The D-G reports this morning that the per-person cost of the private option for the month of April thus far is $496, which is about $18 above the monthly per-person spending cap set by the feds. The “budget neutrality caps” are a topic I’ve been intending a deep-dive post on (and the subject of a recent piece in Forbes taking a look at the Arkansas private option), so let’s jump in! As we reported when the feds first approved the waiver of Medicaid rules allowing the state to pursue the private option — which uses Medicaid funds to purchase private insurance plans for low-income Arkansans — the feds established per-person-per-month (PMPM for “per-member”) costs that the state was supposed to stay below over the course of the waiver’s three years. If the state goes over those costs, the terms of the waiver dictate that Arkansas is on the hook for the difference (this is standard for these Medicaid waivers, known as 1115 “demonstration waivers” — the idea is to try to ensure that experiments like the private option don’t cost more than regular old Medicaid would have). So how are things looking on this front for the private option? 

A post at Forbes from Josh Archambault
 last week raised the alarm bells that Arkansas exceeded those caps in February and March (as the D-G reports, the state is thus far over the cap in April too). It’s aimed at a national audience and there are some nifty charts and graphs (though sadly one employs Fox-style y-axis shenanigans), but Archambault largely covers similar ground as Rep. Joe Farrer — who is quoted at length to begin the piece — did in an article published on the eve of the final private option vote back in March

The question of cost is of course a crucially important one in evaluating the private option policy. But I come bearing good news for Farrer, who says that “Arkansas taxpayers are going to end up on the hook”: Because of the way that the waiver is designed, there is almost no chance of Arkansas taxpayers paying out money to the feds for potential cost overruns during the private option waiver. Of course, the state should nevertheless attempt to keep costs down and will have some policy tools to do just that in future years; we’ll get to that in a moment.


First let’s take a look at what is happening on the cost front thus far. Here are the average costs PMPM projected by Optumas, the actuaries hired by the state to evaluate private option costs, in August*, when the state submitted its waiver proposal

2014: $472.19
2015: $495.79
2016: $520.58

*These predicted costs are higher than the initial Optumas projections from March of 2013 (which estimated an average PMPM of $437 for 2014). According to officials from the Arkansas Department of Human Services, the higher August projection was based on updated actuarial data. On the other hand, there were other cost offsets not included in the March Optumus projection, such as the impact of pregnant women currently covered by Medicaid transitioning to the private option  — estimated by DHS to save the state $20 million in fiscal year 2015 alone. I’ll write about this more in a subsequent post, but the state badly needs to create an updated projection looking at net budget impacts now that we have much more in the way of concrete data than we did a year ago.


So, on to that concrete data. The costs thus far have been a little higher than the $472.19 for 2014 that the state projected in its waiver application. In January, the average PMPM cost was $476.59, but that was up to $483.15 in February and $485.77 in March (these will likely need to be bumped up to account for “wraparound” costs). In testimony before the Public Health committee, Medicaid Director Andy Allison said that increase was because of a slightly older population than expected and because of a clever gambit by Ambetter, one of the carriers offering policies available to private option beneficiaries (see this post for an explanation of how Ambetter offered additional benefits not guaranteed by the private option, which helped them nab more customers but at a higher cost to the private option). State officials plan to disallow this practice in future years, but Ambetter’s maneuver is making the PMPM costs in 2014 higher than anticipated. 

The latest, reported this morning: thus far in April, the per-person cost has been $496 (this does include “wraparound” costs — approximately $6 per-person on average). According to DHS officials, a quarterly report will be forthcoming soon that will show the average PMPM costs for the year as a whole. That’s the number to watch in terms of the budget neutrality cap rather than monthly fluctuations.


