Dwelling Age Heteroskedasticity in Repeat Sales House Price Equations

Several authors have attributed the heteroskedasticity observed in repeat sales house price equations to the length of time between sales. Recently, Goodman and Thibodeau (1995) developed a theoretical model that relates heteroskedasticity in hedonic house price equations to dwelling age. Using data for nearly 2,000 repeat sales in Dallas, Texas, this research examines whether repeat sales heteroskedasticity is related to dwelling age, to the length of time between sales, or to both. An iterative generalized least squares procedure that explicitly models the residual variance is used to obtain robust parameter estimates and to increase the efficiency of the usual repeat sales price indices.     

Repeat Sales as a Matching Estimator

The most common approaches for constructing house price indices—hedonic price functions and the repeat sales estimator—focus on changes over time in mean prices. Though the hedonic approach is less wasteful of data than the repeat sales estimator, it relies on an accurate specification of the underlying econometric model. I suggest using a matching estimator as an alternative to the hedonic and repeat sales approaches. Like the repeat sales approach, a matching estimator uses pairs of sales from different dates to estimate the mean difference in sales prices over time. The matching approach preserves much larger sample sizes than the repeat sales estimator while requiring less preimposed structure than the hedonic approach. The matching approach makes it easy to characterize changes in the full distribution of house prices.     

On Choosing Among House Price Index Methodologies

This paper compares housing price indices estimated using three models with several sets of property transaction data. The commonly used hedonic price model suffers from potential specification bias and inefficiency, while the weighted repeat-sales model presents potentially more serious bias and inefficiency problems. A hybrid model combining hedonic and repeat-sales equations avoids most of these sources of bias and inefficiency. This paper evaluates the performance of each type of model using a particularly rich local housing market database. The results, though ambiguous, appear to confirm the problems with the repeat sales model but suggest that systematic differences between repeat-transacting and single-transacting properties lead to bias in the hedonic and hybrid models as well.     

Variation Across Price Segments and Locations: A Comprehensive Quantile Regression Analysis of the Sydney Housing Market

Standard house price indices measure average movements of average houses in average locations belonging to an average price segment and hence obscure spatial and cross‐sectional variation of price appreciation rates even within a single metropolitan area. This article combines penalized quantile regression techniques with the hedonic imputation approach to reveal such kind of variation. The method is applied to house transactions from Sydney between 2001 and 2014. The analysis finds significant variation across sub‐markets over time and in particular during the boom‐and‐bust cycle peaking in 2004. Appreciation rates were highest for suburban, low‐priced and lowest for inner‐city, high‐priced houses.     

The Spatial Proximity of Metropolitan Area Housing Submarkets

An important question related to housing submarket construction is whether geographic areas must be spatially adjacent in order to be considered the same submarket. Housing consumers do not necessarily limit their search to spatially concentrated areas and may search similarly priced neighborhoods located throughout a metropolitan area when making housing consumption decisions. This article examines two alternative procedures for delineating submarkets: one that combines adjacent census block groups into areas with enough transactions to estimate the parameters of a hedonic house price equation and a second that permits spatial discontinuities in submarkets. The criterion used to evaluate the alternative techniques is the accuracy of hedonic house price predictions.     

The Economics of Style: Measuring the Price Effect of Neo‐Traditional Architecture in Housing

Vintage effects have received considerable attention from economists in the context of house prices. Although strongly related, the impact of architectural building styles on prices has not been studied yet. Using a cross‐sectional hedonic price analysis including building styles of recently developed homes in the Netherlands we find a significant price premium for housing with neo‐traditional architecture. Extensive intervention by local authorities on the supply side of the housing market seems the most probable explanation of this effect. The decreasing price premium over time reflects the impact of supply restrictions on price, but also indicates that style does matter.     

Predicting Spatial Patterns of House Prices Using LPR and Bayesian Smoothing

This article is motivated by the limited ability of standard hedonic price equations to deal with spatial variation in house prices. Spatial patterns of house prices can be viewed as the sum of many causal factors: Access to the central business district is associated with a house price gradient; access to decentralized employment subcenters causes more localized changes in house prices; in addition, neighborhood amenities (and disamenities) can cause house prices to change rapidly over relatively short distances. Spatial prediction (e.g., for an automated valuation system) requires models that can deal with all of these sources of spatial variation. We propose to accommodate these factors using a standard hedonic framework but incoporating a semiparametric model with structure in the residuals modeled with a partially Bayesian approach. The Bayesian framework enables us to provide complete inference in the form of a posterior distribution for each model parameter. Our model allows prediction at sampled or unsampled locations as well as prediction interval estimates. The nonparametric part of our model allows sufficient flexibility to find substantial spatial variation in house values. The parameters of the kriging model provide further insights into spatial patterns. Out–of–sample mean squared error and related statistics validate the proposed methods and justify their use for spatial prediction of house values.     

