by Katherine Ralston, Elizabeth Beaulieu, Jeffrey Hyman, Matthew Benson, and Michael Smith
USDA Economic Research Service
What Is the Issue?
USDA’s Farm to School Program was established in 2012 to improve access to local foods in eligible schools through grants and technical assistance. Farm-to-school programs bring locally or regionally produced foods into school cafeterias; provide hands-on learning activities such as school gardening, farm visits, and culinary classes; and integrate food-related education into the regular, standards-based classroom curriculum. Promotional activities and experiential learning also serve as an additional form of nutrition education to increase student awareness and interest in healthier foods. Frequent use of local foods in school meals has also been suggested as a way to expand the market for local agricultural producers.
What Did the Study Find?
This study is the first to focus on the prevalence of districts serving local food daily and the characteristics of those that do so more versus less frequently to identify potential targets for future technical assistance. While previous studies have examined the prevalence of farm-to-school activities, this study uses a “daily access” target, which provides a benchmark for assessing the depth of farm-to-school programs at different stages in their development.
The major findings in the study are:
- More than a third (35 percent) of all U.S. school districts reported serving local food in school meals during the 2011-12 school year.
- Nineteen percent of all school districts served at least one locally sourced food item daily. These districts, which tend to be larger, include 30 percent of all students.
- Locally produced milk—offered daily or more than weekly by 15.4 percent of school districts—and locally produced fruit, offered by 14.5 percent, were the local food categories served most frequently.
- The prevalence of serving local food daily differed significantly by region. Daily use of local food was least prevalent in the Southwest (8 percent of districts) and most prevalent in the Northeast (41 percent). After other factors that vary by region were accounted for, differences in the probability of serving local food daily were still statistically significant across regions.
- Holding other district characteristics constant, school districts with enrollment above 5,000, urban districts, and those districts in counties with higher density of farmers’ markets were more likely to serve local foods daily, as were districts with higher per capita income, higher levels of college attendance, and those in States with more legislated policies supporting farm-to-school programs.
- Districts with the lowest probability of serving local food daily had enrollments at or below 5,000, were in rural counties with lower density of farmers’ markets, or had lower per capita income, lower levels of college attendance, and fewer State-legislated farm-to-school policies.
- When the associations between school district characteristics and daily use of local food were modeled separately by region, we found that district size, locale type, and farmers’ market density were associated with daily use of local food in almost all regions. However, per capita income, school spending per student, foodservice labor costs, and certification rates for free and reduced-price meals were not significantly associated with daily use of local food in most regions or differed in their effects by region.
How Was the Study Conducted?
In this study, researchers examined the association of daily access of locally produced foods with region, fruit and vegetable acreage, county-level per capita income, school district expenditures per student, foodservice labor cost per student, locale type (city, suburb, town, and rural), school district size (enrollment), percentage of students certified to receive free and reduced-price meals, and State-level educational attainment. Also in this study, researchers considered the role of State-legislated policies supporting farm-to-school programs and the density of farmers’ markets as indicators of interest in local foods in the surrounding communities.
To identify characteristics of school districts that serve local food more frequently, an ordered probit model of the frequency of serving local foods was estimated using data from the 2013 Farm to School Census. Additional school district characteristics were merged from the National Center of Education Statistics’ Common Core of Data (Public School District Universe Survey for the 2011-12 school year and School District Fiscal Survey for the 2010-11 school year) and State and county attributes from the USDA’s Economic Research Service (ERS) Food Environment Atlas. The survey was distributed to all public school districts, including multisite charter districts, and included questions on farm-to-school activities in the 2011-12 school year as well as frequency and categories of local foods served in school meal programs.
Read the entire report. (.pdf)