Never one to let anything speak for itself, the president made his case for the Affordable Care Act on Tuesday, the day its exchanges launched. To listen to his speech, the health-care law has fixed the U.S. economy, will save tens of thousands of lives, and has solved the U.S.’s long-term debt problem. Needless to say, none of these things are true.
He began with a sort of Clintonian random factoid he said he’d heard the other day: that people who get cancer and have health insurance are 70 percent more likely to be alive after five years than people who get cancer and don’t have health insurance. I have no idea if this is true or not, statistically, because I couldn’t find the study — but (1) there are plenty of lurking variables here (people who have health insurance are more likely to survive it because they’re probably wealthier, have healthier habits, etc., and it’s impossible to control completely for all of this), and (2) if you get cancer, you’re significantly more likely to live another five years if you live in the U.S. than you are in any other wealthy country (there are problems with citing five-year survival rate, but he said it).
Then the president made a rather more dramatic claim, dispensing with statistical nuance altogether: “This is life-or-death stuff. Tens of thousands of Americans die each year just because they don’t have health insurance.” This isn’t true in the way most people would understand it: Obviously, tens of thousands of Americans do not die in ditches or outside the doors of hospitals because they lack insurance that will save their lives. But the claim isn’t entirely without a statistical basis: A 2009 study by Harvard Medical School found what a number of other studies had argued over the years, that uninsured people, adjusted for a wide range of factors, have higher death rates than people with insurance (its result was much stronger than those of other studies, though). Taking its finding that mortality rates for the uninsured are 40 percent higher than the insured, the authors calculate that in 2005 there were “approximately 44,789 deaths among Americans . . . associated with lack of health insurance.” (Note: Extrapolating the HMS numbers to the CBO’s predicted rates of insurance coverage after Obamacare is fully implemented at the cost of $200 billion a year, tens of thousands of Americans will still be dying for lack of insurance coverage.)
The study has some problems — for one, it looked at people who were uninsured at one point in time, rather than measuring whether they were uninsured over the period when their mortality was higher; the authors can’t remove the fact that people who choose to have health insurance are more likely to care about their health; etc. But most important, there are lots of things that are associated with dramatically higher rates of mortality — having a job that involves manual labor, owning a motorcycle, drinking alcohol, smoking tobacco, etc. All of those risks are real, but we wouldn’t exactly call each a “life-or-death issue” worthy of a trillion-dollar national effort. Increasing the number of Americans with health insurance probably will reduce mortality rates, but blaming the deaths of tens of thousands of Americans on “just” the lack of health insurance now is a serious stretch of statistics.
Even to the extent that lack of health insurance does cause deaths, we should consider whether the policy being proposed is the best “treatment” to affect the bad variable (mortality) and the best way to apply that treatment. It’s not clear the health-care law is well designed to reduce mortality. For one thing, while the exchanges will hopefully provide reasonably good insurance, a good bit of Obamacare’s expansion in coverage (one third, according to the CBO, but that might change a lot) comes via the Medicaid expansion — and the best assessment of Medicaid that’s been done, a randomized control trial in Oregon, foundno evidence that it improves physical-health outcomes. It did find, though, that people were better off mentally, because they had less financial stress — so I suppose the president is partially right when he boasts that we’re “free[ing] millions of our fellow Americans from [the] fear” “that they’ll go broke if they get sick.”
But whether we’re freeing them from that problem rather than just the fear of it is also a matter of real debate: Will expanding insurance coverage, as the ACA does, make medical bankruptcy a thing of the past?
Not necessarily. There are a number of provisions that will, of course, make more Americans more financially secure: no more lifetime limits on an insurance plan’s payments, subsidies to people who couldn’t afford insurance but want it, etc. But there are also countervailing factors: People will be buying a health-care plan that’s going to cost them a lot of money, and it won’t necessarily protect them from bankruptcy if those costs spiral out of control. Inasmuch as health-care bankruptcies can be accurately measured, Massachusetts, a state with high levels of health-insurance coverage, has plenty of them; and the implementation of a plan similar to Obamacare in 2005 doesn’t seem to have lowered them. Health-care problems cause all kinds of financial stress that’s not affected by insurance coverage, so medical bankruptcies happen even in social democracies with single-payer health-care systems. It’s possible that the ACA will, overall, reduce the rate of medical bankruptcy, but it will be very far from eliminating the problem, as the president suggests.