The purpose of this project is to develop a way to value encroachment effects from transport infrastructure (noise, barrier effects, vibrations etc.), in order to be able to include such effects in cost-benefit analyses. At present, such effects are mostly not included, with the occasional exception of noise. There is still no agreement on how the such effects should be measured and valued. In this project, stated preference methodology is used to measure how the willingness to pay for residential housing is affected by proximity to roads and railways of different types. Respondents were sampled from two areas just outside Stockholm where major roads and railways cut through residential areas. Some of the respondents lived close to these roads or railways, but not all of them. All respondents had bought their house within the past five years. Thus, it can be assumed that most of the respondents have some knowledge about potential problems with living near roads and railways. It can also be assumed that most of the respondents have considered what decrease in purchase price they would need in order to be willing to live near a road or railway. They were asked to imagine themselves in the situation when they bought their house, and the prices of the alternatives presented in the SP exercise were based on the price the respondent had paid at the time of the purchase. Valuations are strongly non-linear in the distance to the road or railway. We compare several different non-linear and nonparametrid estimation methods. The most robust result come from a weighted average of several piecewise linear valuation functions, essentially emulating a non-parametric method. Valuations differ significantly across the population, making it important to distinguish between how intrusion effects are valued by the average individual, and how intrusion effects affect market prices, which will be determined by the valuations of the 'marginal buyer'. We show that there is a "marginal market segment" with potential buyers where the willingness to pay for a house decreases about 20% if there is a railway within 100 m from the house, while the average decrease in willingness to pay across all respondents decreases 20% already at 400 m distance. Corresponding results hold for motorways and "large urban roads" (70 km/h, 2 lanes each direction). The results are used in an empirical example to value a significant increase in railway traffic through a village. Our results are of similar magnitude with those obtained through noise valuation methods (obtained through an independent hedonic price study) and also with actual relative price changes over the last 30 years. These conclusions were also reached in an earlier study we conducted a few years earlier. Compared to that study, the design and estimation methods have been improved, making our new results more stable. This makes us inclined to conclude that stated preference methods are a viable way to value intrusion effects in cost-benefit analyses in many cases. The method is, however, limited to measuring intrusion effects in residential areas, and is not applicable to intrusion if e.g. parks and forests. For the covering abstract please see ITRD E135207.
Samenvatting