Patuelli / Arbia Spatial Econometric Interaction Modelling

1. Auflage 2016
ISBN: 978-3-319-30196-9
Verlag: Springer Nature Switzerland
Format: PDF
Kopierschutz: 1 - PDF Watermark

E-Book, Englisch, 466 Seiten

Reihe: Advances in Spatial Science

This contributed volume applies spatial and space-time econometric methods to spatial interaction modeling. The first part of the book addresses general cutting-edge methodological questions in spatial econometric interaction modeling, which concern aspects such as coefficient interpretation, constrained estimation, and scale effects. The second part deals with technical solutions to particular estimation issues, such as intraregional flows, Bayesian PPML and VAR estimation. The final part presents a number of empirical applications, ranging from interregional tourism competition and domestic trade to space-time migration modeling and residential relocation.

Roberto Patuelli is Associate Professor for Economic Policy in the Department of Economics, University of Bologna, Italy. He holds an M.A. from George Mason University, USA, and a Ph.D. from Vrije Universiteit Amsterdam, Netherlands. Giuseppe Arbia (Ph.D. Cantab) is Full Professor of Economic Statistics in the Department of Statistical Sciences, Università Cattolica del Sacro Cuore Roma, Italy; Lecturer in Statistics at Università della Svizzera Italian, Lugano (CH), and Research Professor at the College of William & Mary, Wiliamsburg, Virginia (US).

Autoren/Hrsg.

