Final Thesis - Leiden University
Protecting the Primacy of Member-States:
On the Role of Informal Institutions
in Limiting the Information Exchange
Between Europol National Units and Europol
Joris J. Broeders S1412809 11-06-2018 Word Count: 9997
Thesis submitted in partial fulfillment of the requirements of the degree
Of
Political Science: International Organisation [Msc] Faculty of Social and Behavioural Sciences
Leiden University
Submitted for review to:
Dr. N.J.G. van Willigen &
Abstract:
Since the European Council Decision that established Europol as a European Agency, there has been variance in the success within different crime areas. This thesis contributes to understanding of this variation. It argues that the phenomenon can best be understood as the result of actors strategically applying informal norms when exchanging information. Using internal document analysis and the review of relevant literature, it analyses four crime areas: Drug Related crimes, Financial Crime, Human Trafficking and Counter Terrorism during the Europol Strategy 2010-2014. The thesis aims to provide a systematic application of a
rational-choice institutionalism argument that can be used for other research on the topic. Given the novelty of applying this approach, it provides us with new insights into the behaviour of actors at the European National Units. It provides an overview of the functioning of the organisation in these four areas with a focus on the role of information exchange and existing theory. It finds that actors strategically create(d) informal institutions to counteract the European formal institutional demands placed upon them. It concludes by encouraging more work on the subject, both in terms of expanding theory and doing practical research.
I. Introduction 3
II. Literature Review 6
II.A: The Europol Mandate: Competences, Tools and Scholarly Review 6
II.B: Alternative Explanations 10
II.C: Organisational Theories: Rational-Choice/Functionalism, Social Constructivism and
Neo-Institutionalist Theory 11
III. Argument and Research Design 18
III.A: Theoretical Argument 18
III.B: Operationalisation 22
III.C: Data-selection and Evidence 24
IV. Analysis & Results: Institutional Characteristics; Data Value; Domain and
Enforcement. 25
IV.A: Characteristics 26
IV.B: Data Value 27
IV.C: Domain 28
IV.D: Enforcement 30
V. Conclusion 31
I. Introduction
In 2009, Europol was given a larger mandate which increased competences, responsibilities and also eased future legal amendments (2009/371/JHA); (Europol Annual Review (AR) 2010; 2011; 2012; 2013; 2014). This entailed an improvement in data processing capability, an increase in the number of requests, and an influence on policy-making (Carrapiço and Trauner, 2013; AR 2012; Disley et al., 2012). As such, Europol was to be the organisation for exchanging information and analysis related to organised crime in the EU. To achieve this, Europol was meant to receive information from member-states and vice versa, and provide operational support (e.g. Joint Investigative Task forces (JITs), training, best practices). While the overall consensus on the mandate was positive (Disley et al. 2012), the success was not uniform. The performance of Europol in counterterrorism was found to be weak, its performance in human trafficking medium with challenges remaining, its performance in financial crime medium with significant improvements, and its performance in drug
trafficking excellent. (Rozée, Kaunertal and Léonard 2013; Disley et al. 2012). As Justice and Home affairs (JHA) had to prepare a new strategy for the next mandate , understanding the variance and how to address it was necessary (AR 2012). As part of these preparations, Europol was evaluated in the Disley-report (2012), which identified problems with
staff-management and information exchange as important (Ibid., p. xxiii ). This thesis will analyse information exchange due to it underpinning all Europol tasks. The Disley-report (2012) identified two main types of barriers to information exchange: a) practical, such as insufficient time or resources and b) different (informal) standards across the various Europol
National Units (ENUs) and between the staff-members. However, despite the importance of the problem, the report was unable to reach any definitive conclusions . 1
In order to investigate the reason for the lower quality of information exchange in certain criminal areas, I examine the role of cultural norms. This thesis pursues three
arguments: Firstly, that the quality of information exchange in the different crime areas plays a key role in explaining the overall performance of Europol. Secondly that the interests of actors at ENUs can better explain the differences in what information is shared than practical barriers. Thirdly, that the ‘different (informal) standards’ can best be understood as emerging from the interests of rational actors that seek to limit the amount and type of information that gets shared. For these arguments to be true, key qualities of a these norms are that they are socially enforced but hidden from plain sight.
This thesis applies a subset of rational choice neo-institutionalist theory. In particular, it categorises the rules and norms that govern information exchange in the four crime areas into formal and informal institutions, and studies how they interact. While the exchange of information between and across member states - a key feature of the work of Europol (AR 2010-2014) - is bound by the same formal rules; ENUs have different norms and informal arrangements in place to decide what information is actually shared. These norms vary both on the basis of the subject matter and among the members of the Europol network. This research will be guided by the following research question:
What explains the difference in the quality of information exchange in the areas of counterterrorism, human trafficking, financial crime and drug trafficking during the first
mandate of Europol as a European Agency (2010-2014)?
1 “It is beyond the scope of this evaluation to assess whether, and the extent to which, these factors play an
Studying this subject can help us address that Europol currently performs worse in Financial Crime (Paravicini 2018) and performs significantly better in the context of counter terrorism (Ring and Wild 2018).
This thesis has five parts. The first part, is the introduction. The second is a literature review, that firstly covers the state of Europol during its first multi-annual strategy. In particular, it establishes the competences and working methods of Europol to show the importance of information exchange. Secondly, it reviews existing explanations on the challenges that Europol faced and their suggested causes. Thirdly, it covers the theoretical basis for the argument by examining different theoretical perspectives for studying outcomes at an (international) organisational level. This section will discuss formalistic or structural accounts and contrast these with social constructivist approaches. It proceeds by arguing that the case of information exchange at Europol requires insights from both perspectives and how the theory of (in)formal institutions from a rational choice perspective provides this. The third part discusses the transition from theory to practical analysis by covering the research design. It looks at the theoretical and practical dimensions of doing the research, including data-selection. The fourth part contains the analysis of informal institutions at Europol and is grouped into four sections: 1) their characteristics 2) their interaction with different types of information, 3) their domain; and 4) their enforcement. Part five concludes by discussing the implications of the results on Europol and contains recommendations for doing future research on the organisation and/or by employing the same theoretical framework for research into similar cases.
II. Literature Review
This section provides an overview of academic and professional literature in three sections: Firstly, it establishes an understanding of Europol, including its competences and working methods, during its first mandate (2010-2014) to show the key role of information to the proper functioning of the organisation in practice. Secondly, it reviews theoretical arguments that the literature has already provided to explain the the challenges. Thirdly, it covers theoretical perspectives for studying international organisation performance. The insights from these perspectives can be combined by using a novel approach to the subject in the form of rational choice institutionalism (RCi), a subset of neo-institutionalism.
