ABSTRACT
A major challenge in the analysis of micro-level spatial interaction is to distinguish actual interactions from the effects of spatially correlated omitted variables. We propose extending the simple spatially lagged explanatory (SLX) model to include two spatial weighting matrices at different spatial scales to reduce omitted-variable bias. The approach is suitable when actual interaction takes place on a smaller local level, while the omitted variables are spatially correlated at a larger regional level and correlated with the included characteristics. We provide an empirical motivation and use Monte Carlo simulation to illustrate the bias-reduction effects in certain settings.
ACKNOWLEDGEMENT
The authors are grateful to Grete Stokstad and Svein Olav Krogli from the Norwegian Institute of Bioeconomy Research (NIBIO, Ås, Norway) for providing the farm coordinates used in the analysis. They also thank two anonymous reviewers and the editor, Paul Elhorst, for valuable comments and suggestions that greatly strengthened the paper.
DISCLOSURE STATEMENT
No potential conflict of interest was reported by the authors.
ORCID
Hugo Storm http://orcid.org/0000-0003-3791-3615
Thomas Heckelei http://orcid.org/0000-0002-6251-3480
Notes
1 An appropriate instrument would be correlated to but independent of
.
2 In Norway, farms receive subsidies in form of various direct payments based on the number of animals and the area under production as well as the output produced. These subsidies account for a substantial amount of farm income.
3 The correlations are calculated as Pearson’s linear pairwise correlation coefficients between and
considering all observations, where
denotes column
of row
.
4 The application is based on the MATLAB® Statistics and Machine Learning Toolbox routine ‘fitctree’ and ‘cvLoss’.