House Republicans have proposed a massive expansion of an obscure job training program as a way to get millions of people off of food stamps — notching a welfare reform win as part of the farm bill.
There’s just one problem: There’s little evidence the training program actually works, let alone that it can be scaled up quickly to enroll hundreds of thousands, if not millions, of new participants.
House Speaker Paul Ryan, who has long sought to shrink the safety net, lauded the farm bill advanced last week by the House Agriculture Committee calling it “the precise thing we need to do to get people from welfare to work." Ryan said Thursday he sees the farm bill as part of “the final installment” of the House GOP agenda.
But standing up training capacity for as many as 3 million people so quickly would be roughly equivalent to building almost half the existing U.S. community college system from scratch, said Kermit Kaleba, director of federal policy at the National Skills Coalition.
"You could have a real paradigm shift in how we invest in low-income workers," said Kaleba, a big proponent of expanding training. "But in order to have that conversation, you have to have a realistic sense of scale.”
And that scale dwarfs what exists now.
The House bill would impose stricter work requirements on between 5 million and 7 million recipients of the Supplemental Nutrition Assistance Program, while also pouring more money into training and education programs — with the promise of guaranteeing enough slots in all 50 states for everyone who’s eligible. SNAP, still known to many as food stamps, was created to combat hunger amid American plenty and helps feed some 42 million people, or one in eight households.
Under the plan, the Department of Agriculture’s SNAP Employment & Training program — which is little-known even in food and agriculture policy circles — would explode from about $90 million in annual funding to $1 billion in annual funding over three years. States, which run their own training programs with USDA funding, would have two years to ramp up to handle all the new cases.
“It’s a great time, quite frankly, from the economy’s standpoint because new jobs are being created every day and wages are higher,” House Agriculture Chairman Mike Conaway (R-Texas) said recently as he unveiled the bill. “We think this is a great time to dedicate significant resources toward the E&T portion of the SNAP program."
Ordinarily, Democrats and anti-poverty advocates might cheer such a massive expansion of a program that serves low-income people. But even the most enthusiastic supporters think most states aren’t ready to take on millions of new trainees. They fear that requiring people to spend 20 hours a week working or preparing to work before training programs are built out might force millions off the rolls.
While there have been promising results in places like Washington state, SNAP training programs, which are mostly voluntary, haven’t been rigorously evaluated since 1994, according to USDA. Back then, researchers concluded they had no significant impact on participants’ wages, hours or job retention.
President Donald Trump tweeted on Monday that “since January 2017, the number of people forced to use food stamps is down 1.9 million. The American people are finally back to work!” That’s true — the number of Americans on SNAP has fallen every year since 2013 as the economy has slowly improved since the Great Recession.
Despite that improvement, though, there is no workforce system aimed at serving the millions of people who still rely on the program, a reality that runs counter to the rhetoric in Washington about giving people a springboard out of poverty and into well-paying jobs.
Several administrations, both Democrat and Republican, simply failed to build a real welfare-to-work training program even as the rolls swelled. Launched during the Reagan administration, SNAP E&T plugged along for years with few staff members and little attention. After the Great Recession fueled a massive increase in SNAP participation, however, the Obama administration became interested in the program to help get people back to work. They expanded staff and encouraged states to apply for more funding. But despite increased interest, local governments are still leaving funding on the table.
“I think the SNAP program has been told for the last 10 years that work [training] really isn’t their job,” said Robert Doar, a poverty policy expert at the conservative American Enterprise Institute, noting that states have been more focused on enrolling eligible residents. “If Washington is telling you, ‘Really, your job is program access’ — you do that and you don’t pay attention to work.
“That’s unfortunate,” he added, “because SNAP and Medicaid are not enough to bring people out of poverty.”
Doar and other conservatives argue that expanding the program will help low-income people become more self-sufficient.
But some Democrats regard the Republican farm bill as a massive expansion of bureaucracy. Rep. Collin Peterson of Minnesota, a moderate Democrat who serves as ranking member on House Agriculture, has railed against expanding SNAP E&T as part of tougher work rules.
“It mystifies me how the party that doesn’t like government wants to make it so much bigger,” Peterson said. “From what I can tell, the money that’s being put into work training is a sham. It’s designed to be a hassle factor so people will just drop out of the program.”
The stricter work requirements wouldn’t apply to most food stamp recipients, the majority of whom are children, elderly or disabled. They are targeted narrowly at adults between 18 and 59 who are not disabled, pregnant or caring for a child under 6 — a pool estimated to be between 5 million and 7 million people.
