Thomas Northcut/DigitalVision
Health Care

We Don't Know Whether Obamacare Was a Net Gain

It's not obvious whether people got healthier when they received more care.

Last week was a wild adventure on the health-care front, as the House seemed to inch closer to passing a bill that repealed or altered significant chunks of Obamacare, before shying away at the last minute. With Democrats proudly declaring that they saved millions of Americans from the brink of disaster, it seems an appropriate time to consider the latest assessment of what Obamacare has actually done for those people.

There's a new paper out looking at how the Affordable Care Act has transformed health-care access, and in turn, what that has done for health. The authors' first answer probably won’t surprise you: when millions more people became insured, more got checkups and primary care doctors. But it’s not obvious that these people got any healthier. As the paper puts it: “No statistically significant effects on risky behaviors or self-assessed health emerge for the full sample.”

Other studies have found substantial effects on self-assessed health, but on the harder markers of health -- like blood sugar, cholesterol and blood pressure -- the famous Oregon Medicaid Study found no significant improvement when government gave people health care.

It turns out the link between “health care” and “health” is not as close as we would like. Health insurance is not magic. Insurance can give you access to a doctor, but it cannot make you take your medicine. An absolutely astonishing percentage of people don’t take their hypertension medication, even though the side effects are minimal, the medications are cheap generics, and the benefits are large; overall, about 50 percent of patients seem to fail to take their medicine as recommended. Insurance cannot make you stop drinking or smoking or overeating. Just ask the primary-care physicians charged with bullying people out of those behaviors. Yes, people with health insurance are more likely to be healthy. But people with good health insurance are also more likely to be successful folks who are remuneratively employed, which seems to be independently correlated with health, even when everyone is getting their health care through Britain’s National Health Service.

Even as some people use health insurance to get their blood pressure down or their lung disease treated, others will fail to comply with treatment regimens, or will get treatments that turn out in hindsight to have been a bad idea, or will spend money on unhealthy things now that they're spending less on health care. Which effect ends up being larger in aggregate is an empirical question, even though individuals may be able to point to a very clear benefit or detriment to themselves.

That empirical answer ends up being … kind of cloudy. Some studies show large health benefits from health insurance, others little, none, or even negative effects.

Consider a study that set out to determine whether giving people health insurance reduced their later risk of dementia. (Entirely plausible: One source of dementia is small strokes that slowly degrade brain function, and hypertension drugs prevent strokes. It seems reasonable to think that long-term control is better than attacking the problem when you turn 65, and hypertension is something we’re very good at controlling, so gaining access should really and clearly help reduce rates of dementia.)

The problem is that studying the problem would take decades, and tracking all those people would be a herculean administrative labor as they moved, married, died or simply stopped opening your envelopes. Moreover, you’d have a terrible time even sorting out who was uninsured, because most of the uninsured don’t stay that way for very long; do you count someone who as uninsured who lacked insurance the first year of the study, but then had good insurance for the next 20? These are the sort of quandaries that plague health-care researchers every day. The things that are easiest to measure, like short-term responses to drugs, or surveys asking people how healthy they feel, are often not really the things we’d most like to know. We see but through a glass darkly, and maybe we always will.

Seven years ago, I raised the question of whether health insurance actually saved lives, and ended up concluding that my best guess was “Probably, but probably not that many, or the data wouldn’t be this noisy.” That was my prediction for the Affordable Care Act, and so far I’ve had no reason to revise it. A substantial number of null results, with some positive, and a few negative, is basically what I’d expect from noisy variance around a relatively low mean improvement in health. Or measurements that can’t give us the information we’re really looking for.

Wherever you stand on Obamacare, and however you feel about the Republicans' failure to replace it, let's all maintain a little perspective about what we don’t know, including: how beneficial insurance actually is.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

    To contact the author of this story:
    Megan McArdle at

    To contact the editor responsible for this story:
    Philip Gray at

    Before it's here, it's on the Bloomberg Terminal.