Please use this identifier to cite or link to this item: http://hdl.handle.net/10419/177552
Authors: 
de Carvalho, Alexandre Ywata
de Mendonça, Mário Jorge Cardoso
da Silva, José Jaime
Year of Publication: 
2017
Series/Report no.: 
Texto para Discussão 2336
Abstract (Translated): 
The aim of this study is to evaluate the effect of broadband on the GDP taking into account regional differences. This is done by using the econometric model of panel data based on database of 5564 municipalities from 2007 to 2014. In order to incorporate the regional differences we use information derived from cluster analysis that isolates homogeneous groups of municipalities. Simultaneous causality between broadband and GDP is identified through instrumental variable method. Regional differentiated effects of broadband among municipalities were observed. The greatest impact of broadband occurs on those municipalities with the highest income per capita and high urban concentration followed by ones with a strong concentration in the service, commerce and construction sectors. The impact of broadband is also significant in the municipalities linked to the dynamic agricultural sector. In poorer municipalities the effect of broadband also occurs although with less intensity. Finally, we test the hypothesis that the effect of broadband on the output has an "inverted U" form that indicates that the effect rises to a certain level of access, but decreases after it is exceeded. The econometric results do not indicate that this hypothesis can be denied so that the impact is more intense in the intermediate levels of access of broadband.
Subjects: 
broadband
cluster analysis
K-means
panel data
instrumental variable
JEL: 
C21
O33
Document Type: 
Working Paper

Files in This Item:
File
Size
3.16 MB





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.