Abstract:
In this paper we estimate social mobility rates, free of measurement errors, using register data for Denmark and Sweden, 1968 to 2021. To correct for measurement error attenuation, we take ratios of the correlation of relatives at different locations in family trees, such as cousins relative to siblings. Three things emerge from these estimates. First social mobility rates in both Denmark and Sweden are much lower than conventionally estimated. Second these countries, despite their reputation for high social mobility rates, have nearly the same degree of persistence as in modern England, and also nineteenth century England or Sweden. Finally in all the cases observed marital assortment is much stronger than conventionally estimated, and this helps explain the low rates of intergenerational mobility.