This dissertation consists of three chapters. In the first chapter of the dissertation, I propose a new method for estimating the racial tipping points of US neighborhoods. Applying this method to US census data, I find that the average neighborhood tipping point has increased from 15% in 1970 to 42% in 2010. Moreover, I find that from 1970-2010 that differences in racial preferences play a diminishing role in explaining cross city differences in tipping points. Differences in the outside options of white household’s, on the other hand, play an increasingly important role in explaining cross city differences in tipping points from 1970-2010. In the second chapter of the dissertation, which is joint work with Kent Smetters, we document an im...