Common Questions

How does Stanford calculate 34 AI jobs lost per layoff?

Stanford Digital Economy Lab compared the JOLTS 3.1% February 2026 hiring rate with the 2023 baseline, finding 950,000 to 1 million fewer annual hires, against Challenger's 27,645 declared AI layoffs — a 34-to-1 ratio.Source: Stanford Digital Economy Lab / ProCap Insights

Why are young workers hit hardest by AI job displacement?

Workers aged 22 to 25 in AI-exposed occupations are 16% below late-2022 employment levels, while colleagues over 30 in the same roles are up 6 to 12%. Entry-level positions are more easily automated; senior roles require contextual judgement current models lack.Source: Stanford Digital Economy Lab

What is the JOLTS hiring rate and why does it matter for AI?

JOLTS (Job Openings and Labor Turnover Survey) is the BLS measure of monthly hires. Its February 2026 reading of 3.1% — the lowest since April 2020 — is the starting point for Stanford's estimate that AI is suppressing nearly 1 million annual hires.Source: Bureau of Labor Statistics

Background

The Stanford Digital Economy Lab is a research centre at Stanford University directed by economist Erik Brynjolfsson, focused on how digital technologies reshape labour markets, productivity, and economic growth. Brynjolfsson co-authored The Second Machine Age and has tracked AI labour displacement since the mid-2010s, making the Lab the most cited academic source on AI workforce impact in the US policy debate.

The Lab's most consequential measurement to date came in April 2026, applying the JOLTS 3.1% hiring rate — the lowest since April 2020 — to the 158.6 million nonfarm workforce. The result: AI is preventing roughly 950,000 to 1 million annual hires against the 2023 pace, a ratio of approximately 34 hires suppressed for every one declared AI layoff. That ratio reframes how displacement is measured: the dominant channel is hires that never happen, invisible to standard labour statistics until cohorts show up missing years later. The 34:1 finding has been reasserted as each subsequent Challenger record is published — including the May 2026 record of 38,579 cuts — reinforcing the framing of declared counts as a floor, not a ceiling.

The Lab's earlier work corroborated the age-concentration pattern: workers aged 22 to 25 in AI-exposed occupations are 16% below late-2022 employment levels, while colleagues over 30 in the same roles are up 6 to 12%. Young software developers are approximately 20% below their 2022 peak. The figure arrived the same week the Bureau of Labor Statistics skipped its scheduled GenAI workplace publication, shifting the evidence burden onto academic and regional bank sources — and giving the Lab an outsized role in a debate that central data agencies have largely vacated.