Does Inequality cause Crime? Evidence from a Latin American Panel

Sammanfattning: In this thesis, I investigate if inequality has a significant impact on crime rates in Latin America and the Caribbean for the period from 1950 to 2010. Several studies in the past have found robust correlation between indicators of inequality and crime incidence, even when accounting for confounding factors and country-fixed effects. However, a major drawback of those studies is that they ignore the possibility of spurious regression in panel data. By pre-testing the data for unit roots and adjusting the model accordingly, I avoid the possibility of nonsense regression. My results show that once taking into account the possibility of spurious regression, it is no longer possible to find any significant correlation between inequality and crime rates. Applying fixed effects methodology and including additional control variables does not change the results; additional tests likewise suggest that inequality does not Granger-cause crime incidence. The findings lead me to the conclusion that inequality and crime are possibly unrelated in Latin America and the Caribbean and that the significant correlation found by previous studies might have been driven by the commonality of stochastic or deterministic trends in both series.