Traffic congestion is one of those rare consensus issues: most people agree that it is a problem, that it costs individuals, business, and government money, and that it seems to keep getting worse despite attempted remedies. For example, Tucson averaged 52 hours of delay at an average cost of $759 per auto commuter in 2017. According to the Texas A&M Transportation Institute (TTI), the nationwide cost of traffic congestion in 2017 (the last year of available data) was $166 billion, averaged to $1,010 per auto commuter. In terms of time, commuters wasted 6.9 billion hours, or an average of 54 hours per commuter – a whole extra work week of time spent traveling due to delays caused by bad weather, traffic accidents, or road construction. The TTI estimates that this is “more than the time it would take to drive to Pluto and back, if there was a road.”
The MAP Dashboard usually compares Tucson to peers at the Metropolitan Statistical Area (MSA) level; however, this traffic congestion data is at the “urban area” level. Urban areas are located within a metropolitan region and have a population density of more than 1,000 persons per square mile. This analysis includes several traffic congestion measures such as annual delay per auto commuter, congestion cost per auto commuter, and total congestion cost. Annual delay per auto commuter is the extra time (in hours) spent traveling at less than free-flow speeds by private vehicle drivers and passengers in typical peak periods. A commuter is considered one of these “peak-period travelers” which is a person who makes a trip between 6 – 10 a.m. or 3 – 7 p.m., with the number of commuters in a given urban area being estimated from the National Household Travel Survey. However, annual delay per auto commuter considers both peak period delay and off-peak period delay, because commuter trips are not always limited to peak periods. Congestion cost is the annual value of delay time and wasted fuel by all vehicles. Congestion cost is presented as both a total cost and per auto commuter in 2017 dollars. The researchers estimated value of delay time for 2017 at $18.29 per person-hour and $54.94 per truck-hour. Excess fuel consumption is estimated using the average cost per gallon of fuel by state.
While exploring the performance of MAP Dashboard urban areas in Figure 1, we found that Tucson had the fifth highest annual delay at 52 hours per auto commuter in 2017. Seven cities’ annual delay was less than the national average of 54 hours per commuter, including Tucson. Portland and Austin tied for the most in the group with annual delay hours per auto commuter at 66, while El Paso had the least at 41 annual delay hours per commuter.
Figure 1: Annual Delay Hours per Auto Commuter (2017)
However, a snapshot of the most recent year’s data congestion might not tell the whole story. A 2004 report from The Brookings Institution identified population growth as a major factor of transportation challenges. Without suggesting that correlation is causation, annual population growth rates of the most congested urban areas in 2017, Portland and Austin, are consistently far higher than the population growth rate of El Paso. The long-term trend of annual delay hours in Figure 2 illustrates this story, particularly for those geographies with population growth rates consistently higher than 2.0%: Tucson rarely exceeded that threshold in the early 2000s and experienced small changes in traffic congestion.
While Tucson’s congestion did not dip in 2008, as many other peer urban areas did at the beginning of the Great Recession, its post-recession population growth and traffic congestion have both been noticeably less than other urban areas that experience higher population growth. For example, the congestion trend line for cities maintaining population growth through the recession (e.g., Austin, Denver, and Portland) shows some recession related drop-off, but also sharp post-recession recovery. Recession decreases in traffic congestion are expected not only due to decreased population growth, but also due to nationwide unemployment increases that mean less commuters in almost all areas. Traffic congestion might also decrease during a recession with drivers making less trips in general to mitigate the cost of driving.
Figure 2: Long-Term Trend in Annual Delay Hours per Auto Commuter
This chart is interactive: Click on and off geographies to update the graph and compare between urban areas.
Additional hours on the road translate to substantial congestion costs, which include the value of both extra time and excess fuel consumed by all vehicles on the road. Figure 3 breaks down the increasing trend of congestion costs from 1982-2014 in select western urban areas by effect on individual commuters. Portland again claimed the top (worst) spot in 2017, with the highest congestion cost per auto commuter at $1,193 for the year. Tucson was third lowest among the 12 peer western urban areas and was also one of eight urban areas, in this analysis, where the cost for the average commuter was less the national per commuter congestion cost of $1,010. Colorado Springs had the lowest congestion cost at $716 in 2017. Congestion cost rank order does not always match annual delay rank order, mostly due to fuel price variance across states.
While total congestion cost is not directly comparable due to substantial population differences between urban areas, looking at total cost provides insight into the magnitude of traffic congestion effects. Tucson’s congestion cost of $759 per commuter in 2017 resulted in a total congestion cost of $598 million for the urban area. Of the 12 urban areas, Phoenix had the highest overall congestion cost of more than $3.0 billion in 2017; however, when adjusting for population, Phoenix had the fifth highest cost per commuter.
Figure 3: Annual Congestion Cost per Auto Commuter
This chart is interactive: Click on and off geographies to update the graph and compare between urban areas.
Long-term trend comparison also reveals that all urban areas in the peer group have experienced substantial congestion cost increases over time at both the micro and macro levels. Considering just individual cost effects, how have these western urban areas performed in the last decade? From 2008-2017, all 12 urban areas in the peer group have increased congestion costs per auto commuter (Figure 4). Tucson had the lowest percentage increase over the decade with 2017 costs that were 6.5% above 2008 costs, while Phoenix had the third lowest percentage increase at 12.4% over the decade. Six of the analyzed urban areas experienced more than 20.0% rise in congestion cost between 2008 and 2017, with Austin increasing 44.0% and Portland increasing 38.2% over the decade.
Figure 4: Percent Change in Congestion Cost per Auto Commuter 2008-2017
Figure 5 maps the variation in 2017 congestion cost per auto commuter in the 12 urban areas relative to each other, as well as indicating whether the urban area experienced less than 20.0% increase in congestion cost from 2008-2017 (blue) or more than 20.0% increase in congestion cost in that time period (red). While there is relatively high correlation between changes in congestion cost and population growth over the same time period, other factors such as local transportation policy may also play a role in congestion increases or decreases. The data used in this analysis specifically measures mobility, which is the speed of moving vehicles, rather than individual transportation preferences or even the effectiveness of alternative modes of transportation. Finally, it is worth considering, as the Victoria Transport Policy Institute suggests, that vehicle owners do not actually factor congestion into travel decisions, but instead consider it an external cost of operating a vehicle.
Figure 5: Annual Congestion Cost per Auto Commuter
This map is interactive: Click on and off geographies in the legend to compare between urban areas.
How is it measured?
Traffic congestion data comes from the 2019 Urban Mobility Report, a joint report from TTI and INRIX published every few years. This report provides congestion estimates for 494 U.S. urban areas. Some known criticisms of the UMR include lack of peer review, the definition of “congestion” as less than free-flow speeds, and the focus on mobility rather than access. INRIX also produces an annual Global Traffic Scorecard.
