The Guide to the American Dream


Why We Defend the American Dream




Land Use

Open Space


Smart-Growth Disasters


Public Health & Safety

Evaluating Congestion in Your Region

The Texas Transportation Institute (TTI) publishes an annual congestion report on the nation's forty largest urban areas along with another thirty-five urban areas of various sizes. The report's calculations are based on data published by the U.S. Department of Transportation on the number of miles of driving and the number of lane-miles of roads to drive on in each urban area.

If you live in one of the seventy-five regions reported by the Institute, you can use its data as a guide to local congestion. Some of the most important of the Institute's various measures of congestion include:

  • The travel time index, which measures the amount of time it takes to make a trip during rush hour compared to the amount of time required with no congestion. An index of 1.2 means that a 10-minute trip without congestion takes 12 minutes at rush hour.
  • Per capita (or per driver) hours of annual delay. People's time is valuable and time wasted is one of the most important costs of congestion.
  • Gallons of fuel wasted. Cars consume far more gas (and emit far more air pollution) in congested traffic.
  • The annual cost of congestion, which is the cost of the wasted fuel, based on local prices, plus the cost of wasted time based on local wage rates.

TTI's congestion data are not based on actual measurements of congestion in every city. The Institute calculates the travel time index and other congestion costs using formulae that assume that freeways, arterials, and other roads have certain flow capacities. When reported uses approaches or exceeds those capacities, the Institute's formula projects that traffic will slow and travelers will be delayed.

This system isn't perfect. Freeways, for example, are not all built alike. The newest roads, such as the Los Angeles Century Freeway, which opened in 1993, can have much higher flow capacities than older roads, such as Connecticut's Wilbur Cross Parkway, which opened in the 1930s. TTI's formulae do not take these differences into account.

This means that TTI's congestion measures are more reliable as a time series for any given urban area than as a comparison across urban areas. While the publicity accompanying each annual update to the mobility report usually focuses on the rankings of urban areas, this ranking is not very reliable.

In the report for 2000, Los Angeles has a travel time index of 1.9, which is significantly higher than San Francisco's index of 1.59. Los Angeles' freeways are so much more heavily used than freeways in most other regions that it is probably reasonable to conclude that Los Angeles has the nation's most congested roads. The next ten urban areas, however, all have indices between 1.40 and 1.47. There is no reason to think that the ranking among these areas is particularly accurate or even that San Francisco's score of 1.59 proves that its congestion is worse than, say, Seattle's (1.45) or New York's (1.41).

TTI's data are more useful for comparing the changes in congestion over time. In most regions, the road network that existed in 1982 (the first year reported by TTI) is still pretty much in place today, so the problem with differences in capacities among roads is less important.

This also means that the best way to compare regions is compare the change in congestion over time. For example, Los Angeles not only has the highest travel time index, it has the greatest percentage increase in this index since 1982.

This measure is also superior because increases in congestion are more stressful than congestion itself. If congestion were constant, people would adjust their travel habits, job locations, or things to compensate. But if congestion is continually increasing, people have to continually adjust or lose more of their time each year.

Unfortunately, the Institute made some changes in its calculations starting with 2000 data. These changes assume that coordinated traffic signals, freeway ramp metering, and similar measures, where they exist, will reduce congestion. Since the Institute did not recalculate the numbers for years prior to 2000, the pre-2000 numbers are not strictly comparable with the 2000 and later numbers in regions that had these measures prior to 2000.

Another problem with the recent mobility reports is that they have made unjustifiable claims about the benefits of transit for congestion. The Institute makes the assumption that all transit riders would, in the absence of public transit, be driving cars. The fact that many transit riders can't drive, and that without public transit many private transit options would become available (options now illegal in many areas as a way of protecting the government transit monopoly), are ignored by the Institute. The report thus significantly overestimates the benefits of transit. It may only be a coincidence that these overestimates first appeared in the mobility reports in the year that the American Public Transportation Association began funding the Texas Transportation Institute.

