Thanks to the debt-ceiling showdown, Republicans appear to have a fighting chance to add a work requirement for those receiving SNAP (food stamp) and Medicaid benefits. They should also turn their attention to another major federal program that fosters dependency and discourages work: housing assistance. Local experiments across the country suggest that requiring those in public housing or getting housing vouchers to enter the labor market has positive effects, including upward mobility.

“Public and other assisted housing” is not a minor program. Housing vouchers are the largest item in the $73 billion Housing and Urban Development budget. At $30 billion, vouchers are almost twice as large an “outlay” as cash welfare ($16 billion), which has its own work requirement (though blue states have been finding ways around it.)

Unlike Temporary Assistance to Needy Families, housing benefits come with no time limit: about 32,000 New York City public-housing residents have lived in the projects for more than 40 years. What’s more, housing-benefit rules include a strong disincentive to work or raise earnings. Public and assisted-housing residents pay 30 percent of their income in rent—which means that, as their income rises, so does their rent. No private tenant would sign a lease like that, but it’s the rule for some of the poorest Americans.

By contrast, a handful of public-housing authorities—including Chicago and Atlanta, two of the nation’s biggest—have had work requirements for more than a decade. They’ve been among a small group of local public-housing agencies (39 of nearly 3,000) included in Moving to Work, a Clinton-era initiative.

The work-requirement experiment has been closely evaluated, including by the left-leaning Urban Institute. In Chicago, which exempted the elderly and disabled from the program and focused on public-housing residents rather than housing-voucher recipients, more than half of project residents (51 percent) had “no wage income” in 2010; by 2017, that proportion had declined to 38 percent. (Job training and other “good faith efforts” also satisfied the work requirement.) Average annual incomes rose from $11,568 in 2010 to $14,205 in 2015. Housing Authority staff interviewed for the report were enthusiastic, saying that, in their experience, “People want to do better. They want to work. They want money. They want to buy their children things.” The staff also noted that residents were not just working but, over time, getting better-paying jobs.

Work rules also resulted in higher rent payments, though that could also be seen as a negative, since it would be better for residents to keep more of what they earn.

Enforcement is serious. Residents looking to continue in assisted-housing programs must show pay stubs. Officials contact those known to be “noncompliant” weekly and offer help in finding work. Residents covered by the policy are given 90 days to comply or face the possibility of eviction (though that is unlikely).

The effects, by the standards of social science research, must be viewed as stunningly positive, as summarized in the bland language of the Urban Institute report:

The share of residents working more than 25 hours a week increased after the agency began enforcing its work requirement policy. The report found no increase in evictions and a modest increase in the rate of positive move outs because of gains in income attributed to compliance with the work requirement policy. A subsequent study that examined self-reported health and wellbeing outcomes found mixed effects associated with the agency’s work requirement policy; it found that residents wanted to work, and that increases in income decreased stress.

“Positive move outs” mean leaving the projects. Upward mobility means that those stuck on waiting lists may get a place—but be subject to the work requirement. This approach and others promise to change the culture of housing assistance.

Moving to Work has expanded from 39 housing authorities to 126. They will have good examples other than Chicago’s to emulate. In Atlanta, as I’ve written, high-rises were demolished and tenants were “vouchered-out” with a work requirement; labor-force participation rose from 18 percent to 62 percent. In San Bernadino, California, a five-year time limit has yielded strong results, according to an evaluation by Loma Linda University. Even without an explicit work requirement (though the city has begun to experiment with one), the imperative to prepare for an end to assistance led to a 26 percent increase in employment and a 145 percent increase in earned income after the five-year period. Education levels rose, too. All this will be news to Representative Pete Aguilar, who represents San Bernadino; the congressman had said that the work requirement for food stamps would “take food out of the mouths of kids.”

In insisting on linking work requirements to the social safety net, Kevin McCarthy and House Republicans are on to something—namely, that work requirements work.