UCLA

My dissertation consists of three essays on U.S. regional economics. The aim of the research is to understand the variations in economic outcomes across regions and how those persistent differences can be reduced. In particular, I study the geographic patterns and the determinants of internal migration, and examine the development of social capital.

In the first chapter, I document U.S. internal migration between the period from 1960-2000 using a newly collected data set. I find that the recent decline in migration is driven by lower migration across states, while within state migration has increased during the observed periods. To explain this, I use a gravity framework and estimate the effect of state borders on migration flows. I find that the border effect is strongly significant, and within state migration is 3.2 times higher than across state migration. Furthermore, the border effect has increased from 2.7 in 1960 to 3.6 in 2000. I show that the differences in social and economic characteristics between areas contribute to a larger state border effect, and the increase of the border effect over time is associated with the rising differences in house prices as states implement more restrictive land use regulations. I find that this is due to lower in-migration in states that are highly regulated in land uses. For high income destinations, the rise in regulations can explain all of the increase in the border effect.

In the second chapter, I examine the effect of political differences on migration decisions and provide empirical evidences of partisan geographic sorting among American migrants. Using presidential election returns and the same migration data for the period from 1960-2000, I show that political differences between areas decrease migration, as Americans prefer to live in areas with similar ideology and political views. I find that there are a lower number of migrants between places that supported different political parties, and migration flows decrease as differences in vote shares increase. In addition, I do not find evidence that Americans are increasingly sorting by partisanship over time, which is previously known in the literature as the “big sort” hypothesis.

The last chapter studies the level of social trust across regions in the United States, and how it arises using natural disasters as exogenous shocks between 1973 and 2010. Every year, the United States is hit by natural disasters that take away lives and cause property damage. In the event of a natural disaster, the victims are more likely to have increased interaction with others. If the experience of increased social interaction is positive, this can positively affect the level of social trust. To test this, I combine two U.S. survey data sets with the natural disaster data, and find that the individuals who have recently experienced natural disasters are more likely to have higher level of trust. One standard deviation increase in natural disasters is associated with an increase of 0.014 standard deviations in trust. I also provide evidence of positive social experiences by showing that natural disasters are associated with an increase in cooperation with neighbors.