HIGHWAY INFORMATION UPDATE

What happens when the peak-hour traffic volume approaches or reaches saturation? The traveler makes changes to adapt to the situation, such as traveling at another time of day, perhaps an hour earlier or later than before, or taking transit, or changing job location. The employer may change the location of the job because of congestion; both the employee and employer may take other steps to minimize travel during the peak hour.

Thus, a leveling off of the v/sf ratio does not necessarily indicate a leveling-off of congestion. It may mean that the severity of congestion has reached a peak, and that the duration of congestion is increasing. Congestion may occur for 4 hours a day, or 6 hours. The v/sf ratio, after all, is limited to a maximum of about 1.0 for any individual section of highway. For nearly 70 percent of peak-hour travel to occur under congested conditions means that much of the peak hour travel is occurring at or near a v/sf ratio of 1.0. It says nothing about how much travel at other hours is occurring under congested conditions.

The major decrease in the percent of travel that appears for the year 1995 is an artifact of the change in capacity calculation procedures based on the Highway Capacity Manual (HCM). A revised HCM was issued in 1994, and subsequent calculations of capacity have been based on the revised procedures. The change in procedures was based on research that showed that drivers were willing to follow each other more closely and at higher speeds than previously. Of course, this change in driving habits occurred over a period of years, but the change in procedure occurred abruptly. This caused the break in the trend that invalidates comparisons between 1995 measures of capacity and previous years.

No doubt our freeway system, especially the Interstate System, will accommodate more vehicles per lane per hour than was the case 20 years ago because of the change in driving habits. Does this suggest a decrease in congestion? Not likely. It does show a willingness to take greater risks in driving, or improvements in driver ability, or improvements in vehicle braking systems--perhaps all three.

Despite the discontinuity in the v/sf ratio figure for 1995, note that another measure unaffected by the HCM change, travel per lane-mile on urban Interstate highways, continues to increase. This is a measure of travel intensity throughout the day. To improve our understanding of congestion, one must increasingly look for measures in addition to the peak-hour v/sf ratio. The Federal Highway Administration's Congestion Management Coordinating Group is currently considering a number of other measures such as delay and travel time. Another need is to look at the issue of non-recurring congestion, which is not measurable by the peak hour v/sf ratio, and which many estimate is as large a problem as recurring congestion.

The Texas Transportation Institute (TTI) has developed time series estimates of the cost of congestion for 50 cities in the Nation. For 1988, this cost was reported to be $34 billion; for 1993, this cost was reported to be $50 billion. (See Urban Roadway Congestion--1982 to 1993, November 1996, by David L. Schrank and Timothy J. Lomax.) This study indicates that congestion costs continue to increase significantly. Note that the TTI report does not use capacity calculations, but bases its estimates of the cost of congestion on traffic volume values per lane.

For additional information, contact Cliff Comeau, HPP-20, at (202) 366-4051 or Beverly Harrison, HPM-20, at (202) 366-4048.