Job Market Paper

The event of initial childbirth is associated with a highly unequal and persistent adjustment to female labour supply and earnings. It is also associated with an increase in the demand flexible, `family friendly’ employment among mothers. My job market paper explores the hypothesis that under an individual tax structure the interaction of these two factors results in the reorganization of household labour supply around co-employment with childbirth. This is because, as self-employed workers these households benefit from both reducing their taxes through income-splitting and from the flexibility of self-employment.

I show that in Canada, an individual tax jurisdiction, the event of initial childbirth is associated with a discontinuous increase (3%) in joint self-employment among married households. As in the literature, fathers show no sign of reducing their employment with family formation but do switch from the wage-paying sector into self-employment (4%); even pre-empting the event of childbirth. Maternal self-employment increases discontinuously after childbirth (5%), with the majority of this explained by women married to self-employed men. In a parallel analysis of longitudinal survey data I show that this is a real labour supply phenomenon, where co-employed mothers benefit from the same flexibility of other self-employed mothers: increased part-time work and the ability to work form home. This transition to self-employment represents the creation of a `family friendly’ firm.

I then connect this labour supply decision to the income tax structure by means of a novel simulated instrument research design. I exploit exogenous variation in the income tax structure as a shock to the potential value of income-splitting across different birth cohorts. As such, I can estimate a separate elasticity for each event-time: it is both stable and symmetric across men and women, consistent with the income-splitting mechanism. The estimated elasticity of 0.5 and potential savings of 3% suggest that income-splitting can account for a 1.5 %-point increase in male self-employment or half the estimated increase in co-employment with childbirth. This is approximately the difference-in-difference in Canadian and US self-employment among men of the same age with and without children. Thus, income-splitting is not simply a tax avoidance exercise. It is an incentive to be self-employed and co-employed with family formation, that has a long-lasting impact on both male and female labour supply beginning with childbirth.