Forecasting Road Traffic Accidents in Metro Manila Using ARIMA Modeling

Abstract

In this paper, we have determined and analyzed the behavior of road traffic accidents (RTAs) in Metro Manila, Philippines over the period of 2012-2021, and created a forecast for the next 5 years using ARIMA modeling. This study used 10-year historical monthly data collated through the Metro Manila Accident Recording and Analysis System program for the years 2012 through 2021. Our result suggests that the total RTAs in Metro Manila gradually increased until the first quarter of 2020, then it plummeted and reached its lowest point in April 2020 due to COVID-19 lockdown. As lockdown eases, it bounced back but only halfway. Similarly, RTAs resulting in damage to property also bounced back to only halfway as lockdown eases, while RTAs resulting in injuries (fatal and nonfatal) bounced back to their normal range as before the lockdown. Despite the decrease in total RTAs, the ratio of RTAs resulting in injuries drastically increased during the lockdown due to reckless driving behaviors. Using Box-Jenkins methodology of ARIMA modeling, this study identified ARIMA (1, 1, 12) as the best model. With this model, the forecast shows that the total RTAs will stay halfway in 2022 and gradually decrease for the next 4 years.