A comment on “Specification searches in spatial econometrics: The relevance of Hendry's methodology”

Florax et al. [Florax, R.J.G.M., Folmer, H., Rey, S.J., 2003. Specification searches in spatial econometrics: the relevance of Hendry's methodology. Regional Science and Urban Economics, 33, 557–579] undertook a simulation study designed to assess the properties of various selection strategies applied to a spatial modelling problem. Unfortunately, a serious flaw in their experimental design vitiates their main conclusions. A proposal for how to rectify the problem suggests that the opposite conclusion might well result, particularly for more interesting settings than the simple one they considered.