Department of Human Services officials told me earlier in the week (and stated publicly previously) that any change in income of 10 percent or more would flag a beneficiary and lead to an income verification letter. These letters, of course, have a 10-day window of response time, and beneficiaries have their coverage terminated if they don’t reply in time with the necessary documentation (pay stubs, a letter from an employer, etc.). 

DHS Director John Selig got in touch this evening and stated that the previous explanations from DHS were “too general” — not all +/- 10-percent changes would be flagged, he said. Instead, beneficiaries were flagged if the updated wage data (from the Arkansas Department of Workforce Services) showed a greater-than-ten-percent change up or down and they moved across one of the eligibility lines. In addition, anyone whose wage data showed zero income was flagged. In other words, anyone currently unemployed would be flagged — likely a large portion of those losing coverage, a group that DHS for some reason hasn’t mentioned before. 

This does not change the fact that many, and probably most, of the beneficiaries who had coverage terminated were actually Medicaid eligible according to the very same data that triggered the verification letters in the first place. And it does not change the fact that a 10-day window (of questionable legality) put an unnecessary burden on both beneficiaries and the bureaucracy trying to serve them, leading to catastrophe. It does, however, mean that the DHS system has at least a coherent rationale, and one that appears to plausibly and properly apply the federal reasonable compatibility standard (if you wanna know, see this explanation of that standard from Medicaid expert Tricia Brooks, a Senior Fellow at the Center for Children and Families at Georgetown University, at the bottom of this post). 

Okay, friends, for the sake of accuracy, time to venture into the policy weeds. Here is how the verification system actually operates, according to Selig. There are two relevant Medicaid eligibility lines: 1) Parents of dependent children who make less than 17 percent of the federal poverty level (FPL) are covered under the old, pre-existing Medicaid program, whereas Medicaid-eligible parents who make more than 17 percent would be covered by the new Medicaid expansion program (in Arkansas, that’s what we call the “private option”). The main difference between these two is that the federal match rates are different between the old Medicaid and Medicaid expansion—although because of the private option, it can also be the difference between a private plan and the traditional Medicaid program. 2) The income eligibility line for the private option itself is 138 percent of FPL — if you make less than that, you qualify; if you make more than that, you don’t. The state’s verification system is flagging anyone who, according to the Workforce Services wage data, had a change in income of ten or more percent that moved them across one of those lines. So if a dependent parent went from 10 percent FPL to 50 percent FPL (or 50 to 10), they would be flagged because they crossed a  line (17 FPL); if a private option beneficiary moved from 100 FPL to 150 FPL, they would be flagged because they crossed a line (138 FPL). A change within a category — say, from 50 to 90 FPL — would not be flagged.


Got all that? The idea here is to identify beneficiaries who are changing eligibility categories, which is reasonable (the previous explanation from DHS, flagging anyone whose income changed by 10 percent, was not). They also flagged people who had showed no income at all (which would apparently suggest that they’re Medicaid eligible!), which we’ll get to in a moment. 

The question I still had was why the state didn’t just collect data on income eligibility itself rather than flagging people as needing verification (without capturing the wage data) based on the standards above. After all, that would allow them to make the relevant determinations. A mother whose very-low income shifts a bit is still clearly Medicaid eligible, she just moved categories. Why not verify her income with the wage data itself? It seems, well, overzealous to send out an income-verification letter with a 10-day deadline and the threat of coverage termination to someone who is Medicaid eligible according to the very data being run through the system. Meanwhile, in the case of the data suggesting that someone now actually made too much money to still be Medicaid eligible, it would be nice if, along with an income-verification letter, DHS could also send information that explained new options for coverage for those potentially transitioning out of the program (something that never happens if someone’s coverage is terminated under the 10-day rule). 

According to Selig, the wage data itself would be insufficient to verify someone’s income if the beneficiary’s income moved  by 10 percent across an eligibility line (even though it was perfectly sufficient to verify someone’s income if they did not move across an eligibility line). Therefore, he said, there was no need to capture the wage data alone—the only relevant thing to capture was a change across a line. He said that the state was required to do additional checking if someone had one of these income shifts, even if the wage data clearly established them as Medicaid eligible in a particular category. When I spoke with Brooks, she disagreed with this interpretation of federal requirements, saying that the state could (and should) choose to simply enroll the beneficiary in the new category based on the wage data—the state could send a letter offering the beneficiary the chance to send income documentation if the change was inaccurate, but otherwise the enrollment change could go through automatically. Well, it’s late on a Friday — we’ll carry on this dispute another day. 

The bigger issue here is that DHS also flagged anyone who showed no income at all according to the updated wage data. That includes both people who were previously unemployed and still are, and folks who previously showed a little income and now show none. Again, Selig said the state is mandated by federal law to demand additional documentation from these beneficiaries. And again, when I asked Brooks, she questioned this interpretation of federal requirements. “In the case of someone who has zero income, if that person enrolled [and had no income], they would have already done due diligence to make sure that person doesn’t have a source of income,” she said. “Now they’re getting information that does not conflict with the information that they have on file. I am aware of no federal regulation where someone with zero income has to be investigated differently than anybody else.” The state would have more of an argument, she said, for doing further investigation and due diligence of a scenario in which someone’s original file listed some income and the new wage data came back zero—but she reiterated that was a policy choice by the state rather than a federal requirement. “There’s no requirement that I’m aware of,” she said. 

In any case, it’s these zero-income people that very likely make up a bulk of the folks whose coverage has been terminated after the ten-day window (it’s curious that DHS has never mentioned this category of flagged beneficiaries before now). It’s impossible to know for sure, because DHS says that the raw wage data for those dropped is too difficult or impossible to produce. But approximately 40 percent of private option beneficiaries have previously attested to no income (per DHS) — given that volume, and the volume of people whose coverage has been terminated, it’s reasonable to surmise that more of the terminations went to people with no income at all than to people who had actually transitioned out of the program. Someone who legitimately has no income, of course, is Medicaid eligible. 

One last thing: there’s another apparent problem with the system Selig describes. Under federal law, the state cannot terminate someone’s Medicaid coverage until they have checked to see if they are eligible in other Medicaid categories. According to Selig, some people are being flagged (and ultimately having their coverage terminated after the 10-day window) precisely based on evidence that they are eligible in a different Medicaid category — parents whose income moves up or down across that 17-percent-FPL eligibility line. “If you can determine someone eligible you cannot cancel someone’s coverage until you exhaust your opportunity to determine them eligible for other categories of coverage,” said Brooks. “That’s really clear.” It does not appear that that’s happening for any of these beneficiaries facing termination of coverage, including those low-income parents who are clearly eligible for another category according to the data triggering the letters. 


And so, while the the DHS clarifications are certainly important (and the system that Selig explains above is certainly more reasonable than what DHS officials erroneously described earlier this week), we’re back to where we started: It is likely that most of the people who lost coverage because they didn’t or couldn’t respond within the ten-day window (or tried to but weren’t processed in time by DHS) are in fact eligible for Medicaid. The ten-day window is questionably legal and a terrible policy choice given the rocky rollout of this verification system and the flood of plan terminations. And it is a choice. Selig stressed, as others have, that Gov. Asa Hutchinson inherited the policy regime already in place. Okay. But no one would say, in retrospect, that this has gone smoothly. The governor could now decide to extend the 10-day window to give beneficiaries a more reasonable amount of time to respond — and to give state officials, DHS employees, insurance carriers, brokers, and community leaders more time to communicate with them about just what the hell is going on. It remains the right thing to do.