UPDATE: Want to know the latest? See our follow-up post here.
Some very smart people are either really angry or really stumped these days, for reasons that have nothing to do with the sequester.
You read about the Arkanfuffle in our latest edition of Wonkbites, right? Arkansas Governor Mike Beebe just received permission from the Department of Health and Human Services to sidestep the Medicaid expansion—and get the federal government to pay for private coverage instead. Curiously, HHS has declined to go on record about the legal foundation for the move—but after some tremendously enlightening correspondence with Timothy Jost and other legal scholars, we think we’ve got it down. The legal grounds might be shaky, and that’s a big question. But there’s one that’s even bigger: if it is legal, can we handle it?
Before we throw down the legal gauntlet, let’s start by clarifying what the Arkansas option isn’t. Behold, the myths:
It’s not what Florida did. Originally, some commentators thought this looked like a Florida copycat. Florida received a waiver from HHS allowing them to move Medicaid recipients into Medicaid plans run by private insurers—key redundancy being that they’re still Medicaid plans. Arkansas, by contrast, will be putting newly-eligible citizens directly into the insurance exchange where they can shop among private plans. The federal government said they’ll pick up the tab at the same rates—100% for the first three years, phased down to 90% by 2020—designed to support “traditional” Medicaid expansions in other states.
It’s not a “loophole” or “drafting error” in the Affordable Care Act. We might hate Congress more than lice and Donald Trump, but this isn’t something they screwed up. This isn’t even new. Under specific conditions of the Social Security Act (Section 1905), states have had the ability to do this for years. It was just too expensive to do on this scale before; average federal match has steadily hovered around 55-57% since 1970. But the ACA has the federal government picking up the whole newly-eligible tab for the first three years, which gets phased down to 90% into theoretical perpetuity. That changes the ball game.
It’s not written in stone—or law—yet. Yes, Gov. Mike Beebe presented this to the Department of Health and Human Services. Yes, HHS approved the plan. And yes, this move seems to be placating previously resistant Arkansas legislators, who are starting to look favorably at this new “expansion.” But it hasn’t passed the state legislature yet—until it does, it’s just a figment of our collective wonky imagination. The legislative session ends April 19, but doesn’t formally adjourn until May 17, so we might have to sit tight.
So where’s the rub? Okay, this is where it gets confusing. The Social Security Act has a segment better known as the “Medicaid statute.” Section 1905(a) of the Medicaid statute says that states can use Medicaid funds to subsidize private insurance. Furthermore, Section 1905(a) doesn’t seem to articulate a cost-effectiveness requirement for using Medicaid money on exchange plans—but in law, omission and permission aren’t always the same thing. This seems to be ambiguous. With us so far?
When a statute is ambiguous, an agency—HHS in this case—often proposes regulations that clarify those ambiguities. And HHS did just that. In fact, there’s an explicit cost-effectiveness requirement in rule 435.1015:
“… the cost of purchasing coverage under an individual health plan for a Medicaid-eligible individual in the private market, including coverage in a [qualified health plan] in the Exchange, must be comparable to the cost of providing direct coverage under the state plan (or waiver of the state plan).”
So, it sounds like Medicaid dollars can be used to pay for plans on the private exchanges as long as those plans don’t cost more than Medicaid would have. But is that even possible? Private plans cover more. Private plans pay providers more. Those numbers don’t seem to add up.
So now we’ve reached the crux of the legal question: does Arkansas have to obey that HHS regulation, or not? Here’s the thing about law: statutes trump regulations. They just do, that’s how it works. So if rule 435.1015 (the HHS regulation) conflicts with the Medicaid statute in some way, the Medicaid statute wins and 435.1015 loses .
Top health law experts we communicated with are not certain Arkansas’s expansion plans are on sound legal footing. Insofar as we’ve wrapped our heads around this, there appear to be three possibilities. Submitted for your wonky analysis, with brief commentary:
- The Medicaid Statute expressly permits Arkansas’s move. It trumps the recent HHS regulation. This seems unlikely. To our knowledge, no one has been able to point to specific text in the Medicaid statute indicating “hey guys, cost-effectiveness doesn’t matter.” The omission is what creates ambiguity. And why would HHS knowingly write a resolution that conflicts with an established statute?
- Though not expressly permitted, Arkansas has discretion to make their own decisions about using Medicaid funds on the exchange. This similarly trumps the recent HHS regulation. This is more likely than the first option, and suggests that the ambiguity actually creates flexibility for the states. But it does make one wonder: what was the point of HHS drafting a regulation on cost-effectiveness if states have the authority to casually disregard that regulation on a whim?
