The impact of land use effects in infrastructure appraisal

When benefits of proposed infrastructure investments are forecasted, residential location is usually treated as fixed, since very few operational transport models are able to forecast residential relocation. It has been argued that this may constitute a source of serious error or bias when evaluating and comparing the benefits of proposed infrastructure investments. We use a stylized simulation model of a metropolitan region to compare calculated benefits for a large number of infrastructure investments with and without taking changes in residential location into account. In particular, we explore the changes in project selection when assembling an optimal project portfolio under a budget constraint. The simulation model includes endogenous land prices and demand for residential land, heterogeneous preferences and wage offers across residents, and spillover mechanisms which affect wage rates in zones. The model is calibrated to generate realistic travel patterns and demand elasticities. Our results indicate that ignoring residential relocation has a small but appreciable effect on the selected project portfolio, but only a very small effect on achieved total benefits.