At NBER, there is a working paper with very strong results that shows health insurance buys health.
In the winter of 2017, the IRS engaged in a massive randomized control experiment. 4.5 million people had to pay the individual mandate. The IRS randomly selected 3.9 million people to send letters to remind people to enroll. The researchers found two major things:
Beginning with the effect on coverage, we and that among individuals who were uninsured for some portion of the prior year, those in the treatment group were 1.3 percentage points more likely to enroll in coverage in the year following the intervention than those in the control group, a 2.8% relative increase. On average, each letter increased coverage among this group by 0.14 months during 2017, or one additional year of coverage per 87 letters sent. We document larger effects among individuals who lacked any coverage during the prior year and among older nonelderly adults. The effect appears to operate through new enrollments in the individual marketplace as well as through Medicaid take-up. Although there is some attenuation, coverage rates remains higher in the treatment group than in the control group in the two years following the intervention.
finding #1. The letters were a cheap and effective outreach effort. 87 letters to generate one enrollment is inexpensive. There is also enrollment inertia as once people sign up once, they are more likely to sign up and maintain coverage in the future. Coverage could be in either the individual market or Medicaid.
The more important finding is that health insurance buys health as well as financial insurance:
We present evidence that it did. In the two years following the intervention, the rate of mortality among previously uninsured 45-64 year-olds was lower in the treatment group than in the control by approximately 0.06 percentage points, or one fewer death for every 1,648 individuals in this population who were sent a letter…
Exploiting treatment group assignment as an instrument for coverage, we estimate that the average per-month effect of the coverage induced by the intervention on two-year mortality was approximately -0.17 percentage points… With these caveats, our results provide the first experimental evidence that health insurance reduces mortality
That is an incredibly important finding. In some ways it is intuitive. Being able to pay for health care leads to better health. But there has been no good randomized control work to show that is the case. And there has been a cottage industry arguing that the public paying for health care is buying no health. This won’t stop that argument but it renders it incoherent.
takebakawashi
what are your thoughts on the RAND Health Insurance Experiment?
(Canadian, fan of single payer, not pushing a viewpoint, no dog in this fight)
Major Major Major Major
Always happy to see more studies pointing to this outcome; thanks!
Brad F
David
How solid an instrument is a letter type to mortality for the IV? Also, the paper has different types of letters (Fig A1)–and 84% received “treatment” and 14% control overall
Brad
Tom Levenson
If I’m doing the arithmetic right 1/1648 fewer deaths over a population of 3.9 million who received letters comes out to 2366.5 deaths avoided/lives saved.
Or, about three quarters of a 9/11. Expand that to everyone facing the mandate and you get to over 2700, or damn near exactly the number killed in the World Trade Center attack.
These are the potential victims of the GOP health terrorists.
jl
Thanks for important post.
Hob
There’s a particularly combative commenter on Lawyers Guns & Money, of the “down with establishment Dems / Obama sucks” variety, who’s made some specific claims about the ACA that I find confusing; their response to requests for citations or clarifications tends to be along the lines of “fuck off, you Obama-worshiper” so I’m inclined to think there’s not much to their claims, but I’d like to at least understand the factual issue better.
So: they stated that, compared to before the ACA, “Medical inflation is way up, thanks to the public subsidy coupled with uncontrolled provider consolidation. Hence a mass shift of employer-provided plans to high-deductibles and an overall decline in health care utilization. Plus, a net increase in deaths from underinsurance.” Now, on the first point there (medical price inflation), I’m thinking the argument that ACA subsidies have made providers greedier is probably not possible to prove one way or the other, since we have no control world with which to compare the rate of price increases without the ACA. But on the second, “net increase in deaths from underinsurance”… I’m not sure what they’re talking about there, or what numbers one would look at specifically for that. I mean, “they totally pulled it out of their ass” is always an option but in my experience it’s more common for something like that to be at least distantly based on a real statistic that someone is interpreting in a particular way.
So basically I’m wondering, how would the statistic “deaths from underinsurance” be calculated, because I don’t think it is quite the same as what the study mentioned in this post is talking about. That is, this post mentions evidence that people who gained insurance have a lower mortality rate, but that would be separate from the question of whether the not-having-insurance group is growing faster, or whether the people in that group are dying faster than they used to. Does that make sense and can someone enlighten me? I do try to follow the posts in this series but I know I’ve missed a lot.