Here’s a number you’ll be hearing a lot about at the Capitol in the coming months: $835 million. That’s the benchmark for savings to the traditional Medicaid program (that is, everything other than the private option) that Gov. Asa Hutchinson has demanded from the Health Reform Legislative Task Force. The task force duly rubber stamped the governor’s plan last week, recommending that 1) the private option continue with some conservative tweaks and 2) the state initiate reforms to the traditional Medicaid program that project to hit the magic number of $835 million in savings over five years, between 2017 and 2021. Just how it will reach that target remains to be determined; the task force and the governor kicked the can down the road on the difficult work of agreeing to any actual plan. Instead, what they agreed on was simply that target — one way or another, they need to come up with $835 million.
But how did the governor conjure that number?
Hutchinson has said, absurdly, that these are the savings necessary to pay for the private option. In fact, the state’s own consultant, the Stephen Group, found that the private option saves more than $400 million over the five-year period in question and continues to save money even once the state has to start chipping in its full 10-percent share. I’ll have more detail on that shortly. But first, let’s take a look at that $835 million figure. Hutchinson claimed that it was “what the Stephen Group report [found] and we’ve independently verified. That’s the consensus on what we’ve got to achieve on savings.” In fact, the Stephen Group initially had no idea where the governor had come up with $835 million, and the only way you can arrive at that number is with some selective editing of what’s in the Stephen Group report.
Here’s how Hutchinson invented this benchmark. The first thing you have to keep in mind is federal match rates. In Arkansas, the federal government pays for about 70 percent of the traditional Medicaid program, while the state chips in 30 percent. That means, for example, that if the total spending in the traditional Medicaid program falls by $835 million, the state of Arkansas would save 30 percent of that, or $250 million. Meanwhile, the match rates for Medicaid expansion — in Arkansas, that’s the program called the “private option” — are more generous: the feds pick up the entire tab from 2014 to 2016, then 95 percent of the costs in 2017, slowing falling to 90 percent in 2020 and beyond. This is what the governor says he’s worried about: the costs to the state once it has to start paying its full 10 percent (in terms of the state fiscal calendar, the state would begin paying its full 10-percent share in fiscal year 2021).
Over a five-year period, $835 million averages out to $167 million per year. If you just focus on the state savings (30 percent of the total), that’s $250 million or $50 million per year. This is the number that Hutchinson has zeroed in on: $50 million in average annual state savings. What’s the significance of that number?
The governor argues that $50 million per year is what the total net cost of the private option will end up being when the state has to start chipping in its full 10 percent tab in fiscal year 2021. This might seem curious since the Stephen Group report found that the state would come out ahead by $25 million even in fiscal year 2021, when the state will be paying its full share, not in a $50 million hole. Here’s where Hutchinson has done some revising. The Stephen Group factored in the gross costs to the state of paying for the private option, as well as offsetting savings and revenues: savings on reduced state spending on uncompensated care, savings because the private option covered certain populations at the higher match rate (and the state was able to phase out certain programs for people now covered by the private option), revenues from premium taxes assessed on insurance companies selling plans for the private option, and revenues collected from state taxes on the billions of federal spending coming in to the state. See here for a detailed account of all of those categories of costs and offsets.
Hutchinson uses those same numbers but erases one line altogether — the revenues accrued from state taxes applied to the tens billions of dollars of federal spending pouring into Arkansas because of the private option. See the figure below: in fiscal year 2021, the Stephen Group estimated that those revenues would be $77 million. If you get your eraser out and pretend that those revenues don’t exist, voila: the state goes from about $25 million in the black to just over $50 million in the red.
Hutchinson has never come right out publicly and said that he’s pretending this line in the Stephen Group spreadsheet doesn’t exist (although I’m told that his camp has recently told the Stephen Group that he doesn’t “believe” in these revenues). Instead, the governor insists that the “consensus” view is the one he holds. But let’s be clear about the governor’s assumptions here. The governor is saying that, despite the fact that the federal government is going to put around $2 billion per year into the Arkansas health care system, somehow not a single penny will be subject to state taxes that raise revenue for the state coffers. That’s a convenient assumption if you want to understate the benefits to the state bottom line of Medicaid expansion. But it also strains credulity, to say the least. Try a thought experiment: If the private option ended tomorrow, and the uninsured rate doubled, and the billions in federal funds vanished — does anyone doubt that there would be less health care spending in the state and thus that Arkansas would collect less state taxes on that health care spending?
Ultimately, this is a somewhat in-the-weeds dispute about just how to score budget projections, but it’s worth highlighting what the governor’s implicit claims are here and the politicking he’s doing on the accounting. The Stephen Group — just like the independent actuarial firm that the state hired in 2013 and just like the Department of Human Services in its own original estimates — attempts to account for tax revenues that come in because of the private option. That includes the direct taxes that the state levies on insurance companies on each private option plan sold, known as a “premium tax.” The governor has no objection to counting these revenues, which amount to $200 million over five years (see the line in the figure above that says “increase in premium tax revenue”).
But these estimates also account for the state taxes that will up being collected on the billions of dollars of federal spending (see the line above that says “increase in collections from economically sensitive taxes”). Remember, the insurance companies don’t get to just pocket those premiums, they have to spend that money (at least 85 percent of it by law) on medical care, and that additional health care spending gets subject to state taxes. For example, income taxes on providers, property taxes on new hospital facilities, sales taxes on new equipment and supplies, etc. Could $20 billion in federal money come into the Arkansas health care system without it being subject to any state state taxes at all? If so, Hutchinson should explain how that’s supposed to work.
