| Summary: | The first chapter proposes a geographically decentralized Diamond-Mortensen-Pissarides model of the labor market in which workers in one region have the option to migrate to a different region. Migration is modeled as an extreme value process in ordered space: regions are related bilaterally via a distance matrix. The model aims to replicate stylized facts about both the macro-level and disaggregated labor market over the business cycle, as well as migration. The paper presents novel data on both geographical migration patterns and labor market flows, and together they explain observed spatial heterogeneity and correlation of labor market outcomes at a higher regional level than the local labor market. The implication is that the economy has an underlying spatial structure based in part on differing migration opportunities across regions and which is observable through the migration data. The data then enable calibration of the structural parameters related to the economic geography of a decentralized economy. The model is simulated to match data on labor market flows by MSA and the macro-level predominance of quantity fluctuations over wage fluctuations. The second chapter proposes a preliminary model of international immigration as a dynamic social network in the tradition of Jackson and Rogers (2007). The evolution of international immigration since 1960 is characterized by "super-diversity," that is, the increasing importance of a larger number of smaller bilateral immigrant flows, relative to a past in which colony-metropole flows predominated. That super-diversity arises naturally in a type-biased social network as the out-degree distribution of newer cohorts converges to the overall network type profile.
|
|---|
