Abstract
Scholars of party politics have recently examined the influence of niche parties on the dynamics of party competition. In particular, drawing on the ‘Position, Salience and Ownership’ (PSO) model, it has been argued that when a new niche actor enters the political arena, it seeks to introduce a new policy dimension into political debate and simultaneously affects the competitive strategies of mainstream actors. By using manifesto data, this article analyses the impact of the Lega Nord on political discourse in Italy relating to the territorial dimension. Its findings challenge the assumptions of the PSO theory on niche-mainstream dynamics of competition. The article argues that, firstly, niche parties can mobilise on a policy dimension that has long been present on the political agenda; secondly, that niche party influence on mainstream party strategies is limited, at least when compared with the influence of other systemic variables. Nevertheless, the role of niche parties can be re-evaluated by considering a further strategic tool that parties can use to define their strategies, namely framing. In particular, the article argues that the Lega Nord has introduced a new language into political debate on the territorial dimension, which has required rival parties to react by reframing and redefining these issues differently from the past. Using an original coding scheme for the measurement of party attitudes in party manifestos, the article provides empirical grounds for arguing that a consideration of rhetoric should be included in any analysis of the competitive strategies of political parties in a two-dimensional space.
Introduction
On 22 May 1990, the Italian newspaper La Repubblica, referring to the first party conference of the newly founded Lega Nord (LN), declared: ‘He [Umberto Bossi, movement leader] did not say anything new: fight against centralism, legislative and fiscal autonomy for the regions. But he said it in a new way’ (emphasis added). 1
Italian parties have discussed the redistribution of competencies downwards to subnational levels since the early 1960s. In the early 1990s, however, a new ethno-regionalist actor, namely the LN, introduced a new language into political discourse relating to the territorial dimension 2 (Giordano, 2000) whilst, at the same time, posing a serious electoral threat to mainstream parties. In response to such a challenge, the latter were compelled to adjust their strategies on territorial issues, by adapting their political discourse to the ethno-regionalist rhetoric. The evolution of political discourse on the territorial dimension in Italy after the breakthrough of the LN represents a paradigmatic example of the likely impact of new, electorally relevant, niche actors on the strategies of mainstream parties in a multi-dimensional competitive context (Elias et al. 2015).
Taking as its starting point the conceptual framework provided in this Special Issue, this article provides new insights into the dynamics of niche-mainstream competition in a two-dimensional space. In particular, it argues that niche parties are likely to affect the strategies of mainstream actors in two complementary ways: firstly, by challenging them in electoral competition and, secondly, by introducing a new language into political discourse reflecting a secondary policy dimension.
The article builds on Meguid’s Position, Saliency and Ownership (PSO) theory (2005, 2008), which offers a unique model for conceptualising the impact of the electoral challenge of niche actors on the strategies of mainstream parties. In particular, Meguid argues that mainstream actors respond to the new issues introduced by niche challengers by either dismissing these issues or incorporating them into their own agenda (manipulation of the salience of niche issues). In the latter case, mainstream actors must also decide whether to pursue an accommodative or adversarial strategy (manipulation of their position on niche issues). The scholarly literature on issue framing, however, has pointed out that parties do not only compete against each other by manipulating the salience of, and their strategic positioning on, a policy dimension. They can also discuss, justify and report policy issues using different frames, or arguments. Indeed, when a niche party enters the political arena, it is likely to introduce a new language dealing with its core policy dimension. As a consequence, mainstream actors are often compelled to redefine, or to ‘reframe’, their own political discourse on those issues to account for a niche party’s rhetoric.
Within this theoretical framework, this article will consider political discourse on territorial issues in Italy as a theory modifying case study (Rohlfing, 2012: 10). Accordingly, a set of baseline hypotheses, grounded in Meguid’s PSO theory, will be tested and, drawing on the insights of the issue framing literature and the empirical analysis, reformulated. Through a detailed study of the Italian party manifestos between 1963 and 2013, the article will examine whether (issue salience), to what extent (issue positioning) and how (issue framing) the territorial dimension has been addressed by mainstream actors, both before and after the entry into the electoral arena of a niche ethno-regionalist party (the LN).
The empirical analysis uses an original scheme for the coding of party manifestos, designed for the analysis of single policy dimensions. The instrument is conceived to measure party strategies in issue competition, here defined as the result of the combination of three different, yet complementary tools: salience, position and frames. Data resulting from the content analysis is examined by using descriptive quantitative analysis, which is further substantiated by qualitative evidence.
