Female labor force participation in census microdata

Gustafscenen Session 3: Labour, living standards and inequality organized by Erik Bengtsson, Kathryn Gary, Tobias Karlsson, Malin Nilsson and Jakob Molinder


Jørgen Modalsli


The use of historical census microdata to study individual outcomes and intergenerational mobility has become mainstream in the economic history literature. However, because of uncertainty about how female occupations were reported, most papers focus solely on men.

This project will use a combination of historical data sources to assess female labor force participation in the past. As a starting point, I plan to use the 1910 Norwegian census, from which individual records are available, in combination with other statistical sources such as industry censuses (there was a comprehensive census of industries (fabrikktelling) in 1909) and labor market statistics. Moreover, there are contemporary written sources - the issue of ‘correct’ measurement is frequently discussed in the publications accompanying the printed tables. The project will build on existing economic-historical work on female labor force participation (e.g. Goldin, Humphries, Meerkerk) in conjunction with more recent economics work on comprehensive historical microdata.

Based on preliminary analyses, I expect to find a substantial level of female labor participation even when relying on the census alone; for example, 80 per cent of single and 11 per cent of married 40-year-old women in 1910 are registered with an occupation in census microdata. However, from previous literature, it is expected that combination with industrial and labor statistics will find that women’s formal labor force participation is under-represented in the census, in particular in industrial occupations.

The article will focus on (a) the ‘true’ female formal labor force participation in Norway in 1910, and how the definition of ‘formal’ affects this measurement, (b) the distribution of labor force participation by occupation, geography and marital status, as well as a comparison to male labor force participation and (c) an assessment of how well-suited census microdata is as a source of women’s position in the labor market in the early 1900s.

At this point, the focus of a the work is Norway circa 1910. However, I do not rule out a later extension to other countries (for comparative purposes) or to other Norwegian censuses (to assess a time trend).


No PDF available.