Friday, March 25, 2016

Are some Medicaid-eligible Indianans still in subsidized private health plans?

Phil Galewitz of Kaiser Health News has a progress report on Indiana's conservative-flavored Medicaid expansion, which obligates all enrollees to make small monthly payments ($1-29) into a health savings account. If they don't pay, they either lose benefits for a period or lose dental and vision coverage provided to paying enrollees.

Critics decry the bureaucratic complexity and also worry that the payments discourage some from participating. And enrollment has apparently not been as gangbusters as in some states:
Since Indiana’s expansion began in February 2015, more than 235,000 previously uninsured able-bodied adults have signed up. As of late February, the plan covered more than 370,000 people total, many with extremely low incomes. Another 190,000 adults are eligible but not enrolled, according to state estimates, though some who are above the poverty line may be in subsidized private plans.
It's true that some states swiftly exceeded their enrollment targets upon expansion, Michigan and Ohio among them. Galewitz relays some possible inhibiting factors:

Joan Alker, executive director of the Georgetown University Center for Children and Families, said the red tape in Healthy Indiana exceeds that of any state’s Medicaid expansion. Few third parties, such as employers and nonprofit groups, have offered to help individuals cover their monthly contributions, as the state had hoped, Alker noted.

She questions why so few eligible people above the poverty level have not enrolled. Many may have signed up for subsidized Obamacare marketplace plans in 2014 and could now be paying more than necessary, she said
More detailed enrollment data suggests, I believe, that the second problem cited here is not having a strong effect. That is, I doubt that there a lot of people still enrolled in subsidized private plans who should be in the Medicaid program -- though I explored the same worry for Pennsylvania when the state launched a rather rocky expansion in December 2014.

Here's the evidence. In states that refuse the Medicaid expansion, eligibility for subsidized private plans begins at 100% of the Federal Poverty Level (FPL). In expansion states, eligibility begins at 139% FPL; those below that level are eligible for Medicaid, and not for private plan subsidies.

HHS breaks out private plan enrollment by income in states that use, as Indiana does. The relevant income band is 100-150% FPL. About two thirds to three quarters* of enrollees within that band have incomes under 139% FPL and so are eligible for Medicaid if the state has embraced the expansion.  (Unfortunately, HHS does not break out the 100-138% FPL group separately in the nonexpansion states.)

In states that have not expanded Medicaid, an outsized percentage of private plan enrollees are in the 100-150% FPL income band. In Florida this year, it's 55%; in Mississippi, 60%. In all 38 states on the platform, half of which have expanded Medicaid, 38% of enrollees have incomes in the 100-150% FPL range (Florida and Texas, where 44% are in the 100-150% FPL band, carry outsize weight).

In Indiana in 2016, just 21% of enrollees are in the 100-150% FPL income range. That's slightly higher than the norm in expansion states, which average 17.3% of enrollees in this band. It's also lower than Indiana in 2015, when 29% of enrollees were in the 100-150% FPL band. Enrollment for that year lasted from Nov. 1 to February 15 (with a few days' extension), while the Medicaid expansion became effective on Feb. 1. So it's likely that many soon-to-be-Medicaid-eligibles did not enroll in subsidized private plans. Unfortunately, HHS did not provide an income breakout for 2014,so we have an imperfect comparison.

Normally, a Medicaid-eligible applicant would not be able to enroll in a subsidized private plan. Because the Indiana expansion began in February 2015, however, while open enrollment extended from Nov. 1, 2014 to mid-February 2015, some Medicaid-eligibles may indeed have stayed in private plans. Because the Indiana expansion did not begin on Jan. 1, enrollees who either auto-renewed or actively re-enrolled in late 2014 were told that they were not eligible for Medicaid and instead enrolled in subsidized private plans. In early 2015, the state of Indiana sent out notices warning that those who were eligible for Medicaid needed to change their enrollment or else pay back their premium subsidies. But the federal Treasury and HHS reversed that order, ruling that a determination of ineligibility for Medicaid would stick throughout the year if the person did not apply for Medicaid.

It's possible that if such an enrollee allowed her 2015 insurer to auto-enroll her in 2016, without logging onto and updating her data, she might still be enrolled in a private plan rather than Medicaid. But the number of such enrollees is probably. small.


*  I've estimated that slightly over two thirds of enrollees in the 100-150% FPL band are under 139% FPL, extrapolating from 2015 HHS data that contrasted the rates in nonexpansion states with those of expansion states. Charles Gaba has estimated that three quarters are in Medicaid range, on the simple basis that 38 is three quarters of the way from 100 to 150 -- but I can't find the post. For all I know the simple method is better.


  1. "In Indiana in 2016, just 21% of enrollees are in the 100-150% FPL income range. That's slightly higher than the norm in nonexpansion states, which average 17.3% of enrollees in this band"

    Confused. Prior paragraph, FL and TX with 44% with agreed, outsize wt. Yet, when you avg them in, 17.3% is the composite? That seems off.

    Also, on Indiana Mcaid POWER accts. Its interesting how against the grain their early response has been with contributions and durability of response. Whole of empiric lit seems to be OOP costs take a bite. Time, ? 1 yr or so, might decay enrollment. If economy tanks more so, but most of these folks have already been and are still being hit hard, so who knows. Want to keep an open mind. If their data is clean, maybe many observers, including me, will change their priors.

    1. Oops! That was supposed to be "expansion states," Fixed, thanks.