Does being uninsured kill you?

A bit of a blog-o-war has been cooking regarding the evidence that being uninsured results in excess deaths. Megan McArdle who blogs for the Atlantic and Ezra Klein who blogs for the Washington Post were the principals in the debate with many chiming in. I am not going to link to all the back and forth or all the various folks who chimed in. I think the main thing that McArdle deserves to be criticized for is previously arguing that the Democractic health reform plans were bad because they would lead to rationing and a limiting of access to medical innovation. Then she says not having health insurance isn't so bad because health care is not so effective at extending life. Either of these two positions are questionable, I think. Holding both of them at the same time doesn't make so much sense.

In any event, two bloggers have weighed in with guest posts from researchers who have specialized in this area in very convincing, peer review fashion. J. Michael McWilliams from Harvard Medical School, and Stan Dorn who worked for Urban Institute and was principal for one of the key studies. Austin Frakt with a bit more.

The upshot is that it is hard to measure precisely, but being uninsured is not so good for your health, on average, and does lead to early death, though the exact number of hard to quantify. Mcardle cherry picked the one study that took her where she wanted to go (a study done by a reputable researcher, but one that did not account for the endogeneity of illness in the decision to seek insurance), but the totality of literature is in the opposite direction. No study is perfect, and observational/epidemiological studies of real people living their lives are hard to do well. You have to look at the totality of the evidence. The only other thing I would note is that all the folks weighing in on the controversy with their words have health insurance....