Prioritizing public values in e-government policies: A document analysis

Three public administrations were selected, two traditional national public administrations and the European Union (EU). It was decided to select Belgium and the United Kingdom for three reasons. Firstly, the two countries have a different governance approach, ranging from a more traditionally-oriented Weberian bureaucracy to strongly NPM-oriented structures, sometimes influenced by more recent governance approaches. An overview is provided in Table 1. Belgium has a stronger legacy of the traditional Weberian bureaucracy, with some influences of NPM and NPG. The UK was originally a Weberian bureaucracy, but evolved into an administration with strong NPM characteristics. Looked at from a European perspective, it is also clear that both countries have different governance traditions: Belgium has a continental tradition, whereas the UK has an Anglo-Saxon tradition. The different ways in which these public administrations are governed make an analysis of both cases relevant. A second reason for selecting those two public administrations is the fact that both countries differ in terms of their e-government developments. Since 2004 the United Kingdom has always scored higher than Belgium on the United Nations E-Government Index where Belgium has consistently scored slightly lower. Finally, both countries were selected for practical reasons as well. The authors have an extensive knowledge of the e-government policies, the overall governance approaches and political situations in the countries and the authors have a mastery of the languages of both countries.

In addition to the United Kingdom and Belgium, it was also decided to analyse the e-government policy of the EU. The EU cannot be regarded as a traditional public administration and can therefore not be compared with the other two public administrations. The Union has no competence in defining the e-government policies or strategies of member states (Publications Office of the European Union, 2016). It does, however, have an important impact on e-government developments for EU member states via legally binding and nonbinding actions in other domains where the EU has competencies (De Coninck & Van Hecke, 2018). EU member states do not function as self-standing public administrations that are not influenced by other public administrations. The EU is one key influence on public administration as shown by Hooghe and Marks (2001, 2003). The EU is to this end relevant to analyse, as there is the expectation that the PVs balance at EU level also influences member state administrations. At the EU level the horizontal and strategic e-government policies developed by the European Commission, the Council of Ministers, the European Parliament and the Ministerial Declarations are examined. This can, for example, include Minister Declarations such as the 2009 Malmö Declaration or 2017 Tallinn Declaration, the Directives and Regulations that influence e-government development at the level of the member states or the nonbinding EU Action Plans on E-Government or the Interoperability Framework developed by the European Commission.

The period of analysis runs from 2000 to 2018. This has been a deliberate choice as it seems that the year 2000 was an important year for public administrations around the world initiating strategic and horizontal e-government policies. At the Belgian level, for example, it was decided in that year to launch a horizontal e-government organisation, In addition, countries such as Singapore and Brazil started to invest more funds in their e-government policies (Ke & Wei, 2004; Musafir, 2018; SPF BOSA, 2019). Whereas the UK and the EU already published a number of relevant policy documents before 2000, it was considered necessary to cover a period in which the three public administrations would have published horizontal e-government policies. Otherwise it would not be possible to make a comparison between the two national public administrations and the EU. Therefore we decided to focus our research on this period.

Documents defining the horizontal strategic direction of the e-government policy were selected for the analysis. The concept of ‘horizontal’ e-government policies, refers to policies that are developed for the entire public administration. The policies are supposed to cover all different policy areas in which the public administration is active and are as such not developed for a specific policy area. For the two traditional public administrations (Belgium/UK) this implied a focus on non-legally binding documents as those documents define the strategic direction of the e-government policy. All the documents focus on the e-government policy as a self-standing horizontal policy domain and where prepared or co-prepared with the ministerial cabinets, by the administrative organisation that is responsible for the e-government policy of the administration. The documents are all of strategic nature and set the direction for the future developments of e-government activities. None of the documents have a technical or operational nature. The strength of working with strategic documents is the fact that they are more abstract and therefore often include more information on the prioritized PVs. At the EU level, both legally binding documents (Directives, Regulations) and non-legally binding documents (Action Plans, Ministerial Declarations etc.) were selected. The first category has a defined impact at the national level and is therefore of crucial importance for national e-government developments, the second category is more strategic and includes as such also a clear orientation to PVs.

