Five years on, the COVID-19 pandemic is best understood not as a temporary disruption to work but as the moment a set of changes already underway compressed from a decade-long arc into about eighteen months. The pandemic did not invent remote work, e-commerce logistics, or worker leverage. It accelerated each of them past the inflection point at which they became default rather than exception. The honest analytical question is which of those accelerations stuck — and which the labor market is, slowly, reverting back from.
Three changes look durable. Three look like they are unwinding. Sorting the two categories matters more than another retrospective about Zoom fatigue.
What stuck: remote work, worker leverage in services, and labor-shortage-driven wage compression
The most durable change is the existence of remote work as a category. Stanford economist Nick Bloom's WFH Research project, which has surveyed U.S. workers monthly since 2020, finds that roughly 30% of full paid workdays were performed from home in 2024 — down from a 2020 peak above 60%, but stable above 25% for the past three years. That 30% number is essentially flat across recent surveys, suggesting the system has settled at a new equilibrium that is roughly 5–7 times the pre-pandemic baseline.
The wage effect is also durable. Bureau of Labor Statistics Current Employment Statistics data show that the largest real-wage gains during the 2021–2023 labor-market tightness accrued to workers in the lowest wage quartile — what the Realtime Inequality Tracker maintained by Emmanuel Saez and colleagues at Berkeley documented as the first significant compression of the U.S. wage distribution in roughly four decades. The pandemic-era leverage of low-wage service workers shifted relative pay in ways that have not reversed even as the labor market has cooled.
The third durable change is the structural acceleration of e-commerce and home-delivery logistics. U.S. Census Bureau e-commerce data show online sales settling around 16% of total retail in 2024, up from 11% in 2019 — a permanent step change. The supporting warehouse and last-mile employment base grew accordingly, and the AI-and-robotics investment cycle in those sectors is downstream of that base.
What is unwinding: total flexibility, the strongest worker leverage, and pandemic-era benefit largesse
The reversals are real and worth naming honestly. Return-to-office mandates have intensified through 2023–2025; the WFH Research project's tracking shows that the share of fully remote workers in office-job categories has declined, with most large employers settling on three-day in-office hybrid policies. Amazon's January 2025 five-day in-office mandate is the most visible recent inflection.
Worker leverage has also softened. The "Great Resignation" peaked in late 2021–early 2022, with monthly quits rates reaching 3.0% per BLS Job Openings and Labor Turnover Survey data. By 2024 the quits rate had fallen back to roughly 2.0%, close to the pre-pandemic norm. Pay growth, while still real, has decelerated. The employee-friendly conditions of 2021–2022 are not the conditions of 2025.
The pandemic-era benefits surge — expanded mental-health programs, "Netflix accounts as benefits"-style perks, paid social-justice leave — has largely been quietly trimmed. The benefits that survived the cuts were the ones with measurable productivity or retention returns; the ones that read as PR have mostly been retired.
The most consequential shift that hasn't been priced in yet
The change with the largest long-run implications is one that almost nobody talks about directly: the permanent normalization of working from home moved the locus of labor-market matching from local geography to a wider radius. The implications for housing markets, regional inequality, and city tax bases are still working themselves out. Brookings Institution research on the "geography of remote work" has documented that high-skill knowledge workers have dispersed from the most expensive metros, with measurable population gains in mid-size cities and the urban hinterlands of Sun Belt metros. Whether the dispersal continues or reverses will determine more about the next decade's urban policy than any zoning debate.
The OECD's Employment Outlook 2023 identified a related second-order effect: the rise of cross-border remote work, where a worker in one country is employed by a firm in another, creates tax, social-insurance, and labor-rights questions that no developed economy has resolved cleanly. The EU Platform Work Directive of 2024 is the first major attempt at a framework; the U.S. has not produced an equivalent.
The harder lesson for managers
The pandemic also exposed a set of management practices that had been protected by inertia and proximity rather than by their actual usefulness. Microsoft's Work Trend Index research, drawing on telemetry from hundreds of millions of users, found that the "productivity paranoia" gap between manager perceptions and worker self-reports widened during the pandemic — managers consistently believed productivity had dropped while workers reported it had risen — and that the gap correlated with employer surveillance investment. Companies that responded with more dashboards generally saw declining engagement; companies that responded with clearer goals and asynchronous documentation generally saw the opposite.
That distinction connects directly to NWLB's research on Productivity Theatre →, which argues that much of what white-collar work measures as productivity is actually visibility — and that the pandemic made the difference between the two newly legible.
The pandemic was not the cause; it was the catalyst
The most useful way to read this period is that it accelerated decisions employers and workers had been deferring. Remote work was technically feasible for many roles before 2020; what was missing was the social permission to take it seriously. Pay compression for the lowest-wage workers was structurally overdue; the labor shortage created the condition for it. The reckoning with management practices that depended on co-located visibility was always going to come; the pandemic forced it.
What did not happen, despite the breathless commentary at the time, is a fundamental reordering of capital and labor. Worker leverage rose and then partially receded. Hours did not collapse. The five-day workweek survived. The U.S. did not adopt portable benefits, sectoral bargaining, or any of the structural reforms the 2020–2021 moment briefly suggested. The world of work was not turned upside down. It was nudged onto a new track — and the next decade will be about which of those nudges become permanent.
The pandemic did not invent the future of work. It compressed five years of overdue decisions into eighteen months and then quietly handed managers back most of the leverage workers briefly had.
Updated May 21, 2026. This piece was substantively rewritten as part of NWLB's 2026 editorial refresh.
