Migration is often thought of as a risky endeavor in which a migrant trades a known low return for an unknown but potentially higher return. However, migration has been empirically linked to insurance mechanisms through remittances. Chapter 1 unifies the risk-taking and insurance-seeking behaviors of migration into a single framework by framing the migration decision as one of income diversification in which multiple agents within a household to decide whether or not to migrate. Each migration strategy (no migration, partial migration, and full household migration) has its associated risks which are weighed against the returns the household could gain through choice of that particular migration strategy. I test the framework by estimating the probability of each migration strategy for Indonesian households during the period 1993-1998. The framework performs reasonably well in the case of urban households. However, the framework's predictions do not hold as well for rural households, which may be linked to the fact that they function within a larger insurance network than the nuclear family.

In Chapter 2, I find that the response of return migration to GDP per capita can differentiate migrants who are seeking increased consumption for their household (i.e., consumption-oriented migrants) from migrants with intentions to invest at origin (i.e., investment-oriented migrants). Each type of migrant should have differential responses to GDP per capita at destination and may have differential responses to GDP per capita at origin. Using data on Mexican households between 1992-2002, I show that migrants returning from the USA exhibit characteristics of consumption-oriented migrants and migrants returning from internal locations exhibit characteristics of investment-oriented migrants.

Chapter 3 is a published work in collaboration with Sandra Decker, Jalpa Doshi, and Daniel Polsky which uses Medicare claims data linked to two different surveys--the National Health Interview Survey and the Health and Retirement Study--to describe the relationship between insurance status before age 65 years and the use of Medicare-covered services beginning at age 65 years. Although we do not find statistically significant differences in Medicare expenditures or in the number of hospitalizations by previous insurance status, we do find that individuals who were uninsured before age 65 years continue to use the healthcare system differently from those who were privately insured.