DEMONSTRATION DATA FOR U.S. CENSUS BUREAU DISCLOSURE AVOIDANCE SYSTEM
The U.S. Census Bureau is using the 1940 Census full enumeration from IPUMS USA as one of the test microdata files supporting the development of the differential privacy method for confidentiality protection system for the 2020 Census. When the Census Bureau releases algorithms and code for testing the 2020 Census system, those programs will also work on this extract from the 1940 Census. The Census Bureau urges all external testers of its algorithms to also use these 1940 microdata files so that differences in the inputs will not be a source of uncontrolled error in such analyses. Further instructions from the Census Bureau will accompany any public algorithm or code release.
Citation of IPUMS-USA
Publications and research reports based on this extract of IPUMS-USA data must cite it appropriately.Steven Ruggles, Sarah Flood, Ronald Goeken, Josiah Grover, Erin Meyer, Jose Pacas, and Matthew Sobek. IPUMS USA: Version 8.0 Extract of 1940 Census for U.S. Census Bureau Disclosure Avoidance Research [dataset]. Minneapolis, MN: IPUMS, 2018. https://doi.org/10.18128/D010.V8.0.EXT1940USCB
IPUMS-USA Microdata Files for Download
Use the links provided below to download the 1940 full-count dataset, documentation, and statistical command files. Please note that you must be a registered IPUMS-USA user to access the data file. When you click the dataset link, you will be directed to the IPUMS login page where you will be able to login with your IPUMS user information or register as a new user.
1940 FULL-COUNT DATASET (File size: 6.9GB compressed, 39.5GB uncompressed) -- This is a fixed-width data file and contains all variables available for this dataset (about 150). This file can be read into SAS, SPSS, Stata or R with the following command files. The complete list of variables made available are listed in the codebook link below.