Road Traffic Fatalities Forecasts Based on Cointegration Analysis

ABSTRACT Road accident-forecasting models are indispensable modules within any efficient strategy that gauge and enhance road safety. These models are used to assess the effectiveness of road safety policies that have been introduced. The greatest share of the annual 1.25 million global road fatalities occur in developing countries, yet very little research is done in this context. This paper presents a simple approach for forecasting traffic accidents in developing countries where minimal data is collected. The model applies cointegration analysis of the road accident and economic data. Cointegration relationship is tested through Engle-Granger two-step method. The error correction model is formed using the fatality rate as dependent variable and economic growth as predictor variable. Using this approach, spurious regression is avoided. The forecasts provide the numerical context through which a mechanism is created to set new road safety targets in Turkey with the intention of assisting the monitoring of potential safety countermeasures and thus promising reduction in road traffic fatalities.


  • English

Media Info

  • Pagination: pp 135 - 142

Subject/Index Terms

Filing Info

  • Accession Number: 01729975
  • Record Type: Publication
  • ISBN: 9780784482575
  • Files: TRIS, ASCE
  • Created Date: Aug 28 2019 3:01PM