Spatial Regression Analysis of Commercial Land Price Gradients July 2001 David Dale -Johnson* W. Jan Brzeski** Abstract Commercial land price gradients for an emerging real estate market are estimated using spatial regression techniques. Spatial statistics are used to explore the extent of spatial autocorrelation in the residuals of an OLS land price gradient model. Spatial autocorrelation is present but not to the same degree for all time periods or commercial land uses. Maximum likelihood estimates of land price gradients are as one would expect in mature real estate markets. Key Words: Spatial regression, spatial autocorrelation, land price gradients, emerging markets. To be presented at the Asian Real Estate Society Sixth Annual Conference at Keio University, Mita Campus, Tokyo, Japan, July 31 to August 2, 2001. *Associate Professor of Finance and Business Economics, Marshall School of Business and Lusk Center for Real Estate, 701 Hoffman Hall, University of Southern California, Los Angeles, California 90089-1427. Voice: (213) 740-6526 Facsimile: (213) 740-6650 E-mail: ddj@marshall.usc.edu. **Adjunct Professor, Jagiellonian Business School, Cracow, Poland and President of the Foundation, Cracow Real Estate Institute, 31-002 Cracow; ul. Kanonicza 7, Poland E-mail: jbrzeski@kin.cc.pl We are indebted to the Lincoln Institute of Land Policy, the Cracow Real Estate Institute and the USC Lusk Center for Real Estate for support of this research. Chris Redfearn was a valuable sounding board during the genesis of the paper. The authors are responsible any remaining errors that remain.