- Nothing in the Medicaid statute expressly forbids this. HHS has discretion to interpret the law and pass regulations that states must follow. Assuming HHS has discretion, they used that discretion to pass a regulation requiring cost-effectiveness for the use of Medicaid dollars on the exchanges. As we discussed, it seems impossible that private insurers could offer plans at or below Medicaid’s costs. So, offering Arkansas this deal seems to invalidate their own rule. What’s up with that? [ETA: based on this legal brief from Sara Rosenbaum, this appears to be the legal reality—but the interpretation of cost-effectiveness is still murky; it may be more generous than “at or below Medicaid costs,” but we don’t know how generous.]
Sick of legal nuance? Let’s get to the good stuff: if it turns out that this kind of expansion is totally legit, what are the implications?
Don’t forget who’s picking up the tab. So yes—the cost question is huge, not only in what it means for this deal’s legality, but also in what it means to all of us who don’t live in Arkansas. Odds are that includes you. If their plan costs the federal government more than it would’ve for Arkansas to expand Medicaid the old-fashioned way, well, that money has to come from somewhere. It comes from taxpayers, obviously, but the rub is that those taxpayers are all over the country, and the folks who benefit are those in Arkansas. In other words: if Mike Beebe’s plan costs more, we’re all paying for it.
This reveals yet another quirk of federalism. Right now Arkansas gives the feds a bit more in taxes than it gets back in spending. But Karan’s state, New Jersey, gives way more than it gets back—150% of GDP, making it by that measure the third most shafted state in the union. And NJ decided to expand Medicaid the old-fashioned way; so in a way, they’re getting penalized for playing fair with the ACA. So Karan’s mad as hell, which is fine, but it shows an important point: Arkansas is getting more from the feds without giving them anything extra in return, and every other state is on the hook for it.
By rough estimate, the most the feds would have pay extra for the Arkansas deal comes out to $700 million yearly . It could be money well-spent, because commercial plans provide better coverage than Medicaid (back to this in a bit). But the amount is substantial (0.7% of 2011 state GDP), it has to come from somewhere, and more importantly, it would get way, way bigger if other states decide they want in, too.
And why wouldn’t they? It’s an awesome deal—way more federal money entering the pockets of your own insurance companies and healthcare providers, shiny commercial insurance plans, plus the swag of supporting state autonomy and private enterprise. The feds will foot patients’ entire commercial premiums, and co-pays can’t be any higher than Medicaid’s. This deal is almost as good as getting Chipotle when you order Taco Bell.
But if Texas wants Chipotle, then we’re talking about a Texas-sized $4.5 billion bump in the federal price tag . If even more states realize how much better Chipotle is than Taco Bell… well, you get the point. This could be limitless, since from our research there’s no cap on how many states could get deals like these. If commercial plans cost extra, states that choose plain-old Medicaid expansion get the shaft. Depending on how many states follow the AR model, that could be a Samuel L. Jackson SHAFT. We don’t yet know whether HHS could say no to other states after creating this precedent with Arkansas—but we can’t imagine what they’d say to justify it.
There is one pie in the sky that could make this work: but it’ll require a temporary suspension of disbelief that many experts aren’t okay with. What if the cost-effectiveness requirement stands, and by some stroke of God, commercial plans make themselves cheaper than Medicaid? The savings are out there—Don Berwick estimates that 34% of what we spend on healthcare is waste—but achieving them would require a ruthless, wholesale re-engineering of how our healthcare works. It isn’t impossible. It certainly would take time, and it wouldn’t be easy. But encouraging state-level innovation is a huge part of the ACA implementation strategy, and hey, we can dream right?
Are we creating yet another tier of healthcare access? As we mentioned earlier, the AR deal applies to those newly eligible for Medicaid. Since states have historically restricted Medicaid to the extremely poor, or to parents (as we’ve explained before) Obamacare’s Medicaid expansion means a lot of people (who were less poor, or childless, for example) will be eligible for Medicaid who weren’t before. Except in Arkansas, where those new folks will get subsidies for shiny commercial insurance plans (Chipotle).
How does that matter? Commercial plans provide far more access to healthcare than Medicaid, because they pay doctors better. So what could emerge is a situation where Arkansans getting Medicaid today (working parents under 16% of the poverty line) have worse access than those who were supposed to gain Medicaid eligibility under Obamacare. Again, these aren’t simply issues of income. This deal means one adult under 16%FPL could have better, commercial coverage than another adult, also under 16%FPL, on Medicaid because he or she has a kid . Those arguably in the most need of affordable healthcare will have the least desirable coverage. It’s probably an unintended consequence, but that doesn’t make it any less perverse.
