Marriage Formation and Dissolution in the United States
This dissertation consists of two essays that examine marital formation and dissolution in the United States. The first chapter highlights the roles that both the availability of men and competition from women within a marriage market play in the low marriage rates of uneducated black women. Black women who drop out of high school are far less likely to marry than those who do not; however, they also, counterintuitively, face much more favorable marriage markets than more educated women if we define marriage markets as independent by education level, as is standard. Using a simple model of the marriage market with men and women of different quality levels that allows for marriage market integration across education levels, I show that the marriage prospects of any woman should depend not only on the availability of men, but also the competition from more educated women. Additionally, this model predicts that any gender imbalance should disproportionately affect the marriage prospects for the least educated. Using data from the 1979-2004 waves of the NLSY79, I estimate discrete-time hazard models of first marriages for black women, capturing a woman's marriage prospects in four ways: (i) using a flexible specification that includes five ratios for the relative availability of men as well as the prevalence of competing women at each education level, (ii) using the ratios for the availability of men and women at adjacent education levels, (iii) using an education-specific simple sex ratio from the educationally segmented marriage markets that dominate the literature, and (iv) using a "cascading'' sex ratio implied by the simple model. The results emphasize the importance not only of the supply of men, but also of the competition from other women for the least educated women. Thus, marriage market measures that do not account for this cross-education competition greatly overstate the favorability of the marriage markets for uneducated black women.
The second chapter is joint work with my advisor Marjorie McElroy and Tongyai Iyavarakul which presents a new approach to the reduced-form estimation of dynamic models using aggregated panel data. With forward looking behaviors, exogenous changes in laws or rules give rise to selection effects on those considering entry and surprise effects for those who have already entered. We develop a model, the Cohort Panel Data Model (CPDM), to examine the effects of divorce law changes on divorce rates. Our analysis has several key features. First, we introduce the concept of a marriage cohort, a group of people married under a certain set of laws. The calculation of the shares of the population in each marriage cohort is a key element of our aggregation protocol. Second, the model includes floodgate effects, a spike in the divorce rate followed by a decline, resulting from heterogeneity within these marriage cohorts. Lastly, we develop a wait-time index for the cost of divorce and carefully code divorce laws based on two dimensions: costs and rights. Thus the model allows for an unbiased test of the Coase Theorem since our model gives the surprise effect of a change in the rights to divorce holding costs constant. Results strongly support the idea that unanticipated divorce law changes affected divorce rates, but finds that these effects are operating through changing costs and not changes in rights (supporting the Coase Theorem). Additionally, lower divorce costs are found to increase divorce rates through the selection of less well-matched couples into marriage. Finally we reconcile our results with previous studies by showing that previous studies that do not differentiate between costs and rights suffer from omitted variables bias and improper aggregation over marriage cohorts.
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.
Rights for Collection: Duke Dissertations