Anti-System Parties Revisited: Concept Formation and Guidelines for Empirical Research

Abstract

On the occasion of the fiftieth anniversary of the concept of anti-system party it is time to ask whether it enjoys good health in addition to longevity. Reflecting on what constitutes an anti-system party appears to be of unprecedented relevance, particularly in the light of the electoral success of populist parties such as the French Front National, the Five Star Movement in Italy and Syriza in Greece. This article highlights two crucial questions that remain unsolved if we follow existing conceptualizations: What are the boundaries of the concept? When does a party cease to be anti-system and how can it be reclassified thereafter? In order to overcome such limitations, this article develops a revisited concept of anti-system party and provides a set of guidelines for its empirical application. Furthermore, a novel typology capable of investigating the evolution of anti-system parties and classifying political parties in general is presented.

The first usage of the concept of anti-system party in a comparative perspective can be traced back to a seminal book chapter by Giovanni Sartori published in 1966, and on the occasion of its fiftieth anniversary it is time to ask whether the concept enjoys good health in terms of theoretical consistency and analytical power in addition to longevity. This article aims, firstly, to review the existing conceptualizations critically by emphasizing their theoretical shortcomings and the subsequent problems for empirical analysis. Secondly, it provides a minimal definition of a revisited concept of anti-system party by focusing on two constitutive dimensions: the ideological orientation of a party towards the status quo and its visible systemic interactions. Although my revisited conceptualization differs from classical approaches to anti-system parties (Capoccia 2002; Sartori 1976, 1982) in several respects, it shares with the latter an emphasis on the importance of conceiving the positive term ‘system’ and its negation ‘anti-system’ as ‘neutral’ and ‘relative’ terms. This is a point worth emphasizing as in the scholarly debate the terms ‘anti-system’ and ‘anti-democratic’ are often considered to be synonyms, yet this represents a serious misconception, often bemoaned by Sartori himself throughout his writings (e.g. Sartori 1982: 300).

Whereas terms such as ‘populist’, ‘extreme’ or ‘radical’ refer only to the ideological profile of political parties, my revisited concept of anti-system party is a multidimensional concept assessing not only the ideological features of an actor but also its functional role in the party system.¹ Significantly, the relationship between two such crucial dimensions is only superficially explored in the classical Sartorian perspective and is subject to problematic assumptions in Giovanni Capoccia’s (2002) approach to anti-system parties. At the same time, a bi-dimensional perspective also characterizes the vast majority of the alternative ‘anti’ concepts introduced by scholars, such as anti-establishment (e.g. Abedi 2004), challenger (e.g. Hobolt and Tilley 2016), outsider (e.g. McDonnell and Newell 2011) and protest parties (e.g. Smith 1989). However, such competing concepts fail to clarify what an ‘anti’ party becomes if a variation occurs in only one of the concepts’ constitutive properties, and they appear to be inadequate from a terminological point of view (cf. Gerring 2012: ch. 5).

The article subsequently develops a typology with four categories for political parties – anti-system, pro-system, halfway house and complementary parties – and enabling their classification across space and time. It then illustrates the payoff of my revisited conceptualization in comparison with existing approaches by assessing their different classificatory power and analytical utility in the analysis of controversial cases from Greece, Hungary and Italy. Using my typology, it is possible to sharpen our analytical tools in the study of party systems, especially if we adopt a comparative perspective, and to shed light on party stability and change. Significantly, my revisited concept of anti-system party and the broader typology of political parties can be employed to tackle important research questions by following a new and rigorous perspective, and to open new research avenues, particularly in relation to two crucial phenomena of interest for party politics scholars: the overcrowding of Sartori’s systems of moderate pluralism (see Mair 1997), and the increasing number of populist parties achieving integration in national political systems (cf. Mudde 2016b: 16).

CLASSICAL PERSPECTIVES ON ANTI-SYSTEM PARTIES

The concept of anti-system party constitutes a fundamental element of Sartori’s (1966, 1976, 1982) seminal works on parties and party systems. The narrow definition refers to a party abiding by ‘a belief system that does not share the values of the political order within which it operates … that would not change – if it could – the government but the very system of government’ (Sartori 1976: 132–3). The broad definition, on the other hand, is meant to cover the variations over time and space ranging ‘from alienation to protest’: ‘a party can be defined as being anti-system whenever it undermines the legitimacy of the regime it opposes’ (Sartori 1976: 132–3; original emphasis). The positive definition of the ‘system’ that Sartori (1982: 300) had in mind was the regime understood as a ‘neutral’ and ‘relative’ term, and in his view, ‘anti-system’ were not only communist and fascist parties, but also actors ‘of other varieties’. According to Sartori, anti-system are those parties that occupy a distant position from the ‘centre’, defined in political-constitutional terms: for example, with the establishment of the Fifth Republic, the Gaullists ‘switch[ed] from being anti-regime to being the new regime’ (Sartori 1976: 154, 163).

More recently, Capoccia (2002) has suggested a distinction between two different forms of ‘anti-systemness’. If the focus is placed upon the study of party systems, the yardstick is the ‘relational anti-systemness’, a property indicating the party’s ability to trigger polarization and centrifugal mechanics. Conversely, if the analytical focus is the comparative study of democracies, the decisive feature is the ‘ideological anti-systemness’, defined as a party’s opposing any one of the defining properties of the democratic system, according to the different level of abstraction with which the latter is reconstructed (Capoccia 2002: 22–3).