The article is structured as follows. Section 1 examines the PSO model, the baseline hypotheses derived from this theory and their drawbacks. Section 2 introduces the variables, as well as the method and data used. Section 3 then tests the baseline hypotheses and proposes their reformulation, before Section 4 focuses on the role of frames in the dynamics of party competition. The Conclusion summarises the impact of the new language introduced by niche actors on the strategies of mainstream parties in a secondary policy dimension, a finding that constitutes a significant development of the PSO theory.
1. A dwarf among giants?
1.1. Party competition between niche and mainstream parties: The baseline model
The scholarly literature on party competition has paid growing attention to the likely impact of niche parties on the strategic behaviour of mainstream actors. The underlying idea is that even the ‘giants’ of the political systems are compelled to adjust their strategies when challenged by political ‘dwarves’ on the terrain of a secondary policy dimension.
Most of the existing research dealing with niche-mainstream dynamics draws on Meguid’s Position, Saliency and Ownership (PSO) theory (2005, 2008) and its implications for party competition (Alonso, 2012). The core idea of the model is that new, electorally relevant, niche parties refuse to compete on extant dimensions of policy. Rather, they seek to place ‘new’ issues, representing the very raison d’être of these actors, onto the political agenda (Meguid, 2008: 3–4). However, the success of such an attempt, as well as the electoral fortunes of niche parties, will depend on the electoral strategies of mainstream actors. In order to avoid losing votes to the niche challenger, mainstream parties must decide whether to ignore niche issues (a dismissive strategy), or recognise and emphasise niche issues in their own agendas; the decision here is one about whether to give niche issues salience or not. In the latter case, mainstream actors are expected to choose between the two possible policy solutions (that is, a decision on issue position): they can either accommodate the demands of the niche actor (accommodative strategy), or oppose them (adversarial strategy) (Meguid, 2008: 27–30).
As Elias et al. (2015) argue, Meguid (2008) conceives of the political space as one-dimensional; furthermore, her work focuses mainly on the electoral fortunes of niche parties, rather than on the strategies of mainstream actors. Nevertheless, the PSO model can be used to introduce a baseline set of hypotheses concerning the dynamics of the niche-mainstream competition, although conceived in a two-dimensional, rather than a one-dimensional, space:
1.2. Refining the PSO model: Italy as a theory modifying case study
The aforementioned hypotheses provide an insightful interpretation of the dynamics of party competition between niche and mainstream actors. When testing them on concrete cases, however, some limitations to Meguid’s theory, on which the baseline model is grounded, emerge.
Firstly, the PSO theory relies on the assumption that niche actors strive to introduce new issues into the political debate. However, it does not contemplate the case in which the secondary policy dimension is not new to the political debate, but, rather, has long existed. Recent advances in the literature on niche parties have pointed to the fact that these actors do not necessarily campaign on new issues. Indeed, they are likely to revitalise old topics that have been neglected by the competitors (Meyer and Miller, 2013: 3), such as decentralisation (Wagner, 2012: 852) (see also section 3.1, below).
Secondly, Meguid limits the strategic options available to parties to a clear-cut choice between accommodation and opposition to niche parties. However, sometimes political actors either express purposefully vague and ambiguous positions on a certain policy dimension or they take contradictory stances on various issues on the same dimension (for example, pro-decentralist positions on some issues and anti-decentralist on others). This ‘blurring’ (Rovny, 2013) or ‘broad-appeal’ (Somer-Topcu, 2014) strategy can be adopted by parties when they have not yet developed a clear preference on a secondary policy dimension, or when adopting a clear stance in either direction might deter a share of potential voters holding different preferences on those issues (see section 3.2, below).
Thirdly, the PSO theory only conceptualises changes to issue salience and issue position as being among the strategic tools available to parties when deciding on their strategies. The theory thus overlooks a further complementary instrument at parties’ disposal, namely the definition and justification of the policy problem (issue framing). As argued by Elias et al. (2015), by using these arguments and justifications, parties frame issues belonging to the secondary dimension (for example, territorial issues) in core-dimension terms (for example, using socio-economic or cultural arguments). In this way, mainstream parties may ‘subsume’ new or neglected issues raised by niche actors into their preferred, primary dimension of competition (Elias et al., 2015; on how parties embed new cleavages into existing dimensions, see also Kriesi et al., 2006 and section 4, below).
The rest of this article tests the baseline model outlined above on a country featuring a centre-periphery cleavage, namely Italy. As argued in the Introduction, this approach constitutes a theory-modifying case study (Rohlfing, 2012: 10), which proceeds by testing their hypotheses as well as pointing out their limits, before reformulating them. In Italy, the territorial dimension was already present in the political discourse well before the electoral breakthrough of the LN in the early 1990s. However, the extraordinary electoral success of this ethnoregionalist actor, as well as its harsh and original rhetoric, posed a serious challenge to state-wide parties. As a result, the latter were compelled to rethink their strategies on territorial issues in a competitive arena that is two-dimensional in nature.
