The missing link(s): Women and intergenerational mobility
The missing link(s): Women and intergenerational mobility
Authors: Lukas Althoff, Harriet Brookes Gray, Hugo Reichardt
Abstract: Research on intergenerational mobility in US history has focused on father-son income correlations. We build a new linked census panel to include daughters (18501940). To also incorporate the role of mothers, we propose a mobility measure that considers parental human capital alongside income (R2) and a semi-parametric latent variable method to estimate this measure from historical data. Our approach reveals increasing mobility, overturning conclusions based on income alone. Mothers’ human capital was more predictive than fathers’ and accounted for the increase in mobility. Aligning with their historical role in homeschooling, mothers were especially important when school access was limited.
Seminar Notes
Venue
· UC Davis 2024 Spring Public/Labor Workshop
Objective
· To understand the role of mothers in intergenerational mobility in the US 1850-1940
Importance
· The literature on intergenerational mobility has focused on men because intergenerational datasets have been missing women and women don’t always have income (the primary measure of economic status)
Background
· An underdiscussed contribution of women to the US economy is the education of children at home, which could vary by mothers’ human capital.
· Outside of New England public school very rare before 1900
· Data & Key Variables: Panel dataset 1850-1940 that includes women.
· Link Census records, adding historical administrative SSA data with maiden names and married names. Match SSN application of Census child to SSN that gives father’s name and mother’s maiden name. Link this backwards to get parent’s background.
· Covers around 1/4 of population in 1930 and 1940 Censuses, good population representativeness
Methodology
· Mobility measured on joint predictive power of parents’ income and human captial (R-squared from regression of parent variables on child outcomes)
· Separate predictive power of maternal vs paternal inputs
· Semi-parametric latent variable method to infer rank-rank mobility using binary literacy indicator
Results
· Maternal human capital key variable before universal access to public schools
· Mobility rises as school access spreads over time (replacing the education provided by mother at home).
· Change role of mothers key driver of rising mobility (lower R-squared=higher mobility)