Despite their appeal in the scholarly debate, both Sartori’s and Capoccia’s conceptualizations, to use John Gerring’s words (2012: 128), possess a very limited ability to ‘establish . . . clear contrasts with what lies outside the boundaries of a concept’. In this respect, the boundlessness of the Sartorian concept clearly emerges in his own discussion of the Italian Communist Party (PCI) between the 1970s and the 1980s. Since the adoption of the strategy of the ‘Historic Compromise’, the party consistently moderated its public propaganda (Capoccia 2002: 29), and this represents a crucial point as, following Sartori (1976: 194, n.8), the ‘delegitimizing impact’ of a party needs to be primarily assessed in terms of ‘verbal behavior’. At the same time, if it is recalled that it is precisely the ‘belief system’ of a party that constitutes the ‘hard core of the concept’ (Sartori 1976: 133), it becomes incomprehensible why Sartori (1982: 302), while emphatically arguing that the ‘primary loyalties’ of the PCI had not changed (i.e. its belief system) and that there is no proof to argue otherwise, argues a few pages later that the ‘nature’ of the PCI was changing in the same period: from anti-system to halfway (Sartori 1982: 326, n.75). Unfortunately, halfway parties are never really discussed, and the most extensive definition reads: ‘parties that accept the system as long as it exists . . . ranging between yes and no (swaying between verbal refusal and conditioned acceptance)’ (Sartori 1982: 21; 19). Here again: how can we determine when a party accepts a system only ‘as long as it exists’? What should we understand by ‘swaying between verbal refusal and conditioned acceptance’? Furthermore, according to Sartori (1976: 127–8, 142; 1982: 208), the integration of anti-system parties needs to be assessed in terms of ‘reciprocal re-legitimation’ between anti-system and pro-system formations, by distinguishing ‘invisible’ (what is done in low visibility areas, such as parliamentary commissions) from ‘visible’ politics (what is said in public). However, Sartori (1976: 194, n.8) simply does not provide guidelines on how to assess such a ‘reciprocal process’ beyond suggesting the ‘content analysis of the daily press’.

On the other hand, Capoccia (2002: 10) argues that although relational anti-systemness points to the ‘ideological difference of one or more parties from the others in the system’, we should perform ‘a general evaluation of a party’s coalition and propaganda strategies’ in order to determine it. Accordingly, this would logically imply that if a party experiences a change in its coalition or propaganda strategies it would cease to possess the property of relational anti-systemness. Nevertheless, Capoccia (2002: 26) allows for a party to retain relational anti-systemness if it is ‘disloyal in coalitions’. However, this results in equating two very different scenarios, namely being ‘not-coalitionable’ and being ‘disloyal’ yet coalitionable, which carry substantially different implications for the functioning of party systems. ² This is evident if we consider, for example, Capoccia’s (2002: 25, n.18) reference to the Austrian Freedom Party (FPÖ) in the early 2000s: the party ‘certainly contributed to the overall polarization of the system while in opposition, it is much less so now that the party is in government’. It is unclear, then, to what extent the FPÖ presented the property of relational anti-systemness during such a period, and the problem of determining conceptual boundaries thus remains open: how should we classify parties whose coalition and propaganda strategies do not vary in the same direction and obtain contradictory impacts on the functioning of party systems, such as the Northern League in Italy?

BEYOND CLASSICAL CONCEPTUALIZATIONS: A REVISITED CONCEPT OF ANTI-SYSTEM PARTY

In the following sections I shall explore in detail the constitutive properties of my revisited concept of anti-system party, which are identified by adopting a classical, or Aristotelian, approach which includes its necessary and jointly sufficient properties (Sartori 1984). According to my revisited conceptualization, a party can be defined as anti-system when the two following properties are simultaneously present:

1. (1) its ideological orientation towards the status quo does not simply result in the articulation of a conventional anti-incumbent and policy-oriented opposition, but also in questioning the established metapolicies, and;

2. (2) it has not taken part in very visible cooperative interactions at the systemic level, whether because of its own antagonistic self-perception and/or the attitudes of the other parties in the system; or despite a previous involvement in such interactions it deliberately favours a return to the margins of the party system through the process of radical disembedding.

At this point, it is crucial to clarify why the term ‘anti-system’ is still superior to the alternative ‘anti’ labels introduced by scholars, despite the departure from classical conceptualizations. The online supplementary material provides a list of the most influential of these competing concepts, and, although it is not possible to discuss them in detail here for reasons of space, it suffices to say that they usually adopt a bi-dimensional perspective, assessing both the features of the party in itself (e.g. party ideology, propaganda or style) and its role in the party system (usually defined in terms of coalition potential). ³

Hanna Pitkin (1972: 173) argues that ‘the meaning of a word . . . is what one finds in a good dictionary’, and, following Gerring (2012), this can be defined as the ‘resonance’ of a concept. By checking the meanings of the competing ‘anti’ concepts in the authoritative Oxford English Dictionary, it becomes evident that they fail to establish a consistent link between concepts’ labels and language as well as between the concepts’ labels and their constitutive dimensions. For example, the term ‘outsider’ may be appropriate to refer to the status of a party in the party system only, but not to evoke a specific set of ideological features. The vast majority of the competing ‘anti’ concepts also suffer from limited ‘fecundity’ because ‘a coherent term’ allows ‘us to infer many things (the common characteristics of the concept) with one thing (the concept’s label)’ (Gerring 2012: 124–6). Indeed, virtually all of the ‘anti’ concepts listed in the online supplementary material – including, inter alia, the concepts of anti-establishment (e.g. Abedi 2004), challenger (e.g. Hobolt and Tilley 2016), outsider (e.g. McDonnell and Newell 2011), protest party (e.g. Smith 1989) – fail to clarify what an ‘anti’ party becomes if a variation occurs only in one of the two constitutive properties; interestingly, the limited fecundity of such concepts appears to be influenced by the lack of linguistic resonance (see below).

In my conceptual exercise, the very decision to adopt the term ‘system’ and its negation ‘anti-system’ is made to overcome these serious weaknesses. Drawing on the Oxford English Dictionary (Soanes and Stevenson 1989), the word ‘system’ – from Latin systēma, from Ancient Greek σύστημα (sústēma) – can be defined, inter alia, as:

1. (a) ‘a set of things working together as parts of a mechanism’;

2. (b) ‘the prevailing political or social order’;

3. (c) ‘a set of principles or procedures according to which something is done’.

