Consistent Misspecification Testing in Spatial Autoregressive Models
58 Pages Posted: 3 Sep 2020
Date Written: August 27, 2020
Spatial autoregressive (SAR) and related models oﬀer flexible yet parsimonious ways to model spatial or network interaction. SAR specifications typically rely on a particular parametric functional form and an exogenous choice of the so-called spatial weight matrix with only limited guidance from theory in making these specifications. The choice of a SAR model over other alternatives, such as spatial Durbin (SD) or spatial lagged X (SLX) models, is often arbitrary, raising issues of potential specification error. To address such issues, this paper develops an omnibus specification test within the SAR framework that can detect general forms of misspecification including that of the spatial weight matrix, functional form and the model itself. The approach extends the framework of conditional moment testing of Bierens (1982, 1990) to the general spatial setting. We derive the asymptotic distribution of our test statistic under the null hypothesis of correct SAR specification and show consistency of the test. A Monte Carlo study is conducted to study ﬁnite sample performance of the test. An empirical illustration on the performance of our test in the modeling of tax competition in Finland and Switzerland is included.
Keywords: Conditional moment test, Misspecification test, Omnibus testing, Spatial AR, Weight matrix misspecification
JEL Classification: C21, C23
Suggested Citation: Suggested Citation