Frontier and Remote Area Codes - Users’ Guide

The methods used to create the USDA, Economic Research Service’s (ERS) Frontier and Remote area (FAR) codes data product remain the same for the vintage 2020 data as for earlier vintages of the data. However, there were a couple of changes that impact the data product: new input sources and a census tract version.

This Users’ Guide describes these changes to the FAR codes data product and their implications in two sections. The first section compares the newly released census tract version of the FAR codes to the ZIP Code version, using the vintage 2020 data. The second section compares the 2010 and 2020 vintages of the FAR code data.

The third and final section of this Users’ Guide discusses when the FAR codes may be preferred to other measures of rurality, with a focus on those produced by the USDA, ERS.

Comparison of ZIP Code and census tract data

Census tract versions of the 2010 and 2020 FAR codes were added with the release of vintage 2020 data. Census tracts are a statistical unit that the U.S. Department of Commerce, Bureau of the Census (Census Bureau) uses to present data. Census tracts are subdivisions of counties and are delineated to have an optimal population size of 4,000 people. This means that the land area of tracts can vary greatly across the United States, with larger census tracts in rural areas of the country and smaller census tracts in urban areas. Census tracts are the smallest statistical unit for which the Census Bureau regularly produces data on a variety of demographic and socioeconomic factors. This makes them a particularly important unit for research and policy.

Previously, the FAR codes were only released for ZIP Code areas. ZIP Codes are a useful statistical unit for both policymakers and researchers, as it is easy to locate addresses by ZIP Code. Similar to census tracts, the size of ZIP Code areas varies greatly throughout the United States, with larger footprints in rural areas and smaller footprints in urban areas. However, the areas were not created for statistical purposes. Instead, the U.S. Postal Service developed ZIP Codes to make mail delivery more efficient. As a result, ZIP Code areas change according to mail delivery needs and may even span county or state boundaries.

The census tract and ZIP Code versions of the FAR code data product were created using the same underlying grid cell data and method (see the Documentation page for more information). The difference is that the grid cell data were aggregated to different units. Each unit is classified as FAR if the majority of residents are considered FAR, so there may be differences between the census tract and ZIP Code versions of the data. The table below compares the distribution of census tracts, ZIP Codes, population, and area across 2020 FAR code levels.

For both census tracts and ZIP Code areas, the share of the population that lives in locations classified as FAR for each level is essentially the same. While slightly more area is classified FAR by ZIP Codes, the difference is less than 1 percentage point for each FAR level. This indicates that, overall, the census tract and ZIP Code versions of the 2020 FAR code data are very similar.

However, differences may exist at a smaller scale. For example, about 13 million people reside in FAR Level 1 census tracts, while about 12.8 million people reside in FAR Level 1 ZIP Code areas. Because the boundaries of census tracts and ZIP Code areas are different, not all the residences considered FAR Level 1 by the ZIP Code version of the data are necessarily considered FAR Level 1 by the census tract version. It is the same for residences considered FAR Level 1 in the census tract version—they may not be considered FAR Level 1 in the ZIP Code version.

The share of each unit (census tracts and ZIP Codes) considered frontier and remote at any level is different between the census tract and ZIP Code versions of the 2020 FAR code data. These differences highlight the differences between census tracts and ZIP Codes described above. Nearly 20 percent of ZIP Code areas are classified FAR Level 1 (the most expansive level), while only 5 percent of census tracts have that classification. And more than 10 percent of ZIP Code areas are classified as FAR Level 4 (the most restrictive level), while less than 2 percent of census tracts have that classification. However, despite their differences, the two geographical units ultimately lead to similar findings for the share of population and the total area in each FAR code.

Overall, both the census tract and ZIP Code versions of the FAR codes use the same methods and underlying grid cell data. Whether census tracts or ZIP Codes should be used depends on the policy or research objectives of the user. Census tracts only change once every decade and are a common geographic unit for demographic and economic research, while Zip Codes may change at any time but are well known to individuals and are easily linked to a mailing address.

Comparison of 2010 and 2020 FAR code data

There are slight differences between the vintage 2010 and vintage 2020 versions of the FAR codes data product. These differences are due to changes in the data inputs, as the methods and the geographic scope of coverage remained the same.

