Journal of Real Estate Finance and Economics, 29:2, 167±191, 2004
# 2004 Kluwer Academic Publishers. Manufactured in The Netherlands.
Modeling Spatial and Temporal House Price Patterns:
A Comparison of Four Models
Board of Governors of the Federal Reserve System, Washington, DC 20551
Centre for Real Estate, University of Connecticut, Storrs, CN 06269-1041, U.S.A.
Department of Economics, Weatherhead School of Management, Case Western Reserve University,
Cleveland, OH 44106
TCU Box 298530, Fort Worth, TX 76129
This research reports results from a competition on modeling spatial and temporal components of house prices. A
large, well-documented database was prepared and made available to anyone wishing to join the competition. To
prevent data snooping, out-of-sample observations were withheld; they were deposited with one individual who
did not enter the competition, but had the responsibility of calculating out-of-sample statistics for results
submitted by the others. The competition turned into a cooperative effort, resulting in enhancements to previous
methods including: a localized version of Dubin's kriging model, a kriging version of Clapp's local regression
model, and a local application of Case's earlier work on dividing a geographic housing market into districts. The
results indicate the importance of nearest neighbor transactions for out-of-sample predictions: spatial trend
analysis and census tract variables do not perform nearly as well as neighboring residuals.
Key Words: kriging, out-of-sample prediction, data snooping, local polynomial regression, smoothing
regression, semiparametric models, cluster analysis, nearest neighbors, hedonic models
It is well known that real estate values are highly dependent on their locational and market
characteristics. Over the past decade, technological advances have led to a signi®cant
increase in the amount of available data with spatial attributes that allow investigators to
more readily account for these important factors in their pricing models.
Indeed, over the
past several years, researchers have made great strides accounting for spatial and temporal
factors in real estate pricing models.
However, it is not at all obvious how best to account
for space and time in empirical investigations. This research contributes to the existing