This should diminish commercial/Medicaid “churn.” But if we look a bit higher up the income scale, another interesting wrinkle emerges. Under Obamacare, Medicaid is at risk of a “churn” problem—because of normal fluctuations in income, up to 50% of eligible people could flip back and forth between Medicaid and the insurance exchanges in a given year. But as we mentioned above, not every doctor who takes health insurance will take Medicaid, so people “flipping” down to Medicaid will lose healthcare options and continuity in the process.
Beebe’s plan, however, shrinks that problem. By making a much broader group of people exchange-eligible, through simple math, he reduces the amount of people likely to flip back and forth into Medicaid. If you support healthcare access and continuity of care, this is a good thing. But again, what if everyone says they want Chipotle? We do think that Beebe’s plan will improve the healthcare of those getting commercial plans instead of Medicaid—but can we afford it if other states follow suit?
Also, despite providing broader access to care, it’s possible that the insurers providing subsidized plans might just not be as good at covering new low-income populations. Since they’re used to catering toward better-off people than those getting Medicaid, their networks might not have healthcare providers with what their new low-income members need (language services, etc.). As Sara Rosenbaum notes, we have an “Essential Community Provider” designation to make sure those providers are part of every insurance network, but as always, the devil is in the details (which we respectfully leave to her).
Whew. Are you tired? We’re tired. But we emerge from our tour of law, policy and economics with three big questions:
- Is their deal legal? If HHS approved it, it should be, but someone has some explaining to do. We see policy conflicts on three levels—state, congressional, and regulatory—and they leave us, as well as the prominent legal scholars we talked to, stumped.
- Is it good policy? For Arkansas—yes, in a sense. As we discussed, Chipotle comes with a lot more goodies: more federal money, better healthcare coverage, and ideological purity. On the national level, Beebe’s deal smacks of preferential treatment. But even for the state, it’s still not necessarily the right thing to do: Paul Krugman equates these moves to “welfare for the medical-industrial complex,” assuming—as we’re inclined—that commercial plans will, by definition, cost taxpayers more. And Adrianna, who is too wonky for her own good, still has some theoretical concerns .
- Does it matter? Maybe. It all depends on how many other states turn into Beebeliebers. In the scheme of national health expenditures, $700 million yearly is money, but it’s nothing close to the $2.6 trillion we spend on health as a nation. Twelve states have yet to announce what they’re doing with Medicaid, and many others could probably change their minds and decide the deal’s too good to pass up. Then this would really matter.
We might not have answered these questions in full, but hopefully we’ve highlighted how important each of them is. This story is constantly evolving, but we’ve attempted to give you the all the what-ifs and wonkery you need to understand it. We will provide updates as news comes our way, but we’d appreciate your feedback in the comments if you hear something we don’t!
Footnotes (Yes, footnotes!)
1. CFR 435.1015 is currently a “proposed” rule—which is to say, it needs to be finalized following an open comment period before having the full effect of law. It is possible that HHS will amend the rule before finalizing it. That does not change the agency’s inconsistency between having drafted the rule and permitting Arkansas to pursue this expansion. Policymakers closely track differences between “proposed” and “final” rules; to change or omit this part would telegraph to other states that they, too, could seek an expansion with few financial constraints. More to the point, this rule may not be finalized until after the Arkansas legislature adjourns, potentially creating an environment of serious uncertainty.
2. Why $700m? 233,000 projected new enrollees, each costing $3,000 more than they would’ve in Medicaid comes out to $700 million. This is of course on the high end, assuming that the subsidized commercial plans aren’t any cheaper than what’s already on the exchanges.
3. For Texas, estimates say 1.5 million potential new enrollees, again multiplied by $3,000 extra. This is both because Texas is enormous, but also because it’s had one of the more restrictive Medicaid policies in the country.
4. Since childless Arkansans today aren’t Medicaid-eligible at all, and under Obamacare you don’t need a kid, a mom under 16%FPL will have Medicaid, while a childless adult under 16%FPL will have a commercial plan.
5. Adrianna wrote about another legal challenge involving the state insurance exchanges. If that comes to pass, there are important interactions to consider._____________________________
Adrianna works in clinical research and is a graduate student in public policy & public health at the University of Michigan. Follow her on Twitter @onceuponA.
Karan is a first-year student at Robert Wood Johnson Medical School and Duke graduate who previously worked in strategic research for hospital executives. Follow him on Twitter @KRChhabra.