Journal of Real Estate Finance and Economics, Vol. 17:1, 61±85 (1998)
# 1998 Kluwer Academic Publishers, Boston. Manufactured in The Netherlands.
Analysis of Spatial Autocorrelation in House Prices
Statistics Department, Southern Methodist University, Dallas, TX 75275
THOMAS G. THIBODEAU
E. L. Cox School of Business, SMU, Dallas, TX 75275
This article examines spatial autocorrelation in transaction prices of single-family properties in Dallas, Texas. The
empirical analysis is conducted using a semilog hedonic house price equation and a spherical autocorrelation
function with data for over 5000 transactions of homes sold between 1991:4 and 1993:1. Properties are geocoded
and assigned to separate housing submarkets within metropolitan Dallas. Hedonic and spherical autocorrelation
parameters are estimated separately for each submarket using estimated generalized least squares (EGLS). We
®nd strong evidence of spatial autocorrelation in transaction prices within submarkets. Results for spatially
autocorrelated residuals are mixed. In four of eight submarkets, there is evidence of spatial autocorrelation in the
hedonic residuals for single-family properties located within a 1200 meter radius. In two submarkets, the hedonic
residuals are spatially autocorrelated throughout the submarket, while the hedonic residuals are spatially
uncorrelated in the remaining two submarkets. Finally, we compare OLS and kriged EGLS predicted values for
properties sold during 1993:1. Kriged EGLS predictions are more accurate than OLS in six of eight submarkets,
while OLS has smaller prediction errors in submarkets where the residuals are spatially uncorrelated and the
estimated semivariogram has a large variance.
Key Words: hedonic house prices, spatial autocorrelation, semivariogram
House prices are spatially autocorrelated for two reasons. First, neighborhoods tend to be
developed at the same time, so neighborhood properties have similar structural
characteristics such as dwelling size, vintage, interior and exterior design features and
so on. Second, neighborhood residential properties share location amenities. For example,
the same police and ®re departments protect area residents, and neighborhood children
have access to the same public schools.
Hedonic house price equations attempt to explain variation in house prices using
property structural and location characteristics. The residuals produced by these equations
are frequently spatially correlated. While most structural characteristics are relatively easy
to measure and are typically included in publicly available data, location characteristics
such as distance to CBD, distance to major transportation arteries, quality of public
schools, crime rates, proximity to nonconforming land uses, and so on are more dif®cult to
measure and are rarely included in publicly available data.
Identifying relevant neighborhood boundaries within metropolitan areas is also dif®cult.
Obvious neighborhood boundaries include municipal and school district boundaries and
geographic features like highways, major roads, rivers, and parks. Other neighborhood