Crime in the United States of America: testing the ‘Broken Window’ hypothesis

This study wishes to examine and validate the ‘broken window’ hypothesis among the fifty one states in the United States. The chosen method for this analysis is Johansen cointegration test to test for cointegration, and if any cointegrating vector is found, we proceed to test for Granger causality based on VECM. We test whether property crime (proxy for minor crime) leads to violent crime (proxy for major crime) in the fifty one states with respect to the period 1960 to 2007. Result of the study indicates that violent crime and property crime are cointegrated in forty eight states out of fifty one states. Further analysis to test for the validity of the broken window hypothesis provides stunning result whereby we found that the hypothesis is indeed valid in forty four out of forty eight states.

Preview PDF (Abstract) Crime in the United States of America.pdf Download (267kB) | Preview

Actions (login required)

View Item