To be clear, the test that the federal Center for Medicare and Medicaid Services (CMS) will use to enforce budget neutrality isn’t based on how many people enroll, but on the per-enrollee cost of coverage. The cap changes each year, and the state is at risk for any additional costs if the actual PMPM price tag is above the cap. The budget neutrality test is cumulative over the three-year life of the waiver (the state is not penalized for being over the target in a given year — the feds will tally things up at the end of 2016).  Here are those caps…these are the targets to keep an eye on if you’re worried about state liability:  

2014: $477.63
2015: $500.08
2016: $523.58

As you can see, they’re a bit higher than the state’s projections in its waiver application, giving the state a little cushion. The actual January average for the private option ($476.59) was right on target, but the February and March figures ($483.15 and $485.77) were a little high, and April has been even higher. I don’t say “a little” lightly — being a little high can be a big deal: At last count, around 106,000 Arkansans had enrolled in private option plans, which means each additional dollar in PMPM costs represents a total of $1.27 million in additional costs for the year. And of course the number of people enrolling is growing quickly. If it hits 150,000, each additional dollar amounts to an extra $1.8 million; if it got to 200,000, $2.4 million. 

The March numbers were $13 above the cap. April is $18 above the cap so far. If things keep going the way they’re going, say Farrer and Archambault, the state will be on the hook for tens of millions. 

But here’s the thing: the waiver is designed to give the state plenty of wiggle room to request changes in those caps. The caps are not the firm limits that Farrer and Archambault are making them out to be. Archambault makes passing reference to this, writing that the state can ask to raise the caps but that’s “supposed to only be granted for mistakes in participation rates and similar factors.” That phrase is doing a lot of work. Here’s what the waiver actually says: 


If the State’s experience of the take up rate for the new adult group and other factors that affect the costs of this population indicates that the PMPM limit described above…may underestimate the actual costs of medical assistance for the new adult group, the State may submit an adjustment….

This represents incredibly wide latitude. If the demographic makeup of the group makes it costlier than the caps…the state can request to move the caps up. If the targets “underestimate the actual costs,” then the targets can be moved. It’s frankly hard to describe them as “caps” at all. 

“If age mix of enrollees continues to be slightly higher than predicted,” Allison said, “we will simply get the waiver budget cap changed, which the waiver terms specifically allow for.” Allison said that he was confident based on his conversations with CMS that the state could get the caps adjusted if necessary. “We will get an adjustment if we need an adjustment for purposes of demographic misprediction,” he said. The state has until October 1 to make an adjustment request for the cap in a given plan year. Allison said they would look at what the population makeup is rounding up to be this summer, as well as taking a look at expected premiums in Year 2, before deciding whether to make that request for 2014  (Allison said he is still aiming to come in below the unadjusted budget neutrality caps over the life of the waiver and has expressed cautious optimism that this remains doable).

Now, it’s true that ultimately the feds have to approve the request. They could decide to hold the state’s feet to the fire and refuse to budge. But the chances of that happening are vanishingly small. Arkansas will have a reasonable argument within the extremely wide latitude articulated by the waiver. The feds, under both Democratic and Republican administrations, have been deferential to the states on budget neutrality in 1115 waivers (if you’re feeling generous, the feds have been flexible; if you’re feeling critical, they’ve been lax). 

I asked CMS whether a state had ever had to pay back money to the feds for exceeding budget neutrality caps under an 1115 Medicaid waiver during the Obama administration? Or, for that matter, under any prior administrations? They haven’t been able to give me an answer yet. I’ve asked various experts on Medicaid waivers; everyone agrees that a state paying back money for breaching the caps is at least exceedingly rare, and may never have happened. “I can only remember one instance years ago where Tennessee exceeded its cap and I believe they went to Congress to get it fixed,” said Joan Alker, executive director of the Center for Children and Families at the Georgetown University Health Policy Institute and one of the nation’s foremost experts on Medicaid waivers and the use of premium assistance in Medicaid.

Of course just because that’s how things have operated in the past doesn’t mean that the Obama administration won’t crack the whip (it will be the Obama administration in charge of okaying requested adjustments over the three year life of the waiver). But here — even aside from the confidence expressed by Allison, who is in frequent communication with CMS — you have to use a little common sense. The Obama administration is very invested in the success of the Arkansas private option, in part because they’re very invested in reeling in other states, 24 of which have thus far refused Medicaid expansion, including some high-population big fish like Florida and Texas. As Archambault notes “the idea has spread,” with “some Republican legislators hoping to secure their own version of the Arkansas model.” The administration’s goal is to convince more states to expand Medicaid and the “private option” may well be the way forward; it would be very odd for the Obama administration to suddenly break precedent on budget neutrality enforcement in order to stick Arkansas with millions in penalties, spooking any state thinking about following the Arkansas path.

