This article is part of series of blog posts that details the strengths and weaknesses of prominent labor market and education data sources. Find the whole series here.
The American Community Survey (ACS) program was developed by the US Census Bureau. It provides detailed information about the American people and workforce, including data on industries and occupations, educational attainment, veterans, demographics, whether people own or rent their home, and other topics.
Public officials, planners, and entrepreneurs use this information to make decisions, including whether to build hospitals and schools, support school lunch programs, improve emergency services, and build bridges, and how to properly inform businesses looking to add jobs or expand to new markets.
- ACS has a relatively short lag time; it collects data monthly and releases it within 12 months of the survey date.
- ACS collects a wide variety of data, providing information on national demographic, social, economic, and housing characteristics.
- ACS is mandatory for the housing units that are selected, so the response rate is strong.
- ACS is a survey, meaning it is subject to various forms of measurement error such as sampling error, misclassification (industry/occupation) error, and even incomplete or misleading responses.
- Because it is designed to ensure good geographic coverage and does not target individuals, the Census Bureau selects only a small, random sample of about 295,000 addresses (of more than 180 million) to be included in ACS each month.
- The full implementation of ACS began in 2005, so historical data is limited.
How EMSI Incorporates ACS
EMSI’s self-employed dataset (our third class of worker) includes all people who consider self-employment a significant part of their income and/or taking a significant part of their time. EMSI largely bases job counts, hourly earnings, and projections for these unincorporated self-employed jobs on responses to the American Community Survey (with additional input from other sources).
EMSI’s extended proprietor dataset (our fourth class of worker) represents jobs that generate miscellaneous labor income, such as very small self-employment income and partnerships with many partners having limited involvement. EMSI derives job counts and hourly earnings for extended proprietors from differences between ACS and other proprietor counts—the latter of which are based on tax returns and other data compiled by the Bureau of Economic Analysis as well as local personal income reports.
EMSI also uses ACS to construct our demographic data for non-QCEW employees, self-employed, and extended proprietors. In addition, EMSI uses ACS to build national staffing patterns for self-employment (class 3 and 4) and for industries that OES does not cover.
EMSI collects data from more than 90 public sources, harmonizes it, and delivers it so you can use it effectively. To learn more about EMSI data, visit our data page or contact us. Follow EMSI on Twitter (@DesktopEcon) or check us out on LinkedIn and Facebook.