Note: This going to get a bit nerdy

By my estimation Megan McArdle has one the most complete understandings of the obesity epidemic in the econ blogosphere. It seems to be Megan’s contention, however, that the declining cost of calories is central to the rise in obesity. People like to eat and eating has become cheaper, she says. This doesn’t seem quite right to me.

Lets think about what we need to make a model like that work.  First we are assuming a calories-in / calories-out model of obesity. People choose to eat more calories than they choose to expend. There are some basic problems with that obesity model including, how people managed to balance this calculation for 1000s of years without the benefits of nutritional science or how this jives with the evidence that exercise does not induce weight loss.

Lets gloss over that and just get to what it says about weight gain. The issue is that if we use a calorie model we have to suppose that people stop gaining weight because the extra weight they are carrying burns off so many calories that it counterbalances their extra eating.

Remember, the calorie equation is a flow equation. Thus eating more calories than you expend doesn’t imply being fat. It implies being fatter. Unless at some point calories-in falls or calories-out rises, the person will grow fat without bound. This doesn’t seem to describe most people. Indeed, most overweight people are in equilibrium. Their weight is stuck where it is and thus they must be burning exactly as many calories as they are ingesting.

So implicit here is that the fattening process is halted by the extra calories burned by being fatter. The standard calculations imply that a sedentary woman has marginal burn rate of 5 calories per pound. Now suppose that our this woman decided to have one Grande 2% Cafe Mocha on her way to work.

The Cafe Mocha contains 330 calories which raises her equilibrium weight by 66 pounds. In BMI terms, if she started out as 5’4” and 130 lbs she would have a BMI of 22.3, well in the normal range. Indeed, its normal by the official standards, most people today would probably describe her as thin.

In equilibrium she would wind up with a BMI of 33.6, well into the obese range. In the visual chart below she would move from the third position to the the seventh position.

Note that these calculations imply that she has become no less active as a result of the increase in her size. She lives exactly the same life as before except now she has a Cafe Mocha and is 66 pounds heavier.

So far this may or may not seem realistic. Having slowly watched my entire generation become addicted to Starbucks, 10 –15 pounds seems closer to reality than 66. Still, there is more.

To follow the cheap calorie hypothesis we have to assume that a drop in the price of calories could have induced a move like this. It seems unlikely that the change in the price of Cafe Mochas was enough to compensate someone for this change in physique. Would we all gain 100 pounds if Cafe Mochas were a quarter each? I am skeptical but maybe its not Cafe Mochas that are doing it. Lets look at food prices more generally.

According to the USDA the monthly cost of a moderate food plan to woman aged 19 –50 is \$238. The Consumer Price Index for food is 2.18 times higher than it was in 1983. Disposable income is 4.04 times higher. This implies that food is 54% cheaper as a percentage of income today than it was in 1983. Thus, a moderate food plan would have cost \$517 or \$279 more in income adjusted terms back in 1983. That works out to an additional cost of food of \$3384 a year.

This then suggests that having to pay an extra \$3384 was enough to prevent our hypothetical slender woman of 1983 from becoming an obese woman today. In economic terms the net (of food benefit) utility cost of obesity is less than \$3384.

It does not seem that are many women for whom \$3384 would be sufficient for them to choose to become obese. Nor does it seem likely that very many obese woman would be able loose all of their extra weight if offered \$3384. Now perhaps all of the obesity epidemic is being driven by a very small fraction of Americans who face a very low net utility cost of obesity.

It doesn’t make sense, however, that Americans would spend so much time and money on exercise – which they believe will make them thinner – as well as diet books and supplements. After all, their inducement to eat more is only being driven by a savings of a few thousand dollars per year. If having to pay a few thousand dollars less for food makes it worth it to get fat then why is it worth it to spend so much time and money to get unfat?

It seems likely to me that the connection between food consumption and obesity occurs in ways that are not directly observable to consumers. That is, consumers believe in a calories-in / calories-out model of obesity and form their expectations of the costs and benefits of eating based on it. If this is not indeed the correct model then consumers will consistently err in their consumption decisions.

The corollary to that is that obesity is proceeding through some other food consumption choice channel. Something about the mix of food is changing and the causal chain is not immediately clear.

My guess is that it has something to do with the glycemic load of our food. Larger portions, higher glycemic carbs, “unbalanced” snacks and perhaps less dairy cause sharper spikes in insulin levels and eventually lead to hyperinsulimia. This frequent readers know, is how I would explain the strong connection between obesity and diabetes even while liposuction does nothing to help diabetes.

All of this of course assumes that changes in the market for food are at the heart of the obesity epidemic. I strongly suspect that is true but it is not certain.