II.A: The Europol Mandate: Competences, Tools and Scholarly Review
To improve the work of Europol and ease future (legal) amendments, the EC expanded the mandate of the organisation. This decision entered into force on 01-01-2010. The mandate expansion enhanced the ability of Europol to process and distribute information; to provide operational support; and endowed the organisation with greater resources. As the Lisbon treaty abolished the pillar system in favour of a more comprehensive approach towards internal security and policing, Europol was to eventually occupy a central role in the
EU-structure (AR 2013). As part of the first multi-annual strategy 2010-2014; Europol made the excellence of the Criminal Intelligence Hub (CIH); Support Centre for law enforcement operations (ScLEO); and Centre for Law enforcement Expertise (CfLEE) its primary goal. However, the organisation faced a number of difficulties with information exchange (Disley et al. 2012). The majority of crime areas actually held constant before and after the mandate
expansion, despite the tools given to Europol to expand into new areas (Deflem 2006; AR 2
2010; AR 2011). Moreover, in spite of some (informal) rights to initiative, the work of Europol was (and is) overwhelmingly on the basis of requests from the member-states, third 3
party states or international organisations . This dependence on third parties distorts the 4 5 expected comprehensive analysis as a) the type of information and b) the amount of
information is provided varies by third party. The member states were aware of the problem, but remained consistent in their resistance to delegating authority and certain types of
information to Europol (Karns, Mingst and Stiles 2015, p. 193).
To investigate the precise effects of the quality of information exchange, I will now look at three important tasks of Europol in greater detail: publishing the ‘main’ documents; facilitating intelligence exchange & analysis; and providing operational support. The first task is to publish the four regular main documents that Europol produces (Strategy 6
2010-2014; Europol 2018): the IOCTA, (S)OCTA, TE-SAT and the Annual Reviews. These 7 documents present the ‘state-of-the-art’ of European policing. The first three provide
knowledge and policy recommendations and are considered benchmark reports. They are written by Europol analysts on the basis of data and analysis that is allowed to be published in this way and used by politicians and policy-makers to define priorities and policies (Rozée, Kaunert and Léonard 2013). For this reason, the intelligence behind these reports needs to be as accurate and complete as possible. The ARs are made publicly available, and serve two
2The exception being the 2013 opening of the European Cybercrime Centre in 2013 (AR 2014).
3In 2014, the distribution of actors starting new cases in SIENA was 92% member states, 4% Europol and 4%
other actors (AR 2014).
4If Europol had an agreement with them.
5Idem; the number of non-Eu countries and organisations with which Europol had agreements changes over the
period under discussion and not part of the scope of this research.
6 Europol produces many more reports including those dedicated to regions (West African Organised Crime
Threat Assessment (WA-OCTA) and Russian Organised Crime Threat Assessment (ROCTA)), but the ones discussed are the main documents.
7Internet Organised Crime Threat Assessment, (Serious) Organised Crime Threat Assessment and the EU
official functions: to publicly ‘market’ their successes; and to keep Europol accountable to European political actors (AR 2010). Unofficially, they are one of few the ways in which Europol can put pressure on countries for them to increase their information sharing. This can be done by pointing out (earlier) challenges with them or making other countries in the same region look better by comparison. The contents of an AR may be thus used as a basis to question the national police forces and their cooperation with Europol, though their actual impact is likely limited.
The second task is facilitating intelligence exchange and analysis. The most important tool for this is the European Information System (EIS), which is is accessible through the Secure Information Exchange Network Application (SIENA). The information is gathered through member-states, third parties or organisations with which Europol has an exchange agreement (‘strategic’: when all information excluding personal data can be shared and ‘operational’: where personal information can also be included (AR 2012)). The satisfaction with the technical quality of the system remained (and remains) consistently high (User Survey 2016). A lack of trust in the system is therefore unlikely to be the reason for countries not to share information. The majority is procured via ENUs and their associated liaison officers (individuals who perform a linking function between Europol and a particular country or organisation, most of whom are stationed at Europol (AR 2010; AR 2014)). All information is given handling codes, which are precise instructions on how and by whom the information may be used. Herein, the most important decision is whether Europol can add the data to its databanks and use it for analysis, or whether it merely passes through. These are decided on by the (original) owner of the information. Even though article 8(4) of the ECD requires the ENUs to provide Europol with the information that is ‘necessary for it to carry out its tasks’ including ‘to respond to Europol’s requests for information, intelligence and
advice’ and ‘[shall] supply Europol with information for storage in its databases” , only half 8 of the information exchanges facilitated by Europol allowed for the information to be
incorporated in its own analysis working files and/or projects (Disley et al. 2012). This means that, even though many countries combine their own intelligence with Europol data, the comparison and its implications are not always shared with Europol (Ibid.) thereby decreasing the added value Europol can provide.
The third task is operational support. This comprises activities including, but not limited to, setting up Europol Mobile Units (which provide on the spot assistance and expertise); Joint Investigation Teams (JITs); hosting of meetings; exchanging expertise; and advising and/or requesting third parties to commence an operation or project (ECD 2009; 9 Disley et al. 2012). It goes without saying that it is in the interest of both Europol and the country for Europol to have the most relevant information.
As I have shown, the extent to which Europol has access to information is the primary component of its success: without information, Europol cannot perform comprehensive analysis for its reports, nor provide its clients with high quality intelligence. A lack of access also adversely affects the quality of its operational support. This ultimately harms the ability of Europol and national police to fight organised crime. It is therefore unsurprising that the problems with information exchange and the willingness of member states to cooperate are flagged extensively by evaluators and scholars (Disley et al. 2012; Carrapiço and Trauner, 2013).
8Clauses a, b and f.
