This page provides access to delineation files for a range of labor-shed delineations as described in Fowler Jensen and Rhubart (2016), Fowler and Jensen (2020), and for 2020 delineations Fowler (2024). Each delineation includes county-level and labor-shed level fit statistics as described in the 2020 and 2024 papers and the delineations are available in several formats to facilitate use by researchers in both standard statistical software and Geographic Information Systems. Descriptions of the columns included in each file are provided in the table at the bottom of this page and further described in the above mentioned paper.
BEA Delineations
The Bureau of Economic Analysis has delineated ‘Economic Areas’ since the 1960’s. Their delineations rely on Census commuting flows with additional information drawn from newspaper circulation information. More information on the most recent (2004) delineation is available on their web site . The delineations presented here are for 1977, 1995, and 2004 (the same 2004 delineations are associated with 2000 and 2010 commuting flows in the files below). In the analysis by Fowler and Jensen these delineations consistently perform well on measures of containment, have complete U.S. coverage, and generally contain an economic core. However, the large size and comparatively small number of Economic Areas delineated may prove to be a limitation for some analyses.
ERS Delineations
ERS Delineations produced in house are available for 1980 through 2000 (see ERS web site here). A replication of the ERS methodology using 2010 data (REP10) is available with the files below. However, Fowler et al. (2016) identified some discrepancies between the method described by ERS and the results reported for previous years, so a second delineation for 2010 (OUT10) is also included. This second delineation reproduces the intention of prior (1980,1990, 2000) delineations and is preferred to REP10 for longitudinal work using the other ERS-originated delineations. REP10 is included here primarily for completeness with respect to results reported in Fowler et al. (2016).
Commuting Zones for 2020 are based on the work described in Fowler (2024). And complement the earlier work with a shared methodology and continued reporting of fit statistics.
FCC Delineation
The Federal Communications Commission provides ‘Partial Economic Areas’ for 2010. More information available on the FCC web site here. They are divisions of the BEA Economic Areas designed to facilitate auctioning communications licenses. Fit statistics indicate that these county groupings do not do well at matching the concept of a labor market (nor was that their intended purpose). They should not be used to represent labor markets, but are presented here for completeness.
Census CBSA Delineation
Census Core Based Statistical Areas or CBSA’s are one of the principle ways researchers use to divide up the country into different metropolitan regions. The 2010 delineations are available from the Census but are provided here with associated fit statistics so the delineations can be compared with alternatives, particularly that of Tong and Plane (see below).
Tong and Plane Delineation
In a 2014 paper Daoqin Tong and David Plane proposed an alternative to the Census CBSA delineation that was designed to explicitly consider the polycentric configuration of U.S. metropolitan areas. In their paper they demonstrate the particular strengths of this delineation. It is retained here for completeness and so that it can be considered with a consistent set of fit statistics. The delineation files are also available from the authors on request.
File Format Description
This page contains links to the data files in .csv, .xlsx, and .shp formats.
The .csv and .xlsx file formats are presented at the county level with each record containing a county. They currently contain just the raw data used in the analysis (population, wages, and delineations to which a county is assigned), not any of the county-level fit statistics. This may be remedied in the future.
The .shp files are presented in both county-level form and labor-shed level form. The .shp files are delivered in zip archives and include the .shp, .shx, .dbf, and .prj files necessary for these to be used in most common Geographic Information Systems (GIS). Data is delivered in Albers Equal Area Projection.
For the most part the data included in each format is the same, but the following table clarifies the column names and meanings as well as indicating slight variations in what columns are present. For further detail on the method used to generate the fit statistics and the rationale for doing so see Fowler and Jensen (preprint | final print version). Population values are drawn from the decade appropriate decennial Census.
County File Column Names and Descriptions
| Column Name | Description |
| FIPS | County FIPS code, a unique identifier |
| [Prefix] | The five character identifier specific to a delineation and a given year (e.g. BEA80). The numeric code indicates the labor market number or the labor market to which a county is assigned. |
| Pop[yr] | The [yr] specific population of a county |
| Wage[yr] | Year specific average wage for county (from BLS data) |
County Data
County Shapefiles:
County Shapes 1980 .zip | County Shapes 1990 .zip | County Shapes 2000 .zip | County Shapes 2010 .zip | CommutingZones2020_County_GIS_files
County xlsx files:
Counties 1980.xlsx |Counties 1990 .xlsx | Counties 2000 .xlsx | Counties 2010 .xlsx
County .csv files:
Counties 1980 .csv | Counties 1990 .csv | Counties 2000 .csv | Counties 2010 .csv | Counties 2020.csv
Labor Market Files Columns and Descriptions
The rationale and detailed description for these columns are provided in the accompanying paper unfortunately many of these have been updated as the manuscript went through numerous revisions in the process of getting approval. I still haven’t found time to update all the data below to match the new values so please reach out to me if you can’t find what you need.
| Column Name | Description |
| Source | The five character identifier specific to a delineation and a given year (e.g. BEA80). The numeric code indicates the labor market number or the labor market to which a county is assigned. |
| Year | The year for the delineation |
| LM_Code | The code identifying a specific labor market. May be real (as with CBSA delineation) or arbitrary (e.g. ERS 2010 codes are simply sequenced from 1…n) |
| SplitMetro | Does the labor market split a metropolitan area (as defined by OMB for that Year) |
| Has_Core | Does the labor market contain a ‘Core’ county as defined by OMB for that Year |
| Min_Core | For counties within the labor market, what is the minimum share of residents commuting to a core county in the labor market |
| Mean_Core | Averaged across counties within the labor market, what is the share of residents commuting to a core county in the labor market |
| MinFrmCor | For counties within the labor market, what is the minimum share of a county’s workforce that commutes from a core county in the labor market |
| MeanFrmCor | Averaged across counties within the labor market, what is the share of the workforce commuting from a core county in the labor market |
| MinWgCorr | For counties within the labor market, what is the minimum average pairwise correlation in wages over a six year period. |
| MeanWgCorr | Averaged across counties within the labor market, what is the average pairwise correlation in wages over a six year period. |
| MinCtn | For counties within the labor market, what is the minimum share of residents living and working within the labor market. |
| MeanCtn | Averaged across counties within the labor market, what is the share of county residents who also work within the labor market. |
| MinCorCon | The minimum value for counties within the labor market where share of residents working in a core county and share of workforce commuting from a core county are added together. |
| MeanCorCon | Averaged across counties within the labor market, the sum of residents working in a core county and workforce residing in a core county. |
Labor Market Files shapefiles:
BEA 1980 .zip | BEA 1990 .zip | BEA 2000 .zip | BEA 2010 .zip
ERS 1980 .zip | ERS 1990 .zip | ERS 2000 .zip | ERS 2010 .zip | CommutingZones2020_GIS_files
Flows:
By popular demand I am making available the data files documenting county to county flows used in these papers. Most of these are available elsewhere (Census), but the 1980 flows were only found on an old CD held by a retiring employee at ERS, so these need to get out in the wild. The file I am making available has been substantially modified from the original format because that format had some outdated ID codes (I think it used an old BEA format that arbitrarily combined some counties). I can share the details of the changes with anyone who is interested. In the interest of making these available now I am providing them in the Rdata format I used in their creation. They can be opened in R with the load(‘filepathname’) command and written to csv format with write.csv(flows80,file=”yourfilename.csv”)