Why Millions of American Dads Stopped Working — and Can’t Start Again

In 1954, 98 out of every 100 American men between 25 and 54 were either working or looking for work. Today, about 11 out of every 100 aren’t working and aren’t even looking. That’s roughly 7 million men — enough to fill America’s third-largest city. The unemployment rate doesn’t count them. They’ve dropped off the map entirely.

Maybe your dad is one of them. Maybe it’s your brother, your partner, or you. Maybe he’s home right now, and you’re not entirely sure what went wrong, or when, or whether it was even his fault.

This essay is about what the data actually shows — and it’s more complicated and more troubling than anyone on cable news wants to admit.

A note on terms: throughout this essay, “prime-age men” means U.S.-born men aged 25–54 unless otherwise noted. That’s the demographic economists consider the core of the workforce — past school age, before retirement, in the years when work is most expected.

What We Know For Sure

These facts come from federal data: the Bureau of Labor Statistics, Census surveys, and peer-reviewed research. None of them are controversial among economists.

The decline is real and enormous. The labor force participation rate for prime-age men has been falling since the mid-1960s. In April 2025, 11.2 percent of U.S.-born prime-age men were not in the labor force — nearly triple the rate in 1960. The political economist Nicholas Eberstadt, who wrote the landmark book Men Without Work, documented that the work rate for American men aged 25–54 was actually lower in 2015 than it had been in 1940 — before World War II, at the tail end of the Great Depression. An important difference: in 1940, most non-working men were actively looking for jobs. Today’s non-participants have mostly stopped looking altogether. That’s a fundamentally different phenomenon.

It’s concentrated among men without college degrees. Of the roughly 18 million working-age U.S.-born men not in the labor force as of 2025, 87 percent did not have a bachelor’s degree. Among prime-age men specifically, 81 percent of those outside the labor market lacked a four-year degree.

Health and pain are major factors. A 2024 Bipartisan Policy Center survey found that 57 percent of prime-age men not in the labor force cited physical or mental health as their main reason, with 55 percent specifically pointing to disability, serious illness, or receiving disability benefits. Princeton economist Alan Krueger found that nearly half of these men take pain medication every day, and about two-thirds of those take prescription painkillers. Forty percent say pain prevents them from working full-time on jobs for which they’re otherwise qualified.

Criminal records create a wall. A 2022 RAND Corporation study found that by age 35, 64 percent of unemployed men had been arrested as adults and 46 percent had been convicted of a crime. A criminal record cuts your chance of getting a job callback by roughly 50 percent. For Black men, the penalty is even steeper — about 60 percent. The U.S. has catalogued more than 38,000 laws that impose consequences on people with criminal convictions, and over 80 percent of those statutes restrict employment.

The jobs that used to absorb these men vanished. Manufacturing, mining, and construction historically employed men without college degrees at relatively high wages, partly because of union protections. The share of manufacturing jobs has been declining for decades. A 2025 Federal Reserve Bank of San Francisco study found that many manufacturing workers who lost their jobs simply never transitioned to other industries.

The Pattern That Should Bother You

Each of those facts by itself has a reasonable, straightforward explanation. People get hurt, people make mistakes, industries change. But when you look at how they interact, something darker comes into focus.

Imagine a man — it might be your father, your uncle, your neighbor, or the guy you were in high school with. He didn’t go to college. He worked at a factory or drove a truck or did construction. At some point, one of several things happened: he hurt his back, or he got arrested for something he did when he was 22, or his plant closed, or — increasingly common — all three.

Now he’s in a system that looks, from the outside, like it’s designed to help him. He can apply for disability benefits. He can get Medicaid. He’s technically “supported.” But from inside, the system works like a trap. On paper, it looks like a net stretched beneath a falling person. In practice, the mesh is so sticky that once you hit it, you don’t climb back up.

Once you go on disability, you almost never come off. Brookings Institution research confirms this as a well-documented pattern. For many men, disability programs are not the initial cause of their labor force exit — but they often become the mechanism that makes exit permanent. The benefits are enough to survive on but not enough to live on, and they come with rules that punish you for earning too much. If you try to work part-time, you risk losing your health insurance — the one thing keeping you alive if you’ve got chronic pain or a health condition. Economists call these “welfare cliffs.” They mean that for many people, taking a low-wage job actually makes you poorer than not working.