II.B: Alternative Explanations
Apart from the willingness of third parties to exchange information, scholars and
policymakers have suggested a number of alternative reasons as to why Europol had varying success in completing its aforementioned tasks. Fägersten (2010) argues that Europol itself did not consistently share relevant information, due to Europol staff believing that they are uniquely qualified to process the information thereby violating their first task. Furthermore, some have suggested that the problem may lie in the relative strength of the ENUs within their national police hierarchy (Bures 2016, Disley et al. 2012). The argument goes that greater levels of centralisation within the national structures allow the ENU greater access to data from different regional departments. This then allows them to share more information with Europol (Ibid.). Finally, two types of organisational reconfiguration that occured
simultaneously are suggested as an explanation. Firstly, the organisation itself became subject to European Agency standards for hiring and was continuously reorganised during the first four years, including the addition of new liaison officers (AR 2010; 2011; 2012; 2013; 2014). These changes require(d) time to implement properly due to having to adapt working
methods, re-establish working hierarchies, etc. thereby affecting the efficiency of the organisation. Secondly, as part of the Lisbon treaty, the roles and competences of many European organisations such as Frontex, Eurojust and Europol were repeatedly altered, leading to double work in certain areas, with others falling behind; a problem that had already been flagged before Europol became a European Agency (Fägersten 2010).
However, the current arguments from the literature do not fully address two concerns. Firstly, identifying that member states may not wish to share information is insufficient in a context wherein enforcement and monitoring are high. Those who wish to go against the
formal structures would face the likely repercussions. Secondly, arguments about the
extensive reconfiguration of the organisation and its staff are convincing when describing an overall change in the level of performance, but cannot explain why the performance varies structurally per crime area. Therefore, I now move on to examine different theoretical perspectives for studying international organisational performance in general in order to find an angle to address these two concerns.
II.C: Organisational Theories: Rational-Choice/Functionalism, Social Constructivism and Neo-Institutionalist Theory
Different perspectives can be used to study the variance of performance during the first mandate of Europol. In my case, I require a theoretical perspective that allows me to explain the highly formalised and rule-bound nature of police work in the context of European legislation. Simultaneously, I require the tools to help me understand why representatives of member-states deviate from said rules to which they have bound themselves. To provide an overview of relevant academic approaches I review three theoretical perspectives. First, positivist accounts in the form of structuralism and rational-choice. Second, social constructivist approaches in the form of bureaucratic culture and intra-organisational processes. Third, neo-institutionalism, in the form of rational-choice-, historical- and sociological institutionalism. I justify my choice for neo-institutionalism by showing that it incorporates insights from both positivist and social constructivist approaches.
Firstly, functional (IR) or rational-choice accounts (economics) emphasise that the identity and functioning of an organisation can best be understood by looking at the (rational) interests of the actors that maintain and set up the organisation (Karns, Mingst and Stiles
2015; Shepsle 2010). As such, other actors are irrelevant as they are not in positions of power that can directly challenge the rules. Consequently, a (formal) change to the competences of an organisation should be studied by looking at its function and those who control the mandate (Shepsle 2010). Functionalism is particularly adamant in its belief that “(...) organisations arise from the basic, or functional, needs of peoples and states.” (Karns, Mingst and Stiles, 2015, p. 49). Similarly, rational-choice approaches emphasise the fact that individuals seek information to decide between different options at their disposal and
ultimately pick the one that is most in their favour . From this point of view, the expansion 10
of the mandate of Europol would be understood as a manifestation of the interest of
Member-states, European Political actors and actors within Europol to enhance the operations and activities that take place at the level of the organisation (Cini and Borragán 2013). While this viewpoint can effectively explain why the mandate was set up in a certain way, it has greater difficulty explaining why parts of it failed. The differences between the crime areas are particularly puzzling: if the mandate was not achievable, member-states would have set it up differently. Before and during the expansion of the mandate, the importance of the four crime areas remained constantly high (AR 2010-2014) which eliminates the possibility that an onset lack of interest in a certain crime area explains the difference in performance. In summary, this theory is competent at explaining why formal rules are set up in the way they are, but is unable to parsimoniously explain the both subsequent deviance from these rules and its structural nature.
Alternatively there are theoretical perspectives that emphasise the importance of studying how and what individuals value beyond simple rational calculations and how this influences their actions. These approaches, based in social constructivist paradigms,
emphasise that (members of) organisations have ideological values of their own.
Organisations therefore have their own agency and agenda, which interact with the actors that maintain and set up the organisations. Furthermore, organisations (re)produce and spread certain norms, which scholars have argued to be of great influence (e.g. Barnett and
Finnemore 1997; Barnett 2010). From this point of view, the expansion of the mandate would therefore be understood as a successful attempt by the organisation to convince the relevant actors of the importance of strengthening its structural support (materialistic and legal). Simultaneously, the fact that members create norms of their own next to the demands placed upon them by the official rules helps us explain why individuals may formally bind
themselves to these rules, but contradict them in practice. This theory has its own theoretical and practical limitations, meaning I cannot apply it solely. The theoretical perspective can only analyse structural factors of an organisation (such as material factors or legal
competences) through the value that its members attach to them (Ibid.). This leads to the practical problem that we cannot infer the motivations and actions of agents beyond our knowledge of them at a particular time and place. This kind of data could be collected through extensive fieldwork, which is difficult in this specific case (see also the next section on methodology).
I will thus employ a model that seeks to combine the insights from the two perspectives above. Like the social constructivists, I argue that both the members of the organisation and those who set it up matter. Moreover, they create their own values and norms. However, they do so on the basis of their rational interests, which are to compete with the formal legislation without being caught (rational-choice theory). Therefore, I choose a rational-choice institutionalism (RCi), which is a school within the broader theoretical perspective of neo-institutionalism. As neo-institutionalism is interested in how rules and
norms generally interact with one another, identifying the types of institutions at Europol can help us better understand why the organisation did not perform as well as it could. In
particular, it can help us understand how these norms change over time and how they might be addressed by Europol.
There are three main schools of institutionalism: rational-choice; historical-; and sociological- (Hall and Taylor 1996; Lowndes and Roberts 2013). To explain RCi, I begin with the seminal work on economic exchange by North (1990) . Herein, he argues that an 11
informed viewpoint on (economic) exchange must take into account transaction costs and 12 the subsequent attempts by actors to reduce these. The central idea of institutional theories is that humans seek to establish rules and norms that limit certain types of behaviour: by limiting or placing constraints on negative behaviour, such as contract-violation or violence. North (1990, p. 4) defines institutions as:
“(...) the humanly devised constraints that structure political, economic, and social interactions. They consist of both informal constraints (sanctions, taboos, customs, tradition,
and code of conduct) and formal rules (constitutions, laws, property rights).”
As will be discussed further in my theoretical argument, the emphasis on both formal and informal constraints is key: neo-institutionalism seeks to balance the highly formalistic world of functionalism; the meaning-centered approach of social constructivism and the interaction between them (Lowndes and Roberts 2013). The theoretical logic behind
neo-institutionalism is as follows: groups of individuals that are unified by a common set of
11While North was later associated with historical institutionalism (Lowndes and Roberts 2013), I apply his
original approach by assuming that actors exclusively make rational choices.
