Wednesday, December 10, 2014

Costs and Benefits of Desire

"Accounting for the facts that healthy foods are otherwise less desirable and that consumers already have some information about health, the net benefit to consumers possible from consuming healthier foods is 30-40% of the value of the gross health benefit from switching to the healthiest possible diet."
What "facts"? A Reuters report on Monday told the story of the $5.27 billion in "lost pleasure" estimated in a U.S. Food and Drug Administration analysis of product labeling. According to the report, to arrive at that estimate, "the agency relied almost solely on a 2011 paper by then-graduate student Jason Abaluck."

In all fairness to Abaluck, the paper strikes this reader as an earnest and diligent graduate student exercise in mathematical modeling. Of course quantifying the "otherwise less desirable" characteristics of healthy foods is sheer nonsense. But that's not an issue for mathematical modeling. Do the conclusions follow rigorously from the assumptions? That's all that counts. Assuming that healthy foods are otherwise less desirable... But why would you?

3 comments:

Thaomas said...

The student seems to have made an error. On the margin "healthy" and "unhealthful" foods are of equal value. If they were less desirable, consumers would purchase more unhealthful foods until they are equal. Now labeling will have some costs which will lead to some decline in purchase of both healthful and unhealthful food and this decrease in purchase could be looked at as a cost of the health benefits that would accrue to the shift from less to more healthful foods, assuming that would occur. But applying cost benefit analysis to the decision about labeling does not seem wrong in principle.

Thaomas said...

The student seems to have made an error. On the margin "healthful" and "unhealthful" foods are of equal value. If healthful foods were less desirable, consumers would purchase more unhealthful foods until they became equal. Now labeling will have some costs which will lead to some decline in purchase of both healthful and unhealthful food and this decrease in purchase could be looked at as a cost of the health benefits that would accrue to the shift from less to more healthful foods, assuming that would occur. Notwithstanding the student's error, applying cost benefit analysis to the decision about labeling does not seem wrong in principle.

Thaomas said...

Sorry. The attempt to "preview" seems to have resulted in a "publish."