A new study by Federal Reserve economists Leland D. Crane and Paul E. Soto has found that employment growth among U.S. programmers dropped roughly 50% after ChatGPT launched in November 2022 — providing the first Fed-level research to directly connect AI adoption to a measurable, occupation-specific decline in developer hiring.
Before ChatGPT's arrival, programming-intensive jobs were growing at around 5% annually, well above the overall labor market. Since then, growth has fallen sharply. In sectors most concentrated with programmers, such as IT services and software development, job growth has essentially flatlined.
Skeptics have long argued that interest rate hikes, the end of the pandemic-era digital boom, and the crypto crash of 2022 alone explain the developer slowdown. Crane and Soto addressed this directly by constructing a counterfactual — modeling how many programmers would exist if their share within each industry had remained constant. Even after stripping out those broader economic effects, programmer employment was still falling by about 3% per year. Non-AI-exposed occupations showed no comparable dip.
Stretched over three years, the gap amounts to roughly 500,000 jobs that would likely have existed without the rise of large language models. The authors strongly caution against reading this as a direct count of lost positions, noting that many affected workers probably found employment in adjacent fields and that the study does not capture broader macroeconomic feedback. Nevertheless, the signal is clear.
Notably, the employment gap did not open until mid-2024 — approximately 18 months after ChatGPT's launch. The researchers suggest companies needed time to see LLM capabilities improve sufficiently before pulling back on headcount. Whether that reflects actual productivity gains or simply the expectation of them, the data does not resolve.
The study identifies programmers as the most AI-exposed occupational group in the country, a finding consistent with actual usage data. Anthropic's Economic Index shows that computer and mathematical tasks — coding, debugging, and software architecture — account for roughly a third of all Claude.ai conversations and nearly half of enterprise API traffic.
A separate Harvard study of 62 million automatic data processing payroll workers found that junior developer employment drops roughly 9–10% within six quarters when companies adopt generative AI, while senior employment barely moves. "If A.I. disproportionately affects junior positions, it could have lasting consequences for the college wage premium, upward mobility and income disparities," Harvard researchers wrote.
Broader expert opinion is also shifting. A recent multi-university survey of 69 economists, 52 AI experts, and 38 superforecasters found broad agreement that faster AI progress means lower labor force participation — including among researchers who previously held the "augment, not replace" consensus.
The Fed researchers stop short of framing their findings as catastrophic. Wages for programmers have not declined; the effect has shown up in headcount rather than pay. Job postings stabilized in 2024 and have ticked slightly upward since. The authors also note that cheaper AI-assisted programming could open new markets and grow total demand for developer labor over the long run.
Crane and Soto describe their work as "only a first step." The study was published under a preliminary designation, meaning it has not yet completed the full Federal Reserve review process. Still, it represents the first analysis of its kind produced within the Fed, carrying the methodology and institutional weight that entails.
Why it matters
The employment gap appeared roughly 18 months after ChatGPT launched, suggesting a lag between AI tool availability and firms actually restructuring headcount — a pattern that could repeat as newer models are deployed across other occupations.
Because the effect has shown up in hiring volume rather than wages, standard wage-based labor market indicators may not capture AI-driven displacement in real time, complicating how policymakers and analysts read labor market health.
The Harvard finding that junior roles absorb the bulk of the decline has structural implications for career entry points in tech, since junior positions traditionally serve as the training ground for senior expertise.
