajectory yet.