Marriage, Child-Bearing and Women's Earnings: Evidence Based on Complete Longitudinal Profiles

IZA Logo

IZA Seminar

Place: Schaumburg-Lippe-Str. 9, 53113 Bonn

Date: 24.07.2007, 12:15 - 13:30


Presentation by 

Ross Finnie (University of Ottawa)


This research exploits the unique qualities of a Canadian longitudinal tax-based datafile (the “LAD”) to present new empirical evidence on the effects of marriage and childbearing on women’s earnings. It differs from earlier research, first, by following a cohort of young women from their late teens through their early forties and controlling for the unobserved heterogeneity across those who marry and/or have children and those who remain unattached by conditioning the analysis on completed marriage-fertility profiles. The approach then consists of estimating separate earnings equations for each marriage-childbearing group which allow shifts in the underlying age-earnings profiles at the point of marriage or childbirth within each group. The approach also allows the effects of marriage and childbearing to evolve over time, thus capturing the effects after one year, after two years, and so on, these profiles of the relevant earnings effects replacing the single parameter(s) for being married and/or having children present typically found in the literature. Marriage effects are separated from the effects of having children, and these latter are broken down by the number of children the woman has and the birth order of the child. The results indicate that i) marriage per se seems to have relatively little effect on women’s earnings, ii) women who marry generally have moderately higher earnings than those who remain single until they begin to have children, iii) married women’s earnings decrease significantly with the birth of each child and then recover to decreasing degrees with each additional child the woman has (even though pre-marriage earnings levels are relatively similar across all the married groups), iv) these effects differ to some degree for women in common law relationships, v) single mothers who have just one child have somewhat lower earnings than others before they have their child, but then fall behind and recover only partly over time, while lone mothers who have two children are characterised by low earnings throughout their careers and quite flat earnings trajectories after the sharp declines that occur with the birth of the first child, vi) these patterns are surprisingly similar for women with and without university degrees, and even more alike for those who have their children at a younger versus older age.

Download complete paper   
For more information, please contact