Paul Krugman's column today, on the "cost-shifting" in U.S. health care, is very thought-provoking, but let's highlight just one part:

Yet the cost of providing medical care to those denied private insurance doesn't go away. If individuals are poor, or if medical expenses impoverish them, they are covered by Medicaid. Otherwise, they pay out of pocket or rely on the charity of public hospitals.

Indeed, I've often wanted to make a clever case for covering the uninsured by arguing that society ends up paying for these costs anyway. Surely all that charity and uncompensated care going on right now is coming out of everyone else's pocketbooks eventually, either through increased government spending or hikes in premiums. And surely the fact that the uninsured can't get preventive care when they need it means that they (or someone else) ends up paying far, far more later on for diseases that could've been headed off early.

Sadly, though, after rooting around for numbers on this, neither of these arguments seems particularly compelling.

First, the uncompensated care issue. This interesting paper by Jack Hadley and John Holohan show that the uninsured receive about $34.5 billion per year in "uncompensated care," i.e. care that they receive but do not pay full price for. Of this amount, the federal and state governments eventually ended up paying for about $30.6 billion. So there's not a huge cost-shift onto those with private insurance policies. "Ah, you say, but if the government pays for the vast majority of uncompensated care anyway, through various subsidies and other weirdness, doesn't it make sense just to spend that money on coverage for the uninsured instead?

Well, not quite. It would cost a good deal more than just shifting that $34.5 billion to cover the uninsured. First of all, Hadley and Holohan have noted elsewhere that the uninsured currently receive about $98.9 billion in health care per year, which includes uncompensated care, out-of-pocket costs, and private or public insurance sources. If all of those uninsured people were to be brought in to existing public insurance programs (Medicaid, etc.), then total health care spending would rise by $34 billion. However, total government health care spending would increase by about $100 billion when all is said and done (even after subtracting the drop in subsidies for "uncompensated care")—in part because many people currently with private insurance would drop what they have and then sneak into the newly expanded government program. So most taxpayers would be paying far, far more for the uninsured than they currently do.

Here's another way to look at it: in 2001 the full-year uninsured received $1,253 in care, about half what privately insured people received ($2484). So from all appearances, and as cynical as this sounds, it's much cheaper to keep fifty million Americans uninsured than to spend a little extra and draw them into the program.

Okay, now the other economic argument for covering the uninsured is that, without insurance and hence, without preventive care, these poor folks—and they mostly are poor—are more likely to develop chronic diseases and whatnot that lead to expensive problems later on. So in the end, we as a society pay more by not insuring these folks early on. This sounds good—intuitively, it seems that it's cheaper to treat hypertension or diabetes earlier, rather than pay for hospitalization costs later. But there are a couple of reasons why this argument may not tell the whole story. One, as Phillip Longman has reported, only about 10 percent of all premature deaths in the last 30 years can be attributed to shortfalls in medical care. This doesn't exactly address my question, but it suggests that medicine plays a very small role in preventing serious illness. (Far more effective would be changes in behavior, like more exercise.)

The other possibility is that more preventive care may catch more "pseudo-diseases," i.e. "disease that would never become apparent to patients during their lifetime without testing." In other words, more care can increase costs by catching, say, benign tumors that would otherwise not be a problem. Finally, some hard research: Studies by RAND and economist Louise Russell showing that preventive care doesn't really lower overall health care costs in the aggregate, and may even increase them.

Okay, so this is a long post, but the brunt of it is: there probably isn't a good economic case to be made for covering the uninsured, although I'm obviously open to hearing one. Certainly there's a moral case to be made, and I think a overridingly powerful one, though I'm not sure how effective that's going to actually bringing change about. As Uwe Reinhardt once noted, the last time we had a budget surplus to spend, Americans chose tax cuts over helping the uninsured; no matter what the polls might say, our actual priorities seem pretty clear.