Background: A growing number of children and their households live in food insecurity (i.e., lack of consistent access to adequate food). The food insecurity rate among households with children increased to decade-highs between 2008 and 2011 due to the recent economic crisis, and reached 20% in 2012. More than 8 million children lived in households experiencing food insecurity. The National School Lunch Program (NSLP) is one of the most important food assistance programs to ensure that low-income, school-aged children have access to adequate nutrition and to mitigate the problems associated with malnourishment among children. However, it is unclear to what extent the NSLP reduces food insecurity among children. One challenge to assessing the impact of the NSLP is the potential selection bias that households with a high risk of food insecurity are more likely to participate in food assistance programs.

Methods: The study takes advantage of seasonal differences in the NSLP participation—i.e., the program does not provide services in summer months when school is not in session—to examines its impact on food insecurity and food insufficiency for low-income children (aged 5-18 years) using four panels of data from the Survey of Income and Program Participation (SIPP). The SIPP sample is randomly divided into four rotation groups, and each wave of interview is conducted in four consecutive calendar months for these groups, respectively. Therefore, the reference period of each wave is fixed to four months, but covers different calendar months for four rotation groups. This difference in calendar months by rotation group creates variation in the number of summer months (defined as June-August) included in the reference period, and can be used to identify the NSLP’s impact on food insecurity. The study has two dichotomous dependent variables, including household food insufficiency (Yes = 1 and No = 0) and household food insecurity (Yes = 1 and No = 0). The major independent variable is the dummy indicator of summer months (Yes = 1 and No = 0).

Results: The monthly food insufficiency rate is nearly 4% and the food insecurity rate is about 25% in the four-month reference period among low-income households with children aged 5-18. Fixed-effect analysis suggests that the monthly food insufficiency rate is 0.6% higher in summer months for households with children receiving free/reduced-price lunch (p<.01). The different-in-different analysis finds that one more summer month in the reference period increases the difference in food insecurity rates between recipients and non-recipients eligible for free/reduced-price lunch by 1.5 percentage points (p<.01). 

Conclusion: This study suggests that the NSLP has positive impacts on food sufficiency and food security among low-income households with children. Results indicate a 14% reduction in the probability of food insufficiency and a 20% reduction in the probability of food insecurity among recipients of free/reduced price lunch. Given the growth of food insecurity among households with children, it is important to increase their access to school meal programs and summer food service program.