The U.S. Census Bureau conducts several data collection programs about the U.S. population which have their own language, geography, and data portals. This guide is based on a similar guide created by Kelly Smith at UCSD.
The Integrated Public Use Microdata Series (IPUMS-USA) consists of more than fifty high-precision samples of the American population drawn from 15 federal censuses and from the American Community Surveys of 2000-2010. IPUMS is not a collection of compiled statistics; it is composed of microdata which requires statistical analysis software (SAS, SPSS, Stata, R, etc.) to view and manipulate.
ICPSR archives and disseminates census data acquired from the United States Census Bureau as well as files prepared by ICPSR and other principal investigators. Both microdata and aggregate data constitute ICPSR's census holdings. The microdata comprise individual responses to census questions while the aggregate data contain tabulations of the individual responses at various aggregate levels of observation such as states, counties, places, and census tracts.
These are public-use samples randomly selected from all of data collected by a census. ICPSR has microdata from the censuses of 2000, 1990, 1980, 1970, 1960, 1950, 1940, and some earlier censuses, with sampling fractions ranging from 10 to 0.01 percent. Selected digital files for 1790-1960 Censuses also available.
Users who do not have access to SAS, SPSS, etc. may want to try PUMS on DataFerrett, which is a Census Bureau free online tool that can analyze and extract data from the ACS PUMS files and the ACS Summary File.
Raw Data Illustration
This image is taken from the 2000 IPUMS data, 1% sample of Illinois. This data is not compiled into statistical tables. It is machine readable, raw, data for use with statistical analysis software.