A first comparison of the innovation processes of introducing electronic identities on a national level in Austria, Belgium, Germany and Spain, based on extensive expert interviews with key actors, has been amended by four more country reports from Denmark, Finland, Estonia and Sweden in order to check the validity of generalisations derived from the first four cases. The extended comparison with the four additional countries increases the variance between the eID systems in Europe by showing differing technical and organisational features, such as purely softwarebased solutions, e.g. in Denmark, or complete outsourcing of the eIDMs, e.g. in Sweden. In the second part of the paper, the conceptual framework of the comparative study, a combination of path analysis, institutional actor theory and policy field analysis will be reflected. It has resulted in a fruitful approach allowing for the explanation of some, but by no means all, of the differences between the national eIDMs in Europe.

This chapter provides an analysis of the long process of introducing an electronic identity for online authentication in Germany. This process is described as a multi-facet innovation, involving actors from different policy fields shifting over time. The eID process started in the late '90s in the context of eGovernment and eCommerce with the legislation on e-signatures, which were supposed to allow for online authentication of citizens. When after 5 years it was recognized that this was not the case, a new digital ID card, which had meanwhile been announced, was chosen as token for the eID. This process was dominated by the concerns for visual inspection and border control, including the storage of digital fingerprints. Under the leadership of the Ministry of the Interior (BMI) and technical guidance of the Federal Agency for Information Security (BSI), technical specifications have to a large extent been adopted from the electronic passport, which had been smoothly introduced 2 years before. However, in the legislative process some concern regarding digital fingerprints on the eID card was raised and led to an opt-in solution. In 2009, a bill on the new ID card was passed which regulates the eID function for online authentication as well. This is characterized as a radical innovation by introducing a double-sided, mutual authentication of the citizen and the service provider and implementing the principle of proportionality regarding the access of service providers to data on the chip.

This paper compares the four national electronic Identity Management Systems (eIDMS), which have been described in the previous chapters. The section "Similarities and differences between four national eIDMS" will highlight the differences between these systems conceived as socio-technical systems with regard to the eID itself, the eID cards as tokens, the authentication processes as well as the procedures for distribution and personalisation, the support provided for installing the technology and any provider-related regulation. The section "A three-fold path dependency", according to the conceptual framework presented in the introductory chapter to this special issue, compares the new electronic systems with the previous ones in each country, in order to assess the continuation or changes with regard to the organisational, technological and regulatory path of development. The following sections explain the differences between the paths chosen and the path-related changes by analysing the actor constellation of the institutional actors, in particular the policy field and the power structure, as well as the context in which the policy makers made their choices, looking at privacy and "Staatsverständnis" in particular. Finally the diffusion and usage of the eID function will be compared and analysed, discussing to what extent the new institution has made a contribution to solving the policy problem it was developed for, e.g. providing a stronger authentication in order to meet security concerns regarding e-government and e-commerce transactions and avoiding new privacy infringements. Using grounded theory, the explanations provided have the status of generalisations derived from the four cases. They have to be considered as hypotheses, which will be checked for other countries in the following papers of this special issue. The comparison of the four cases in this article shows a high degree of path dependency. Most of the differences between the new systems are just a continuation of differences between the previous systems although they are to solve the same problem and can draw on the same technologies. But most IDIS (2010) 3:111-153