Joan Alker
, Executive Director of the Center for Children and Families (CCF) at the Georgetown University Health Policy Institute, is one of the nation’s foremost experts on Medicaid waivers and the use of premium assistance in Medicaid. She has taken a keen interest in the Arkansas “private option,” which proposes to use Medicaid funds to purchase private health insurance for low-income Arkansans. Alker was one of the authors of a letter to the feds on behalf of 26 national health-related organizations, commenting on the Arkansas plan as part of the public comment period for the federal waiver necessary to pursue the “private option.” Alker also helped pen a longer and more comprehensive set of comments, which you can see here

Alker expresses general support for the Arkansas plan but raises some questions and concerns about some of the details of its implementation. We spoke by phone earlier this week to chat all things “private option.” Oh, and “Back to the Future” and Robin Thicke! Well, sort of. Q&A after the jump…


You titled a recent blog post “I Support Arkansas’ Effort to Move Forward (But Am Holding My Nose).” Can you explain why?

We’re happy that the state is moving forward. It’s critical that these folks get coverage. There are certain ways in which the rhetoric around it is troubling. If the state had gone ahead and done this for folks [between 100 and 138 percent of the federal poverty level] rather than the whole population, I think I’d feel a lot more comfortable. But in any event, the bottom line is they’re going forward and that’s good news. There’s just some ways in which the discussion of Medicaid is inaccurate. Essentially — “private is good, Medicaid is bad” — is grossly oversimplified in my view and has some conceptual errors.


Is that what you mean by troubling rhetoric?

Yeah, and there are some logical inconsistencies when you read the waiver application.


How so?

On the one hand they’re saying the exchange will be better — private is going to better — because they’re going to pay the providers better. And that’s true: Medicaid typically is not the best payer of providers. And sometimes that causes access problems (sometimes those access problems are exaggerated by those that don’t like the program). On the other hand, that means Medicaid is very efficient, and in fact more efficient than the private sector. You can’t have it both ways.

Another issue is they’re taking the medically frail out, which is a good plan and consistent with the federal requirements [see our post earlier this week for more discussion of the medically frail and Alker’s concerns about the state’s plan to screen for that population]. Those are the sickest people. Medicaid is serving the sickest people that are going to be more expensive to cover. That raises two questions. One is, when they look at the budget neutrality, they’ve got to compare apples to apples. Hopefully they’ll do that. A waiver is an experiment, so it’s important that they really test this approach, which is a novel approach.

The other question is if private is so much better and more wonderful, why aren’t you putting the sick people in it, who really need the healthcare? 

You said you’d be more comfortable if this experiment was for the 100-138 population. Why?


I think that’s where a lot of the churning happens. That’s one of the laudable goals of what the state’s trying to do, to reduce that churning. And also, the income fluctuations. A lot of the very low-income adults who are going to be expanded to as a result of this — think of the folks below 50% of poverty. I used to work with homeless people. A lot of those very, very low-income people have a lot of residential instability. Many of them are going to have mental health and serious issues. To the extent they’re getting care it’s probably from community health centers right now. I’m not sure that private insurers have the experience or the right tools to deal with that very low-income population.

Let’s talk about budget neutrality [for more on this topic, see our post from last month]. DHS’s argument is that in order to achieve sufficient access under traditional Medicaid expansion, Medicaid would have to reimburse more. How much more? They claim the answer to that is whatever the rates would be if you expanded via the private market. The rates necessary to achieve expansion, they argue, would be the same either way, and thus the cost of covering people would be the same either way.

A lot of the groups that signed the letter, we’ve been working on issues around section 1115 waivers for years, and the need for greater transparency. There recently was a GAO report about problems in the section 1115 waiver process. What they said was a common issue — frankly with administrations of both sides of the aisle — is that the budget neutrality agreements are lacking rigor in some cases. I think that the rationale that you described that the state is putting forward falls within that tradition. They’ve made it by definition that it will cost the same either way, which doesn’t really make sense. 

