Skip to main content
Skip to main content

Official websites use .gov
A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS
A lock ( ) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.

Food Access Research Atlas - Documentation

Scope and Coverage of the Data

The data presented in the Food Access Research Atlas (FARA) cover the 50 U.S. States and the District of Columbia. U.S. territories are not included.

Estimates in the Supplemental Nutrition Assistance Program (SNAP)‑authorized Retailer Access Map (SRAM) for 2025 are based on information obtained from the Store Tracking and Redemption System (STARS), which has stores authorized to accept SNAP (Supplemental Nutrition Assistance Program) benefits; the 2020 Decennial Census; the LandScan USA 2020 nighttime dataset; and the 2020–24 American Community Survey (ACS). SNAP-authorized farmers markets and delivery route store types are excluded.

Estimates in the Large Retailer Access Map (LRAM) for 2019 are based on a 2019 list of supermarkets, the 2010 Decennial Census, and the 2014–18 ACS. The LRAM estimates for 2015 are based on a 2015 list of supermarkets, the 2010 Decennial Census, and the 2010–14 ACS.

FARA Data and Methodology

A complete list of indicators and further details on their definitions and data sources in the Food Access Research Atlas are available in the 2025 SRAM Reference Guide and the 2019 LRAM Reference Guide. The methods used to create certain variables depend on their underlying data sources and are outlined in the reference guides. A detailed technical discussion of the methods used to create and develop the maps and estimates in the Atlas is provided in both the SRAM Technical Methods webpage and the LRAM Technical Methods webpage.

Shared Definitions Used in Both SRAM and LRAM

Low-Income Neighborhoods

The criteria for identifying a census tract as low income in the 2025 dataset are from the U.S. Department of the Treasury’s New Markets Tax Credit (NMTC) program (Community Development Financial Institutions (CDFI) Fund, 2000)

This program defines a low-income census tract as any tract where:

  • The tract’s poverty rate is 20 percent or greater; or
  • The tract’s median family income is less than or equal to 80 percent of the State-wide median family income; or
  • The tract is in a metropolitan area and has a median family income less than or equal to 80 percent of the metropolitan area's median family income.

Urban-Rural Classification

Census tracts are classified as urban or rural based on the location of their population-weighted center points relative to 2020 Census Bureau Urban Area boundaries. A census tract is classified as urban if its center point falls within an urban area boundary; tracts located outside these boundaries are considered rural. Center points were obtained from the Census Bureau’s tract-level Centers of Population dataset.

These shared definitions apply across both SRAM and LRAM to ensure consistency in classification and interpretation.

SRAM: Data and Methodology Details

The SRAM integrates SNAP-authorized food retailer locations, gridded high-resolution population data (i.e., LandScan), and census tract-level socioeconomic characteristics to identify populations with limited geographic proximity to SNAP-authorized food retailers using both Euclidean-based (straight-line) and network-based (driving paths along roads) distance measures.

Data Sources and Geographic Area

SNAP-authorized Food Retailer Locations

The Food and Nutrition Administration (FNA; formerly Food and Nutrition Service) provided the complete list of food retailers authorized to accept SNAP benefits as of June 2025 from the Store Tracking and Redemption System (STARS). ERS excluded farmers markets and delivery routes because they do not necessarily operate throughout the year or with regular hours (farmers markets) or have a fixed location (delivery routes).

The STARS data include retailer addresses and geographic coordinates. For the SRAM, retailer addresses were geocoded using Environmental Systems Research Institute, Inc. (Esri)’s ArcGIS StreetMap Premium. When StreetMap Premium did not return a matching store location, the original STARS coordinates were retained.

Store coordinates were also validated by comparing the State and county listed in the STARS data to the Census boundaries (i.e., mathematics applied geographies). Any geocoded point outside the expected state boundary was flagged for review. This process helped:

  • Identify geocoding errors or mis‑assigned coordinates,
  • Catch incorrect or incomplete address inputs,
  • Improve the reliability of subsequent spatial analysis and reporting, and
  • Maintain consistency across datasets that rely on accurate location information.

In addition to automated checks, manual review methods were also used to verify food retailer locations flagged during the validation process.

Edits and improvements to the original food store locations were made based on geocoding processes described above. Location issues in the original dataset ranged from legitimate geocoded results differing from the latitude/longitude within the provided data to potential human error during initial data entry.