Patuelli, Roberto

Arbia, Giuseppe

Inhaltsverzeichnis

1;Preface;6
2;Contents;8
3;Contributors;10
4;1 Spatial Econometric Interaction Modelling: Where Spatial Econometrics and Spatial Interaction Modelling Meet;13
4.1;1.1 The Spatial Interaction Model: An Established Regional Economics Workhorse;13
4.2;1.2 Spatial Interaction and Flow Dependence;15
4.3;1.3 Towards a New Class of Spatial Interaction Models;16
4.4;1.4 The Structure of the Book;19
4.5;References;22
5;Part I General Methodological Issues;25
5.1;2 Spatial Regression-Based Model Specifications for Exogenous and Endogenous Spatial Interaction;26
5.1.1;2.1 Introduction;26
5.1.2;2.2 The Log-Normal (Independent) Spatial Interaction Model;28
5.1.3;2.3 Exogenous Versus Endogenous Spatial Interaction Specifications;30
5.1.3.1;2.3.1 An Endogenous Spatial Interaction Specification;30
5.1.3.1.1;2.3.1.1 A Theoretical Motivation for Endogenous Interaction;31
5.1.3.2;2.3.2 An Exogenous Spatial Interaction Specification;33
5.1.4;2.4 Interpreting Estimates From Spatial Interaction Specifications;34
5.1.4.1;2.4.1 Interpreting Estimates from Non-Spatial Interaction Specifications;35
5.1.4.2;2.4.2 Interpreting Estimates from Exogenous Interaction Specifications;38
5.1.4.3;2.4.3 Interpreting Estimates from Endogenous Interaction Specifications;39
5.1.5;2.5 An Applied Illustration Involving Movement of Teachers Between School Districts;41
5.1.6;2.6 Conclusion;46
5.1.7;References;47
5.2;3 Constrained Variants of the Gravity Model and Spatial Dependence: Model Specification and Estimation Issues;48
5.2.1;3.1 Introduction;48
5.2.2;3.2 Unconstrained and Constrained Classes of Gravity Models: The Classical View;51
5.2.3;3.3 Poisson Versions of the Constrained Gravity Models;55
5.2.4;3.4 Accounting for Spatial Dependence in the Model Specifications;58
5.2.5;3.5 Equivalency Relationships Between the Balancing Factors, Fixed Effects, and Random Effects;61
5.2.5.1;3.5.1 Comparisons Between Balancing Factors and Fixed Effects;61
5.2.5.2;3.5.2 Comparisons Between Balancing Factors and Random Effects;63
5.2.6;3.6 An Illustrative Example;65
5.2.6.1;3.6.1 Model Specifications Ignoring Spatial Dependence in origin-destination Flows;66
5.2.6.2;3.6.2 Spatial Filter Model Specifications Accounting for Spatial Dependence in Flows;71
5.2.7;3.7 Concluding Remarks;75
5.2.8;Appendix: Results for the Estimation of Singly Constrained Random Effects Specifications;75
5.2.9;References;76
5.3;4 Testing Spatial Autocorrelation in Weighted Networks: The Modes Permutation Test;78
5.3.1;4.1 Introduction;78
5.3.2;4.2 Spatial Autocorrelation in Weighted Networks;79
5.3.2.1;4.2.1 Local Covariance and the Exchange Matrix;79
5.3.2.2;4.2.2 Spatial Filtering and Spatial Modes;80
5.3.2.3;4.2.3 The Modes Permutation Test;82
5.3.2.3.1;4.2.3.1 Heterodasticity of the Spatial Field;82
5.3.2.3.2;4.2.3.2 Homoscedaticity of the Spatial Modes;83
5.3.2.3.3;4.2.3.3 The Test;83
5.3.2.4;4.2.4 Illustration: Swiss Migratory and Linguistic Data;85
5.3.3;4.3 Adjacency Graphs and Accessibilities;87
5.3.3.1;4.3.1 Proposal A: Simple Rescaling with Diagonal Adjustment;87
5.3.3.2;4.3.2 Proposal B: Time-Embeddable Exchange Matrices;88
5.3.3.3;4.3.3 Illustration: The Distribution of Blood Group a in Eire;90
5.3.4;4.4 Conclusion;90
5.3.5;Appendix;91
5.3.6;References;92
5.4;5 Effects of Scale in Spatial Interaction Models;95
5.4.1;5.1 Introduction;95
5.4.2;5.2 Literature Review;97
5.4.3;5.3 A Formalization of MAUP in Spatial Gravity Models;97
5.4.3.1;5.3.1 Case a: Ô and D? Are Non Stochastic;99
5.4.3.2;5.3.2 Case b: Ô and D? Are Stochastic and Independent;100
5.4.3.3;5.3.3 Case c: Ô and D? Are Stochastic, Spatially Autocorrelated and Mutually Independent;101
5.4.3.4;5.3.4 Case d: Ô and D? Are Stochastic, Spatially Autocorrelated and Mutually Dependent;102
5.4.4;5.4 Analysis of Artificial Data;104
5.4.5;5.5 Conclusions and Research Priorities;110
5.4.6;References;111
6;Part II Specific Methodological Issues;112
6.1;6 Dealing with Intraregional Flows in Spatial Econometric Gravity Models;113
6.1.1;6.1 Introduction;113
6.1.2;6.2 The Spatial Econometric Gravity Model;114
