Abstract
There is little doubt that both urbanization and industrialization changed the way people live and interact. However, even though family structure has long been considered as the best indicator of the changes induced, little is known, empirically, about its evolution. We take advantage of a large dataset of matched censuses in a fast industrializing city to investigate how families function in a new environment. We show that family formation confronted two structural forces: the sheer numbers of migrants and the company that dominated the labor market. The company tried to promote a new family model by allowing only some kinds of migrants, selected through housing and labor, to settle in the city. Many aspects of their lives were thus constrained by the firm's paternalistic organization. This process did not occur without resistance but it contributed to the integration of migrants in the city of Le Creusot.
Acknowledgements
We are sincerely thankful to Patrice Bourdelais and Michel Demonet for sharing with us their wonderful database on Le Creusot. Carole Treibich provided excellent research assistance for census linkage.
Notes
1. Information on place of birth is only available in the 1872 and 1876 censuses and, unfortunately, only at the département level. Nationality is recorded more often but the share of foreigners is insignificant before 1870 and still very limited after that date.
2. On the Whipple index see (Mokyr, Citation1983; Spoorenberg, Citation2007; A'Hearn, Baten, et al., 2009).
3. And this bias probably also varies with time; in the very first censuses there is no room on the form to enter the occupation of those who are not head of household.
4. Unfortunately marital status is not given in censuses starting from 1881 so we cannot perform this analysis for the last two dates considered in this paper.
5. We cannot rule out that those recently arrived in the city have a higher mortality rate than individuals of similar age and social condition, for instance as a result of working or housing hardships. This would not be sufficient, however, to explain the high number of people not reappearing from one census to the next.