12 The costs associated with gathering information and enforcing any agreements between an individual and a
goals (organisations) seek to constrain negative behaviour from their members. In order to reduce uncertainty that arises due to possible negative behaviour from another party, they seek to bind themselves to common rules (institutions). In a hypothetical scenario where no rules exist before, the members of organisations will begin the process of institutionalisation by generating informal institutions (such as traditions, taboos or norms surrounding good and bad behaviour) and will mutually enforce them through social sanctions. Over time and on the basis of the goals of the organisation, the number of informal institutions is likely to grow as institutions generate further institutions. Indeed, norms of behaviour require norms on what the punishment is, and subsequently who carries out the punishments and how. This leads to a complex, mutually interacting ‘rule-set’, that governs the activities of the organisation. As the size and range of activities of an organisation grows, so does the difficulty of understanding and applying the ruleset for new members. This uncertainty is then reduced by the creation of formal institutions. These are constraints that are written down and sanctioned by an authority that enforces them . Eventually, the behaviour of 13
individuals becomes subject to a complex interaction between both formal and informal institutions which discourage certain behaviour and comparatively encourage others.
Given this state of affairs, two further arguments emerge. Firstly, individuals will seek to change institutions over time: this is necessary for institutions to continue functioning in the face of contextual change. It is also in the interest of individuals to reduce constraints on actions they want to perform, or increase constraints on actions they do not want others to perform. Secondly, as actors with different interests change institutions, some will strengthen or come into conflict with one another. In practice, this means that the effect of an institution is relative to how well it is known and enforced within organisations (Helmke and Levitsky
2004). Scholars find that informal institutions can be effectively used to change the outcomes expected from agents that set formal institutional arrangements: they are difficult to pinpoint, operate through mechanisms that are difficult to monitor when one is not part of the exchange process itself and are often automatically learned by individuals when they access a new formal institutional environment (Crossland and Hambrick 2011; Lowndes and Roberts 2013; Gel’man 2004).
This leads us to the differences between the three main schools of institutionalism. While they share a common interest in the concept of institutions, they disagree substantially about their exact nature, working and origin of change. The first school of institutionalism is historical institutionalism, which takes the theory of institutions and applies it to historical accounts of organisations to understand changes in law, policies, customs, or ideologies. In particular, scholars of this tradition emphasise the notion of path-dependency, which describes the phenomenon where the interactions between institutions and the cost of changing them produces a situation whereby small incremental changes become desirable over large and abrupt changes. In turn, this means that a small change at a particular point in time can gradually exert an ever-larger effect (Pierson 2004). Due to its interest in
change-over time and its theoretical focus, historical institutionalism is most suitable for studying large-scale organisations over substantial periods of time. Moreover, it does not place strong theoretical demands on how preferences arise , preferring instead to gain this 14
understanding through historical analysis (Lowndes and Roberts 2013).
The second, sociological institutionalism argues that, rather than dictated by rationality or inferred from history, institutions can best be understood as being culturally
14See also the neo-institutionalism sub-theory of “Analytical Narratives”; which generates rational-choice
models and then tests these by applying them to historical cases (e.g. Bates et al. 1998; Thelen 1999). By doing so, many scholars consider it an approach that bridges rational-choice- and historical-institutionalism.
constructed. Scholars of this school argue that institutions serve a symbolical function and are part of an institutional culture. They also often focus on notions of legitimacy that are
constructed by organisations and their members. The driving factor behind why certain institutional arrangements take their form can therefore be hypothesised; but generating true insight requires scholars to perform qualitative research, most often within the organisations themselves (Ibid.). This limits the comparability of this type of research and practical feasibility of carrying it out.
Finally, I will now elaborate on RCi as it is the most relevant to this research.
Rational-Choice Institutionalism, as the name implies, sees institutions as the outcomes of the (rational) interests of individuals. In order to obtain the best outcomes, individuals require information and certainty that the other party will honour their side of an agreement. This requires individuals to spend time and effort to find this information and enforce the
agreement. As both can benefit from exchange, it is in their interest to reduce the costs. They therefore bind themselves common, efficient rules. As long as structural factors and
preferences remain the same, an equilibrium will emerge. This is a state where a change to institutions makes transactions less efficient, either in terms of reducing, changing or increasing the (number of) institutions.
Neo-institutionalism has previously been applied to the case of Europol, but in this case the author (Fägersten 2010) focused on the notion of asset specificity to explain how 15 the exchange of sensitive data in the context of counterterrorism was resisted by bureaucrats on the grounds that they should maintain control over the information both because they had invested most into its current state and could use it most effectively. Institutionalism has been successfully applied to studying democratic transition (North, Wallis and Weingast 2013),
15A concept from economics that states that certain goods are only valuable in a context where extensive
Corruption in the (former) USSR (Gel’man 2004) and inequality (Savoia, Easaw and McKay 2009) emphasising its explanatory power. In particular, neo-institutionalism excels at
explaining why certain negative trends occur, even when the rules and preferences of actors seem to be in favour of a more positive trend (North 1990; Lowndes and Roberts 2013)
I choose RCi because it combines valuable insights from the dominant perspectives on international organisations and the most applicable of the three neo-institutionalisms. It allows us to make assumptions about the interests of actors before beginning the analysis. While the overall approach has been applied to the organisation before, I have not found literature that discusses the specific interaction between the formal and informal institutions of Europol. This in spite of the literature backing up the general strength of informal
institutions and the Disley-report (2012) specifically hypothesising that cultural norms affect the information exchange of Europol. I will now discuss the transition from this theory to my analysis.
III. Argument and Research Design
In this section I discuss the transition from the previously discussed theory of RCi to applying this in practice on Europol. This is used to generate a systematic approach for studying the effects of informal institutions (at Europol).