The first step of the policy document selection was to define a list of policy documents that would be as complete as possible. To this end it was decided to apply a mix of purposive sampling and snowball sampling (Bryman, 2016). None of the three public administrations has a structured overview of e-government policy documents that were published by the public administration during the analysed period. As a result of this the authors were forced to make use of this combined sampling strategy: Some e-government policy documents of each public administration were, because of earlier research, known to the researchers. This can be labelled as purposive sampling. Those documents then served as an entry point for finding other policy documents to be included in the analysis: Based on references in the documents that were known in advance to the researchers, extra documents were selected. All documents were selected on the basis of three criteria: The focus had to be the horizontal e-government policy of the public administration, the documents had to be of a strategic nature and the documents had to be prepared by the public administration or in collaboration with the public administration.

Once a list of documents was created for each administration, random sampling was carried out on each of the three document lists whereby half of the documents were selected. Half the documents were selected for two reasons. First, there is a practical argument: It would have been impossible to analyse all documents due to limited resources. Secondly, and more importantly, each document requires a thorough analysis in the reporting. Analysing all documents would lead to an overload of information to be reported. An overview of the numbers of selected and analysed documents can be found in Table 2. A complete overview of all documents can be found in the Annexes 1–3.

Choosing for a policy document analysis is a deliberate choice. Since research focuses on the period 2000 to 2018, there would be a strong risk for a memory bias if this research is conducted via interviews with civil servants. Besides the fact that it is impossible to recollect what and why certain decisions were taken over such a long period, also the ideas of individuals on values might change and would influence the research results. Secondly, we defined PVs as being ‘distinctive of an individual or characteristic of a group’. Policy documents fit this definition in an excellent way: They have been prepared by groups of people and can as such be regarded as disconnected from the individual level.

The PVs typology presented in Table 3 was used for the coding of the policy documents. The applied PVs typology has been obtained on the basis of one group discussion and four bilateral discussions. Each PVs was defined and in this way a comprehensive coding scheme was obtained. The coding itself took place in NVivo 12, a qualitative text analysis software. A closed coding approach was applied. Only those PVs which are part of the typology as presented in Table 3 were used for the coding in order to ensure that the document results are comparable. A sentence or word group could refer to one PVs or to several PVs, if the latter was the case then different codes where assigned to the same sentence or word group. In one sentence or word group references could be made to two or more PVs. Making the decision on assigning this sentence or word group to only one PVs would lead to biased results. So if this situation appeared, it was decided to assign the sentence or word group to the different PVs that related to it. The authors completed the coding: The first author was responsible for the coding of all the documents and the co-authors reviewed the coding process. In this way the authors were able to ensure that the data was always coded by the same person and that there was no difference in the way that the different documents were coded

Once the policy documents had been coded, the data for each document was grouped, weighted and interpreted. For each of the three PVs categories (hierarchy, market and network), the PVs were added up to get an overview of the total number of references to PVs in one category. After this, the data was weighted in order to make the results of the three categories comparable. This was a necessity since there are three PVs categories that each consist of a different number of PVs. The weighting was conducted as follows: The total sum of PVs references for each category was multiplied by the factor representing the balance between the total number of PVs (35) divided by the number of PVs for each category (hierarchy: 16 PVs, market: 12 PVs and network: 8). On the basis of the weighted data, the percentages were calculated to represent the results in a relative way, that is, one category represents a percentage of the total amount of PVs references found in each analysed document. By making use of this approach, the data for each analysed document could be compared and analysed.

2.5 Assumptions and limitations

There is a number of important limitations and assumptions which we would like to clarify before presenting the results of the analysis. Firstly, it has to be underlined that the number of documents related to the e-government policy of public administrations was too high for a well-executed qualitative text analysis. A first list of documents was created on the basis of purposive sampling and snowball sampling, as explained above, and on the basis of this list a random selection was made. The list which resulted of this random selection was then analysed. A first risk is that the purposive and snowball sampling strategy did not detect all relevant policy documents. A second risk is that the random sampling led to a loss of relevant information which is not included in the analysis. Secondly, PVs are a complex study object: We have for this purpose, on the basis of one group discussion and four bilateral discussions, made a precise selection of the PVs, left out PVs which did to not relate clearly to a certain governance approach and defined the PVs. However, the typology developed in this article is open for criticism and debate. Finally, a document analysis requires a certain level of interpretation: The PVs are all defined but there remains a risk for a certain interpretation which might in turn lead to an interpretative bias in the final research results.

