Pasar al contenido principal

page search

Biblioteca SPATIAL RELATIONSHIPS IN RURAL LAND MARKETS WITH EMPHASIS ON A FLEXIBLE WEIGHTS MATRIX

SPATIAL RELATIONSHIPS IN RURAL LAND MARKETS WITH EMPHASIS ON A FLEXIBLE WEIGHTS MATRIX

SPATIAL RELATIONSHIPS IN RURAL LAND MARKETS WITH EMPHASIS ON A FLEXIBLE WEIGHTS MATRIX

Resource information

Date of publication
Diciembre 2002
Resource Language
ISBN / Resource ID
AGRIS:US2016210082

Empirical research has shown that location and economic development are important factors in rural market. With more and more rural land being converted at the urban fringe, buyer, sellers, planners, appraisers, tax assessors, and others are expected to have an increasing need for information related to the effect of location on rural land values.In econometric land value estimation, misspecification of the variance-covariance matrix results in loss of efficiency, predictive accuracy, and biased inference. The form of the weight matrix is also important. Additional flexibility can be introduced by including a decay parameter, along with different number of neighbors to provide improved estimates. The general objective of this paper is to develop an empirical rural land value model. Specifically, this study emphasizes the use of a more flexible spatial weight matrix by including a decay parameter along with different number of neighbors.Data for this study are based on rural land market sales from the southeast area of Louisiana that were collected using mail survey techniques for the period January 1993 through June 1998. Hedonic models are estimated by SAR using two types of weight matrices: the Delaunay, a rigid form of a weight matrix, and a flexible nearest neighbor asymmetric matrix that includes a decay parameter that lies between 0.4 and 1, along with different number of neighbors m ranging from 6 to 30. Likelihood ratio tests are used to test for statistical fit between spatial and OLS models. Results also suggest that the SAR model based on the Delaunay matrix and nearest neighbor matrix performed better than the OLS model. We also found that this particular set of data, one cannot conclude that the flexible nearest neighbor matrix outperforms the use of a more rigid spatial weight matrix. Further research should continue to test for other forms of spatial weight matrices.

Share on RLBI navigator
NO

Authors and Publishers

Author(s), editor(s), contributor(s)

Soto, Patricia
Vandeveer, Lonnie R.
Henning, Steven A.

Publisher(s)
Data Provider