Production technology and agglomeration for Japanese prefectures during 1991–2000

We analyse the seemingly unrelated regression (SUR) models with spatial dependencies from a Bayesian point of view and estimate the parameters of the models using a Markov chain Monte Carlo (MCMC) method. Further, we analyse the production technology and the economics of agglomeration in Japanese prefectures from 1991 to 2000, simultaneously taking into account spatial and serial correlation. Model comparison is done via log-marginal likelihoods, and it is found that the spatial error SUR model is the best model and that the economics of agglomeration and spatial heterogeneity decreased over this decade. Resumen. Analizamos modelos de ecuaciones aparentemente no relacionadas (SUR, siglas en ingles) con dependencias espaciales desde un punto de vista bayesiano y estimamos los parametros de los modelos utilizando el metodo de Monte Carlo basado en cadenas de Markov (MCMC). Ademas, analizamos la tecnologia de la produccion y las economias de aglomeracion en prefecturas japonesas desde 1991 a 2000, teniendo en cuenta simultaneamente la correlacion espacial y serial. La comparacion de modelos se realiza mediante verosimilitud log-marginal, y se encontro que el error espacial del modelo SUR es el mejor modelo y que las economias de aglomeracion y la heterogeneidad espacial disminuyeron durante esta decada.

[1]  Hajime Wago,et al.  Small-sample Properties of Panel Spatial Autoregressive Models: Comparison of the Bayesian and Maximum Likelihood Methods , 2008 .

[2]  S. Chib,et al.  Analysis of multivariate probit models , 1998 .

[3]  S. Chib Marginal Likelihood from the Gibbs Output , 1995 .

[4]  David F. Percy,et al.  Prediction for Seemingly Unrelated Regressions , 1992 .

[5]  L. Tierney Markov Chains for Exploring Posterior Distributions , 1994 .

[6]  P. Speckman,et al.  Posterior distribution of hierarchical models using CAR(1) distributions , 1999 .

[7]  A. Gelman,et al.  Weak convergence and optimal scaling of random walk Metropolis algorithms , 1997 .

[8]  Bhavani Shankar,et al.  Bayesian Spatial Probit Estimation: A primer and an application to HYV rice adoption , 2002 .

[9]  Y. Kanemoto,et al.  AGGLOMERATION ECONOMIES AND A TEST FOR OPTIMAL CITY SIZES IN JAPAN , 1996 .

[10]  A. Zellner An Efficient Method of Estimating Seemingly Unrelated Regressions and Tests for Aggregation Bias , 1962 .

[11]  S. Chib,et al.  Marginal Likelihood From the Metropolis–Hastings Output , 2001 .

[12]  Siddhartha Chib,et al.  Hierarchical analysis of SUR models with extensions to correlated serial errors and time-varying parameter models☆ , 1995 .

[13]  Adrian F. M. Smith,et al.  Sampling-Based Approaches to Calculating Marginal Densities , 1990 .

[14]  Shuangzhe Liu,et al.  Spatial system estimators for panel models: A sensitivity and simulation study , 2014, Math. Comput. Simul..

[15]  Wolfgang Polasek,et al.  Spatial interaction of crime incidents in Japan , 2008, Math. Comput. Simul..