According to Kluckhohn (1952) a value is ‘a conception, explicit or implicit, distinctive of an individual or characteristic of a group, of the desirable which influences the selection from available modes, means and ends of actions’. Bozeman (2009) argues in this respect that it is questionable if a definition of PVs is necessary. According to the author there is no necessity in defining it as it is closely related to other concepts such as public interest, public value criteria or public goods (Bozeman, 2009). Nevertheless, the author does state that PVs are ‘those providing normative consensus about (a) the rights, benefits, and prerogatives to which citizens should (and should not) be entitled; (b) the obligations of citizens to society, the state, and one another; and (c) the principles on which governments and policies should be based’ (Bozeman, 2009; Bryson et al., 2014). What, however, is missing in the description of Bozeman (2009) is the broader link to society: The attention is only on citizens, whereas society encompasses more than just citizens. Government policies also target enterprises, non-profit organisations and other societal actors which are distinct from citizens.

In order, however, for the reader to have a clear understanding of what we mean with PVs, we can state, in line with Kluckhohn (1952) that public values are ‘concepts, distinctive of an individual or characteristic of a group, of the desirable which influences the public services offered by the public administration’. The elements which Bozeman (2009) brings forward will, despite rigid focus on citizens, function as guiding principles: They underline the relation which is established between state and society.

It is essential to clarify the differentiation between ‘public values’ (plural) and ‘public value’ (singular) in order to avoid confusion between both concepts. The concept of ‘PVs’ (plural) points to the input side and refers to the driving forces behind a certain decision or behaviour. This line of thought is represented by, for example, Bozeman (2009), and focuses on the policy or societal level. The public value (singular) line of thought focuses on the values of public managers themselves and the creation of public value by them. This line of thought is represented by Moore (1995). It refers to the output side and is focused on the performance of administrations (Bryson et al., 2014; Ho et al., 2018; Moore, 1995).

From an empirical point of view, a number of relevant studies have been published on the potential conflicts that might emerge between PVs (de Graaf et al., 2016; Hubert & van Hout, 2011; Reino & Jaakson, 2014; van der Wal et al., 2011). de Graaf et al. (2016) state in this respect for example that “in daily practice, multiple public values that are all desirable will conflict in such a way that choices have to be made”. This ‘conflictualization’ connects to our research: The prioritization of PVs can be considered as a first step towards a potential conflict. Public administrations will in their daily practices be forced to make certain choices on the PVs they strive for. Societal actors, the public administration, public servants, politicians or citizens might desire multiple values, but not all can be fulfilled. Therefore, a balance in the PVs will be aimed for. Some PVs will receive more attention than others, whereby the prioritization of certain PVs can change over time and certain values become more important than others (de Graaf et al., 2016; Hubert & van Hout, 2011).

3.2 Public governance approaches: Hierarchy, market and network

Pollitt and Bouckaert (2017) state that ‘governance’ is a concept that “appears in almost as many versions as there are authors writing about it”. Rhodes (1996) for example describes governance as “a change in the meaning of government, referring to new processes of governing; or changed conditions of ordered rule; or new methods by which society is governed”. Although it gives a first indication of what the concept can mean, thereby referring clearly to the role of government towards society, it remains vague and imprecise. It points in different directions and can include various aspects. What we retain from the description is the focus on the organisation of government in relation to society. Rijke et al. (2012) are more precise in describing the concept. According to them governance “refers to both processes and structures for steering and managing parts of societies”, entailing “the networks of actors, institutional frameworks and processes that take place within these networks and frameworks.” (Rijke et al., 2012). Once more we note a clear focus on the role of the government and administration in relation to society. However, the definition of Rijke et al. (2012) entails a normative element as it refers to ‘networks’, whereas governance can also take place in different forms, thereby not focusing on the presence and/or development of networks.

What is clear from both definitions is however that the government and administration orientation is a key element of governance. A ‘governance approach’ can as such be defined as ‘a method for steering and managing parts of society by the public administration’. Such a method is linked to certain PVs which are prioritized above others. Before going deeper into the relation between PVs and governance approaches in the next section, the three governance approaches will be briefly introduced. This paper looks at three ideal type public governance approaches: Hierarchical governance, market governance and network governance.

Hierarchical governance can be related to the ideal type bureaucracy developed by Max Weber. Meuleman (2008) attributes the following characteristics to hierarchical governance: “[It] accounts for top-down decision-making, strict internal and external accountability procedures, a hierarchical organisation structure, an emphasis on project management rather than on process management, strategy styles of a planning and design type, and a strong preference for legal measures.” (Meuleman, 2008). Hierarchical governance has been the most used governance approach in Western public administrations, although it has been criticized since the 1970s, leading to an uptake of the two other governance approaches by those public administrations.

