Despite attempts to get around the Jacobian in fitting spatial econometric models by using GMM and other approximations, it remains a central problem for maximum likelihood estimation. In principle, and for smaller data sets, the use of the eigenvalues of the spatial weights matrix provides a...