Tuesday, March 17, 2015

Medicaid expansion means a richer QHP buyer pool on ACA exchanges

In my close look at silver plan selection in 2015 among healthcare.gov customers who were eligible for Cost Sharing Reduction (available only with silver plans). I expressed some disappointment that CSR takeup had apparently dipped a bit on the federal exchange from 2014 to 2015 (HHS did not provide CSR takeup numbers for states operating their own exchanges). Disappointment on that particular point may have been misplaced.

While I continue to believe that too many low income ACA private plan buyers selected bronze plans, CSR takeup probably did not decline from 2014 to 2015. In fact, since silver plan selection across all exchanges ticked up a bit, from 65% in 2014 to 67% in 2015, CSR takeup probably did too.

My perception that CSR takeup dropped on healthcare.gov in 2015 stemmed from a drop in silver plan selection from 76% to 74% by buyers eligible for any kind of subsidy --including those eligible for premium subsidies but not CSR. But that drop may just reflect a shift in the composition of the market using healthcare.gov -- specifically, it may reflect a higher percentage of customers in healthcare.gov states being placed in Medicaid in 2015. This happened because two states that dropped their own exchanges and joined healthcare.gov in 2015, Oregon and Nevada, had expanded Medicaid, while three other states on the federal platform implemented the Medicaid expansion at some point during 2014 (Michigan and New Hampshire) or as of Jan. 1, 2015 (Pennsylvania).

The poorer the buyer pool, the higher the CSR takeup

States that expanded Medicaid generally have lower CSR takeup because their buyer pool for private health plans is wealthier. In states that refused the expansion, the buyer pool starts at 100% of the Federal Poverty Level (FPL); in states that embraced the Medicaid expansion, it starts at 138% FPL. In hc.gov states that refused the Medicaid expansion, consequently, an astounding 50% of private plan buyers had household incomes under 150% FPL; in expansion states, just 25% had incomes below that level.

CSR is strongest at incomes below 150% FPL -- at that income level it raises the actuarial value of a silver plan to 94%, better than a platinum plan without CSR. At the same time, silver plan premiums are most affordable for those under 138% FPL, reduced by subsidy to just 2% of income (rising to 4% at 150% FPL and 6.3% at 200% FPL.  Thus those private plan buyers who were shut out of Medicaid in non-expansion states have the strongest inducements to buy silver plans.

Given these inducements, one would expect CSR takeup to be higher in states that refused the Medicaid expansion than in those that embraced it, and highest in poor states that refused the expansion. And that is usually the case.  In all healthcare.gov states taken together, 60% of buyers accessed CSR -- that is, had incomes under 250% FPL (the eligibility cutoff) and selected silver.  In the fourteen healthcare.gov states that expanded Medicaid, just 49.7% of buyers accessed CSR.

With some exceptions, the poorest states that refused the Medicaid expansion had high CSR takeup. Here are the numbers for the ten non-expansion states on the federal platform with the lowest median household incomes (as of 2013):

Median HH income
% buyers with CSR
North Carolina
South Carolina
Group average
National or hc.gov
$51.9k (national)
60 (37 hc.gov states)

Compare six low-income states on healthcare.gov that have expanded Medicaid:

Median HH income
% buyers with CSR
West Virginia
New Mexico
Group average

While the average median household incomes for the two groups are nearly identical, the nonexpansion states show 12-point lower CSR takeup -- 52% vs. 64%.

Two states that joined hc.gov in 2015, Nevada and Oregon -- both Medicaid expansion states -- had CSR takeup rates of 55% and 47% respectively. Their addition to the healthcare.gov pool must have at least slightly depressed the proportion of buyers from 100-150% FPL in the whole hc.gov pool (HHS did not provide state-by-state income breakouts). Ditto for the states that expanded Medicaid after the Oct. 2013--April 2014 open season ended but in time to affect enrollment in 2015. Taken together, the three hc.gov states that expanded Medicaid since the first open season closed have added about 700,000 to their Medicaid rolls. Its hard to say how many of those came out of the private plan pool. But navigators working with poor populations in Philadelphia have told me that almost all of their 2014 private plan clients (almost all of whom ended up in CSR-enhanced silver) are eligible for Medicaid this year.

None of this proves that  the buyer pool on hc.gov was a little bit richer in 2015 than in 2014  -- or more to the point, that a lower proportion of buyers had incomes under 138% FPL, where they were likeliest to buy silver. HHS did not provide income breakouts or CSR takeup numbers in 2014. The percentage of buyers eligible for any kind of subsidy was nearly identical year-over-year, but buyers may have been slightly more concentrated in higher subsidizable brackets in 2015. If so, a slight drop in the percentage of subsidy-eligible buyers who bought silver would make sense. The higher up the income brackets you go, the lower the proportion of silver buyers.

The most notable across-the-board change in metal level selection from 2014 to 2015 was a rise in bronze plan selection. On the federal exchange, silver selection among all buyers stayed flat at 69%, but bronze selection rose from 17% to 21%. Among subsidized hc.gov buyers, bronze selection rose from 15% to 19% (as previously noted, silver dropped from 76% to 74%) On all exchanges, silver selection rose from 65% in 2014 to 67% in 2015, but bronze takeup also rose, from 20% to 22%.

The bronze increase came mainly at the expense of gold and platinum, which suggests that it probably didn't affect CSR takeup, since CSR-eligible would mostly be unable to afford gold and platinum plans. The one exception is among subsidized hc.gov buyers, where silver selection did drop. That brings us full circle: again, that drop may be due to a lower overall percentage on healthcare.gov of buyers under 138% FPL.

Incidentally the removal of those in the 100-138% FPL income band from the QHP buyer pool in Medicaid expansion states goes a long way toward explaining the lower silver selection rates in state-run exchanges. Not only are those states wealthier on average, their poorest potential buyers are out of the QHP pool.

1 comment:

  1. Very interesting comparison. The difference in CSR uptake is pretty profound. Thanks!