ABSTRACT
A proper census should enumerate the whole population of an administrative area, usually a nation, and ideally be nominative, i.e. include names. Repeated enumerations and published statistical aggregates have become less important ideals, because with computer technology we can analyze the cross-sectional microdata thoroughly. A baseline type are the numeric censuses taken especially in the early 19th century, listing no individuals, only the number of persons by gender, etc. for each geographic location. These pre-censuses resulted in statistics, but lack prospects for reanalysis and are hard to use comparatively. The word “census” means a listing of persons or other entities, while the German “Volkszählung” focuses on the quantitative aspect and the Russian “perepis” stresses the repetitions. The pre-censuses spread from China with the Mongolian 13th century conquests, having their roots among partial enumerations organized to levy taxes and conscript soldiers.
Demographers tried to compute population sizes from vital rates with diverse formulas, since in England and France, regional nobility resisted census taking before the 19th century. The mercantilists took imperial censuses to monitor their colonial settlements after an Indian war in America or a volcano eruption in Iceland. The 1703 census for Iceland was the world’s first nominative aiming to be full count. Casting the census as a dialectic process juxtaposing the state’s classifications and the people’s social categories summarizes the conflicts of the biblical pre-censuses and the enumerations introduced during the 18th-century revolutions.
