Calorie Counts

The idea of mandating calorie counts on restaurant menus touches on a lot of things that interest me--how/why people make the choices they do, government intervention in the economy, food. I haven't been following the issue too closely, but the recent exchange between Jacob Sullum and Ezra Klein caught my attention.

Klein (who favors mandating the posting of calorie counts) tries to counter Sullum's two major arguments, but in so doing makes the classic mistake of listening to what people say instead of what they do.

His article on the subject basically makes two points: The first is that consumers don't want this, because if they did, then the market would already have provided it. As Sullum says, "If customers really were clamoring for conspicuous calorie counts, restaurants would provide them voluntarily." That sentence competes for space with a poll showing 84 percent of Californians support caloric labeling requirements, and the basic reality the article is responding to: Democratically elected legislators who depend on the favor of voters for their jobs are the ones trying to pass a bill. Because they think it popular. The idea that public preferences only have legitimacy if they're strong enough to be heard atop the clamor of the market is an exceedingly odd one.

The second is that they won't work. This appears to be a misread of a new survey from the New York health department. The researchers polled 7,318 customers at nearly 300 franchises of 11 fast food chains. Of these chains, the only one that posted calorie information in a usable space was Subway. At Subway, 32 percent of consumers reported seeing it (it's posted near registers, though not on the menus or menu board), and 37 percent of that 32 percent said it was a factor in their purchasing decision. "In other words," concludes Sullum, "simply making people aware of calorie content is not enough to affect their food choices."

Jacob Grier, guest blogging at the Agitator, shreds Klein's contention against Sullum's first claim:

Well, ok. But the notion that conducting a poll is a more reliable way to gauge consumer preferences is even odder. Answering a question in a poll is not like ordering lunch in a restaurant. Facing no trade-offs, there’s no reason not to give the publicly virtuous answer. Of course most people will say they support posting caloric information. Faced with the actual trade-offs of less menu space, higher costs for testing new products (more significant for small chains than for large), and the sometimes unpleasant reminder of how dense some food is, they might not actually prefer the one-size-fits-all rule of posting calorie counts prominently on the menu. If 84% of consumers were really demanding it, you would think that at least one restaurant chain would have filled this demand. The fact that none has done so voluntarily suggests that the mandate is excessive. (And what of the rights of business owners? They don’t merit concern, apparently.)

As for Klein's attack on Sullum's second claim, 37 percent of 32 percent is about 12 percent, which means that a total of 12 percent of respondents self-reported that the information presented influenced their behavior. First, survey respondents are notoriously bad at self-reporting behavior (especially when there's an inherent moral element to their response), and second the survey Klein cites is was a sort of best-case scenario, since many people choosing to go to Subway do so because they're concerned about their calorie consumption.

Image from here.