Costs of Congestion in Canada’s Urban Areas: Methodological Considerations

‘Congestion’ is commonly cited as a major urban ill by the public, politicians and the media. Urban transportation authorities aim continuously to manage (if not altogether eliminate) the problem through a variety of measures. But what do we mean by congestion? How do we quantify it? What is its cost? Some Canadian urban areas have attempted to answer these questions. However, methods, data, approaches and assumptions have varied. A recent Transport Canada study has provided the first comprehensive and systematic analysis of congestion: The Costs of Congestion in Canada’s Transportation Sector study developed congestion indicators for the nine largest urban areas in Canada (Québec City, Montréal, Ottawa-Gatineau, Toronto, Hamilton, Winnipeg, Calgary, Edmonton and Vancouver). The indicators were based upon data that were derived from each urban area’s travel demand forecasting models. Although the models all produce the same outputs (i.e., simulations of vehicle [and other] trips), there are structural and methodological differences among them. The resultant indicators thus are not easily compared; however, they do provide different perspectives on congestion in these urban areas, which contain over half the population of Canada. Accordingly, the research required the development of common means to measure congestion and extract the requisite data from the nine models. Among other results, the research found differences ranging from how expressways and arterials are defined in model networks to the time periods and trip purposes that are considered in each model, to the ways in which speeds were calibrated. Some of the differences are subtle, while others are more obvious; but all have an impact on the practice and application of travel demand forecasting (as well as to the measurement of congestion). This paper reviews these differences, as well as the points of commonality, and discusses their implications on the analysis of congestion. It explains the role of models in analyzing congestion, and provides a basis for urban authorities to conduct their own congestion analyses. The paper also provides some suggestions for further research.