III.A: Theoretical Argument
My theoretical argument is that actors generate informal institutions when they cannot directly contest formal institutions that are in their favour. As discussed above, RCi predicts that observed outcomes must be in line with actors trying to maximise their respective utility given the institutional constraints (Helmke and Levitsky 2004; North 1990; North, Wallis and Weingast 2013). Institutions that are not in the interest of the actors that engage with them are
either removed or countered by another institution. I will first present my overall argument and then discuss the specific components in greater detail. I argue that Europol and the ENUs have different preferences with regard information exchange. The ENUs are either not
convinced of the benefit of exchanging the information, or are directly opposed to sharing it. However, the ENUs cannot openly state this, as they are required by European legislation to exchange the information. In order to protect their interest, members of the ENU generate a system of informal institutions that fulfill three criteria: a) they are comprehensive enough to provide guidance for members of their organisation when they interact with Europol and b) they are not transparent to an extent where (members of) Europol or their political oversight can identify and tackle them directly; and c) they can be adapted per context. In order to have a comprehensive understanding of how the informal institutions interact with the formal institutions that pressure ENUs to give information to Europol, I look at four categories (Lowndes and Roberts 2013): 1) in what form they are present (Characteristics); 2) how the rules change on the basis of the type of information (Data Value); 3) where they apply in the organisations (Domain); and 4) how they are enforced (Enforcement).
The first category (Characteristics), covers the mechanism of the informal institutions and which types are present. Here, I expect that, if informal institutions are present, actors mainly generate those that downplay the benefits of sharing their information with Europol. However, there are other possible forms for informal institutions. Some may directly compete with the formal institutions, whereas others may increase the disadvantages or reduce the advantages of sharing particular information by the ENU.
Given that information is a key commodity for national police forces, I argue that they will seek to maximise the relative amount of information they have (Bures 2016). This is a two-way process, where ENUs demand information from europol and give little in return.
However, countries cannot openly justify this behaviour as as Europol is generally seen in a positive light. As such, they generate norms that reduce the costs of acquiring information from Europol, and minimise their capabilities of giving information back to Europol. By appealing to commonly found demands for increasing efficiency, nations can justify spending fewer resources on sharing their information with Europol.
The second category (Data Value) shows which institutions are general in the ENUs, and which are specific to information types tied to different crime areas. I expect that, the higher the value of the data, the less likely it is to be shared with Europol. Therefore, the norms covering more sensitive crime areas will be more limiting. As stated in the literature review, Europol has different agreements depending on what information third parties wish to share with them. “Information” is defined as intelligence and data relating to criminal
investigations (AR 2011) and can have different levels of sensitivity. In line with Europol regulations and the strategic and operational agreements, a distinction should be made between non-personal (e.g. specific technical knowledge or the whereabouts of particular criminal goods) and personal (home address, private correspondence, etc.). The first difference in sensitivity can be understood in line with the operational and strategic
agreements: personal data is protected to much greater extent under law than non-personal data (AR 2010). The second difference is that information relating to salient crimes, such as those that are politically charged, may not be shared easily due to ENUs wanting to address the issues themselves. Sharing allows other parties to use the information to take credit and pursue their own goals. This fear is exacerbated due to the fact that Europol lacks any formal enforcement mechanisms. This means that, while ENUs have to answer their nation states, Europol can use their lack of executive power as a shield against blames if mistakes are made (Fägersten 2010). On a final note, there is evidence to suggest that information can be used as
a ‘currency’ that can be exchanged for other information or political favours (Bossong 2008). Highly valuable in information might therefore be kept by ENUs for bargaining with other organisations, as shared information cannot be exchanged for favours.
The third category (Domain) shows whether the institutions vary on the basis of hierarchy within the organisation. I therefore expect different norms to apply at different dimensions, as actors within ENUs (vertical) have more direct control over one another than between ENUs (horizontal; as part of bi- or multi-lateral cooperation). This is because actors within ENUs are part of the same police force, and are therefore subject to chain of
command. This means that powerful actors within national police forces can back up their informal institutions through their authority, which they lack in the context of the horizontal dimension. Moreover, even though the formal demands placed on ENUs by Europol are the same for all countries, the ENUs are part of different national police forces and are contained in political, judicial and administrative frameworks that vary extensively (Bures 2016). This means that the informal institutions of ENUs are adapted to these frameworks . I thus also 16
expect that ENUs thus create norms that refer to these differences to create plausible deniability: they can blame a failure to share particular information on a peculiarity of their system as they understand it better than the vast majority of the staff at Europol.
The fourth category (Enforcement) helps us understand the strength of the informal institutions, as it shows how much (non)compliance affects the utility of actors within ENUs. Informal institutions “ are enforced through social processes such as hostile remarks, gossip, ostracism, and other displays of social disapproval (Helmke and Levitsky 2004, p. 733)”. I therefore expect that they derive their strength from appealing to norms that are present in the national police forces, not just the ENU, such as loyalty to the own police force and a drive
16Also due to the fact the higher ranks of national police forces have much more direct forms of oversight and
for increased efficiency. This is due to the fact, that it is more difficult to only use these social processes in a professional and rule-bound environment. On the other hand, actors within ENUs are rewarded for pursuing and strengthening norms that are valued by the national police forces as a whole. These enforcement mechanisms are generally difficult to observe for outsiders as they are often ‘subtle, hidden or even illegal’ (Ibid.)”.
Overall, I expect that an important contributor to how limiting the informal
institutions are is the length of the relationship between ENUs and Europol. This due to the fact that a longer relationship allows them to build more trust. Moreover, more time provides more opportunity for positive experiences, which likely counteract the argument that
exchanging information with Europol is not worth the effort or a violation of norms of loyalty or efficiency.
III.B: Operationalisation
In order to properly establish the effect, if any, of informal institutions, a step by step process is required (Helmke and Levitsky 2004; Voigt 2013; Lowndes and Roberts 2013). This is particularly true for informal institutions, which are harder to pinpoint due to their unwritten and socially, as opposed to formally, sanctioned nature.
In order to isolate the effect of informal institutions, it is first necessary to establish what behaviour would be observed if only the formal institutions are followed. If a deviation is found, this means that the formal institution is either unclear or lacking in de facto
implementation, due to or providing room for an informal institution countering it (Voigt 2013). Second, the nature of the deviation must established in practice: is it structural in nature?; does it cover an identifiable interest?; and which actors deviate? (Helmke and Levitsky 2004). The specifics of each of these factors show whether the deviation is due to
competing institutions or simply a pattern of behaviour. The difference lies in whether there are shared expectations between the actors and whether they reciprocally enforce the
competing institution. Consider for example the reasons a group of workers may not do their work: they might be indifferent combined with insufficient monitoring or enforcement of the rules. In this case, this would be a pattern of behaviour. If, on the other hand, the workers would refuse because they are on a strike, then this behaviour is on the basis of shared social expectations (and likely also mutually enforced, where group members would sanction those who consider giving in to the management). The implications are considerable: whereas the former might be resolved by a simple change to the monitoring or enforcement of the rules, addressing the latter requires dealing with the additional set of social expectations and enforcement. Once a clear picture emerges of the interaction between formal and informal institutions, this allows us to see which types of informal institutions are present
(characteristics).
