In my post last week, I asserted that the premium subsidies provided in Obamacare’s state exchanges would be tied over the longer run to consumer inflation.
The story is actually much more complicated than that, and requires additional explanation. (Fair warning: What follows is very technical.)
The way the law (see page 111) works is that it sets a limit on the percentage of income a household must pay for premiums when it is getting insurance through the state-based exchanges. In 2014, the government’s contribution is determined by subtracting the maximum contribution required from a household from the total premium required for the second-lowest-cost “silver plan.” The percentages of income used to set limits on household premium payments are to be adjusted in the years after 2014 based on factors specified in the law.
What are those factors? From 2015 to 2018, the law says that the percentages “shall be adjusted to reflect the excess of the rate of premium growth for the preceding calendar year over the rate of income growth for the preceding calendar year.” This provision is written poorly and imprecisely. What does it mean to “reflect” the excess of premium growth over income growth? How exactly are the percentages to be adjusted? The law does not specify a mathematical formula. It just says that an adjustment shall be made.
Common sense would indicate that this adjustment is intended to prevent household contributions from becoming an ever smaller share of the total premium. Income is expected to grow less rapidly than health costs. If the percentages of income required for premium payment by households were held constant, the share of the premium paid by households would fall over time (and, conversely, the government’s share would rise). This adjustment is clearly intended to keep the proportion required from households and the government roughly constant over time and prevent the government’s contribution from rising even faster than health costs.
After 2018, the law says that, in addition to the adjustment made to reflect premium growth in excess of income, the percentages shall also be adjusted
to reflect the excess (if any) of the rate of premium growth estimated under subclause (I) for the preceding calendar year over the rate of growth in the consumer price index for the preceding calendar year.
Again, the law doesn’t say how the percentages are to be adjusted, or on what basis. Nor does it define what it means to “reflect” premium growth over the CPI.
One reasonable interpretation is that this second adjustment is intended not to maintain proportionality between the government and households over time but to require households to pay for premium growth in excess of consumer inflation. That would mean limiting the government’s contribution to growth in the CPI, and adjusting the household percentages accordingly to ensure full payment of the balance. This approach would have the virtue of some underlying logic. Household premiums would be adjusted to “reflect” the excess of premium growth over the CPI, which looks to be the objective of the provision.
But, apparently, that is not how the provision is being interpreted by congressional scorekeepers. The Congressional Budget Office’s cost projections show the exchange subsidies growing, on a per capita basis, at a rate that is just below 5 percent annually from 2019 to 2021 — which is above its long-term annual inflation assumption of 2.3 percent, but certainly well below historical health-cost growth rates.
What other way of making this second adjustment is possible? It might be that CBO is assuming an entirely different way of “reflecting” the excess of premium growth over the CPI on household premiums (although CBO has not issued any kind of official explanation of how it is interpreting this provision). An example can help illustrate what might be going on. Suppose premium inflation is 7 percent, and CPI growth is 2.3 percent. CBO could be assuming that the initial adjustment (premium over income growth) requires household premiums in the exchanges to rise by 7 percent. In addition, the second adjustment then requires premiums to rise by the excess of 7 percent over 2.3 percent (or 4.7 percent). Thus, household premiums would be required to rise by a total of 11.7 percent under this example. The percentages of household income used to set premiums would be adjusted accordingly to hit these new premium requirements.
Note that this method doesn’t make a lot of policy sense. It effectively double-counts health inflation in the beneficiaries’ premium payments. Premium growth rates already include health inflation above the CPI; adding the excess of premium growth over the CPI to premium growth simply counts health inflation twice.
Note also that doing the calculation this way means the government’s contribution will grow at widely varying rates, by income. In all cases, the government’s contribution will grow at a rate below health-cost inflation. And, in some cases, the government’s contribution will actually fall below CPI growth. For instance, if the total premium for coverage in an exchange is $18,000 in 2018, and if health inflation is 7 percent and the CPI is 2.3 percent, the government’s contribution would fall below the CPI for any household that was paying at least $9,000 toward the total premium in 2018.
There’s an additional complication. The law states — in a provision called “the failsafe” — that this additional adjustment to reflect premium growth over the CPI will only apply in years (after 2018) in which “the aggregate amount of premium tax credits under this section and cost-sharing reductions under section 1402 of the Patient Protection and Affordable Care Act for the preceding calendar year exceeds an amount equal to 0.504 percent of the gross domestic product for the preceding calendar year.”
CBO assumes that spending on the subsidies and cost-sharing credits will (“probably”) exceed the threshold, and therefore the additional adjustment is operative (see p. 35 of The Long-Term Budget Outlook).
The chief actuary of the Medicare program, however, has stated (see p. 5 of this analysis) that his office estimates the aggregate spending on the credits will be just barely above the 0.504 percent of GDP threshold in 2018. Consequently, according to the actuaries’ projections, the government’s vouchers in 2019 and beyond would rise roughly in concert with GDP, not the CPI.
To sum up: My column asserted that Obamacare’s exchange vouchers would be tied to consumer inflation. That is apparently not CBO’s interpretation of what the law requires (based on a review of CBO’s current cost projections), although the law is so vague and imprecise that alternative interpretations are certainly possible. The larger point of the column remains valid regardless. Critics of Rep. Paul Ryan’s Medicare plan are on the warpath about vouchers for private insurance falling below baseline expectations of cost growth. The same accusation can be leveled against Obamacare’s vouchers. Indeed, some Obamacare participants would get vouchers that grow at less than the rate of the CPI. And even assuming this apparent interpretation of what the law requires, it is easy to see how allhouseholds in Obamacare’s exchanges would be facing double-digit premium increases each and every year after 2018 if health costs continue to rise as rapidly in the future as they have in the past.