Population Distribution Data

Population is not evenly spread within Census tracts, and some tracts cover large areas. To better reflect how people are actually distributed within each tract, Oak Ridge National Laboratory’s LandScan USA 2020 nighttime gridded population data. This is produced at about 90-meter resolution. LandScan models ambient population density (the average number of people in a given area at any time) using a variety of data sets, such as Census counts, land cover classifications, and infrastructure information to estimate where people live. Using gridded population data allows the SRAM to more accurately represent where people are located within a tract, which can then be combined with distance calculations to figure out how many people live within certain distances of SNAP‑authorized food retailers.

Demographic and Socioeconomic Data

Tract-level demographic and socioeconomic variables were acquired from two Census Bureau sources:

  • 2020 Decennial Census: Total population and housing unit counts
  • 2020–24 American Community Survey (ACS) 5-Year Estimates: Detailed demographic, socioeconomic, SNAP participation, and vehicle availability characteristics

Tract-level geography was selected to mitigate the effects of Differential Privacy in the 2020 Decennial Census and to reduce margins of error present in ACS estimates at smaller geographies. The Census Bureau's Differential Privacy approach adds statistical noise to protect individual data, and this noise is larger for smaller geographic areas because the risk of identifying someone is higher. Similarly, ACS estimates for smaller geographies carry larger margins of error because fewer people are included in the sample. Tract-level aggregation provides a good balance by offering data that are statistically reliable while still maintaining useful geographic detail.

Geographic Boundary Data

Geographic boundary files were obtained from the Census Bureau's Topologically Integrated Geographic Encoding and Referencing (TIGER)/Line and cartographic boundary products, including 2020 and 2024 Census tracts, 2020 Urban Areas, 2020 Core-based Statistical Areas, and U.S. States.

Calculation of Distance to Stores for Each Tract

To determine whether a tract was within a specified distance to a SNAP-authorized food store, zones with low access to foodstores were generated using two distance‑based approaches: (1) Euclidean (straight‑line) distance and (2) network‑based (driving) distance. Both methods produced foodstore access polygons (areas) at four distance thresholds: 0.5, 1, 10, and 20 miles, consistent with prior versions of the ERS Food Access Research Atlas (Rhone et al., 2017; 2022). Overlapping polygons within each threshold were merged to create unified Euclidean and network‑based access areas.

Network distances represent the driving path from where people live to the nearest SNAP‑authorized food retailer.

Euclidean Distance Methods

Circular buffer polygons were created around each SNAP‑authorized retailer at the four distance thresholds. Straight-line distances were generated using geodesic calculations, which take the Earth’s curvature into account and provide more accurate results than calculations that assume the surface is flat, especially across large areas.

Network-Based Driving Distance Methods

Driving distances were calculated using Esri ArcGIS Pro Network Analyst with StreetMap Premium Custom Roads 2026 (Release 1) as the network dataset. All network-based distances measure travel distance to the nearest SNAP‑authorized retailer along roadways. A series of three Network Analyst tools were used to calculate areas reachable within each distance threshold to each retailer location.

Since the road network does not include home driveways or store parking lots, buffers were applied to the roadways to capture populations reasonably close to the routed service areas. Specifically, a 50 meters buffer was added for the 0.5- and 1-mile thresholds and a 250 meters buffer for 10- and 20-mile thresholds. Larger buffers were applied to the longer (rural) thresholds to accommodate lower road density and larger parcel sizes typical of rural areas.

The SNAP-authorized foodstores file included 101 stores (97 in Alaska and 4 in the continental United States) that could not be incorporated into the drive‑distance calculations. The primary reasons were that the associated road segments were not accessible year‑round, vehicle traffic was not permitted, or the stores were located more than 1 kilometer from the road network. As a result, they are excluded from drive‑distance calculations and do not contribute to low‑access determinations.

Population Mapping

To locate where people live within each census tract at a more detailed scale, tract‑level population counts were distributed onto a high‑resolution grid using a population mapping approach based on LandScan USA 2020 population patterns. Population mapping uses extra geographic information, such as land cover, to place people more accurately on the landscape. This process creates a detailed grid of population estimates that can be overlaid with low food access areas to determine how many people fall within each distance threshold.

The allocation process consisted of three steps:

  1. Tract-level LandScan totals: LandScan population values were summed within each census tract to produce tract-level population totals based on the gridded data.
  2. Cell-level proportion proportions: For each small grid cell, the share of the tract’s LandScan population that fell in that cell was calculated.
  3. Grid output: Census population and housing variables were then distributed to the grids cells using these proportions to produce gridded census variable datasets.

Tract-level grid data were aligned to LandScan grids so they matched in coverage, cell size, and grid structure.

