In this paper we investigate the effects of new railway stations on house prices using an extensive repeated sales dataset over a period of 13 years. We employ semiparametric panel data techniques allowing for anticipation effects of station openings. We show that a kilometre reduction in distance to the nearest railway station increases property values by about 1.5 - 2 percent. The geographical range of the effect is about 3.5 kilometres. Ignoring anticipation effects in the estimation procedure leads to a large downward bias for short time period datasets. We do not find any significant house price adjustment effects after station openings.
House prices Anticipation effects Repeated sales Railway development Semiparametric