2. Variables, data and methods
As already argued, this article analyses the impact of an electorally relevant ethno-regionalist party on the strategies adopted by mainstream actors on a secondary policy dimension, namely the territorial one. Accordingly, the main dependent variable is constituted by party strategies on territorial issues in a two dimensional space (Elias et al., 2015). Strategies are here conceived as the result of the combination of three tools, namely the emphasis posed on a policy dimension in the political discourse (salience), the concrete policy proposals supported (position) and the way in which such positions are justified (frames).
A coding scheme 3 has been designed in order to measure the salience of, as well as the related positions and frames used on, the territorial dimension in 100 mainstream party manifestos, 4 or their closest equivalent, 5 issued between 1963 and 2013. 6 The coding starts in 1963 because after this time territorial reforms were discussed and adopted in the Italian system; the analysis ends in 2013, the year of the last Italian general elections. The coding framework largely draws on the large body of scholarly work that has undertaken content analysis of party manifestos (Budge et al., 2001; Libbrecht et al., 2011; Maddens and Libbrecht, 2009; Volkens, 2001). It nevertheless introduces original elements such as the coding of neutral statements and the measurement of frames.
Salience is measured by summing up all the quasi-sentences 7 dealing with territorial issues in each political document (Budge et al., 2001; Volkens, 2001); then, they are divided by the total number of quasi-sentences in which the manifesto can be split and multiplied by 100, to obtain a percentage. Position is measured in two steps. Firstly, the quasi-sentences containing favourable references to decentralisation are coded as pro-decentralist (code 101 for institutional decentralist statements, code 102 for statements supporting cultural regionalism); if otherwise, they are coded as anti-decentralist (code 103 for institutional centralist statements, code 104 for statements supporting cultural nationalism). A further neutral code (code 100) is applied when a quasi-sentence generically refers to territorial issues but does not express any clear supportive or adversarial stance on it. Secondly, directional certainty is calculated as the sum of quasi-sentences coded as pro-decentralist, minus the sum of those coded as anti-decentralist, divided by the total number of quasi-sentences dealing with territorial issues (Libbrecht et al. 2011; Maddens and Libbrecht, 2009; see also p. 11 of the online appendix). The values of directional certainty range from –1 (clear centralist direction) to +1 (clear decentralist direction), through 0. In particular, values close to zero are obtained either when a party uses only neutral statements to deal with territorial issues (that is, ambiguous and vague direction), or when it uses an even number of pro-decentralist and anti-decentralist statements, which cancel each other out in the calculation of the directional certainty (in other words, contradictory direction) (see also section 3.2.1).
Finally, according to Helbling et al. (2010) and Elfrinkhof (2010), frames are the concrete arguments used by political actors to define, represent and justify their position on a specific dimension. The list of possible frames was developed inductively by looking at a sample of manifestos, in order to identify a set of recurrent topics and arguments related to the territorial dimension. Frames are then associated to specific codes which, in turn, can be neutral, pro-decentralist or anti-decentralist. When the quasi-sentences coded as territorial also contain a reference to one of these frames, then the associated code is applied. For each frame, the salience and the directional certainty are calculated according to the same formulas described above. The list of relevant frames includes arguments used to justify different alternatives of the territorial redistribution of powers (for example, national unity and territorial solidarity) and references to pressures that are deemed to trigger the process decentralisation (for example, the need to enhance the state efficiency) (see also below, section 4).
The reliability of the data obtained has been assessed by comparing the early coding with a further round of coding conducted by the same author after some time and after having improved coding abilities with practice (Krippendorff, 2004: 215). This ‘intra-observer reliability’ test (Krippendorff, 2004; Neuendorf, 2002) merely addresses the stability of the measurement and it is the weakest reliability test, as compared to the ‘inter-coder reliability’ test. Nevertheless, in light of practical constraints, it represents an acceptable ‘second best’ solution (Krippendorff, 2004; Neuendorf, 2002). The test confirms a high level of consistency between the early and the latter coding choices.
3. Empirical analysis: Towards a reformulation of the baseline model
3.1. Who placed territorial issues on the agenda?
The baseline hypotheses rely on the premise that electorally relevant niche parties are likely to introduce new issues from scratch, such as the territorial ones, when they enter the political arena; consequently, mainstream actors are compelled to select a strategy to deal with this secondary policy dimension. A likely scenario occurred in Italy after 1992, when the LN achieved sudden and unexpected electoral success whilst prioritising the issue of territorial reform in its political discourse. It can be considered the first electorally relevant ethno-regionalist party in the Italian party system, whose electoral appearance could reasonably have been expected to introduce the territorial dimension into Italian political debate, as the hypotheses lead us to expect. Although other ethnoregionalist parties contested Italian elections before 1992 (for example, the Partito Sardo d’Azione, the Südtiroler Volkspartei and the Union Valdotaine), they obtained a marginal share of the vote at state level and never represented a serious threat to state-wide parties.
