Tuesday, September 20, 2016

A few statistical anomalies in healthcareville

Struggling with a difficult post, I just made a little list in my mind of some statistical oddities I've stumbled across in the past year or so in the wonderful world of U.S.  health insurance. Some are mysteries, some are apparent errors, and some involve frames that may give the wrong impression. In no particular order:

1)  In 2016, CBO changed the way it counts Medicaid enrollees under age 65. No, there was not a surge of 16 million unanticipated enrollments, as reported more than once. Rather, CBO started counting "dual eligibles" under age 65 in the Medicaid total, along with enrollees in various limited benefit programs.  More here.

2. Oft-quoted stat: In 2017, the ACA marketplace may have only one insurer in almost one third of U.S. counties. Less quoted: 19% of the population lives in those counties. 72% of the population lives in counties with 3-8 insurers (Kaiser). This is not to sugar-coat bad news, but the effect of recent pullbacks on people should take priority over its effect on acreage.

3) The Commonwealth Fund's 2016 ACA Tracking Survey apparently undercounts the relatively affluent uninsured. According to Commonwealth, just 6% of the uninsured are in households with incomes over 400% FPL. According to the Census Bureau's latest Current Population Survey, 20% are over that income threshold (See Table A-4, here).

4) The CPS, in its turn, appears to undercount the percentage of poor people who have gained public-program insurance since full ACA implementation.  That is, from 2013 to 2014, it shows bigger gains in private insurance among the nation's poorest than in public insurance, notwithstanding an increase of about 6 million Medicaid enrollees as of April 2014. I explored the anomaly in some detail here (last year's numbers*).  Emma Sandoe, formerly an HHS spokeswoman and now a public health grad student, speculates that many enrollees in managed Medicaid programs report that they have private insurance because such insurance is branded with the MCO's name, e.g., United Healthcare or Aetna.

5) The Kaiser Family Foundation's twice-yearly estimate of the percentage of marketplace enrollment as a share of potential marketplace population stuns at first glance: the current nationwide count is 40%. But that's because Kaiser counts in its estimate of the potential market "all individuals eligible for tax credits as well as other legally-residing individuals who are uninsured or purchase non-group coverage, have incomes above Medicaid/CHIP eligibility levels, and who do not have access to employer-sponsored coverage"  The enrollment estimate, however, does not include those enrolled off-exchange, most of whom are in risk pools combined with the insurer's marketplace customers.   A better indicator is Kaiser's estimate of potentially subsidy-eligible enrollees who are currently enrolled. That stood at 64% as of March 31. The  ACA-compliant market as a whole is at about 70% capacity.-- not good enough, but a lot better than 40%.

6. While a lot of marketplace enrollees are in high deductible plans, a lot aren't. Thanks to Cost Sharing Reduction (CSR) subsidies, slightly less than half of enrollees are in plans with an actuarial value of 94% or 87%; add in gold and platinum enrollment, and a bit more than half are in plans with AV over 80%, roughly the ESI norm. On the other hand, high AV enrollees almost all have incomes under 200% FPL, the cutoff for strong CSR. Off-exchange, higher bronze plan selection is partly offset by higher gold and silver selection, and average AV hovers around 70%. Average off-exchange AV may go down this year as unsubsidized premiums spike -- unless bargains appear off-exchange. More here.

Well, that was fun. Teeth, back to tougher nut.

--
**This year's CPS ASEC actually shows a drop in the number insured with incomes under 100% FPL, but that's presumably because the number of people below the poverty line dropped. The uninsured rate under 200% FPL dropped two percentage in 2015, with gains recorded as evenly split between private and public insurance.

First published 9/19

No comments:

Post a Comment

Share