Nevertheless, if you live in a region included in the Institute's mobility report, you can still find valuable information in the report. This disk includes a spreadsheet that has many of the raw data and calculated numbers for every urban area in the report from 1982 through 2003. The disk also has individual reports for each urban area in the 05tti file folder.

If you live in a region that isn't included in TTI's annual report, you can still get a rough idea of local congestion using the same federal highway data that form the basis of TTI's reports. These data are published in tables HM-71 and HM-72 of Highway Statistics, which is published annually by the Federal Highway Administration. The federal government, in turn, relies on state transportation departments for raw data, so you may be able to get even more information from the states.

Table HM-71 classifies roads as interstates, freeways, other major arterials, minor arterials, collectors, and local. The table gives the number of miles and miles driven on each type of road. Table HM-72 provides population, land area, and freeway lane miles for each urban area. Unfortunately, the tables do not give the lane miles of other major arterials (most of which are probably four or more lanes wide), but you can may be able to get these data (which are in the TTI reports for the urban areas reported by TTI) from the state.

Once you get these data, you can divide the miles of travel on each type of road by the number of lane miles of that road type. As shown in the table below, the most heavily used freeways are in Los Angeles, where each lane mile of freeway supports more than 23,000 vehicle miles of travel a day. The most heavily used arterials are in Washington, DC, where each lane mile supports 8,324 miles of travel a day.

These numbers are considerably higher than the average of the nation's twenty or forty largest urban areas and more than twice as high as the twelve least congested areas in TTI's survey. These twelve areas, including Oklahoma City, Spokane, and Anchorage, all have travel time indices less than 1.1.

Daily Vehicle Miles of Travel Per Lane Mile
                                           Other Major
                               Freeways     Arterials
Most heavily used               23,425        8,324
 Region                       Los Angeles   Washington
Los Angeles (TTI=1.9)           23,425        6,621
San Francisco (TTI=1.59)        20,548        7,047
TTI=1.40 to 1.49 (9 areas)      16,987        7,378
TTI=1.30 to 1.39 (10 areas)     17,043        6,249
TTI=1.20 to 1.29 (21 areas)     14,750        6,386
TTI=1.10 to 1.19 (20 areas)     12,294        6,013
TTI=1.00 to 1.09 (12 areas)     10,360        4,941

The table suggests that urban areas with around 17,000 miles of driving per freeway lane mile and 7,000 miles of driving per arterial lane mile are very congested, while 10,000 miles of driving per freeway lane mile and 5,000 miles of driving per arterial lane mile produces very little congestion. The table also suggests that the differences in traffic levels are found mainly on freeways, as the arterials in most regions have about the same amount of traffic.

It is also interesting to note that freeways can produce far more than twice as many miles of travel per lane mile than other arterials. Yet freeways tend to cost only about twice as much per lane mile as arterials, most of the difference being due to the cost of over- and underpasses. Freeways may be the best investment most regions can make in transportation improvements.

The 2001 Nationwide Personal Transportation Survey estimates that an average of 1.6 people occupy private passenger vehicles. So the above numbers can be multiplied by 1.6 to get daily passenger miles of travel per lane mile.

Passenger miles of travel per dollar of investment might be a worthy criterion for comparing highway and transit projects. But it can be misleading because it doesn't necessarily measure real improvements in transportation productivity.

A better measure is the cost per hour of reduced delay. Most regional transportation planning agencies use computer models that allow them to calculate the effects of various road and transit projects on the total daily or annual hours of delay experienced by local travelers. Proposed highway projects in the San Francisco Bay Area are expected to cost anywhere from $5 to $313 per hour of reduced delay. Bus transit projects were expected to cost an average of $11 per hour saved, while rail transit projects cost an average $52 per hour saved.

Ideally, transportation planners would make this calculation for all possible transportation projects. They would then rank the projects from lowest to highest cost and fund those projects with the lowest costs. While other criteria such as safety are also important, projects with a very high cost per hour saved should not be funded when low-cost projects remain unfunded.