Investments in rail are usually expected to improve accessibility for individuals. The value of the additional accessibility is often measured by the impact on housing prices nearby. This paper analyses the City Tunnel in Southern Sweden, a major railway investment that resulted in two new train stations in 2010: the Triangeln and Hyllie train stations. A station may affect housing prices both directly and indirectly. Direct effects are the transport-related benefits (e.g., improved accessibility), while indirect effects are additional non-transportation investments that the new station attracted to the area. The direct and indirect effects generate together the total effect on housing prices. To accurately estimate the value of the accessibility provided by the new stations it is necessary to disentangle the two effects.
In this paper, a hedonic difference-in-differences model is used to estimate the total and the direct effects of the train stations on housing prices. The main results indicate that (i) the total effect on housing prices within 1,000 m from the Hyllie and Triangeln train stations shows a price increase of 17% and 8%, respectively. Yet, (ii) when removing part of the indirect effects the estimates fall to 7.2%-10.4% for Hyllie and to 5.6%-6.2% for Triangeln. The results show that: (i) the value of the accessibility may be misspecified if only the total effect on housing prices is estimated, and (ii) the value of train stations may increase if combined with other investments.