• Insurance and health

    I gotta get something off my chest.

    Occasionally I read something that suggests, or outright states, that the author believes that health insurance does not improve health. (No, I won’t link to any of them.) Fine, they can believe that. Let’s ignore the massive amount of research to the contrary for a moment and just look at the logical, rational consequences of this belief:

    If insurance does not improve health then it either (a) worsens health or (b) it is irrelevant to health (neither improving or worsening it).

    In the case of (a), it worsens health, then it is rational to refuse health insurance, public or private, even if it is provided at no cost. That requires paying for all of one’s care out of pocket. It may make financial sense to do this, but that’s not the point I’m arguing right now. I’m talking about health. If insurance is bad for health and one rationally refuses it on those grounds–independent of cost–how does one also justify using health care at all? What’s the full argument that health care is of some value while health insurance is not? (I once offered one. I’m not linking back to it. If you want to make this argument, figure it out for yourself.)

    In the case of (b), insurance is irrelevant to health, then it doesn’t matter whether one is insured or not. Now I’m willing to enter into financial considerations to break the tie. Should one buy insurance or not if it is irrelevant to health? One should if it is a less expensive way to obtain the health care one would otherwise buy without it. (If one wants to include risk aversion in the cost-benefit analysis, that’s fine.)

    Pursuing this idea, one finds that the whole notion of insurance unravels. If only those who would financially gain by purchasing it do so, premiums must increase to balance expenses. But then those for whom the higher premiums outweigh expected costs would refuse insurance. Continuing, we enter the classic adverse selection death spiral. Insurance eventually makes financial sense for no one. (It does not matter that this is not common in the real world. We’re talking about theoretical arguments here.)

    A world could exist with no insurance at all. But if health care has any positive effect on health, we’d all be fools to live in such a world. Health care costs something, a lot if one is sick enough. Dying is a certain end of one’s ability to enjoy life. Most people would (and do) want to obtain care to prolong life or improve the quality of it. Who will finance that, possibly expensive, care? How does one transfer the risk that one will want care beyond one’s current means? How does one protect one’s family from the financial hardship of accidents and disease? Insurance! It makes sense even if it doesn’t itself improve health, so long as it doesn’t itself endanger health (case (a) above). (Note, I am not suggesting that insurance means one can get something for nothing. It’s a question of how to finance an acute episode. One may be able to “pay it back” or prepay it in premiums. It’s fine with me if you want to think in terms of catastrophic insurance only.)

    Actually, insurance can’t logically play the financial role just described without also being good for health. If one thinks that insurance has financial value because it facilitates the access to expensive care, spreading its cost over time and/or over a broad risk pool, what does that mean about the relationship between insurance and health? Well, if the care bought with insurance is good for health and that care could not be obtained without insurance, then insurance is good for health. It’s impossible to claim that insurance plays purely a financial role if the use to which insurance is put is to buy health-improving care that could not otherwise be purchased.

    Thus, the only logical end of the claim that insurance isn’t good for health is to opt out of insurance and to opt out of health care altogether and forever. How many of those who claim insurance doesn’t improve health are willing to follow the logic? Insisting they walk the walk, and not just talk the talk, makes that small crowd a lot smaller. So small, I conjecture, it’s not worth paying attention to. They can’t possibly have any effect on policy. They’re not just wrong, they’re wasting their time and anyone who pays a moment’s notice to them is too.

    With that, I’ll move on to something relevant.

    • Something has been gnawing at me and you’re barking up the same tree here.

      It’s a somewhat agreed upon fact that people die without insurance. The Urban Institute’s numbers on the matter seem solid to me on this matter.

      Now, given that, when Texas threatens to drop out of Medicaid, I think the outrage needs to start now. Not later, not after babies are dying from lack of prenatal care or people are showing up in ER’s with cancerous tumors.

    • Of course, by that logic then lotto winnings, tax credits, and no-interest loans also improve health.

      Personally, I’m much closer to your B above. Health Insurance itself has nothing to do with your health. It can improve your access to something that does but, as above, it isn’t alone in that. Phrasing things to conflate first and second order effects, as you seem to be doing, doesn’t really improve the general debate. It’s one of the rhetorical tricks that makes rationally discussing this issue pretty much impossible.

    • “Thus, the only logical end of the claim that insurance isn’t good for health is to opt out of insurance and to opt out of health care altogether and forever.”

      I largely agree with you. But this particular point seems wrong. Assume that I have read statistics that convince me that, on balance, having affordable access to medical care is not good for my health. (Perhaps I know–assuming for the sake of argument that it’s true–that some care episodes will improve my health, but others will harm it, making a lifetime of care a wash.) Still, I also know that I am weak-willed, and I know that whenever I feel like I could benefit from going to the doctor, I will go, despite my rational belief that this will not benefit me on balance. (Here the anxiety associated with not knowing what’s going on medically when I refrain from seeing a professional is surely relevant.) In this case, having insurance may be useful, because it protects me from having to pay the total costs associated with inevitable behavior that nevertheless runs against my best judgment. If my ongoing desire to seek medical care is stronger than my aversion to spending out of pocket each time I am anxious about my health and wish to see a doctor, then on balance it might be worth it to me to buy the insurance.

      Isn’t this view “logically” possible?

      • @Paul – Is it “logical?” I don’t know but it sure seems irrational. Anyway, it’s so finely balanced it doesn’t seem like it’d be broadly applicable. My best hypothesis for how insurance can harm health (rejected by the data, by the way) pertained to Medicaid. I didn’t put it in the post but now I’m feeling more generous. Take a look.

