Abstract:Spatial autoregressive (SAR) and related models offer flexible yet parsimonious ways to model spatial and network interactions. 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. Also, 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 a new specification test within the SAR framework that can detect general forms of misspecification including that of the spatial weight matrix, the functional form and the model itself. The test is robust to the presence of heteroskedasticity of unknown form in the disturbances and the approach relates to the conditional moment test framework of Bierens ([1982, Journal of Econometrics 20, 105–134], [1990, Econometrica 58, 1443–1458]). The Bierens test is shown to be inconsistent in general against spatial alternatives and the new test introduces modifications to achieve test consistency in the spatial setting. A central element is the infinite-dimensional endogeneity induced by spatial linkages. This complexity is addressed by introducing a new component to the omnibus test that captures the effects of potential spatial matrix misspecification. With this modification, the approach leads to a simple pivotal test procedure with standard critical values that is the first test in the literature to have power against misspecifications in the spatial linkages. We derive the asymptotic distribution of the test under the null hypothesis of correct SAR specification and prove consistency. A Monte Carlo study is conducted to study its finite sample performance. An empirical illustration on the performance of the test in modeling tax competition in Finland is included.

Robust Testing Procedures in Heteroscedastic Linear Models.

A robust specification test in linear panel data models

Robust testing in generalized linear models by sign flipping score contributions

A new class of robust two-sample Wald-type tests

A NEW PERSPECTIVE ON ROBUST M-ESTIMATION: FINITE SAMPLE THEORY AND APPLICATIONS TO DEPENDENCE-ADJUSTED MULTIPLE TESTING

Testing for Heteroscedasticity in Semiparametric Random Effect Model

A Non-Parametric Box-Cox Approach to Robustifying High-Dimensional Linear Hypothesis Testing

Heteroskedasticity-robust inference in linear regression models with many covariates

Consistent Specification Testing Via Nonparametric Series Regression

Detecting Neglected Nonlinearity in Dynamic Panel Data with Time-Varying Conditional Heteroskedasticity

Robust Hypothesis Testing with Wasserstein Uncertainty Sets

On the Robustness of Kernel Goodness-of-Fit Tests

Heteroskedasticity Robust Specification Testing In Spatial Autoregression

Robust tests for treatment comparisons based on recurrent event responses

Kernel Robust Hypothesis Testing

Testing Many Restrictions Under Heteroskedasticity

Robust Wald-Type Tests under Random Censoring with Applications to Clinical Trial Analyses

Robust Inference under the Beta Regression Model with Application to Health Care Studies

Testing Heteroskedasticity in High-Dimensional Linear Regression

Goodness-of-fit Testing in Linear Regression Models

Robust Hypothesis Testing Using Wasserstein Uncertainty Sets