The Comparative Value of Health Care
By Mike Dorf
In my latest Verdict column, I argue that the legal battle over the constitutionality of the individual mandate in the Patient Protection and Affordable Care Act raises a relatively unimportant constitutional question: Whether or not Congress has the authority, either under the taxing power or the commerce power, to mandate that individuals purchase a service (or good) from a private actor, Congress will remain nearly omnipotent. As I explain in the column, even if the Court invalidates the mandate, Congress could accomplish essentially the same thing using other powers.
Of course, I also explain, Congress won't, because of the politics. I then go on to lament that result as a matter of policy. The large-n studies by health economists show that being within the "flat of the curve," i.e., having health insurance at all, makes a substantial difference relative to not having health insurance, but having different kinds of health insurance doesn't make a noticeable difference. This result is generally thought to support two policy prescriptions: 1) we should try to lower overall health care costs for people who have health insurance because the additional fancy care for which we're paying doesn't seem to do any good; and 2) we should insure everybody. As a first approximation, that's pretty much what the PPACA attempts to do, which makes it beneficial.
Here I want to see if I can quantify the expanded-coverage benefits and put them in a little bit of perspective. This is incredibly tricky because the data we have are often non-comparable. Moreover, the medical literature contains all sorts of stupid errors. For example, it is not uncommon to see claims that early detection of cancer leads to higher survival rates. That might be true, if early detection led to effective early intervention, but it also could just mean that survival is defined as 5-year or 10-year survival, and by detecting cancer earlier, doctors start the survival clock ticking earlier. E.g., someone who lives 2 years after detection of a stage-3 cancer or 5 years after a stage-1 cancer detected three years earlier. The early detection isn't leading to a longer survival.
What would be really useful then is some number that cancels out all the noise. Life expectancy should be adequate to the task. To be sure, it's hardly perfect. Effective medical care could greatly improve quality of life without affecting life expectancy. E.g., a successful drug treatment for Alzheimer's might have that profile. Still, if one is measuring broad inputs, life expectancy will correlate positively with overall health.
So, what is the impact on an average American's life expectancy from going from a state of no health insurance to having health insurance? Searching large databases, I couldn't find anything definitive, but my overall sense is that it's something on the order of a year, probably less. It's hard to isolate the effect because health and longevity correlate strongly with socioeconomic status, which also correlates strongly with health insurance.
A year of longevity is nothing to sneeze at. Indeed, even something much smaller is substantial. Suppose that the average longevity benefit of going from no health insurance to health insurance is only a quarter of a year. Still, getting the roughly 50 million uninsured Americans a quarter of a year each would mean an extra 12.5 million person-years. Or, dividing by life expectancy, that comes to a savings of over 150,000 human lifetimes (over the course of a human lifetime)--which is an enormous difference.
That's still well short of the per person benefit from quitting, or better yet, not starting smoking, which is about 6 years of life expectancy. Or obesity, which appears to reduce life expectancy by about 2 years. These numbers are contested and highly dependent on the specific study, but I'm pretty confident that I have the relative shape right: Smoking is worse for longevity than obesity, which is in turn worse than not having health insurance.
What that all says is that government policy that aimed at promoting longevity could be very effective at low cost if it reduced smoking and eliminated agricultural subsidies for commodity crops (mostly grown for animal feed), which artificially lowers price of, and thus increases demand for, unhealthy foods. Those are worth doing regardless of the long-term fate of the PPACA.
In my latest Verdict column, I argue that the legal battle over the constitutionality of the individual mandate in the Patient Protection and Affordable Care Act raises a relatively unimportant constitutional question: Whether or not Congress has the authority, either under the taxing power or the commerce power, to mandate that individuals purchase a service (or good) from a private actor, Congress will remain nearly omnipotent. As I explain in the column, even if the Court invalidates the mandate, Congress could accomplish essentially the same thing using other powers.
Of course, I also explain, Congress won't, because of the politics. I then go on to lament that result as a matter of policy. The large-n studies by health economists show that being within the "flat of the curve," i.e., having health insurance at all, makes a substantial difference relative to not having health insurance, but having different kinds of health insurance doesn't make a noticeable difference. This result is generally thought to support two policy prescriptions: 1) we should try to lower overall health care costs for people who have health insurance because the additional fancy care for which we're paying doesn't seem to do any good; and 2) we should insure everybody. As a first approximation, that's pretty much what the PPACA attempts to do, which makes it beneficial.
Here I want to see if I can quantify the expanded-coverage benefits and put them in a little bit of perspective. This is incredibly tricky because the data we have are often non-comparable. Moreover, the medical literature contains all sorts of stupid errors. For example, it is not uncommon to see claims that early detection of cancer leads to higher survival rates. That might be true, if early detection led to effective early intervention, but it also could just mean that survival is defined as 5-year or 10-year survival, and by detecting cancer earlier, doctors start the survival clock ticking earlier. E.g., someone who lives 2 years after detection of a stage-3 cancer or 5 years after a stage-1 cancer detected three years earlier. The early detection isn't leading to a longer survival.
What would be really useful then is some number that cancels out all the noise. Life expectancy should be adequate to the task. To be sure, it's hardly perfect. Effective medical care could greatly improve quality of life without affecting life expectancy. E.g., a successful drug treatment for Alzheimer's might have that profile. Still, if one is measuring broad inputs, life expectancy will correlate positively with overall health.
So, what is the impact on an average American's life expectancy from going from a state of no health insurance to having health insurance? Searching large databases, I couldn't find anything definitive, but my overall sense is that it's something on the order of a year, probably less. It's hard to isolate the effect because health and longevity correlate strongly with socioeconomic status, which also correlates strongly with health insurance.
A year of longevity is nothing to sneeze at. Indeed, even something much smaller is substantial. Suppose that the average longevity benefit of going from no health insurance to health insurance is only a quarter of a year. Still, getting the roughly 50 million uninsured Americans a quarter of a year each would mean an extra 12.5 million person-years. Or, dividing by life expectancy, that comes to a savings of over 150,000 human lifetimes (over the course of a human lifetime)--which is an enormous difference.
That's still well short of the per person benefit from quitting, or better yet, not starting smoking, which is about 6 years of life expectancy. Or obesity, which appears to reduce life expectancy by about 2 years. These numbers are contested and highly dependent on the specific study, but I'm pretty confident that I have the relative shape right: Smoking is worse for longevity than obesity, which is in turn worse than not having health insurance.
What that all says is that government policy that aimed at promoting longevity could be very effective at low cost if it reduced smoking and eliminated agricultural subsidies for commodity crops (mostly grown for animal feed), which artificially lowers price of, and thus increases demand for, unhealthy foods. Those are worth doing regardless of the long-term fate of the PPACA.