In their latest “farm bill,” Republicans on the House Agriculture Committee have proposed a simple rule for the food-stamp program: If you’re an able-bodied adult between the ages of 18 and 59, and have no children under six, you must either work or participate in a work program to get benefits. This is stricter than the two-tiered system we have today, in which most nonworking adults have to register for work and accept a job if offered one, and those under 50 with no dependents are limited to three months of aid if they don’t find a job or join a program.
And most important, the bill would give states additional funding to ensure that everyone who wanted a spot in a work program could get one, something the current system doesn’t achieve. “Only an individual who chooses not to participate in a guaranteed [employment and training] slot will lose eligibility,” wrote House Agriculture Committee chairman Mike Conaway in a recent USA Today op-ed (co-authored with Lee Bowes of America Works). States could exempt up to 15 percent of their caseloads to reflect barriers to employment, and the changes would not apply to the disabled or children, including children whose parents fail to meet the new work rules. Nonetheless, the reform is expected to double the number of food-stamp recipients subject to serious work requirements, to about 5 to 7 million.
This is a laudable approach, even if the bill is highly flawed in some particulars (more about which in a bit). It’s in line with American values: It offers help to the poor but insists that they make an effort to support themselves as well. And it’s part of a broader movement by Republican officials to make sure that the safety net promotes work for those capable of it. Some call it “Welfare Reform 2.0.”
Welfare reform’s first iteration is more than 20 years old — and it has worked. The 1996 law put pressure on states to reduce the number of people receiving cash welfare, and to increase work rates among people who do receive it, by enforcing work requirements and time limits.
The states’ implementation of the law left much to be desired. But at a very basic level it made work much more attractive than welfare, especially when combined with other pro-work benefits that expanded around the same time, including the earned-income tax credit. Overall spending on poverty programs actually continued its decades-long rise, but the able-bodied had to work to earn taxpayer support.
Single mothers — cash welfare’s main beneficiaries — flooded into the work force in unprecedented numbers. Poverty rates fell for these families and remained lower than their previous levels even during the Great Recession.
The major question is to what extent these lessons can apply to other poverty programs. Food stamps are an especially ripe target, because the program has expanded dramatically over the past two decades, and not just while the economy was in crisis. In 1998 enrollment stood at 19.8 million — a number that swelled by a third, to 26.3 million, by 2007. Enrollment peaked after the recession, in 2013, at 47.6 million; in 2017, four years of recovery later, it had fallen barely more than 10 percent, to 42.2 million. That last number amounts to one in eight Americans.
The increase results, in part, from a combination of loose standards and states’ creativity in finding loopholes in the law. Geographic areas can get work-requirement waivers if they have unemployment rates sufficiently above the national average, even if the economy is booming everywhere. “States have been able to essentially gerrymander areas by unemployment rate,” a House Agriculture Committee aide tells me. “And some entire states are excluded.” Many states also make some of their residents “categorically eligible” for food stamps by giving them small benefits, often just a brochure, through another program.
Further, there is evidence that many of the able-bodied adults on the program would leave if simply asked to find work or engage in work-related activities for a reasonable amount of time, whether because they found jobs or because they had other sources of income and didn’t really need the help. When Maine introduced a work requirement for able-bodied, childless, adult food-stamp recipients, it saw an enrollment drop of 80 percent among this population — even though enrollees could satisfy the requirement with just six hours a week of community service. Other states and counties have seen similar results when their waivers ended.
Broadly speaking, the House bill is a decent opening bid on these issues. In addition to the new work requirements, it cracks down on “categorical eligibility” and narrows the scope of geographic waivers. Today 30 percent of the population lives in a waived area; that would fall to about 12 percent under the new legislation, according to the House Agriculture Committee. The Congressional Budget Office estimates the bill would reduce enrollment by about a million over ten years.
But it’s come in for legitimate criticism from both the right and the left. Heritage Foundation senior fellow Robert Rector, for example, paints the current draft as simultaneously too weak and too harsh. He doesn’t believe the geographic-waiver provisions will be as effective as the committee claims, and points out that some aspects of the work requirements simply don’t add up. An able-bodied adult might get $200 a month in food stamps, while the bill would require him to spend 20 to 25 hours a week at a training program to get the benefit. “It was easy defending the Maine program, because I could say it only asked for six hours of community service,” he says. “But it has to be proportionate. The idea here is not to punish these people.”
