In previous posts, I have been skeptical of work requirements for non-cash welfare programs in general, and for Medicaid in particular. One reason for my skepticism is that the majority of able-bodied recipients of Medicaid already work. Another is that past experience with work requirements, such as the exhaustively studied results of the welfare reforms of the 1990s, show that work requirements fail to move more than a tiny fraction of beneficiaries to self-supporting employment.

Now the results of the latest round of work requirements are starting to come in. New rules for Medicaid recipients in Arkansas are one of these. Writing in the New York Times, Margot Sanger-Katz emphasizes still another reason why work requirements don’t work: Those subject to the requirements are often not aware that they even exist.

Sanger-Katz reports that Medicaid authorities in Arkansas have been realistic in recognizing that many Medicaid recipients cannot be expected to work because of disability, family responsibilities, student status, and so on. Taking those factors into account, they exempted two-thirds of all beneficiaries, leaving just 20,000 people subject to work requirements. But of those, just 1,200 reported to the state that they had completed enough work related activities (which can include job search or training in addition to actual employment) to continue their Medicaid eligibility.

Conservative supporters of work requirements might fantasize that is because the rest just sighed, gave up their benefits, and got jobs. But that is not what Sanger-Katz found.

Instead, she found that huge numbers of beneficiaries were not even aware that they were subject to work requirements.

State officials said they worked hard to get the word out — mailing letters, sending emails, placing phone calls, briefing medical providers, putting posts on social media sites and distributing fliers where Medicaid patients might find them. . . .

But it seems that not everyone opened or read their mail. Ray Hanley, the president of the Arkansas Foundation for Medical Care, which ran a call center for the state, told my colleague Robert Pear that many people never answered their phones. The state said the open rate on emails was between 20 and 30 percent.

And that is not the end of it. Even if Medicaid beneficiaries learn of the work requirements, they may find it difficult to make monthly on-line reports of their work activities, as the system requires. Arkansas has one of the lowest levels of internet penetration in the country, and the website where reports must be made can be confusing even for those with solid computer skills.

The bottom line: If success is to be measured by reducing Medicaid roles, without worrying what happens to those who are dropped, the Arkansas approach looks like it is working. If success means getting beneficiaries into self-supporting employment, it is a dismal failure.