Participation in the National School Lunch Program and Food Security: An Analysis of Transitions Into Kindergarten

Year: 2013

Research Center: Southern Rural Development Center, Mississippi State University

Investigator: Arteaga, Irma, and Colleen Heflin

Institution: University of Missouri

Project Contact:
Colleen Heflin
University of Missouri
Truman School of Public Affairs
112 Middlebush Hall
Columbia, MO 65211
Phone: 573-882-4398


There is ample evidence that food insecurity has negative consequences for childhood development, including, but not limited to, neurological damage in the prenatal period, susceptibility to disease in infancy, and poor school performance in the elementary years. The National School Lunch Program (NSLP) was designed as a nutritional safety net for low-income children who are at risk for these outcomes. The NSLP is the second largest Federal nutrition program behind the Supplemental Nutrition Assistance Program (SNAP), formerly the Food Stamp Program. In 2011, 97 percent of public schools participated in the NSLP and over 31 million students received a free or reduced-price lunch daily. Two-thirds of households with food insecurity among children report participation in the free or reduced-price school lunch program in the last 30 days.

The prevalence of the NSLP among food-insecure families raises an important question: Does the NSLP not only provide meals during school, but also help to protect whole households from food insecurity more generally? This study uses two different datasets and research methods to examine the change in household and child food security as children enter kindergarten and are able to access the NSLP. The study is unique because it focuses on the NSLP, while most other food policy studies examine SNAP, and because it focuses on the transition to kindergarten, which is a time period that is especially important for future school success. A few prior studies have examined the NSLP and food insecurity, but the results have not been clear.

Data are drawn from the Early Childhood Longitudinal Survey, Birth Cohort (ECLS-B), a nationally representative dataset drawn from a diverse range of families in the United States. The ECLS-B is a sample of 10,700 children conducted by the National Center for Education Statistics of the U.S. Department of Education. It examines the development, health, and learning environment of a single cohort of U.S. children born in 2001, including information about the home, parenting practices and behavior, and educational experiences. From these data, the study draws a sample of 3,600 children to study the impact of the NSLP on household food insecurity among households with incomes at or below 185 percent of the Federal poverty level and with a kindergarten-aged child. Additionally, the study uses a smaller sample (n=650) to examine whether participation in the NSLP increase transitions into food security, given food insecurity prior to kindergarten entry and upon kindergarten entry.

The outcome variable of interest is food insecurity in households with children. The analysis uses a 12-month and 30-day food security variable, constructed from the U.S. Department of Agriculture’s food security module that was administered in the ECLS-B. The independent variable of interest is school lunch participation. In the ECLS-B, the parent reported participation for the focal child. All models include controls for the child’s age, gender and race; and maternal education, marital status and household income.

Because families with higher need may be more likely to enroll in the NSLP, participants might differ in important ways from non-participants, interfering with the ability to compare the two groups. To correct for this potential selection bias, the study uses a measure of age in months from the State cutoff age for kindergarten entry. State policies differ on kindergarten-entry age, and children are born randomly throughout the year, making this a good source for naturally occurring exogenous variation.

The analysis uses instrumental variable models to estimate the NSLP’s causal relationship to food insecurity: The first strategy uses age relative to the kindergarten eligibility cutoff point; the second strategy uses the longitudinal data to estimate the probability of becoming food secure associated with NSLP participation, conditional on being food insecure prior to kindergarten entry. The study finds participation in the NSLP improves household food security at least two percentage points. Sensitivity analysis demonstrates that school entry is not associated with reductions in food insecurity among families whose incomes are above 185 percent of the Federal poverty line and ineligible for the NSLP. This increases confidence that the observed change in food status in the study’s sample is due to participation in the NSLP. Additionally, controlling for the reduction in child care hours among low-income households on school entry does not diminish the size of the NSLP effect.

The results of this study suggest the NLSP helps protect low-income households from food insecurity. Specifically, a child’s participation in the NSLP reduces the families’ probability of being food insecure by at least 2 percentage points. This finding is significant because it supports the growing literature demonstrating the efficacy of the Federal food program, and because it identifies a protective factor that is activated at kindergarten, a critical developmental stage when experiences occur and habits are formed that affect the life course.

While there may be other factors correlated with school entry that were not included here that may affect food security status, those remain unspecified and should be identified and examined in subsequent analyses.