To Be or Not to Be? The Importance of Group Size and Homogeneity for Union Membership in Sweden

Maria Stanfors, Centre for Economic Demography/Dept of Economic History, Lund University
Tobias Karlsson, Lund University

Industrialization brought significant economic and social changes. As a response, the birth of modern industrial society came with social movements addressing that progress, prosperity and well-being did not benefit all equally. For long, the emergence of trade unions was regarded as a sign of growing “class consciousness” appearing as workers became aware of their interests. Mancur Olson (1965) challenged this view, arguing against collective action due to the collective goods produced, emphasizing the importance of group size and homogeneity, arguing that unions appeared in small workplaces where they could use a variety of social mechanisms to restrict free riding. This was backed up by evidence from the US and the UK. Recently, Schmick (2018) established a hump-shaped relationship between collective action and group size in late-nineteenth century US; workers were most likely to form unions and launch strikes at medium-sized workplaces (50 workers), indicating that the size distribution of establishments can explain a lot of between-county variation within the US, and, potentially also explain cross-country differences in union density. In this paper, we test Olson’s theory, investigating the origins of the Swedish labor movement using matched employer-worker data. Data cover all workers and firms in three industries with different features around 1900 and are linked to regional and community-based indicators of relevance for collective action. We ask: was workplace size important for union density? Did unions attract more members from homogeneous than heterogeneous groups? The case of Sweden is an interesting contrast to the much explored Anglo-Saxon contexts featuring high levels of unionization. The importance of group size and homogeneity has not been investigated from a micro-perspective. Preliminary results indicate a hump-shaped relationship and the importance of group homogeneity for risk sharing.

No extended abstract or paper available

 Presented in Session 110. Scandinavian Labor Markets