The Incidental Economist

I’m sure people are going to start rolling out the old “Medicaid is worse than being uninsured” line, so here is an innoculation:

Let’s start with a basic fact. Having health insurance is better than not having health insurance. Here’s Michael McWilliams saying it saves lives in a guest post. Here’s another post full of links to others who argue that insurance saves lives. Not only that, but health insurance is good for health. Medicaid is health insurance. Therefore, it shouldn’t surprise you that studies show Medicaid improves health.

But wait! You heard that Medicaid actually hurts people. Well, it turns out those studies some interpret as showing Medicaid is bad for health are showing correlation, not causation. Here’s some more examples of that.

Want to get in the weeds? Austin describes how the use of instrumental variables can improve research into Medicaid. Here’s a post on Medicaid and mortality for HIV patients. Here’s Medicaid and child health. Here’s Medicaid and saving babies. Here’s Medicaid expansion and health care utilization. Here’s Medicaid expansion and the technology of birth. Here’s a summary of that series:

My take-away from the Medicaid-IV literature review is: there is no credible evidence that Medicaid results in worse or equivalent health outcomes as being uninsured. That isMedicaid improves health. It certainly doesn’t improve health as much as private insurance, but the credible evidence to date–that using sound techniques that can control for the self-selection into the program–strongly suggests Medicaid is better for health than no insurance at all.

Don’t forget, of course, the Oregon Health Study, an actual randomized controlled trial of Medicaid. Here’s Austin on the results. Here’s me on the results. Guess what? Medicaid was good for health.

Medicaid is not bad for health. That’s a zombie idea. It just won’t die.

@aaronecarroll