Recent Trends in the School Lunch Program: Selection, Nutrition, Health, and Achievement

Year: 2016

Research Center: Institute for Research on Poverty, University of Wisconsin-Madison

Investigator: Schanzenbach, Diane Whitmore and Elizabeth Debraggio

Institution: Northwestern University

Project Contact:
Diane Schanzenbach
Northwestern University
School of Education and Social Policy
2120 Campus Drive, Room 205
Evanston, IL 60208
Phone: 847-491-3884


In response to rising rates of childhood obesity, the first decade of the 2000s saw significant policy and media attention directed to the epidemic, resulting in various local, State, and Federal initiatives aimed at stemming or reversing the trend. Serving meals to over half of the Nation’s school-aged population each school day, the National School Lunch Program (NSLP) is in a unique position to launch far-reaching, nutrition-based efforts. Among these efforts, Congress passed the Healthy, Hunger-Free Kids Act of 2010 (HHFKA), making dramatic reforms to nutrition standards in the National School Lunch Program and the School Breakfast Program (SBP), and enhancing requirements for local school wellness policies governing nutrition and physical activity.

Building on past research that found school lunch to be a contributor to child obesity rates in the late 1990s and early 2000s, this study assesses the relationship between school lunch participation and obesity status for an updated cohort of children who entered kindergarten in 2010-11 (ECLS-K: 2010-11). Although HHFKA was not fully implemented until the studied cohort was in second grade, there is evidence that some schools initiated changes to improve the nutritional quality of their school meals prior to the full implementation of the law, potentially in response to heightened national attention to the issue of childhood obesity or in anticipation of full operation of the new school meal standards.

This study estimates the relationship between school lunch participation and the rate of weight gain from the beginning of kindergarten through third grade. The study uses the most recent wave of a panel dataset, the Early Childhood Longitudinal Study (ECLS-K) from the National Center for Education Statistics, that surveys information from a nationally representative sample of students beginning in their kindergarten year and through subsequent years, providing data on a child’s early life and school experiences, contexts, and outcomes. HHFKA implementation occurred in the studied cohort’s second grade. Therefore, data are available both pre- and post-HHFKA implementation. The paper uses a difference-in-differences approach, comparing the change in obesity rate from kindergarten to subsequent grades across children who participate in the school lunch program and those who bring lunch from home. The analysis is limited to students who are ineligible for free or reduced-price meals.

The study also measures changes in the nutrient content of school meals during the 2000s by comparing data from two waves of school lunch menu data from the School Nutrition Dietary Assessment (SNDA) conducted by the U.S. Department of Agriculture Food and Nutrition Service. This survey provides data on school food environments, including characteristics of the school meal program, information on competitive foods and other nutrition-related wellness policies, and the quantity and quality of nutrition content of school meals and snacks. Data are used to compare the nutrient content of meals across school poverty levels and over time.

The findings of this study show that prior to participating in the school lunch program, non-poor students who go on to eat school lunch are more likely to be obese at kindergarten entry compared with their schoolmates who brown bag their lunch. The study controls for selection into the program by comparing rates of growth in obesity after kindergarten entry.

By the end of third grade, the increase in obesity rates is statistically the same between school lunch participants and their schoolmates who do not participate, with an estimated coefficient of school lunch participation on obesity of 2.5 percentage points. Results from models where outcomes are body weight and incidence of overweight also show no differential gain in these measures for school lunch participants. These results contrast with work from an earlier time period that found a persistent statistically significant relationship between school lunch and obesity growth. There is some evidence of heterogeneous treatment effects, with increases in obesity among White students who participated in school lunch programs, but no impact among nonwhite students. Comparisons of obesity growth before and after HHFKA were inconclusive but suggested that improvements in school meals were evident prior to implementation of HHFKA.

Analysis of school lunch menu data shows a substantial decline in calories served during lunch between data collected in 1998-99 and 2009-10, with larger declines in high-poverty schools.

Recently, childhood obesity rates have plateaued. Though the study cannot directly measure the impact of HHFKA in changing the nutritional quality of school meals, the research finds some evidence that school meals became healthier over the study period and may be contributing to the flattened obesity rates among children.