Now that we're getting ready to spend a whole lot of money providing health care for other people, obesity is becoming an increasingly pressing topic.  There's a fair amount of controversy over how much upward pressure obesity puts on health care costs, but the wear and tear on joints alone heralds millions of dollars worth of arthritis drugs and knee replacemetns in our public health plan's future.

So it's not surprising that writers and bloggers are discussing our rising rates of obesity, and possible ways to counteract that.  Matthew Yglesias writes:

But there's a very profound problem of evolutionary psychology here. For the vast majority of human existence people were engaged in much more daily physical activity than is the typical member of a contemporary rich society and it was impossible to be certain that food would be available in the future. Consequently, people are largely designed with the instinct to err on the side of eating more food rather than less. Especially if the food is tasty. These days, of course, we're in a very different situation. Nobody starves to death in the contemporary United States, but lots of people have problems related to poor dietary habits.

Hardly an original point on my part. But the sign made me think of it. And I suppose I would make the point that at the margin expenditures of funds to fight this tendency are going to do a lot more to improve public health than will expenditures of funds to treat people's diabetes.

That presumes you can find a marginal dollar that will reduce peoples' tendency to eat more than they burn just as effectively as we treat diabetes.  Seven years ago, when I investigated fast food lawsuits, I found very little evidence for that proposition.  Pretty much every public health effort to get people to eat less has proven a dismal failure.  As Paul Campos has noted,  telling people to eat less and exercise more is the most exhausitvely attempted experiment in the history of science.  And we have 200 million data points that prove just how badly it works at keeping Americans slender.

I'm thus pretty skeptical that we're going to do much about obesity through the sort of mild nudges that a lot of the discussion about changing peoples' eating habits implicitly imagines.  For example, I'm a big fan of Brian Wansink's Mindless Eating, which details all the ways that we take in more calories than we think.  But I'm skeptical that in the long run, these factors make all that much difference.  If you think about it, taking in an extra fifty calories a day more than you need--half a piece of bread, or a few cocktail nuts--is enough to pack on an extra five pounds a year.  If we really at that mindlessly all the time, we'd all be morbidly obese.  I expect that in any given sitting, things like portion size, or calorie counts, can cause people to reduce their intake.  But over time, I doubt they'll have much impact.

That leaves more illiberal options, like forcing manufacturers to change their foods in order to make them less apealling, or massive taxes on fat, sugar, and salt.  Even if I thought there was a practical, and politically acceptable, way to carry this out, I'd be against it.  Enacting national health care, and then declaring that it now gives you the right to dictate how people will eat . . . well, that's exactly the sort of thing libertarians are talking about when they bemoan the creepign nanny state.