This dataset contains annual measures of training and vocational schools in the United States from 1990 through 2022. The data include counts, per capita densities, area densities, and employment figures for twelve categories of training and vocational establishments: business and secretarial schools, data processing schools, general educational services, beauty schools and barber colleges, vocational schools, dance schools, instruction schools and camps, arts and crafts schools, music and drama schools, vehicle driving schools, reading and speaking schools, and personal development schools.
The unit of analysis is either the Census Tract or ZIP Code Tabulation Area (ZCTA), with separate files standardized to 2010 Census Tract boundaries, 2020 Census Tract boundaries, 2010 ZCTA boundaries, and 2020 ZCTA boundaries. Each file covers all census tracts or ZCTAs in the fifty United States, including Alaska, Hawaii, and US island territories.
Business establishment data were drawn from the National Establishment Time Series (NETS) database, which was cleaned and geocoded to Census Bureau TIGER/Line shapefiles. Population denominators came from the American Community Survey and Decennial Census. The data cleaning protocol addressed known NETS limitations by re-geocoding addresses, standardizing SIC codes across time, collapsing duplicate records, and removing businesses located at residential addresses using Zillow's ZTRAX data.
Key variables include count (e.g., count_businesscho), per capita density per 1000 population (e.g., den_datascho), area density per square mile (e.g., aden_vocationscho), and total employment (e.g., emps_beautyscho) for each establishment category. The Census Tract 2020 dataset includes both tract_fips20 and tract_fips22 variables to accommodate Connecticut's 2022 county boundary changes.