Continuous-Time Modeling with Spatial Dependence

Tinbergen Institute Discussion Paper No. 11-117/3

27 Pages Posted: 11 Aug 2011 Last revised: 6 Jul 2012

See all articles by Johan H. L. Oud

Johan H. L. Oud

Radboud University Nijmegen

Henk Folmer

University of Groningen

Roberto Patuelli

University of Bologna - Department of Economics

Peter Nijkamp

VU University of Amsterdam - Department of Spatial Economics; Tinbergen Institute

Date Written: August 11, 2011

Abstract

(Spatial) panel data are routinely modelled in discrete time (DT). However, there are compelling arguments for continuous time (CT) modelling of (spatial) panel data. Particularly, most social processes evolve in CT, so that statistical analysis in DT is an oversimplification, gives an incomplete representation of reality and may lead to misinterpretation of estimation results. The most compelling reason for a CT approach is that, in contrast to DT modelling, it allows adequate modelling of dynamic adjustment processes. The paper introduces spatial dependence in a CT modelling framework. We propose a nonlinear Structural Equation Model (SEM) with latent variables for estimation of the Exact Discrete Model (EDM), which links the CT model parameters to the DT observations. The use of a SEM with latent variables makes it possible to take measurement errors in the variables into account, leading to a reduction of attenuation bias (i.e., disattenuation). The SE M-CT model with spatial dependence developed here is the first dynamic structural equation model with spatial dependence. The spatial econometric SEM-CT framework is illustrated on the basis of a simple regional labor market model for Germany made up of the endogenous state variables unemployment change and population change and of the exogenous input variables change in regional average wage and change in the structure of the manufacturing sector.

Keywords: continuous-time modelling, structural equation modelling, latent variables, spatial dependence, panel data, disattenuation, measurement errors, unemployment change, population change, Germany

JEL Classification: C33, E24, O18, R11

Suggested Citation

Oud, Johan H. L. and Folmer, Henk and Patuelli, Roberto and Nijkamp, Peter, Continuous-Time Modeling with Spatial Dependence (August 11, 2011). Tinbergen Institute Discussion Paper No. 11-117/3, Available at SSRN: https://ssrn.com/abstract=1908147 or http://dx.doi.org/10.2139/ssrn.1908147

Johan H. L. Oud (Contact Author)

Radboud University Nijmegen ( email )

Postbus 9108
Nijmegen, 6500 HK
Netherlands

Henk Folmer

University of Groningen ( email )

P.O. Box 800
Groningen, 9700AV
Netherlands
+31 503633897 (Phone)

Roberto Patuelli

University of Bologna - Department of Economics ( email )

via Anghera' 22
Rimini, 47921
Italy
+39-0541-434276 (Phone)
+39-02-700419665 (Fax)

HOME PAGE: http://https://sites.google.com/site/patuellihomepage/

Peter Nijkamp

VU University of Amsterdam - Department of Spatial Economics ( email )

De Boelelaan 1105
1081HV Amsterdam
Netherlands
+31 20 4446091 (Phone)
+31 20 4445611 (Fax)

Tinbergen Institute

Gustav Mahlerplein 117
Amsterdam, 1082 MS
Netherlands

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
64
Abstract Views
820
Rank
623,067
PlumX Metrics