The Origin of the Marriage Premium. A Long-Term Analysis of the Evolution of the Association between Marriage and Earnings in Post-War Sweden

Martin Dribe, Lund University
Maria Stanfors, Centre for Economic Demography/Dept of Economic History, Lund University

The wage premium married men receive is well-documented. Together with the family gap it is an important contributor to the gender gap, and an obstacle to achieving gender equality in earnings. Family responsibilities obviously pull in opposite directions for men and women; helping wages and careers for men, while being detrimental for women. It is, however, not clear why married men earn more than otherwise comparable unmarried men. Neither are the historical roots of this phenomenon established. The main issue is if the marriage premium related to selection processes into marriage or to real productivity differences between married and unmarried? In this paper we add a historical component to the marriage premium, using data on Swedish men and women between 1947-2015. The data come from a regional population sample, with register data before the beginning of the comprehensive and digitized population registration in Sweden. We have detailed data on earnings and occupation for both the men and their wives, which enable to test important hypothesis for a previously unexplored long period. Using annual data on earnings pre- and post-marriage we estimate marriage premiums using distributed fixed-effects models, which estimates earnings by time to/since marriage to assess the extent to which earnings are causally impacted by the marriage, rather than the result of some endogenous self-selection process or reverse causality. We study the marriage premium during distinct historical periods when dramatic changes took place on the labor market affecting gender relations in order to determine the historical roots of the marriage premium.

No extended abstract or paper available

 Presented in Session 133. The Impact of Cohesion: Single Households, Marriage Premium and in-Laws.