Recently, there has become more insight into the determinants of housing prices. However,most research almost exclusively focuses on economic and financial conditions. Spatial factors are only implicitly included, e.g. in new housing supply. Moreover, there is no theory that is capable to explain the changing impact of different conditions over time and especially over periods longer than a few decades.   

An important suggestion of such a theory has been made by Knoll et al. (2014), but is almost certain that their idea cannot be right. When analyzing home prices adjusted for cpi in fourteen countries since 1870, they found a clear hockey-stock pattern. Their explanation is that a reduction in transport costs triggered an expansion of developed land, suppressing land prices. They point to the development of railway infrastructure as well as efficiency gains in maritime transportation. Their theory is creative, but also amazing as it will be mainly relevant for agricultural land. Urban land supply is also determined by the availability of transport, but in a totally different way. To gain insight into this problem, we will use a model proposed by Manning (1984). He shows that there have been different phases in urban development, each phase based on different transport solutions and costs. Growth of urban areas can become constrained by the capacity of its transport systems. After the introduction of a transport innovation land supply suddenly improves, resulting in increased housing production and lower prices. 

We will try to show the basic soundness of this theory by analyzing developments in several urban areas in different countries. Due to lack of data, we are unable to produce formal tests, but there are clear differences between cities that can be related to developments in their transport infrastructure. In our view, this approach is also able to explain rising house prices after 1870, but also the period of relatively low prices after 1910. This alternation of rising and falling prices is overlooked by Knoll et al.  Another conclusion of our research is that it is important to study long-term house price developments on a regional scale, in order to understand the impact of local and regional spatial developments.