This approach relies on two key assumptions:

  1. Population distribution accuracy: LandScan accurately represents how people are distributed within each tract.
  2. Demographic uniformity: the spatial distribution of demographic subgroups within each tract mirrors the same general pattern as the overall population.

Low Food Access Area Identification

After the food access zones were created (using both straight line and road network driving distances), these zones were combined with the gridded population data to identify where people do and do not have nearby geographic access to SNAP authorized food retailers.

Each LandScan grid cell contains an estimated number of people or homes and has a known distance to the nearest retailer. Grid cells were compared with the food access zones:

  • Cells inside an access zone (within 0.5, 1, 10, or 20 miles of a store) were marked as “having access.”
  • Cells outside all access zones were marked as “low access.”

This process was repeated for each of the three regions (the continental United States (CONUS), Alaska, Hawaii), each distance method (straight-line and driving), and each distance threshold. A fourth region, Alaska-East region, was not processed because it had no SNAP-authorized retailers, meaning the entire region is classified as low‑access.

Population Summary and Tract-Level Aggregation

Finally, low-access population and housing counts from the grid cells were added up for each census tract. A tract’s low‑access values reflect the number of people living in grid cells that fall outside the access zones.

When LandScan showed no population in a tract but Census data showed at least one person, for example, on military bases or university campuses, low access was calculated based on the share of the tract’s land area that was classified as low access. This process was repeated for each combination of census variable, distance method (Euclidean or network), distance threshold, and region.

Tract-level outputs from this step were then used in later processing to assign tract-level access classifications.

Data Sources

Information on the location of supermarkets, supercenters, and large grocery stores was obtained from two directories—Store Tracking and Redemption System (STARS), which has stores authorized to accept SNAP benefits, and stores in TDLinx, a commercial retail store directory maintained by NielsenIQ. For this analysis, the directory only included authorized stores in the system as of June 15, 2019. Data on population and characteristics were obtained from the Census Bureau’s 2010 Decennial Census. Information on income and household vehicle availability was obtained from the 2014–18 ACS.

Food Access Methods

Note: All distance measures in the LRAM are based on Euclidian, or straight line, access.

Spatial analysis, string matching, and manual review methods were used to merge the STARS and TDLinx datasets to construct a combined-store directory. This combined directory encompasses all the supercenters, supermarkets, and large grocery stores from each dataset, with duplicates eliminated to avoid double counting. This matching process identified STARS and TDLinx stores within a 1/3-mile radius of one another or within the same ZIP Code. An automated string-matching algorithm was used to identify exact or similar store name-address matches, which were subsequently verified manually. Foodstores from either the STARS or TDLinx systems—without a match in the other system—were included in the final combined directory, totaling 45,233 foodstores in the 2019 merged directory. Most foodstores (36,425) were in both data sources. Of the remaining stores, 5,170 were exclusive to TDLinx and 3,638 were found only on the STARS list. Military commissaries and warehouse club stores, such as Sam’s Club, Costco, and BJ’s were excluded. Although such stores offer a wide variety of foods and accept SNAP benefits, military commissaries are only accessible to a select group of individuals and club stores are only available to those who pay an annual membership fee, which may be a barrier for people with income constraints. Drug stores, dollar stores, and convenience stores were also excluded. Even though some of these store types sell a variety of fruits and vegetables, their offerings vary widely. Excluding these types of food retailers from our store directory is likely to result in an overestimate of the number of people lacking access to nutritious food.

Population data were from the 2010 Census because ACS data for these characteristics—though available at the census tract level—are less precise. Population counts, occupied housing unit counts, and other population characteristics (i.e., age, race, and ethnicity) from the 2010 Census were allocated based on an area to 0.5-kilometer-square grids (Rhone et al., 2022; Rhone et al., 2017; Ver Ploeg et al., 2012). For income and vehicle access, ACS tract level 2014–18 share estimates of housing units without vehicles and the share of individuals below 200 percent of poverty were multiplied by the 2010 housing-unit counts and population counts, respectively, to estimate the number of households without vehicles and the number of people with income at or below 200 percent of the poverty level. These numbers and shares were then similarly aggregated down to the 0.5-kilometer-square grid level to provide a fuller picture of population distribution within a census tract so access and populations could be described. From here, the methods to estimate distance to the nearest foodstores for the overall population and for subgroups have been the same as in previous reports.

To estimate whether a tract is low income, 2014–18 ACS tract data were used to directly measure whether the tract: (1) has a poverty rate of 20 percent or greater; (2) is at or below 80 percent of the greater Metropolitan Statistical Area (MSA) median family income or the State’s median family income; or (3) has a median family income at or below 80 percent of the State’s median family income if outside of a MSA (Community Development Financial Institutions Fund, 2000).

