The way European populations are enumerated has considerably changed in the last fifty years. The monopoly the exhaustive and periodical census still had in the 1960s has been continuously eroded by alternative forms of counting that give an increasing role to population registers or to surveys, or to mixed forms. It has been paralleled by a diversification of techniques to collect information on individuals and households.

Beyond diversity in modes of data collection, European censuses are also characterized by an important diversity in content, i.e. diversity in census questionnaires when classical forms persist or diversity in information extracted from population registers when they are the basis of “virtual censuses”.

After years of efforts aiming at systematization of census material through European space, diversity in the modes of data collection and contents has been recognized by a 2008 EU regulation, which leaves countries free to choose their own procedures but imposes common concepts, so that comparable results can be produced everywhere. This decentralized but coordinated system of access to comparative macro data concludes forty years of changes back and forth through different combinations of ex ante and ex post data harmonization.

To document this long-lasting movement we use not only legal texts issued by European authorities but also reports from preparatory and expert meetings archived at Eurostat, the Conference of European statisticians and national statistical institutes. In parallel, we interview actors of this piece of history, in particular statisticians involved in the decision process from the 1970s to now. That will give us a full picture of a process which is also shared by most other fields of European statistics.