For the vintage 2010 FAR Codes, ERS projected census-block-level population from the 2010 Decennial Census to ½ x ½ kilometer grid cells using the assumption that people are equally distributed throughout the census block’s area. For the vintage 2020 FAR Codes, nighttime population data from LandScan USA were used. The LandScan USDA data are higher resolution (use smaller grid cells) and use the built environment to project census-block-level population to likely places of residence.

The input road network has also changed substantially. The road network data source is the same (NAVTEQ/HERE Technologies) for both the vintage 2010 and 2020 data. However, there have been changes to the footprint of the road network due to growth or loss of roads available for public use, as well as an overall increase in data quality due to technological advancements. Similarly, the Census Bureau’s urban areas were used for both the 2010 and 2020 vintages of the FAR code data. However, changes to their urban area delineation, particularly changes to the minimum urban area size, affect the calculation of the FAR codes.

Finally, between the 2010 and 2020 vintages of the FAR code data, the number of ZIP Code areas in the 50 States and Washington, DC increased from 30,337 to 32,158. In addition to the new ZIP Code areas, changes to the boundaries of some existing ZIP Codes have also occurred, due to changes in postal routes and post office locations.

The table below shows the distribution of ZIP Codes by FAR code levels for vintage 2010 and vintage 2020 data. At each level of remoteness, a slightly larger share of ZIP Code areas (1 to 2 percentage points more) is considered FAR in the vintage 2020 data than in the vintage 2010 data. However, due to the frequent changes in the number of ZIP Code areas and their boundaries, this does not necessarily indicate an increase in remoteness.

The population data tell a slightly different story. While the share of the population considered FAR Level 1 remained the same (3.9 percent) in the vintage 2020 data as it was in the vintage 2010 data, there were slight increases in the other FAR levels. The largest increase in the share of the population considered frontier and remote was in FAR Level 4, from 0.7 percent in the vintage 2010 data to 1.1 percent in the vintage 2020 data. This increase is due to changes in the definition of urban areas.

In 2010, the minimum population requirement for urban areas was 2,500 residents. But in 2020, the Census Bureau required urban areas to have at least 2,000 housing units or 5,000 residents. This increase in the minimum population requirement decreased the number of urban areas with a population less than 10,000 residents from 2,181 in 2010 to 1,114 in 2020. In other words, with fewer areas qualifying as a small urban area, distances to the next qualifying small urban area increased, and a larger share of the population was designated as FAR Level 4.

When to use the FAR codes to measure rurality

The USDA, ERS produces several measures of rurality to suit the different needs of users. The FAR codes are used to identify frontier areas in the United States—places that are extremely rural and remote. The four FAR code levels allow users to differentiate between different levels of remoteness (distance) from the goods and services available in urban areas of various sizes. But for each level, locations are divided into one of two categories: “FAR” or “not FAR.”

The Rural-Urban Commuting Area (RUCA) codes, instead, focus on classifying places into a continuum of urban and rural areas by population size and economic connection. This results in 10 categories indicating whether a location is part of an urban core or connected to it through commuting, as well as the size of the urban core and the strength of the commuting connection.

The biggest difference between the FAR and RUCA codes is how they define locations. The FAR codes focus on how remote a location is, grouping all other locations into a single “not remote” category that ignores the urban character of a location. Conversely, the RUCA codes focus on classifying locations by how urban they are, grouping all other locations into a single rural category. Because both of these approaches have their value, researchers have created a rural-urban categorization that combines them, the Integrated Metropolitan-to-Frontier Area Codes (IMFAC), as another way to define rurality (Gibson, et al. 2025).

Another consideration for determining whether the FAR codes are the appropriate rurality measure to use is geographic level. The FAR and RUCA codes are sub-county measures that provide a geographically detailed classification to distinguish the rurality of a location more accurately than can be achieved with a larger geographic area. They are available for census tracts and ZIP Codes. However, many users may not have such geographically detailed information. In these cases, USDA, ERS county-level measures of rurality are more appropriate. These measures focus on classifying locations based on how urban they are, as well as their economic connections to an urban core. See the Rural Classifications page for more information on the other rural-urban delineations available from the USDA, ERS.

Reference

Gibson, B. J., Dobis, E., Cromartie, J., Grieshober, L., Ulrich, C., Onega, T., & Doherty, J. (2025). A novel approach to incorporate frontier areas into urban–rural geographic classifications: Integrated Metropolitan‐to‐Frontier Area Codes. The Journal of Rural Health, 41(4), e70102.