Since the COVID-19 pandemic began, food insecurity has remained persistently elevated at record levels. Typically, measures of household food insecurity overstate the extent to which children experience food insecurity because parents try to shield their children from want. But today, household measures understate the extent to which food hardship has increased among its most vulnerable members: children.
Fellow - Economic Studies
Associate Director - The Hamilton Project
Former Research Analyst - The Hamilton Project
Historically, food insecurity rises in the summer, especially among households with school-age children (including those eligible for free- or reduced-price school meals). To combat this, the U.S. Department of Agriculture’s (USDA) summer meals programs were established to provide meals to students even when school is not in session. The program has been successful, but has uneven saturation: evidence suggests that participation rates in summer meal sites “is heavily dependent upon the location of the sites.”
Because school meal programs are the frontline in combating child hunger in the United States, many entities, including Congress, the USDA, states, and schools, took action in response to COVID-19-related school closures to reconstitute the food safety net for children. In the wake of COVID-19-related school closures in March, schools and districts pivoted to a summer strategy: serving prepared meals at a limited number of sites. Meanwhile, Congress authorized a new program, Pandemic Electronic Benefit Transfer (Pandemic EBT), which provided grocery vouchers to replace the value of lost school meals. Meal sites remained at the center of the food safety net for children early on as all 50 states stood up Pandemic EBT.
In this analysis, we provide data on the number of children who live in a Census Tract (a geographic unit containing roughly 4,000 people) where a meal site was in operation during the school shutdown period in the spring. We find that more students lived in a Census tract with a meal site in the spring of 2020 than during the summer of 2018.
However, while more families had proximal access to a prepared meal site during the spring school shutdown than during a typical summer, take-up rates remained low. Our analysis of the first three months of the Census Household Pulse Survey finds that between 13 and 17 percent of households with incomes below 185 percent of the federal poverty level (income below which students are eligible for free or reduced price lunches) reported that they received a free meal from school or from a program directed to children between April and July.
Access to School Meal Sites
In many ways, the operation of meal sites during the spring of 2020 mimicked a typical summer when school is out. Summer Food Service Program meal sites are eligible to operate in locations (Census block groups or school catchment areas) where at least half of students are income-eligible for free or reduced-price meals. Schools and community partners may additionally operate meal sites during the summer. Due to the COVID-19 pandemic, states have been allowed to open meal sites for “grab and go” meals, parents could to pick up meals for their children to reduce the public health risk while maintaining access to food, in addition to expanding area eligibility, and in some places, delivering meals. USDA has also extended a series of waivers that allow more families to more easily participate in prepared meal programs through the end of 2020.
Through a data interactive, the USDA allows people to search for meal sites in their area: “Find Meals for Kids.” We downloaded these data for the last time in July 2020 and had previously scraped these data for the spring of 2020 and summer of 2018, allowing us to geolocate meal sites for these two time periods. Additionally, we used National Center for Education Statistics data to geolocate every school in the country. In combination with American Community Survey data, which provides data on the number of children in a particular location, we present data on the share of children who live in a Census Tract with:
- spring 2020 access to meal sites;
- summer 2018 access to meal sites; and,
- hypothetical access to a meal site if one were in operation in every public school or preschool.
We find that schools and districts have increased access to meal sites during the pandemic relative to a typical summer: we find that more students have had access to a meal site during the pandemic than in summer 2018. In the spring of 2020, about 53 percent of children lived in a Census tract with a meal site, more than the 43 percent who lived in a tract with a 2018 summer meal site (figure 1). However, there was still room to expand—if meal sites had been in operation at every public school, more than three quarters of children would have local access.
Moreover, local access varies materially by urbanicity. In the summer of 2018, about 40 percent of children living in metropolitan Census Tracts and 45 percent of rural children living in rural Census Tracts had access to a summer meal site; 54 and 56 percent of children in micropolitan and small-town tracts, respectively, had access. While more meal sites were in operation during the spring of 2020 than in the summer of 2018 across the board, access expanded most among rural (62 percent more children in rural areas) and least among metropolitan (24 percent more children in metropolitan areas). The highest share of access in the spring of 2020 was in small towns (74 percent).
Local access to these meal sites varies widely across states. Figure 2 shows the share of children who live in a Census Tract where school meal site was in operation during the spring of 2020 by state. In general, access to meal sites is higher in the south and mountain west. Just 20 percent of children live in a tract with a site in Nevada, compared with 92 percent in North Dakota.
Take-up of School Meals
Research on summer meals programs has found that about one in seven eligible children participate in the program; in 2018, almost 50,000 summer meal sites served 2.7 million children. Evidence from surveys fielded since the pandemic suggest similarly low take-up rates of grab-and-go school meals during the pandemic as during a typical summer.
Early in the pandemic, a survey of hourly service workers found that only 11 percent had picked up a meal at a school. In our analysis of the Survey of Mothers with Young Children, 14.1 percent of mothers overall—and 18.8 percent of mothers whose children were eligible for free or reduced priced meals—reported leaving home to pick up a school meal in late April. The share of mothers who reported taking up a grab-and-go meal and whose children were eligible were statistically indistinguishable from the share who reported going to a food pantry (15.2 percent) but far fewer than the share who reported going to a grocery store (79.3 percent).
The Census Household Pulse Survey asked in surveys fielded between April 23 and July 21 about where households received free groceries or meals. In these data, between 8 and 10 percent of households with children reported getting free meals “through school or other programs aimed at children.” This is three to four times higher than the share reporting that they went to a food pantry or food bank in the past week over the same time period. Between 13 and 17 percent of households with incomes below 185 percent of the federal poverty level reported that they received a free meal from school or from a program directed to children between April and July.
Research shows that even in a typical year, families cannot absorb the loss of the value of school meals in the transition from the school year to the summer. The combination of pandemic-instigated school closures and a severe recession has been devastating for families with children: between 9 and 17 million children live in a household where the adults say that their children do not have enough to eat, and they do not have the resources to purchase more food. These levels of food hardship among children are orders of magnitude higher than during the Great Recession. Despite this level of need, prepared meal programs are reaching a fraction of the eligible population.
Since April, we’ve argued that while prepared meal programs have operated heroically under difficult circumstances, evidence suggests that providing grocery vouchers to eligible families would be more effective at reducing food insecurity. This evidence comes from summers past: an experimental pilot program for providing EBT during the summer has been proven to dramatically reduce food insecurity and very low food security among children. What’s more, we have direct evidence from this past summer: Pandemic EBT reduced food hardship experienced by children by thirty percent in the week of its disbursement.
The 2020-21 school year has started, but many families are facing continued uncertainty about whether and when schools will be open for in-person instruction and meal service. Pandemic EBT expires on September 30, 2020 and there are indications that it will not be a part of the stopgap funding measure. States invested heavily to stand up this program over the spring and summer. The program was effective and reached more families than the prepared meals program when schools were closed. Extending the Pandemic EBT program would provide much-needed certainty for school leaders and reduce food insecurity among children.