Building Social Science Data Infrastructure: The 1900, 1910, 1940 and 1950 PUMS

Diana Magnuson, University of Minnesota
Steven Ruggles, University of Minnesota

This paper documents the creation of the 1900, 1910, 1940 and 1950 public use samples. The US Census Bureau created the first compatible public use microdata sample series for the 1960 (USCB 1962) and 1970 (USCB 1972) decennial censuses of population. The availability of these compatible microdata products invigorated the social science research community (Ruggles and Menard 1995; Magnuson and Ruggles, forthcoming). In the 1970s, two separate research teams independently embarked on projects to extend the series backward by creating historical public use samples for earlier decennial censuses. Samuel H. Preston directed the 1900 and 1910 public use sample projects at the University of Washington and the University of Pennsylvania, respectively (Graham 1980 and Strong et al. 1989). Halliman H. Winsborough at the University of Wisconsin, in collaboration with the US Census Bureau, created public use samples for the 1940 and 1950 censuses (USCB 1984a and USCB 1984b). With support from federal grants, the effort to construct these public use samples outside of the federal government involved enormous organizational, logistical, and methodological hurdles. Drawing on historical evidence from the National Archives and Records Administration (NARA), including US Census Bureau internal memoranda, technical census advisory committee minutes, and correspondence, as well as the Halliman H. Winsborough Collection (IPUMS Archive at the University of Minnesota), oral histories, and a variety of secondary and primary source materials gathered from private and public collections, our research aims to uncover the history of the people and social science behind the creation of the historical 1900, 1910, 1940 and 1950 public use samples.

No extended abstract or paper available

 Presented in Session 193. Building and Interpreting Censuses III