AI, Young Entrants, and the Limits of the Headline
Labor productivity is often recruited into a larger story about artificial intelligence: if output per hour is rising, perhaps machines are replacing people; if young graduates struggle to find work, perhaps language models have already closed the entry ladder. This chapter examines that story through the OECD evidence on young labor-market entrants, AI hiring, job postings, occupational exposure, and productivity. The useful judgment is narrower and more demanding: current data show real difficulties for young entrants and suggest some technology-related changes, but they do not make large language models the main explanation for the observed unemployment gaps so far.
The distinction matters because “AI,” “jobs,” and “productivity” are not single measures. AI hiring can be a proxy for firms seeking people with AI-related skills. Occupational exposure describes how much an occupation's tasks may be affected by language-model capabilities. Online postings record advertised opportunities, not completed hires. Labor productivity relates real output to hours worked. The unemployment gap for young entrants describes a labor-market outcome. Evidence at one point in this chain cannot automatically prove the next link.
Why young entrants are unusually exposed to a cooling labor market
The OECD reports that young people entering the labor market have become increasingly exposed to unemployment relative to the rest of the working-age population. The pattern differs by education. The unemployment gap for young college graduates had been rising since before the COVID-19 pandemic in all the countries analyzed. For young people without a college degree, the gap increased more recently in Canada and the United States, but not in Australia or the euro area.
An unemployment gap is a relative measure: it compares the unemployment rate of a defined group of recent entrants with the rate for the broader working-age population. That comparison matters because a young entrant can face a worsening position even when overall unemployment remains low. If experienced workers keep their jobs while employers reduce new hiring, the total unemployment rate may move only modestly while the path into work becomes much harder for people without an established employment history.
Entry-level workers are exposed to this margin for reasons that do not require an AI explanation. They are more likely to be searching for a first job, moving between short contracts, or competing for roles in which employers can pause recruitment quickly. When labor demand weakens, a firm may leave a vacancy unfilled or reduce the number of new positions before ending the employment of experienced workers. The result is a bottleneck at the point of entry: employment can remain resilient for people already inside organizations while new entrants wait longer for a first opportunity.
The OECD points to several conditions that may compound the problem. Hiring weakened in some sectors, including technology, information, and business services after heavy pandemic-era hiring. Real wages remained below their 2021 levels in many countries, which can make available work less attractive than expected. Some of the flexibility associated with remote work had been rolled back. Fixed-term and entry contracts can be especially sensitive when employers adjust to uncertainty. These factors do not have identical effects, but they all change the opportunity structure facing a young worker.
Longer-run changes can operate alongside the cyclical slowdown. The OECD discusses digital offshoring and the rapid evolution of skill requirements. Hiring models are increasingly emphasizing demonstrated skills rather than formal credentials in some settings.