Some Methodological Problems in the Study of Multigenerational Mobility

A number of recent studies by sociologists have sought to discover whether a person’s status (typically their social class, education, or socio-economic status) is directly affected by the status of their grandparents, once the effects of parents’ status are controlled. The results have been ambiguous, with some studies finding a direct effect of grandparents on their grandchildren, while others find no effect. I use causal graphical methods to demonstrate some of the methodological problems that occur in trying to identify this direct effect, and I offer some suggestions as to how they might be addressed.