To estimate if a tract is low access, the number and share of people more than 0.5 or 1 mile (urban areas) from a foodstore or 10 or 20 miles (rural) was estimated based on the location of foodstores relative to the grids. Urban census tracts are areas with more than 2,500 people, and rural areas are sparsely populated areas with fewer than 2,500 people. These estimates were aggregated at the tract level for all grids within a tract. The same criteria for demarcating low access used in the previous FARA were applied for each of the four measures of low income and low access (LILA).

Strengths and Limitations

Both SRAM and LRAM are components of FARA, and some strengths and limitations apply to both mapping applications while others are specific to each. FARA has the following strengths and limitations in providing details on the incidence and prevalence of limited food access:

Strengths

  • FARA provides estimates of low-income and low-access tracts using 0.5- and 1-mile demarcation options, which can be used for urban areas, and 10- and 20-mile demarcation options, which can be used for rural areas. These added measures give users additional ways to consider food-access limitations for census tracts.
  • FARA also provides access measures for populations of interest, such as low-income families, households with vehicle access, group quarters (e.g., dormitories and institutions), SNAP-receiving housing units, children, the elderly, race, ethnicity, and veterans. Group quarters populations are included because they are available in Census data used for population of interest measures; however, group quarters may not always be fully represented in the gridded population used in earlier processing steps.
  • SRAM focuses on SNAP-authorized food retailers, which includes all store types (e.g., smaller grocery stores, specialty food stores, convenience stores, neighborhood markets, dollar stores), except farmers markets and delivery routes, from the Store Tracking and Redemption System (STARS), provided by FNA.
  • FARA’s LRAM measurements use a comprehensive list of supermarkets in the United States developed from two separate national-level directories of foodstores, combining the list of SNAP-authorized retailers from STARS with additional supermarket data from Trade Dimensions TDLinx (a NielsenIQ company), a proprietary source of individual supermarket store listings for the analysis year.

Limitations

  • FARA does not include restaurants, fast food outlets, and food service establishments because their prepared foods have higher costs per unit relative to food retailers. The costs of the food in the prepared food items at these outlets is a small share (about one-third) of the total cost to consumers, reflecting labor and other factors.
  • FARA’s area-based access measures that consider the share of low-income residents do not measure access for all the residents in the given geography.
  • FARA exclusively considers access to stores from consumers’ homes and does not consider the potential access to food that consumers have when they travel to school, work, or other activities. Because this measure is based on residential locations, individuals without a fixed address, such as unhoused populations, are not directly represented in the underlying data and may therefore be undercounted.
  • FARA measures only one dimension of food access, geographic proximity, so these indicators do not capture other important factors such as food affordability, household resource constraints, store quality, and product availability.
  • Because FARA’s measures rely solely on geographic distance, they do not account for physical barriers or travel constraints such as steep terrain, unsafe routes, lack of sidewalks, or other environmental factors. As a result, areas that appear to be close to food retailers by distance may still face significant access challenges.
  • The precision of tract access designations based on driving distance is contingent upon the quality and thoroughness of the road-network dataset incorporated into the model. Errors in the foundational road data such as inaccurate geometries, omitted roads, or misassigned classifications can lead to erroneous classification of census tracts.
  • There may be inconsistencies resulting from temporal misalignment, where dataset years may not match. Integration of data from multiple periods (e.g., 2020-24 American Community Survey, 2025 STARS data) may not accurately reflect current conditions in areas experiencing rapid population or retail market changes.
  • SNAP-authorized retailer locations in STARS are primarily provided by the retailer at a specific point in time. Store addresses and geolocations may occasionally be inaccurate and may not accurately represent current SNAP-authorized retailer availability. In an effort to quantify the potential for inaccuracy in the STARS data, a review of 480 randomly selected retailers was completed, with reviewers randomly assigned to 195 of the selected retailers to ensure interrater reliability of the sample. The final edited STARS dataset still contains an estimated error rate of approximately 1 percent (down from 5 percent) where store locations may fall in the wrong county.
  • The methodology assumes LandScan accurately represents within-tract population distribution patterns and assumes the spatial distribution of demographic subgroups within each tract mirrors the overall population distribution represented in LandScan. This assumption may not hold in tracts with spatially clustered demographic patterns.
  • FARA employs a fixed buffer around each store location, which does not necessarily accurately account for road network configuration, terrain barriers, or other obstacles affecting actual travel routes for each store.
  • The network-based driving distance measures focus on automobile access and do not capture public transit availability, pedestrian infrastructure quality, or individual mobility limitations.
  • There is no partial allocation of population within a LandScan grid cell; each cell is classified entirely inside or outside an access zone based on the location of its center.
  • FARA exclusively considers an area-based concept of food access, where the data are reported at the census-tract level.
  • The low-income, low-access measures do not capture all low-income people or households with limited access because not all low-income people live in areas with concentrated poverty but may still lack the resources or transportation to access affordable food.
  • FARA doesn’t account for online purchasing, food pantries, and food banks as a way to increase access to food.
  • SRAM, which focuses exclusively on brick-and-mortar SNAP-authorized retailers, does not fully measure all food access because not all foodstores in the United States are SNAP-authorized retailers.
  • LRAM, which focuses on supermarkets, supercenters, and large grocery stores as consistent sources of the full range of foods that comprise a healthy diet (Franco et al., 2008; Neckerman et al., 2009; Rose et al. 2009; Sharkey & Horel, 2009). Since excluded stores (e.g., smaller grocery stores, neighborhood markets, or dollar stores) may carry the full range of food products necessary for a healthy diet, the LRAM may not accurately identify areas with low access to healthy food.