These three main meanings of the word ‘system’ can easily be tailored to the comparative study of political parties and party systems, and provide a solid bridge between the constitutive dimensions of my revisited concept of anti-system party and language without introducing a neologism (cf. Gerring 2012: 118). Indeed, these meanings are particularly appropriate to refer to the phenomenon under investigation for the following reasons:

• ∙ meaning (a) is the most general and can be used to determine the level of analysis, namely the party system;

• ∙ meaning (b) can be employed to refer to the ideological orientation of a party towards crucial issues of the status quo;

• ∙ finally, meaning (c) makes it possible to refer to a very visible set of interactions at the systemic level.

DETERMINING THE IDEOLOGICAL ORIENTATION OF A PARTY TOWARDS THE STATUS QUO

Whereas the classical Sartorian conceptualization focuses on ‘constitutional policy positions’ (Sartori 1976: 154), Capoccia’s (2002) reassessment and the majority of the alternative ‘anti’ concepts extend the perspective to encompass other crucial dimensions of political conflict, such as major economic and social issues (e.g. Abedi 2004; McDonnell and Newell 2011). Broadening the perspective is crucial if we are interested in capturing the full range of possible sources of substantial ideological differentiation between political parties and the existing ‘system’. In this respect, I fully share the view of Duncan McDonnell and James Newell (2011), who, following Maurizio Cotta (1996), evoke the concept of metapolicy. Metapolicies are: ‘the choices that concern the basic arrangements of the political regime, of the political community or of the social and economic system, or else the country’s location in an international system of alliances expressing fundamental conflict between two sides, or, lastly, support for all-encompassing visions of the world’ (Cotta 1996: 29, cited in McDonnell and Newell 2011: 445).

‘Metapolicies’ refer to a qualitatively different and higher level of partisan competition in comparison with mesopolicies (e.g. the transformation of the electoral system; a reform of the pension system) and micropolicies (e.g. patronage), as a party questioning one or more metapolicies challenges crucial elements of the status quo. In other words, metapolicies refer to crucial values and/or practices of the political, social or economic system that are enshrined by the existing order.

Nevertheless, as virtually any party can question specific metapolicies for tactical reasons at specific and limited points in time, it is crucial to distinguish the former from the actors for which such an antagonistic orientation is a consistent and long-standing feature. In particular, in order to avoid conflating different levels of analysis which are often interrelated but that do not necessarily vary together (e.g. verbal propaganda, rhetoric or political style), I argue that it is necessary to focus on the core ideological concepts of a party (see Mudde 2007). As Terence Ball (1999: 391–2; original emphasis) underlines: ‘a core concept is one that is both central to, and constitutive of, a particular ideology and therefore of the ideological community to which it gives inspiration and identity’ (cf. Freeden 1996).

Therefore, according to the specific configuration of their own ideological morphology, political parties can question none, one or multiple metapolicies at the same time. In order to guide empirical research, it is necessary to unpack the definition given by Cotta (1996) and to clarify the meaning of each metapolicy. Although, for reasons of space, the discussion in the next pages is conducted with liberal-democratic systems in mind, it is worth recalling that each broad metapolicy can be identified (and operationalized) across time and space.

The political regime

As Pippa Norris (2011: 26) argues, the regime evokes ‘the legitimacy of the constitutional arrangements and the formal and informal rules of the game’. Here, it would certainly be an exaggeration to consider any proposal for constitutional reform as a discontinuity in the configuration of the metapolitical system; on the contrary, only the ideological positions that question crucial elements constituting the sources of legitimation upon which the political regime itself is built qualify as instances of metapolitical opposition. For example, populist parties of different varieties (see Mudde 2007: 30) and New Politics parties in their early phase are prominent instances of actors which do not oppose democracy as an ideal but instead question decisive elements of real existing liberal-representative democracies (i.e. the political regime), especially their intermediate structures. Furthermore, if our analytical focus is placed on contemporary EU member states, the European Union can certainly be considered as an integral and constitutive part of the national political regime. Accordingly, a party advocating the withdrawal of its own country from the eurozone, from the EU, or systematically questioning the very legitimacy of the latter on ideological grounds can be classified as an example of ideologically inspired opposition to a crucial element of the existing political regime.

The political community

This metapolicy does not simply refer to the criteria for the determination of the demos, but, in particular, it evokes the legitimacy of the boundaries of the national state. Ceteris paribus, not all non-state-wide parties qualify as instances of metapolitical opposition to the political community but, rather, only secessionist parties do – that is, actors that ‘clearly and unambiguously manifest their will to break away from the state, whether to form a new independent state or to (re)join another’ (Massetti and Schakel 2016: 63).

The social and economic system

If our analytical focus is placed on the study of Western democracies, this involves referring to the basic values and practices of contemporary capitalism. Radical left parties are the most prominent examples of actors questioning the metapolicy represented by the established socioeconomic system, as they oppose ‘the underlying socio-economic structure of contemporary capitalism and its values and practices’ (March 2011: 19).

The country’s location in an international system of alliances expressing fundamental conflict between two sides

Membership in international alliances can be considered as a metapolicy only if being a member or a non-member bears fundamental implications for the political and socioeconomic status quo at the national level. This point was exemplified by the contraposition between NATO and the Warsaw Pact, and evokes the case of orthodox communist parties in Western Europe during the Cold War.

Support for all-encompassing visions of the world

A qualitatively distinct group is constituted by extreme parties, as although such actors technically oppose the political regime, their ‘prognosis’ is no longer within democracy (Mudde 2006: 183) but rather evokes non-democratic solutions. Notable examples here are extreme right parties, especially historical fascist formations, and extreme left parties, in particular historical communist parties of the Stalinist, Maoist and Trotskyist variants.

As the examples above suggest, in the majority of cases the most immediate way to determine whether a party challenges the established metapolitical structure on ideological grounds is to adopt a party family approach (Mair and Mudde 1998). However, scholars can resort to a variety of research methods and techniques to explore this point, especially in the analysis of borderline cases.

DETERMINING THE VISIBLE INTERACTIONS OF A PARTY AT THE SYSTEMIC LEVEL: A TWO-STEP APPROACH

Step One: Has the Party Achieved Systemic Integration?

