A troubled process: determining the depreciation in house prices as a result of the earthquakes in Groningen

How to make a good estimate of the depreciation of Groningen homes as a consequence of the earthquakes? Maarten Bosker, Professor of International trade and Economic development at Erasmus School of Economics, contributed to developing the model used to enable this depreciation to be determined. It proved to be no simple task.

Loppersum, Appingedam and Eemsmond: three municipalities seriously affected by the consequences of the extraction of natural gas that has generated over 400 billion euro for the Dutch government in recent decades. The Groningen gas field exploitation has resulted in the ground becoming unstable, and local residents have been confronted by earthquakes that resulted in considerable damage to homes.

Research on flood risks by Maarten Bosker formed the core of this depreciation scheme

The physical damage to homes as a consequence of the earthquakes is currently being entirely compensated by the oil and gas exploration & production company NAM and the Dutch government. However, as well as physical damage, residents in the earthquake zone are suffering from house price depreciation, as the earthquake risk has made their homes less attractive to buyers. How should this depreciation be calculated? The independent depreciation Advice Committee, established to determine the best method to compensate home owners, compared seven different models suggested by scientists, research institutes and consultancy agencies. Ultimately, just one model passed the test: that of Bosker and his co-researchers. This model was converted into policy by the government last year. Approximately 95,000 homes are eligible for compensation and so far 43 million euro has already been awarded to some 3,000 of the initial applicants.

Bosker states: ‘Our method is based on previous scientific research into the effect of flood risk on house prices, which we published in 2019 in the renowned scientific journal, the Journal of the European Economic Association. The difficulty in determining the depreciation due to earthquakes lies in the fact that you would ideally do this by comparing the value of two identical homes, one with and one without risk of earthquakes. Such a combination of homes is very hard, if not impossible, to find in practice. Moreover, the Arnhem-Leeuwarden court had determined that the compensation scheme must be readily implementable. It was explicitly indicated that a separate valuation of ten thousand homes, both in the earthquake zone and reference homes without risk outside this area, was not an option. The method we proposed uses the same statistical method that we developed in earlier research into assessing depreciation in flood risk areas, modified according to the situation in Groningen. The first step in this method was to determine the boundaries of the risk zone where homes would be eligible for the depreciation scheme. Ultimately, the best way to determine this proved to be using the percentage of awarded damage claims in the total housing stock of a postal code area: for all homes in the postal code areas in which this percentage was greater than 20%, a (substantial) decline due to the earthquakes cannot be excluded. All houses in these areas are now eligible for the depreciation scheme.’

Interestingly, in the Groningen postal code areas that had under 20% of the awarded damage claims, the value of homes had even increased or had fallen less significantly than in comparable regions. This is known as the waterbed effect; as many people would like to stay living in the area, they seek homes in the outlying areas where earthquake disruption is minimal.

The next step in the research was to determine the exact level of the depreciation, taking explicit account of the differences in earthquake problems that are found within the Groningen earthquake zone.

Bosker continues: ‘To determine the level of depreciation, we used so-called ‘matching’ and regression techniques. These statistical methods are used to select the most comparable homes outside the earthquake zone for each home in the established risk zone, using a very large number of property, local, and regional characteristics. This enabled us to determine the so-called ‘image effect’; the average depreciation of homes caused by their location in the Groningen earthquake zone. To do justice to the large differences in earthquake frequency and intensity in various parts of the earthquake zone, we then refined this result using the earthquake history at the site of each property. This ultimately resulted in depreciation valuations that are much higher in Loppersum or Appingedam than in, for example, Zuidhorn or Oldambt, two municipalities that have suffered less from the earthquakes.’

The independent Advice Committee established to determine the depreciation in house values ultimately chose the method of Bosker and his co-researchers primarily because of the model’s sound scientific basis. Another decisive factor was that the model easily accommodates the final scheme to take into account the statistical uncertainty surrounding the calculated depreciation. This is reflected in ‘assurance compensation’, an additional percentage of the house value paid on top of the carefully determined most likely depreciation in the postcode area in which the properties are situated. Using the model developed by Bosker and his co-researchers to calculate the depreciation leads, in the words of the Advice Committee, to an outcome that is just, generous, effective, and reasonable.

Bosker concludes his thoughts on the depreciation scheme: ‘My co-researchers and I are proud that our method was selected. Hopefully, the depreciation scheme will contribute to a successful, correct, and widely supported settlement of a drama that has unfolded over the past few decades.’

More information

You can read more about how the depreciation scheme was established here.

You can read more about the justification of the methodology by the Minister of Economic Affairs and Climate here.

Compare @count study programme

  • @title

    • Duration: @duration
Compare study programmes