The second governance approach, market governance, is related to NPM. A number of characteristics are related to market governance, such as the inclusion and focus on efficiency principles, the adherence to private sector approaches as well as market mechanisms, a customer orientation, increased attention for management skills, output-driven policy making, increased administrative competition instead of hierarchy when providing public services and contractual provision of public services (Hood, 1995; Kickert, 2001; Levy, 2003; Meuleman, 2008). Just like with hierarchical governance, also market governance cannot however be captured within a single definition. It is a broad approach which resulted out of the economic recession of the 1980s that has been (partially) implemented in various forms in different Western public administrations (Hood, 1991, 1995). However, what is always present in this governance approach is the ‘market’ element which “refers to market mechanisms and market thinking” (Meuleman, 2008).

The third and final governance approach discussed in this paper is network governance. It can be defined as “the ‘management’ of complex networks, consisting of many different actors from the national, regional and local government, from political groups and from societal groups (pressure, actions, and interest groups, societal institutions, private and business organisations)” (High et al., 2005; Kickert, 1997; Meuleman, 2008). Meuleman (2014) argues that network governance “is characterized by cooperation rather than coercion or competition, by trust rather than authority or price, and by interdependency than dependency or independency”. This governance approach will for example focus on the inclusion and consultation of the broader public in policy making and service development. Concepts such as co-creation and coproduction can be seen as part of network governance (Van Eijk & Steen, 2016; Steen et al., 2016).

3.3 The relation between PVs and governance approaches

Although the literature on PVs and governance approaches is elaborate, much less attention is given to the relation between PVs and governance approaches. This section aims to clarify this relationship by looking at currently existing empirical and theoretical literature, in order to present at the end of this section a typology that can be applied in this research.

From an empirical point of view, we see that some authors start from the potential conflict that can emerge between PVs to build a connection to related governance approaches. When such a conflict emerges, the actor confronted with this conflict will need to make a decision on which PVs are prioritized over other PVs. The research of Jaspers and Steen (2018) on PVs and coproduction starts from a need for prioritization of PVs and the potential conflicts that can emerge between PVs. Specifically the authors look at the use of coproduction – which can be considered as being part of network governance – as a governance mode to prioritize PVs and to deal with potential conflicts between PVs (Jaspers & Steen, 2018; Steen et al., 2016). The PVs are, in the research of Jaspers and Steen (2018), considered as the dependent variable on which the coproduction governance approach impacts as an independent variable. This causal relation is not put forward by Vrangbæk (2009) who analysed the PVs of Danish civil servants. In his analysis on the PVs of civil servants, the author finds a number of PVs that appear among different public organisations: responsibility, transparency, maintenance of judicial values, adherence to professionalism and innovation. The author found a clear connection to governance approaches and states that there is mostly adherence to hierarchical governance and clan-based governance – an approach not included in this study as there is discussion in literature on whether or not this governance approach can be considered as a separate approach (Meuleman, 2008; Vrangbæk, 2003, 2009).

From a theoretical point of view the relation between PVs and governance approaches remains also largely understudied. An example is the article of Jørgensen and Bozeman (2007) in which the authors identify a list of 68 PVs on the basis of a systematic literature review. The authors make no connection to governance approaches. Instead the authors make a connection to the different relations between public administration actors as well as between the public administration, the political layer and society. Other authors have however established a connection between PVs and governance approaches, such as Hood (1991) in his work on NPM. Hood (1991) argued in this respect that “different administrative values have different implications for fundamental aspects of administrative design – implications which go beyond altering the ‘settings’ of the systems” (Hood, 1991). Bannister and Connolly (2014) connect their typology, which is presented in Table 4, also to the research conducted by Hood (1991). The PVs typology of Bannister and Connolly (2014) and Vrangbæk (2009) are, because of their connection to several governance approaches, considered as highly relevant for this research.

There are however a number of difficulties in both typologies. The meaning of certain PVs (‘proper use of public funds’, ‘facilitating the democratic will’, ‘respect for the citizen’) of the Bannister and Connolly (2014) typology are unclear. Furthermore, the connection to the governance approaches remains partially undefined. The duty oriented values are the only category of values that can be connected to a governance approach. Those values can be traced back to the sigma-type values of Hood (1991), which the last author connects to NPM. However, other authors connect some of the PVs in this category to the Weberian bureaucracy (Hondeghem, 2017). It has therefore been decided to leave out the PVs that had no clearly definable meaning and to rework both typologies to a new typology.

