Climate Variability and Theft in Colombia

The objective of this study is to estimate the causal impact of the most recent extreme weather event (EWE) in Colombia (‘La Nina’ between 2010-2011), labelled as the ‘winter wave’ by the local media, on persons, houses, business and car theft rates in municipalities subject to the treatment of this EWE. Using a novel annual municipal panel dataset (2007-2012, inclusive), and measuring the affected areas according to the number of houses damaged and destroyed, this study relies on a Difference-in-Difference (D-i-D) model to show that the concurrent year of the winter wave brought a decrease in theft rates, especially, theft from persons. This may be perhaps attributable to the emergence of pro-social behaviour in the municipalities most affected. We also find an increase in theft from houses possibly linked to a ‘survival mechanism’, rather than one that one that seeks reward like the type the BECKER (1968) model of crime and punishment. In addition, the D-i-D estimates also reveal that the presence of conflict, in general, discourages theft perhaps due to the establishment of coercive institutions by illegal armed groups.