• And, quietly, we slip down the wonk hole

    Last week I wondered about this passage in Henry Aaron’s recent NEJM paper:

    The ACA’s subsidies fully protect people receiving them through 2017 against growth of health insurance costs beyond specified fractions of income; thereafter, they provide partial protection against premium growth.

    I’m almost sorry I asked about this. As I tried to unravel it, I felt like I was slipping down a hole, a deep, dark wonk hole. I’m nearing the bottom and the shaft of light is so dim it barely illuminates the complex details and is certainly not sufficient to tease out any meaningful intuition.

    Less whimsically, here’s what I did: I corresponded with a former CBO analyst who worked on health reform. In addition, I exchanged email with Henry Aaron who himself had corresponded with another former CBO analyst. Aaron (which, in this post, means Henry Aaron, not co-blogger Aaron) sent me a very long and detailed email exchange that included an explanation from his CBO contact. I also poked around in the law and other official reports, but that wasn’t terribly helpful. Aaron was enormously helpful.

    Conclusion: it’s very ugly sausage. I will not even attempt to describe it. Not now. In time I might, but I’ll do it in charts, not words. In truth, I’m secretly (now publicly) hoping someone else does the work. For now, I’ll just say that Aaron has a point. I think he’s right. Here are some stylized facts based on the details:

    • Through 2017, subsidies are forced to grow at the pace of premiums, so there is no erosion of their purchasing power.
    • Subsidies may grow more slowly after that, but, if so, the rate of growth would vary by income level. They would fall behind premiums more quickly over time at higher incomes. So, it’s a progressive erosion of purchasing power.
    • This erosion doesn’t happen at all unless spending on exchange subsidy premiums exceeds a certain fraction of GDP.

    Whether or not subsidies will erode over time relative to premiums is a crucial question, particularly for those of us who have critiqued subsidies proposed by others (for Medicare beneficiaries) as doing just that. At this point, whether the ACA’s exchange subsidies will erode, shifting more costs to individuals, is uncertain. A conclusion would have to be based on a simulation analysis, and I’ve not seen one. But, it is possible for them to erode. That may be problematic and is, no doubt, something that will be revisited in the future.


    • so glad someone finally wrote about this. been worried about this for months! glad i’m not the only one.

    • “•Through 2017, subsidies are forced to grow at the pace of premiums, so there is no erosion of their purchasing power.”

      This isn’t true from the consumer’s perspective (which differs from the federal government’s perspective). Take someone at 200% of FPL, in 2014 and again in 2015. This person’s income rises in proportion to the poverty rate (which is not directly related to purchasing power). For the reference silver plan, the law says they owe 6.3% of their income, with the subsidy making the difference.

      S. 1401 of PPACA as amended by HCERA adjusts the 6.3% target upward to reflect the excess of premium growth over income growth in the preceding calendar year. In the 11/30/09 Preimium analysis produced by the CBO, that’s why they used a *6.5%* of income contribution in 2016 (Note: that CBO 11/30 model result didn’t deal with the Manager’s amendment or HCERA changes, but the structure of the clawback pre-2018 is materially the same). To illustrate at the extreme, at some ridiculous point in the future this approach may have them paying more than 100% of their income toward premiums.

      • Right. Proportion of premium owed and proportion of one’s income paid are different. Part D’s approach is to keep the former constant. I’m not aware of any existing US health program that keeps the latter constant. Are you?

        UPDATE: I guess in a trivial sense Medicaid or the VA keep premium as a proportion of one’s income constant: $0. But I’m trying to think of a health program that sets subsidy levels such that the beneficiary pays a nonzero contribution that stays a fixed proportion of one’s income. Does one exist in the US?