Given the literature on Europol discussed so far, two of the analytical steps have been fulfilled. First, there is an observed deviation from a well-defined rule (i.e. the ECD 2009). Second, the deviation varies structurally per crime area, covers the identifiable interest of information exchange; and is consistently performed by the ENUs. The characteristics can be operationalised by looking at whether ENUs emphasise that the benefits do not outweigh the costs or have automatic information exchange in place or whether they limit this. The second category can be studied by looking at the relative amount of cooperation per crime area (e.g. amount of information shared, participation in shared projects or task forces). The third category can be operationalised by looking at the strategic partners of member states, and whether they are (both) members of competing information networks (e.g. the Budapest club or ISTC (Svendsen 2011)). The final category can be examined by carefully studying the
findings of evaluators and what (former) employees of ENUs describe when discussing their work environment. This can be used to generate a holistic understanding of how strictly the norms are applied and what the types of sanctions are that play a role. I will now discuss my approach data-selection and evidence to complete my toolset for doing the analysis.
III.C: Data-selection and Evidence
Due to the intangible nature of informal institutions, I use evidence in two ways: either, it strengthens the argument directly, or weaken alternative explanations. I will now discuss the data-selection and the possibility of triangulation.
As stated, evidence for informal institutions is either found directly or by comparing the observed behaviour to the expectations on the basis of formal institutions and structural factors. While it would be preferable to measure the interests of the actors directly, there are two barriers: there is no way to measure the thoughts of actors; and even if I could ask directly, actors have strategic and personal reasons to present biased information (Carapico and Trauner 2013). As such, there are two types of evidence that I use. Firstly, document review takes place on the basis of the Annual Reviews, Publicly Available Europol Data ; 17 and the multi-annual strategy. Indeed, despite some likely bias, the ARs provide a systematic historical overview of the growth of the organisation, its changing priorities and relationships. Moreover, each AR provides comparison with previous years which helps see the changes in performance over time. Given that the annual review for 2017 is already out, this provides three years of data beyond the cut-off point for this research. Furthermore, the multi-annual strategies of Europol and European political actors shows how they sought to address the effects of informal institutions.
The second type of data that will be used is academic and professional literature, which serves two goals: firstly, to expand to scope and control for bias that would be inherent in only using documents from Europol itself. Secondly, to provide better insight into the specific type of information and its usual sensitivity per criminal area.
Originally, I intended to interview analysts that work at Europol, ideally with one representative from each of the four crime areas. They would reflect on the tentative
outcomes and indicate their (dis)agreement. The exchanges would have covered a reflection on overall theory, the specifics of their area of crime and the tentative findings specifically related to their area. Unfortunately, it was not possible to secure a meeting with any analysts. Future research should consider addressing this gap.
For the two data-types, my analysis is guided by a qualitative approach, with an attempt to weigh information from both data-types equally. In particular where disagreement between pieces from both data-types is found, I will seek to find more pieces from either category to establish which is most applicable and why. Where data is lacking, I will identify this clearly. Needless to say, the specific research goal of this thesis makes it unlikely that I find extensive evidence in all categories. As argued, the evidence for many of the informal institutions is hidden by the actors themselves, and often confidential or strategic in nature. Nevertheless, the different sources of data allow me to test the relative strength of my arguments, and can at the very least provide insights for future research into the subject.
IV. Analysis & Results: Institutional Characteristics; Data Value; Domain and Enforcement.
This part of the thesis will present the analysis and results in the order they were introduced in the theoretical argument.
IV.A: Characteristics
As discussed during the operationalisation, establishing whether there are any informal institutions present requires that the observed deviation from the formal institutions is on the basis of shared expectations between actors. The first line of evidence that there are shared expectation within and among national units comes from the fact that many actors involved with Europol felt that the expansion of the mandate increased the political control by the European Union, at the cost of the member states: “Now you already see that the European Commission has instructed Europol to follow the same list of countries which are on the political list of European Commission … the operational needs are no longer [the prime consideration] … we will … have agreements with countries which are much less relevant for operational … cooperation.” (From Disley et al. 2012, p. 42). This has led to resistance by actors who feel the member states should retain their primacy (Ibid., p. 40). Moreover, it is noted by both the Disley-report and in several ARs (2011; 2013) that Europol has a distinct multi-faceted culture of its own and that it tries to spread and establish best practices. One of the main ways in which it does this is the Europol Platform for Experts (EPE) which is a web platform that facilitates “sharing of best practices, documentation, innovation, knowledge and non-personal data on crime” (AR 2014, p. 34). This push has led to a counter-effort from member-states, who remained unwilling to grant any executive power to Europol. As a member of the Europol directorate noticed, there is a lot of scrutiny from the member-states on Europol, but not vice versa (Disley et al., p. 59). However, the report also notes that Europol communicated information about the ‘strengths and weaknesses of the relationship between Europol and particular countries’ (Ibid.). This means that ENUs need to have clear internal lines of defense in place. All of these factors strengthen the notion that the variation
in success is not simply due to patterns of behaviour. ENUs have specific interests that they need to defend in face of political and legal pressure from Europol and other parties. There is also evidence (Ibid.) that certain member states have specific informal institutions that emphasise that exchanging information is not worth the effort. This is due to the fact that, when Europol marketed its successes, clients were willing to share more information. The highly rule-bound nature of European Police Work (Deflem 2009), combined with the pressure that emerged due to the new environment therefore led countries to generate systematic responses and resistances in favour of their own interests (Casson et al. 2009). While ENUs likely also employ common forms of resistance such as foot-dragging,
performing additional but unnecessary tasks to stall, I am unable to find any specific evidence for this. Having established the evidence for the presence of specific informal institutions, I will now discuss the relationship between the rules and the particular type of information.
IV.B: Data Value
The information that is exchanged and shared with Europol covers a wide range of interests and levels of sensitivity. There is evidence to suggest that the different crime areas have different levels of sensitivity, including the AR (2016, p. 7) directly stating that : “It [the year of 2016] also provided impetus for Member States to cooperate on sensitive issues, such as those related to terrorism or irregular migration”. As identified in the literature, this means that information relating to counterterrorism (which is often personal and politically
sensitive) is most protected, whereas drug-related information is the least protected. Human trafficking and financial crime fall in between: the more personal nature of human trafficking information compared to financial crime helps us explain why the former was still faced with significant challenges. These helps us explain the different rates of performance in these
areas. The differences in sensitivity can also be seen when looking at the number of searches and objects per category. In 2014 (AR), nearly half the data-objects in the EIS were related to drug trafficking (29%) and robbery (20%), with few objects on human trafficking (5%) and no mention of counterterrorism. While there is some information on counter-terrorism (Ibid., p. 31) there is no indication of the relative scale compared to other crime areas, nor how many cases were handled by competing authorities. Finally, the majority of operations mentioned in the annual reviews are related to drugs, followed by general organised crime (AR 2010; 2011; 2012).
