What's Government Insurance Really For? Lessons From the Medicaid Study

Medicaid didn't conclusively protect patients from diabetes. But it did protect them from financial ruin.

800 medicaid1.jpg

In a landmark study of Oregon's Medicaid patients, the results section begins ominously: "We found no significant effect of Medicaid coverage on the prevalence or diagnosis of hypertension or high cholesterol levels or on the use of medication for these conditions."

Uh oh. Does that mean Medicaid doesn't work? Does that mean Obamacare, which expands Medicaid, is doomed? Hold your horses.

A Mixed Finding, in Every Way
The very fact the Oregon study is considered one of the most important public policy studies of the decade is a mixed blessing for Oregon and health care wonks. The bad news (for Oregon) is that the state only had enough money to expand Medicaid coverage to a few needy families, so it did so randomly, by lottery. The good news (for wonks) is that this created the perfect conditions for a random controlled trial to determine whether Medicaid actually improved health outcomes for those lucky few.

And the results? They were mixed blessings, as well. Families randomly-selected for Medicaid spent more on health care, saw more doctors, and received more preventative care. But they weren't measurably healthier judging by blood pressure or blood sugar. Still, they were far less likely to report depression and far more likely to be financially secure.

I'm willing to bet that your reaction to these findings -- and mine -- will be determined by two things: (1) your prior opinions about Obamacare, Medicaid, and the importance of expanding health access; and (2) your deeper views on what health insurance is for. People don't like to change their mind, and based on my reading of WonkWorld's reaction to this study, practically nobody has.

But the more interesting question to me coming out of the report is the most basic one: What the point of government health insurance?

Medicaid and Health
The most obvious answer to the question would be that health insurance is for health -- like car insurance is for cars, or home insurance is for homes. So Oregon's results come as something of a shock.

But there are a few mitigating factors. First, improvements in mental health "are still improvements in health," as Aaron Carroll and Austin Frakt write. One hardly needs to elaborate on the importance of mental health in the wake of Newtown.

Second, among the randomly selected Medicaid population, the researchers actually observed fewer patients with high blood pressure; more patients on hypertension meds; fewer people with diabetes; and fewer people with high total cholesterol. Medicaid improved numbers in all of these categories, but the results were not statistically significant.

I'm not trying to sugar-coat finding that should be disappointing to every health care access advocate. But there is a big difference between (a) a study conclusively showing Medicaid doesn't improve health, and (b) a study showing Medicaid's health improvements aren't statistically significant.

And that's enough on confidence intervals for today.

Medicaid and Poverty
Medicaid didn't conclusively protect people from diabetes. It did conclusively protect people from financial ruin. The most important positive finding from the study was that the share of Oregonians on Medicaid who faced out-of-pocket medical expenses exceeding 30 percent of their income fell from 5.5 percent to 1 percent.

This point can scarcely be understated. Beyond keeping us safe from attack, the primary role of modern government is to keep its citizens safe from poverty -- poverty from circumstance, poverty from retirement, and poverty from medical bankruptcy. Social Security, Medicare, Medicaid, income security, unemployment benefits, and veterans benefits .... these programs, and their close relatives, make up about two-thirds of the federal budget (not to mention a majority of practically every advanced country's budget in the world). All these functions can be seen as insurance. They take money from Americans, pooling risk from 150+ million families, and distributing the funds to protect at-risk families from financial shocks.

Successful government programs successfully fight poverty. Consider Social Security, which smooths consumption and provides a safety net for retired Americans. A paper from Gary Engelhardt and Jonathan Gruber found that Social Security benefits accounted for the entire 17-point decline in elderly poverty between 1967 and 2000.

Screen Shot 2013-05-02 at 9.13.49 AM.png

Health insurance should provide similar protection. "The primary purpose of health insurance is to protect you financially in event of a catastrophic medical shock," Amy Finkelstein, an author of the study, told Jonathan Cohn in an interview, "in the same way that the primary purpose of auto insurance or fire insurance is to provide you money in case you've lost something of value."

If we want to improve the way we spend money to improve health outcomes and alleviate poverty, we should compare Medicaid to money rather than compare Medicaid to nothing. As Matt Yglesias writes, the most useful random study would pit $1,000 in Medicaid against $1,000 in cold hard cash and see what happens. I tend to think government's purchasing power in the health care market is not only useful, but essential, since its "consumers" aren't really consumers at all, but rather patients who need protection and bargaining power. But who knows. Just because the U.S. federal government is the best insurance company in the world doesn't make it the perfect insurance company for everything.