Today, census data is consistent across geographies. The set of data available at the state level, for instance, is the same as the set of data available at the tract level. Between 1940 and 1960, however, the Census Bureau generated independent data for tracts and counties, each with a somewhat different set of variables. To facilitate historical analysis and simplify research, we have created a “compatibility set” for 1940 and 1950 that represents all the variables which appear at both the county and tract levels. Subscribers have the option of viewing all county and tract data sets, while the public version includes only a subset of the compatibility sets.