Keywords

1 Introduction

Across the globe, demands on public services are increasing at a fast pace, partly due to the widespread availability of new technologies and higher expectations from digitally-savvy citizens. Citizens expect personalized customer journeys at all levels of government, as they have become accustomed to smartphone-empowered lives  [1]. New digital data exchange infrastructure are essential for broader service access as well as the provision of significant benefits to service users at a reduced cost. However, even large budgets are no guarantee for successful digital government transformations  [2, 3]. From a purely technological perspective, all of this is hard to explain.

This paper argues that, in order to understand what contributes to the success of e-government, we also need to consider the institutional design as well as the design process of developing data-exchange infrastructure. Aiming to learn from successful examples, this paper conducts a comparative case study on two leading countries in e-government: the Netherlands and Estonia. Both the Netherlands and Estonia are in the group of high performers in the e-Government Development Index  [4], and both countries have widely adopted data exchange infrastructures  [5, 6], enabling for instance a pre-filled tax return form that takes minutes to electronically check and submit. For the sake of this paper, a data exchange infrastructure is defined as the whole of standards, technical components, services and governance framework in place for data exchange. These are by nature socio-technical constructs  [7, 8], which makes them hard to understand from a single point of view. Data exchange infrastructures are essential when it comes to the successful delivery of e-government services, since they facilitate process, application and data integration across the various government silos. Therefore, the maturity of data exchange infrastructure can be used as a proxy for e-government maturity.

While there is a growing body of knowledge on the technical design and the governance of data exchange infrastructures, we lack insights from a combined institutional, technical and process perspective. The goal of this paper is to analyse the development of data exchange infrastructures from such a combined design perspective. We conduct the comparative case study by adopting the theoretical framework for multi-facet design of socio-technical systems by Koppenjan and Groenewegen  [9]. Koppenjan and Groenwegen state: “Institutions concern different levels of analysis like laws and regulations as well as contracts and organisations which regulate and coordinate the behaviour of actors in complex networks”  [9]. Therefore, the framework is suitable for analysing complex socio-technical situations for policy making.

This paper proceeds as follows. Section 2 describes the research approach, which centres around a four day workshop in Estonia. Section 3 discusses the theoretical analysis framework of Koppenjan and Groenewegen that we use to compare data exchange infrastructure policies in the Netherlands and Estonia. Section 4 provides a high-level description of the selected cases the Netherlands and Estonia with respect to their e-government background. The findings of this paper are discussed in Sect. 5. Section 6 concludes the paper and provides recommendations for policy makers.

2 Research Approach

To analyse the development of data exchange infrastructures, we conduct a comparative case-study on the design of the data exchange infrastructures in the Netherlands and Estonia. The research draws on an international collaboration facilitated by DigicampusFootnote 1, a quadruple-helix-based innovation partnership for public service innovation in the Netherlands. One of the missions of Digicampus is to facilitate international collaboration on designing the next generation of public services. One of the vehicles for realizing this is an international collaboration agenda, focusing on current issues in e-government, learning from each other’s e-government agendas and pinpointing topics that are suited for collaborative research and prototyping. The first concrete research activity as part of this agenda was a four-day workshop in Tallinn from November 18 to 21, 2019. Table 1 provides an overview of this workshop.

Table 1. Overview of the workshop in Tallinn, Estonia.
Full size table

3 The Theoretical Analysis Framework

Koppenjan and Groenwegen  [9] introduce an analysis framework for a certain class of large-scale technological systems that do not consist merely out of technological assets, but involve institutions as part of their solutions. Institutions regulate behaviour and are essential components of socio-technical systems. Socio-technical systems are characterized by their complexity due to the many dependencies between the institutional and technology parts shaped by change processes. Koppenjan and Groenwegen suggested to analyze such system as a technological design that is teamed together with an institutional design  [10], see Fig. 1 (“co-design perspective”). Moreover, they suggest to make the design of the design process (process design) explicit in the analysis of such system (the design process in Fig. 1 is simply the process that yields the technological design and the institutional design, the process design is actually a ‘design process’ design). For our purposes, it is important to understand all designs (technological design, institutional design and process design) as continuously recurring endeavours that evolve and improve over time. Now, technological design is about “demarcation, components, relations, processes”  [9] of/in the technological system; institutional design is about “arrangements between actors that regulate their relations: tasks, responsibilities, allocation of costs, benefits and risks”  [9]; process design is about “who participates in the design process; what are the conditions, rules, roles, items, steps, etc.”  [9].

Fig. 1.
figure 1

A institutional design model by Koppenjan and Groenewegen  [9].