“The Obama Administration I believe is quite invested in the success of the Arkansas model so I don’t believe Arkansas will be the state to break the tradition here,” Alker said. 

I can understand the political upside of warning about the dangers of what Archambault calls “a federal per-person cap on what is typically an open-ended entitlement,” though it’s a little striking to see conservatives like Archambault and Farrer complain about that (in fairness to Archambault, he likes block-grant-style funding, but only if states have maximum flexibility). For his part, Rep. John Burris, one of the key Republican backers of the private option, said that the budget neutrality requirement “is something that conservatives should celebrate. …We set a bar for success in terms of government spending, and we hold ourselves accountable.”

The real issue here is that the budget neutrality agreement itself is fudgey, not just in terms of adjusting the caps but in arriving at those figures to begin with. That is the issue that conservatives should be upset about (and liberals too!). The Government Accountability Office released a report on 1115 waivers last summer and found a pattern of a less-than-rigorous budget neutrality process, and that was true under both Democratic and Republican administrations. Remember, what we’re talking about here in the first place is the question of whether the private option is budget neutral compared to traditional Medicaid expansion. The fudginess inherent in the budget neutrality process means that states can pursue what in practice still more or less amounts to an open-ended entitlement via a process that may well be more expensive (put another way: Arkansas is unlikely to owe the feds money even if most observers conclude that traditional Medicaid expansion would have been cheaper). Opponents of the private option often complain that the state doesn’t have enough flexibility, but in this sense, the feds were very flexible: giving the state some budget room to try this experiment.

On the other hand, if you want a sexy headline or if you want to scare Arkansans (or other states looking at Arkansas) about the private option, go with “State Taxpayers To Pay Tens Of Millions.” It’s just that that is very unlikely to happen. 


I’ve focused thus far on the potential liability of Arkansas taxpayers not just because that was the meat of the Forbes piece but because Arkansas lawmakers opposed to the private option like Farrer have been shouting about it from the rooftops. But that issue aside, cost is important! Allison and lawmakers like Burris are still hoping to beat the caps. Can they do it?

The first thing to keep in mind is that the state has three years to meet the budget neutrality requirement. Farrer got this wrong in his article back in March; Archambault knows it, but his readers might not — he chooses not to mention this and writes that “state taxpayers could be on the hook for millions of dollars for 2014 alone.” (For the very careful reader, that’s for, not in!) 

It’s true that there is not much the state can do to change the per-person costs in 2014. The premiums are set and the private option beneficiaries can choose any Silver-level plan they want to. Archambault correctly points out that the way the policy is currently set up, beneficiaries aren’t price sensitive at all. There is nothing stopping a beneficiary from simply picking the most expensive Silver-level plan available — the government is the one footing the bill. In addition to “demographic misprediction,” that is driving up costs in 2014. 

But going forward, state officials will seek to address these issues, and bring down costs, by making changes to the private option. 

One big one for 2015, already in the works, is only allowing private option beneficiaries to select among plans that exclusively offer the ten essential health benefits (EHBs) mandated by the law, with no additional bells and whistles. That will solve one factor that contributed to higher per-person costs: the Ambetter plans with vision and dental benefits. It could also lead to other lower-cost plans in the private option: every single plan except for one available in 2014 had benefits beyond the EHBs; next year, every  carrier must offer an EHB-only plan for the private option. If the private option exclusively purchases EHB-only plans, that will introduce a form of price sensitivity: beneficiaries will choose the cheaper EHB-only plans (premiums fully covered by the private option) rather than more expensive plans with additional benefits (not covered by the private option at all).  

