This is quite interesting. Two studies say that government subsidies appear to be associated with increased childhood obesity. Of course, one has to be careful of endogeneity, since it is likely that poor kids are more likely to be obese, and poor kids are more likely to receive subsidies. (May I point out that the fact that poverty goes with obesity would have seemed like a bizarre claim just 50 years ago? Things can't be THAT bad if our poor people are fat, right?)
Does the National School Lunch Program Improve Children’s Dietary Outcomes?
Benjamin Campbell et al.
American Journal of Agricultural Economics, July 2011, Pages 1099-1130
Abstract: The National School Lunch Program’s effect on children’s diets has been
extensively studied. Results have tended to be inconclusive regarding the effectiveness of the program. Utilizing more specific treatment groups, we find that participants in the National School Lunch Program do not consume a higher-quality diet at lunch than children choosing not to participate, even though the program is offered — but rather consume a higher quantity of foods while consuming similar amounts at other meals. Furthermore, children attending schools not participating in the National School Lunch Program have dietary outcomes that are not significantly different from program participants.
Child care subsidies and childhood obesity
Chris Herbst & Erdal Tekin
Review of Economics of the Household, September 2011, Pages 349-378
Abstract: In this paper, we study the impact of child care subsidy receipt on low- income children’s weight outcomes in the fall and spring of kindergarten using data from the Early Childhood Longitudinal Study, Kindergarten Cohort. Our results suggest that subsidy receipt is associated with increases in BMI and a greater likelihood of being overweight and obese. Using quantile regression methods, we find substantial variation in subsidy effects across the BMI distribution. Specifically, child care subsidies have no effect on BMI at the lower end of the distribution, inconsistent effects in the middle of the distribution, and large effects at the top of the distribution. Our results point to the use of non-parental child care, particularly center-based services, as the key mechanism through which subsidies influence children’s weight outcomes.
I am going to go out on a limb here, and guess that the answer from P-Kroog is "The program needs to be bigger! Much BIGGER!"