In his letter to the Health Reform Legislative Task Force last week, Gov. Asa Hutchinson noted that the state’s consultant flagged around 40,000 Medicaid beneficiaries who had an out-of-state address in a public records search. Hutchinson, in an impressive display of logical gymnastics, argued that this proved that everything had been hunky-dory with the state’s disastrous Medicaid renewal process, which booted tens of thousands of people, many of them eligible, from the rolls: 

While there may be various scenarios that explain an out of state address, it seems apparent to me that the reason thousands failed to respond to the request for income information and were subsequently stricken from the Medicaid rolls is that they were in fact not eligible.

This statement is preposterous. Let’s unpack it. 


The background here is that the Stephen Group, the state’s consultant, subcontracted with LexisNexis and found 42,891 names on the rolls that for whatever reason had an out-of-state address in public records (there are around 900,000 beneficiaries total in the state’s Medicaid program). Meanwhile, separately, the state recently conducted a troubled income-verification check as part of its required annual renewal process for the program. This led to more than 50,000 people being removed from the rolls even though many were eligible for the program according to the state’s own data (people were removed for failing to respond to vague eligibility verification letters within a 10-day deadline that was later scrapped because it violated federal regulations). 

So the state removed tens of thousands of people from the rolls based on a poorly planned and unworkable system that failed to comply with federal rules. That led to some bad press for the governor, particularly since, again, many of these folks were actually eligible, and they faced gaps in coverage — including in some cases the inability to get needed medication. In addition to risking dire consequences for beneficiaries, this led to an administrative mess


But wait, the governor says now. The Stephen Group looked through the rolls (for the most part prior to the governor’s purge) and found out-of-state addresses. Thus, “it seems apparent” to Hutchinson that those purged from the rolls via the improper 10-day deadline “were in fact not eligible.” No harm, no foul! 

Unfortunately, this makes absolutely no sense. First of all, it’s important to understand that the “flags” that the Stephen Group used to look for risk factors and that the Department of Human Services used are completely different and have nothing to do with each other. The state’s system flagged people who had certain changes in income, or who had no income at all (many of them, according to this data, were eligible, but they got income-verification letters with the 10-day deadline anyways). The Stephen Group, meanwhile, flagged “risk indicators” — such as the out of state addresses found in public records — that are unconnected to income. 


The second thing to keep in mind is that many of those flagged as “out of state” by the Stephen Group are likely in fact Arkansas residents who are eligible for the program. The Stephen Group, by their own admission, did not find anything definitive or conclusive. We have no evidence at this point whether the “out-of-state” flag even has meaningful predictive value or identified a significant number of people who don’t belong on the rolls. The DHS system flagged people who may or not be eligible…and that’s what the Stephen Group did too. 

In other words, the fact that the Stephen Group found 40,000 out-of-state addresses via LexisNexis (which, again, does not prove that someone is an out-of-state resident) tells us absolutely nothing at all about the eligibility of those flagged by the state’s income verification system.

There is zero evidence that the tens of thousands people flagged for income changes (or for having no income at all) by the DHS system also happened to move out of state. Indeed, it strains credulity to conclude that, say, the Medicaid beneficiaries without income targeted by the state’s income-verification system just so happened to leave the state en masse in the preceding months. 

In fact, the information reported by the Stephen Group suggests that only a small minority of those flagged with out-of-state addresses moved out of state after they were found eligible — the sort of people that Hutchinson would be referring to when he he concludes that the out-of-state addresses are “the reason thousands failed to respond to the request for income information.” Three quarters of those flagged in the traditional Medicaid program — and 87 percent of those flagged in the private option — had the out-of-state address before they enrolled. Aside from the fact that this calls into question whether Lexis was simply picking up an old address, it certainly doesn’t line up with Hutchinson’s story that he purged people who are no longer eligible because they left the in-state address they had when they signed up.


Hutchinson’s contention seems to be based on the fact that there is at least some overlap between those purged and those flagged by the Stephen Group. DHS has looked at a small sample of those with “out of state” flags; according to Mark White, deputy director of DHS,  a “significant number” (although not necessarily a majority) of this sample was removed from the rolls for failing to respond to the 10-day letter on time. “That shows that the renewal process is doing what it was intended to do,” he testified to the task force. “It’s identifying beneficiaries who are no longer eligible for benefits and getting them off the rolls.”

Well, no. The problem here is that someone showing up on both the DHS list and the Stephen Group list could very well still be eligible, since both lists inevitably flag eligible beneficiaries. This isn’t far-fetched — given the way that the two lists were constructed, you would expect significant numbers of eligible beneficiaries to be flagged by both systems. A student without income who still had an old family address turn up on Lexis would be on both lists; someone who had to leave her job to temporarily take care of an out-of-state relative would show up on both lists; a low-income parent who started making a little bit less money could easily show up on both lists if she moved around a lot prior to settling in Arkansas.

I asked White about this and he acknowledged that some of those flagged by the Stephen Group and purged from the rolls are likely eligible. “If they are, they will get reinstated if they come back to us and go through the process,” he said.  

Of course, many have already come back: more than 15,000 beneficiaries who were purged from the rolls for failing to reply in time to the 10-day letters turned out to be eligible and have gotten reinstated. This is what the governor’s statement skips over: That’s 15,000 Arkansans who were eligible for the program but lost coverage — threatening access to care and needed medication — because of a system that was in clear violation of federal regulations on Medicaid renewals. Obviously it’s good news that they have since been reinstated, but the state’s disastrous policy unnecessarily created gaps in coverage and an administrative nightmare. It’s worth noting that state officials who defended the 10-day deadline argued at the time that a more reasonable time window wouldn’t have made any difference in terms of response — but it sure looks like it would have made a difference for these 15,000 Arkansans!

Meanwhile, the state has no idea how many of the remaining tens of thousands purged are actually eligible. “Some of them will be coming back on,” DHS Director John Selig told the task force last month. “We don’t know how many at this point.” DHS has struggled to get in touch with most of these purged beneficiaries, suggesting that it needs to commit to more robust outreach for this transient population (and certainly should have had an outreach plan in place before setting up an unworkable 10-day deadline that led to a massive purge). 

Did some of the folks who were purged move out of state? Surely, yes. Did others legitimately move out of eligibility because their incomes went up? Again, certainly. Here’s where Hutchinson’s language is sneaky. Yup, the reason that “thousands failed to respond…is that they were in fact not eligible.” That’s true enough if he’s referring to some subset of the tens of thousands purged. But no one has ever doubted that the purge successfully removed thousands of ineligible beneficiaries. The problem is that it also removed thousands of people who were eligible — 15,000 that we know about, and many thousands more that remain without coverage. 

Nothing has changed on this point. The Stephen Group numbers do not tell us a single thing about just how many eligible beneficiaries were purged by the policy trainwreck last summer. 

Hutchinson’s allies have been peddling this same revisionist history, by the way. “The legislature said we must have a scrub,” Rep. Charlie Collins commented at a recent task force meeting. “Separate and apart from that the governor said, we must have a scrub…that was essentially doing the same thing. … While Stephen was discovering that there were thousands of people that needed to be addressed, [the state was] in real time, two months ago, throwing those people off the rolls.”

Nope. The governor was (belatedly) instituting a renewal process mandated by federal law, but did so via an overzealous policy with disastrous results — and we have no idea how many ineligible beneficiaries were identified by the Stephen Group, or, for that matter, by the DHS purge. And now Hutchinson is trying to use the irrelevant Stephen Group findings to bamboozle the public about what happened.