Full size image

In order to adequately grasp institutional design, Koppenjan and Groenewegen introduce a second perspective that adapts Williamson’s four-layer model of economics of institutions  [11], see again Fig. 1 (“institutional analysis perspective”). Layer 4 is the layer of the “informal institutional environment of socio-technical systems”, which is about “norms, values, orientation, codes (informal institutions, culture)”  [9], see also  [12], compare with  [13]. Layer 3 is the layer of the “formal institutional environment”  [9], which is about “formal rules, laws and regulations, constitutions (formal institutions)”  [9], see also  [14], compare with  [15]. Layer 2 is the layer of the “formal and informal institutional arrangements”  [9], which is about “gentleman agreements, covenants, contracts, alliances, joint-ventures, merges, etc.”  [9] and informal “rules, codes, norms, orientation, relations”  [9], see also  [16], compare with  [17]. Layer 1 is the layer of the “actors and games”  [9], which is about “actors/agents and their interactions aimed at creating and influencing (infrastructural) provisions, services, outcomes”  [9], compare with  [18]. The systems that Koppenjan and Groenwegen address are large-scale systems: “energy networks, water management services [...], waste treatment, transport systems (rail, road, water, tube), industrial networks, information systems and telecommunication networks, city service”  [9]. Information systems are among those systems, however, the model becomes actually relevant only if an information system is beyond the scope of usual enterprise architecture  [19], i.e., involves an ultra-large-scale software system  [20]. Therefore, e-government systems are typical instances of the system class characterized by the model of Koppenjan and Groenewegen  [9], the model is a suitable candidate as an analysis framework for e-government systems, e-government ecosystems and interoperability solutions alike. This is why we have chosen the model as the theoretical basis of our comparative case study. In our analysis, we exploit both the perspective of relationships between technological/institutional/process design (that we call “co-design perspective” for short) and the four-layered model of institutional design (that we call “institutional analysis perspective” for short).

4 Background in the Netherlands and Estonia

4.1 The Netherlands

The Netherlands consistently performs well in e-government rankings  [4]. The country has a high level of decentralized governance and public private collaboration in public service delivery. This decentralization results into various government agencies that use different infrastructures. For instance, the Tax administration uses a different data exchange infrastructure with the private sector and other government agencies (called Digipoort)  [21] than the customs authority (called Single Window for Maritime and Aviation)  [22]. Across different sectors such as health, energy and education, different data exchange infrastructures are used. Table 2 provides an (incomplete) overview of the main data exchange infrastructures in the Netherlands.

Table 2. Overview of various data exchange infrastructures in the Netherlands.
Full size table

Table 2 is by no means exhaustive, but does highlight the variety in data exchange infrastructures. While some are only web-portal based, most infrastructures support application-to-application data exchange. Municipalities also have different infrastructures, which provides autonomy, but is not cheap to operate and maintain. On a municipal level, there is a growing tendency to combine forces and to develop and use a shared infrastructure. Inspired by the X-Road approach in Estonia, municipalities are currently piloting with the “common ground”.

4.2 Estonia

Estonia is signee of the D9 (Digital Nine) charter, i.e., a member of the “Digital Nations” network, also known as D9 or “Leading Digital Governments”. In media, on international conventions on digital transformation as well as in policy maker circles it is often perceived or presents itself as leader when it comes to e-government; actually, stakeholder from many countries, as from the Netherlands, have visited Estonia with the aim to learn from their success. The perception of Estonia as a digital leader might be, in large parts, due to communication strategy  [23] and nation branding  [24]; still: the technological, legal and organizational assets in Estonia have been designed with and for each other and evolved over time into a particularly stable e-government ecosystem. At the centre of the Estonian e-government ecosystem lies the interoperability framework X-Road  [5, 25]. Estonia created X-Road – an application network for exchanging data among agency systems so that all government services are effectively available in one spot. In addition to offering querying mechanisms across multiple databases and supporting the secure exchange of documents  [26], X-Road seamlessly integrates different government portals and applications. Also the private sector can also connect with X-Road to make queries and benefit from access to a secure data exchange layer  [27].

5 Findings

The main objective of the workshop in Estonia was to learn from each other’s approaches to e-government design and look ahead which challenges could be dealt with in a collaborative manner. Table 3 outlines the comparison based on the layered institutional analysis of the institutional design model of Koppenjan and Groenewegen  [9].

Table 3. Institutional analysis based on the Koppenjan and Groenewegen model  [9].
Full size table
Table 4. Institutional comparison from the co-design perspective (Koppenjan and Groenewegen  [9]).
Full size table

During the workshops, multiple user/citizen oriented contexts were used to compare the e-governments in Estonia and the Netherlands: using the life event approach  [30], user-centred scenarios were discussed in depth, specifically comparing the steps users need to take in order to achieve their goals. This provided rich case descriptions as substance for the institutional analysis frameworks. Next, Table 4 outlines the comparison based on the technological/institutional/process co-design perspective of the institutional design model of Koppenjan and Groenewegen  [9].

Fig. 2.
figure 2

High-level technical design of the Dutch data exchange infrastructures (please note that not all the different infrastructures outlined in Table 2 are displayed).

