Highway travel times increased between 1990 and 2000 in 70 of the 75 urban areas studied by the Texas Transportation Institute. The average Travel Time Index (TTI) for these areas in 2000 was 1.39, an increase from 1.31 in 1990 . This means that in 2000, it took 39 percent longer, on average, to make a peak period trip in urban areas compared with the time it would take if traffic were flowing freely (see box).
Travel times tend to deteriorate as urban area size increases (figure 13). For instance, Los Angeles, California, had the highest TTI (1.90) in 2000, while Anchorage, Alaska, and Corpus Christi, Texas, had the lowest (each 1.04). Of the urban areas with the highest index in 2000, only three had a population under 1 million: Austin, Texas (1.27); Charlotte, North Carolina (1.27); and Albuquerque, New Mexico (1.26). At the other end of the spectrum, urban areas of over 1 million people with low indices include Buffalo-Niagara Falls, New York (1.08) and Oklahoma City, Oklahoma (1.09).
Between 1990 and 2000, the greatest increases in TTI generally occurred in small- and medium-sized metropolitan areas, while the increases were more moderate in the very large and small areas1 (figure 14). Overall, the average index for large urban areas increased by 10.2 percent, while that for medium urban areas was up by 8.3 percent. In small and very large areas, the increases were 4.7 percent and 4.1 percent, respectively.
The Texas Transportation Institute analyzed congestion for the Federal Highway Admin-istration (FHWA) for almost 400 urban areas between 1987 and 2000 . In 2000 for those areas, an average peak period trip required 51 percent longer than the same trip under nonpeak, noncongested conditions, equivalent to an index of 1.51. The values in the FHWA report differ from those in the Texas Transportation Institute annual study, due to differences in the scope of the two analyses.
In urban areas, where highway infrastructure is typically well developed, the principal factor affecting travel times is highway congestion resultings from both recurring and nonrecurring events. Recurring delay is largely a phenomenon of the morning and evening commute, although in some places congestion may occur all day and on weekends. National estimates, based on model simulations, of the effect of nonrecurring events on freeways and principal arterials suggest that about 38 percent are due to crashes, followed by weather (27 percent), work zones on freeways(24 percent), and breakdowns (11 percent) .
1. Chin, S.M., O. Franzese, D.L. Greene, H.L. Hwang, and R. Gibson, “Temporary Losses of Highway Capacity and Impacts on Performance,” Oak Ridge National Laboratory, 2002.
2. Texas A&M University, Texas Transportation Institute, 2002 Urban Mobility Report (College Station, TX: 2002).
3. U.S. Department of Transportation, Federal Highway Administration and Federal Transit Administration, 2002 Status of the Nation’s Highways, Bridges, and Transit: Conditions & Performance, Report to Congress (Washington, DC: January 2003).
1 Very large urban areas have a population of over 3 million; large urban areas, 1 million to 3 million population; medium urban areas, 500,000 to 1 million; and small urban areas, less than 500,000.