Luiza Antonie, University of Guelph
Kris Inwood, University of Guelph
Chris Minns, London School of Economics and Political Science (LSE)
Fraser Summerfield, St. Francis Xavier University and RCEA
This paper explores intergenerational occupational mobility in Canada prior using two overlapping samples of linked census data the censuses of 1871, 1881, 1891 and 1901. We identify boys aged 6-15 years in 1871 who can be traced across the four censuses. We compare the occupations of boys in 1901, when they were 36-45 years old, with the occupations held by their fathers at similar ages in 1871. Thus, we are able to consider how the family’s social position in 1871 influenced the occupational attainment of their sons in middle age. We link the census records using two methods on a decade by decade basis. First a machine learning algorithm (SVM) classifies pairs of records to identify potential individual histories (Antonie et al 2014, 2015). We link across each ten year interval; we restrict ourselves to only those boys who can be linked uniquely across all three intervals. The second samples increases the sample with adding additional links that are obtained through a process of disambiguation. Cases with multiple matches are re-examined using family relationships to identify additional unique pairs following the Jacquard-based disambiguation procedures described in Richards (2013) and Antonie et al (2018). Thus, we have two samples, one of which is much larger, at the price of over-representing boys whose family members remained together for decadal intervals. We wish to answer two questions. 1. Does our understanding of intergenerational movement up and down the occupational ladder depend upon whether or not we disambiguate? 2. Do the patterns of movement off the farm appear to be sensitive to the choice of sample/linking method?
No extended abstract or paper available
Presented in Session 203. Linking 19th Century Census Records