Full size image

The high-level technical design of the Dutch data exchange infrastructures is sketched in Fig. 2.

In accordance with Table 2, citizens work with multiple data exchange infrastructures in the Netherlands. The demarcation is clear for public services and private services. When consuming public services, citizens can use the state-issued eID called DigID (digid.nl/en). The ‘MijnOverheid’ portal and app provides data access to public agency messages (pdf files). This is a one-way data flow; for service consumption or data entry users need to go to the website/portal of the respective public agency. Here, they can use DigiD. Since law dictates that DigiD (as well as other public sector data exchange infrastructure components such as the ‘MijnOverheid’ portal and app, the citizen message box and the Digipoort) can only be used by public organizations, private organizations have to use their own infrastructure or use sector-specific infrastructures (see Table 2 for an overview).

The high-level technical design of the Estonian data exchange infrastructure is sketched in Fig. 3. Estonia embraces an integrated design, which extends the technology. X-Road consists of technical, legal and organizational assets that are teamed together as described in the sequel. In a narrow sense, the data exchange infrastructure is nothing but the data exchange layer X-Road (dark grey in Fig. 2); in a broader sense, it encompasses also crosscutting services that are built on top of X-Road such as the document exchange center  [26]. The data exchange layer relies on a PKI (public key infrastructure) and a time stamping service. A PKI is itself a combination of technological assets (such as the certification server) and an institution, i.e., the CA (certification authority). Next, the X-Road consists of security servers, which are software components. Each organization (public agency or private company) that wants to exchange messages over X-Road must become an X-Road member first, by application and registration at the Estonian IT Authority RIA (Riigi Infosüsteemi Amet). Then, each X-Road member needs to install the security server. The basic task of the security server is to encrypt and decrypt the data exchange messages sent among the X-Road members. For this, the security server teams together with a signature device, which must be obtained from the CA by each X-Road member. Furthermore, the security server allows for access rights management: each X-Road member can determine itself, which other X-Road members can access its services. It is regulated that each X-Road member announces its information systems and services to RIA before it is allowed to launch them. This way, RIA has the chance to streamline the service offering, in particular, to enforce the the once-only principle  [31]. Similarly, timestamping by the installed security servers as well as timestamping service provision are subject to regulations. All citizens can see all communications about them. This creates transparency and ensures that mistakes or fraud can be detected immediately.

Fig. 3.
figure 3

High level, integrated technical design of the Estonian data exchange infrastructure.

Full size image

The Estonian e-government ecosystem is a network consisting of different players with well-defined roles. As a fundamental task, the e-government ecosystem implements the Estonian eID  [28, 32], which provides also the basis for any system interoperability, compare again with Fig. 2. The Information System Authority (RIA) and the Estonian Police and Border Guard Board (PBGB) are the main authorities in the e-government ecosystem. RIAFootnote 2 operates on behalf of the Ministry of Economic Affairs and CommunicationsFootnote 3. RIA coordinates the development and administration of the state’s information system. It oversees the functioning of the Estonian PKI, organizes activities related to information security, handles security incidents that occur in Estonian computer networks, and serves as the technical eID competence centre.

6 Conclusions and Policy Recommendations

Estonia and the Netherlands had different starting points and used different paths in developing their e-government. The main differences in approaching e-government boil down in the following categories:

  • Strong centralized government institutions in Estonia versus decentralized and market-oriented institutions in the Netherlands.

  • Institutional boundaries in the Netherlands between the use of state-owned digital infrastructures (or components) for public services (only allowed for citizen-to-government interactions) and semi/non-public services (not allowed for citizen-to-business and business-to-business interactions) versus no boundary in Estonia (single infrastructure, seamless use across public and private services).

  • Focus on experimenting and learning by doing in Estonia, whereas the Netherlands is risk-averse and leaves innovation to the market.

These differences make it difficult to copy each other best practices as the institutional settings are different, and due to the different data exchange infrastructures that are currently in place. Nevertheless, both countries can learn from each other’s approaches. The capacity of reliable institutions to meet performance expectations, perceptions of competence and effective public service delivery for all, along with public accountability, should be among the leading concerns in developing e-government further. It is recommended that governments exploit the potential of digital technologies through coherent national policies that are closely aligned with the knowledge of user needs that is available at the respective public agencies. Being successful requires a whole-of-government approach across ministries and agencies and between levels, as well as partnerships with non-government actors. Such an approach needs to be supported by a high-level political will, an example of which is an effective cross-government institution with clearly earmarked financial resources and decision-making powers. This demands a shift from inward, disjointed and process-oriented organizational structures to highly collaborative frameworks for seamless delivery of services towards citizens and entrepreneurs. Maximizing the potential of digital technologies also demands appropriate data exchange infrastructures for interoperability and digital transactions across the public sector, dependent on common standards, data sharing, highly skilled staff, as well as knowledgeable organizational capacity.