MIT Study: 57 Million American Workers Now Classified as 'Disposable'

MIT Sloan professor emeritus Paul Osterman published his findings this month in a book titled "Disposable Workers: The Transformation of Employment," out from Harvard University Press. The number driving headlines: 57 million workers, or roughly one in three, are what Osterman calls "disposable."

Osterman based the book on a survey of more than 6,000 civilian, nonagricultural adults conducted in late 2022, plus nearly 100 interviews with workers, employers and staffing agencies, according to MIT Sloan's own release. This is survey data collected almost four years ago, packaged into a book released now.

Three buckets, not one gig economy

Osterman breaks disposable work into three groups. Contractors employed by staffing firms but working at a client's site make up 13% of the workforce, according to his research as reported by Fortune and MIT Sloan. Think temp office workers, cleaners, security guards, and the contractors PepsiCo hires to taste-test Doritos for addictiveness.

Freelancers who work for organizations without being employees, like Uber drivers or freelance programmers, are 5% of the workforce. The biggest bucket is "marginal workers," 17% of the workforce, people technically on staff but never fully integrated: adjunct professors passed over for tenure, staff attorneys at law firms with no partner track, and large numbers of part-timers.

Osterman's own framing pushes back on a popular narrative. "When people talk about the gig economy, they're picturing Uber drivers and DoorDashers. But that's a small slice of this, less than two percent of the workforce," he said, according to MIT Sloan. His point: America isn't becoming a gig economy, it's becoming a "disposable" one, with far more people affected than app-based delivery work suggests.

Why employers do it

Osterman's research attributes the shift to straightforward cost-cutting. Contractors and part-timers let companies scale staff up or down without paying for training, health insurance or severance. Some employers, he found, deliberately structure part-time roles to stay under legal thresholds that would otherwise trigger paid leave or anti-discrimination protections.

If a company is engineering shift schedules specifically to dodge benefits law, that's a legal-threshold-gaming problem regulators and legislators can actually address.

Where coverage diverges

IBTimes ran with a global framing, describing the shift as one that "altered the employment landscape across the Atlantic," despite the underlying data being entirely U.S.-based. This is an American labor story, not a transatlantic one.

Naked Capitalism, which republished Osterman's original essay, added an editorial framing tying the findings to Amazon's subcontracting model in New York City and New Jersey, and argued the trend explains stagnant real wages. That's Osterman's data plus an added policy argument the professor himself didn't make in the excerpted material.

Parriva took the numbers and asked a narrower, useful question: are Latino workers disproportionately represented among the 57 million? The outlet correctly notes that Bureau of Labor Statistics contingent-worker categories are narrower than Osterman's definition, so a clean answer isn't available in the existing data.

The manufacturing counter-narrative

None of this squares neatly with parallel reporting on U.S. manufacturing. Breitbart's Business Digest highlighted that durable-goods manufacturing employment has risen every month since December 2025, with payrolls up 18,000 in July following gains of 12,000 in June and 14,000 in May, according to Labor Department data cited by Breitbart. The Labor Department also reported manufacturing output rose at a 4.6% annual rate in the second quarter, with productivity up 1.9%.

The Epoch Times' Steve Cortes pointed to factory and trade-related construction jobs rising from 9,300 added in the first quarter of 2025 to 93,000 added by the second quarter of 2026, framing it as evidence of a broader "America First" industrial revival.

Those manufacturing gains are real and sourced to Labor Department figures. But they don't cancel out Osterman's findings, because they measure different things. One is job creation in a specific sector; the other is job quality and security across the entire economy, based on 2022 survey data. A factory can add headcount while still classifying a growing share of those workers as contractors rather than permanent employees, and neither Breitbart nor the Epoch Times piece addresses whether the new manufacturing jobs are full, secure positions or staffing-agency placements.

What's unresolved

Osterman's survey is nearly four years old. No update has been published showing whether the 35% disposable-worker share has grown, shrunk, or held steady since 2022, including through the recent manufacturing hiring numbers. Whether Congress or state legislatures, following New York City and New Jersey's moves against Amazon's subcontracting model, will pursue federal rules on staffing-agency liability or part-time benefit thresholds remains an open question with no scheduled action reported.