Resources

Community Development Financial Institutions Fund. (2000). New Markets Tax Credit Program: Low‑income community eligibility criteria. U.S. Department of the Treasury.

Dutko, P., Ver Ploeg, M., & Farrigan, T. (2012). Characteristics and influential factors of food deserts (Report No. ERR-140). U.S. Department of Agriculture, Economic Research Service.

Franco, M., Diez Roux, A. V., Glass, T., Caballero, B., & Brancati, F. (2008). Neighborhood characteristics and availability of healthy foods in Baltimore. American Journal of Preventive Medicine, 35(6), 561–567.

Neckerman, K. M., Bader, M. D., Purciel, M., & Yousefzadeh, P. (2009). Measuring food access in urban areas. Built Environment, 35(3), 293–301.

Rhone, A., Ver Ploeg, M., Dicken, C., Williams, R., & Breneman, V. (2017). Low-income and low-supermarket-access Census tracts, 2010-2015 (Report No. EIB-165). U.S. Department of Agriculture, Economic Research Service.

Rhone, A., Ver Ploeg, M., Williams, R., & Breneman, V. (2019). Understanding low-income and low-access Census tracts across the nation: Subnational and subpopulation estimates of access to healthy food (Report No. EIB-209). U.S. Department of Agriculture, Economic Research Service.

Rhone, A., Williams, R., & Dicken, C. (2022). Low-income and low-foodstore-access Census tracts, 2015–19 (Report No. EIB‑236). U.S. Department of Agriculture, Economic Research Service.

Rose, D., Bodor, J. N., Hutchinson, P. L., & Swalm, C. M. (2009). Neighborhood food environments and Body Mass Index: The importance of in‑store contents. American Journal of Preventive Medicine, 37(3), 214–219.

Sharkey, J. R., & Horel, S. (2009). Association between neighborhood need and spatial access to food stores and fast food restaurants in colonias. International Journal of Health Geographics, 8(9), 1–17.

U.S. Department of Agriculture, Economic Research Service. (2022). 2019 Food Access Research Atlas: Interactive guide.

Ver Ploeg, M., Breneman, V., Dutko, P., Williams, R., Snyder, S., Dicken, C. & Kaufman, P. (2012). Access to affordable and nutritious food: Updated estimates of distance to supermarkets using 2010 data (Report No. ERR-143). U.S. Department of Agriculture, Economic Research Service.

Ver Ploeg, M., Breneman, V., Farrigan, T., Hamrick, K., Hopkins, D., Kaufman, P., Lin, B.-H., Nord, M., Smith, T.A., Williams, R., Kinnison, K., Olander, C., Singh, A., & Tuckermanty, E. (2009). Access to affordable and nutritious food: Measuring and understanding food deserts and their consequences: report to congress (Report No. AP-36), U.S. Department of Agriculture, Economic Research Service.

The Food Environment Atlas provides users with a wide set of statistics at the county level on food choices, health and well-being, and community characteristics for all communities in the United States.

Recommended Citation(s)

U.S. Department of Agriculture, Economic Research Service. (2026). Food Access Research Atlas (FARA).

U.S. Department of Agriculture, Economic Research Service. (2026). Food Access Research Atlas. SNAP‑authorized Retailer Access Map (SRAM).

U.S. Department of Agriculture, Economic Research Service. (2021). Food Access Research Atlas. Large Retailer Access Map (LRAM).