There’s been some confusion about this part of the estimates and foes of Medicaid expansion have occasionally derided it as “funny money.” But it’s actually straightforward. This is not “dynamic scoring” in the sense that the term is typically used — this revenue impact does not include any potential economic stimulus effect. No multipliers or sketchy growth assumptions or fancy macroeconomic modeling here. It simply estimates the state taxes that inevitably and unavoidably end up being collected when the feds spend $2 billion a year in Arkansas, calculating at a 4 percent blended tax rate. That’s it. That makes this an inherently conservative estimate, since pumping billions into the state economy will probably lead to some growth not factored into the estimate. Here’s TSG:
The increase in collections from economically-sensitive taxes is the additional state taxes collected from the addition of new federal funds to the state economy. A typical approach to modeling the economic impact of new programs or investments is to apply a multiplier to the size of the anticipated expenditure, to capture the fact that some proportion of the new funds will be expended through local economic activity, and then the providers of that local activity will expend the received funds on other local goods and services, etc. In the calculations above, no multiplier is applied, which should result in a conservative estimate. Marginal tax revenues due to the additional federal expenditures are calculated as the total private option spending, less the state match, times a percentage factor representing a blended tax rate.
Again, if Hutchinson thinks that somehow all of those billions in federal dollars will be spent without being subject to any state taxes at all, he should make that case. Given Arkansas state tax laws and the amount of health care spending we’re talking about, that seems…implausible. Or if the governor simply has a philosophical objection to scoring it this way, he should explain — but what are we talking about other than the funds available in general revenue? In any case, his insistence that he is merely echoing the “consensus” view articulated in the Stephen Group is just not so.
Of course, projections, however they’re scored, are just that — projections. They’re not a crystal ball. Even if the governor’s explanation doesn’t stand up to scrutiny, perhaps he simply wants to proceed with caution. Prudence in budgeting and forecasting is a virtue. But if he’s building a nest egg, we’d want to see evidence that these Medicaid savings are actually being directed toward private option costs. Note that the private option trust fund, created by the original law to help pay for costs down the road, has just $20 million in it now despite the fact that the private option has already saved hundreds of millions of dollars for the state budget.
A few other curiosities about Hutchinson’s $835 million magic number (again, that’s $50 million annually in state funds):
* However you do the accounting, the governor agrees that the private option is saving the state more than $100 per year on net in the first three years of the program, 2014 to 2016 (that’s factoring in the premium tax revenues collected, but not factoring in the additional state tax revenues discussed above). Again, there is a private option trust fund that those savings could be put in to help pay for the costs once the state has to start chipping in. The Republican-led legislature and the governor decided to plow almost all of those savings into tax cuts instead. That’s their prerogative (elections have consequences!), but that’s a policy choice. It’s a choice that makes the governor’s claims — that $50 million annually in state savings on traditional Medicaid is necessary to pay for the private option — ring completely hollow.
* The years in question for hitting the $835 million target also produce net savings for the private option. The governor’s deadline is fiscal year 2021, when the state is fully on the hook for its 10 percent share. The Stephen Group found that the private option would save more than $400 million between 2017 and 2020; if you use the governor’s eraser to pretend there are no additional state tax revenues, it’s still $131 million in net state savings. That money could also be put in the trust fund to help pay for the state’s tab in 2021 and beyond. Even using the governor’s fuzzy math, that alone would amount to more than half of the governor’s targets for cuts in traditional Medicaid spending. Hutchinson presumably plans to slate it for tax cuts instead.
* The governor has fixated on the $50 million number (in state funds, which is $117 million in total Medicaid spending) for fiscal year 2021. But his benchmark is $835 million between 2017 and 2021 — averaging $117 million per year total, $50 million per year in state funds. In other words, the the governor is aiming to average $50 million per year well before the state has to chip in 10 percent. That would net an additional $200 million in state savings between 2017 and 2020. Nothing wrong with that! But I’d be curious what the governor plans to do with that money, since he keeps claiming this exercise is to fund the private option.
* This is probably just the nature of benchmarks, but it’s worth noting how arbitrary this number is. The governor, using selective editing, is focusing on what he claims will be the net cost to the state of the private option in 2021. But that’s one particular year. It’s not as if these numbers are static. Likewise, the Medicaid reforms proposed by the Stephen Group would likely see an increase in savings over time. If anything, setting a flat figure as a made-up “pay for” may encourage provider groups like nursing homes to skimp on needed reforms. The aim could easily become meeting the specific annual savings target instead of the big-picture changes that would produce higher quality of care and potentially much higher savings in the long run.
Short version of this wonky dive: the governor is picking a number, basically out of a hat, that is convenient for his purposes. The governor wants to save money in the traditional Medicaid program because that’s what he wants to do (and to be fair, there are real arguments on the merits for reforming the program for high-cost populations!). In throwing out a big round number, Hutchinson is hoping to avoid getting himself in the politically tricky muck of the actual details on achieving those savings, as lobbyists howl. And tying it to the private option is a handy trick for rounding up votes, both for the Medicaid reforms and for the private option. But the governor’s claim that the only way to afford continuing the private option is via $835 million in savings to the traditional Medicaid program? Sorry, the numbers don’t add up.
p.s. By the way, Hutchinson has actually changed his tune a little bit on this accounting question. When he gave his first big health care speech earlier this year, he used a $200 million per year figure for costs in the out years — the estimated gross costs of the state chipping in without including any of the offsetting revenues and savings. That’s bananas — you want to know the net cost if you’re trying to plan the budget. Somehow along the line, he switched to the $50 million per year figure he’s settled on now (the gross costs including most of the offsetting revenues and savings but erasing one category altogether). His techniques, apparently, are fluid. He could simply use the report from the independent consultant the state paid more than a million dollars for. But Hutchinson wants a clear benchmark useful for his political messaging, so here we are.