Jed Graham has an important piece in the new Investor’s Business Daily, and a detailed follow-up blog post, on an important, little-understood provision in PPACA. Many PPACA advocates have criticized Chairman Ryan’s Medicare reform proposal on the sensible grounds that it doesn’t grow premium support at a sufficiently high rate. PPACA, in contrast, grows premium support at a rate above CPI-U, though somewhat lower than CPI-M. Many see this as a sensible, difference-splitting growth rate that creates an incentive for cost-consciousness and business-model innovation while making realistic assumptions about what it will cost to deliver a reasonably comprehensive benefit.

But as Graham explains, the law’s indexing provision is much more complicated than that:

Here is an example to understand how the provision works: The rate of premium cost growth (say 4.3%) is added to the amount by which premium costs exceed inflation (say 4.3% minus 2.3%) for a total of 6.3%. Then increase the prior year’s premium contributions by that 6.3%.
In this example, individual payments would rise 6.3%, well above the 4.3% rate of health cost growth. How much would subsidies rise? It all depends on the percentage of the overall premium that the government is paying. For the lowest income exchange members as well as for the oldest, subsidies would continue to rise faster than inflation.
But once the individual share of the premium tops 50%, subsidy growth would begin to lag inflation. And once individuals cover about 70% of the total premium, subsidies would shrink in nominal terms.
On its face, then, the indexing provision seems to be progressive. But that’s only if you ignore how the exchange is meant to work — by pooling the risk of the younger and healthier with the older and less healthy.
Under the example in the chart above — that assumes 4.3% premium cost growth, perhaps not a best-case scenario but not too far from it — the premium subsidy for a 42-year-old earning 300% of the poverty level would be sliced in half in nominal terms between 2014 and 2034. Meanwhile, the share of income paid for premiums would double from 9.5% to 19% — even before any cost sharing for utilizing medical services.
If premiums get too expensive — and far more costly than the mandate penalty capped at 2.5% of income — the relatively young and healthy will simply opt out, driving up costs for everyone else. That of course would spur more people to bow out.

We’re assuming that the broader U.S. health system won’t adapt to cost pressure, which seems unlikely. But I found Graham’s analysis very interesting all the same. (To clarify, I mean that keeping premium support growth lower than the underlying level of cost growth could create constructive pressure to create lower-cost approaches to delivering care.)