The Rise and Fall of Labor

Abstract

Labor unions exist everywhere, but their strength varies widely by time and by place. It is widely known that levels of unionization have fallen from their peak in the advanced world, and our best evidence suggests that unions never attained these peak levels of influence in late developers. This paper reconsiders explanations of union strength with a view to accounting for these longitudinal and cross-sectional differences. I identify four broad accounts, but note empirical, conceptual, and methodological weaknesses in existing research. I marshal as-yet unused data from the first half of the 20th century, propose an improved measure of ordinary individuals’ disruptive capacity, and employ counterfactual exercises in an effort to improve upon past work. In panel regressions spanning the modern history of the labor movement, I find strong evidence for institutionalist and especially structuralist explanations of union success. The counterfactual exercises illustrate the primacy of the structuralist account. The loss of disruptive capacity explains a larger share of union decline in the advanced world than does the loss of enabling institutions, and the structural handicaps of late industrialization explain the majority of the gap in unionization between early and late developers.