Junior Developer Hiring Fell 73% Last Year. The Industry May Not Feel It for Five More.

Entry-level hiring at the top 15 tech firms dropped 73% year-over-year while overall hiring fell just 7% — a gap that Juan Cruz Martinez, writing in The Long Commit newsletter, argues signals something more dangerous than a hiring cycle correction.

New graduates have gone from roughly a third of all hires at major tech companies in 2019 to about 7% today. Many of the decisions behind that slide have been explicit. Salesforce announced a complete halt to engineering hiring for 2025, citing AI agents. Block cut 40% of its workforce, with CEO Jack Dorsey framing it as <a href="/news/2026-03-14-tech-layoffs-45000-march-ai-attributed">an AI-enabled efficiency move</a>. Klarna froze developer hiring in late 2023, then reversed course after the strategy proved unworkable — a rare admission that the math doesn't always hold.

Academic research backs the trend. Stanford's Digital Economy Lab tracked millions of payroll records and found software developer employment for workers aged 22 to 25 declined nearly 20% from its late-2022 peak, while employment for those over 30 held steady or grew. A Harvard study of 62 million workers across 285,000 firms found generative AI adoption correlates with a 9 to 10% drop in junior employment within six quarters, with senior employment largely unaffected. A LeadDev survey found 54% of engineering leaders plan to hire fewer juniors because of AI copilot productivity gains. On paper, avoiding $80,000 to $100,000 salaries and six-month ramp-up periods looks like <a href="/news/2026-03-15-who-captures-ai-productivity-gains-labor-vs-capital">a clean trade</a>.

Martinez's argument is that those calculations miss what juniors actually do for senior engineers. On a functioning team, lower-risk tasks flow down the hierarchy, giving seniors cognitive space for architecture and design. When AI substitutes for that layer, seniors don't gain time for deeper work — they absorb an unmanageable review burden generated by AI's own high output volume. A Harness survey found 67% of developers already spend more time than expected debugging AI-generated code. LeadDev puts 22% of developers at critical burnout levels. The mentorship loop that once transferred institutional knowledge has been severed at the same moment the pressure valve preventing senior burnout has been removed.

What Martinez calls the three-to-five-year talent cliff isn't speculative — it's an accounting of decisions already made. When today's senior engineers eventually exit, there's no developed cohort beneath them. Years of foregone hiring and knowledge transfer don't reconstitute quickly. The sharper risk, he argues, is that most companies are making this bet simultaneously without recognizing the shared exposure they're creating. One firm cutting junior hires is a headcount decision. The entire industry doing it in lockstep is a structural problem with no easy unwind.

Klarna found that out early and reversed. Martinez, a 20-plus-year industry veteran, thinks most companies won't get the signal until the cliff arrives.