The Proximity Tax
When a Fortune 500 tech company analyzed feedback patterns among its software engineers before the pandemic, it discovered something nobody had bothered to measure: employees sitting near their colleagues received approximately 20% more feedback than those working at a distance, according to research from the Federal Reserve Bank of New York. This wasn't a feature of the company's training program. It was the training program, operating invisibly through physical proximity. Then the pandemic arrived, remote work grew fourfold, and feedback to young workers plummeted.
The company's response revealed what corporate America had been hiding in plain sight. Instead of building systems to replicate that feedback remotely, the firm simply stopped hiring recent college graduates. It shifted to hiring workers roughly a decade older on average, per the Federal Reserve analysis. The invisible infrastructure of corporate training, it turned out, had always been accidental. When proximity disappeared, so did the only onboarding process that ever existed.
This wasn't an isolated failure. The unemployment rate among college graduates under age 29 rose 20% after the pandemic, comparing pre-pandemic levels from 2017 through 2019 with post-pandemic rates from 2022 through 2024, according to the Federal Reserve Bank of New York. During that same period, unemployment among older college graduates fell slightly. By March 2026, the unemployment rate for young college graduates reached 5.6%, up from 3.6% in March 2019, nearly a 60% increase in the rate itself.
The Remotability Trap
The pattern becomes stark when you separate jobs by whether they can be done from home. Emma Harrington, an assistant professor of economics at the University of Virginia who co-authored the Federal Reserve report, found that remote work explains 64% of the recent increase in unemployment among young college graduates. The research paired federal employment data with proprietary information from the Fortune 500 tech company to isolate the mechanism.
In jobs classified as remotable, the unemployment rate for younger college graduates jumped by almost a full percentage point after the pandemic, according to the Federal Reserve analysis. For older graduates in those same remotable positions, unemployment fell marginally. The gap in joblessness between recent graduates and older workers grew significantly wider in remotable jobs than in positions requiring physical presence.
This timing matters. The surge in youth unemployment predates the workplace arrival of generative AI, suggesting that remote work itself, not technological displacement broadly, explains the bulk of the rise, per the Federal Reserve research. The jobs young people can't get are precisely the ones they could do from anywhere. That's not a coincidence. It's a filter.
What Training Always Was
The Fortune 500 case study exposes the mechanism. Harrington and her co-authors found that employers are reluctant to place recent college graduates in remote settings where it is harder to absorb lessons from coworkers. But "reluctant" undersells it. The data shows companies didn't just prefer in-person training. They had no alternative to offer because they'd never formalized the knowledge transfer that happened when junior employees sat near senior ones.
That 20% feedback gap when workers sat apart, measured before anyone had heard of COVID-19, wasn't a bug in the system. It was evidence that proximity was the system. Corporate training programs, with their orientation schedules and learning management systems, were window dressing. The actual work of turning graduates into productive employees happened through hundreds of micro-interactions: overhearing a phone call, watching someone troubleshoot a problem, asking a quick question without scheduling a video meeting.
When the Fortune 500 company eventually implemented a return-to-office policy, it resumed hiring new graduates, according to the Federal Reserve research. The company hadn't suddenly developed better remote training tools. It had simply restored the physical proximity that made training possible. The ladder was back, but only for those willing to climb it in person.
The Preference Paradox
Here's the cruel irony: 71% of Gen Z workers prefer a hybrid work arrangement, and only 6% want fully on-site work, according to a Gallup survey from May 2025. The generation that grew up digital, that was promised flexibility as the future of work, now faces a labor market where that flexibility has become a barrier. They want remote work. Remote work is why they can't get hired.
This mirrors a pattern visible in other systems where upward mobility metrics looked fine while the actual ladder broke for younger generations. California's recent housing crisis showed the same dynamic: incomes rose, but homeownership rates for young people collapsed because the informal mechanism that once made housing accessible, a functioning market with reasonable price-to-income ratios, had disappeared. In both cases, the infrastructure was never intentional. It was emergent, dependent on conditions that changed.
The difference is that companies can choose to rebuild the training infrastructure they never knew they had. The Fortune 500 firm proved it by returning to the office and resuming graduate hiring. But that choice reveals the stakes: either formalize the knowledge transfer that proximity once provided, or accept that entry-level employment now requires physical presence while experienced workers enjoy flexibility. One generation gets the ladder. The next gets locked out.
When Guardrails Were Always Just Proximity
This isn't really a story about remote work. It's about what happens when we discover that systems we thought were designed were actually just accidents of physical space. Corporate training joins a growing list: institutional norms that protected Federal Reserve independence, housing markets that once enabled broad homeownership, career ladders that seemed sturdy until tested. All of them functioned through informal mechanisms that looked like intentional design.
The unemployment gap between young and old college graduates in remotable jobs now stands as a measurement of something we never built. Not a training system. Not an onboarding process. Just proximity, mistaken for infrastructure, until the day everyone went home and the scaffolding disappeared. What's left is a generation of credentialed graduates and a corporate America that can't figure out how to teach them anything through a screen.
The question isn't whether remote work is good or bad. It's whether we're willing to replace the invisible systems that proximity provided with actual infrastructure, or whether we'll simply reserve flexibility for those who already made it up a ladder that no longer exists for anyone else. Right now, the data suggests we've chosen the latter. The Fortune 500 company that restored office work and resumed graduate hiring didn't solve the problem. It just opted out of it.
