CESifo Study Finds Little Direct Evidence AI Is Driving Graduate Unemployment
A CESifo report covered by TechRadar says there is little direct evidence that AI is causing recent-graduate unemployment to rise, though it links remote work to some job risk and says future effects cannot be ruled out.
The authors argue that any existing graduate unemployment has many causes beyond AI, making it difficult to attribute job losses to the technology. They note that unemployment typically rises during summer months, when large numbers of graduates enter the market, and falls again in the fall. Since 2022, just before ChatGPT's public preview launch, summer unemployment averages among graduates have ranged from 6.3% to 7.8%, the report says, indicating little change.
CESifo's research acknowledges that the most AI-exposed roles have changed in ways that might be expected if AI were harming employment. However, the report says those shifts are not statistically significant enough to conclude that AI is replacing younger human workers. The paper also correlates graduate unemployment with remote working, suggesting that work-from-home jobs could be more at risk than AI-exposed roles.
Because AI-exposed roles are typically the most desk-based jobs and are also often suitable for remote setups, earlier research could have confused or blended the two impacts, according to the report. The authors concluded that "there has been little direct evidence" that AI will replace jobs, while admitting that the world is still in the very early years of AI and that the study is merely "a useful first test." They stressed that "these findings do not rule out larger effects in the future."