IV.C: Domain
In order for institutions to have their effect, it must be clear to those they affect where they apply. The structure of Europol generates two types of relationships where actors from ENUs generate their norms: vertically (within the ENU) and horizontally (between ENUs).
The first domain is the hierarchy within the ENU (vertical). The national units are usually located at the national police headquarters of the respective country (Disley et al 2012 p.54). Moreover, each one has a head (HENU) and at least one liaison officer, the vast
majority of whom are stationed at the headquarters of Europol in the Hague (AR 2011; 2012). This means that the largest part of an ENU is based in the member-state, with only the liaison staff available for direct contact with Europol. Furthermore, Europol does not have formal power over liaison officers, who instead fall under their national laws. This has a number of effects: ENUs answer firstly to their own police force and then secondly to the requests from Europol (Fägersten 2010). In fact, it is found that ENUs feel primary loyalty to their national police force (Den Boer 2015). This also strengthens their preference to share less than more (Bures 2016). Furthermore, the Disley-report (2012) found that some interviewees had
complaints about the seniority and reputation of HENUs: “The national unit … is an
obligatory gateway for Europol, which is good – but it should be strengthened. So it would be more helpful for Europol if the role of the national unit at national level would be stronger, so that they can fulfil their role better back home (p. 59). This echoes the alternative
explanation discussed in the literature review that greater centralisation of the national police force can strengthen the position of the ENU. On the other hand, it is also possible that the lack of seniority of the HENU is used as an excuse by the national units to resist requests from Europol. Due to not having the chance to speak with analysts at Europol, I can not definitively show this either way.
The second domain is the horizontal relationship between the ENUs. As covered in the literature review, Europol allows nations to share information directly on a bilateral basis (meaning that it does not end up in the Europol database or AWFs). Given the value of information and the desire of ENUs to limit the access to their information where possible, many nations prefer to either share the information bilaterally, or via ‘competing’ (informal) networks such as the budapest club, the Counter Terrorism Group or INTCEN (Den Boer 2015, Svendsen 2011). The fact that the number of these networks and informal arrangements themselves respond to the sensitivity of the information, combined with the preferences of the ENUs therefore help us explain the variance between the different crime areas. Not only is Europol unable to establish a progressive working relationship with the ENUs, but depending on the crime area, they are undermined by the competition.
However, there is an opposing trend in the form of an ever increasing networking that crosses conventional boundaries: the aforementioned EPE puts over 6000 experts in contact with one another, meaning that it becomes more difficult for ENUs to monitor how members of their staff communicate with others national representatives. Moreover, in 2011 (AR, p. 69),
Europol introduced three regional support officers (RSOs) whose task is to ‘(...) strengthen dialogue with Member States’ competent authorities (...) namely: western Europe, north-east Europe and south-east Europe”. Finally, the annual convention of police chiefs which grew consistently in popularity, also represents a forum wherein new connections can be made by actors that are not directly under the control of the ENUs. For this reason, the relative strength of the informal institutions that limited the information they shared on the basis of their horizontal relationships has reduced over time. Having discussed the domain of the informal institutions and to whom they apply; I will now turn to the final section of analysis.
IV.D: Enforcement
On this topic, evidence is particularly scarce: these processes are not discussed neither in the ARs (2010-2014), nor in the User Survey (2016) and when interviews are carried out on the subject, interviewees either refuse to comment or keep their explanations purposely vague (e.g. by stating ‘it’s not a conspiracy’ (Disley et al. 2012, p. 50)). While it would be preferable to have insight into how these processes work as this might allow scholars to explain individual cases, the absence of evidence in favour of one enforcement mechanism over another does not necessarily harm my overall argument: the behaviour shown by ENUs is in line with other identifiable components of informal institutions. Moreover, the different forms of enforcement would not influence the aggregate effect of the informal institutions on the levels of information exchange across the different crime areas.
V. Conclusion
This thesis has sought to address the issue of the variance in the performance of Europol during its first mandate (2010-2014) by providing a systematic analysis on the basis of RCi. Over the course of the thesis, three main points have been pursued. First, that the quality of information sharing plays a key role in explaining this variance. Second, that this difference in quality is due to interests of actors, and not due to practical barriers. Finally, that the way in which actors pursue these interests can best be understood through the notion of RCi by looking at formal and informal institutions.
The literature review was used to establish a dual gap in the literature on current explanations. Firstly, that a change in preference would manifest itself in the form of a different mandate, rather than in the current mandate failing. Secondly, that the explanations could explain general success and lack thereof, but failed to address the systematic nature of the variance. Ultimately, the results were partially inconclusive, but generally in favour of my line of argument. Moreover, on the basis of the Disley-Report (2012), Europol documents and Academic sources, I did not find any strong evidence against my argument. There were no clear indicators that a possible alternative explanation was comparatively stronger. Nevertheless, due to the lack of access to internal sources and the nature of the informal institutions themselves, I cannot make definite claims about the informal institutions and their influence on the quality of information exchange. Ultimately, my results show that first, ENUs create norms that allow them to decide what information to share with Europol. Second, that they do so on the basis of the value of data. Third, that there are several types of norms that occur at a vertical level (within the ENUs) and others that hamper the information
exchange at a horizontal level (between ENUs). The strength of of the latter has declined. Finally, I do not have definitive results in the fourth category.
Given that both scholarship and evaluations identified cultural norms and the sensitivity of data as key barriers to the full exchange of information, I therefore conclude that applying the theory of informal institutions has given us insight into the specific mechanisms and methods that were employed by the ENUs to evade the formal demands placed upon them. While extensive policy recommendations on the basis of these results are beyond the scope of this work, I make a number of points. Firstly, given the intangibility and complexity of the norms that govern the information exchange, a solution to the challenges faced by Europol is likely not found by an exclusive focus on strengthening its formal rules. Rather, the process of marketing the successes should continue, as this makes it easier for ENUs to begin the exchange of information. This generates a positive process of building trust. Secondly, a fruitful method for addressing the informal institutions is the process of strengthening relations between members of ENUs and Europol. Examples thereof are regular meetings of police chiefs, exchange of expertise and the facilitation of the exchange of personnel. Finally, as the organisation is young in its current form, addressing the informal institutions may simply take more time.