6.1.3;6.3 Issues with Intraregional Flows;116
6.1.3.1;6.3.1 Intraregional Distance;116
6.1.3.2;6.3.2 Unobserved Intraregional Flow Data;117
6.1.3.2.1;6.3.2.1 The Heckit-Type Model;117
6.1.3.2.2;6.3.2.2 EM Algorithm Model;118
6.1.4;6.4 Empirical Application;119
6.1.4.1;6.4.1 Data;119
6.1.4.2;6.4.2 Results;120
6.1.5;6.5 Experimental Analysis of the Difference in Nature;123
6.1.6;6.6 Conclusion;125
6.1.7;References;126
6.2;7 A Bayesian Spatial Interaction Model Variant of the Poisson Pseudo-Maximum Likelihood Estimator;128
6.2.1;7.1 Advantages of PPML for Spatial Interaction Modeling;128
6.2.2;7.2 A Spatial Extension of the PPML Model;130
6.2.3;7.3 Estimating the PPML Model;133
6.2.4;7.4 An Application to Commuting Flows;137
6.2.4.1;7.4.1 Higher-Order Neighbors;144
6.2.5;7.5 Conclusion;148
6.2.6;References;149
6.3;8 The Space of Gravity: Spatially Filtered Estimation of a Gravity Model for Bilateral Trade;151
6.3.1;8.1 Introduction;151
6.3.2;8.2 The Gravity Model and Autocorrelation;153
6.3.2.1;8.2.1 Theoretical Gravity;154
6.3.2.2;8.2.2 Practical Gravity;155
6.3.2.3;8.2.3 Consistent Estimation and Autocorrelation;157
6.3.3;8.3 Recent Developments in Estimating the Theoretical Gravity Model of Trade;158
6.3.4;8.4 Proposed Methodology: Spatial Filtering Estimation;159
6.3.5;8.5 Spatial Autocorrelation Diagnostics;162
6.3.6;8.6 Empirical Application;164
6.3.6.1;8.6.1 Data and Model Specification;164
6.3.6.2;8.6.2 Estimation Results: Spatial Filtering and Benchmark Models;165
6.3.6.3;8.6.3 Testing for Spatial Autocorrelation;168
6.3.7;8.7 Conclusions;171
6.3.8;Appendix;172
6.3.9;References;173
6.4;9 A Spatial Interaction Model with Spatially Structured Origin and Destination Effects;176
6.4.1;9.1 Introduction;176
6.4.2;9.2 Empirical Modeling of Commodity Flows;179
6.4.2.1;9.2.1 Conventional Gravity Models;179
6.4.2.2;9.2.2 A Bayesian Hierarchical Gravity Model;181
6.4.2.3;9.2.3 Related Spatial Effects Models;184
6.4.2.4;9.2.4 MCMC Estimation of the Model;185
6.4.3;9.3 A Data Generated Example;188
6.4.4;9.4 An Application to Commodity Flows Between Spanish Regions;190
6.4.4.1;9.4.1 The Data;191
6.4.5;9.5 Conclusions;199
6.4.6;Appendix: Details Regarding the MCMC Sampler;199
6.4.7;References;201
6.5;10 Bayesian Variable Selection in a Large Vector Autoregression for Origin-Destination Traffic Flow Modelling;203
6.5.1;10.1 Introduction;203
6.5.1.1;10.1.1 A VAR Model for Traffic Flows;205
6.5.1.2;10.1.2 Bayesian VARs;205
6.5.1.3;10.1.3 Bayesian Variable Selection;206
6.5.2;10.2 Bayesian Variable Selection in VAR;207
6.5.2.1;10.2.1 The Model;207
6.5.2.2;10.2.2 The Latent Binary Selection Vector;211
6.5.2.3;10.2.3 The Priors;211
6.5.2.4;10.2.4 The Posterior;214
6.5.2.5;10.2.5 The Sampling Procedure;215
6.5.3;10.3 A Simulation Study;217
6.5.4;10.4 Concluding Remarks;222
6.5.5;Appendix 1;223
6.5.6;Appendix 2;225
6.5.7;References;226
6.6;11 Double Spatial Dependence in Gravity Models: Migration from the European Neighborhood to the European Union;228
6.6.1;11.1 Introduction: 150 Years of Gravity;228
6.6.2;11.2 Theory;232
6.6.2.1;11.2.1 Origins and Destinations;232
6.6.2.2;11.2.2 Spatial Gravity;234
6.6.2.3;11.2.3 Spatial Multilateralism;235
6.6.3;11.3 The Econometrics of Spatial Gravity Modeling;235
6.6.3.1;11.3.1 Double Spatial Lagged Dependent Variables;235
6.6.3.2;11.3.2 Spatial Autocorrelation;236
6.6.3.3;11.3.3 Spatially Robust Standard Errors for Gravity Models;236
6.6.3.4;11.3.4 Spatial Autoregressive Conditional Heteroskedasticity (SpARCH);237
6.6.4;11.4 The European Neighborhood;238
6.6.5;11.5 Immigration Theory;240
6.6.5.1;11.5.1 Stocks and Flows;243
6.6.5.2;11.5.2 Push and Pull Factors;244
6.6.6;11.6 Results;245
6.6.7;11.7 Conclusions;247
6.6.8;Appendix: Data Sources;248
6.6.9;References;253
6.7;12 Multilateral Resistance and the Euro Effects on Trade Flows;255
6.7.1;12.1 Introduction;255
6.7.2;12.2 Literature Review;257
6.7.3;12.3 Cross Sectionally Dependent Panel Gravity Models;259
6.7.4;12.4 Empirical Results;263
6.7.5;12.5 Conclusion;276
6.7.6;12.6 The Data Appendix;276
6.7.7;References;278
7;Part III Applications;281