The identification of the ideological orientation of an actor towards established metapolicies needs to be complemented by another property capable of clearly distinguishing political parties according to their qualitatively different functional roles in the party system. Accordingly, I suggest focusing on a second property, defined as systemic integration, which allows us to shed light on the qualitatively different interactions that can occur between a political party and the broader metapolitical system.

Determining whether a party has achieved systemic integration requires the identification of the two possible ways in which it can cooperatively interact with the broader metapolitical system: indirectly or directly. The most common path to achieving systemic integration is indirect, and it occurs when a party takes part in visible, formalized and reciprocal cooperative interactions with the actors variously associated with the status quo. The latter are mainstream parties – that is, ‘typically governing actors’ (Meguid 2010: 46) that occupy an ‘overall advantageous position in the system’ (De Vries and Hobolt 2012: 250). The direct path to systemic integration, on the other hand, occurs when an actor, despite its principled refusal to cooperate with the parties variously associated with the metapolitical structure, achieves governmental relevance and directly contributes to the continuity of the established metapolitical system, to the level of representing a functional equivalent of coalescence with the former group of parties. Here, a paradigmatic example is constituted by the recent experience of Syriza in Greece (see below).

Although in many cases achieving systemic integration corresponds to the development of coalition potential, by referring to the former concept we can capture a broader set of interactions that the latter concept is not able to capture, irrespective of whether we perceive it as a status assigned by core system parties (Bolleyer 2008) or the capacity to determine the formation of government majorities (Sartori 1976) (see below).

At this juncture a crucial question arises: what is comparable across space and time? Given the peculiarities of national political systems, we need information which is sufficiently precise to be meaningfully comparable and capable of highlighting the different ways in which parties can be very visibly integrated into the existing ‘system’. To appropriately determine the visible interactions of a party at the systemic level the focus needs to be placed at the national level, and this is crucial as many parties may be ‘coalitionable’ at the subnational or regional level but may not have achieved the property of systemic integration as defined above. This point is especially salient, in particular but not exclusively, in the case of secessionist or irredentist parties.

Hence, six different scenarios suggest that a party has achieved systemic integration:

Scenario I: It belongs to the group of core-system parties

The ‘core’ of the party system refers to the ‘the party or parties that over a substantial period have been in leading positions; those parties that have been especially influential for the functioning of the system and the particular pattern of party alignments, especially the coalitional line-ups, that has evolved’ (Smith 1989: 161). ⁴

Scenario II: It has participated in coalition governments and/or pre-electoral coalitions with mainstream parties at the national level

Participation in governing coalitions and/or in pre-electoral coalitions with mainstream parties – which are usually potential governing coalitions – indicates not only that a party is willing to cooperate with others but also that it is integrated in visible cooperative interactions with the actors variously associated with the status quo.

Scenario III: It has participated in formal minority governments led or supported by mainstream parties at the national level

A political party participating in a formal minority government with mainstream actors is not only available for cooperation, but is also accepted by the other parties in such a role as ‘there are grounds to believe [that] this support was based on some sort of agreement with the leaders of government parties’ (De Swaan 1973: 143).

Scenario IV: Its public relationships with one or more mainstream parties suggests that it is integrated in visible and mutual cooperative interactions

In general, the most immediate test to assess whether a party has achieved systemic integration is provided by an assessment of the public relationships between political parties. In this respect, Peter Mair (1997: 210) maintains that ‘what matters is whether there are parties that . . . are regarded by the other parties in the system as unacceptable allies’. This represents an important insight; however, such a perspective overlooks the fact that a party may deliberately characterize itself as antagonistic in behavioural terms to the level of rejecting the very possibility of engaging in visible cooperative interactions while, at the same time, the other parties may perceive it as Koalitionsfähig, as exemplified by the case of the Italian Five Star Movement (M5S) in the aftermath of the 2013 general elections. ⁵ In other words, achieving systemic integration through Scenario IV corresponds to acquiring coalition potential, which is understood here as follows: on the one hand, mainstream parties perceive the actor z as a potential coalition partner; and, on the other, the actor z is equally available to bear the costs and benefits of a potential, reciprocal and formalized cooperation with one or more mainstream parties. As Nicole Bolleyer (2008: 24–5) underlines, this corresponds to a potential ‘in the sense of the word’ whose concretization in pre-electoral coalitions and/or coalition governments becomes simply dependent on bargaining strength or programmatic compatibility (cf. Sartori 1976: 122).

Scenario V: Mainstream parties in fluid contexts

In some contexts, for example recent democracies, political systems characterized by low levels of party system institutionalization and/or in major transitional phases, the identification of mainstream parties can be, admittedly, problematic. Here, a more flexible approach is required, and in such fluid contexts it is necessary to identify the actors that play a predominant role in the coalitional line-ups oriented at the construction of governmental majorities and that represent the major (credible) electoral alternatives within a given system (e.g. Forza Italia in 1994, see below).

Scenario VI: The party has contributed to the continuity of the metapolitical system through very visible and direct actions while in office

The participation in government by a party that ideologically challenges crucial features of the status quo does not necessarily imply that it adopts a behaviour oriented to the overthrow or paralysis of crucial features of the existing metapolitical system, although in some historical cases this has certainly occurred (e.g. the German National Socialists in 1933). Such a possibility is highlighted by the example of Syriza since July 2015 which, by accepting and directly enacting the Third Memorandum, has made it clear that its contribution to the overall functioning of the system has substantially and qualitatively changed (see below).

Step Two: Has the Party Experienced the Process of Radical Disembedding?

Once a political party crosses the threshold of systemic integration it overcomes a critical juncture, and its functional role in the system qualitatively changes. However, it is wrong to simply assume that such a property is necessarily maintained in omne tempus. Indeed, an additional scenario suggests that a party, despite having previously achieved systemic integration, has deliberately favoured a qualitative reversal of its functional role in the system.

