Policymakers need detailed but clear information about the results of proposed transportation policies in order to choose the best ones for a region and its population. However, current tools available to assist them--specifically, forecasting models--fall short. An urgent need exists for more comprehensive forecasts for predicting the effects of various policies, as well as for tools that can measure and evaluate user satisfaction with a system rather than simply describing what it looks like.

This is one conclusion reached in Understanding Urban Travel Demand: Problems, Solutions, and the Role of Forecasting, a report by CTS Research Associate Gary Barnes and Associate Professor Gary Davis of the Department of Civil Engineering. The recently published report is the second in a series of the Transportation and Regional Growth Study.

Forecasts play a vital role for policymakers in addressing such complex and varied transportation problems as congestion, pollution, and urban sprawl. The difficulty lies not only in defining problems and potential solutions, but also in predicting what influence a policy will have on the people it affects. Because policymakers rely on forecasts for this, the results and implications of these forecasts must be easily understood and appropriately interpreted by non-experts. And good forecasts are critical because policy, once implemented, is often costly and long lasting.

By examining the standard traffic forecasting models in use, the report finds they are not well suited for evaluating many of the policies of greatest current interest, particularly those aimed at reducing the overall amount of travel through changes in land use or travel behavior. These standard forecasting models were developed decades ago, when policy objectives sought to accommodate travel behavior (such as planning and developing highways for more auto travel), rather than to modify it. Little research went into understanding what determined the total amount of travel, such as trip quantities, destinations, and home and work locations--factors that are quite relevant today. Because these factors are still not well understood, improving the models to account for them is a difficult problem to solve.

A potentially easier improvement would be to find ways to measure accessibility (how well people can access destinations) and describe the transportation system in terms of user satisfaction, rather than usage measurements. Since the goal of policy is to make systems better for the people using it, tools should report whether or not this is being accomplished. Also, a well-defined measure of system quality that isn't limited to a particular mode of travel would make it possible to directly compare the benefits of different policies, such as highway versus transit or bicycle proposals.

As a result of its findings, the report recommends continued research to acquire a better understanding of travel behavior and how it responds to different types of incentives, particularly the role of land use; more useful and understandable output from forecasting models; and the development of tools for describing and measuring the overall quality of systems and how this quality varies for different places or groups of people.

The report appears in its entirety on the study's Web site at www.umn.edu/cts/trg. To order a printed copy of the report, call 612-626-1023.