In May I wrote that America’s War on Poverty may have helped create a permanent underclass — not because the intentions were evil, but because transfers replaced market income and left too many people dependent on the state instead of work.

The post kept getting visits. Nobody left a sermon in the comments. Fair enough. The question still stands: has anything happened since May 11 that actually strengthens the case?

Short answer: no generation got trapped between May and September. Anyone selling that is selling junk. What did happen is smaller, and more interesting. A few numbers moved. Washington started treating the dependency argument as something you legislate, not just debate. That is not proof. It is not nothing either.

Start with work, because that is the whole point of the proverb.

U.S. labor force participation fell to 61.5 percent in June 2026 — the lowest reading outside the pandemic since 1976. It slipped again in July, to 61.4 percent. Part of that drop is boring: a Bureau of Labor Statistics population revision and an aging country. It is not only boring. Participation among prime-age adults slipped too.

A July report from the Center for Immigration Studies put the longer trend in ugly English. The share of U.S.-born men aged 16 to 64 who are not in the labor force was 11.3 percent in April 1960, 16.9 percent in April 2000, and 22.2 percent in April 2026. Households where nobody works collect SNAP, TANF, SSI, Medicaid, and housing aid at far higher rates than households with at least one worker. Poverty among non-working adults is not a mystery. Work is still the best anti-poverty program we have. We just keep paying people as if it were optional.

Then look at the program that was supposed to be “temporary.”

A June 2026 brief from the Administration for Children and Families found that only 8 percent of TANF families had enough hours in real, unsubsidized jobs to count under the federal work measure. More than half of families with a work-eligible adult reported zero hours. On paper, states face a 50 percent work-participation target. In practice, the caseload reduction credit let 39 states face a target of zero in FY2024. Transfers continue. Exit does not. That is the Corinth–Burkhauser pattern in one paragraph: you can push people over a poverty line without making them independent.

Congress has started acting as if that pattern is real.

The 2025 reconciliation law tightened SNAP work rules. Enrollment has fallen by roughly five to six million since last summer, from about 42 million to about 36.6 million by May 2026. Medicaid is getting an 80-hour-a-month “community engagement” requirement, with states required to implement it by 2027 and a few moving earlier. This is not a new academic paper. It is the political system admitting that work disincentives are not a pamphlet invention.

Now the part I am not allowed to skip, because I told you I would be honest.

The Corinth and Burkhauser study runs from 1939 to 2023. Summer 2026 does not rewrite it. A July Census Bureau working paper on SNAP work requirements found that the rules cut caseloads without raising labor supply, and that they disproportionately screen out the poorest. So the naive slogan — impose requirements and the fish learn to fish — does not survive contact with the bureaucracy. The system is clumsy. Clumsy can punish the needy without producing workers.

The last official Welfare Indicators series still showed a few percent of people getting more than half their income from TANF, SNAP, and SSI — 3.7 percent in 2023. Official child poverty improved in 2024, from 15.3 percent to 14.3 percent. The trap did not suddenly deepen in the four months after my post.

Europe is not a clean exhibit for the prosecution either. EU employment is at record highs. Poverty and social exclusion are a bit lower than in 2019, and still nowhere near the 2030 target. Finland cut benefits. The Labour Institute for Economic Research estimates almost 100,000 more people on low incomes than under the old rules. Less fish, more measured poverty in the short run. That does not prove that more benefits create a caste. It proves that taking the fish away hurts before anyone has learned to fish.

So what is left, if you strip out the sermon?

Material poverty can fall while dependency and weak work among people who could work remain. That was the narrow, defensible point of the original study. American policy has started tightening because of that point. Four months of data do not prove a permanent underclass. Sixty years of substituting paychecks with programs still deserve a harder look than a slogan.

We have spent trillions on fish. Teaching people to fish is slower, ruder, and easier to get wrong. It is still the only ending that does not require a permanent clerk in Washington.

The May post is here: The War on Poverty That May Have Created a Permanent Underclass. If four months changed your mind either way, say so — preferably with a number, not a sermon.