Scenario VII: The party deliberately engages in a process of radical disembedding

The process of radical disembedding occurs when an actor whose core ideological concepts question crucial issues of the status quo emphasizes its antagonistic stance towards established metapolicies and simultaneously favours its return to the margins of the political scene (i.e. by antagonistically distancing itself from the previous involvement in Scenarios I–VI). Although such a process is relatively rare because the acquisition of systemic integration normally has important consequences for the credibility of a party as a (potential) governing actor, a party may deliberately engage in the process of radical disembedding in order to underline its own differentiation in the competitive market as well as to distance itself from the ‘system’. ⁶ Paradigmatic examples of the process of radical disembedding include the Austrian FPÖ following the change of leadership from Jörg Haider to Heinz-Christian Strache (Heinisch 2008: 83), the Dutch Freedom Party following the collapse of the Rutte I cabinet (Vossen 2016), and the Italian Northern League during the secessionist phase (Giordano 1999: 220) (for details, see below).

To summarize, a party achieves systemic integration in the presence of at least one of Scenarios I–VI. This property is then usually maintained over time except in the case of a party deliberately favouring a return to a status of non-integration through the bi-dimensional process of radical disembedding (Scenario VII).

A final specification is needed: the presence of systemic integration does not necessarily imply that a party is positively integrated in ‘the system’ (cf. Ieraci 1992). On the contrary, systemic integration can be either positive, if the core ideological concepts of a party are not in contrast with the metapolitical system, or negative, in the case of actors that ideologically challenge one or more established metapolicies. As I shall discuss in the next pages, negatively integrated parties qualify as halfway house parties, a distinct type of political actors in comparison with anti-system parties, in terms of their visible interactions at the systemic level, as well as in comparison with pro-system parties – in terms of their ideological core concepts. For example, important contemporary cases such as Syriza, the Northern League and the Swiss People’s Party are instances of negative integration and qualify as halfway house parties: on the one hand, they are very visibly integrated in the ‘system’ (that is, they present the property of systemic integration); on the other, they ideologically question one or more crucial features of the status quo (i.e. as indicated by their own core ideological concepts).

These considerations suggest that the typical perspective adopted in the little industry of ‘anti’ concepts proposed by scholars is indeed inadvisable. For example, McDonnell and Newell (2011) speak of ‘outsider parties’ in government, yet the key question remains open: what does an outsider party become once it participates in government? As the concept of outsider party is built upon two necessary and jointly sufficient properties – having gone through a phase as not coalitionable and the articulation of a metapolitical opposition – it is clear that we are talking about a different phenomenon once such actors enter the group of potentially governing parties. At the same time, clear problems emerge following the dichotomization by Amir Abedi (2004: 11) between establishment and anti-political-establishment parties, because it is an exaggeration to believe that an actor belonging to the latter group – that is, a party challenging ‘the status quo in terms of major policy issues and political system issues’ to echo one of his own criteria – becomes tout court a party ‘like the others’, an establishment party, simply as a consequence of the acquisition of ‘governing potential’ or ‘governmental relevance’ and in the absence of substantial moderation of its core ideological concepts. These problems, also shared by the vast majority of the competing ‘anti’ concepts listed in the supplementary material, suggest that a dichotomous approach is indeed inadvisable, and invite us to undertake the path of multidimensionality.

ESTABLISHING CONCEPTUAL BOUNDARIES

By focusing on the ideological orientation of party towards the status quo (anti-metapolitical vs anti-incumbent/policy-oriented) and its visible systemic interactions (presence or absence of systemic integration), it is possible to craft a conceptual typology allowing for the identification of four distinct types of political actors: anti-system, halfway house, pro-system and complementary parties (see Table 1).

Table 1 A typology of political parties

Anti-system parties

Although the most famous historical cases were the fascist formations in the interwar period and communist parties during the Cold War, many different variants of anti-system parties have developed over recent decades. In contrast to the historical cases, the vast majority of contemporary anti-system parties do not question democracy as such (that is, as an ideal, see above): as Gregory Luebbert (1986: 12) underlines, ‘many parties have been anti-system without being anti-democratic’. A political party qualifies as anti-system when it presents a double image of externality in comparison to the ‘system’: in terms of its core ideological concepts as well as in terms of its direct and indirect visible interactions with the system itself. This double image of externality characterizes parties such as the French Front National, the British UK Independence Party (UKIP) and the Italian M5S, while other parties such as the Greek Syriza and Podemos in Spain have displayed such a double image of externality until recently. Although many of the anti-system parties on the rise in recent years do display a populist core (Mudde 2004), it is important to underline that not all populist parties qualify as anti-system parties under my revisited concept, as they would following the classical Sartorian perspective. On the contrary, they represent instances of ‘halfway house parties’.

Halfway house parties display systemic integration, yet they decisively differ from the more conventional actors possessing such a property because their core ideological concepts constitute the ‘source of inspiration’ for a metapolitical opposition. ⁷ The choice of the term ‘halfway house’ to refer to this type of political actors seems particularly suitable as it points to ‘a compromise between two different or opposing views or courses of action’ (Oxford English Dictionary (Soanes and Stevenson 1989)), namely being visibly integrated in ‘the system’ while simultaneously questioning one or more of the crucial features of the status quo ideologically. Notable examples of halfway house parties are the Italian Northern League since 2000; the Swiss People’s Party since the early 1990s; Syriza, following the acceptance of the Third Memorandum in July 2015; and Podemos, following the 2015 Spanish general elections and the (failed) negotiations with the Socialist Party which nevertheless indicated its acquisition of coalition potential (see Simón 2017: 12–13; cf. Bolleyer 2008: 24–5).

Pro-system parties are characterized by a ‘conventional’ anti-incumbent and policy-oriented ideological profile and present the property of systemic integration. The most common pro-system parties are ‘core system parties’ (Smith 1989: 161), which are not simply insiders par excellence but are often the actors that have decisively contributed to the establishment and/or the continuity of the constellation of metapolicies in place in a given context (e.g. the German Social Democratic Party (SPD) and the Christian Democratic Union-Christian Social Union (CDU-CSU)). Other pro-system parties are non-core system actors that present a moderate ideological profile and whose public relationships with mainstream parties suggest that they have been integrated in visible cooperative interactions at the systemic level (e.g. the Austrian Greens and To Potami in Greece).