What Medicaid Director Andy Allison argues is that it’s naïve to assume that you could get the providers necessary for 200,000+ more folks in Arkansas at the old Medicaid rates. From the perspective of providers, he says, it doesn’t matter whether it’s a public or private payer. His line is “we can’t beat the market.”

The people are still there either way. I think the underlying question is: Is there enough capacity to serve them, whatever approach you’re taking? So that’s one question. Obviously Arkansas is not in a managed care environment. In other states where they’re already in a managed care environment, we’ve seen a lot of big interest with the big insurers to come in to the Medicaid marketplace. So they feel like they can make it work. Obviously Arkansas, as I say, is not in managed care environment. But I think it’s a larger question of what is provider capacity. Because it’s still the same number of people either way, however you’re going to serve them. Are the providers underpaid in Medicaid? In a lot of places they are. But market share is growing. Because so many more people are going to be on Medicaid, that gives the state more leverage in negotiations, whether they do it through the exchange or they do it through Medicaid.

To the extent that there is an underlying problem, the state has 100 percent federal funding and a lot more lives to negotiate their way out of the problem, whichever way they go.

We know that Medicaid has typically been much cheaper than private insurance. Will the Arkansas “private option” cost more?

I think we really have to wait and see. We just don’t know. This isn’t exactly private insurance.

The hypothesis that costs are going to be the same by definition is not a very compelling rationale. But one could say there’s just not enough information in the application to really assess it. It’s asserted without really being backed up. 

What is a good way to test budget neutrality? It seems like these what-if counterfactuals get in to kind of “Back to the Future” territory.  It’s not necessarily an easy empirical question — there’s likely still going to be room to have political fights around this no matter what.


Yeah. Budget neutrality is always going to be theoretical. It is really hard to get a good handle on it. In a way the more interesting question to me with respect to Arkansas is: Is it cost effective? Which is different than budget neutrality. Budget neutrality is asking the question, “Is the federal government paying more under this approach than it would have under other the approach?” The cost-effective question is “Are the costs comparable to buying Medicaid coverage through the private option, as they would be just covering them in Medicaid?” I think that’s going to be a really interesting question to try to study going forward. To do that, it’s going to be important to have good data on what the administrative costs are because we know the administrative costs tend to be higher in private insurance than in the Medicaid program. It’s not too hard to get per capitas for the Medicaid program and presumably won’t be too hard to get per capitas for the new approach.

But the per capitas for the remaining Medicaid program in Arkansas will be a much higher risk pool, right?

Yes, so they’re going to have to be adjusted so that it’s an apples to apples comparison. But I think it’s going to be doable. If the state really wants an honest answer to that question, I think that’s doable. That’s a separate question from budget neutrality, which in a way you can never really answer.

Insofar as you give any stock to the DHS theory about Medicaid having to raise reimbursements under a traditional expansion, you’re still going to run in to the problem of a counterfactual that didn’t happen.

Yeah, you are, but I think probably somebody could do some modeling to account for that.

The obvious thing would seem to be to compare Arkansas to another state that did traditional Medicaid expansion.

Yeah, they mention that a couple places in the evaluation section and we did comment on that in our longer letter. It would make sense to compare it to a state that had done traditional expansion and has similar demographics. 

Ultimately it will be up to CMS to determine how to evaluate cost-effectiveness and budget neutrality. How rigorous or strict will CMS’s testing of these questions be?

On the cost effectiveness test, which I blogged about, my guess is they’re going to use what’s in the regulations for the waiver. In general their approach, which I think is the right approach, is to use existing policies and rules except for the things that are being waived. For budget neutrality, I think the administration wants to get to “yes” on this. Everybody wants to get to yes on this. In our letter, all of the groups felt strongly that we want Arkansas to move forward. We’re pointing things out that we think should be fixed, but we think can be fixed without compromising the experiment. Since budget neutrality is really an art rather than a science, I think they’re going to be able to get to “yes” on that. 

Yeah, on the one hand the waiver is supposed to be an experiment, an empirical test to see if something new works, but on the other hand, politically CMS seems very invested in seeing that this does work.

