That's a flashy headline but this is going to be a nerdy post.
All of the empirical and theoretical evidence suggests that expanding coverage via private health insurance instead of Medicaid will cost more. The hundreds-of-millions-of-dollars question is how much more?
The number that has been floating around is $3,000 per person per year, based on a Congressional Budget Office estimate. That’s almost certainly not applicable to this deal in Arkansas. To begin with, the CBO projects the per-person, per-year (PPPY) costs of Medicaid to be $6,000. That's too high for Arkansas, and is in fact a little higher than the PPPY cost of the current state Medicaid pool, which is dramatically more sick/poor than the expansion pool would be. In fact, DHS has done already done an estimate of the PPPY cost of Medicaid expansion. Using census-based survey estimates of insurance coverage from the Urban Institute, they determined that the PPPY costs of the new eligibles would be $3,900. Without getting too deep in the weeds, I'll just quickly say that even that number may be on the high side.
So you might be thinking, wow, Medicaid in Arkansas is cheap, so this is going to cost a lot more. But in fact health care costs in Arkansas in general are among the lowest in the nation. According to both DHS and Arkansas Surgeon General Joe Thompson, we should also expect the PPPY cost of insuring someone on the exchange to be significantly less than $9,000.
So throw out $6,000 and $9,000. Here's where my attempts to do a back-of-the-napkin estimate run in to a wall. I feel comfortable using DHS projections for the cost of traditional Medicaid expansion. But to guesstimate the higher number, I need to know: What is the approximate cost of a premium for private insurance in Arkansas on an exchange with community rating and government-regulated standards of coverage? How much compared to public health insurance? Nobody, as far as I can tell, knows.
Here's Surgeon General Joe Thompson:
There's no question that private delivery of care costs more than Medicaid’s delivery of care primarily because the payment rates to providers are drastically different....
There’s no doubt that us choosing to use the private delivery system to deliver care for the expansion population is likely to cost more. How much more will remain relatively uncertain until the insurance exchange in April puts out there defined essential health benefits and asks plans to report back what their actuarial assessment of what that’s going to cost is....
We don’t know what the plans are going to come in, what their bid is. And that bid is going to affected by two things that we know. Essential health benefits covers new services that private insurance historically has not....The other thing is that the Affordable Care Act limits the amount of underwriting that health plans can do.
What about the ratio? Even if we know that $6,000 and $9,000 are each too high, the CBO's projection of percentage increase is 50%. Is 1.5 a generally good multiplier to determine the cost difference Medicaid of v. private insurance? Thompson thinks that might be a little too high. But presumably the CBO has some theoretical backing, and I can't really think of any reason that would make Arkansas have cheaper private insurance relative to public insurance than other places.
So I think 50% is a pretty good number to use, but we could use 25% if you want. Again, we're just in the back-of-the-napkin mode here. Well, DHS used that Urban Institute data to project enrollment numbers and PPPY costs and determined that the total cost of covering the expansion population was around $700 million in 2015 and $835 million a year once ramp-up is complete. That would make the cost of moving from expansion to the "private option" around $400 million a year if we use 50%. If you think the CBO ratio is high (and you need to explain why) and think it will be more like 25%, it would be $200 million a year.
George Washington University’s Sara Rosenbaum said in the D-G this morning that the reduction of "churn" between coverage plans would be a money-saver. The total yearly administration costs estimated for traditional Medicaid expansion in Arkansas (shared by the state and the feds): $14 million a year. Administration costs moving to the exchange won't go down to nothing, but even if we lop off that $14 million entirely due to reducing churn, it doesn't make much of a dent. I know Rosenbaum suggests that reducing churn will also be a money-saver by keeping folks from dropping coverage (though the higher the enrollment the more gross costs if the government is buying the premiums), but by hundreds of millions?
Gov. Mike Beebe suggested to me that the feds could make some of the additional spending back via the 3.5 percent fee they're going to impose on premiums sold on federally run exchanges. But that fee disappears if Arkansas moves to a state-run exchange, as it's already angling to do. Meanwhile, I'll spare you the algebra, but suffice to say that the revenue from a 3.5 percent fee on the plans in the expansion pool is very small compared to a 50% increase in costs of providing that coverage.
Despite the higher costs, this is still going to be a good fiscal deal for the state because of various offsetting revenues and savings. It's not impossible that some of those offsets will go up to make the whole thing a wash for Arkansas. We'll see.
But what about the feds? I just don't see any way around the fact that this is going to cost hundreds of millions of dollars more yearly (I'd say somewhere between $300 and $400 million is a decent guess — that's in present-day dollars so would grow down the road with inflation adjustment). Now if we're talking just Arkansas, that's nothing but a baby carrot. But what if other states get a deal like this? The governor and other state officials have said that this is not a special "Razorback Rule" for Arkansas, that their expectation is that other states can and will use some version of the "private option."
If Florida and Texas and Ohio and others come knocking? Then things get dicey on the cost end.
Now, HHS might well have something up their sleeve. There may be a nuance they haven't made public. I've also heard state officials speculate that the ACA overall has more wiggle room than expected to remain revenue neutral, opening up the feds' options.
But based on what we know, going to the "private option" in Arkansas looks like it will cost hundreds of millions of dollars more for the feds. I've joked that maybe Republicans like this idea because it will drive up the overall cost of the hated Obamacare! Certainly strange to see them leading the charge on a maneuver that looks to saddle the federal government with higher spending.
*UPDATE: Meant to include this: the DHS numbers use constant 2012 dollars. CBO's estimate is for 2022.