This paper continues from the discussion of Florax et al. (Florax, R., H. Folmer and S. Rey, 2003. Specification searches in spatial econometrics: the relevance of Hendry's methodology. Regional Science and Urban Economics, 33, 557-579.), regarding the properties of various specification strategies for spatial econometric models. Habitual practise has popularised a technique based on the well-known Lagrange Multipliers, characterized as a Specific-to-General approach, and which seems to give good results. In our work, we contemplate other alternatives, some of which may be seen as slight variations of this proposal, including the selection tests of Vuong (Vuong, Q., 1989. Likelihood ratio-tests for model selection and non-nested hypotheses. Econometrica, 57, 307-333.) and of Clarke (Clarke, K., 2003. Nonparametric model discrimination in international relations. Journal of Conflict Resolutions, 47, 72-93.). We also examine an approach of the General-to-Specific type, as clearly opposite to the others. The comparison of the two strategies is carried out through a Monte Carlo experiment, the results of which are quite diffuse, in the sense that we do not find conclusive evidence in favour of either of these two approaches. However, it should be recognized that the General-to-Specific strategy seems to be more robust to the existence of anomalies in the Data Generating Process.
Download Info
To download:
If you experience problems downloading a file, check if you have the
proper application to
view it first. Information about this may be contained
in the File-Format links below. In case of further problems read
the IDEAS help
page. Note that these files are not on the IDEAS
site. Please be patient as the files may be large.
As the access to this document is restricted, you may want to look for a different version under "Related research" (further below) or search for a different version of it.
Volume (Year): 39 (2009) Issue (Month): 2 (March) Pages: 200-213 Download reference. The following formats are available: HTML
(with abstract),
plain text
(with abstract),
BibTeX,
RIS (EndNote, RefMan, ProCite),
ReDIF