Do House Price Levels Anticipate Subsequent Price Changes within Metropolitan Areas?

This research examines the relationship between hedonically controlled housing price levels and subsequent changes in those prices across locations within MSAs. Are areas with a high price relative to an “imputed rent” paying for higher appreciation? In an efficient market (e.g., Gordon Growth Model), as fundamentals (impute rent) differ across locations and change over time, anticipation of these should generate a positive correlation between (residual) price levels and subsequent price changes. We undertake these tests in four different MSAs using a panel of repeat‐sale house price indices that have been scaled to price levels with the hedonic attributes of the house and ZIP code. In three markets we find that identical houses in higher priced ZIP codes subsequently appreciate faster. In one market we find that there is little statistical difference.     

The Gender Gap in Real Estate Sales: Negotiation Skill or Agent Selection?

This study examines differences in net selling price for residential real estate across male and female agents. A sample of 2,020 home sales transactions from Fulton County, Georgia, are analyzed in a two‐stage least squares, geospatial autoregressive corrected, semi‐log hedonic model to test for gender and gender selection effects. Although agent gender seems to play a role in naïve models, its role becomes inconclusive as variables controlling for possible price and time on market expectations of the buyers and sellers are introduced to the models. Clear differences in real estate sales prices, time on market and agent incomes across genders are unlikely due to differences in negotiation performance between genders or the mix of genders in a two‐agent negotiation. The evidence suggests an interesting alternative to agent performance: that buyers and sellers with different reservation price and time on market expectations, such as those selling foreclosure homes, tend to select agents along gender lines.     

Marking Single–Family Property Values to Market

This paper is the text of the 2001 Presidential Address for the American Real Estate and Urban Economics Association. House price prediction accuracy is a function of: (1) the procedure used to identify within–metropolitan–area housing submarket boundaries, (2) the econometric technique used to estimate the parameters of the house price model and (3) the characteristics of the local housing market. This paper empirically examines several procedures for delineating within–metropolitan–area housing submarket boundaries. In addition, the paper examines the increase in prediction accuracy that can be achieved by employing spatial econometric techniques to spatially adjust ordinary least squares house price predictions. Finally, the paper examines the influence that local housing market characteristics have on house price prediction accuracy. Housing market characteristics examined here are: (1) structural characteristics of neighborhood properties, (2) heterogeneity of the neighborhood housing stock and (3) housing market liquidity. Prediction accuracy is examined using hedonic house price equations with over 40,000 single–family transactions for Dallas, Texas.     

Risk Segmentation of American Homes: Evidence from Denver

This article empirically examines the segmentation of house price risk across 99 ZIP‐code‐delineated neighborhoods in metropolitan Denver. The house price risk in each neighborhood is measured with the temporal variation of quarterly appreciation rates of the neighborhood house price index over the 2002–2007 period. Cross‐sectional regressions of neighborhood house price risk on the median household income and the percentage of population in poverty from the 2000 census data for the same neighborhoods provide strong evidence that the house price risk is significantly higher in low‐income/poor neighborhoods. Subperiod analyses further indicate that the risk segmentation exists in both a booming period (pre 2005:2) and a busting period (post 2005:3). The results indicate that homeownership can be a much riskier investment for low‐income/poor households.     

Modeling Spatial Variation in Housing Prices: A Variable Interaction Approach

The absolute location of each real estate parcel in an urban housing market has a unique location-value signature. Accessibility indices, distant gradients and locational dummies cannot fully account for the influence of absolute location on the market price of housing because there are an indeterminable number of externalities (local and nonlocal) influencing a given property at a given location. Furthermore, the degree to which externalities affect real estate values is not only unique at each location but highly variable over space. Hence, absolute location must be viewed as interactive with other determinants of housing value. We present an interactive variables approach and test its ability to explain price variations in an urban residential housing market. The statistical evidence suggests that the value of location, as embodied in the selling price of housing units, may not be separable from other determinants of value. It is recommended that housing valuation models, therefore, be specified to allow site, structural and other independent attributes to interact with absolute location—{ x , y } coordinates—when accounting for intraurban variation in the market price of residential housing. This approach is especially useful when estimating the value of housing for geographic areas where very little is known a priori about the neighborhoods or submarkets.     

