Abstract:For each class of transportation control measures (TCMs), the relevant travel behaviors expected to change are identified and techniques for simulating these changes are listed. Then, the latest round of analysis of TCMs in each of the four largest urban regions in California is studied carefully to see whether the relevant behaviors were modeled in a credible fashion, on the basis of local data. In modeling TCMs that change travel time and costs or expand transit options, models were found to lack automobile ownership steps and accessibility variables in some steps. Intersection capacity and delay should be entered into the road networks, and the networks need to be more detailed. In addition, more cost data are needed. Household income should be retained in final trip tables to allow for equity evaluations of changes in travel patterns. In simulating policies that change land uses, walk and bicycle modes should be explicit, and better land use data are needed. For analysis of clean vehicle incentive programs, vehicle types should be linked to trip purposes. Most agencies did a poor job evaluating TCMs; in some cases, they did not even use their travel demand models but instead used spreadsheets with generalized default values. Many improvements are being made to these models, and practice will be improved.