Figure 1 reports the average salience of territorial issues in Italian electoral manifestos, calculated for each electoral contest between 1963 and 2013. Mean scores are displayed both with and without outliers. 8 The saliency of the territorial dimension in the documents issued by the LN is displayed separately from the average for mainstream parties.
Mean salience of the territorial dimension in Italian electoral manifestos (1963–2013).
Figure 1 clearly shows that the territorial dimension has been present in all the electoral manifestos analysed, including before 1992, although the level of mean salience has not been even over time. These data, in line with the recent developments on niche parties (Meyer and Miller, 2013; Wagner, 2012), suggest that the territorial dimension has been long present in the Italian party system agenda, well before the electoral breakthrough of the LN. This finding challenges the PSO theory’s theoretical premise, according to which niche actors are likely to introduce those issues they prioritise.
Furthermore, variation in the salience of this policy dimension cannot be explained exclusively as a response to an ethno-regionalist challenge; rather, it is associated with major policy changes and reforms. In particular, it emerges that periods of heightened attention to the territorial dimension have usually set the pace for a round of territorial reforms, which have been generally followed by a declining interest in those issues (Basile, 2014: 122–128). For instance, there was a peak of attention in the elections of 1963 (mean salience of 5.3%, without outliers), which preceded the first watershed territorial reform in Italy. Since then, the average attention given to territorial issues slowly declined until 1987 (1.8%), as the reform was approved and implemented. In 1992, the year of the LN’s electoral rise, the mean saliency score of mainstream parties considerably increased by 1.97 percentage points and it remained almost constant in 1994 and 1996. Once again, these elections preceded a period of major territorial reform culminating in 2001 in the reform of Title V of the Constitution, which transformed the territorial redistribution of competencies in Italy. After that reform, attention declined again in the 2001 election (2.2%). In 2006, only two coalition documents were issued, so a straightforward picture of the overall political debate is not possible. However, in the subsequent elections in 2008, the average salience was the same as in 2001. Lastly, in 2013, 9 there is a further increase of salience, although it is too early to say whether this will have an impact on the path of reforms in the current legislature.
In light of these findings, and contrary to the expectations derived from Meguid’s PSO model, the emergence and variation of salience of the territorial dimension cannot be considered uniquely as a reaction to the ethnoregionalist challenger. However, as Meyer and Miller (2013: 3) suggest, the emergence of a niche actor such as the LN may have at least contributed in the 1990s, among several other determinants, to the revitalisation of an already politicised issue that had been neglected over time.
Accordingly, the main premise of the baseline model can be reformulated as follows: H1- Reformulated: A secondary policy dimension is not always introduced into party competition from scratch by a niche actor; nevertheless, niche parties may play a role in contributing to the revitalisation of a debate on ‘old’ policy issues.
3.2. Ethno-regionalist actors and mainstream strategies
3.2.1. Mapping party strategies
The baseline set of hypotheses, drawing on the PSO model, introduces the strategies that mainstream parties are likely to adopt when confronted with a niche challenger, namely: one-dimensional, two-dimensional/accommodative or two-dimensional/opposing strategies. Before evaluating the impact of niche actors on mainstream party strategies, however, it is necessary to consider whether this baseline model captures the full repertoire of strategic alternatives that are available to mainstream parties.
In order to provide a synoptic picture of party strategies, the two tools of salience and position can be recoded and then combined into a typology. Saliency scores, ranging from 0 to 100, can be recoded by setting boundaries between each category to consider the median and the percentiles as parameters, thus obtaining the following categories: ‘no interest’ (0%), ‘minor interest’ (from 0.1 to 3.3%) and ‘major interest’ (from 3.3% upwards). Values of directional certainty ranging from –1 (clearly centralist attitude) to +1(clearly decentralist attitude), through zero, can be recoded as follows: values between –1 and –0.5, and 0.5 and +1 correspond to, respectively, anti-decentralist and pro-decentralist attitudes; values comprised between –0.5 and 0 or between 0 and +0.5 indicate documents that are leaning in either a centralist or decentralist direction. Documents holding a value of 0 represent ambiguous or contradictory stances (section 2).
