Spatial Econometric Analysis Kuan-Pin Lin Portland State University Spatial Econometric Models Spatial Exogenous Model Spatial Lag Model Spatial Mixed Model Spatial Error Model Spatial AR(1) Spatial MA(1) Spatial ARMA(1,1) Spatial.
Download ReportTranscript Spatial Econometric Analysis Kuan-Pin Lin Portland State University Spatial Econometric Models Spatial Exogenous Model Spatial Lag Model Spatial Mixed Model Spatial Error Model Spatial AR(1) Spatial MA(1) Spatial ARMA(1,1) Spatial.
Spatial Econometric Analysis 2 Kuan-Pin Lin Portland State University Spatial Econometric Models Spatial Exogenous Model Spatial Lag Model Spatial Mixed Model Spatial Error Model Spatial AR(1) Spatial MA(1) Spatial ARMA(1,1) Spatial Error Components Model Spatial Exogenous Model Lagged Explanatory Variables The Model y Xβ WXγ ε E (ε | X, W ) 0 2 I n ' wij x j Var (ε | X, W ) E (εε ') WX j 1 i 1, 2,..., n Spatial Lag Model Lagged Dependent Variable The Model y Wy Xβ ε E (ε | X, W ) 0 wij y j Wy j 1 i 1, 2,..., n 2 I Var (ε | X, W ) E (εε ') n (I W )y Xβ ε y (I W ) 1 Xβ (I W ) 1 ε Var (y ) 2 [(I W ) '(I W )]1 Cov(Wy, ε) 2W (I W ) 1 0 Spatial Mixed Model The Model y Wy Xβ WXγ ε 2 I E (ε | X, W ) 0, Var (ε | X, W ) E (εε ') (I W )y Xβ WXγ ε y (I W ) 1 ( Xβ WXγ ) (I W ) 1 ε Var (y ) 2 [(I W ) '(I W )]1 Cov(Wy, ε) 2W (I W ) 1 0 Spatial Error Models ε Wε υ Spatial AR(1) ε Wυ υ Spatial MA(1) Spatial ARMA(1,1) ε Wε Wυ υ E (υ | X,W ) 0 Var (υ | X,W ) E (υυ ') 2I Spatial Error Components Model The Model ε Wψ υ E (ε) 0 Var (ε) WW ' I 2 2 E (ψ) E (υ) 0, Cov(ψ, υ) 0 E (ψψ ') I, E (υυ ') I 2 2 Spatial Econometric Models The General Model: SARAR(1,1) Allowing spatial weights matrix to be different in the regression and in the error. The special case is W = M. y Wy Xβ WXγ ε ε Mε Mυ υ E (υ | X,W , M ) 0 Var (υ | X,W , M ) E(υυ ') I 2 Spatial Model Specification Tests Moran Test Moran’s I Test Statistic Asymptotic Theory Bootstrap Method LM Test and Robust LM Test Spatial Error Model Spatial Lag Model Hypothesis Testing The Basic Model y Xβ ε ε Wε υ or ε Wυ υ E (υ | X, W) 0 Var (υ | X, W) 2I υ ~ normal iid (0,2I) H 0 : 0 or 0 H1 : (not H 0 ) Moran-Based Test Statistics Moran’s I Index εˆ 'Wεˆ εˆ 'Wεˆ I ~ normal iid ( E ( I ),V ( I )) 2 εˆ ' εˆ nˆ εˆ y Xβˆ βˆ ( X ' X) X ' y Can not distinguish between spatial lag or spatial error trace( MW ) E(I ) , where M I X( X ' X) 1 X nK trace(MWMW ' ) trace[( MW )2 ] [trace( MW )]2 V (I ) E ( I )2 (n K )(n K 2) LM-Based Test Statistics LM Test Statistic for Spatial Error 2 1 εˆ Wεˆ LM Error 2 ~ 2 (1) b ˆ ' yˆ Xβˆ , ˆ 2 εˆ 'εˆ / n b trace(WW W 'W ) Can not distinguish between spatial AR or spatial MA LM-Based Test Statistics LM Test Statistic for Spatial Lag 2 1 εˆ Wy 2 LM Lag ~ (1) 2 ˆ b ˆ ' (Wyˆ )' M (Wyˆ ) ˆ ˆ 2 b trace(WW W 'W ) LM-Based Test Statistics Robust LM Test Statistic for Spatial Error 2 εˆ Wεˆ b εˆ Wy 1 1 2 LM Error 2 ~ (1) 2 ˆ ˆ b ˆ ˆ b ' ' * Robust LM Test Statistic for Spatial Lag 1 LM Lag ˆ * 2 2 εˆ Wyˆ 1 εˆ Wy εˆ Wεˆ 2 ~ (1) 2 2 2 ˆ ˆ ˆ ˆ ' ' ' LM-Based Test Statistics Joint LM Test for Spatial Correlation (Spatial Lag and Spatial Error) 2 2 1 εˆ Wεˆ 1 εˆ Wyˆ LM 2 2 ~ 2 (2) b ˆ ˆ ˆ ( LM Error ) ( LM Lag * ) ' ' ( LM Lag ) ( LM Error * ) Hypothesis Testing Example Crime Equation (Crime Rate) = a + b (Family Income) + g (Housing Value) + (numbers in parentheses are p-values of the tests) Moran-I LM-err LM-lag Crime Rate 5.6753 (0.000) 26.902 (0.000) 26.902 (0.000) Family Income 4.6624 (0.000) 17.841 (0.000) 17.841 (0.000) Housing Value 2.1529 (0.031) 3.3727 (0.066) 3.3727 (0.066) 2.954 (0.003) 5.723 (0.017) 9.363 (0.002) Robust LM-err Robust LM-lag Hetero. 0.0795 (0.778) 3.72 (0.054) 1.058 (0.589) References L. Anselin, and A. K. Bera, R. J.G.M. Florax, and M. Yoon (1996), “Simple Diagnostic Tests for Spatial Dependence,” Regional Science and Urban Economics, 26, 77-104. L. Anselin, and H. Kelejian (1997), “Testing for Spatial Autocorrelation in the Presence of Endogenous Regressors,” International Regional Science Review, 20, 153–182. L. Anselin, and S. Rey (1991), “Properties of Tests for Spatial Dependence in Linear Regression Models,” Geographical Analysis, 23, 112-131. H. Kelejian, and I.R. Prucha (2001)., “On the Asymptotic Distribution of Moran I Test Statistic with Applications,” Journal Econometrics, 104, 219-257.