Junior Developer Employment Down 19% Since 2022 Peak as AI Automates Entry-Level Coding Tasks
Tags AI · Enterprise · Labor Market · Developer Tools
Analysis of ADP payroll data by Stanford's Digital Economy Lab shows US software developers aged 22–25 have declined 19% from their late-2022 peak, while developers 41–49 grew 14%. Entry-level job postings are down 28%; CS graduate unemployment (6.1%) now exceeds liberal arts majors. The decline accelerated in 2024–2025 as coding assistants shifted from line autocomplete to ticket-level agents. Meanwhile, total developer employment rose 4.4% and GitHub added 36M accounts in its latest Octoverse year — but 63–80% of new 'vibe coding' users identify as non-developers (marketers, analysts, founders), so the labor-market signal is diluted across occupational categories. IBM is tripling entry-level hiring around a redesigned junior role; Salesforce hired zero engineers last fiscal year.
Technical significance
The data reveals a k-shaped labor market: aggregate developer employment grows while the traditional apprenticeship pipeline collapses. The 'computer programmer' BLS category (code-to-spec) fell 16% YoY vs. a projected 6% per decade. This breaks the junior-to-senior talent pipeline — AI writes the mediocre code juniors used to cut their teeth on, so nobody hires juniors, so no one becomes the senior reviewer. The surge in AI-native non-developer builders (GitHub: 36M new accounts, 80% using Copilot week one; App Store submissions +24% YoY, +80% in Q1 2026) creates a judgment gap: more code ships with less review. Organizations face a strategic choice: redesign junior roles around specification and customer contact (IBM model) or accept a future seniority cliff.