CASEY COPELAND’S addiction to heroin landed him in jail, but he came out scared straight. Without a job, he signed up for health insurance through Medicaid, the government health-insurance programme for the poorest, and took up volunteering at a charity that helps the homeless. Mr Copeland thought that was that. He was unaware of the work requirement Arkansas had recently put on the programme and didn’t notice the letters from the state that were piling up. After three months of non-compliance his insurance was cancelled. Mr Copeland is reapplying, but in the meantime he is uninsured. He had to return the machine to treat his sleep apnea, a condition which causing breathing difficulties. Mr Copeland is sanguine about this even as he recounts that without the machine he once stopped breathing 17 times in a single night.
In January 2018 the Trump administration signalled that, for the first time since Medicaid was introduced in 1965, it would grant waivers to states allowing them to place “community engagement” conditions on the programme. Able-bodied adult recipients would need to work, volunteer or study for a set number of hours to keep their coverage. It is the most significant change to welfare policy of Donald Trump’s presidency. According to estimates by the Kaiser Family Foundation, a think-tank, if similar requirements were implemented nationwide, between 1.4m and 4m people would lose coverage. Fifteen states, almost all Republican-led, quickly applied. Arkansas became the first to implement the new rules, starting in June 2018.
The big reforms to cash welfare during the 1990s came about in a similar way. States were granted authority to experiment with making benefits conditional on work and introducing lifetime limits. Eventually these were codified nationwide under Bill Clinton. The arguments in favour are the same now. “This is an effort to essentially be compassionate and not to trap people onto government programmes or to create greater dependency on public assistance,” says Seema Verma, the administrator for the Centres for Medicare and Medicaid Services (CMS). “If you’re living in poverty, you need more than just a Medicaid card. You need a pathway out of poverty,” notes Ms Verma. Asa Hutchinson, the governor of Arkansas, takes a similar line. “It’s balancing that compassion with the other value of our country, which is responsibility,” he says.
The preliminary results from the Arkansas experiment look alarming: 18,000 people lost their health insurance in the first six months because they did not comply with the requirements. Confusion seems widespread. Many only realise they have lost insurance in the pharmacy, after trying to pick up a prescription they can no longer afford. In some months more than 90% of those required to report their activities did not. For the first few months reporting could only be done online. More than 20% of those affected did not have access to the internet; those that did found the website, which shuts down between 9pm and 7am, clunky and complicated.
In theory, placing work requirements on welfare programmes can result in higher employment and less government spending. In Arkansas, though, the labour-market effects are hard to detect. State officials point to a report showing that over the first six months of the new policy 4,400 Medicaid participants found work. But it is unclear whether people are moving from unemployment to work or merely switching jobs. Similar numbers before the work requirement went into place, which would allow for comparison, are unavailable. “There is no baseline data, and that lack of data is really concerning,” says Kevin De Liban of Legal Aid of Arkansas, which is suing the state to reverse the policy.
In practice people who are eligible can fail to jump through bureaucratic hoops and end up with neither work nor welfare. One of the plaintiffs in the lawsuit against the state is Adrian McGonigal, a 40-year-old chicken-plant worker with respiratory problems. Without a computer, smartphone or access to transport to a public library, he failed to meet the work requirements and lost his health coverage—which he only learned after trying to fill the prescription for his medication. Without insurance this would have cost $800, which he did not have. Mr McGonigal went without, got sick and missed several days at work, for which he was then sacked.
Because of the volatile nature of low-wage work—in which earnings and hours change seasonally or erratically—the chances of someone working insufficient hours to meet the requirement or having an income that is temporarily over the limit, and thereby losing health coverage, is fairly high. More than 60% of able-bodied adults who receive Medicaid already work. Most of those who do not are typically in poor health, taking care of young children or disabled relatives, or in school—all of which exempt them from the work requirements. Another analysis from the Kaiser Family Foundation finds that only 6% of adult Medicaid recipients are currently not working and unlikely to fall into these exempt categories.
Understanding whether the Arkansas experiment is successful requires knowing whether those 18,000 people who lost their coverage after the new rules came in have moved on to other health insurance or employment. Yet that is strikingly difficult to find out, and the state is not trying too hard. “You’re asking who they are: I don’t have the statistical information, it hasn’t been broken down,” says Mr Hutchinson, the governor. “There’s no doubt in my mind that of those 17,000, somebody out there is healthy, has received a notice, understands the responsibility but just doesn’t do it. And what do you do at that point?” he asks.
State officials did launch an outreach campaign but found that many people in the Medicaid programme were not contactable. These people could have already moved up the income ladder, received insurance through an employer or spouse or moved out of state, says Cindy Gillespie, the director of the Arkansas Department of Human Services. Because the coverage lock-out ends every calendar year, those barred from Medicaid last year can reapply. Only 1,300 have done so—which state officials and Ms Verma see as evidence that only a few legitimately claimed the coverage. Ms Gillespie also points out that hospitals are not reporting increased uncompensated care. “We would expect that if there were a lot of people who were actually using their insurance, that we would see a rise in uncompensated care,” she says.
Mandy Davis, the director of Jericho Way, a day centre for the homeless, sees it differently. The people she helps “get a letter and they don’t understand it, or they try to fill their medication and are denied,” says Ms Davis. “There’s the assumption that people are computer literate, or just literate to begin with.” She has helped read aloud the official letters giving notice of lost coverage to those who have trouble understanding them. “These are hard letters to read,” she says. “We’re having to find the nurses and doctors who will provide medical care for free—the same ones we used to call ten years ago.”