        @Komori – There’s no conflation of first and second order effects or rhetorical trickery. (Why do people even write such things? It’s silly to assume I’m trying to trick you. Just ask honest questions and I’ll answer them as I can.) There’s a genuine causal effect of insurance on health. Obviously it works through actual receipt of health care. This whole line was the point of my Medicaid-IV series, though there is a vast literature on it.

    • I like to measure health against health and longevity against longevity rather than to measure health and longevity against dollars.

      So we know that money can be used to improve health and longevity.

      So if you has x amount of dollars to spend to improve your health would the most effective use for that money be health insurance? I contend that this is not an easy question to answer. consider safer cars, fly more drive less etc. At one point congress was going make a requirement for airlines but a study showed that though it would make flying safer it would cause more death by making more people drive. Life is tricky.

      This analysis sidesteps a lot of moral issues like whether a shorter happier life better. Like if fewer people had health insurance would it be healthier to not carry health insurance.

      On the more complicated subject I delegate may comment to Tyler Cowen:

    • Austin,

      I agree the reasoning I laid out (it’s not *my* reasoning, mind you) is surely finely balanced, as you say. We philosophers are fine going in for the overwrought in order to quibble with claims about what’s logical or not. Granted.

      But my line of reasoning, which seems so overwrought when presented from a first-person perspective, is indeed commonly invoked to justify universal insurance coverage at the population level. Indeed, it was precisely Krugman’s response to Michael Cannon in this debate about HCR: . Krugman claimed that *even if* Cannon is right that health care/insurance doesn’t improve health, the financial considerations (which you conceded might break the tie) can win the day, precisely because people are not going to refrain from seeking medical care, *regardless* of what they believe about the quality they’ll receive.

      But at the end of the day I’m with you: I don’t see any good argument for thinking that insurance doesn’t promote health.

      • @Paul – Maybe I need to think about it further, but I don’t like Krugman’s argument. If people are stupid enough to keep using health care that is not effective, the solution is not to make it easier to do so, and certainly not to subsidize it. But we can just jettison this whole line. It isn’t true that health care is not effective. It is true that some health care isn’t, but it is hard to tease apart what is and is not and when. Doing so requires a lot of research, only a little bit of which has been done. We need to do more, and fund it.

        In the interim, expanding insurance access is, on net, good for health and worth doing. That means someone has to pay for it or some groups get a little less (e.g. Medicare beneficiaries). I don’t have a problem with that, as an interim solution. In time I’d like to cut out the useless care, or some of it. See my recent posts on prostate cancer as an example of an area where there is good evidence of wasteful spending, or at least arguably so (one has to draw a line somewhere so there is subjectivity in this).

    • Austin comments: “If people are stupid enough to keep using health care that is not effective, the solution is not to make it easier to do so, and certainly not to subsidize it.”

      This sounds right off the cuff, but Krugman’s point, I take it, was that *if* health insurance on balance didn’t promote health (because health care doesn’t promote health), and *if* people would be still be “stupid enough” to want care anyway, then there are nevertheless distributional considerations that could justify subsidization. After all, the well-off would still buy insurance, and would therefore be protected against the human folly of seeking care. Those who are worse off would not have such protection against the financial ruin they visit upon themselves.

      Of course, all this is counterfactual, since insurance does promote health. But the counterfactual situation is not clearly one where subsidization is stupid. It would be more like a rule of rescue where we rescue hikers stupid enough to get trapped on the side of a mountain they had no business climbing. If the poor were more likely to end up stranded on a metaphorical mountain, it’s not clear that bailing them out would be stupid. Moreover, if the counterfactual is one where insurance neither harms or promotes health (on balance), there could still be non-health benefits to health care (e.g. a calming presence in a dismal hour). Again, if the poor were differentially likely to face financial ruin for indulging such a common human desire, it’s not clear to me that subsidization would be “stupid.” —-But please note that these are free-flowing musings provoked by your provocative post. Keep ’em coming!

      • @Paul – I still think the solution to that hypothetical is to tax the crap out of health care use, maybe even means-tested taxation, not subsidize the poor. If it doesn’t have objective, measurable value I don’t want to spend money on it, I want to tax it. Some people like cigarettes and soda too. I don’t want to subsidize those either. (No, I don’t think one should be able to buy the latter with food stamps. I’ve heard the arguments and my opinion is pretty firm on this.)

        Feel free to keep this going. It’s interesting.

    • Whether health insurance improves health depends on your definition of health. If health is recovery from sickness or injury, then catastrophic health insurance certainly seems valuable (the car equivalent is insurance for someone crunching your fender and your insurance paying for the repair).

      On the other hand, if you define health as a state of being healthy, i.e., your ability to play sports, keep a job, sleep well, etc., then I am not so sure. The car equivalent would be buying insurance to change your oil, replace your tires, etc. To carry on the analogy, if you had insurance that would pay for oil changes AND replace your engine if you failed to do so, I think it is arguable that some people would not take the time for the oil change. Even more so for tires, as people would have little incentive to drive conservatively and save tread life. We see people all the time who have good ‘health’ insurance, who act in ways that reduce their healthy state, so I think you can make a strong argument that at least some forms of health insurance do not improve health.

      • @SteveSC – Which is not what I was arguing against. If someone wants to say, “Sometimes and for some people insurance can be less than helpful,” I would agree. However, in my rant I permitted the reader to think exclusively of catastrophic coverage if (s)he wished.