A Primer for Spatial Econometrics With Applications in R, STATA and Python

Foreword to the First Edition
Preface to the Second Edition
Preface and Acknowledgments to the First Edition
Contents
List of Figures
1 The Classical Linear Regression Model
	1.1 The Basic Linear Regression Model
	1.2 Non-sphericity of the Disturbances
	1.3 Endogeneity
	1.4 Computer Codes
		1.4.1 Running a Regression with R
		1.4.2 Running a Regression with STATA
		1.4.3 Running a Regression with Python
	References
2 Some Important Spatial Definitions
	2.1 The Spatial Weight Matrix W and the Definition of Spatial Lag
	2.2 Testing Spatial Autocorrelation Among OLS Residuals Without an Explicit Alternative Hypothesis
	2.3 Computer Codes: R
		2.3.1 Creating and Managing W Matrices
		2.3.2 Calculating Moran’s I Spatial Correlation
		2.3.3 Some Useful Spatial R Databases
	2.4 Computer Codes: STATA
		2.4.1 Creating and Managing W Matrices
		2.4.2 Calculating Moran’s I Spatial Correlation
		2.4.3 Some Useful Databases to Be Used in STATA
	2.5 Computer Codes: Python
		2.5.1 Creating and Managing W Matrices
		2.5.2 Calculating Moran’s I Spatial Correlation
		2.5.3 Some Useful PySAL Databases
	References
3 Spatial Linear Regression Models
	3.1 Generalities
	3.2 Pure Spatial Autoregression
	3.3 The Spatial Lag of X Model (SLX)
	3.4 The Spatial Error Model (SEM)
		3.4.1 Introduction
		3.4.2 Maximum Likelihood Estimator
		3.4.3 Feasible GLS
	3.5 The Spatial Lag Model (SLM)
		3.5.1 Generalities
		3.5.2 Maximum Likelihood Estimator
		3.5.3 Two Stages Least Squares Estimators
	3.6 The Spatial Durbin Model
	3.7 The General SARAR(1,1) Model
		3.7.1 Generalities
		3.7.2 Maximum Likelihood Estimator
		3.7.3 The Generalized Spatial Two Stages Least Squares (GS2SLS)
		3.7.4 The Lee Fully Efficient Estimators
	3.8 Testing Spatial Autocorrelation Among the Residuals with an Explicit Alternative Hypothesis
		3.8.1 Testing Spatial Autocorrelation Among the Residuals Using SEM or SLM as Alternatives
		3.8.2 Testing Spatial Autocorrelation Among the Residuals Using a Spatial Models as an Alternative: The Modified Moran’s I Test
	3.9 Interpretation of the Parameters in Spatial Econometric Models
	3.10 Estimating Linear Spatial Models with R
	3.11 Estimating Linear Spatial Models with STATA
	3.12 Estimating Linear Spatial Models with Python
	References to the Chapter
4 Further Topics in Spatial Econometrics
	4.1 Heteroscedastic Innovations
		4.1.1 Generalities
		4.1.2 The SARAR Model with Heteroscedastic Disturbances
		4.1.3 Spatial HAC Estimators
	4.2 Spatial Models for Binary Response Variables
		4.2.1 Introduction
		4.2.2 The A-Spatial Logit and Probit Models
		4.2.3 The Spatial Logit and Probit Models
		4.2.4 The Spatial Tobit Model
		4.2.5 Further Spatial Discrete Choice Models
	4.3 Spatial Panel Data Models (Written by Giovanni Millo)
		4.3.1 Generalities
		4.3.2 Unobserved Heterogeneity and Individual Effects
		4.3.3 Spatial Panel Models with Random Effects
		4.3.4 Spatial Panel Models with Fixed Effects
		4.3.5 Estimation
		4.3.6 Further Modeling Frameworks in Spatial Panel Data Modeling
	4.4 Non-stationary Spatial Econometric Models
		4.4.1 Generalities
		4.4.2 Geographically Weighted Regression
		4.4.3 Further Developments
	4.5 Bayesian Spatial Models
	4.6 Stochastic and Endogenous Weighting Matrices
		4.6.1 Stochastic Weight Matrices
		4.6.2 Endogenous Weight Matrices
	4.7 Computer Codes
		4.7.1 Estimating Heteroscedastic Linear Spatial Models with R
		4.7.2 Estimating Heteroscedastic Linear Spatial Models with STATA
		4.7.3 Estimating Heteroscedastic Linear Spatial Models with Python
		4.7.4 Estimating Spatial Probit/Logit/Tobit Models with R
		4.7.5 Estimating Spatial Probit/Logit/Tobit Models with STATA
		4.7.6 Estimating Spatial Panel Models with R
		4.7.7 Estimating Spatial Panel Models with STATA
		4.7.8 Estimating Spatial Panel Models with Python
		4.7.9 Estimating Geographically Weighted Regression Models with R
		4.7.10 Estimating Geographically Weighted Regression Models with STATA
		4.7.11 Estimating Geographically Weighted Regression Models with Python
		4.7.12 Estimating Bayesian Spatial Econometric Models with R
	References to the Chapter
5 Alternative Model Specifications for Big Datasets
	5.1 Introduction: Spatial Econometrics and the Big Data Revolution
	5.2 The MESS Specification
		5.2.1 A MESS Spatial Lag Specification
		5.2.2 A MESS Spatial Error Specification and Further Extensions
	5.3 The Unilateral Approximation Approach
		5.3.1 The Importance of Asymmetries and Anisotropies in Spatial Econometrics
		5.3.2 Testing Isotropy in Spatial Lag Models
		5.3.3 Inference for a Unilateral Spatial Lag Model
	5.4 A Composite Likelihood Approach
		5.4.1 Generalities
		5.4.2 A Bivariate Marginal Likelihood Approach to Spatial Error Model Estimation
	5.5 Handling the Second “V” of Big Data (Velocity) in Spatial Regressions
	5.6 R Codes
	References
6 Conclusions: What’s Next?
	References
Index