Local Spillovers in Bankruptcy

The objective of this dissertation is three-fold. First, I improve on spatial cluster tests. These tests identify geographic clusters where data are spatially correlated. I develop a new test called the Local Modified Moran's I, which performs better than existing tests under certain scenarios. Second, I use new data to understand the role of spillovers in bankruptcy filings. I examine how someone - an individual or a business - may decide to file for bankruptcy based on the bankruptcy filings of others around them. Third, I begin to bridge the gap between spatial econometrics and spatial statistics by using spatial statistics, including the Local Modified Moran's I, to examine bankruptcy spillovers.