Criminal behavior in Medellin in times of the pandemic crisis, a spatio-temporal model

Abstract

This paper aimed to identify the effect that the implementation of preventive mandatory social isolation had on the dynamics of security and coexistence in Medellín, Colombia. Using a Bayesian structural time-series model, the weight of this preventive measure in the variation of the indicators was calculated: homicides, forced intra-urban displacement, robbery, motorcycles theft, car theft, sexual crimes, family violence, fights, and personal injuries. The results showed this measure had a direct and significant impact on the behavior of the indicators considered; there were increases between 9.8% and 173%, and decreases between 22.9% and 85%, compared to the projections without this measure.