By combining the recoded variables of saliency and directional certainty, eight categories can be obtained. Table 1 applies the typology to the coded Italian documents. A first ‘neglectful’ group is comprised of parties that do not address the territorial dimension at all, thus corresponding to the one-dimensional strategy. Then, there are two groups of ‘strong’ parties, whose documents emphasise territorial issues and express a clear stance on the territorial dimension. In contrast, the two ‘mild’ groups represent those cases in which manifestos do not put much emphasis on territorial reforms, but the direction of party attitudes is expressed in a definite and unambiguous way. Mild and strong attitudes are included under the broader two-dimensional strategy, which, in turn, can be either accommodative or adversarial. It should be noted that more than half of the coded texts show a clear pro-decentralist direction, be it mildly (21%) or strongly (30%) expressed, while nearly 5% and 2% of the documents express, respectively, a mild or strong adversarial strategy (see also sub-section 3.2.2, below).
Typology of party strategies and distribution of Italian documents across each category (%, N in parentheses) (1963–2013).
Finally, the ‘grey zone’ of the typology corresponds to the ‘blurring’ (Rovny, 2013) or ‘broad appeal’ (Somer-Topcu, 2014) strategy. It includes three groups of ‘reluctant’ documents and three groups of ‘empty talkers’. In both cases, parties either assume an ambiguous and vague direction, or they take decentralist positions on some issues and centralist on others within the same manifesto (for example, decentralist on health but centralist on education); they can lean towards a centralist, decentralist direction or assume a totally neutral direction, but they never exhibit a clear-cut preference in either direction. 10 Then, the difference between the two groups is just in the amount of attention paid to the secondary dimension: ‘reluctant’ documents pay little attention on the territorial dimension, while the ‘empty talkers’ put a lot of emphasis on it. The blurring strategy represents a significant share of the coded documents, with 24% expressing a blurred support for decentralisation, while nearly 8% used this strategy in an adversarial way.
Table 1 demonstrates that mainstream parties have at their disposal a wider range of strategies than those suggested by the PSO theory. Accordingly, the baseline hypotheses should be extended as follows: Besides strategies H1a–c- Parties may also respond to the niche electoral threat by: H1d- New. Introducing the secondary policy dimension into their discourse but blurring their position on it.
3.2.2. Does the dwarf affect the strategies of the giants?
The reformulated set of hypotheses can be tested in the Italian case in order to assess whether and to what extent the LN has actually affected mainstream parties’ strategies on the territorial dimension. Table 2 compares the distribution of party manifestos across the eight strategies, both before and after 1992.
Distribution of Italian documents of state-wide parties across the different party strategies, before and after 1992 (%, N in parentheses).
It is noteworthy that the one-dimensional strategy has seldom been adopted in both periods, and mostly by small parties. Overall, only eight out of 100 documents do not contain any statement on territorial issues. The two-dimensional/adversarial strategy, be it strongly or mildly expressed, seems on the whole to be underrepresented. It appears in 12% of the documents issued before 1992, and is used mainly by minor actors both from the extreme right and extreme left such as the Movimento Sociale Italiano and Democrazia Proletaria in 1976, with the exception of the liberal Partito Repubblicano Italiano in 1979. In the aftermath of the LN’s breakthrough, however, no mainstream party clearly opposed territorial reform until the recent general election of 2013, when the Movimento 5 Stelle articulated a (mild) centralist strategy.
After 1992, scepticism towards the ongoing process of decentralisation was mainly manifested through blurred decentralist strategies (11% of documents issued after 1992). This was the case, for instance, of Rifondazione Comunista in 2001 and Unione di Centro in 2008. This strategy was nevertheless also used before 1992 (6%) by those actors that were traditionally critical about territorial reforms, such as the Movimento Sociale Italiano and Partito Liberale Italiano.
The neutral-blurred strategy, featuring a complete opacity and vagueness of statements on the territorial dimension, was rare and was employed by only a few minor parties. In contrast, the blurred-decentralist strategy was relatively widespread before 1992 (30%); it was often used by the main governing party, the Democrazia Cristiana, mostly to express its ‘catch-all’ approach to territorial reforms, shifting from convinced support for decentralisation to a prudent defence of the central government. On the contrary, after the rise of the LN, it was used less frequently (20%) and on exceptional occasions. An example is in the manifesto of the centre-left coalition of Unione in 2006, whose heterogeneous composition made it difficult to find any point of clear convergence on several policy dimensions.
Finally, as already argued, the largest share of manifestos features a two-dimensional/accommodative strategy in both periods, although it was more frequent after 1992 (55%) than before the emergence of the LN (44%).