In the typology of Vrangbæk (2009), presented in Table 5, all PVs are connected to a governance approach. However, also clan-based governance is taken on board in this typology. According to us this cannot be considered as a separate governance approach, and the focus lies in this research on three ideal-type governance approaches: Hierarchy, market and network governance (Bouckaert et al., 2010; Meuleman, 2008). There are two reasons for not including clan-based governance in the applied typology. In the first place there is discussion in the literature on whether or not this can be considered as a separate governance approach, whereas the three other approaches are generally accepted as the three ideal-types of governance (Meuleman, 2008). Secondly, the argumentation applied by Vrangbæk (2009) on why clan-based governance was included in the typology is unclear in comparison to why the three other governance approaches were included. It was therefore decided to not include the clan-based governance approach. Furthermore, some PVs (‘public insight and transparency’) presented in this typology did not belong solely to one of three categories or combined different PVs (‘innovation and service orientation’).

Recognizing however the strengths of both the typologies of Bannister and Connolly (2014) (Table 4) and Vrangbæk (2009) (Table 5), we decided to rework both typologies to a reworked version which is presented in Table 3.1

The strength of this typology lies in the fact that it continues on the work conducted by other researchers in this domain, who build their typologies either for the specific domain of e-government or have applied their typology in practice. Furthermore, the strongly added value of the typology is the clear relationship between on the one hand PVs and on the other hand the governance approaches. As stated earlier, this theoretical relationship between PVs and governance approaches was until now largely neglected and via this typology we aim to contribute to this relationship.

We wish to underline that the relation between PVs and e-government policies has only received marginal academic attention. A systematic literature review revealed that in most of the scientific literature on this topic there is, first of all, a conceptual confusion between ‘PVs’ and ‘public value’. Secondly, the PVs perspective is mainly used as a starting point, but without a clear introduction or clarification of their meaning, and finally, the authors often do not continue using the concept in their papers or for the argument that they aim to make. An exception to this are the results of Jørgensen (2007) and Palmhøj Nielsen (2003) who focused on Danish public sector organisations. The authors argue that the PVs of the Danish ICT Office shifted from traditional bureaucracy oriented PVs towards more NPM-oriented PVs. Interestingly enough they connected the PVs to governance approaches and used a highly similar typology as Vrangbæk (2009). Whereas at the start, the balance of the Danish ICT Office’s PVs was more inclined towards citizen equality, service quality and openness, those values were gradually replaced with increased attention for efficiency and productivity growth via e-government. According to the authors, this shifting balance in the PVs was a result of the changing governance structure in which the Ministry had more influence: It prioritized different PVs and was, because of its governance structure, able to prioritize also different PVs in the ICT Office. This is a highly relevant finding as it shows that the original values in the e-government policy of the organisation were replaced by NPM values (Jørgensen, 2007; Palmhøj Nielsen, 2003).

4.1 Belgian and United Kingdom public administration

4.2 European Union public administration

The third set of analysed policy documents comes from the EU. As discussed earlier, the public administration of the EU cannot be treated as a public administration similar to the national public administrations of Belgium and the UK. The EU is a collaboration between different countries with a public administration possessing various tools to influence the policies of member states, as has been demonstrated by, among others, Hooghe and Marks (2001, 2013).

The results for the different analysed EU policy documents are presented in Fig. 6. Market related PVs are four times dominant, twice having a percentages above 50% (EU2013 and EU2016). Network related PVs are three times dominant, also twice with a percentage above 50% (EU2006 and EU2007) and finally, the hierarchy related PVs are dominant in one document, also with a percentage above 50%. The hierarchy related PVs are however also six times the PVs with the lowest presence in the documents. A first factor influencing the balance of PVs in those documents seems to be the specific topic dealt with. The first set of documents that have a similar PVs balance between the categories are the EU2006, EU2007 and the EU2010 documents: All documents put network related PVs first, followed by market related PVs and hierarchy related PVs. This does not come as a surprise as the EU2006 and EU2010 documents deal with the development of interoperable e-services in the EU, which requires coordination and the development of cooperation networks in order to ensure the exchange of data and service building blocks. Highly relevant is the result of EU2004, the decision of the European Parliament and the Council on interoperable delivery of e-services, in connection to the documents EU2006 and EU2010 – both also focus on interoperable service delivery. Market related PVs are dominant in EU2004, followed immediately by network related PVs. The document focuses on improved service delivery towards internal and external governmental actors. EU2006 and EU2010, in contrast, put network related PVs first. The focus on improved service delivery seems to result in the need for more cooperation – with both internal and external government actors –, resulting in a shift towards network related PVs. Finally, similar to EU2006 and EU2010, is the document EU2007 on the establishment of a common ‘Infrastructure for Spatial Information’. This refers to the creation of a network – in this case for the exchange of data and information – and ways to stimulate cooperation.

