Two big headlines from the DHS release of actuarial findings on expansion options Wednesday. They project that 1) The "private option" saves $670 million to the state bottom line over ten years. 2) The "private option" not only doesn't cost the feds more than traditional Medicaid expansion, it's cheaper — by almost $600 million over ten years.
That first one is not surprising. We've been explaining why expansion is revenue positive for the state for some time, and the "private option" doesn't change that (DHS projects that traditional expansion would save the state a little less—$610 million over ten years).
But #2 is a bit of a doozy. Most outside observers have speculated that the "private option" will be significantly more expensive. When Arkansas first got the go-ahead for the new framework, Arkansas Medicaid director Andy Allison wrote to Cynthia Mann, the director of the federal Center for Medicaid Services, "Cost effectiveness is the obvious hurdle for folks who didn't expect this" (e-mail acquired by the Times via FOIA request).
DHS challenged the consensus about cost last week with its release of general findings that it could be close to a wash; the fully enumerated projections this week suggest significant federal savings.
Of course we cannot know for sure, but predictions about federal costs matter a great deal. If the "private option" costs significantly more, it's hard to justify it over a traditional Medicaid expansion (outside of political necessity). It also might create a significant problem for the federal government if other states want to hop on board the "private option" bandwagon. Some started worrying about billions in additional costs.
DHS is swooping in now with an actuarial study that says not so fast. If they're right, then the case for the "private option" could be even stronger than the case for Medicaid expansion. But their arguments go against the general consensus about the costs of the government paying for health coverage through private companies instead of directly. They've been greeted with some skepticism outside the state so far.
I interviewed Allison yesterday to try and tease out where these numbers came from and why they think they're on to something that the consensus view has missed. After the jump, see a summary of DHS's key arguments, plus a response from a health economist who remains skeptical.
Let's look at the core components behind the theory of DHS and their actuaries. We'll start with the basics and then get to the core theory behind their findings which, if accurate, would be a game-changer.
1) Obviously, all of the numbers are only as good as the actuaries who compiled them. The Insurance Department contracted with Optum, an actuarial company in Arizona. "This firm has experience consulting both private firms engaged in the exchange as well as for a long period of time Medicaid programs and managed care programs," Allison said. "They're very used to developing these sorts of estimates. I have found them to be even conservative." In addition to standard actuarial methods and experience nationally in the public and private sectors, they also had the most recent three years of comprehensive claims data from DHS.
2) The baseline difference that Optum found between private rates and Medicaid rates was around 24 percent. They also projected that private rates would go down further under the "private option," by around five percent, because of competition and because of better health plan management. The competition is achieved not through lots of consumer choices in the marketplace, but through a competitive bidding process for the carriers. The rules haven't been set yet, but the plan will be something like this: consumers only get the premium paid for if they pick one of the two cheapest plans (the same competitive bidding structure will exist for the 138-400 population getting subsidies on the exchange). Since almost everyone will do that, there will be intense competition among the carriers to be one of the cheapest two. Again, this means not that many options for consumers if they want to be subsidized, but it's a mechanism to make them extremely price sensitive. As for plan management, Allison offered the example of the small cost sharing allowed by the law. It will be "tailored and focused" by private insurers in a way that's "just too awkward for the state to try to do," he said.
3) The biggest, and likely most controversial factor, is that they assume that under a traditional Medicaid expansion, Medicaid rates and private rates would be the same. That's why they apply a 24 percent up-charge to the Medicaid rates under a "traditional expansion" counter-factual: we know that Medicaid rates are cheaper than private rates, but Allison says that if we expanded Medicaid, that would cease to be true.
"Much more in reimbursement would need to be made available to providers and to the system to increase and secure access for this new population," he said. Not only higher than before, but higher even than Medicare — all the way up to private rates. "We're suggesting that you can't continue to cross-subsidize in a world of full insurance."
Some of this may be specific to Arkansas — which has a low differential between public and private rates now and also has extremely stingy Medicaid, so that there is "a missing market in the state." But Allison is also arguing that the trend of private and public rates moving toward each other is what we should expect to see if places move toward universal coverage. That trend is observable in other states who already cover this population, he said, as well as reports from private insurers developing bids for exchanges looking ahead to 2014. Whether private rates go down or public rates go up is "a bit of an open question" but "the point is that they need to be the same."
