<p>I was amazed by the magnitude of the mortality drop in Massachusetts after Romneycare was enacted. A 3% drop in mortality, all of it in adults 20-64, is huge. It’s extending the life expectancy in MA by something like two years.</p>
<p>All results like this can be challenged, but this one looks strong. Rather than sampling in Massachusetts, the researchers looked at all deaths, which makes a big study. If the death rate had been from something other than uninsured people getting insurance, then we might expect to see a drop in death rates for all ages, but the drop was only in the 20-64 age group that contained the newly insured people. The death rate drop was bigger in counties that have more poor people who would have become newly insured. They matched counties in Massachusetts with similar counties in other states, and the counties in other states did not see any death rate decrease. The mortality decrease might have been caused by something else, but there are not a lot of obvious candidates.</p>
<p>Some people might wonder why the Oregon Medicaid study didn’t show a similar mortality decrease. Turns out, it couldn’t; it was underpowered. If there had been a similar mortality drop in Oregon, the Medicaid study would not have been big enough to detect it.</p>