Thursday, November 05, 2015

An NHIS non sequitur

Back in February, I complained that Gallup had conflated the effects of a state expanding Medicaid with the effects of a state running its own ACA exchange. All states running their own exchanges but one (Idaho) accepted the Medicaid expansion, making the conflation easy. But their superior performance in reducing the uninsurance rate was due entirely to the Medicaid expansion, not the decision to create their own exchanges. State-run exchanges got good press early on because most were not as dysfunctional as healthcare.gov in fall 2013. But some were even worse, and by the end of open season 2015 they had not collectively outperformed the federal exchange with regard to enrolling their potentially eligible populations.

CDC today released results of the National Health Interview Survey for the first 6 months of 2015. It's a trove of information about who's gained insurance over the last two years and who's still uninsured. But it contains one apparent non sequitur that recalls Gallup's:
Health insurance coverage by state Health Insurance Marketplace type

Under provisions of the ACA, states have the option to set-up and operate their own Health Insurance Marketplace,rely on a Federally Facilitated Marketplace operated solely by the federal government, or have a hybrid partnership Marketplace that is operated by the federal government but within which the state runs certain functions and makes key decisions. In the first 6 months of 2015, adults aged 18–64 in states with a Federally Facilitated Marketplace were more likely to be uninsured than those in states with a state-based Marketplace or states with a partnership Marketplace...In the first 6 months of 2015, there were no differences in the percentage of adults aged 18–64 with private coverage by Marketplace type (emphases mine). 
The blue text effectively negates any implications one might be tempted to draw from the red text.

Perhaps the equal rates of private insurance in the three state categories mask higher rates of employer-sponsored insurance and lower takeup in the ACA marketplace in states with federally run exchanges? No. Kaiser Family Foundation provides an estimate of the percentage of potentially eligible marketplace customers each state has enrolled. The national average is 36%. The average in the fourteen states (including D.C.)  that ran their own exchanges in 2015 is...36%. In the partnership states, it's 32%.  (The average are low because Kaiser counts all adults under 65 who don't have public or employer-sponsored insurance as potentially eligible, including those who buy insurance off-exchange in the nongroup market.)

The state-run exchanges have been a mixed bag. Some have failed, some have provided a consistently better experience than healthcare.gov, some a consistently worse one. Some are better in some respects and worse in others. You would think that the states that run their own exchanges, where state government is committed to maximizing enrollment, would do a better job reaching out to the uninsured, but that has not always proved to be the case. Other factors are at work: marketplace plan takeup is highest at the lowest income levels, and in states that refused to expand Medicaid, the eligibility threshold is lower (100% of the Federal Poverty Level instead of 139% FPL).

Whatever the mix of factors, comparing the overall drop in uninsurance in states that run their own exchanges with that of those who rely wholly or partly on the federal government does not seem particularly illuminating. Again, the great divide is Medicaid expansion.


No comments:

Post a Comment