Daily Labor Mobility and the Timing of First Childbearing

The data present a supplementary material (SM) to the paper "Daily labor mobility and the timing of entry into motherhood" published in Applied Econometrics. Estimation strategy addresses the problem of potential endogeneity of commuting decision. The main estimation approach is a bivariate probit model developed by Heckman (1978) for a model with a binary choice and an endogenous binary regressor. Two other methods, probit model in assumption of exogeneity of commuting decision, and a instrumental variables approach are employed for robustness check. The direct effects of commuting are estimated as the average treatment effects in the age-specific FBRs. Spillover effects of daily labor mobility on the entry into motherhood are tested through the link between FBRs and total fertility rates, relative earnings, and matching indicators in the marriage market, namely, the sex ratio in the night-time and daytime population, and by the proportion of commuters in the population.
The study uses individual register data from Sweden for women born in 1974 and residing in the Stock-holm area following them from 19 to 32 years old. The study uses data from an open source, Swedish Statistics, scb.se, and Individual register data collected by Statistics Sweden and compiled into the longitudinal database of Sweden on Health Insurance and Labour Market Research (Longitudinell Integrationsdatabas för Sjukförsäkrings och Arbets-marknads studier, LISA). The data obtained through the Demographic Data Base, Umeå University.

The SM contains 4 tables:
Table S1. Marginal Effects at Means for Commuters in the Bivariate Probit Model
Table S2. Marginal Effects at Means for Non-Commuters in the Bivariate Probit Model
Table S3. Estimates in the Linear Probability Model with Endogenous Treatment
Table S4. Marginal Effects at Means in the Probit Model