Dual career couples often face the problem of finding two jobs close to each other, mostly in the same metropolitan areas, which is called the co-location problem. The co-location problem has significant implications in the study of the distribution of human resources across cities. In Chapter 1 of this thesis, we employ the data set of National Study of Postsecondary Faculty (NSOPF), which consists of 4 cycles of comprehensive surveys about the faculty and/or institutions in postsecondary education, to study the demographical changes and co-location problem of the postsecondary faculty in USA from year 1987 to year 2004. We document some demographical changes of the postsecondary faculty. Our work shows that the co-location problem has more impacts on the faculty in smaller cities. We find evidences of ability sorting of the faculty among cities of different sizes. Chapter 2 of this thesis develops a parsimonious model that examines equilibrium behaviors by universities as well as by academics. The model has a deterministic structure with cities differing in the variety of schools. Schools in cities with more varieties of schools have an advantage over schools of the same type in cities with fewer varieties of schools, because schools of the other type in the same cities help accommodate spouses with different ability levels. Schools can make hiring decisions based the average of a couple's abilities, which is referred to as couple accommodation policy in this thesis, or based on the merit of each individual in isolation. We establish in this chapter that it is better off for any school to adopt the couple accommodation policy regardless of its location and quality. However, the impact on each school of the couple accommodation policy in the equilibrium where every school implements such a policy differs by school location and quality. High quality schools in large cities always get worse off and low quality schools in small cities always get better off.
Chapter 3 of the thesis provides a probabilistic model of the co-location problem for any labor market in general. In the model, cities differ solely in the number of firms. Firms are ex ante the same to all workers. We establish the theoretical results that dual career couples are more likely to locate in larger cities as compared to single career couples and that larger cities get higher ability workers. We also simulate the job choices by academics, and examine the quantitative implications of the co-location problem for the distribution of ability across cities in the academic labor market. The results show that as the labor market gets thinner or the marriage rate gets higher, the impact of co-location problem becomes stronger, i.e., the ability gap between faculty in large cities and those in small cities gets larger.