All Unemployment Is Local

Why It Matters

The findings signal that AI‑enabled productivity gains are reshaping labor demand for educated workers, creating localized pockets of higher joblessness that could strain regional economies and policy responses.

Key Takeaways

• College‑educated metros see rising unemployment rates 2022‑2025.
• FIIPB sectors output up, employment flat, driven by AI.
• AI compresses routine cognitive tasks, reducing entry‑level jobs.
• Historical automation patterns suggest delayed job creation after disruption.
• Policy focus needed to smooth AI‑driven labor market transition.

Pulse Analysis

The latest Burning Glass Institute analysis overturns the conventional wisdom that metropolitan areas with dense college‑educated populations are immune to job loss. By comparing local unemployment percentiles from 2022 to 2025, Levanon reveals that once‑thriving hubs like San Jose and Boston have slipped into the middle of their historical ranges, while smaller metros with fewer degree holders remain relatively stable. This divergence reflects a broader shift in the knowledge economy, where AI and automation are no longer confined to manufacturing but are now targeting routine cognitive tasks traditionally performed by entry‑level professionals.

At the heart of the trend are the FIIPB sectors, which together account for over 40 percent of U.S. GDP. Since 2022, real output in finance, insurance, information, and professional services has continued to climb, yet employment in these industries has flattened or even declined. Levanon estimates that, absent AI‑driven efficiency gains, the economy would have added roughly three million knowledge‑industry jobs. This pattern mirrors classic skill‑biased technological change, where productivity advances initially displace workers before new, higher‑skill roles emerge, a cycle observed during past automation waves in manufacturing.

For policymakers and business leaders, the localized rise in college‑worker unemployment raises urgent questions about workforce development and safety‑net design. While the national unemployment rate hovers at a healthy 4.4 percent, the erosion of the degree premium in key metros could exacerbate regional inequality and slow overall economic dynamism. Targeted interventions—such as reskilling programs focused on AI‑augmented tasks, incentives for firms to create new knowledge‑intensive roles, and regional coordination of labor market data—will be essential to ensure that the productivity gains from AI translate into broad‑based employment growth rather than concentrated job loss.