Spatial Dependence and House Price Index Construction (original) (raw)
Related papers
Journal of Geographical Systems, 2007
Hedonic house price models typically impose a constant price structure on housing characteristics throughout an entire market area. However, there is increasing evidence that the marginal prices of many important attributes vary over space, especially within large markets. In this paper, we compare two approaches to examine spatial heterogeneity in housing attribute prices within the Tucson, Arizona housing market: the spatial expansion method and geographically weighted regression (GWR). Our results provide strong evidence that the marginal price of key housing characteristics varies over space. GWR outperforms the spatial expansion method in terms of explanatory power and predictive accuracy.
Spatial Correlation in Housing Prices Indexes
The housing price per square meter and its evolution in the last years is one of the issues that more worry Spanish citizens, and therefore, also to their political and economic authorities. Nevertheless and in spite of the importance that the space has in the real estate market, the averages obtained by the official authorities do not take into account the spatial correlation of the house prices. To solve this handicap we propose a new estimator, the kriging the mean, the best unbiased linear one, which provides more realistic estimates of the mean price of housing per square meter.
An Application of Spatial Econometrics in Relation to Hedonic House Price Modelling
2010
This paper applies spatial econometrics in relation to hedonic house price modeling. Some basic spatial model alternatives are used for a battery of relevant tests. Geographically-weighted regression, semiparametric analysis, and the mixed spatial Durbin model are also applied. The purpose is to detect missing spatial variables, misspecified functional form, and spatial heterogeneity in estimated parameters. Such misspecifications have been shown to give spurious results in relation to some frequently used directional based tests. Significant model improvement is achieved, so the paper should be of general interest as an example of practical econometric modeling within the field.
An alternative to the standard spatial econometric approaches in hedonic house price models
Hedonic models are subject to spatially correlated errors which are a symptom of omitted spatial variables, misspecification or mismeasurement. Methods have been developed to address this problem through the use of spatial econometrics or spatial fixed effects. However, often spatial correlation is modeled without much consideration of the theoretical implications of the chosen model or treated as a nuisance to be dealt with holding little interest of its own. We discuss the limitations of current standard spatial approaches and demonstrate, both empirically and theoretically the generalized additive model as an alternative. The generalized additive model is compared with the spatial error model and the fixed effects model. We find the generalized additive model to be a solid alternative to the standard approaches, having less restrictive assumptions about the omitted spatial processes while still being able to reduce the problem of spatial autocorrelation and provide trustworthy estimates of spatial variables. However, challenges connected with spatially varying data remain. The choice of flexibility in the spatial structure of the model affects estimated parameters of some spatially varying characteristics markedly. This suggests that omitted variable bias may remain an important problem. We advocate for an increased use of sensitivity analysis to determine robustness of estimates to different models of the (omitted) spatial processes.
Measuring the Importance of Location in House Price Appreciation
Journal of Urban Economics, 1996
This paper examines the variation in the rates of price appreciation within an individual metropolitan market. A methodology is developed to examine the Ž. locational variation in house price changes in Dade County Miami Florida, from 1971 to 1992. House price appreciation appears to be somewhat spatially related; that is, it varies by municipality, with distance from the CBD, with local changes in population and housing units, and by ethnic mix. However, these relationships have minimal explanatory power. Controlling for the census tract group location of each Ž. home explains only around 12% of the residual variation in the appreciation of individual homes that is not explained by metropolitan-wide changes in house prices. The effect of tract group location appears to be dominated by the idiosyncratic influences of individual homes and their immediate environments.
A methodology to compute regional housing price index using matching estimator methods
The Annals of Regional Science, 2011
This paper proposes a methodology for a spatial cost index of housing that considers spatial heterogeneity in properties across regions. The index is built by combining three different techniques to reduce the spatial heterogeneity in housing: Quasi-experimental methods, hedonic prices and Fisher spatial price index. Using microdata from the Chilean survey CASEN 2006, it is shown that the quasi-experimental method called Mahalanobis metric within propensity score calipers (MMWPS) leads to a significant reduction in the potential bias. The technique matches dwellings of a particular region with other properties of similar characteristics in the benchmark region (Metropolitan region). Once the houses are matched, a hedonic price model is computed, and a regional housing price matrix is created using Fisher spatial price indices. The paper concludes the existence of price differentials for homogeneous houses across regions in Chile.
