County level inequities in the ACA

Health wonks like to say that the ACA is not a single program but fifty one programs.  It works well in some states (California) and poorly in others (Arizona) and muddles through in most.  This is a good insight but it is an incomplete insight.  The ACA as experienced has some state level components such as the decision to expand Medicaid or not and risk pool/risk adjustment interactions but the Exchanges are experienced best at the zip code and county level.  Carriers can be in a state but only sell in a select number of counties.  Some zip codes can span multiple counties.  This matters.

Tennessee is an extreme example of the differing local experiences of the ACA that depends on the county a person resides in.  The map below is a county level map of the 2017 differences between the least expensive Silver plan and the benchmark Silver plan for a single 40 year old non-tobacco user.  A big gap means the cheapest Silver plan for a subsidized individual is a better deal after the subsidy.  Green is a big gap, red is a small gap.

tn-gap-and-spam

Perry County  is one of a cluster of counties in Central Tennessee that has the biggest spread in the country.  It is an extreme Silver gap at $97.62.  There is a single carrier, Blue Cross and Blue Shield of Tennessee that offers two Silver plans.  Roane County is the other extreme.  A single carrier is also in the market, Humana.  However Humana is offering only a single Silver plan which means the spread between the benchmark and the cheapest plan is zero.  There is no chance of a good deal for a Silver below benchmark.

What does this mean?

 

This is an extreme example. In Perry county, a 40 year old earning $25,000 a year, which is slightly more than 200% Federal Poverty Level (FPL), saves $100 a month or roughly 5% of their income because of the Silver Gap compared to the same individual in Roane County. Going to a broader scenario, a married pair of 40 year olds plus a ten year old kid can earn up to $38,600 (~190% FPL) before they have to pay a dollar for Silver plan with Cost Sharing Assistance. In Roane County, that same family is paying $182 per month or 5.6% of their monthly income for their Silver plan.

If that family of three is a pair of 50 years olds and a 10 year old, the Perry County family can earn up to $45,000 a year before they have to pay a dollar for their Silver plan.  Bronze plans are zero dollar premiums for this family when they earn just north of $75,000.

The same family in Roane County pays 7.2% of their monthly income for a Silver plan if they earn $45,000 a year.  Their zero dollar Bronze plan ends once they earn more than $43,500.

From a risk pool perspective, Perry County should be extremely healthy.  The subsidies are rich enough that almost everyone can afford a plan even if it is a minimal Bronze plan.  Roane County will see good uptake among people who earn under 150% FPL and decent uptake to 200% FPL.  Above those income levels, there will be significant adverse selection as the deals just aren’t too good for healthy people, so the population will be fairly sick and expensive.

More importantly, people in Perry County who are getting subsidized will see the ACA working really well.  They have good, cheap health insurance.  However their cousins across the state are getting a raw deal compared to the great deal that they get in Perry County.  This is especially true as we move up the income scale which means moving up the likely voter scale and influence scale.

 

 

Share On Facebook
Share On Twitter
Share On Google Plus
Share On Pinterest
Share On Reddit

9 replies
  1. 1
    Lit3Bolt says:

    Roane county.

    I know, it’s auto correct.

    Thanks for writing about my home state! Remember, we’re responsible for everything evil in the world since November 2000, when we refused to vote for our own Senator for President.

  2. 2
    satby says:

    Richard, what’s the solution to such wide discrepancies? Are there any policy proposals discussed in the trade about reducing or eliminating the variation?

  3. 3

    @satby: My proposal is to really strengthen meaningful difference regulation. That would not matter in these cases where there is a single carrier on the market in each county. Honestly, unless there is a rewrite of the law on how subsidies are calculated AND the net minimum subsidy increases, I am not sure what the solution is for this set of corner cases.

  4. 4
    karen marie says:

    While the numbers I am reading are bad news, the bottom line is, people have access at all. But that’s small comfort when it’s taking half of whatever small cushion you might have, if any, at the end of the month, after housing, utilities, food, clothes and transportation. Until the cost structures of everything related to health care are addressed, I don’t see the individual costs changing. This really is a shitty country. Selfish beyond belief.

  5. 5
    karen marie says:

    @Lit3Bolt: It never occurred to me to blame Tennessee but now that you mention it …

  6. 6
    greennotGreen says:

    My home state and Perry County, my mother’s family home. However, you left a couple of factors out. Perry County is extremely poor; probably a lot of families are under the poverty line. And you say they have “good, cheap insurance.” But what they don’t have is good available health care. I didn’t retire there for the exact reason that the nearest good hospital is an hour and a half away.

  7. 7
    greennotGreen says:

    @Lit3Bolt: Yes, and it makes me sad. But it’s the logical conclusion of Republican policies: Don’t fund schools, create an ignorant electorate who will vote against their own interests, they keep you in power, lather, repeat.

  8. 8
    Unknown known (formerly known as Ecks, former formerly completely unknown) says:

    Seems a bit odd that subsidy levels were made so contingent on price offerings. Surely it’s the classic confusion of “thing that is a good proxy measure for X” and “making important outcomes contingent on that thing.”

    For instance: Standard tests are a pretty good proxy for how much students are learning. But if you make teacher rewards heavily contingent on standard tests, then you create all kinds of weirdly distorted teaching behaviours, many of which are not compatible with the learning that you are trying to create.

    Or do I have the wrong end of this?

  9. 9
    eclare says:

    I’m here in Shelby county, TN, and I’m going to a sign-up session tomorrow to see what my options are for next year. Seen in the news the premiums are going up substantially, this year I didn’t qualify for subsidies, we’ll see about next year.

Comments are closed.