account the types of traffic effects that are and are not reflected in the elasticities (i.e., the elasticities developed by Ruiter et al. represent the net effects on total highway system use, whereas the elasticities developed by Hansen et al. represent the effects on the improved facilities and include diversion from other facilities).

A recently released report for the Department of Transport of the United Kingdom (SACTRA 1994) reviewed the evidence for the existence of induced traffic. As in many metropolitan areas in the United States, the Department of Transport uses transportation models that assume fixed trip patterns (but allow for the effect of general economic growth on traffic growth) in assessing new road projects or major highway capacity additions (SACTRA 1994, iv).

The Standing Committee's review of case studies of major European highway projects, as well as of the Department of Transport's own monitoring studies of before-and-after traffic flows on road improvement projects, found that traffic increases on newly expanded road segments more than exceeded traffic reductions on unimproved segments. This finding provides evidence of induced traffic, that is, of growth in traffic beyond route shifts (SACTRA 1994, 80). However, the Standing Committee indicated that the studies are not helpful in identifying the components of this traffic growth (SACTRA 1994, 76, 77). It is not possible, for example, to distinguish changes in the time-of-day of travel or separate travel growth attributable to improved economic conditions from growth attributable to the road improvement itself, nor is it possible to rule out as a source of traffic growth broader shifts in travel routes than are captured by study control corridors and screenlines (SACTRA 1994, 76).

The Standing Committee also reviewed the evidence for induced traffic using transportation models that allow demand to vary. Predicted estimates of induced traffic from major highway capacity additions in congested urban areas were found to be small when viewed at the network level; the effects were more significant in the corridors directly affected by the road projects (SACTRA 1994, 160). Because the scale of effects depends on the size of the study area modeled the scale of the project, and the behavioral responses modeled, among other factors, the Standing Committee noted the circumstances in which induced traffic is most likely to be large (SACTRA 1994, 169, 170):