Background: Ever since the Princeton European Fertility Project, structural and diffusion effects on fertility behavior have been juxtaposed. However, we still hardly know what the relative effects were of shifting socio-economic conditions and shifts in sociability in explaining the historical fertility decline.

Objective: To what extent and how did structural and diffusion effects play a role in the adoption of fertility control in the Dutch historical fertility transition?

Methods: A national data set was used with more than 3,000 maternity histories of married Dutch women aged 15-50, whose reproductive careers took place between 1870 and 1940. Apart from husbands' occupations, characteristics of the set of couples' marriage witnesses were included to measure their social networks. Cox regression analyses of age at last birth and negative binomial regressions of net family size were conducted.

Results: Results indicate that unskilled laborers and farm laborers were laggards in the practice of fertility control during the Dutch fertility transition. Besides SES differentials, differences in couples' social networks were important in explaining fertility behavior. Those who had networks consisting of lateral kin, age peers, and people of urban background stopped childbearing earlier and had smaller families than other couples did. Particularly the presence of lateral kin of the bride and of female witnesses was strongly associated with smaller family size.

Conclusions: The evidence lends support for so-called "blended diffusion models" and suggests that the fertility transition must be understood as much from the viewpoint of changed cost-benefit calculations related to structural changes, as from shifting patterns of sociability associated with the decline of patriarchy and the increasing lateralization and age homophily of people's social networks.