Meanwhile, his criminal record follows him everywhere. Even if he was arrested and never convicted, it shows up on a background check. Even if his conviction was 15 years ago for something nonviolent. Even if he’s been clean ever since. Employers see it and move on to the next application. Research by Devah Pager showed this wasn’t speculation — when job applicants with identical qualifications applied for the same jobs, those who disclosed a criminal record were called back about half as often.

And the pain medication? Krueger’s research showed that doctors prescribed opioids more freely in some regions than others, not because patients in those areas were in more pain, but because of different medical practice norms. In counties where more opioids were prescribed, more men dropped out of the labor force. The association is strong enough that many economists now treat opioid overprescribing as a labor-market shock in its own right. But the decline in male participation predates the opioid crisis by decades — opioids didn’t start the trend. They accelerated it, and then locked men into it. The pills didn’t solve the pain. They created a new dependency on top of the old one.

The simple version: each of these things pushes men out of the workforce, and once you’re out, the system’s own rules make it extraordinarily difficult to get back in.

Alternative Explanations — Honestly Considered

Before offering any systemic explanation, we have to take seriously the simpler ones people offer. If these explain the problem, then the structural story is unnecessary.

“They just don’t want to work.” This is the most common dismissal. And for some men, it contains a grain of truth — Eberstadt documented that men outside the labor force spend significantly more time on screens and less time on community activity than employed men. But think about what that actually tells you. If you take away status, structure, and a realistic path back in, of course many men retreat into low-friction digital worlds. That isn’t proof they never wanted to work. It’s evidence of what people do when meaning and opportunity evaporate. And the survey data complicates the laziness story further: 57 percent report health barriers. Many would work if they could do so without losing the fragile support system keeping them alive.

“It’s just automation — these jobs are gone forever.” Partly true. Manufacturing decline is real. But automation doesn’t explain why men in these communities also have higher opioid prescription rates, higher criminal record rates, and worse health outcomes. The jobs vanished, and then nothing replaced them — not just economically, but socially. Other countries that experienced similar manufacturing decline — Germany, for instance — invested heavily in apprenticeships, sectoral training, and short-term work schemes. They saw far less catastrophic drops in male employment. The question isn’t whether automation played a role (it did), but whether the policy response to automation made things worse. The evidence suggests it did.

“Disability benefits are too generous.” They’re not. Average Social Security Disability payments are modest — not enough to live comfortably. But the design of benefits creates perverse incentives: the cliff effects that punish partial work, the risk of losing health coverage, the all-or-nothing structure. The problem isn’t generosity. It’s architecture.

“Why can’t they just drive for Uber?” The gig economy doesn’t absorb this population the way people imagine. Many of these men have criminal records that disqualify them from ride-share and delivery platforms. Many don’t have reliable vehicles or smartphones. And gig income triggers the same benefit cliffs — earn a few hundred dollars and risk losing your Medicaid. The gig economy was designed for people who already have a floor beneath them, not for people standing on one.

Each of these explanations captures a piece of the picture. None of them captures the whole thing. And none of them lets the institutions off the hook for designing a system where falling out of the workforce is easy and climbing back in is nearly impossible.

Why It Looks Different Depending on Where You Sit

Now that we’ve ruled out the simple stories, something harder to see comes into focus.

If you’re a policymaker or a bureaucrat managing disability programs, this situation looks like a coordination problem that’s being handled. People who can’t work are receiving benefits. The system is functioning. The numbers balance. From this angle, the whole arrangement looks like a set of rules doing their job.

If you’re the man living inside this system — the one taking pills for his back, with a record from his twenties, watching his kid study for the SAT while he sits at home — it looks like a trap. He’s not being helped. He’s being managed. His human potential has been written off, and the system is structured to keep him exactly where he is.

Both perspectives are looking at the same facts. The difference is what you stand to lose. When the cost of failure falls entirely on the person at the bottom, everyone above them has the luxury of calling the system “neutral” or “working as intended.” The institution loses nothing from the status quo. The man loses everything.

This gap — between how a system looks from the top and how it feels from the bottom — is the core of the problem. It’s why nobody with power feels urgency. The system is working for them. It is failing him.