The bigger one: rules to enforce price sensitivity for customers choosing between carriers and plans — possibly via some form of competitive bidding — is coming down the pipeline, Allison said. The state decided against using competitive bidding (for example, only letting private option beneficiaries select among the two cheapest plans) in the private option’s first year. The goal was a “transition to market… to attract and retain competitors in the insurance marketplace,” with the focus shifting to introducing price sensitivity in future years, according to DHS officials. “Our focus in this transition to market competition has always been on first insuring that there were competitors and secondly to insure that there is competition,” Allison said.  

That could be as soon as 2015, Allison said, but “the point is, it must begin. The state must adopt some measure of price sensitivity in its purchase on behalf of private option participants. Carriers need to know that it will begin. And I believe that if carriers know that it will begin, we will see the effect of that commitment whether it takes effect in 2015 or 2016.”  

Other changes are coming too. Starting in 2015 (pending federal approval, but the state law mandates getting that approval in order for the private option to continue), the private option will include health savings accounts, which will include cost-sharing. I personally am skeptical that this will do much to reduce private option costs, but someone like Sen. David Sanders, another key Republican backer of the private option, is a firm believer that HSAs will introduce consumerism to the private option, impact utilization, and drive down costs.  We’ll see — at the very least, it’s a factor to take into consideration thinking about the private option going forward. It’s part of the experiment. 

Archambault seems to assume that the the private option policy structure is more or less static. He had a new post up yesterday that does the same thing, complaining about the fact that this year the private option covers non-emergency medical transportation (NEMT) via Medicaid and doesn’t have cost-sharing for beneficiaries below the poverty line, without making even a mention of the fact that an NEMT waiver and cost-sharing below the poverty line are mandated in the law for 2015 (that NEMT waiver is another factor that should push down per-person costs in future years — in April, providing NEMT services cost the private option an additional $6 per person on average, according to DHS). The private option will look different in 2015 than it did in 2014, and will likely look different in 2016 than it did in 2015. 

Meanwhile there are other factors we don’t know going forward. If more carriers come in to the state — or more of the ones here now compete statewide — the Marketplace will likely see more intense price competition than it does today (remember, carriers are competing for non-private option customers too). We don’t even know if the current per-person costs of the private option so far this year are accurate. The medical loss ratio (MLR) provision of the Affordable Care Act means that the carriers must spend 80 percent of their premium dollars on medical care. If they spend less than that, they have to rebate the difference to the customer — in this case Medicaid — at the end of the year. In other words, the per-person costs of the private option could turn out to be lower than current premiums suggest if there’s an MLR adjustment at the end of the year. 

None of this is to say that the future cost picture of the private option is definitively rosy. It’s possible that all of the policy tweaks to come will fail, or unexpected things go wrong. But there is no reason to panic three months into a three-year (federally funded) experiment, particularly given that the private option will likely look very different in 2016 than it did in 2014. 

What we know so far is that per-person costs are a little bit higher than what the state’s actuaries projected for the waiver application in August and significantly higher than the actuaries’ original projection in March of 2013 (we need a new projection on net budget impacts!). They’re also significantly lower than many of the law’s critics predicted, including Farrer himself, who wrote there was “reason to believe” the PMPM cost would be more than $600 a month. Lawmakers and state officials will use every tool at their disposal to contain costs going forward, and they have a three-year window of federal funding to try to prove that the private option is cost effective and that it will be affordable in the state budget in future years.

In some ways, the attention to the caps is a distraction from the bigger question of private option costs. Regardless of what happens over the next three years, Arkansas will eventually have to start chipping in to pay for the private option — 5 percent in 2017, and up to 10 percent in 2020 and beyond. Even aside from the parochial interests of the state, the costs of pursuing this approach have major federal budgetary impacts if other states decide to use the “private option” route. Using this model for Medicaid expansion, or Medicaid generally, has a number of virtues, but the big open question — one of the keys of this whole experiment — is whether it’s cost effective.  

Unless it gets some major help from MLR rebates, the private option is likely going to come in higher on the cost side than hoped in 2014. Going forward, state officials and the key Republican architects of the policy remain confident that the private option can be a cost-effective reform. The challenge in 2015 and 2016 will be to prove it.