The Impact of Food Stamp Program Participation on Adult Obesity and Child Overweight Status

Year: 2009

Research Center: Department of Nutrition at the University of California, Davis

Investigator: Schmeiser, Maximilian

Institution: University of Wisconsin-Madison

Project Contact:
Maximilian D. Schmeiser
Department of Consumer Science
University of Wisconsin-Madison
1305 Linden Drive
Madison, WI 53706
Phone: 608-301-5063


Over the past three decades, the prevalence of child obesity in the United States has more than tripled and the prevalence of adult obesity has more than doubled. The increasing prevalence of obesity has become one of the most pressing public health challenges facing the United States and other industrialized nations given its association with numerous health conditions, higher medical expenditures, and a greater risk of mortality. A clear income gradient exists in the prevalence of child obesity and that of adult females: Low-income children and women are significantly more likely to be obese than those from families with higher incomes.

One suggested cause of the higher prevalence of obesity among low-income families is participation in the Food Stamp Program (FSP), now the Supplemental Nutrition Assistance Program (SNAP). Researchers have documented that obesity rates are higher among FSP participants than among eligible nonparticipants and that the provision of FSP benefits increases expenditures on food. However, participation in the FSP also could decrease the likelihood of being obese, particularly among children, given that it could reduce food insecurity, decrease the number of meals consumed away from home, or increase the purchase of more nutritious, less calorically dense foods. Although several studies have examined the relationship between FSP participation and obesity for both children and adults, none have adequately addressed the possibility that the relationship is driven by unobserved determinates of both FSP participation and obesity, such as experiencing food insecurity and enjoying food.

This study contributes to the literature by estimating the causal effect of FSP participation on body mass index (BMI) and obesity for both adults and children by using data from the National Longitudinal Survey of Youth, 1979 cohort (NLSY79) and the NLSY79 Children and Young Adults cohort. The NLSY79 contains detailed longitudinal information on numerous demographic and economic characteristics, including FSP participation and self-reported weight and height. The NLSY79 Children and Young Adults data contains similar information on the children born to women in the NLSY79. Both datasets also contain State identifiers that allow for the addition of relevant State-level characteristics, including food prices and cigarette prices.

In order to generate causal estimates of the effect of FSP participation on BMI and obesity, a two-stage least squares estimation strategy is employed that exploits exogenous variation in FSP participation driven by the labor supply response to changes in the parameters of State and Federal Earned Income Tax Credit (EITC) programs. Specifically, the number of years in the previous 5 that a person has participated in the FSP program is instrumented for using the average value of combined State and Federal EITC benefits in a given State and year. State-level fixed effects are also employed in order to eliminate sources of time-invariant heterogeneity.

The findings for children suggest that participation in the FSP program actually reduces BMI and the probability of being obese. For boys ages 5-11, an additional year of FSP participation reduces the BMI percentile by 15 points and the probability of being obese by 6.7 percentage points. For girls ages 5-11, an additional year of FSP participation reduces the BMI percentile by 12 points and the probability of being obese by 7.5 percentage points. The similar direction and magnitude of the effect of FSP participation on obesity for young children is contrary to previous research, which found that FSP participation decreased obesity for boys and increased obesity for girls, but is much more plausible since there is little reason to expect differential effects of FSP participation for boys and girls ages 5 through 11.

The results for boys ages 12-18 are similar to those for younger children, with an additional year of FSP participation reducing the BMI percentile by 8 points and the probability of being obese by 7.4 percentage points. However, for girls ages 12-18, FSP participation appears to have an insignificant effect on BMI percentile and obesity. These differential effects for older children may be driven by the divergence in physiology that occurs at puberty, as well as differences in eating habits in adolescence.

The reduction in the BMI percentile and probability of obesity observed for most children as a result of FSP participation likely operates through its expansion of food budgets— reducing household food insecurity and binge eating, which have previously been associated with child obesity.

For adults, FSP participation appears to consistently increase BMI and the probability of being obese. An additional year of FSP participation is associated with a 1.6-unit increase in BMI and a 12.6-percentage-point increase in the probability of being obese for women. For men, an additional year of FSP participation increases BMI by 1.6 units but has no significant effect on the probability of being obese. These results suggest that, for adults, the FSP program serves only to increase food consumption and, thus, weight.

Direct inquiries about this study to the Project Contact listed above.