Document Type

Working Paper

Publication Date



Transportation -- Planning, Traffic congestion -- Management, Nationwide Personal Transportation Survey (U.S.), Travel time (Traffic engineering)


Spreading of the peak is one effect of increased peak-period congestion. Due to peak spreading, the travel-time impact of congestion is mitigated for some travelers, but the inconvenience of traveling at a less-preferred time also has a cost. Alternatively, increases in capacity have their impacts on peak-period congestion mitigated by a narrowing of the peak. This reduces the travel-time savings, but it generates a benefit for those traveling closer to their preferred times. This benefit from capacity improvements has largely been ignored, and one reason is the difficulty of quantifying the effect. This paper reports on some crude attempts to quantify the peak-narrowing effect of increased capacity. Using 1995 Nationwide Personal Transportation Survey (NPTS) data, this study investigates how workers in Metropolitan Statistical Areas (MSAs) in the US respond to roadway capacity differences in terms of changing their departure time to work. It is assumed that there is some distribution of travel times given by a mean and some measure of deviation from the mean. Peak narrowing is then measured by decreases in the deviation from the mean. The peak narrowing effects are modeled for both the morning commute and all departures from home for work at any time of the day. As expected, in urban areas with higher roadway capacity per capita, workers tend to depart from home closer to the peak time. For the morning commute, the model estimates that each one-percent increase in lane-miles per capita results in workers departing from home about 5.75 percent closer to the peak time. The effects of roadway capacity are found to be statistically significant in both models.


Catalog Number DP00-4.

Published by Center for Urban Studies, College of Urban and Public Affairs, Portland State University.

Delivered to the 80th Annual Meeting of the Transportation Research Board.

Persistent Identifier