Geoadditive models for regional count data : an application to industrial location

Auteurs :
BASILE, R.
BENTRAFELLO, L.
CASTELLANI, D.

Description :
A geoadditive negative binomial model is proposed for regional count data that allows to address simultaneously some important methodological issues. It is applied to the study of location determinants of inward greenfield investments during 2003–2007 in 249 European regions. The empirical analysis shows the good performance of Geo-NB-GAM. The inclusion of a geoadditive component (a smooth spatial trend surface) permits to control for spatial unobserved heterogeneity that induces spatial clustering. Allowing for nonlinearities reveals that the positive effect of agglomeration economies fades as the density of economic activities reaches some threshold value.


Type de document :
Article de périodique

Source :
Geographical analysis, issn : 0016-7363, 2013, vol. 45, n°. 1, p. 28-48, nombre de pages : 21, Références bibliographiques : 3 p.

Date :
2013

Editeur :
Pays édition : Etats-Unis, Columbus, OH, Ohio State University Press