Sunday, October 15, 2017

60% of ACA marketplace enrollees with CSR are in nonexpansion states

In February of this year I noted that the 19 states that refused to implement the ACA Medicaid expansion comprised 38% of the U.S. population but 53% of ACA marketplace enrollment and 60% of enrollees who accessed Cost Sharing Reduction (CSR) subsidies.

That seems relevant now that Trump has abruptly cut off federal funding for CSR, stiffing insurers for the remainder of this year and leaving them to price the benefit into their 2018 premiums. I've pasted the whole of the post below. Enrollment stats are as of March 31, 2016. There's a state-by-state breakout at bottom of population, CSR enrollment and APTC (premium tax credit) enrollment. This year's enrollment breakout would be roughly proportionate: nationwide, the same 57% of enrollees access CSR. As of the end of the first quarter 2017, CMS reported 10.3 million total enrollees (which, per Charles Gaba, may well have been an undercount, leaving out up to a half million late enrollees), compared to 11.1 million at the same point in 2016.

Before getting to the repost below, a few notes:

1. Many observers are drawing a somewhat misleading conclusion from the high concentration of CSR enrollees in red states. They will be hurt by the CSR funding cutoff only insofar as insurers pull out of the market -- and, longer term, by the thinning of the risk pool caused by unsubsidized enrollment dropping off, which will reduce competition and so plan choice. Unsubsidized enrollees bear the brunt of the premium increases driven by the cutoff of federal funding for CSR. And most of the damage on that front is already done, as most insurers in most states filed rates assuming that CSR would not be reimbursed in 2018 (and/or accounting for more general politically induced uncertainty).

2. The low income enrollees so heavily concentrated in nonexpansion states will probably be hurt more by HHS's radical cutback in outreach and enrollment assistance than by the shift in who funds CSR, which will remain in place.

3. The cutoff of CSR reimbursement will cost the federal government money, because when insurers price in CSR, subsidies rise to cover the inflated premiums. They rise disproportionately, because most states have told insurers to concentrate the premium hikes in silver plans (the only metal level at which CSR is available), and income-based subsidies are keyed to silver plans. Kaiser estimates that the increased premium subsidy cost exceeds the savings from stopping CSR payments by 23%.

4. The cost to the Treasury of cutting off CSR payments is even more concentrated in the nonexpansion states than the 60%-of-CSR enrollment would indicate. That's because the lowest income enrollees, who get the highest level of CSR help, are concentrated in nonexpansion states, where the income threshold for eligibility for marketplace subsidies is 100% of the Federal Poverty Level, as opposed to 138% FPL in expansion states (where people below that level are eligible for Medicaid). Fully 36% of marketplace enrollees in nonexpansion states are in this "should-have-been-in-Medicaid" category. 43% of enrollees in these states had incomes under 150% FPL, the cutoff for the highest level of CSR - about 80% of enrollees in that category in the 38 states that used In this income bracket, CSR raises the actuarial value of a silver plan (the percentage of the average user's medical costs paid for by the plan) from a baseline of 70% to 94%. Insurers now must price in that extra AV one way or another.

Below is the previous post. Again, note the chart at bottom.

I have noted on multiple occasions that in states that refused the ACA Medicaid expansion, marketplace enrollment has been bolstered by a large contingent of people who "should have" been enrolled in Medicaid. That is, over 2 million marketplace enrollees in those 19 states have incomes between 100% and 138% of the Federal Poverty Level (FPL) -- incomes that would have qualified all of them except for certain legally present noncitizens for Medicaid had their states accepted the expansion.

To review a few facts about these low income enrollees:

  • In nonexpansion states, 36% of enrollees had incomes in the 100-138% FPL range. In mid-2016 that came to about 2.1 million enrollees.

  • Close to 90% (or more*) of those low-income enrollees selected silver plans and so accessed Cost Sharing Reduction (CSR) subsidies that raised the actuarial value of their plans to 94%. That generally translates to a deductible of $0-250.

  • Customer satisfaction in the ACA marketplace is much higher among enrollees who are not in high deductible plans (defined by survey conductor Kaiser as under $1500 for an individual). In Kaiser's 2016 tracking survey, 74% of enrollees in lower deductible plans rated their plans good or excellent, vs. 59% of those in higher deductible plans. And again, the vast majority of enrollees in the 100-138% FPL range are in low deductible plans.
All that said, a fact hiding in plain sight (to me, anyway) is the extent to which nonexpansion states are over-represented in the ACA marketplace. This is not surprising, since a third of enrollees in those states should have been eligible for Medicaid. Nonetheless, I find it rather startling, based on state-by-state data released by HHS in December, that as of the end of the first quarter of 2016, the 19 nonexpansion states contained:
  • 38% of the U.S. population
  • 53% of marketplace enrollees
  • 56% of subsidized marketplace enrollees
  • 60% of enrollees in plans with Cost Sharing Reduction
Here is the state data. I've added population as of 2014.

Enrollment and Population in 19 States that Refused the ACA Medicaid Expansion

Given the income distribution of enrollees in those states, it's doubtless also true that they contain a higher concentration of high-AV CSR enrollees than the expansion states -- that is, higher percentages of enrollees at AV 94% and lower percentages at AV 73%. I'm not so sure about the middle range CSR, AV 87%, available to those with incomes in the 150-200% FPL range.

A few more facts re the lowest income marketplace enrollees, concentrated in nonexpansion states:
  • Notwithstanding decent satisfaction ratings among low-income marketplace enrollees, Kaiser focus group results indicate that lower income ACA beneficiaries are more satisfied in and less confused by Medicaid than by marketplace coverage.

  • The high percentage of low income enrollees in nonexpansion states raises the cost of  insurance. A CMS study found 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.  In expansion states, eligibility for marketplace subsidies begins at 139% FPL, the cutoff for Medicaid eligibility. In nonexpansion states, marketplace subsidy eligibility begins at 100% FPL; those below that threshold are left out in the cold.

  • The low-income marketplace enrollees who "should have been in Medicaid" would be particularly ill-served by Republican ACA replacement plans, such as Tom Price's and Paul Ryan's, that offer tax credits adjusted by age but not income in a deregulated market in which high deductible plans and HSAs figure prominently.
Poor states that did expand Medicaid, most notably Kentucky, Arkansas and West Virginia, massively cut the ranks of their uninsured (by more than half); the 19 nonexpansion states, less so. Nonetheless, thanks to high marketplace enrollment in the 100-138% FPL income band, low income people in nonexpansion states will suffer disproportionately should the ACA be repealed without income-appropriate replacement at low income levels.


* CMS stats reported last summer indicate that 87% of enrollees with incomes in the 0-150% FPL range selected silver plans. The ratio is probably higher in the 0-138% FPL range, as those up to that threshold pay just 2% of their income for a silver plan, as opposed to 3-4% in the 138-150% FPL range. 

No comments:

Post a Comment