They want states to do the Medicaid expansion so they really are doing their best to allow states to try different approaches, without compromising the fundamental integrity of the Medicaid program. And that’s one thing — this is going to sound really nerdy, but for me as someone that looks at Section 1115 waivers, that’s the exciting thing to me. This really is testing something new! A lot of times states come in and want a waiver for something and they don’t really need one. This really is an experiment. That’s why, coming back around to where we started, I think it’s important that the hypotheses be intellectually honest. And that it really is studied and tested going forward because I think it’s a valuable experiment.

In terms of the value of the experiment, let’s talk about “premium assistance” schemes generally. Obviously there’s a political advantage—anyone that’s followed the story in Arkansas is sensitive to that. Does it make sense from a policy perspective, the idea of providing this kind of service or coverage through a private third party as opposed to directly?

I’d say two separate things to that. If you’re thinking about the population that’s over poverty…The marketplaces are coming online shortly. The new subsidies will be available. You could see how in a perfect world — this gets to the issue of children, for example, the state says they eventually want to put children in. So you can see in a perfect world that it makes sense to put the children in with the parents and put them in the exchange. Our concern is that ARKids has been very successful in Arkansas and overall the Medicaid and CHIP programs have been very successful. There are so many things going on right now with reform and changes as a result of health reform that we don’t want to rock the boat right now. We really need to keep CHIP going for the foreseeable future, until things settle down and we can assess. That’s a longer term question. On the issue of using Medicaid dollars to buy private coverage, it’s interesting because this is still government coverage. This is another of the things that lines are murky. You say Medicaid versus “private option.” Well a lot of Medicaid is actually purchased from private insurance companies. That’s how a lot of states run their Medicaid program. They buy a package from Amerigroup or Centene. These are all private, publicly traded companies. They’re buying that coverage for the Medicaid beneficiaries. Now the state is going to be buying from insurers in essentially a government marketplace. The lines are blurry.

You said you see this as a legitimately new experiment. Is this different than managed care in other states?

I see it as different because the marketplaces are new. The kinds of premium assistance that has been done in the past in Medicaid is nothing like this. It’s been about buying somebody’s WalMart coverage for them, it’s been about subsidizing employer-sponsored coverage and seeing if that’s cost effective. Under poverty there’s not a lot of people who have access to employer-based coverage and that’s why it’s been a strategy of limited utility.

That circles back to my earlier question — this idea of using premium assistance to buy coverage for folks on the government-regulated private market as opposed to giving them that direct service…does that have advantages as a policy (as opposed to the political advantage that it was the only way forward)?

It very well might. Because this is going to be a relatively large marketplace with a lot of subsidized lives coming in to it, it’s possible that more providers will participate and that that will be a good thing. It could be a good thing for the Medicaid beneficiaries. Absolutely, it may well be, that’s why it’s an experiment worth doing.

Could better access lead to better health outcomes?

Yes, I think so, although if you look at access that exists today, there are some places where there are big differences between Medicaid and private, like dental — but by and large, at least for kids, it looks pretty much the same for Medicaid and private. But having better access to providers is obviously a good thing, particularly in the area of specialists, which is often cited as an area of frustration for Medicaid beneficiaries. Although you always have to ask the question — as compared to what? I can say it’s hard getting in with my children’s dentist and we have private insurance. So I think it’s important to place all of these questions in a larger context. To coin the song of the summer there are “blurred lines” between public and private, and that’s evident here in ways in which some of the logical inconsistencies come out.

Around half the states aren’t going forward with Medicaid expansion. When the “private option” first came around there was a notion that some other states might latch on. Is this an idea that has legs in other parts of the country [this interview was conducted prior to the announcement that Pennsylvania may pursue expansion via some form of the “private option” approach]?

It got a lot of attention when it first came out. There are a couple of states that are planning something with the 100-138 population. I don’t think it’s going to be a firestorm to do this, and I never thought that. But I do think it’s an option, particularly for the 100-138, that when we look at the second round of states that grapple with this, this definitely is going to be part of the discussion.