A new working paper from economists at Munich’s CESifo challenges the prevailing narrative that artificial intelligence is causing widespread job losses for recent college graduates. Researchers Robert Fairlie and Jane Wu argue that current data shows no significant reduction in hiring for entry-level roles, contradicting a previous Stanford study which suggested employment in AI-impacted occupations was lagging. The authors note that firms are likely reducing new recruitment for simpler, standardized tasks rather than laying off experienced staff who already hold those positions. This distinction matters because it suggests the labour market shock might be delayed or concentrated in specific hiring cycles rather than existing as immediate mass unemployment. The paper highlights that while AI spending and enterprise token usage have risen sharply over the last twelve months, these metrics have not yet translated into a measurable drop in absolute hiring levels for fresh graduates. The findings imply that the transition of tasks to machines is occurring without the immediate contraction of the workforce that many predicted. Consequently, the expected downturn for new entrants appears less severe than anticipated, though risks may remain for those entering the market in future years.
- The study focuses on recent graduates because changes in labour demand often appear first through reductions in new hiring.
- Firms are replacing employee tasks with AI tools rather than firing long-term staff.
- Sharp increases in AI spending and ChatGPT Enterprise token use have not yet caused a drop in hiring.
