Technological unemployment in Victorian Britain: a tasks based approach

Vipond, Hillary Grace (2024) Technological unemployment in Victorian Britain: a tasks based approach. PhD thesis, London School of Economics and Political Science.

Text - Submitted Version Restricted to Repository staff only until 2 June 2026. Download (5MB)

Abstract

This dissertation studies technological unemployment in 19th century Britain, using a tasks-based framework and granular new data on occupation. While the events of the British Industrial Revolution are often invoked in contemporary debates on the future of work, with the Luddites reliably making a cameo appearance, we know little about the true scope of historical labour displacement. The empirical contribution is a dataset of sub-industry, task-level occupations for 180 million individuals in the British census records (1851-1911). I pair dynamic machine learning with natural language processing to parse tasks from the original occupational text strings. The new data enables, for the first time, a systematic quantification of job loss and job creation in this period. I draw on the new data to study one industry, English bootmaking, as it mechanized over the latter half of the 19th century. It was the fourth largest sector in Britain at the time, employing nearly a quarter of a million people. The task-level data illuminates previously invisible changes in the occupational structure. It reveals that 152,000 artisanal bootmaking jobs disappeared as mechanization set in, while 144,000 new jobs, demanding new skills, emerged. Although the artisanal jobs declined in every county across England, the new jobs emerged primarily in Leicestershire and Northamptonshire. Overall, this resulted in a sharp shift in the geographical location of the industry. To examine who was affected, I link individuals across census waves. The panel data shows that incumbents, particularly men, were rarely displaced. Instead, young men increasingly chose not to enter the trade. In this case, decline unfolded through erosion of entry. Because these findings rely on linked data, the final chapter develops a method for evaluating record linking quality in the absence of ground truth. I show that British historical census data, which benefits from detailed information on place of birth, permits highly accurate linking. I demonstrate that my findings on labour displacement in the English bootmaking industry are robust to a range of linking specifications. The speed of adoption shapes who bears the brunt of labour-saving technological change. When adoption is slow, some demand for old skills persists. If incumbents fill that demand, new entrants may have to look elsewhere for opportunity. This shifts the focus of adaptation from retraining displaced workers to creating new opportunities for young people.

Item Type:	Thesis (PhD)
Additional Information:	© 2024 Hillary Grace Vipond
Library of Congress subject classification:	H Social Sciences > HB Economic Theory H Social Sciences > HD Industries. Land use. Labor
Sets:	Departments > Economic History
Supervisor:	Wallis, Patrick and Minns, Chris and Horrell, Sara
URI:	http://etheses.lse.ac.uk/id/eprint/4867

Actions (login required)

Record administration - authorised staff only