S03-S1 The Impact of Earthquakes on Regional Housing Markets and Regional Economic Development
Tracks
Special Sessions
| Thursday, August 31, 2017 |
| 2:00 PM - 3:30 PM |
| AB A8 (0008) |
Details
Conveners: Roderik Ponds, Harry Garretsen, Gerard Marlet / Chair: Harry Garretsen
Speaker
Mr Nicolas Duran
Phd Candidate
University Of Groningen
A study on the effect of earthquakes on housing prices in the north of the Netherlands
Author(s) - Presenters are indicated with (p)
Nicolas Duran (p), Paul Elhorst
Discussant for this paper
Harry Garretsen
Abstract
The largest natural gas field in Europe is located in the north of the Netherlands. Extraction from it has induced many small earthquakes in the region due to the soil subsidence it has elicited.
We estimate the effect that these have had on housing prices employing detailed data on nearly 250.000 transaction spanning between 1993 and 2014 for the three most northern provinces of the Netherlands.
We fit hedonic models for housing prices controlling for 43 house and neighborhood characteristics. Additionally, we control for discrete sub-market heterogeneity (neighborhood) by means of neighborhood fixed effects, as well as for weak and strong cross-sectional dependence typically present in housing market data.
Spatial effects (weak cross-sectional dependence) are accounted for employing weight matrices that lag observations in space and time while weights them according to house-characteristics similarity. Common factors (strong cross sectional dependence), in turn, are accounted for employing average housing prices at the national, provincial and city levels.
Additionally, we account for the endogenous effects of both, declining population in the region and the time the house has spent on the market, by means of instrumental variables. Finally, we propose a new measurement for the impact of earthquakes improving on earlier approaches, as well as we deliver a readily available method to compute compensations to households for declining house values due to these gas-extraction-induced earthquakes.
We find a strong and significant negative effect associated to the discrete sub-market level (neighborhood). Employing the average peak ground velocity hitting houses previously sold in the same neighborhood, we determine that an increase of 1% in the neighborhood's historical PGV average reduces the house price in around 4% , ceteris paribus. Additionally, we find significant spatial effects in the form of global spillovers from prices at which similar houses were previously sold, which increase the overall earthquakes' effect. On average, these spillovers add a 0.5% marginal reduction to the house price.
We estimate the effect that these have had on housing prices employing detailed data on nearly 250.000 transaction spanning between 1993 and 2014 for the three most northern provinces of the Netherlands.
We fit hedonic models for housing prices controlling for 43 house and neighborhood characteristics. Additionally, we control for discrete sub-market heterogeneity (neighborhood) by means of neighborhood fixed effects, as well as for weak and strong cross-sectional dependence typically present in housing market data.
Spatial effects (weak cross-sectional dependence) are accounted for employing weight matrices that lag observations in space and time while weights them according to house-characteristics similarity. Common factors (strong cross sectional dependence), in turn, are accounted for employing average housing prices at the national, provincial and city levels.
Additionally, we account for the endogenous effects of both, declining population in the region and the time the house has spent on the market, by means of instrumental variables. Finally, we propose a new measurement for the impact of earthquakes improving on earlier approaches, as well as we deliver a readily available method to compute compensations to households for declining house values due to these gas-extraction-induced earthquakes.
We find a strong and significant negative effect associated to the discrete sub-market level (neighborhood). Employing the average peak ground velocity hitting houses previously sold in the same neighborhood, we determine that an increase of 1% in the neighborhood's historical PGV average reduces the house price in around 4% , ceteris paribus. Additionally, we find significant spatial effects in the form of global spillovers from prices at which similar houses were previously sold, which increase the overall earthquakes' effect. On average, these spillovers add a 0.5% marginal reduction to the house price.
Prof. Paul Elhorst
Full Professor
Rijksuniversiteit Groningen
Testing for common factors and cross-sectional dependence among individual housing prices
Author(s) - Presenters are indicated with (p)
Paul Elhorst, Nicolás Duran (p)
Discussant for this paper
Roderik Ponds
Abstract
This paper extends the cross-sectional dependence (CD) test of Pesaran (2004, 2015a) and the exponent of cross-sectional dependence test of Bailey et al. (2015) such that they can also be calculated based on individual data. The modified tests are applied on 163,323 housing transactions that took place in the provinces of Groningen, Friesland and Drenthe located in the Netherlands over the period 2003-2014.
Prof. Harry Garretsen
Professor of International Economics & Business
University of Groningen
