Inverted linkages of gross and net premiums in a price linked subsidy scheme

We’ve been talking a lot this week about the ACA price linked subsidy system.  I’m hammering this point home because the subsidy structure as it is currently written into law has weird incentives and linkages.  We need to understand these linkages for spread sensitive cohorts to understand the lived experience on the ground with current policy and proposed future policy.

Let’s build a simple model.  Let us assume that there are two buyers.  The first buyer is not subsidized.  They are level sensitive. The second buyer is subsidized.  They are expected to pay $150 in net premium for the benchmark plan.  There are three classes of plans available. There are plans priced below the benchmark.  There is a plan priced as the benchmark.  Finally there are plans priced above the benchmark.

We’ll run this universe three times.  The first time is our baseline.  The second time there is a general 20% reduction in premium from the baseline.  The third scenario has a 20% increase in premiums for all plans.  (All numbers are hypothetical and we should seldom expect uniform percentage change in premiums but as a teaching example, this is good enough for a first pass to develop intuition)

We get what our intuition tells us to expect for the level sensitive, non-subsidized buyer; when premiums go up, they pay more, and when premiums go down they pay less.  This is simple.

It gets messy with subsidized buyers.  A change in premium level  affects subsidized buyers differently depending on what class of plan they buy.  Subsidized buyers who purchase the benchmark plan pay the same price in all three scenarios.  They are completely indifferent to premium levels.

But they are the only people who are completely indifferent to premium levels as changings in levels becomes a change in spread.

Lower premium levels reduces the spread between the benchmark plan and the below benchmark plan.  Lower premium levels increases the net premium paid by people who purchase lower cost plans.  These folks are worse off.  However, people who want to purchase a plan that is priced above benchmark will see their net premium decrease when there is a general reduction in all plan premium levels.  These folks are better off. They might be buying a richer benefit plan (gold instead of silver) or a broad network PPO instead of a narrow network HMO, but they get more for less premium.

Higher premium levels increase the spread  between the benchmark and the below benchmark plan.  Really high premium levels can theoretically reduce the net premium paid for below benchmark plans to zero.  Below benchmark plan buyers are better off with a high general premium level.  Buyers of above benchmark plans are worse off as they pay more in net premiums every month when premium levels increase.

Within the context of the ACA individual market, most subsidized buyers are purchasing either the benchmark plan or a plan priced below benchmark.  This means that most of the population is either indifferent or actively benefits from higher premium levels.  It is not everyone.  Anyone who buys a platinum plan benefits from lower premium levels, and people who are fairly broad network non-restrictive  and low cost-sharing plans are almost always worse off when premium levels increase.

These dynamics are complex. They are confusing.  And they shape the distributional consequences as well as the operational mechanics of most of the ACA individual market now and under most plausible reform plans.  We need to understand these mechanisms well.