If you've been following the healthcare debate and the "private option" debate, you will recognize how radical — or at least new — this argument is. It a priori assumes that the gross cost of insuring the expansion via Medicaid or the "private option" is about the same (which means a "private option" is cheaper overall because of #2 above). That would change everything! As Allison said, "If it's the same number either way then this question of cost comparability — cost effectiveness — for the private option versus some kind of traditional Medicaid is moot."
"It just didn't occur to folks," he said. "It didn't occur to me or anyone else. It's not a discussion we've had in the country because a pure buy-in like this really wasn't imagined until the last couple of months."
The key point here is that achieving the access mandated by law for the large expansion pool requires upping reimbursement rates (this is precisely why private insurance has better access now). Allison believes that previous estimates failed to take this into account. Allison wrote to CMS director Cynthia Mann on March 7 (email acquired by FOIA request): "CBO seems to have made some sort of naive assumption, apparently, that existing average provider reimbursement rate differentials (Medicaid FFS versus private carrier rates) would be the same with and without all the new coverage."
A rate increase in Medicaid would require legislative action but Allison suggested that a rate increase to meet access needs was easier to achieve politically than you might think — and presumably access requirements mandated by federal law in order to get the match rates could force the legislature's hand.
The theory goes that in a competitive market with near-universal coverage, the price will move toward what is required to achieve the minimum required level of access. That could move Medicaid rates up or private rates down or a little bit of both.
"Look at this from the provider's point of view, from the healthcare system's point of view," Allison said. "Partly what is implied here is that the system itself is not quite big enough. There's two issues — there's how many certified professionals and how many facilities." How they adjust to the new pool of consumers is "a market decision. And that market decision, from their point of view, we don't think is any different depending on who's paying them."
The key to making the "private option" work is that the medically needy people in that pool will go to Medicaid, so there will be a separate high-risk pool (and in fact, some healthy people currently in Medicaid will move to the exchange). "We are explicitly protecting the exchange from the highest risks among that group because the top high-risk 10 percent would likely be withheld," Allison said.
That leaves "a large, stable, and healthy population that could actually stabilize and significantly reduce adverse selection" in the exchange, Allison said. "What it takes to create a competitive insurance market like an exchange is to seed that large group—that marketplace—with a large, stable, and ideally relatively healthy population. That is precisely what the Arkansas option would do." Allison described it as a "perfect antidote" to the potential for higher private premiums suggested by a recent study from the Society of Actuaries. (This is an appealing frame for Republican lawmakers: private option protects us from Obamacare! Silly, but politically valuable.)
Boston University health economist Austin Frakt said that the DHS predictions were possible but should be greeted with healthy skepticism.
These numbers are really hard to know for sure. The probability that everything will work out just perfect that it's going to be a cost savings or cost neutral — it's just hard to know. I think it's an interesting and innovative plan. It's got some great things in it for beneficiaries, great things in it for providers. It's obviously of tremendous political value and that can't be ignored. It definitely comes with some risk of increased costs to the federal government. ... I'm not fully swayed by the analysis that it will be cost neutral or cost savings.
You can (and should!) read more of Frakt elaborating on his questions about the Arkansas plan's costs here on the Incidental Economist blog.
That said, Frakt suggested that those risks of federal costs might be worth it "as long as it's understood that [increased costs] could happen." He also noted that other states might follow the lead of Arkansas and the numbers might work out differently. "It's hard to believe that every state would be able to save money with it," he said. The key is that we go in to this experiment with our eyes open about cost uncertainty. (Cue the joke that the sneaky reason for the "private option" is driving up the costs of Obamacare.)
In the end, I find the experimentation argument pretty convincing, and you have to think that's part of what is motivating HHS. The "private option" will be at least as good for beneficiaries, so if HHS is willing to take a little financial risk to try an approach that could plausibly turn out to be better, that seems like a good thing. As Allison wrote to me, "we could easily end up learning more about Medicaid in the next ten years than the previous 50 combined." That doesn't guarantee better policy but it's a start.
And we can't forget the political reality: right now, the "private option" is the only way forward to getting health coverage for more than 200,000 Arkansans.