Complementary parties are often, although not necessarily, new political actors that, despite the fact that their ideological morphology does not qualify as anti-metapolitical, tactically adopt an antagonistic stance. ⁸ For this reason, they have not been integrated in visible cooperative interactions, either because they present themselves as unavailable for cooperation with mainstream parties or because the latter may view such parties suspiciously (e.g. the Italian Movement for Democracy–The Network in the final years of the First Republic). Actors that have engaged in a process of substantial ideological moderation (i.e. from anti-metapolitical to anti-incumbent and policy-oriented), such as the German Greens in the early 1990s, and are still nevertheless viewed as untrustworthy outsiders by mainstream parties also qualify as complementary parties.

MAPPING CONCEPTUAL AND EMPIRICAL CHANGE: AN ASSESSMENT OF THE CLASSIFICATORY POWER OF THE COMPETING CONCEPTUALIZATIONS

Having outlined the main features of the four types of political parties identified through my novel typology, it becomes crucial to assess its classificatory power in comparison with existing conceptualizations. This section provides a dynamic application of my typology in the analysis of crucial cases from Italy, Hungary and Greece, which share the commonality of being very difficult or impossible to classify following both the existing approaches of anti-system parties and the alternative ‘anti’ concepts listed in the online supplementary material. ⁹

Greece: Syriza and the Independent Greeks

In an analysis prior to the 2015 Greek general election, Takis Pappas (2014a: 109) argued that Anexartitoi Ellines (ANEL – Independent Greeks) and Syriza qualified as anti-system not simply because of their opposition to the Memorandum, but in particular because in ideological terms they represented negations of ‘political liberalism’. However, the adoption of a perspective focusing only on party ideology makes it impossible to reclassify both parties, despite the substantial developments that occurred in the second half of 2015. The problem here is that following Pappas (2014a) and Sartori’s (1976) narrow definition they would remain anti-system despite the acceptance of the Third Memorandum, given the fact that they still qualify as abiding by a ‘belief system that does not share the values of the political order within which [they] operate’. A Manichaean and populist vision of the world remains a central element of the ideological morphologies of both the parties, and after the de facto acceptance of the third bailout the revival of the Syriza-ANEL coalition following the September 2015 elections can be explained only if we maintain a populist framework (Mudde 2016a). At the same time, however, it is clear that both ANEL and Syriza no longer play the same functional role in the broader system as they did before the acceptance of the Third Memorandum, which sanctions their transformation from anti-system to halfway house parties.

This is precisely the point at which the analytical and discriminatory power of the second dimension of my conceptualization, systemic integration, clearly emerges. While remaining ideologically distant from crucial features of the status quo, both Syriza and ANEL have been involved in very visible cooperative interactions with the metapolitical system and achieved systemic integration through the direct path (Scenario VI). In other words, Syriza and ANEL achieved systemic integration without reciprocal and visible cooperation with conventional parties (that is, the indirect path), but rather through the acceptance of harsh austerity measures while in government, which for the very peculiarities of the Greek case represented a functional equivalent to coalesce with mainstream actors. In this respect, it is important to stress that while austerity measures usually qualify as mesopolicies, the Third Memorandum assumed a metapolitical character not simply for Greece itself – as it was at risk of sovereign debt default – but also for the entire eurozone, as Greece was on the brink of expulsion from the euro. This consideration highlights why I previously argued that achieving systemic integration often but not necessarily coincides with the development of coalition potential, and this suggests that a typological effort grounded on the latter only will lead to a serious deadlock, especially if we want our concepts to travel across time and space. For example, following the approach suggested by Bolleyer (2008: 27), the crucial discontinuity following the acceptance of the Third Memorandum is intangible as Syriza still does not qualify as a party willing to enter a ‘bargaining arena regulated by the core of established parties in a party system’, despite openness on the part of PASOK, To Potami and New Democracy, and similar considerations can be made with regard to the Sartorian conception of coalition potential (Sartori 1976: 297).

Unfortunately, the decisive evolutions of Syriza and ANEL remain intangible following classical conceptualizations. This clearly emerges if we follow the broad definition by Sartori, given the central importance assigned to the ‘verbal element’ (Sartori 1976: 133), which, in the case of Syriza, despite a decreasing emphasis on its radical left elements, remains clearly populist and Manichaean; that is, permeated by the dichotomy between the ‘pure people’ and the ‘corrupt elite’ (Mudde 2016a). This also emerges following Capoccia (2002), because it is unclear whether Syriza and ANEL still qualify as instances of relational anti-systemness, not simply because of their propaganda strategies but also because of their very peculiar coalition strategies.

Hungary: Fidesz and Jobbik

Hungary qualifies as an ‘extreme case’ in Europe (Enyedi 2015: 249), as two populist parties, Fidesz and Jobbik, received about two-thirds of the votes in the elections of 2010 and 2014. Fidesz achieved systemic integration shortly after the collapse of Communist rule by engaging in various forms of cooperation with the other parties of the centre-right and became a central player in the Hungarian party system (Fowler 2004); at the same time, following a long ideological journey – first as a liberal party, and later as a national-conservative one – it embraced populism as one of its core ideological concepts, especially since the 17 June 2005 party congress (Enyedi 2015: 240). Therefore, following my typology, the enduring presence of systemic integration combined with the transformation into a populist party, signals the evolution of Fidesz from a pro-system to a halfway house party since 2005.

More recently, Fidesz emerged as the dominant party in Hungary as it obtained a ‘supermajority’ of two-thirds of MPs both in the 2010 and 2014 elections. In particular, following the 2010 electoral victory, the party’s leader Viktor Orbán has repeatedly declared his goal of transforming Hungary into an ‘illiberal state’, and his government ‘has begun a systematic attack on liberal institutions’ (Pappas 2014b: 19). As a consequence, contemporary Hungary can be defined as a ‘populist democracy’, a ‘distinct type of representative democracy’ which is ‘definitely democratic but, to a larger or lesser extent, illiberal’ (Pappas 2014b: 4).

