How Do Land Use Trends Affect CBA Outcome?
M Borjesson, S Berglund, P Almstrom, D Jonsson, Royal Institute of Technology, SE
This paper explores the robustness of the CBA outcome of six rail and road investments with respect to land-use scenario. We model three land use scenarios for Stockholm over 24 years: a trend scenario, a central scenario and a peripheral scenario.
The result of a Cost-benefit analysis (CBA) for transport investments is based on a number of assumptions about how input factors such as population, location, economic development, vehicle characteristics, fuel prices and public transport ticket prices develop over time. These assumptions must be made for a future period of at least 10 to 20 years. This study explores how uncertainty in land-use trends affects uncertainty in the outcome CBA of rail and road investments.
There is a considerable body of research dealing with the uncertainties of CBA results due to model errors (the classical discrepancy between model and reality), but not many studies deal with uncertainties in the outcome of CBA due to uncertainties in the assumptions of the input factors (Chung et al. 2007). According to Widlert (2002) and de Jong et al. (2007), however, the uncertainty in the prediction of travel demand due to assumptions of model inputs is often greater than the uncertainty due to model errors. They conclude that the transport models in general seem to provide a satisfactory description of reality but that errors in the demand forecast due uncertainty about the development of the input factors may be substantial.
To forecast how input variables develop over time is usually a time consuming process and is often controversial. Improved understanding of to what extent different input factors affect the outcome of CBA, make it possible to spend more effort on the most important input variables and sensitivity analysis can be undertaken with respect to these.
The aim of the present paper is to explore the robustness of the CBA outcome with respect to different planning approaches (giving rise to different land use-scenarios) over 24 years. Moreover, the investments, if materialized, have themselves an impact on the land use pattern through the improved accessibility. In a conventional CBA the land use scenario is, however, assumed to remain unaffected by the investment, thus underestimating benefits. The second aim of this paper is to explore to what extent this underestimates the outcome of the CBA.
We focus on six rail and road investments in the Stockholm region. We explore how the CBA outcomes of these investments depend on three general land use scenarios for the period 2006-2030: a trend scenario, a central scenario and a peripheral scenario. We use a land use model developed for the Stockholm region to forecast the land use scenarios A logit based transport model developed for the Stockholm region is used to forecast travel demand.
The study indicates three important results. First, land use effects due to major network changes are relatively small regardless of the land use scenario in the region. Thus, the uncertainty in the CBA outcome caused by uncertainties in future land use trends is limited. Second, land use effect due to new infrastructure is limited in a time perspective of 20 years, and these effects therefore also have a limited impact on the CBA outcome. Third, the planning policy and its effect on land use have a considerably stronger impact on accessibility and travel distance than individual road or rail investments have.
Association for European Transport