mead cohen berger shevtsova garfinkle michta grygiel blankenhorn
Health Care Humility
Health Insurance Doesn't Always Make You Healthier

We missed this when it first came out, but two Northwestern professors and two other researchers recently released a paper on health insurance (h/t @AndrewCQuinn); it affirms that, of course, people with insurance use health care more than people without insurance, but, surprisingly, they aren’t significantly healthier:

Being uninsured significantly reduces healthcare utilization, across a number of measures. The uninsured visit the doctor and are admitted to a hospital less often, are less likely to have prescription drugs prescribed, and are more likely not to take prescribed drugs for financial reasons. However, we find little evidence of an impact of health insurance on health or mortality. Controlling for initial health and other covariates, the uninsured do not become less healthy than the insured over time, across an array of measures, including self-reported health status, heart disease, stroke, cancer, diabetes, lung disease, depression, and limits on daily activities. In our central estimates (uninsured versus all insured), we find no evidence that uninsurance predicts higher mortality for the uninsured through 12-14 years after initial observation and only mild evidence after 16-18 years.

This study echoes another done earlier this year on the Oregon Medicaid population. That study found that a randomly assigned group of people who were given access to Medicaid in Oregon used more health care services than those remaining uninsured, but without any large improvements in their physical health. (Mental health is another issue—the insurance was associated with a drop in depression). Taken together these two studies suggest that insurance can have an inflationary effect on medical spending without producing proportionately better health outcomes.

That doesn’t mean that we should give up on the goal of expanding access to at least some basic forms of insurance to the whole population. There are lots of good reasons for doing that, not the least of which is preventing people who get seriously sick or injured from going bankrupt. But it does point out the costs of the way we currently fund health care. Health insurance is not health care; it’s one way of paying for health care, and it incentivizes people to use more services than they actually need. The fact has to be in the forefront even as we work to expand access.

Still, as we keep a close eye on how insurance mechanisms affect our system, the most important thing to get right in the coming years is not who pays or how, but rather how people receive care. If we can improve delivery options, questions of how to expand access will become easier to answer.

Features Icon
show comments
  • Anthony

    No surprise in research paper finding; yet health care – received and used – encompasses dental, clinical, physical, visual…etc. Resolving all that (in line with how people receive care) plus layers of cost variables (incentives) embedded in current system will certainly complicate question of access/marginal cost going forward; but consequent public policy dialog is well worth it as system under goes closer look.

  • Clayton Holbrook

    “In our central estimates (uninsured versus all insured), we find no evidence that uninsurance predicts higher mortality for the uninsured through 12-14 years after initial observation and only mild evidence after 16-18 years.”

    Didn’t TAI have a post that rightfully cautioned on the use of a single metric, such a mortality rate, as a measure health outcomes? I think the gist was comparing mortality rates of populations under different health care systems didn’t take into account other lifestyle aspects. E.g. Americans tend to drive more, therefore the higher rate of traffic mortality wasn’t accounted for. Can the same principle be applied somewhat here? Mortality rates are dependent more than just healthcare.

    • TommyTwo

      “We find little evidence of an impact of health insurance on *health* or mortality”

      Read the paper itself, which addresses several of the other usual objections. From a first glance, it looks like a worthwhile contribution to the discussion.

  • Self Pay Patient

    Sure, read the second paragraph of Kenneth Arrow’s paper that just about started modern health care economics, it says pretty clearly that health is not the same as health care.

    • NCMountainGirl

      Nor is health insurance the same as health care.

  • mgoodfel

    As shown here:

    not only does the U.S. spend a lot more on health care without an increase in life expectancy, there are a bunch of rich nations that have about the same life expectancy despite variations in their spending.

    And if you look at life expectancy over time, here:

    you see that there’s a lot of improvement early in the 20th century before health care gets very sophisticated. In fact, you get most of the increase (from age 40 to age 70) by 1970, before health care gets really expensive. That’s because the increase in life expectancy is mostly due to a decrease in infant mortality. Life expectancy at age 60 has hardly changed.

    The conclusion I draw is that high-tech health care just doesn’t affect life expectancy that much. All the rich countries do the basics right (clean food and water, sanitation, vaccination, prenatal care, birth in a hospital, etc) and get about the same result. The U.S. then also spends tons of money on high tech care that doesn’t matter much.

    For all the fuss about Obamacare, it’s just not going to make much difference. It is the defacto nationalization of the health care system, and it isn’t going to help control costs, but it’s not going to help or hurt life expectancy.

    • NCMountainGirl

      Life expectancy isn’t the entire story. Many treatments that do not in themselves prolong life increase the quality of life. The greater ability to exercise after receiving a hip replacement may have only a small impact on lif expectancy but the enhanced mobility makes the patient’s later years far more rewarding.

    • Andrew Allison

      Thank you very much for the link to The Incidental Economist above, which debunks several of the myths surrounding the efficacy of US healthcare. It’s nice to have the data to back up the intuitively obvious fact that, just as student loans cause tuition cost to rise, insurance causes the cost of healthcare to do so.
      In neither cases, it appears, do the outcomes change commensurately (or at all).
      The healthcare issue we should be addressing is why the US performs so poorly vis a vis the rest of the developed world in infant mortality and longevity. The former could possibly be explained by access to pre- and post-natal care, the latter by violent death unrelated to disease. These data should not be too hard to come by, and might suggest where we should be spending the money.

  • Calgarystatefarm

    I never thought of it that way, thanks for your perspective.

    Ted Yan

© The American Interest LLC 2005-2016 About Us Masthead Submissions Advertise Customer Service