At this point, a crucial question emerges: how can Fidesz be classified on the grounds of my typology, following the transformation of Hungary into a ‘populist democracy’? Despite the establishment of the ‘System of National Cooperation’, Fidesz remains a halfway house party. This is possible because, although many ideas of Orbán’s ‘illiberal state’ have been concretized, Hungary remains a democracy – albeit a ‘defective’ one (Transformation Index BTI 2016) – and it is precisely the ‘grey’ nature ‘between liberal democracy and fully blown authoritarianism’ (Batory 2016: 17) of the new regime that produces a tension between the ideological profile of Fidesz and crucial features of the status quo, such as the remnants of liberal principles and institutions as well as the procedure and mechanisms of electoral democracy that have remained in place.

Despite the peculiarity of the case of Fidesz, it can be classified as a halfway house party like other parties discussed in this article because they share the commonality of being visibly integrated in ‘the system’ while simultaneously questioning one or more of the crucial features of the status quo ideologically. At the same time, the remnants of liberal institutions and the procedure and mechanisms of electoral democracy, coupled with the constraints placed by the membership of the EU, which is often delegitimized and accused of ‘colonialism’, clearly set a decisive difference between Fidesz and the case of political parties that can be classified as pro-system following the establishment of a new political regime in the presence of within-democracy transitions (e.g. the French Gaullists following the establishment of the Fifth Republic, see Sartori 1976: 163) or towards non-democratic settings in both their authoritarian (e.g. the Italian Fascist Party since the introduction of the so-called ‘leggi fascistissime’ in 1925) and totalitarian variants (e.g. the German National Socialists since the enactment of the Ermächtigungsgesetz in March 1933).

Although many scholars highlight the increasing programmatic similarity between Fidesz and Jobbik, especially in terms of nativism, the latter can be best understood as a populist radical right party (Pirro 2015). As Jobbik has often been stigmatized by the media and associated with fascism, especially because of its paramilitary-like activities (Pirro 2015: 69), it comes as little surprise that Jobbik is far from achieving systemic integration. Jobbik has only engaged in opportunistic and policy-specific cooperation with Fidesz, and both the parties are unavailable to engage in reciprocal and formalized forms of visible cooperation with each other. In particular, Fidesz has continually adopted a strategy of demonization, by defining Jobbik as a party ‘of criminals’ (Lambert 2016).

How do existing conceptualizations perform in analysing Fidesz and Jobbik? Whereas the latter easily qualifies as anti-system following both the broad and narrow definitions offered by Sartori (1976) and represents an instance of relational anti-systemness (Capoccia 2002), things are much more complicated in the case of Fidesz. Following the establishment of the ‘System of National Cooperation’, many ideas of Orbán’s ‘illiberal state’ have been concretized, and while Sartori’s focus on the ‘political-constitutional level’ may lead to the conclusion that Fidesz has transformed into a pro-system party, it is clear that, given the ‘grey’ nature of the new Hungarian regime (Batory 2016: 17), as well as the constraints that come with EU membership, it finds itself in a halfway house position, in contrast to the leading parties in truly authoritarian regimes, such as Yedinaya Rossiya in Russia (Transformation Index BTI 2016), whose ideological profiles consistently mirror the status quo. A similar problem emerges following Capoccia’s conceptualization, with Fidesz presenting one feature of relational anti-systemness (in terms of propaganda strategies) but not the other (in terms of coalition strategies).

Following my novel typology of political parties, on the other hand, Fidesz qualifies as an anti-system actor during the final years of communist rule, a pro-system party from the early 1990s up to 2004, and a halfway house party from 2005 up to now. Jobbik, on the other hand, is classified as an anti-system party throughout its existence.

Italy: Forza Italia and the Northern League

The final test of the classificatory power of the different conceptualizations is represented by a focus on the Italian case, as prominent scholars are divided on whether the party system of the Second Italian Republic presents relevant anti-system parties following classical approaches. Mair (1997: 217) argues that the party system of the Second Italian Republic in the 1990s displayed ‘no relevant anti-system party . . . at least in Sartori’s sense of the term’, while Cas Mudde (2014: 219, 224, n.7) argues that Northern League and Forza Italia (and its successor, the People of Freedom) are the relevant anti-system parties of the Second Italian Republic. ¹⁰ In the light of these diverging views, it becomes crucial to assess how my typology performs in the study of the Northern League and Forza Italia.

In the final years of the First Republic the Northern League constituted the point of coagulation of anti-party and anti-state sentiments to the level of representing ‘an almost ideal-typical incarnation of populism’ (Tarchi 2003: 135), and accordingly qualified as an instance of ideologically inspired opposition to crucial elements of the status quo. At the same time, the Northern League did not present the property of systemic integration, given its non-involvement in either direct or indirect visible cooperative interactions, as previously defined. With the collapse of the traditional party system the party changed its functional role by entering both a pre-electoral coalition and a coalition government with other parties, most notably Forza Italia. The latter party, since its very foundation, showed a clear governing vocation, and represented at that time a case of a mainstream party in fluid contexts (Scenario V). Indeed, Forza Italia ‘ever since its appearance . . . carried out a double function of systemic integration. On the one hand, it has filled the gap left by the disappearance of governing parties; on the other hand, it has made a decisive contribution to bipolar dynamics in a party system that is still fragmentary and in which there are heterogeneous coalitions’ (Raniolo 2006: 450). Following my typology, scholars may classify Forza Italia as either a halfway house or a pro-system party, depending on whether they consider its populism to be an ideological core concept (e.g. Mudde 2007) or essentially a feature of the political style of its leader, Silvio Berlusconi (Tarchi 2003).

