Health Insurance and Health: the Latest Science

There is a long-running debate about whether health insurance really makes people healthier; some studies have found not much effect either on death rates or on whether people with chronic conditions like diabetes do better with insurance. Other studies have found big effects. The problem is complicated because in America having health insurance tends to go together with other stuff that we know effects health, like education and income. Once you correct for these things, some studies have found that people with insurance get more health care but don't end up healthier.

Via Marginal Revolution, here is a big new paper by three top academics reviewing all the recent evidence on the question. I would summarize it as follows:

• health insurance leads to a significant improvement in people's financial situation: fewer bankruptcies, fewer run-ins with bill collectors;
• health insurance leads to a general reduction in life stress;
• health insurance leads to a large (circa 30%) reduction in depression, probably through a combination of stress reduction and treatment;
• health insurance leads to modest improvements in the lives of people with chronic conditions;
• health insurance leads people to feel better about their health, with more reporting good health and fewer reporting bad; this sounds trivial but actually over a ten-year period people who say they have poor health have a mortality rate two to eight times higher than people who report good health, so it is one of the most powerful measures we have;
• health insurance leads to a reduction in mortality rates equal to one fewer death per year for each 239 to 830 people who get insurance

The numbers for mortality are difficult partly because none of these studies has covered more than a decade, and people die for all sorts of reasons that health insurance can't touch. Hence the wide range of the estimates. But even one death/year for each 830 people is a significant number; if the CBO is right and the House health care plan would lead to 20 million people losing insurance, that adds up to 24,000 additional deaths per year. Plus we know that poor Americans have higher death rates than poor Europeans and Japanese who all have health insurance, so the macro-scale picture fits with this conclusion.

None of this is cheap, of course; one study found that one life was saved for each $327,000 to $867,000 the government spends on Medicaid expansion.

Some libertarians have argued that having insurance doesn't matter because poor people manage to get the most important care by showing up at emergency rooms, and this is true to some extent. But I find this to be a bizarre argument, because the cost of their care has to be paid by someone, and under the current system that means it is paid by people who do have insurance. Does that make sense?

Plus, you know, Americans say they are upset about inequality. Pretty much the only significant thing the government has done recently to fight this problem is Obamacare, which takes hundreds of billions from the rich to give the working class health insurance. I think that is a good idea.