Call for proposal

Studying migration through census data


A seminar in Bordeaux (15-16 May 2012)


A joint organization by IEDUB, INED and CED
Population censuses have two essential characteristics for the study of migration. On one hand, they are based on large numbers and provide fine geographical accuracy, thus allowing detailed studies of internal migration. Indeed, most censuses around the world include questions about the previous residence of persons enumerated and/or their place of birth. Moreover, even samples drawn from most of these censuses are numerous enough to study the populations of territories with more than 10 000 citizens. On the other hand, the fact that census dates are clustered lets provide, almost simultaneously, populations categorized by their geographical origin. For example, persons from a given country can be numbered abroad in various countries, so offering the possibility to study diasporas and, more generally, to apprehend the flows of international migration.

Up to recently, statistical institutes did not provide sufficiently detailed data for external researchers. The study of migration from census was essentially limited to internal migration and immigrant populations in one’s country. Over the past ten years, the provision of microdata by the statistical offices themselves, or through the IPUMS/IECM project, has considerably enlarged the possibilities for migration analysis from censuses. It has become possible to compare the factors of internal residential mobility in several countries, or to study in more detail the structure of some diasporas despite their relative dispersal.

However, those analyses present many methodological problems:

- Differences in definitions of basic geographic levels across countries,

- Variability of the question about the previous residence,

- Imperfect quality of the data associated with this question,

- Inaccurate simultaneity of the dates of censuses,

- Reporting biases concerning the year of arrival for immigrants in their country of current residence...

This seminar aims both to examine the difficulties posed by census data to study internal and international migration, and to present the results that can nevertheless be obtained.

Papers concerning internal as well as international migration are welcome. Those involving an internationally comparative approach (comparison of "systems" of internal migration between countries, diaspora studies using census data in several countries) or methodological (compare census to another source for studying a same migratory behaviour) will be highly praised.


You can send an abstract (1500 characters) before April 16, 2012 to beatrice.valdes@ined.fr.
Acceptance will be notified by 23 April 2012.

Texts, and oral communications, may be submitted in English or French, but the Power Points will be in English at the seminar.