Mining Microdata:

economic opportunity and spatial mobility
in Britain, Canada and The United States, 1850-1911

Supported by

Aims and objectives

The project aims to create representative longitudinal panels of census data in a comparable manner in three countries, and contribute to a long-standing debate on social structure and opportunity in Britain and North America. Given the recent availability of large-scale census databases the challenge now in constructing panel data from censuses is not finding sufficient cases, but ensuring that the panels are representative and unbiased. False links lead to artifactual social mobility, so it is important to ensure high levels of accuracy. We aim to do this in a similar way across Canada, Great Britain, and the United States taking account of differences in census enumeration methods and questions. We will use the resulting panels of data to measure and compare social mobility between 1850 and 1911.

To achieve this aim we need to reach three intermediate objectives

1. Creation of longitudinal panels of men aged 0-15 years old in the initial census in each country for the periods 1850/1 — 1880/1 and 1880/1 — 1910/1, using similar methods in all three countries. While the censuses were conducted in similar ways in all three countries, some differences may affect linkage rates—the proportion of individuals from the initial census linked to the subsequent. The most important differences that might affect the comparability of the samples are the greater precision of birthplace enumeration in Great Britain (parish, compared with province/state in Canada/United States); and the de facto enumeration of people in the British census which enumerated people where they were found on census night, rather than the de jure method of enumeration in Canada and the United States that enumerated people at their permanent residence. The de facto enumeration in Britain may mean that men whose fathers were in occupations more likely to travel away from home are under-represented in the British panel. We will investigate whether there are systematic differences between the countries that may result from this difference in enumeration methods

2. Classification of occupations of fathers and sons into a unified social class schema appropriate for comparison between the three countries. We will start by measuring social class using the widely used Registrar General’s 6-class scale, and the emerging HISCLASS schema, based on the HISCO occupational classification system.

3. Modelling of occupational transitions between generations in a comparable way in each of the three countries. The occupational distribution of the three countries was both different and changing. We will use methods that adjust for these differences over time and space.