Returning to the Northern League, following the very visible cooperative interactions with Forza Italia and the other parties of the centre-right pole, which culminated in the short-lived first Berlusconi government (1994), the party evolved from anti-system to halfway house party. However, this phase was temporally limited and the return to the anti-system party type was very rapid, as the Northern League experienced the bi-dimensional process of radical disembedding (Scenario VII). Indeed, in the second half of the 1990s the Northern League embraced secessionism ‘to try to differentiate itself from the other Italian political parties and (re)create its anti-system image’ (Giordano 1999: 220), and the two conditions required to speak of radical disembedding occurred. First, the party considerably radicalized its antagonistic stance towards the established metapolitical system by openly advocating the break-up of the Italian state. Second, the Northern League deliberately favoured its own return to a status of non-integration, by adopting an isolationist strategy and presenting itself as a distinct pole in the party system. Nevertheless, following a series of electoral defeats, the Northern League relinquished secessionism and progressively developed a new anti-metapolitical blend combining populism, nativism and Euroscepticism (Albertazzi and McDonnell 2005); at the same time, it re-approached the centre-right pole by rebuilding very visible cooperative relationships with the latter, in particular with Forza Italia. For these reasons, since 2000 the Northern League can be reclassified as a halfway house party.

During the First Italian Republic, the Northern League qualified as an anti-system party following both the narrow and broad definitions given by Sartori (1976), and the same applies to the Second Italian Republic. However, such ‘classificatory stability’ overlooks the crucial discontinuities mentioned above. At the same time, things become much more complicated if we follow the conceptualization set out by Capoccia (2002) because the coalition and propaganda strategies of the Northern League in the Second Republic varied in different directions, except for the secessionist phase, with the result of making the party unclassifiable following Capoccia’s (2002) typology. On the other hand, Forza Italia qualifies as anti-system following Sartori’s broad definition thanks to the persistent delegitimizing impact of its propaganda (cf. Mudde 2014), while following the narrow definition a classification is influenced by how we consider its populism (see above). Finally, similarly to the previous cases, it is very difficult to classify Forza Italia following Capoccia’s conceptualization, because its ideology, coalition and propaganda strategies pointed in different directions, and this variation cannot be appropriately captured on the grounds of his property of relational anti-systemness.

CONCLUDING REMARKS

At the beginning of this article we wondered whether the concept of anti-system party enjoys good health as well as longevity. This article has highlighted the fact that classical conceptualizations (Capoccia 2002; Sartori 1966, 1976, 1982) lead to boundless, undelimited and intrinsically problematic analyses. Similar considerations apply to the ‘cottage industry’ of the alternative ‘anti’ labels introduced by scholars over time, which fail to clarify what an ‘anti’ party becomes if a variation occurs in only one of the constitutive properties of the concepts. In order to overcome such limitations, this article has developed a revisited concept of anti-system party and a novel typology of political parties by focusing on two salient dimensions for any political actor: its core ideological concepts and its visible interactions at the systemic level.

As Heraclitus maintains, ‘the only thing that is constant is change’, and our task as social scientists is to be appropriately equipped to understand relevant phenomena not simply statically, but, in particular, dynamically. This is precisely the point at which the existing conceptualizations of ‘anti’ parties show their considerable weaknesses and where the analytical power and empirical utility of my conceptual exercise emerges. In particular, by focusing on the two crucial dimensions of my typology, it is possible to assess whether an actor evolves to become a different type of political party. In this respect, variations in the social structure, such as changing constituencies (Sokhey and Yildirim 2013), and alterations in the structure of political opportunities (Kitschelt 1986) appear to be triggering factors in explaining why political parties evolve over one or both of the dimensions of my typology. As this article suggests, a political party may present the defining features of a specific type only for a short phase of its lifespan, for longer periods or even for its entire history. Political parties may move along the cells of my typology through a gradual and lengthy process, such as the Italian PCI, or suddenly, as the Austrian FPÖ did following the 1999 election. Although previously anti-system parties often evolve into halfway house actors, as in the case of Syriza, they can travel a longer road to become pro-system parties, as in the case of the German Greens. Despite the propensity of anti-system parties to evolve into halfway house parties and that of complementary actors to transit to the pro-system type, party trajectories are neither deterministic nor irreversible. For example, the Swiss People’s Party evolved from a pro-system to a halfway house party since the 1990s, while the Austrian FPÖ can be classified as a pro-system party in the first half of the 1980s, anti-system following the election of Haider as party leader and up to the 1999 elections included, a halfway house actor during the participation in the Schüssel governments in the early 2000s, and again as anti-system under the leadership of Strache as a result of the bi-dimensional process of radical disembedding.

This article has placed focus not simply on the side of concept formation, but also on providing a set of guidelines to help researchers apply the concepts to empirical cases – a point that has always represented a substantial weakness of existing conceptualizations. Following my revisited concept and the novel typology of political parties, important research questions can be tackled in a new and rigorous perspective, most notably opening up new approaches to the comparative analysis of party systems, especially in the light of the substantial overcrowding of Sartori’s systems of moderate pluralism (cf. Mair 1997). Additional new avenues for future research include the study of important questions concerning anti-system parties, such as the conditions favouring their emergence, electoral success or the achievement of systemic integration. More generally, my typology of political parties and its two constitutive dimensions allows us to sharpen our analytical tools for studying parties that differ from more conventional political actors by complementing the assessment of their ideological features with a ‘functional’ perspective. This is relevant in particular, although not exclusively, in the analysis of populist parties which would by default be considered as anti-system following the classical Sartorian approach (see Mudde 2014) and that are, at the same time, often very difficult, if not impossible, to be classified following Capoccia’s (2002) typology. Following my typology, on the other hand, it is possible to operate a major distinction between populist parties by complementing the ideological approach with a broader systemic perspective, which allows us to discriminate the actors that can be considered as anti-system parties on the grounds of their ideological profile and their functional role in the party system (e.g. the French Front National, the M5S in Italy) from halfway house parties – that is, actors that, while ideologically challenging crucial features of the status quo, are visibly, although negatively, integrated in the party system (e.g. the Swiss People’s Party, Syriza in Greece). Significantly, the revisited concept of anti-system party and the novel typology of political parties presented in this article allow us to undertake important steps towards the ‘paradigmatic shift’ advocated by Mudde (2016b: 16) in which populist parties ‘are no longer seen as new outsider-challenger parties, but also as institutionalized and integrated members of the political system’.

SUPPLEMENTARY MATERIAL

To view the online appendix, please visit https://doi.org/10.1017/gov.2017.12.