• Read the Literature

    Not only do I believe McArdle has badly misread (or not read) the literature on the relationship between health insurance and health outcomes, including mortality, Monday I will publish on this blog a guest post that includes a literature review that illustrates it. Of course, one already exists, it’s just not accessible to everyone. In addition, I’ve already posted a very brief one.

    If one knows that literature, McArdle’s statements continue to baffle. In her latest post on the matter she writes,

    I think it is possible that the lack of insurance has no effect on aggregate mortality statistics.  I do not think that this is likely, but I think it’s possible.

    Mostly what I think is that the statistics are really, really flawed.

    And even more stunning,

    The mortality question is really important, but it doesn’t touch non-mortality outcomes, which are even harder to measure comprehensively.

    Not only are the statistics on mortality and it’s relationship to health insurance not flawed (and certainly not “really, really flawed”), but the connection between insurance and non-mortality health outcomes is extremely well established. I cannot fathom how it could be missed by anyone examining the literature. Measuring the effect of insurance on non-mortality health outcomes is not “even harder,” it is far easier. That’s why health services researchers and health economists do it all the time, and publish the results.

    This is incredibly important. People really do suffer and die due to lack of insurance. The empirical evidence bears that out. Meanwhile, policymakers debate (and debate, and debate) what to do. McArdle advises a go slow and/or go small approach based on a misreading of the evidence. If there is one thing I would hope we could agree on it is that that’s a very poor basis for policy prescriptions. My recommendation: read the literature or a credible literature review before claiming to know what it says or what it implies we should do.

    • You’re misreading what I said. I said they’re even harder to measure “comprehensively” — i.e. to get some sort of meaningful index of how much suffering is taking place because people lack insurance. I have in fact mentioned some of the literature on things like blood pressure control, which questions your presumption that I don’t know it exists.

      You, and Ezra, continue to broaden the question beyond the narrow one I addressed, presumably because these are easier arguments to refute. But I haven’t made them. I very deliberately have not made them, precisely because they’re easy to refute. Do you want to stake the maximalist claim on the other side: that we know how many people die within a reasonable margin, that it is at least 20,000, and hence that we will see a >1% drop in the under 65 mortality rate if HCR passes? That’s what would refute me, not the statement that there are studies that show health improves with insurance.

      You and I may disagree whether it’s important to have a rough approximation of how many lives, how much suffering, or how much financial disaster may be prevented by a health care plan before we press forward. But that’s a value judgment, not a scientific one.

      • @Megan McArdle – OK, keep it narrow. Your conclusions are informed by your reading of the literature, and that reading (or that which you cite) is very narrow. A more expansive review gives a very different impression. That review is in the literature (I’ve linked to it) and a summary of it will appear on this blog on Monday.

        I’d like to know to what end you’d use the ideal estimate of the number of deaths. Tyler Cowen made a cost per death calculation today. I responded on this blog that I think it misses a key point. If the cost per avoided death of expanding coverage is too great then that suggests we’re paying too much for care, not that we should not expand insurance.

        Reasonable people can disagree about what to do with the numbers. But I think we ought to agree that an analysis should begin with a complete survey of the literature.

    • I am not the one who has been selling a health care plan by comparing our mortality to that of Sweden’s, or quoting the largest mortality numbers they can find. That’s the only question I’m addressing. I have never claimed that there’s no effect–indeed, went out of my way to say that I was pretty sure there was one. I also went out of my way to note that this was not necessarily an argument against national health care, something I’ve repeated in every subsequent post.

      But I don’t think saying it’s an argument for lowering costs will do. I’m using the figures on the health care plan on the table. If you can do it more cheaply, groovy, but you have to actually do it, not just demand that someone else figure it out while you pass a very expensive plan. I think you’re on stronger ground adding in other effects, not saying that in some ideal world, we wouldn’t be spending $163 billion to prevent these ones.

      We’re back to: analysis of what? In the case I was discussing, the size of the effect was very much at issue. What you seem to be saying is that you’d rather be discussing something else, so why didn’t I?

      • @Megan McArdle – Do you think that regular maintenance contributes to structural integrity of your home? Do you believe that if you did not maintain your home it might one day suffer some bad structural outcomes? I do on both counts, as I would expect most people. Now, can you prove to me that if X people do not maintain their home this year it will lead to Y catastrophic structural failures? That strikes me as exceedingly difficult. Nobody would expect to be able to tease it out. But that’s exactly what you seem to demand we do to some level of confidence beyond that already achieved in the case of insurance and death.

        You claim that your only point is that we don’t know the size of the effect (the number of deaths). That may indeed be what you think your only point is. But that is not what people will take away from what you’ve written. My only point is that you’ve given a strong impression that the evidence does not support a link between insurance and mortality. I do not think that is a fair impression to have given. And the reason is that the full body of literature does not support that view, especially when combined with the fact that it is eminently reasonable (and I dare say, accepted) that bad health contributes to mortality.

        We can continue to disagree here because I am not attacking your main point squarely because that’s not my main point. You can continue to demand that I fight on your turf (your main point), but that’s not my beef. Meanwhile, the bigger picture is that I don’t believe we should wait until we know more precisely or to your satisfaction the number you feel we need to know before extending insurance to the uninsured.

        I disagree that it is not legitimate to conclude that if cost per life saved is high then we should seek to lower costs. Why is that less legitimate than seeking to cover fewer lives? They’re two inputs to the same expression. I’m using the same figures you are just concluding a different thing.

        Was Medicare a mistake because it might be viewed as too costly per life saved? How far do we go with this?