Plan domination in the CSR-94 variants

Earlier this week, Another Scott asked a good question about plan domination:

How many plans are really “dominated” like this – identical in every respect but one costs substantially more?  Is it just the 6-sigma tail?  Do differences in “network” come into play – does one have to pay more to get Dr. Kildare or Dr. House on the team?  Is there an easy way for someone to know if they’re getting a “dominated” plan or not?

I was curious as I knew it was not a six sigma event (1 in a million) on the ACA individual markets.

Yesterday, I sent a draft paper to some colleagues for a friendly pre-submission comment and review.  In that paper, my co-author and I are looking at a domination scenario that is more likely than the Pittsburgh Pirates winning the National League Wildcard next year.  The scenario that my co-author and I are poking at is due to a policy shock.  I think there are numerous dominated plan offerings being made during the normal course of operations.

We’ve longed looked at insurers offering multiple baseline silver plans in a region.  Each of those plans are required to be “meaningfully different.”  I have never thought that the differences are sufficiently different to be meaningful in pragmatic terms but there are at least slight differences in plans offered from the same insurer on the same network with the same plan type (EPO, HMO, PPO etc) that there are chances that one at least one attribute of premium and cost sharing that domination won’t happen.

However silver plans are required to also have three CSR variants at 94%, 87% and 73% actuarial value.  There is a small range of allowable de minimas variation of one actuarial value point.  Some insurers may elect to have multiple silver variants that are different at the base level but only build out a single CSR variant that is then attached to each base level plan.  This is an administrative simplification.  I was curious as to how often this happened for the CSR-94 plan types on Healthcare.gov in 2019.  I conducted a quick and dirty analysis that is biased towards over-finding dominated plans.

METHODS: I used the 2019 HIX COMPARE file, matched on -06 Silver plan variants, on-marketplaces (1&3) and then on HIOS ID, Network ID, Plan Type to generate a list of paired CSR Plan IDS.  I then took each dyad and linked by 14 character PLAN ID to the Healthcare.gov Individual Market Plan Landscape PUF.  A dominated plan was identified if two plans from the same insurer had different premium but same 9 field cost sharing on the Landscape PUF file and same network ID and plan type.  I removed from consideration any plan that had either or both “Dental” and/or “Vision” in their plan names as these characteristics may be meaningful to buyers but would not show up in the Landscape cost sharing attributes.

RESULTS: At least one CSR-94 plan dyad was dominated in about 1 in 6 counties on Healthcare.gov in 2019.

A dominated plan choice for the CSR-94 is presented on-Exchange to individuals earning between 100-150% FPL in all of Hawaii, Maine and both of the Dakotas, most of Illinois and Virginia, large chunks of Georgia and then here and there throughout the rest of the country including Greater Houston and Metro Atlanta so it is not just a rural/low population county story.

Why does this matter?

In general, anyone choosing a dominated plan is paying more money for the same value.

In the CSR-94 context, a dominated plan choice can increase monthly premiums that an individual is paying when that individual is earning only between 100% FPL (~$12,140 for 2019 purchasing decisions) to 150% FPL ($18,210).  An individual buying a dominated plan for an extra $21/month would be spending 2% more of their monthly income than they need to.  For someone who is just getting by a needless 2% of income swing in expenditures to buy no additional value is painful.

So going back to Scott’s initial question, I think dominated plan choice offerings are common enough on the ACA to be pragmatically notable and meaningful.  Plan domination should be  deserving policy attention instead of mere academic attention.