Modeling spatial dimensions of housing prices in Milwaukee, WI
Environment and Planning B: Planning and Design, 2007
Introduction The hedonic housing price model is a powerful econometric tool for capturing important determinants of prices/housing values regarding structural and locational (neighborhood) attributes, and has been widely used in housing and urban studies. Serving as a``joint-envelope of a family of value equations (consumers' preferences) and another family of offer functions (suppliers' technologies)'' (Rosen, 1974, page 44), the hedonic model establishes a formal relationship between housing values/prices and a set of housing attributes (the quantity and qualities embodied in housing). Usually, housing attributes contain not only structural attributes such as floor size, but also locational or neighborhood conditions, such as proximity to certain public facilities. The model is appealing in that the implicit price of various housing attributes can be estimated from the model. Traditionally, the regression is calibrated through the ordinary least squares (OLS) estimator, under the general assumption of independent observation. However, despite the mature OLS technology and its wide application in examining the relationships between housing prices and attributes (for a review see Can, 1992), the full potential of the hedonic model remains to be exploited (Ekeland et al, 2004), and locational attributes in particular have drawn inadequate attention (Orford, 2002). During the late 1980s and early 1990s, largely due to the advancement in spatial statistics and spatial econometrics (
An Analysis of Spatial Dependence in Real Estate Prices
An Analysis of Spatial Dependence in Real Estate Prices, 2022
Real estate properties are naturally location-fixed. When space related factors are not fully incorporated in a standard pricing equation, spatial autocorrelation is likely to exist. This results in inefficiencies in estimations and raises the need for more complex spatial models. This paper analyzes the determinants of spatial dependence and evaluates the performance of the hedonic regression equation when the determinants of spatial dependence are controlled for. Using a novel dataset for a metropolitan housing market, we document the spatial clustering of housing characteristics such as area, total number of floors and the building age. We find support for the hypotheses that the construction process, shared social services and high-rise residential complexes cause spatial correlation. Our findings show that spatial correlation is significantly reduced when the factors of spatial dependence and district level data is controlled for in the standard hedonic regression.
Spatial Variation in Housing Prices: Econometric Analyses of Regional Housing Markets
2008
This thesis consists of six empirically-based papers. Collectively, the papers contribute to the understanding of the spatial variation in housing prices within regional housing markets. The main ambitions have been to identify what contributes significantly to explain the spatial variation in housing prices within such markets, and to account for this variation in econometric models. The thesis focuses primarily on macroscopical and general spatial structural characteristics, rather than on characteristics relevant for a specific region or a specific neighborhood. A summary of the theory underlying hedonic models and a review of the relevant literature are included, in addition to a chapter on various estimators from the spatial econometrics literature. According to the thesis there are two main global factors contributing significantly to explain the intraregional spatial variation in housing prices. These are the urban attraction effect measured by distance from the central business district (cbd), and labor market accessibility. The gravity-based labor market accessibility measure used in the thesis represents a useful approximation towards being able to study how changes in accessibility may manifest themselves and exert a spillover effect on housing prices throughout a region. Relevant kinds of experiments are performed in the thesis. As an example, these experiments show that an increase in the number of jobs in an urban area only marginally influences the spatial distribution of house prices. The impact on local housing prices is predicted to be considerably larger if the job growth is concentrated to the peripheral zones. The decentralization of jobs is hence found to contribute towards levelling out the differences in housing prices between the urban and peripheral zones. Even though the two globally-defined measures of spatial structure characteristics explain a major part of the spatial variation in housing prices, we also find that some locally-defined measures are relevant. The existence of subcenters, for instance, contributes significantly towards explaining spatial variation in housing prices. A model that includes both the urban attraction effect and labor market accessibility is shown to be useful for predictive purposes, particularly in relation to changes in the spatial distribution of jobs. In cases where one does not have detailed information on the spatial distribution of jobs, the cbd gradient captures both the urban attraction and the labor market accessibility effect. As an example, such a parsimonious model-formulation is demonstrated to offer reliable predictions of the variation in housing prices between a centre and the periphery. This conclusion might, however, be changed in a more polycentric area than the one studied here. I am grateful to Professor Roger Bivand at the Norwegian School of Economics and Business Administration. His competence in spatial statistics and willingness to spend time with me discussing statistical questions, programming in R, and even developing algorithms in spdep suitable for the type of data used in the thesis, has been important. I also wish to thank real estate agents in Haugesund for providing data and for taking part in discussions regarding some of the results. I have benefited from discussions with many good colleagues at Stord/Haugesund University College. I also want to thank Arnstein Gjestland and Gisle Kleppe for technical assistance. Paul Glenn has improved the English in various occasions. In general, it has been a pleasure to work with the librarians at the University College. They provide services beyond what could be expected. The deans at my department have been very accommodating, and any problem has been solved with ease. Stord/Haugesund University College has provided the necessary equipment, and has covered all my expenses during the project. More privately, I am grateful to my mother and to Arne for practical ground support. This has contributed to our everyday life being much easier and far more pleasant. My thoughts also go to my husband, and to our two children. With the help of incentives but without arguing, Hanne and Torbjørn have helped with cleaning our house before weekends. Tor, I highly appreciate the delicious dinners that are ready when I come home from work. Thank you for your patience and caring support.