Risk adjustment on Exchange is supposed to minimize the cherry picking incentive of insurers to only try to cover healthy people. Insurers with healthy populations pay net transfer payments to insurers with comparatively unhealthy populations. However there are significant problems with counting on risk adjustment to solve all problems. The biggest issue is the use of average premiums to calculate the size of the transfer payment. This can be functionally decent for Medicare, Medicaid and CHIP where there is a very narrow pricing band for services. However on Exchange, there is a wide spread of provider pricing.
Let’s assume that we need to take care of a serious case of itch elbowitis and that IE is a recognized risk adjustment disease category. On Monday, a surgeon can treat itchy elbowitis with a full cure. The patient has insurance through the Exchange with a low cost narrow network so they have the surgery done at a suburban community hospital. The total bundle of care has a contracted rate of $8,000. On Tuesday, the same surgeon does the same surgery with no complications at the same hospital. The patient this time has a broad network, high cost insurance policy. The total bundle of care has a contract rate of $12,000. The average incremental cost of treating the condition is $10,000.
I’m going to make several massive simplifying assumptions (we don’t need to look at average premium level in the risk pool nor demographic versus medical risk).
Let’s assume that itchy elbowitis creates a $10,000 transfer credit. The low cost insurer pays out more than the cost of their treatment regime for any itchy elbowitis obligation while the high cost insurer receives less than the cost of their treatment regime.
This is primarily a problem for the high cost insurer as they are attracting a less healthy risk pool than the low cost/narrow network insurers. They are getting some compensating side payments in to balance some of the books on treatment but the side payments are not large enough. This means their premiums need to rise at a faster rate than the general premium increase in a given market. Over the long run, I am having a hard time seeing how a high cost, broad network insurer stays on Exchange even with risk adjustment when there is a wide variance in accepted contracted rates for common procedures.
Risk adjustment has other issues (partial year credits is the big one) but as a policy side, as long as there is wide variance in provider payments for similar procedures, the market will converge to a common price point. And given the subsidy structure, that price point will be Medicare plus a little bit in most regions where there is a competitive insurance market.