Housing Attributes Associated with Capital Gain

Housing units are heterogeneous goods. Rates of change in housing prices are typically modelled as if they arise from factors unrelated to the housing unit itself. For example, housing price increases in the latter part of the 1970s and early 1980s are argued to have arisen primarily from demographic factors and the differential effects of inflation on the effective rate of taxation on income from corporate capital and on owner-occupied housing. Cross-sectional variation in price inflation is not addressed. Consumers who purchased housing units are not indifferent to their attributes. To the extent that expectations vary within regional housing markets as a consequence of variation in housing attributes, standard linear hedonic price regression may generate biased estimates of implicit prices. This paper identifies sufficient conditions for the estimates of implicit prices in linear hedonic price regressions to be unbiased and generate unbiased estimates of implied price changes. Finally, this paper identifies living space (house size) as a significant attribute related positively to the increase in individual housing prices in a regional market.     

On the Nature of Self‐Assessed House Prices

In models of optimal household behavior, the value of housing affects consumption, savings and other variables. But homeowners do not know the value of their house for certain until they sell, so while they live in their home they must rely on local house price data to estimate its value. This article uses data from the recent housing boom and bust to demonstrate that changes in households' self‐assessed home values are strongly consistent with the predictions of a model in which households optimally filter available house price data. Specifically, we show that self‐assessed house prices did not increase as rapidly as house price indexes during the boom and did not decline as severely during the bust. A Kalman filter model nearly perfectly replicates these data. These findings have direct implications for economists studying asking prices during booms and busts, optimal default decisions and other key housing‐related phenomena.     

Housing Regulation,Externalities and Residential Property Prices

This article examines the effects of quantity restrictions on residential property prices in the presence of neighborhood externalities. A Brigham Young University policy limiting students’ location choices provides a natural experiment for studying the externality and quantity restriction effects on property values. A flexible hedonic model is used to control for nonstudent population spatial sorting by type. The estimates show significant positive quantity restriction and student agglomeration effects on student housing prices. There are also significant differences in the negative student externality across nonstudent neighborhoods, with the quantity restriction reinforcing (offsetting) the student price premium (discount) at the boundary.     

A Spatiotemporal Autoregressive Price Index for the Paris Office Property Market

This article applies the spatiotemporal hedonic approach to the analysis of office transaction prices in the Paris property market ( i.e ., central Paris and its inner suburbs). The analysis focuses primarily on the market's two main business districts (the Central Business District and the La Défense District). We find that spatial and temporal dependence effects are strongly present in these submarkets. Additionally, we propose a hybrid method for incorporating a temporal regime switch into the spatiotemporal autoregressive model. The regime switching around 1997 ( i.e. , in the presence of temporal heterogeneity) substantially affects the significance of spatial and temporal dependences. Finally, we build a new price index that incorporates both spatiotemporal dependences and temporal heterogeneity. This index differs strongly from the usual hedonic price index.     

Can Tightness in the Housing Market Help Predict Subsequent Home Price Appreciation? Evidence from the United States and the Netherlands

This article assesses the predictive power of variables that measure market tightness, such as seller's bargaining power and sale probabilities, on future home prices. Theoretical insights from a stylized search‐and‐matching model illustrate that such indicators can be associated with subsequent home price appreciation. The empirical analysis employs listings data on residential units offered for sale through a real estate broker in the Netherlands and for certain U.S. regions. Individual records are used to construct quarterly home price indices, an index that measures seller's bargaining power and (quality‐adjusted) home sale probabilities. Using conventional time‐series models we show that current sale probabilities and bargaining power can significantly reduce home price appreciation forecast errors and help to predict turning points in local area housing markets. The measures and approaches in this article help to demonstrate ways in which researchers and practitioners can leverage listings data to gain knowledge about the current and future state of the housing market.     

Transaction-Based Office Price Indexes: A Spatiotemporal Modeling Approach

This study examines the potential of a two-order spatiotemporal autoregressive model with a Bayesian heteroskedasticity robust procedure in modeling strata-titled Singapore office unit transaction prices and in constructing transaction-based disaggregate office price indexes. The model reduces the problems caused by the infrequent trading of individual commercial properties. However, for those office properties that are located outside the CBD and also for those less frequently transacted, the power of the model in capturing these particular office buildings' price dynamics is limited. The significant differences of the office prices across the various office buildings and submarkets show that the model can capture the variation in office prices and track the timing of capital gains and losses that investors may accrue on spatially distributed office properties more accurately than hedonic or weighted least squares estimates.     

An Empirical Examination of Traditional Neighborhood Development

This study analyzes the impact of the new urbanism on single-family home prices. Specifically, we explore the price differential that homebuyers pay for houses in new urbanist developments relative to houses in conventional suburban developments. Using data on over 5,000 single-family home sales from 1994 to 1997 in three different neighborhoods, hedonic regression results reveal that consumers pay more for homes in new urbanist communities than those in conventional suburban developments. Further analyses indicate that the price premium is not attributable to differences in improvement age and other housing characteristics.