New Medicaid expansion + increased silver loading should improve ACA marketplace risk pools

Over the next couple of years, two forces will be at work to make ACA marketplace risk pools somewhat wealthier and therefore healthier -- even as other factors pull the other way.

The factors working to improve the risk pool are Medicaid expansion, which pulls the lowest income  enrollees out of the marketplace, and silver loading*, which creates discounts in bronze and gold plans that mainly benefit (and draw in) enrollees at the upper income range of subsidy eligibility (201-400% of the Federal Poverty Level).

At incomes below 201% FPL, the bronze/gold discounts for the most part are not valuable enough to offset the free, strong Cost Sharing Reduction subsidy available at that income level, and available only with silver plans. At 200-400% FPL, the discounts were attractive enough to offset factors inhibiting enrollment for 2018, such as the huge advertising and outreach cuts implemented by the Trump administration and confusion about repeal and the individual mandate. As I've noted before, compared to 2017 enrollment levels, 2018 enrollment  at 201-400% FPL outperformed enrollment below 201% FPL by about 8 percentage points. That in turn shifted the income distribution of exchange enrollees upwards a couple of percentage points:

Income distribution of subsidized enrollees: HealthCare.gov, 2017-2018

Year	Number enrolled at 0-200% FPL*	Percent enrolled at 0-200% FPL	No. enrolled at 201-400% FPL	Percent enrolled at 201-400% FPL
2017	5,507,246	59.8%	2,851,601	31.0%
2018	5,092,349	58.2%	2,891,851	33.0%

Source: Public Use Files (2018, 2017) published by CMS

Income distribution of subsidized enrollees: Covered California, 2017-2018

Year	Number enrolled at 0-200% FPL	Percent enrolled at 0-200% FPL	No. enrolled at 201-400% FPL	Percent enrolled  201-400% FPL
2017	681,270	49.0%	564,370	40.6%
2018	673,680	46.9%	602,000	41.9%

Source: Covered California, Active Member Profiles**, March 2017/2018

I don't believe that anyone as yet has done a comprehensive comparison of silver loading effects in 2019 vs. 2018. David Anderson, however, has reported that the number of counties in which at last one gold plan premium is lower than the premium for the benchmark silver plan rose from 595 in 2018 to 1136 this year. The number of states instructing insurers to silver load, and/or encouraging them to offer off-exchange silver plans with no CSR load, has also increased this year. 12 states** changed strategy to either encourage/mandate silver loading for the first time or intensify to on-exchange-only silver loading. 

As for Medicaid expansion, a substantial new wave is coming. Virginia is now enrolling those rendered newly eligible for 2019. Maine, having just elected a Democratic governor, is poised to honor its 2017 plebiscite and begin enrollment early next year. Nebraska, Utah, and Idaho all passed referenda last week calling for expansion. And newly elected Democratic governors in Kansas and Wisconsin will push for expansion, with fair likelihood of success.

In states that have accepted the Medicaid expansion, residents with incomes up to 138% FPL are eligible for Medicaid (and therefore not eligible for marketplace coverage). In nonexpansion states, marketplace eligibility begins at 100% FPL, and more than a third of marketplace enrollees  have incomes in the 100-138% FPL range that would render them Medicaid-eligible if and when their states accept the expansion. 

In 2016, CMS broke out marketplace enrollment at 100-138% FPL for the first and only time, and  Charles Gaba charted the results for nonexpansion states. We learned that fully 36% of enrollees in nonexpansion states had incomes that should have qualified them for Medicaid (100-138% FPL), and that 82% of those with incomes in the 100-150% FPL band that HHS/CMS routinely reports were in this "should-have-been-in-Medicaid" category. Gaba's 2016 breakout enables an estimate of how many current marketplace enrollees in the seven states likely to expand in the next two years would be eligible for Medicaid once (or if) expansion is implemented.

The chart below shows 2018 enrollment at 100-150% FPL in these seven states, as of the end of Open Enrollment, and filters that total through a) the percentage of enrollees in that bracket whose incomes were below 139% FPL in 2016, and b) an estimated 15% attrition that should bring the total near to average monthly enrollment. 

Current marketplace enrollees with incomes that would qualify them for Medicaid in states likely to accept the ACA Medicaid expansion

State	Enrollment 100-150% FPL, end of OE 2018	% at 100-138% FPL (2016)	Likely enrollment 100-138% FPL	85% of total (estimated monthly avg enrollment)
Virginia	131,245	81.9%	107,490	91,367
Maine	18,516	81.6%	15,109	12,843
Nebraska	23,980	81.1%	19,448	16,531
Utah	58,383	77.1%	45,013	38,261
Idaho*	31,300	81.8%	25,700	21,845
Kansas	30,592	81.6%	24,963	21,219
Wisconsin	53,554	78.0%	41,972	35,676
Total	347,570		279,495	237,742

* Idaho, the only nonexpansion state that runs its own exchange, does not break out enrollment by income. I've borrowed Gaba's 2016 estimate, which used Kansas as a "comparable" for percentages of total enrollment. I've adjusted down for a 7% enrollment drop in Idaho since 2016.

The Medicaid expansion population is generally assumed to have more acute medical needs than higher income groups. A 2016 HHS study estimated that premiums in states that accepted the ACA Medicaid expansion are 7% lower than in states that refused the expansion, when diverse other factors are controlled for. Expansion in these seven states, if all do expand, would likely increase Medicaid enrollment by some 800,000 -- a major boost to public health that also will likely improve the individual market risk pool (while shrinking it considerably) in those states.

Both factors discussed here, along with reinsurance programs implemented by 7 states and under consideration by others, could have a positive effect on premiums and thus help stop the bleeding in the off-exchange ACA-compliant marketplace, where all enrollees are unsubsidized. Several states have moved to encourage on-exchange-only silver loading, which enables insurers to offer silver plans with no CSR load off-exchange.

Enrollment factors cutting the other way include repeal of the individual mandate, working in tandem with the Trump administration's creation of a parallel market of underwritten, lightly regulated health plans (so-called short-term plans, which now can be offered for terms up to 364 days and renewed twice), and further Draconian cuts to enrollment assistance and outreach. Whether those factors outweigh the effects of silver loading discounts and Medicaid expansion remains to be seen.

----

* Silver loading refers to concentrating the cost of CSR (footed by the federal government as stipulated by the ACA until Trump stopped payment in October 2017) in the premiums of silver plans, since CSR is available only with silver plans. Since premium subsidies, designed so that the enrollee pays a fixed percentage of income, are set to a silver plan benchmark (the second cheapest silver plan), inflated silver premiums create discounts for subsidized buyers in bronze and gold plans. And in states that allowed insurers to offer silver plans off-exchange with no CSR load, unsubsidized enrollees were protected from CSR costs, theoretically at least.

* *Spread sheet on silver loading strategy maintained by Charles Gaba, Louise Norris and David Anderson.