This research note examined the performance of Geographically Weighted Regression (GWR) using two calibration methods. The first method, Cross Validation (CV), has been commonly used in the applied literature using GWR. A second criterion selected an optimal bandwidth that corresponded with the smallest spatial error Lagrange Multiplier (LM) test statistic. We find that there is a tradeoff between addressing spatial autocorrelation and reducing degree of extreme coefficients in GWR. Although spatial autocorrelation can be controlled for by using the LM criterion, a substantial degree of extreme coefficients may remain. However, while the CV approach appears to be less prone to producing extreme coefficients, it may not always attend to the problems that arise in the presence of spatial error autocorrelation.

Three types of market segmentation strategies are available to estimate hedonic housing price models—i.e. no segmentation, segmentation by using statistical clustering methods and segmentation by using a priori information. This research tests the hypothesis of Tiebout theory that individual residential decision-making is determined by equilibrium provision of local public goods in accord with the tastes and preferences of residents, thereby sorting their housing locations into optimal sub-markets. Forecasting accuracies of eight sub-market segmentation strategies and two forecast-combining methods are examined by using housing sales data from Knox County, Tennessee, USA. The results provide empirical support for Tiebout theory of optimal housing sub-market location in that boundaries drawn using a priori information from local government jurisdictions, school districts and expert opinions are more closely aligned with the equilibrium provision of local public goods than boundaries drawn by statistical clustering methods. The advantage of forecast-combining is also demonstrated.

This research suggests that consumers' marginal willingness to pay for environmental landscape attributes, such as water view, developed open space and forest-land open space, decreased during the 2008 recession compared to the 2000–2006 real estate boom. Estimates were obtained from a spatial hedonic housing price model after controlling for household location patterns and structural differences between the periods. Because the decline in amenity values was probably due to a temporary deterioration in economic conditions, the amenity values will probably rebound with economic recovery. Thus, development decisions based on the lower estimated amenity values measured during a temporary deterioration in economic conditions may be determined suboptimal post-economic recovery.

This study applies a hedonic model to assess the economic benefits of air quality improvement following the 1990 Clean Air Act Amendment at the county level in the lower 48 United States. An instrumental variable approach that combines geographically weighted regression and spatial autoregression methods (GWR-SEM) is adopted to simultaneously account for spatial heterogeneity and spatial autocorrelation. SEM mitigates spatial dependency while GWR addresses spatial heterogeneity by allowing response coefficients to vary across observations. Positive amenity values of improved air quality are found in four major clusters: (1) in East Kentucky and most of Georgia around the Southern Appalachian area; (2) in a few counties in Illinois; (3) on the border of Oklahoma and Kansas, on the border of Kansas and Nebraska, and in east Texas; and (4) in a few counties in Montana. Clusters of significant positive amenity values may exist because of a combination of intense air pollution and consumer awareness of diminishing air quality.