Nothing matters a lot in California

Big, geeky technocratic tinkering news out of California last night:

New via @InHealthPolicy: CA to use reserve $ to "cover the $1 premium for abortion coverage required under the ACA." Note: these $1 separate premium payments are not required in statute but rather a burdensome Trump-era regulation. @acog continues to call on @CMSGov to revoke.

— Emily Eckert (@emilyanneck) May 18, 2021

California requires that all plans sold in the ACA individual marketplace have non-Hyde elective/voluntary abortion benefits as a covered benefit. This is considered a “non-essential health benefit.” The ACA has two relevant requirements by law. First, the advanced premium tax credits are only applicable to “essential health benefits.” Secondly, the non-Hyde elective abortion coverage must have a minimum monthly premium of $1 that is put into a dedicated sinking fund so that the money never touches federal funds. The Trump Administration then tried to weaponize administrative burden by requiring two separate payments for individuals who had non-Hyde covered benefits. The first payment was for the essential health benefit and non-abortion related component of non-essential health benefits. This is usually a decent sized check. The second payment was a $1 monthly premium for the abortion related coverage costs.

Under the original ACA subsidy system, a lot of people in California had a $1 plan as their subsidies were big enough to wipe out their obligations to pay for the essential health benefits component of their plan. With the new ARA subsidies, even more people have a $1 plan as their cheapest plan. Now, under this proposal to use state funds to pay for the $1 non-essential health benefit, hundreds of thousands of Californians will now be exposed to a zero premium plan.

Coleman Drake and I found an association between county level presence of a zero premium plan and enrollment among the 151-200% FPL group in a Health Affairs paper from January 2020. We estimated that the presence of a zero premium plan in a county increased enrollment in this subgroup by about 14%.

New work by myself, Coleman, Sih-Ting Cai and Dan Sacks argue that the zero premium effect is not a special price effect (cheaper is better all else equal though) but an administrative burden effect. We looked at zero premium plans in Colorado where there was a tremendous amount of policy driven variation in the prevelance of zero premium plans. We found little differences in enrollment or plan choice among people whose cheapest plan was zero by a little bit and people whose cheapest plan was just over zero by less than the price of a mediocre cup of coffee. Instead what we found was that people who had a zero premium plan were far more likely to start their coverage on January 1st compared to folks who had a nominal premium. Folks who had nominal premiums were more likely to start their coverage in February and March. Colorado’s open enrollment period runs through January, so we think people tried to get coverage, and for folks who had a zero premium plan as their choice, they just had to pick a plan and it started automatically. However, folks who had to pay nominal premiums had to first pick a plan, second make a payment arrangement and then start their coverage. That middle step tripped up a lot of people. McIntyre, Shepard, and Wagner see a similar mechanism in automatic retention in zero-premium plans in Massachusetts.

Zero premium plans get rid of one source of administrative burden and transactional frictions. Given that California, like Colorado, runs their open enrollment through January in non-COVID years, zero might matter less than in other states but it still mattered a lot in Colorado.