Cocaine Seizures and Crime: Data Analytics Using Big Data Tools

Colombia's status as the largest cocaine producer in the world has prompted its government's strategies to combat drug trafficking. One of these strategies is to seize cocaine in the Colombian jurisdictional territory. The unintended consequences of this strategy on crime rates, particularly homicides, remain uncertain. Web scraping methods and big data tools were used to gather and construct a time series dataset on cocaine seizures from three distinct websites, while the homicides dataset was supplied by the Colombian Ministry of Defense (MDN). This study aims to investigate, from a quantitative standpoint, whether there is a link between cocaine seizures and homicides in the Colombian Pacific region, utilizing an exploratory data analysis (EDA) method and machine learning techniques. The study recognizes the constraints of the sample size and opts to reveal valuable insights through data analysis and modeling instead. Despite the constraints, two models were developed to partially explicate the significance of this correlation. The study's findings provide value for policymakers, military personnel, government officials, and academics, offering essential perspectives to devise improved policies and strategies to mitigate drug trafficking in the Colombian Pacific region without exacerbating homicide rates. Future research endeavors could consider expanding the sample size of the cocaine seizure time-series dataset to conduct a more robust correlation analysis.