This paper provides new evidence on the role of the educational system for intergenerational
mobility. I evaluate an educational reform, implemented in Sweden in the 1950s and 60s,
which postponed tracking and extended compulsory education from seven to nine years. The
reform may have influenced intergenerational mobility through several different channels.
First, it is likely that the reform increased education more for children from low educated
households, compared to those children with more educated parents. Second, postponing
tracking to higher ages is likely to reduce the parental influence in the educational choice,
which may weaken the intergenerational economic link between children and their parents.
And finally, recognizing that economic well-being is determined by the income of the
household, and that assortative mating plays a major role in the mobility process, I examine
how the reform has affected mobility through changes in marital sorting. The underlying
hypothesis is that the peer group in which couples form can be affected by the educational
system. Differences-in-differences estimates and sibling-difference estimates indicate that the
reform indeed resulted in a sizeable increase in intergenerational income mobility, but effects
operating through marital sorting seem to play a minor role.