7.1;13 The Effects of World Heritage Sites on Domestic Tourism: A Spatial Interaction Model for Italy;282
7.1.1;13.1 Introduction;282
7.1.2;13.2 Modelling Tourism Flows: Spatial Interaction Models;285
7.1.3;13.3 Cultural Heritage, WHS Endowment and Tourism: The Evidence;287
7.1.3.1;13.3.1 Cultural Heritage and Tourism;287
7.1.3.2;13.3.2 WHS Endowment and Tourism;288
7.1.4;13.4 Model and Estimation Strategy;291
7.1.4.1;13.4.1 Model;291
7.1.4.2;13.4.2 Estimation;296
7.1.4.3;13.4.3 Spatial Sensitivity Analysis: An Interpretative Framework;298
7.1.5;13.5 Empirical Application;301
7.1.5.1;13.5.1 Data;301
7.1.5.2;13.5.2 Model Results;303
7.1.5.3;13.5.3 Spatial Sensitivity Analysis;307
7.1.6;13.6 Conclusions;309
7.1.7;Appendix;311
7.1.8;References;313
7.2;14 Testing Transport Mode Cooperation and Competition Within a Country: A Spatial Econometrics Approach;317
7.2.1;14.1 Introduction;317
7.2.2;14.2 Trade, Transport Flows and Logistics;319
7.2.3;14.3 A Conceptual Framework for Competition and Cooperation Between Transport Modes;321
7.2.3.1;14.3.1 Transport Mode Competition;322
7.2.3.2;14.3.2 Transport-Mode Cooperation;322
7.2.4;14.4 The Empirical Strategy;324
7.2.4.1;14.4.1 Transport Mode Competition;327
7.2.4.1.1;14.4.1.1 A Basic Gravity Equation for Testing Transport Mode Competition;327
7.2.4.1.2;14.4.1.2 An Extended Gravity Equation with Cross-Sectional Dependence;328
7.2.4.2;14.4.2 Hub-Spoke Structures and Transit Flows;331
7.2.5;14.5 Data;334
7.2.6;14.6 Descriptive Analysis;338
7.2.7;14.7 Econometric Analysis;343
7.2.7.1;14.7.1 Aggregate Flows by Modes;343
7.2.7.2;14.7.2 Product-Specific Flows;350
7.2.8;14.8 Conclusions;354
7.2.9;Appendix;355
7.2.9.1;Comments on the Data;360
7.2.9.1.1;First Debugging Procedure for Transport Flows in Tons;361
7.2.9.1.2;Estimating Interregional Trade Prices;362
7.2.10;References;362
7.3;15 Modeling the Effect of Social-Network on Interregional Trade of Services: How Sensitive Are the Results to Alternative Measures of Social Linkages;365
7.3.1;15.1 Introduction;365
7.3.2;15.2 Trade and Social Networks: Background and Definitions;369
7.3.3;15.3 The Empirical Model;372
7.3.3.1;15.3.1 Spatial and Demographic Dependence Affecting Gravity-Model Estimates;373
7.3.3.2;15.3.2 Introducing Spatial and Network Effects in the Gravity Model;374
7.3.3.3;15.3.3 Alternative Spatial Models for Origin and Destination Flows;375
7.3.3.3.1;15.3.3.1 A Spatial Lag Model;376
7.3.3.3.2;15.3.3.2 A Spatial Error Model;377
7.3.3.3.3;15.3.3.3 A Spatial Durbin Model;378
7.3.3.3.4;15.3.3.4 A Spatial General Model;378
7.3.4;15.4 An Application to Spanish Domestic Trade in Certain Service Sectors;379
7.3.4.1;15.4.1 The Data;379
7.3.4.2;15.4.2 Alternative Measures for Capturing the Pro-Trade Effect of Social Networks;383
7.3.4.3;15.4.3 Estimation Results;388
7.3.4.4;15.4.4 Results Using Alternative Measures for Social Networks;406
7.3.5;15.5 Conclusions;408
7.3.6;Appendix;409
7.3.7;References;413
7.4;16 On the Mutual Dynamics of Interregional Gross Migration Flows in Space and Time;415
7.4.1;16.1 Introduction;415
7.4.2;16.2 Neoclassical Migration Model with a Spatial Dimension;416
7.4.3;16.3 Econometric Specification;419
7.4.4;16.4 Data and Regression Results;421
7.4.4.1;16.4.1 Variable Definition and Data Properties;421
7.4.4.2;16.4.2 Regression Results, Summary Measures and Multiplier Analysis;426
7.4.4.3;16.4.3 Robustness Checks;433
7.4.5;16.5 Conclusion;435
7.4.6;Variable Definitions;436
7.4.7;References;438
7.5;17 Residential Relocation in a Metropolitan Area: A Case Study of the Seoul Metropolitan Area, South Korea;440
7.5.1;17.1 Introduction;440
7.5.2;17.2 Literature Review;441
7.5.3;17.3 Study Area: The Seoul Metropolitan Area;443
7.5.4;17.4 Methodology;447
7.5.4.1;17.4.1 Spatial Interaction Models for Residential Relocation;447
7.5.4.2;17.4.2 Model Specification;449
7.5.5;17.5 Results;452
7.5.6;17.6 Conclusions;458
7.5.7;References;459
7.6;18 Conclusions: The Future of Spatial Interaction Modelling;463
7.6.1;18.1 A Reappraisal of the Presented Contributions;463
7.6.2;18.2 Future Roads of Spatial Interaction;464
7.6.3;References;465