Implied marginal rates and the ACA subsidy formula

The marginal rate is how much you spend for the next chunk of income.  Usually we use this framework when talking about taxes.  The next dollar I earn has a North Carolina marginal rate of a little over 5% which means I keep 95 cents and the state of North Carolina gets a nickel.  This is important because decisions happen at the margin and from a social policy perspective, we should want to encourage good things and discourage bad things with marginal rates.  Big jumps create discontuinities and playing hop-scotch is usually a bad thing.

The ACA subsidy formula has weird implied marginal rates.  I am looking at what each additional $1,000 in income costs a family of two who are subsidy eligible and who are looking to buy the benchmark silver plan.

From $17,000 to$20,000 this couple pays an extra $2 per month for every thousand dollars more they earn a year.  Annually this is about a 2% marginal tax rate on the additional income.  And then there is a huge bump from $20,000 to $21,000.  The benchmark premium suddenly becomes $256 more expensive.  This is a 25% marginal rate.  In our income tax system, the 2018 tax year 24% marginal rate applies to couples filing jointly only starts at $165,001.

And then the marginal rate drops again when the family increases their earnings from $21,000 to $22,000.  The marginal rate for this slice is now about 11%.  The marginal rate for couples earning under 300% FPL is in the mid-teens, and then there is a drop in the marginal rate to just under 10% for 300% to 400% FPL and then a potential massive spike as soon as someone earns over 400% FPL.

This is an applied microeconomist’s dream design as it creates tremendous research opportunities as there are numerous notches to shift behavior.  It is bad social-fiscal policy.

It is also surprising because we run a major social welfare program, Social Security, with step functions that smooths out the impact of income changes.  Ping pong can be a fun game to play but it is seldom optimal subsidy design.






6 replies
  1. 1
    Ohio Mom says:

    I have nothing to add, just wanted to say Hi so you wouldn’t feel like you were blogging into the void.

    What I took away from this post was mainly discouragement. Our country has found a gazillion small ways to do health coverage wrong. It’s a certain type of evil genius.

  2. 2
    daveNYC says:

    To be completely fair, it’s not just the ACA subsidies that have this problem. The US has all sorts of safety net type programs that are setup to make improving your income very painful.

    It’s frankly bizarre.

  3. 3
    Ohio Mom says:

    @daveNYC: As the Mom of a disabled young adult who is benefitting from several safety net programs, I’ll add that even when they are not punishing you for improving your financial situation, they are very arcane and Byzantium. You never know if you are doing it right.

  4. 4
    Julie says:

    Very interesting analysis. I would like to see marginal rates for higher incomes, since it looks most of these incomes are below 138% FPL for a couple. My understanding is that in most states, a couple with income below 138% FPL would not be eligible for ACA subsidies, but would instead be directed to the Medicaid program.

  5. 5

    @Julie: Correct —

    138-300% average marginal rates for another $1,000 incremental income bumps around from 11-17% in 2019 until we get to 300%-400% FPL where an extra $1000 incremental income produces a marginal net of subsidy premium of $98.60 annual increase.

  6. 6

    Thanks. Also, a look at the jump between 4xFPL and 4.01xFPL might be worthwhile.

Comments are closed.