Work in Progress
The Evolution of Child Penalties and Exposure to Working Mothers
This paper investigates what factors led to large reductions in the negative impact of children on the labour market outcomes of women but not men – the so-called child penalty – over time. I link women in the Panel Study of Income Dynamics to their parent's labour supply and show that the evolution of the penalty mirrors that of female labour force participation in the prior generation. I then use Census data and the introduction of childcare subsidies during WWII as a source of exogenous variation in exposure to show that women who grew up around more working mothers have better labour market outcomes throughout their childbearing years. I do not find any positive exposure effects for women without children, or men, which suggests that the effect operates through the intergenerational transmission of norms around motherhood which persist across generations. Alternative explanations such as other demographic trends and changes to the labour market incentives available to single women cannot explain the aggregate decline in penalties over time.
Higher Tobacco Taxes Cause More Food Insecurity Among Smoking Households (with Chris Auld)
Using two decades of data from the Panel Study of Income Dynamics (PSID) and an econometric strategy utilizing non-smokers as a control group, we find that the presence of an adult smoker in the household decreases food expenditures and increases the probability the household is food insecure, and that the magnitude of these effects increases greatly with the tobacco tax rate. At the mean tax rate, a smoker in the household decreases food expenditures by about 2% and increases probability of food insecurity by about three percentage points. These effects are negligible at low tax rates, substantially larger at high rates, and occur largely among lower income households. Also at the mean tax rate, the smoking participation elasticity would have to be larger in magnitude than about -0.6 for the food security benefits from deterring people from smoking to exceed the harm imposed on continuing smokers, which is much larger than both our estimates using the PSID and conventional estimates in the literature.