文摘
This dissertation explores catalysts to divorce and the effects of different shocks to marital stability. In order to determine how marriage market participation and job opportunities affect marital stability, a panel data set was constructed of all marriages and divorces (or annulments) granted in each county in the United States from 1965 to 1988. The divorce records are merged with county- level employment and population levels to estimate the employment and divorce rates. Using county level data this dissertation is able to exploit a number of labor-market geographical observation levels, such as state border regions, Statistical Metropolitan Areas (SMA), and Labor Market Areas (LMA).
The first chapter analyzes how changes in the number of available marriage-market participants in a community affect the marital stability of existing couples in the area. The analysis focuses on border regions of neighboring states and assesses the impact of fluctuations in divorcee population in one state on the divorce rates in the neighboring states' border region. Large and statistically significant effects are identified in border regions where the neighboring state's border population is larger than one's own border population, which is consistent with the theoretical models on the subject.
In the second and third chapters, attention is turned to how employment opportunities affect marital stability. In chapter two, I use my unique data to more precisely determine the relationship between employment rates and divorce. Using a fixed-effect panel-data model at the LMA level, the results indicate a strong a pro-cyclical relationship between divorce and the business cycle. Finally, in chapter three, the focus of the research transitions, from temporary employment fluctuations, to how permanent changes in the labor market affect marriage. Exploiting structural changes to the labor markets of steel and coal mining, an instrumental variable approach is used that interacts county-level steel and coal industry-concentrations with a national-level demand measure. The model estimates a strong positive relationship between the real-earnings of low-skilled male workers in the county and the county-level divorce rate in steel regions but finds minimal effects on divorce rates in coal regions.