These findings demonstrate that, since the 1960s, Italian mainstream parties have converged around a two-dimensional/accommodative strategy, regardless of any autonomist electoral threat. The analysis thus reveals minimal differences in the strategic choices of the mainstream actors before and after the appearance of the LN. However, the latter seems to have definitively discouraged even the most resistant bulwarks of opposition to decentralisation. In fact, clear adversarial attitudes almost disappeared after 1992, with the exception of the recent case of the Movimento 5 Stelle.
Accordingly, the reformulated baseline hypotheses are only partially confirmed: although the ethnoregionalist electoral threat may have played a role in altering mainstream parties’ strategic decisions, its actual impact should be evaluated alongside other likely determinants of party strategies, such as institutional reform, ideology and party size (Basile, 2014).
4. The impact of the rhetoric of niche actors on party strategies: The role of issue framing
Thus far, the empirical analysis downplays the significance of the impact of the niche electoral challenge on the strategies adopted by mainstream actors on secondary policy dimensions, contrary to the expectations of the PSO model. However, rather than rejecting the PSO theory outright, this section re-evaluates the importance of the niche threat by considering another tool available to parties to define their strategies, which was overlooked by the PSO theory, namely how parties frame their political concerns.
4.1. Beyond the PSO model: The subsuming strategy
As argued in section 1, parties use frames to subsume issues belonging to the secondary dimension into other dimensions of party competition that are more important for them, such as the socio-economic or cultural cleavages. Accordingly, the subsuming strategy can be added to the set of strategies derived from the baseline model: Besides strategies H1a–d, parties may respond to the niche electoral threat also by: H2- New. Introducing the secondary policy dimension into their political discourse, while at the same time subsuming it into their core issue dimension in party competition.
The adoption of a subsuming strategy is complementary, rather than an alternative, to the other strategies. Indeed, subsuming is mostly a rhetorical strategy which allows parties to provide arguments for, and justifications of, their policy stances on a secondary dimension by using frames derived from their preferred core dimension.
Two scenarios, explicitly referring to the territorial dimension, are helpful for clarifying the role of frames and the subsuming strategy. In a first scenario, political actors agree on the same policy goal but they use different socio-economic arguments to justify it. For instance, decentralising reforms might be supported either by reassuring voters of the need to ensure territorial solidarity among sub-national units, or by claiming the ‘federalist’ principle of leaving resources where they originate. In a second scenario, the same socio-economic argument is used to support a diametrically opposed perspective on decentralisation. For instance, the need to strengthen the efficiency of the state machine can justify either the redistribution of competencies downward or a further centralisation of state powers. In both cases, however, parties clearly express their stances on the territorial dimension (a secondary dimension), but they also subsume it into the primary socio-economic dimension by framing territorial issues in economic terms.
The subsuming strategy can be directly related to the niche-mainstream relationship. As Baumgartner et al. (2008: 100–101 and 137) argue, although there is not one single event or factor to explain why old issues are sometimes revitalised by the introduction of new arguments, issue reframing is usually accompanied by the emergence of parallel movements in the political system, championing those issues.
4.2. New actors, using new frames to tell an old story
A typical example of how an emerging movement might introduce a new language in old, declining debates is represented by the rhetoric on territorial issues triggered by the LN in Italy after 1992. According to Giordano (2000: 446), although ‘political regionalism is not a new phenomenon in Italy’, Bossi’s movement brought to the Italian political debate at least three dominant themes: hostility against ‘wasteful’ Southern Italy; harsh critique of the inefficiency of the central State, epitomised in the slogan ‘robber Rome’ (‘Roma ladrona’); and claims of autonomy for the allegedly historical geographical entity named ‘Padania’. Although some of these themes were not totally original arguments, the LN articulated them in terms of a new and politically more contentious Northern question. Consequently, mainstream parties were induced to change their way of conceiving of, and talking about, territorial issues, thus linking them to different socio-economic or cultural arguments.
Within this framework, the following sections provide a qualitative analysis of the impact of the rhetoric of the LN on the subsuming strategies of Italian state-wide parties, by examining the dominant themes outlined above.