Future research should address that Europol currently performs worse in Financial Crime (Paravicini 2018) and performs significantly better in the context of counter terrorism (Ring and Wild 2018). It should take the insights generated by my theoretical framework and test these in specific areas of crime and cases. Furthermore, field-work or interviews with (ex-)staff should be carried out to triangulate the results that I have found. Finally, the argument should be applied to similar organisations to see to what extent the results are unique for Europol or hold more generally.
VI. References
Barnett, M. and M. Finnemore. (2004). Rules for the World: International Organisations in Global Politics. Ithaca: Cornell University Press.
Barnett, M. (2010). “Social Constructivism”. Chapter in The Globalization of World Politics. 5th revised Ed. John Baylis and Steve Smith. Oxford: Oxford University Press.
Bates, R. H., A. Greif, M. Levi, J-L. Rosenthal. (1998). Analytical Narratives. Princeton: Princeton University Press.
Bossong, R. (2008). “The Action Plan on Combating Terrorism: A Flawed Instrument of EU Security Governance”. Journal of Common Market Studies, 46 no. 1. pp. 27-48.
Bures, O. (2016). Intelligence Sharing and the Fight against Terrorism in the Eu: Lessons Learned from Europol. European View, 15, 57-66.
Carrapiço, H. and F. Trauner. (2013). Europol and its Influence on EU Policy-making on Organized Crime: Analyzing Governance Dynamics and Opportunities. Perspectives on European Politics and Society, 14(3), 357-371.
Casson, M. C., M. Della Giusta and U. S. Kambhampati. (2009). Formal and Informal Institutions in Development. World Development, 38(2), 137-141.
Cini, M., and N. P.-S. Borragán. (2013). European Union Politics. 4th ed. Oxford: Oxford University Press.
Crossland, C and D. C. Hambrick. (2011). Differences In Managerial Discretion
Across Countries: How Nation-Level Institutions Affect The Degree To Which CEOs Matter. Strategic Management Journal, 32, 797-819.
Deflem, M. (2006). Europol and the Policing of International Terrorism: Counter-terrorism in a Global Perspective. Justice Quarterly 23(3), 336-359.
Deflem, M. (2009). “Bureaucratisation”. In The Sage Dictionary of Policing. Eds. Wakefield, Alison and Jenny Fleming. London: Sage Publications.
Den Boer, M. (2015). Counter-Terrorism, Security and Intelligence in the EU: governance Challenges for Collection, Exchange and Analysis. Intelligence and National Security, 30(2-3), 402-419.
Disley, E., B. Irving, W. H., B. Patruni. (2012). Evaluation of the
Implementation of the Europol Council Decision and of Europol’s Activities”. Published by RAND Europe.
Europol Council Decision (ECD). (2009). “COUNCIL DECISION of 6 April 2009: establishing the European Police Office (Europol)” - (2009/371/JHA).
*Europol. (2010-2017). Annual Review. Published annually by the European Police office: the Hague.*
Europol. (2010b). Europol Strategy 2010-2014. The Hague: European Police Office.
Europol. (2016). Europol Strategy 2016-2020. Luxembourg: Publications Office for the European Union.
Europol. (2018). “About Europol”. Available online:
<https://www.europol.europa.eu/about-europol>. Accessed 05/03/2018.
Europol User Survey (EUS). (2016). Available online
<https://www.europol.europa.eu/publications-documents/europol-user-survey-2016>.
Fägersten, B. (2010). Bureaucratic Resistance to International Intelligence Cooperation - the Case of Europol”. Intelligence and National Security, 25 (4), 500-520.
Gel’man, V. (2004). The unrule of law in the making: The politics of informal institution building in Russia. Europe Asia Studies 56, (7), 1021-1040.
Hall, P. A., and R. C. R. Taylor. (1996). Political Science and the Three New Institutionalisms. Political Studies, 44(5), 936-957.
Helmke, G. and S. Levitsky. (2004). Informal Institutions and Comparative Politics: a research agenda”. Perspectives on Politics 2(4), 725-740.
Karns, M. P., K. A. Mingst and K. W. Stiles. (2015). International Organizations: the Politics and Processes of Global Governance. 3rd Ed. Lynne Rienner Publishers: Boulder.
Lowndes, V. and M. Roberts. (2013) Why Institutions Matter: The New Institutionalism in Political Science. Basingstoke: Palgrave Macmillan.
North, D. C. (1990). Institutions, Institutional Change and Economic Performance. Cambridge: Cambridge University Press.
North, D. C., J. J. Wallis and B. R. Weingast. (2013) Violence and Social Orders: A conceptual Framework for interpreting recorded Human History. Cambridge: Cambridge University Press.
Paravicini, G. (2018). “Europol is losing the fight against dirty money”. Politico. Available online
<https://www.politico.eu/article/europe-money-laundering-is-losing-the-fight-against-dirty-m oney-europol-crime-rob-wainwright/>.
Pierson, P. (2004). Politics in Time: History, Institutions and Social Analysis. Princeton: Princeton University Press.
Ring, S. and F. Wild. (2018). “Inside the Google of counterterrorism”. Bloomberg. Available online
<https://www.bloomberg.com/news/articles/2018-04-05/once-snubbed-by-fbi-europol-is-now -google-of-counter-terrorism>.
Rozée, S., C. Kaunert and S. Léonard. (2013). Is Europol a comprehensive Policing Actor?. Perspectives on European Politics and Society, 14(4), 372-387.
Savoia, A., J. Easaw, and A. McKay. (2009). Inequality, Democracy and Institutions: A Critical Review of Recent Research. World Development 38(2),142-154.
Shepsle, K. A. (2010). Analyzing Politics: Rationality, Behavior, and Institutions. New York: W.W. Norton Company.
Svendsen, A.D.M. (2011). On ‘a Continuum with Expansion’? Intelligence Co-operation in Europe in the Early Twenty-first Century. Journal of Contemporary European Research, 7(4), 520-538.
Thelen, K. (1999). Historical Institutionalism In Comparative Politics. Annual Review of Political Science, 2, 369–404.
Voigt, S. (2013). How (not) to measure institutions. Journal of Institutional Economics 9(1), 1-26.
