Population Research Discovery Seminars
Do the growing number of historical intergenerational mobility estimates fail to capture long-run influences from grandparents? Nearly all estimates of multigenerational mobility find that socioeconomic status transmits more strongly between grandparent and grandchild than predicted from a standard intergenerational model. However, a “grandparent effect” could be spurious due to measurement error in the father’s outcome. Using 1850-1940 US census data, we find that accounting for measurement error via instrumental variables almost entirely eliminates evidence for multigenerational drag for white families. The error-corrected estimates suggest that multigenerational persistence for white families is 4 percent stronger than predicted from intergenerational data, as opposed to 45 percent stronger when ignoring error. However, accounting for error also increases the grandfather-grandson association by 40 percent, suggesting that long-run mobility is lower than previously estimated. We also find suggestive evidence that persistent racial disparities across the Black and white population could lead to a small and positive grandfather effect for the Black and white population.
Zach Ward is an Associate Professor of Economics at Baylor University and a Faculty Research Fellow at the National Bureau of Economic Research. He received his PhD from the University of Colorado, and worked at Australian National University until moving to Baylor. His research is in the fields of labor economics and economic history. Most of his work uses linked historical records to answer long standing questions about the history of inequality in the United States, such as the rate of intergenerational mobility and the rate of immigrant economic assimilation. Current research topics include creating large datasets to measure multigenerational mobility for men and women since 1850, and understanding the importance of violence during the Mexican Revolution for migration to the United States.