Our main source of data for generating zip code files comes from the Bureau of Census County Business Patterns (CBP) program, as they release 6-digit NAICs employment data at the zip code level. Rather than employment counts, the CBP data provides information on the number of firms in each of 14 firm size classes (e.g., 1-4 employees, 4-9 employees, 10-19 employees, etc.) for each 6-digit NAICs sector. This allows the CBP to avoid disclosing the exact number of employees at a single firm.
To create our data, we take the mid-point of each of the size classes and multiply this by the number of firms in that size class (e.g. if there are 6 firms reporting under a single NAICS code for 4-9 employees, we get an estimate of ~39 employees for that NAICS code and class). By summing the estimated number of employees within a 6-digit NAICS specification, and aggregating the NAICS code employment values into their appropriate IMPLAN sector, we create an estimate of the employment for that sector at a zip code level. We also add together employment data for the all zip codes represented in a larger region (county or counties) to aid in distribution of non-employment IMPLAN data to individual zip code regions.
After aggregating the 6-digit NAICs data to the 536 IMPLAN Sectors, we can create ratios for distributing the county(ies) level IMPLAN data to the desired zip-code regions. The ratio is derived by dividing the specified zip code employment by the county's zip code employment for each IMPLAN Sector.
Employment ratios are used to distribute all industry data (Employment, Output, Employee Compensation, Proprietor Income, Other Property Type Income and Tax on Production and Imports) into their respective zip code regions.
Census of population data is also available at the zip code level. The ratio of a zip codes population compared to the county's population for the 2000 Census is applied to the county's current population to provide an estimate of the zip code area's current population.
Industry Sectors that CBP Does Not Cover
Proxies are used to distribute industry sectors that CBP does not cover to ZIP Code regions. These sectors, and the proxies used to separate the zip code employment from the county(ies) are shown below:
- Agricultural Sectors: These are estimated from current Census of Agriculture- county level farming is distributed to zip codes based on the Census number of farms by zip code.
- Railroad: These are estimated based on trucking and warehousing distribution employment as a proxy.
- Religious organizations: Population is used as the proxy for determining these employment values.
- Government, except education: Total employment for the zip code is used to distribute county level government employment.
- State & local education: Population is used.
- Construction: Employment for NAICS 22 is used for this proxy.
Final Demand Proxies to Distribute Zip Code Regions
Final demands also need to be estimated. Below are listed the proxy values used to distribute the county(ies) final demands to a zip code area.
- PCE (9 classes): From latest Census American Community Survey by zip code data. This data contains the number of households by income class. E.g. if the zip code area has 10% of highest income households it gets 10% of the highest PCE final demand.
- Fed Military & Non-military: overall Employment
- State & Local Education & non-education: overall Employment
- Federal Sales: overall Employment
- State & Local Gov Sales: overall Employment
- Investment: Employment by Industry
- Foreign Export: Employment by Industry
- Change in Inventory: Employment by Industry
Because of the limitations of the data, productivity data for the zip code area is the same as for the county(ies) containing the zip code area. The following examples are included as part of the productivity data:
- Output per Worker
- Earnings per Worker
- Value Added to Output ratio