4.2.1. Northern selfishness vs territorial solidarity
Since 1992, the LN has displayed hostility towards the wasteful South, which is often portrayed as ‘an area in which people have no real desire to work and are only interested in claiming state benefits’, as opposed to the rich and productive North (Giordano, 2000: 459). For instance, the party has argued that the People of the North would be able to reduce the public debt to sustainable limits by itself, but it obviously cannot carry the burden of all the debt created to accommodate the needs of a politics managed in a wasteful way, in favour of parasites living off state aid. (Lega Nord, 1996, author’s translation)
As shown in Figure 2, before 1992 this argument can be found in just seven documents out of 50 and always displaying a negative sign. It was especially used by the Movimento Sociale Italiano (MSI), a nationalist party employing an adversarial strategy (either blurred or clearly two-dimensional) on territorial issues to warn about the threats to socio-economic cohesion posed by the ongoing process of territorial restructuring:
Use of the territorial solidarity frame before and after 1992. The groups of MSI-DN in the Chamber of Deputies and Senate will present … a group of bills, also on constitutional reform, in order to include the administration of local authorities, municipalities, provinces, regions, within the life of the central state, in order to avoid … the sharpening of the gap at the expense of the central-southern regions. (Movimento Sociale Italiano, 1979, author’s translation)
In the face of such harsh rhetoric, many mainstream parties were compelled to reassure their voters – and especially the Southern Italian electorate – that the proposed decentralisation was something different from the ‘selfish’ autonomy invoked by the LN, and that it would not call into question the mutual solidarity among Italy’s regions. Most of the parties that adopted this frame are also those employing a two-dimensional or blurred accommodative strategy. An example of this ‘reassuring’ use of the territorial solidarity frame can be found in the manifesto of the centre-left coalition of Ulivo in 2001: ‘It is still necessary to ensure forms and instruments of state and/or interregional solidarity, enabling the less prosperous regions to ensure adequate services and to promote the development of their communities’ (Ulivo, 2001, author’s translation).
Against this trend, however, in 2013 the newly founded Movimento 5 Stelle again proposed an adversarial use of the territorial solidarity frame, although limited to the prospects of decentralisation in the field of public health: ‘Two elements are undermining the universality and homogeneity of the National Health System: devolution, which delegates health assistance to regions and sharpens territorial differences, and private health services’ (Movimento 5 Stelle, 2013, author’s translation).
4.2.2. Decentralisation: A panacea for the inefficiency of the central state?
In 1992, the Italian political system was in a phase of political turmoil due to the so-called Tangentopoli corruption scandals that emerged in the early 1990s. Within this framework, the LN widely exploited the rhetoric of the inefficiency of the central state and its crisis of legitimacy. Under the tough slogan of ‘Roma ladrona’ (‘robber Rome’), it blamed the inefficiency of the central institutions, which were often described as a heavy bureaucratic machine marked by corruption and waste of resources at the expense of taxpayers in the North.
This argument resembles the efficiency frame, concerning the alleged influence of territorial reforms on the efficiency of institutions at the central level. When positive, this frame indicates that a party sees territorial reforms as enhancing efficiency and modernisation. On the contrary, those parties maintaining that decentralisation would further jeopardise the efficiency of the central state and supporting a strengthening of the state machinery, employ a negative interpretation.
As Figure 3 shows, this frame was widely used both before and after the emergence of the LN; nevertheless, before 1992, this argument displayed a negative sign in eight out of 30 documents. For instance, the Movimento Sociale Italiano often justified its mainly centralist attitude by arguing that the central bureaucracy would have not benefited from the creation of further layers of public administration. On the other hand, the Partito Liberale Italiano in 1979, notwithstanding its support (although blurred) for decentralisation, warned of the inherent risks of inefficiency: ‘in the current state of public structures, with their tendency to increase their costs and bureaucracy, it would be a serious and dangerous mistake to add the regional structure to other public facilities that already exist’(Partito Liberale Italiano, 1979, author’s translation).
Use of the efficiency frame before and after 1992.
Since the 1990s, however, statements opposing territorial reforms in the name of state efficiency have mostly disappeared from political discourse, with a few exceptions (Partito Popolare Italiano in 1994, Patto Segni in 1994 and Scelta Civica in 2013). Rather, mainstream parties have sought to appease public discontent with the state’s performance by pointing out the potential benefits of territorial reforms. Accordingly, they were able to convey to voters their strong commitment to enhancing the efficiency and responsiveness of the central state machinery.
4.2.3. The Italian Republic is ‘one and indivisible’, but …
Article 5 of the Italian Constitution, which recognises and promotes local autonomy, points out that ‘the Republic is one and indivisible’. Nevertheless, when the LN entered the political arena it broke the taboo of the inviolability of national unity. Claiming the historical roots of an alleged ethnicity based on the invented territory of ‘Padania’, the movement led by Bossi has often sought to achieve the autonomy of Northern Italy. The fragmentation of the Peninsula was now invoked and the ideals of the Risorgimento were no longer untouchable: ‘[the project of reconstruction of the Italian state, based on federalism] implies moving on from a constrained and fictitious unity’ (Lega Nord, 1992, author’s translation).
The national unity frame captures those cases in which parties refer to the need to preserve the integrity of the state. The positive sign applies to the most radical arguments, where national unity is questioned in the name of an extreme form of decentralisation; on the contrary, the negative sign is applied to parties that endorse the principle of the indivisibility of the nation, regardless of its support for territorial reform.
