Let’s Do Lunch: Factors Associated with Student Participation in the USDA’s National School Lunch Program in California, 2010-11 to 2014-15

Year: 2015

Research Center: Center for Regional Development, Purdue University

Investigator: Turner, Lindsey and Carl Siebert

Institution: Boise State University

Project Contact:
Lindsey Turner
Director, Initiative for Healthy Schools
College of Education
Boise State University
1910 University Drive
Mailstop 1745
Boise, ID 83725
Phone: 208-426-1632


The National School Lunch Program (NSLP) is the second largest of the major food and nutrition assistance programs administered by the U.S. Department of Agriculture (USDA), providing meals in over 100,000 schools. The types of foods and beverages served in the NSLP play a crucial role in shaping children’s dietary intake. As directed by the Healthy, Hunger-Free Kids Act of 2010, USDA released updated nutrition standards for school meals with implementation required at the beginning of school year (SY) 2012-13. There has been concern and conflicting information about whether—and how—the implementation of new meal standards would impact student participation rates.

This project assembled a comprehensive data set for all public school districts in California that included the following: 1) administrative claims data on NSLP meals sold, month by month, at each public school district from SY 2010-11 to SY 2014-15; 2) the district’s average full price charged for an NSLP lunch each year; and 3) demographic and contextual covariates. Analytic aims were to: 1) assess time trends in NSLP participation rates; 2) assess trends in NSLP prices; and 3) develop a combined model to examine the association between changes over time in NSLP participation, lunch price, and district characteristics.

Data were available for 868 public school districts in California. Regression models were used to examine variations in trends over time within each district. All models account for covariates (district size, socioeconomic composition, locale, price, and month/year). Participation data showed slight—but statistically significant—declines over time in average annual participation rates at all payment types (free, reduced-price, and full-price) between SY 2010-11 and SY 2013-14, but then a significant upturn in SY 2014-15.

Rural Disparities: Participation rates at all payment levels dipped in both urban/suburban and rural/township districts, but have not yet rebounded in rural/township districts. Further work is needed to understand more of the nuances involved in these patterns, as well as to identify viable strategies to improve participation rates in rural/township districts. However, the data clearly show disparities in rural communities.

Socioeconomic Disparities: After accounting for all other demographic covariates, participation rates—at all payment levels—were significantly higher in districts with 60 percent or more students eligible for free or reduced-priced meals. In other words, where it is the social norm for students to receive lunch, participation rates are higher, even for students paying full price. Over time, lunch participation rates in more disadvantaged districts remained higher than in less disadvantaged districts.

Pricing: In cross-sectional multivariate models, price was significantly and inversely associated with participation rates, with a 1-percentage-point decline in participation with each 5-cent increase in the price charged for a full-price lunch, based on weighted average per-district prices. To examine longitudinal trends, latent class analysis was used, and identified 12 unique trajectories in lunch pricing. Most districts (88 percent) had steady increases, decreases, or flat trends in price; 9 percent had universal-free pricing in all years (that is, prices of $0); and 3 percent transitioned to universal-free pricing at some point during the 5 years.

Among the districts with non-zero pricing, there was a nearly perfect inverse association between participation rates and pricing over time. Lunch participation rates were lowest for the districts with the highest overall lunch prices, and participation rates were highest in the districts with the lowest average prices. Participation varies not only due to the rate of change in price, but also on the absolute price.

Among the 74 districts that offered lunch at universal-free price throughout the entire time period examined here, average total participation rates started—and remained—very high over the 5-year period, at just over 80 percent. This is consistent with other evidence showing that the dropoffs in participation in 2012-13 and 2013-14 were primarily due to fewer meals being served at paid status. Although there were only a few districts (n = 28) that moved to districtwide universal-free pricing, their participation data showed noticeable upturns in participation rates after that change (between 1 and 6 percentage points).

Conclusions: Analyses showed a clear and consistent trend of decreases in participation rates at all price levels, with an upturn in SY 2014-15. This is consistent with other national data on participation, with claims data showing increases in participation between 2014 and 2015 in a majority of States.

More than 97 percent of districts across the country now meet the updated meals standards, and the new standards have significantly improved the nutritional qualities of school lunches. With time, and support from many resources, school foodservice programs across the country seem to be identifying successful implementation strategies that meet students’ taste demands.

Other analyses of school-lunch participation rates also note that declines in participation began long before the revised meal standards, and have concluded that the new standards are not the cause of decreased participation. With time-coinciding events, it is often difficult to identify causal associations, but several patterns of results in this current study can help to parse these connections. Several associations found between pricing and participation suggest that changes in participation have had more to do with national economic factors than with student pushback on the new meals, such as results showing: 1) minimal changes in participation in universal-free pricing districts; 2) a low but nearly flat trend in participation in paid meals at more-affluent schools; and 3) a significant upturn in participation rates among all subgroups in SY 2014-15 as the Great Recession eased. In addition, the new Smart Snacks standards for competitive foods and beverages that went into effect at the start of SY 2014-15 may have had the intended effect of reducing unhealthy competition for students’ lunch dollars, resulting in more students buying school lunch.