Los Angeles, CA
Sections
The LA Globe
AI

Young Coders Aren't Mostly Being Laid Off. They're Not Getting Hired.

Stanford economists tracking ADP payrolls say the drop comes mostly from fewer young hires, and they can't yet pin it on AI.

4 min read
ShareXFacebookEmail
Comments
A female programmer writing Java code with JUnit, photographed in November 2014.
A female programmer writing Java code with JUnit, photographed in November 2014. (Joonspoon / Wikimedia Commons, CC BY-SA 4.0)

Software developers aged 22 to 25 are down about a fifth since ChatGPT launched, on payrolls Stanford economists track. Developers in their 40s went the other way. A chart of both lines was posted on X on Saturday by The Kobeissi Letter (@KobeissiLetter).

Showing it loads X’s own player, and X can see the visit.

The account named the culprit in its first line: “The AI revolution is leaving younger tech workers behind.” It put the 26-to-30 group down 4%. The economists who produce the numbers agree the gap is real and growing, but say in their own notes they can't yet prove AI caused it. And the drop they describe comes mostly from young people not getting hired, not from layoffs.

The picture in the post shows five lines, one per age group, all set to 100 in November 2022. The line for developers aged 41 to 49 climbs to about 120. The youngest group's line slides to about 80, then hooks slightly upward at the very end.

The data behind it comes from the Canaries Dashboard, which Stanford's Digital Economy Lab builds with ADP Research out of payroll records. At launch, its sample held about 25,000 firms with five years of payroll history. The lab says that sample isn't the whole U.S. job market, or even every ADP client.

Software developers sit in the lab's most AI-exposed tier of jobs. The dashboard reports steep declines for the youngest developers, smaller ones for the next age bands up and growth for everyone older.

The mechanism is the part the caption skips. The lab's paper, by Erik Brynjolfsson, Bharat Chandar and Ruyu Chen, was revised Aug. 12 with data through June. It finds the shift works mainly through companies hiring fewer young workers, not through more of them leaving, and that employers are adjusting head count rather than base pay. It's a narrower door, mostly, not people pushed out of the room.

It isn't only coders. Across all highly exposed jobs, the authors wrote in August, employment for 22-to-25-year-olds is about 19% short of the level it would have reached by tracking peers in less-exposed work. A year earlier, that shortfall was 15%.

The revision also floats an idea about who lands on which line. Young workers lost ground in jobs built on codified knowledge, the kind you learn from a textbook, while experienced workers gained in jobs that run on know-how picked up through practice. The authors suggest generative AI reproduces what's already written down better than what's learned on the job, which fits a chart where the youngest developers sink and the oldest climb.

That's where they stop. In a February follow-up, the three wrote that “the precipitous decline in employment for young software developers is a consequence of various factors.” With their strictest controls, they found, the drop in exposed jobs only turns statistically significant in 2024, and the earlier slide likely owes at least partly to other forces.

The August revision piles on caveats. The gaps shrink once education is accounted for, some of the divergence predates generative AI, and the gaps look bigger in the ADP sample than in national surveys.

They also list reasons to think AI does matter, and those are the best case for the post. The divergence kept widening well after interest rates peaked, and it holds up when tech companies and computer jobs are pulled out of the data entirely.

“We do not know whether the patterns documented here will accelerate, stabilize, or reverse,” the authors wrote in August. The dashboard behind the chart updates every month, and the youngest line's last move was up.

Source: X, retrieved November 15, 2026. Other sources: digitaleconomy.stanford.edu, digitaleconomy.stanford.edu (2), digitaleconomy.stanford.edu (3), digitaleconomy.stanford.edu (4).

Have a tip we should know? Email tips@thelaglobe.com.

Stay in the Orbit

The best of L.A. in two emails a week: where to eat, what to do, what everyone’s talking about. Mondays and Fridays. Free. Unsubscribe any time. Privacy.

Or follow the Globe on Facebook.

Comments