Texas Transportation Institute Knows how to Measure Congestion

My friends at Streetsblog, upset with the MAP-21 highway bill that revoked the clause allowing entities to spend highway funds on transportation museums, are turning to the administrative rulemaking phase to create a new definition for congestion. Streetsblog is promoting a report written by Joseph Cortright for CEO for Cities. Mr. Cortright is also President of the Portland consulting firm, Impresa.

At the Council for New Urbanism’s 2011 convention, Mr. Cortright debated Tim Lomax of the Texas Transportation Institute (TTI) on whether TTI’s Urban Mobility Report (UMR) accurately measure congestion. According to Cortright, “As a guide to understanding the problem’s of our urban transportation system and what to do about them, the Urban Mobility Report is a broken compass.”

My colleague Bob Poole previously examined some of Cortright’s claims. Overall, Mr. Cortright makes several valuable points. Some members of the media do not fully understand the UMR report and over exaggerate the congestion issues. Also, while the report provides several policy recommendations it does not provide detailed recommendations for every city. Finally, due to data availability, TTI does not always have perfect data.

However, none of these issues is TTI’s fault. TTI does not control how the media reports the news. The UMR report has never claimed to offer detailed solutions for every metro area. And TTI partnered with INRIX several years ago to improve the data it uses. Further, TTI is constantly updating its methodology typically every 1-2 years to try to present a more accurate picture.

Meanwhile let’s examine Mr. Cortright’s claims. Cortright argues that the travel time index is unreliable because it ignores the distance traveled. However, the UMR travel time index includes the delay per hour per traveler. Calculating the delay requires knowing the distance the driver travels. Distance is factored indirectly into the calculations. Results show that there is a strong relationship between the hours delayed and the travel time index. To use a real-world example, a commuter could be delayed on Lakefront Drive in Chicago or on I-77 in Charlotte but either way the driver is delayed. And while it takes 25.2 minutes to commute one-way work in Charlotte, it takes 31.0 minutes in Chicago. Further Cortright’s procedure to estimate trip lengths may be unreliable since it makes unproven assumptions about the data. Some of his points are valid, but his claims that TTI’s measures are flawed are not substantiated by the data. Cortright’s claim that his methodology is better is also not substantiated by the data. Chicago may have a better transit system than Charlotte, but that is not what TTI’s travel time index is trying to measure.

What is really occurring is that Cortright is trying to use analytical methods to prove a new-urbanism ideal: Charlotte’s problem is that a higher percentage of residents live in suburbs. If only people lived closer to their workplaces in denser communities we could solve all congestion problems.

However, this premise is not proven by facts. Further, it is unrealistic. Many households have two or three wage earners each with employment in a different area of the metropolis. Many of these families live in a specific suburb because it is closest to all three jobs. Second, in many metro areas the highest concentration of jobs is no longer in downtown. In some places it is no longer in the principal city. For example, the largest concentration of jobs in metro Atlanta and in fact in the entire southeast region is in the Perimeter Business district. This district is located in the suburban cities of Sandy Springs and Dunwoody not Atlanta. Third, many families consider other factors such as parks and schools. Unfortunately, many central cities have problematic schools and a major lack of park space. Parents number one concern is often their children, and the trade-off of a longer commute is worth it. Making congestion worse is unlikely to change their equation. Fourth, the central city is often the most expensive. Not everyone earns a six-figure salary and can afford to live there, even in an apartment or condo. Fifth, in some metros it is physically impossible, short of becoming Mumbai, for everyone to live very close to their workplace. This is particularly true for cities with a super dense concentration of jobs.

In reality, Cortright is doing, what he encourages Lomax of TTI of doing. He is proposing a one-size fits all standard. Using Cortright’s logic we should never widen a highway, never increase the speed of transport and never consider the costs of travel time on society. Obviously this does not make sense. One reason for worsening congestion in most places including both Pre-World War II cities like Chicago and Post-World War II cities like Charlotte is that despite substantial population growth we have not added much new capacity. When more vehicles are forced to squeeze onto the same amount of road capacity, congestion is going to get worse. We have been using Cortright’s methods in many places and congestion has still worsened. While transit is part of the solution, decentralizing job centers and multi-worker households make creating a transit network more challenging.

Some of Cortright’s claims are designed to appeal to economic development arguments. However, studies repeatedly show that traffic congestion leads to economic losses. And transit may offer a choice but it does not decrease congestion. At a certain level, the congestion has to be addressed. If a business is choosing between two cities equal in other respects, the business will almost always choose the city with less congestion.

Further Cortright’s report fails to offer its own policy claims. But one solution, managed lanes, can benefit all transportation users not just drivers. Managed lanes, which TTI vigorously supports, enables mobility by charging a market price for travel instead of the one-size-fits all approach. Additional lanes can be part of the solution. But the biggest benefit is to transit users who have more reliable, frequent service.

Streetsblog and other environmental groups have historically been weak on analytical claims. Their arguments have typically focused on normative issues such as urban design and non-economic based preference surveys. It is a step forward to see such an analytical report. Unfortunately, while the numbers add up, the underlying premise behind the numbers is flawed. It is not enough to provide a model, the underlying pretense has to be correct. Models deal with human data and if the inputs are junk so are the outputs.

I understand why politically Streetsblog is trying to punch holes in TTI’s methodology. Streetsblog wants less money to be spent on roads and more money to be spent anywhere else. If Streetsblog can actually show that improving roads, specifically using managed lanes, does not improve travel, I will be the first to support a different approach. Unfortunately, this model appears more geared for political than transportation purposes.