The eu’s Legitimacy in the Eye of the Beholders

Chapter 3 – On Legitimacy, Democracy and Governance

[There is an] inherent tension between the rules for representational democracy and the empirical practice and implicit rules of governance networks[, which] challenges the hegemony of the state and, by implication, the construction of ‘the people’ as the self-regulating sovereign.

Erik-Hans Klijn and Chris Skelcher³⁷
In this chapter, a multi-faceted understanding of legitimacy will be developed. We have seen that legitimacy of the nation-state is based on two versions of republicanism, but since then ideas on democracy have moved on and circumstances have changed. The EU wants to be a legitimate democratic order that uses modern governance techniques. The problem is that legitimacy, democracy and governance are what Bekkers et al. refer to as ‘fuzzy concepts’, which like most central concepts in social science have no clear meaning, but an inherently positive connotation (2007: 5). Before unpacking democracy and governance in relation to legitimacy, we will first take a look at public administration’s usual definition of legitimacy.

3.1 – Public Administration on Legitimacy

In the previous chapter, legitimacy was defined as a consensus, which diverges from the usual ‘public administration definition’. Legitimacy is more often defined as a threefold concept. Consisting of legality, rightfulness or normative justifiability, and finally, shared acceptance of the process or legitimation (e.g. Beetham and Lord 1998: 4-5; Bekkers et al. 2007: 40-41; Bursens and Baetens 2004: 4; Thomassen 2007: 6).

The first condition of legality is that a political order is legitimate if ‘political authority is acquired and exercised in according to established rules’ (Bursens and Baetens 2004: 4).³⁸ C.W. Morrison makes an interesting addition when he stresses that decisions have to be ‘recognized as lawful, just or rightful’ (1998: 37; Emphasis added). Secondly, there is normative justifiability, which refers to the ideological context in which a political order takes decisions. A political order must govern in accordance to broadly accepted social beliefs (Bursens and Baetens 2004: 4). Thirdly, there is legitimation as the explicit expression of approval and confirmation of legitimate positions of powers by power-subjects and other legitimate authorities (Beetham and Lord 1998).

Legitimation comes closest to our definition, but approval need not be explicit. One might even argue that tacit consent implies stronger legitimation, because legitimacy surfaces either when it is threatened (Bernard 2001: 26)³⁹ or nonexistent, as is too some extent the case with the EU, in both cases legitimacy is weak. On the other hand, when a political order is under threat from external dangers it is strong, as we saw in the aftermath of 9/11 in George W. Bush’s approval ratings.

More importantly, this threefold concept hides a division between objective and subjective legitimacy. Legitimacy can be objectively determined by analysing if it meets standards – legal or normative – or whether it is perceived as right by the relevant beholders. In this research, the idea of objective legitimacy of political order is dismissed as unrealistic, because a state cannot be objectively legitimate if it is not perceived as such: no legitimacy without perception thereof. Therefore this conception of legitimacy is rejected.⁴⁰

3.2 – Legitimacy and Representative Democracy: Components

Democracy⁴¹ is what Deborah Stone calls a ‘motherhood issue’: ‘In the abstract every one supports them, but the struggle begins when we start to ask what they mean’ (2002: 12).⁴² There are numerous definitions, accounts and classifications of democracy circulating in academic literature.⁴³ Founding Father of the United States of America Abraham Lincoln eloquently described the basic principle of a democratic political order in the Gettysburg address as: “… government of the people, by the people, for the people” (In: Friedrich 1974: 37).

When in discourse ‘democracy’ is mentioned it often refers to constitutional democracy (Friedrich 1974).⁴⁴ Further, democracy almost by definition means one specific institutional framework: representative democracy (Bekkers et al. 2007: 57; Klijn and Skelcher 2007; Thomassen and Schmitt 2004: 337). Although the exact institutional framework differs per country, the central idea is that public (government) offices should be filled by elected representatives. The authority is then exercised in accordance with judicial principles in order to protect individual rights – here constitutional democracy comes into play. In order to determine in how far this traditional idea of constitutional representative democracy influences the current discourse on the EU’s legitimacy a components framework of legitimacy is developed.

3.2.1 – Scharpf’s and Bernard’s Components of Legitimacy

One of the most influential accounts of the components of legitimacy is given by Fritz Scharpf (1999). In Governing in Europe: Effective and Democratic,⁴⁵ he distinguishes between two components of democratic self-determination: input and output legitimacy, which he describes as follows:
Input-oriented democratic thought emphasizes ‘government by the people’. Political choices are legitimate if and because they reflect the ‘will of the people’ – that is, if they can be derived from the authentic preferences of the members of a community. By contrast, the output perspective emphasizes ‘government for the people’. Here, political choices are legitimate if and because they effectively promote the common welfare of the constituency in question (Ibid., 6; Emphasis in original).
Input legitimacy is a combination of two identities. The first identity, which is often overlooked in accounts of his theory, is a ‘thin’ political identity. This identity is weak and therefore there are the dangers of the tyranny of the majority and premises of ‘normative individualism’ and as a result a breakdown of legitimacy becomes highly likely (Ibid., 7-10). This brings us to Scharpf’s second and most famous notion: the thick identity. People need to believe in the sameness between the members of the community⁴⁶ in order to accept state rulings that are in their disadvantage.⁴⁷ Here we return to the nationalist republican legitimacy based on the idea of the nation. Scharpf (implicitly) distinguishes between the ‘demos’ (thin) and das Volk (thick), but as Yack and Hont argue this division has become conceptual rather than actual.⁴⁸

