Please use this identifier to cite or link to this item: https://hdl.handle.net/10419/205206 
Year of Publication: 
2019
Series/Report no.: 
30th European Conference of the International Telecommunications Society (ITS): "Towards a Connected and Automated Society", Helsinki, Finland, 16th-19th June, 2019
Publisher: 
International Telecommunications Society (ITS), Calgary
Abstract: 
Digital technologies sometimes create digital divides. One of the remedies for certain divides in Europe is the creation of the Digital Single Market, of which e-commerce is one of the main elements. The focus of this work is e-commerce in Spain. The current study improves substantially on existing international literature by using a large and representative panel data set on individuals, with 133,420 observations for the period 2008-2016. Moreover, it uses economic models that incorporate previously omitted variables, and employs a variety of panel techniques. This paper starts bymeasuring digital divides and their evolution along time. Next, a model that incorporates previously neglected explanatory variables, such as income and digital skills, is formulated. Individual demand models are estimated using panel logistic regression techniques. This allows quantifying the impact of each of the socioeconomic and geographic characteristics on the adoption of the service. The resulting models have high explanatory power. Newly incorporated variables like income and digital skills are highly significant. Age, education, gender and geographical variables are also significant. The results also allow novel regional comparisons. Policy recommendations are derived, suggesting effective and affordable measures targeted at specific socio-demographic groups.
Subjects: 
e-commerce
digital divide
TIC-H survey
individual annual panel data
logistic regression
dynamic model
Heckman selection model
Document Type: 
Conference Paper

Files in This Item:
File
Size





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