Arlington County Food Security

This repository contains the quantitative part of the Arlington Food Security grant. We will be working to produce analysis to examine the distribution of food insecurity and other risk factors at the sub-county level. To address the Task Force's interest in understanding families' unmet needs and financial pressures, we will leverage administrative and public data to understand the food environment and understand the financial and policy landscape. These will help the Task Force identify geographic areas for future programmatic targeting. The quantitative analysis will also inform the study’s qualitative sampling approach to ensure we hear from residents living in low-access neighborhoods, even if those neighborhoods are not in the same ZIP codes.

Specifically, the quantitative study will use spatial mapping techniques to assess:

1. The demographic and geographic populations where food insecurity exists and has unmet needs, with consideration to race, ethnicity, age, household size, employment status, and income;
2. Where the Task Force could focus efforts to improve equity;
3. The degree with which households experience marginal food security and food insecurity; and
4. How well existing assistance programs (such as food pantries and benefits programs) are meeting households’ food needs, including spatial/geographic access.

Using predicted food insecurity and location data on food resources (e.g., grocery stores, food banks, etc.), we will identify areas where low access and high need intersect. This analysis aims to identify how well existing assistance programs meet households’ food needs in terms of spatial/geographic access and areas with unmet needs. Specifically, we will examine:

1. High-need neighborhoods with low access to retail food: These are census tracts with high rates of predicted food insecurity and low access to retail food locations. To develop this measure, we will map addresses for all retail food locations, including major supermarkets, local stores, and farmer’s markets using data from Google, the Task Force, and other lists.
2. High-need neighborhoods with low access to SNAP retailers: These are census tracts with high rates of predicted food insecurity and low access to retailers that accept SNAP. Using the list of retail food locations assembled above, we will assess geographic accessibility with the subset who accept SNAP. If applicable, we will also map farmer’s markets that accept SNAP if they have regular operating hours.
3. High-need neighborhoods with low access to SNAP: These are census tracts/ZIP codes with high rates of predicted food insecurity and low utilization of SNAP. This will only be possible if we can collect data on SNAP utilization at the sub-county level in collaboration with the Task Force. We believe this might be feasible, since Virginia’s SNAP program is administered at the county level.
4. High-need neighborhoods with low access to charitable food: These are census tracts with high rates of predicted food insecurity and low access to charitable food locations. To map charitable food locations, we will geocode address data of charitable food agencies, including food pantries, mobile distributions, meal programs, child nutrition programs, and mutual aid. We would also ask the Task Force and its partners to help in compiling this list. We will also include available information on days/times of operation to inform the extent to which community members are able to access charitable food throughout the month.

To identify the high-need and low-access neighborhoods described above, we will use spatial analysis. We will use the public and address data sources detailed above to classify census tracts as high need and low access if they have a relatively high predicted food insecurity rate, the centroid of the census tract is more than 0.5 miles from a food resource, and a substantial share of the population lacks access to a car. We will also identify neighborhoods in the county where residents of color struggle to access food resources using a similar method that examines food access in neighborhoods with a large share of residents of color.

This analysis will generate maps and appendix tables that we will compile into a PowerPoint presentation and final report with insights from both the quantitative and qualitative research. Specifically, we intend to include maps and tables of:

1. High-need neighborhoods with low access to retail food
2. High-need neighborhoods with low access to SNAP retailers
3. High-need neighborhoods with low access to SNAP (pending data availability, as described above)
4. High-need neighborhoods with low access to charitable food
5. Communities of color with low access to retail food
6. Communities of color with low access to SNAP retailers
7. Communities of color with low access to SNAP (pending data availability, as described above)
8. Communities of color with low access to charitable food

In addition to the spatial analysis that will identify high-need, low-access areas and areas where residents of color struggle to access food resources, we will present a descriptive analysis that will characterize:

1. Socioeconomic characteristics of residents, using administrative secondary data, we expect to characterize household income, wages, unemployment rates, SNAP recipiency rates, language proficiency, nativity, internet access, child care costs, and poverty levels at the ZIP code or census tract level, where data available.
2. Household financial pressures, using administrative public data we expect to characterize the financial reality of households using data on rent and mortgage expenses and annual utility costs.
3. Add-on transportation and retail special analysis, leveraging the Urban Institute’s existing work on transportation equity, mapping capacities, secondary administrative data, and address data on food retailers, we will evaluate gaps in public and private transportation, food retail, travel time to food resources, and other relevant neighborhood characteristics which may impact access to food resources.