How Should We Measure Traffic Congestion?

Transportation planning is undergoing a paradigm shift which is changing the way we define transport problems and evaluate solutions. The old paradigm evaluated transport system performance based primarily on motor vehicle traffic speeds, using indicators such as roadway level-of-service, the Travel Time Index and their variants such as the INRIX Congestion Scorecard and the TomTom Traffic Index, which measure congestion intensity, that is, the amount that motor vehicle traffic speeds decline during peak periods. Such indicators are useful for individuals making short-term travel decisions, such as how to travel across town during rush hour, but they are unsuited for strategic planning purposes since they fail to reflect congestion exposure (the amount that people must drive during peak periods), and so ignore the potential congestion reduction benefits from shifting modes or more accessible development patterns which reduce travel distances. For planning purposes it is important to use indicators which measure congestion costs, such as per capita annual congestion delay.

The table below summarizes various traffic congestion indicators. Some only measure congestion intensity, while others are more comprehensive (they consider total traffic delay, taking into account travellers’ exposure to congestion as well congestion intensity) and multi-modal (they consider delays to all travellers, not just motorists), and so measure total congestion costs. It is important that everybody involved in transportation planning understand these distinctions.

Common Congestion Indicators

Various indicators are used to evaluate congestion. Only a few are comprehensive and multi-modal.

More comprehensive and multi-modal indicators are needed to evaluate planning decisions that involve trade-offs between traffic speeds and other accessibility factors. Let me give two examples:

• Consider the evaluation of converting a general traffic lane into a bus-lane. If evaluated using congestion intensity indicators, such a conversion only seems desirable if the bus lane attracts enough former motorists that traffic speeds increase on the remaining general traffic lanes; travel time savings to transit passengers are ignored. Congestion cost indicators evaluate roadway performance based on person-travel rather than vehicle-travel, and so recognizes the overall time savings that result from giving priority to higher occupant vehicles, such as buses, on congested roads. This is fairer and more efficient because a typical urban-peak bus passenger uses less than a tenth as much road space as an automobile passenger, so they should not be delayed by congestion generated by motorists.

• Consider the evaluation of the best location for a large commercial building. If transportation impacts are evaluated using roadway level-of-service, an infill location is often found to degrade local traffic conditions, since local roads are at capacity, so the project would be burdened with a transportation impact fee to help finance roadway expansions. This discourages urban infill development, despite its much greater overall accessibility, and favors urban-fringe, automobile-dependent development, although this increases total per capita vehicle travel, travel time and congestion costs. Evaluation based on per capita congestion costs or overall accessibility, which account for the lower vehicle trip generation and shorter trip distances of infill location, and the resulting savings and benefits, favor more compact, multi-modal development.

Shifting from congestion intensity to congestion cost evaluation is likely to result in more bus lanes, more compact development, and more use of transportation demand management strategies than currently occurs. This supports strategic planning objectives to increase transport system efficiency and diversity.

These changes are causing a lot of confusion and debate. For example, the state of California is developing new methods for measuring transport system performance in environmental impact statements, as discussed in the paper, Preliminary Evaluation of Alternative Methods of Transportation Analysis. California wants to replace roadway level-of-service with more comprehensive and multi-modal indicators. This has stimulated considerable debate on the Institute of Transportation Engineers professional discussion form; some members, those who apply the old automobile-oriented planning paradigm, are aghast.

California proposes using relatively simple indicators including total vehicle miles traveled (VMT), automobile trip generation, multi-modal level-of-service, total fuel consumption, total vehicle hours travelled, and an assumption that infill development improves overall accessibility. I don't think that any of these proposed indicators is perfect, but I believe that this is an important discussion. Of all transportation performance indicators available, roadway level-of-service is the worst for achieving strategic planning objectives because it favors automobile-oriented improvements, particularly roadway expansions, undervalues improvements to alternative modes, and discourages infill development. All of the proposed alternatives are better for identifying projects and mitigation strategies that improve overall accessibility and transport system efficiency.

One of my recent blogs, Smarter Congestion Evaluation – An Example, criticizes the report, Transit Utilization and Traffic Congestion: Is There a Connection?, in part because it evaluated congestion impacts based on the Travel Time Index. One of the report’s authors, Tom Rubin and I have had subsequent correspondence about my critique, particularly concerning this issue. Rubin admits that the Travel Time Index has “issues,” but continues to defend their use. He does not seem to understand the degree that this indicator biases analysis results.

What do you think?

• Which performance indicators do you think most appropriate for evaluating urban transport projects (such as transit service improvements) and land use developments?
• What other indicators or approaches do you recommend?
• What importance do you think traffic congestion should play when evaluating overall urban transport system performance? What other impacts should be considered?

For more information see:

DfT (2006), Transport Analysis Guidance, Integrated Transport Economics and Appraisal, Department for Transport.

Richard Dowling, et al. (2008), Multimodal Level Of Service Analysis For Urban Streets, NCHRP Report 616, Transportation Research Board.

Jonathan Levine, Joe Grengs, Qingyun Shen and Qing Shen (2012), “Does Accessibility Require Density or Speed?” Journal of the American Planning Association, Vol. 78, No. 2, pp. 157-172. 

Todd Litman (2012), Toward More Comprehensive and Multi-modal Transport Evaluation, Victoria Transport Policy Institute.

Todd Litman (2012), Smart Congestion Relief: Comprehensive Analysis Of Traffic Congestion Costs and Congestion Reduction Benefits, Victoria Transport Policy Institute, presented at the Transportation Research Board Annual Meeting.