Exactly how many of those people would drop out of SNAP, versus how many would ultimately participate in training or work, is unknown. The Congressional Budget Office estimates that around 1 million fewer people would be in the program over a decade due to the stricter work requirements, but that figure counts those who leave because they get better-paying jobs, as well as those who decline to or are unable to comply with the new rules.
Right now, voluntary SNAP training programs serve about 700,000 people annually. The House Agriculture Committee estimates that number would be three or four times higher if the bill took effect, but others think it could be significantly higher than that.
Russell Sykes, a director of employment and economic well-being issues at the American Public Human Services Association, which represents state and local agencies that administer assistance programs, thinks the number of participants could be as high as 3 million per year.
“There’s no doubt that going from $90 million to $1 billion in three years is huge,” said Sykes, who gave credit to the committee for recognizing that SNAP E&T expansion would require a federal investment. He called the bill "a step in the right direction," although his association is still reviewing the legislation.
"The biggest thing we don't know is how big is the pool [of participants] going to be?” he said. “Is $1 billion enough?"
No one seems to know.
Many questions also surround what states would have to provide — and their capacity to do that. House Agriculture Republicans said their bill would expand the use of case managers — a model that would give SNAP recipients a point person to help them navigate what they need. But Republicans don’t know how many more people would have access to case managers if SNAP E&T were expanded. The government doesn’t even know how many participants currently have such support because states don’t have to track or report that.
Of those participating in SNAP E&T now, only a small share get the kind of intensive services the committee is touting, whether skills training such as culinary education or apprenticeship programs, or access to services such as free child care or help with transportation. The vast majority of SNAP E&T participants instead receive what are called “low-touch” services, such as support for job searches, where individuals get access to a computer, for instance, to search for a job.
Back in 1994, USDA published a review of the effectiveness of the training program, formerly known as “Food Stamp E&T.” Researchers followed 12,000 people across 12 states. They found there was “no evidence” the program, which was primarily focused on low-touch services like job search, “increased the likelihood of participants finding jobs.”
In 2015, USDA published another report on SNAP E&T, bluntly admitting: “No rigorous evaluation study of the SNAP E&T program has been completed since the 1994 net impact study.”
Washington state is widely considered to have one of the best programs — officials from dozens of states have traveled there to learn more about its approach. The state tracked one group of voluntary SNAP E&T participants from 2015 to 2016 and found that nearly two-thirds were working a year later, with a median hourly wage of more than $13 per hour.
Experts see those results as promising, but say the scale is still small. The group tracked in 2015, for instance, was only about 2,500 participants. Additionally, the most successful training programs tend to rely heavily on additional money from nonprofits and local donors.
While the 2014 farm bill gave USDA $200 million to launch 10 state pilots to test new approaches to SNAP E&T — an effort backed by both Republicans and Democrats — the effort doesn’t wrap up until 2019, which means there are no new national data about what works and what doesn’t.
That point was pounded home during a tense five-hour markup of the House bill last week.
“If you’re concerned with getting the policy right, then why aren’t we waiting to consider [the results] of these pilot programs?” asked Rep. Lisa Blunt Rochester (D-Del.), a committee freshman who previously served as secretary of labor in Delaware. “I think it’s ill-advised to scale up a massive training program until we know more.”
Blunt Rochester noted that Delaware is among the states participating in the USDA pilot, but that it’s too early to know how well their training programs are working.
Conaway said it doesn’t make sense to wait for the results, noting the pilots have been slow and the final report won’t be available until 2021.
“It is time to forge ahead so that everyone can be provided an opportunity — not just a selected population in 10 states,” he said in a statement to POLITICO. He noted that states would be ramping up their programs as the pilots wrap up, so they “will also have the ability to incorporate and use the learnings and information.”
Anti-hunger advocates also worry about how stricter work requirements would pan out in rural areas. What if your state offers you a training slot 50 or 100 miles away? Would a low-income person be dropped from SNAP if they didn’t make it to the training site?
House Agriculture Republicans contend that states would have plenty of flexibility in how they provide training, subject to USDA review.
Still, with so many outstanding questions, some advocates see the House GOP plan as a high-risk gamble that risks cutting off aid that helps low-income people buy groceries each month. The average monthly benefit per person is $126.
“I think there’s broad-based support for expanding job training, skills building and opportunities through the workforce system for very low-income people,” said Stacy Dean, vice president for food assistance policy at the Center on Budget and Policy Priorities, a liberal think tank. “The question is how do you go about that?”
Dean estimated that it would probably take closer to $4 billion to $5 billion to stand up a SNAP E&T program on the scale needed to serve everyone who would be mandated to either participate or work 20 hours per week. States currently vary wildly in their capacity, she noted, which makes it difficult to predict how such a rollout would work.
“There are lots of places where what Mr. Conaway needs doesn’t exist,” Dean said.