Depreciation of dwellings

In this study we use data for the sale of used dwellings to estimate how the values of dwellings depend on various qualitative features, including the age of the dwellings. By comparing the value of dwellings that are similar according to a variety of characteristics, we can estimate how important age is for the price. The sales values, which are the modelled variables in our analyses, consist of two components, namely the value of the physical housing structure and the housing unit's share of the value of the land on which it is built. It would have been desirable to be able to decompose the sales value in these two components, but given the data we have had available for this study this has not been possible. While the physical structure may be assumed to depreciate over time, this is not the case for the land value. Therefore, an estimation error occurs because one cannot distinguish the land value from the physical housing structure. We have tried to correct this problem by including the land area as one of the explanatory variables. Although the dataset is rich in many ways, it also has some weaknesses. When it comes to renovation, only a dummy variable is available that tells whether modernisation has occurred or not. We do not know exactly when the modernisation was done or how extensive it was. The econometric analyses estimate the "net depreciation rate" to be between 0.1 and 0.9 percent of the building value per year. The net depreciation rate is equal to a gross structure depreciation rate less an average renovation or modernisation appreciation rate. Based on the Norwegian national accounts, we estimate the renovation or modernisation rate to 1.8 per cent in 2014. By summing the net depreciation rate and the renovation rate, we derive an estimate of the gross depreciation rate amounting to between 1.9 and 2.7 per cent per year. This is in line with what is assumed in the national accounts calculations in most OECD countries, including Norway.