What This Means For You

If your father doesn’t work, you’ve probably absorbed some version of what society thinks about him. That he’s lazy. That he made bad choices. That he just doesn’t care enough. Maybe some of those things are partly true — people are complicated, and nobody’s story is entirely clean.

But the data says something else is also true: the system he’s inside was not designed to get him back on his feet. It was designed — not maliciously, but through decades of fragmented policymaking — to manage him. Disability programs that were built to provide income support ended up trapping people. Criminal records that were meant to protect public safety became permanent barriers long after any risk had passed. Pain medication that was prescribed too freely turned into a crisis. Jobs that disappeared without any plan for what happens to the people who relied on them.

None of this is your fault. None of it is entirely his fault either. It is the result of institutional choices, made over decades, that prioritized managing a problem over solving it.

What Would Actually Help

Regardless of which explanation carries the most weight, certain actions would address the problem. None of them require agreement on who’s to blame.

Redesign benefit cliffs — so that the man who tries a part-time warehouse job doesn’t have to gamble his pain meds and his kid’s health coverage on a scheduling experiment. Allow gradual transitions back to work without catastrophic loss of support. Several states have piloted programs that let disability recipients earn income without immediately losing their benefits, and early results are promising. This requires no new philosophy — just better engineering.

Reform criminal record barriers — so that the guy who shoplifted at 22 isn’t still unemployable at 42 in a completely different life. Automatic record sealing for nonviolent offenses after a set period, like Pennsylvania’s Clean Slate law, removes a barrier that research consistently shows is disproportionate to any public safety benefit. Economists at RAND have noted that most employers dramatically overestimate the risk posed by applicants with old records.

Invest in health infrastructure in affected communities — so that the man whose back went out at 35 has options beyond a bottle of pills. The opioid crisis was partly a failure of medical practice, not just individual choice. Communities with high non-participation rates need pain management alternatives, mental health services, and flexible workplace health benefits.

Create transitional employment for displaced workers — so that when a plant closes, the workers don’t just get a pamphlet about “retraining.” They need actual jobs, immediately, that bridge the gap while they transition. Germany and Denmark do this. The United States mostly does not.

The Question No One Will Answer

Here is the question that hangs over all of this, and that no institution with the power to answer it has been willing to address:

If roughly 7 million men in the prime of their lives are neither working nor looking for work, and the systems meant to support them are structured in ways that keep them from re-entering the workforce — why are the costs of that arrangement borne almost entirely by the men and families at the bottom?

Not by the disability bureaucracies, which face no accountability for whether anyone ever returns to work — their incentive is to avoid approving someone who might later be judged “not disabled,” not to help anyone re-enter the workforce. Not by the pharmaceutical companies, which profited from prescribing patterns that rewarded volume over outcomes. Not by the employers, who benefit from a surplus labor pool that suppresses wage pressure. These aren’t conspiracies. They’re incentives — incentives that reward short-term cost control and risk avoidance, even when the long-term human cost is enormous.

A safety net worthy of the name would take on some of that risk itself, instead of offloading it onto the very people it claims to protect.

And the real cost isn’t only economic. It’s generational. When millions of fathers disappear from the world of work, their children inherit the consequences — not just the poverty, but the message that the system has no place for people like them.

Your dad’s situation is not a personal failure. It’s a system working exactly as designed — just not for him. And not for you.

Notes on Evidence: The facts in this essay are drawn from federal sources (BLS Current Population Survey), peer-reviewed research (RAND Corporation/Science Advances, Brookings Institution working papers), and national surveys (BPC–Artemis, 2023). Key studies: Nicholas Eberstadt, Men Without Work (2016, updated 2022); Alan Krueger, “Where Have All the Workers Gone?” (Brookings, 2017); Bushway et al., “Barred from Employment” (Science Advances, 2022); Devah Pager, audit studies on criminal records and employment; Bengali et al., “Pulled Out or Pushed Out?” (SF Fed Working Paper 2025-07). Where this essay draws inferences — connecting benefit cliffs, criminal records, and opioid dependency into a self-reinforcing cycle — those connections are marked as inferences, not documented as a single integrated system. Where it proposes structural hypotheses (e.g., U.S. vs. European safety-net designs), those are flagged as hypotheses. This analysis uses a Deferential Realism interpretive framework — an analytical model, not empirical measurement. The documented facts stand independently of the framework.