Output legitimacy is orientated on the performance of the system. Scharpf distinguishes four different output legitimizing mechanisms: electoral accountability, independent expertise, corporatist and intergovernmental agreement, and pluralist policy networks (Ibid., 13-21). Note that these mechanisms are not actually outputs of the system, but demands made to the system and the way it produces outcomes. ‘[T]he legitimizing force of these mechanisms tends to be more contingent and more limited than is true of identity-based majority democracy’ (Ibid., 11). Scharpf thus distinguishes between two components of democratic legitimacy:

Two Components of Legitimacy
1	Input	Identity
2	Output	Interest

Figure 3.1: Scharpf’s Components of Democratic Legitimacy
In discourse democracy often refers to constitutional democracy, therefore it is necessary to distinguish a third component next to the input-output legitimization: throughput legitimacy. Frederick Bernard’s account of the components of legitimacy is useful in this regard. He describes legitimacy as follows:
The concept of legitimacy as a political concept is inherently complex in that it involves at least three distinguishable components: the electoral (constitutional) right to rule; the procedural (normative) rightfulness in the exercise of rule; and the substantive (teleological) rightness in the ends of rule (2001: 26).
These three components together form the legitimacy of a political order. Bernard seems to define input as constitutional, hence judicial, and throughput as normative, but this is somewhat misleading terminology. The constitutional refers to the electoral process and popular mandate, and the normative to the procedural judicial rules of the game. Finally, the output refers to the actual performance of the political system.

Three Components of Legitimacy
1	Input	Right to Rule
2	Throughput	Exercise of Rule
3	Output	Rightness in the End of Rule

Figure 3.2: Bernard’s Components of Legitimacy
Bernard’s framework adds the output legitimacy to the different components. This might come as a surprise, because it seems like Scharpf already distinguishes between input and output. In my view there is a common misinterpretation of Scharpf’s theory in the public administration literature. In the literature, for example Bekkers et al. (2007), outcome is defined as performance, but as we saw Scharpf actually makes procedural demands. The added value of Bernard thus lies in that he distinguishes substantive performances – output – from procedural demands – throughput. Scharpf, on the other hand, also adds to Bernard’s account with the distinction in the input between thin and thick identities. In other words, an identity based on popular support/mandate (thin) and one based on cultural identification (thick).

3.2.2 – Combining the Accounts into a Framework

We will merge these two accounts into a single ‘components framework’. The different components of the legitimacy of a (democratic) political order one can distinguish are input, throughput and output. In order to steer free from any confusion: the following is not based on Easton’s policy model, even though that model also distinguishes between input, throughput and output legitimacy (e.g. Bekkers et al. 2007; Thomassen 2007).

Let us now walk through the different components. The input category combines two important elements: identity and support. Identity encompasses Scharpf’s thick identity (or at the least the feeling thereof).⁴⁹ The thin identity is political in nature and therefore related to popular support for the system (Friedrich 1974). In a democratic system this is closely related to electoral mandate. Thus Scharpf enhances Bernard’s account with identity claims and Bernard Scharpf’s with the idea of popular support to give more body to the thin identity.

The throughput is an interesting component, because what Scharpf classifies as output legitimizing mechanisms all fall into this category. In this framework, throughput legitimacy is defined as how rule is exercised in order to achieve outputs.⁵⁰ In democratic, pluralist societies how output – performance – is achieved is theoretically more important than its actual aims/performance, because the question: ‘what is good?’ is seen systemically unanswerable in a society with diverse value systems.⁵¹

Still actual performance – output – matters, because we might not agree or even know why something is good or evil, some acts are evaluated by a large majority of society as such (e.g. public opinion). Government is expected to perform services for its people. A negative example is the Vietnam war that almost brought down the USA’s entire constitutional democratic system (Friedrich 1974). Performance matters, at least more than one might expect from idealistic accounts of democratic legitimacy. In this category fall the interests of the people and the state, and the state’s ability to achieve these interests. Thus, the framework developed for this research looks as follows:

Three Components of Legitimacy
1	Input	Identity & Support
2	Throughput	(Legal-)Procedures
3	Output	Interests & Performance

Figure 3.3: Components of Legitimacy
Let us see how this framework works out with regards to the conceptualisation by people of the legitimacy of representative constitutional democracy. Democratic theory accepts that there cannot be complete agreement among the population on values and interests, but on the basis of a common identity and representatives with a popular mandate – input – and following certain (judicial) procedures – throughput – people will accept the outcomes of the process. If people thus judge a political order on its democratic legitimacy they will focus on the input and throughput (Nispen and Posseth 2007: 223).

The distinction in the input – thick-thin – is artificial, but stressed because it becomes useful in the analysis of the republicanisms in discourse later on. Democratic theory though is as much a child of both republicanisms as the nation-state’s legitimacy and it does not actually distinguish between universalist and nationalist dimensions of the components.