Before 1992 very few mainstream parties (nine out of 50) included this frame in their documents (see Figure 4), indicating that the unity of the nation was beyond question for the majority of political forces in the country. When present, however, it always displayed a negative sign and was used in parallel with both adversarial and accommodative attitudes. Indeed, the national unity frame represents an intriguing example of how the same argument can be used to support diametrically opposed perspectives.
Use of the national unity frame before and after 1992.
For instance, the Movimento Sociale Italiano often invoked this frame to oppose projects for territorial restructuring by arguing that they risked jeopardising the integrity of the country: the MSI has undertaken in the media and in Parliament a clamorous and brave battle against regionalist reform … to challenge the ultimate and immediate goals that the left … pursued and still intend to pursue, with the acquiescence of the Christian Democrats, in order to disintegrate the State. (Movimento Sociale Italiano, 1968, author’s translation)
However, when the LN started to demand greater Northern autonomy, successfully invoking the prospect of secession, it was clear to all mainstream parties that national unity was no longer an unquestioned value, at least for a part of the electorate. As a consequence, state-wide parties were forced to defend the integrity of the state in their documents; they had to reassure those voters, worried by demands for secession, that to support decentralisation would not mean jeopardising the traditional nation-state and its values. For this reason, the use of this frame to justify the endorsement of a decentralist position became much more frequent among mainstream actors than in the past, featuring in half of the documents analysed for the period 1992–2013. A clear example of this use of the national unity frame, with the explicit aim of marking a clear distance from the secessionist demands of the LN, can be observed in the manifesto of the Partito dei Democratici di Sinistra in 1994: ‘Here there is a need to organise a shift towards a new regional state inspired by federalism, which, by creating strong powers of self-government, will emphasise the vocations, cultures, resources of this ancient country, thus consolidating national unity’(Partito dei Democratici di Sinistra, 1994, author’s translation).
Conclusion
This article has analysed the dynamics of party competition between niche and mainstream actors in a two-dimensional space, by testing a set of hypotheses grounded in the PSO model developed by Meguid (2008). These hypotheses were tested by means of a theory-modifying case study, namely party competition on the (secondary) territorial dimension in Italy between 1963 and 2013. The empirical analysis highlighted several shortcomings with the baseline model and provided grounds for its reformulation as follows.
Firstly, niche actors do not always introduce secondary policy dimensions from scratch; rather, the latter often precede the emergence of niche actors. The territorial dimension, for instance, was present in Italian political debate well before the electoral emergence of the LN. This finding is contrary to the main premise of the PSO theory. Secondly, in the repertoire of available strategies, parties may also blur their position on the secondary dimension, which is not considered by the PSO theory. Data from the Italian case reveals that parties often prefer to avoid expressing any clear stance on the secondary dimension of policy or they show contradictory stances on different aspects of the same dimension.
Thirdly, the impact of the electoral niche threat on mainstream party strategies is not as straightforward as the PSO theory suggests. The empirical analysis examined whether Italian state-wide parties altered their strategies on the territorial dimension after the emergence of the LN. The findings reveal limited differences in mainstream party strategies before and after 1992, so that it is difficult to identify the clear-cut impact of the ethnoregionalist actor on the dynamics of party competition in Italy’s two-dimensional competitive arena. The empirical analysis thus downplays the impact of the niche actor on the dynamics of party competition, as compared with the expectations of the baseline model. However, rather than dismissing the PSO theory in its entirety, this article argues that it requires modification in order to take into account a further strategic tool that parties can use in their strategies, namely framing.
The mixed quantitative and qualitative analysis conducted on the political discourse in Italy revealed that the LN has introduced a new language into political debate relating to the territorial dimension. Mainstream parties have reacted to this new rhetoric by reframing and redefining these issues differently from the past. Issue framing provides an additional way of conceiving the influence of niche actors on the dynamics of party competition. Specifically, it points to the impact of the new language introduced by niche actors on the strategies of mainstream parties in a secondary policy dimension. This finding offers a substantial development of the PSO theory as conceived by Meguid (2008), and further comparative work should focus on testing the theoretical insights of this study in other multidimensional and multinational settings.
Footnotes
Acknowledgements
I would like to thank the editors of this special issue, Anwen Elias, Edina Söcsik and Christina Zuber, for their support and assistance during the writing of this article. I would also like to thank the two anonymous reviewers for their constructive comments. A previous version of this article was presented at the Department of International Politics, University of Aberystwyth, 12–13 December 2013. I would like to thank all the participants for their feedback and comments.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Notes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
