Modelling crimes as a time series analysis in Uganda

Abstract

This study models and analyzes crime rates in Uganda as a time series from 2019 to 2024 to understand their trends, patterns, and variations. The pervasive nature of crime in Uganda presents a significant challenge to public safety and socio-economic development, yet there is a notable absence of empirical research that systematically examines the temporal dynamics of crime. This gap hinders the ability of policymakers and law enforcement agencies to transition from a reactive to a proactive stance. The main objective of this research is to bridge this knowledge gap by applying time series analysis to national crime data, with a specific focus on analyzing long-term trends and identifying seasonal patterns. The study utilizes a quantitative research design and a deductive research approach, analyzing secondary data from the Uganda Police Force's annual crime reports. The analysis employs a sample of national crime data spanning the specified five year period, with the key dependent variable being the crime rate per 100,000 people. The research methodology involves an exploratory time series analysis, followed by model estimation, selection, and evaluation. Findings from the analysis confirm a significant long-term trend in crime rates and reveal significant seasonal patterns and variations. The results from the ARIMA model provide a foundation for forecasting future crime rates, which can be invaluable for strategic resource allocation and targeted interventions. In conclusion, the study demonstrates the critical utility of time series analysis in understanding crime dynamics within the Ugandan context. The findings offer statistically validated insights for evidence-based policymaking, recommending that law enforcement agencies and policymakers utilize these temporal insights to optimize their resource deployment, enhance public safety initiatives, and move towards a more proactive and preventative approach to crime control.