Discussion Paper no. 04-07


The advent of molecular genetics has brought invaluable information, which is now routinely used by anthropologists in their attempt to reconstruct our demographic past. Since the mitochondrial DNA loci are much more similar between populations than are the Y chromosome loci, it has been suggested that women had a much higher migration rate than men throughout history. Based on an examination of intergenerational migration patterns in three large demographic databases, we bring this inference into question. In some early Canadian settlements (St. Lawrence Valley and Saguenay), and in the past Krummhörn region of Northwest Germany, men whose father was a migrant were more likely to migrate, while the migration probability of women was largely independent of that of their mothers. As a result, men’s movements were less effective in preventing genetic differentiation between populations than women’s movements. If it is largely prevalent among human societies, this male-specific transmission of migration propensity could partly explain the geographical clustering of Y chromosome distributions. In order to account for its impact, we propose a slight modification of Wright’s Island model. We also address the relevance of this model with respect to previously reported measures of population differentiation and we discuss the supporting historical and anthropological literature. We conclude that the widespread patrilocal rules of post-marital residence have generated both a higher female migration rate and a patrilineal dependency in the propensity to migrate.