Essays on housing and family economics



Journal Title

Journal ISSN

Volume Title



This dissertation consists of three essays in Housing and Family Economics. In the first chapter, I analyze the interstate migration patterns of families and the effect of labor force attachment of women on joint migration decisions. I show that as the earned income of spouses become similar, the probability of migration falls substantially. This observation is robust in the sense that 1) it holds even after controlling for a rich set of factors that are strongly correlated with relative income, 2) it yields qualitatively similar results when I model the incidence of attrition as another exit, 3) it consistently disappears for the shorter distance moves. I also find that the negative relationship between income similarity of couples and interstate migration is especially strong for supposedly more settled families and couples that have similar labor market characteristics beyond income levels. In the second chapter, I quantify the contribution of women's labor force attachment to the declining trend in interstate migration. I first document that for families in which both spouses have similar incomes, the propensity to migrate is significantly lower than for families with unequal spousal earnings. I then construct a labor search model in which households make location, marriage, and divorce decisions. I calibrate the model to match aggregate U.S. statistics on mobility, marriage and labor flows and use it to quantify the effect of a fall in the gender wage gap on interstate migration. Narrowing the gender wage gap increases women's contribution to total family income; it induces a higher share of families with both spouses working and more couples with similar incomes. The model predicts that the observed change in the gender wage gap accounts for 35% of the drop in family migration since 1981. Finally, in the third chapter, I examine the effects of homeownership on individuals' unemployment durations in the USA. I take into account that an unemployment spell can terminate with a job or with a non-participation transition. The endogeneity of homeownership is addressed through the estimation of a full maximum likelihood function which jointly models the competing hazards and the probability of being a homeowner. Unobserved factors contributing to the probability of being a homeowner are allowed to be correlated with unobservable heterogeneity in the hazard rates. Tentative results suggest that unemployed homeowners are less likely to find a job which is especially stronger for outright owners. I also find that homeowners' nonparticipation hazard does not significantly differ from that of renters' although having a mortgage lowers the chance of exiting the labor force.