Further, those who failed to meet the requirements would be kicked off the program for an entire year the first time and for three years for subsequent failures. “That’s exactly what you don’t want to do,” Rector says. “You want to have a work program where it’s very firm but it’s very forgiving: If you didn’t do what you were supposed to do last month, okay, we’re not going to give you the benefit, but if you want to do the right thing this month we’ll put you back on the rolls.” Rector adds that the bill does nothing to address the program’s marriage penalty and indeed imposes separate work requirements for spouses.
Meanwhile, the Left has raised questions as to how far the additional funding for employment programs will go — whether it can provide decent training or just low-cost options such as a supervised job search — and how well the states will implement those programs, especially given their record with the 1996 reform. (States receive block grants to fund their welfare and work programs but manage to redirect much of the money to other uses; they also have gamed the law to escape its requirement that they engage a certain percentage of welfare recipients in work or work-related activities.) At $1 billion per year, the funds would amount to roughly $170 to $200 per person covered by the requirement, not all of whom would enroll.
The bill needs to be improved, certainly. But moving able-bodied adults from welfare to work is worth the effort — and yet the Senate side of the GOP shows little indication of being willing to pursue this project, especially as part of the farm bill. Senator Pat Roberts, who heads the Senate Agriculture Committee, vows that “this will be an evolutionary — not revolutionary — farm bill in the Senate,” has said that any changes to the food-stamp program will be minor, and has publicly agreed with Senate Democrats to pursue a bipartisan bill.
One can’t fault politicians for worrying about the optics of tightening restrictions on food-stamp recipients so soon after slashing the corporate-tax rate — in an election year to boot, and in a bill famously larded with pork for agricultural interests. In addition, the bill will require 60 votes in the upper chamber to pass, meaning Democratic votes will be needed even if Republicans are unanimous in their support.
It would be possible, of course, for Congress to pursue reforms outside the farm bill, though it’s unclear whether the closely divided Senate has any appetite for that, either. Ideas abound, from fixing the state-level implementation problems that still plague the previous round of welfare reform, to instituting work requirements in public housing, to reworking the earned-income tax credit to cut down on fraud and get aid to those most in need.
But should Congress fail, we will be limited to what the executive can accomplish by itself. In January the Trump administration announced it would grant waivers for states to experiment with work requirements for Medicaid; the next month, it hinted that it was exploring options to make it harder for states to get geographical waivers from work requirements for food stamps. And in April it ordered an across-the-board review of work requirements in safety-net programs, with an eye toward strengthening them within the limits of the law.
Medicaid work requirements are particularly contentious, drawing some criticism even from the right. Rector has pushed back against the idea of instituting them, pointing out the political difficulty of cutting off people’s health insurance as well as the practical difficulty of enforcing such requirements when medical providers often prospectively sign up new Medicaid beneficiaries on the spot when they show up for care.
But others are more supportive. “The way I look at it is this: Say someone comes in, and they’re not on disability and have no other income, and they just want Medicaid. The agency should be concerned about that — you can’t eat Medicaid or pay your rent with it, and frankly work is more likely to lead to better health outcomes,” says Robert Doar, the Morgridge Fellow in Poverty Studies at the American Enterprise Institute, who in his previous job administered twelve public-assistance programs in New York City under then-mayor Michael Bloomberg. “With a program providing such a significant amount of assistance, the agency should not be allowed to say, ‘That’s not my job.’”
Doar notes that only a few states are trying waivers at the moment — experiments that can inform practice in the rest of the country — and that individuals who don’t get Medicaid will still be treated at emergency rooms. He also emphasizes that work requirements can be effective even when cash isn’t at stake: “Most people come in and say, ‘What do I need to do to get this?’”
Meanwhile, it’s uncertain what will come of the top-to-bottom review that Trump has ordered, a process that could affect numerous different programs, from housing benefits to traditional cash welfare, depending on how aggressive agencies are willing to be. But the order’s biggest function might be to send a strong message to executive agencies as to how they should go about their daily business.
“It’s a big change from the previous regime under Obama, and even a little under George W. Bush,” says Doar. “Work was secondary or even ignored — mostly ignored, in fact. Now they’re being told, ‘That’s not what we’re doing anymore. The measure of success isn’t how much you relieve poverty by giving people benefits, but how much you relieve poverty by getting them into work.’”
And whether such changes are pursued through legislation or the executive branch, it will be a tough slog in the face of hysterical accusations of heartlessness. “Republicans have to have a thick skin and keep their wits about them. But the public thinks people on welfare should work, as long as the requirements aren’t punitive,” says Ron Haskins, an architect of the 1996 reform who is now a senior fellow at the Brookings Institution. “They’re with the Republicans on this: Requiring people to work is a good thing, not a bad thing.”