Monday, October 30, 2017

Trump's accidental health care experiment

The Trump administration's ham-fisted attempts to sabotage Obamacare are going to make this open enrollment one of the biggest natural experiments in health care, and we should learn from it.

Trump has done two big things to try to hamstring the Affordable Care Act exchanges.
  • One was to dramatically cut funding for Healthcare.gov outreach ads and navigators to help people select insurance plans, but many blue states that run their own exchanges will continue to spend as much on outreach as they did in the past.
  • The other was to end cost sharing reduction payments to insurers, which is having a very weird impact. In some states it has resulted in big increases of affordable tax credits, making insurance much more affordable for those making between 200-400% FPL. How big the impact is varies significantly across states.
What this has created is a massive natural experiment on two metrics. Some states will see a large increase in premium tax credits for this group but very little outreach spending. Some will see large increases in premium tax credits and strong outreach spending. Some will see only small increases tax credits and little outreach spending, and others will see small tax credit increases and strong outreach spending.

It will be a real chance to see what impact the tax credit level and outreach spending have on enrollment, as well as to test just how rational individuals are when it comes to buying health insurance. We will get to see how much impact outreach efforts really have and how big a difference more subsidies make.

Trump has created needless chaos, but it would be stupid not to learn from it. I would hope experts in health care would make predictions now that can compared against the real world data.

 (I personally think the impact of outreach spending is being overstated and the impact of providing more tax credits will be far more significant. The CBO predicted the short term impact of ending the cost sharing reduction payments would be a modest drop in the number of insured, but I suspect the number could easily increase.)

No comments:

Post a Comment