House prices and relative location

Location is known to be a main determinant in people's efforts to estimate the value of a house. However, the type of location, and subsequently how the value of that location is estimated, has not been investigated to the same degree. In hedonic price modelling, a well-used method of estimating housing as a composite good, locational proxies such as postcodes or census tracts are often used to control for location. This notion of location is what geographers refer to as absolute location, a fixed position in space, as opposed to relative location, a position in space relative to other positions. In this paper, we look at the difference in explanatory power of absolute versus relative location in a hedonic model for apartment sales, using the city of Oslo as our case. The main finding is that the added explanatory power of postcode dummies significantly diminishes when introducing relative location explanatory variables such as walking distance to key places like the metro and parks. As house prices correlate with consumer preferences, these findings will have implications for urban planning insofar as different neighborhood designs vary with respect to their ability to harvest relative location potentials.