OPEN IN VIEWER

The EU2013 document, i.e. the Directive on re-use of public sector information or the PSI Directive, shows a dominant focus on market related PVs. This data is in line with the overall aim of the EU institutions concerning the reuse of public sector information. The aim in the first place is not to give societal actors or citizens access to data to participate in policy making or to hold government accountable, but to support enterprises in their search for information (European Commission, 2018). This public sector information is in turn expected to lead to economic benefits and stimulation of the economic growth. This explains the market orientation when it comes to the PVs in the document.

Another relevant document is EU2014, that is, the eIDAS Regulation. It is the only document in the analysis with a dominance of hierarchy related PVs. This regulation aims to “enhance trust in electronic transactions” and relies thereby strongly on the necessity of complying with the law and the need to work on citizen, enterprise and societal security. The eIDAS Regulation is focused on access to public services in other EU Member States and is in that respect related to cross-border cooperation and interoperability. The analysis did not include key EU documents on interoperability, such as the ISA and ISA 2 Decisions, but is needs to be emphasised that those documents too are expected to have an influence on the national governance of Member States.

Finally, the documents EU2016 and EU2017, that are respectively the EU eGovernment Action Plan 2016–2020 and the Ministerial Tallinn Declaration on eGovernment,2

OPEN IN VIEWER

OPEN IN VIEWER

Besides the topic focused on in the policy document, there seems to be another factor influencing the balance of PVs. Figures 7 and 8 respectively present the PVs balance for the policy documents according to the authors and the legally binding status. Table 6 provides an overview on the legal status and authorship of the EU policy documents. Figure 7 shows that, when the European Commission is the sole author of the document, there is substantially more emphasis on network related PVs and much less on hierarchy related PVs. When the member states are involved, via the Council of Ministers of the EU or via the unofficial Ministerial Meetings, then the market related PVs remain almost stable in comparison to when the European Commission is the sole author, but the emphasis on hierarchy related PVs increases at the expense of the network related PVs. Also the results presented in Fig. 8 are highly relevant: Documents with a legally binding status have a much stronger focus on hierarchy related PVs then documents with a non-legally binding status. Indeed, when a document is non-legally binding then the attention for network related PVs increases at the expense of the hierarchy related PVs. The market related PVs remain stable. Those results can be explained by the powers of the different actors at the European level: When the European Commission is the sole author, or when the document is non-legally binding – and is as such connected to the European Commission, except for the unofficial Ministerial Declarations, then the authors cannot rely as strongly on hierarchy related instruments as when the document is legally binding. So at the EU level, the powers possessed by the actors seem to influence the balance of PVs as well as the topic on which the policy document is focused.

The relation between the EU and its member states is complex. The EU and the member states have a distinctively different approach to policy making and to the preparation of policy documents. Whereas Belgium and the UK position their e-government policies in non-legally binding documents, the situation is different at the EU level. Different factors influence and explain the PVs balance in the policy documents. The Belgian and UK policy documents follow the governance evolutions and trends that are discussed in literature – with a growing diversification of the PVs balance in the policy documents, but those trends and evolutions are much less visible in the EU policy documents (Pollitt & Bouckaert, 2017). The influencing factor here seems to be the power distribution among the different EU institutions. As Hooghe and Marks (2001) argue, the EU member states – in this case Belgium and the UK – no longer function as self-standing blocks. Both the EU and national levels influence each other. The results show indeed that when the member states are involved in the EU policy documents the balance of the PVs is different from when the member states are not involved. Member states are at the same time also being influenced in their policies by EU policy documents – for example, the open data policy, which is legally binding at the EU level and needs to be implemented as policy in member states.