Abstract
Using Sweden as an example, we propose the hypothesis that local candidate lists rest on a different rationale compared to national party lists, more specifically a rationality that has developed as a result of an altered relationship between the local and national levels and is expressed in terms of localism. The emergence of new local lists is most likely an effect of the challenges to the welfare state posed in the past 40 years. However, although local lists are essentially a recent phenomenon, our examples suggest that their existence can also be explained in terms of historical patterns of dissent from and resistance to state intervention at the local level.
The problem
The appearance of new local lists of independent candidates (henceforth ‘local lists’), 1 reflects fundamental changes in the Western European party systems, including the erosion of established party identities (Dalton and Wattenberg, 2000; Mair et al., 2004; Reiser and Holtmann, 2008). Four different models are usually applied to explain the dynamics and rationale of the national party systems: institutional factors, including the role of local self-government; organizational and structural factors, including so-called core-periphery models; and actor-oriented models of explanation. All of these contribute to a general understanding of local lists (Erlingsson, 2005). However, we argue that they fail to explain their rationality. For instance, Wörlund (1999, 2008) has suggested that local lists in post-war Sweden were generally considered ‘historical artefacts’, that is, anomalies illustrating that the municipal level had not yet been properly politicized by the dominant parties in parliament (also cf. Lipset and Rokkan, 1967). This interpretation accounts for organizational features (the stress on the ‘failure’ of the nation-state-level parties) and structural aspects (the inclusion of a core-periphery perspective). This does not, however, explain the typical pattern of the last 20 years – the numerous new local lists.
As historians, we propose the hypothesis that local lists follow a different rationale compared to national party lists, and more specifically a rationale that has emerged from altered state–local relations. This is partly implied by the definition of local lists used here: their characteristics are that they are ‘solely locally organized’; that they compete for voters in ‘one single local jurisdiction’; and that they attend to local issues only (Boogers, 2008; Boogers and Voerman, 2010; Holtmann, 2008: 11). Furthermore, their presence is most likely related to welfare state challenges in the past 40 years.
During this period there were demographic changes and new socioeconomic challenges that put many municipalities under increased economic pressure, aggravated by the simultaneous restructuring of traditional nation-state institutions and the gradual withdrawal of the nation-state from the local level as a result of de-regulation and privatization. Although these changes in part are legitimized by European Union (EU) regionalization policies, they still stand in sharp contrast to the historically typical pattern of Northern Europe: extensive welfare institutions largely implemented by the municipalities. This structure was grounded in a long tradition of local self-government in most of these countries (cf. Esping-Andersen, 1990). Local lists appeared in response to these changes. Analogous to the definition proposed by Holtmann (2008: 11), they are political expressions of localism and cannot be interpreted according to the rationality of nation-state-level parties. But whereas Reiser (2008: 293–294) concludes that local lists, at least in Western Europe, are a modernizing feature of existing nation-state-level party systems, our hypothesis suggests that they signify a new type of political mobilization that is neither easy to incorporate into existing party systems, nor easily conceptualized.
Given the scenario outlined, historical contextualization and ideological analysis is required to understand the role of local lists in different institutional settings and countries (see below). This research note does not allow for a systematic testing of our hypothesis, but by presenting it in greater detail, and giving examples, we hope to encourage further research and renewed debates on local lists as a historically contingent phenomenon.
Local lists in the European context
Local lists may compete for voters on the basis of an environmentalist agenda, or they may be part of the new radical right, as suggested by Kitschelt’s (1989, 1995) analysis of emerging political divisions in Western Europe. In this respect they do, indeed, reflect ideological dimensions typical of late modern society. However, most local lists do not fit this description. Mair (1997: 28–31) has also argued that increased electoral volatility is still contained within established party blocks, most notably traditional left/right class-cleavage blocks. Despite this, the vast majority of local lists focus their programmes on various specific local development and planning issues, some of them environmentalist in outlook, others not.
Local lists may be of minor but, possibly, increasing importance in some countries, such as England (cf. Heath et al., 2003; Swenden and Maddens, 2008). They may also, as suggested by election results, play a more prominent role not only in countries such as the Netherlands, but also in federal states such as Germany, at least in some Länder, for instance Nordrhein-Westfalen, or Schleswig-Holstein: 8.1 per cent of all votes in the 2009 local elections were cast in favour of various independent ‘Wählergruppe’ in the first case, and closer to 5 per cent in the second case (2008 elections) (Gemeindewahl…, 2008: 5; Kommunalwahlen…, 2009: 13). A similar pattern applies to Estonia, viz. an example of a post-communist state with democratic institutions, still maturing: at least judging by the large number of independent lists competing in the local elections (Elections in Estonia 1992–2011, 2012: 90, see also Reiser and Holtmann, 2008).
The Nordic countries are interesting in this context because of their long traditions of local self-government, unlike the situation in southern Europe. But there are also differences between the Nordic states. For instance, the case of Norway lends some support to the above-mentioned ‘artefact-hypothesis’, since the share of votes for local lists – despite an increase in the early 1990s – has actually decreased in recent elections (suggesting consolidation of the national-level parties at the local level). 1 The Finnish pattern also indicates that variation between different sub-national regions is possible (as measured in turnout for local lists). 2 In Sweden, the increase in electoral support occurs with regional variations (see below).
The Netherlands and Sweden
Two countries invite more detailed scrutiny, that is, the Netherlands and Sweden – in a sense they are opposites with a diverse party system in the Netherlands and a party system characterized by great stability, at least until recent years, in Sweden. In the Dutch case local lists received an average of 18.9 per cent of all votes in the 1974 local elections (Boogers and Voerman, 2010: 79). In the 2002 and 2010 elections this percentage had gradually risen and stabilized at a level of 23.1 per cent (Table 1). Many of these parties appear in the statistics under the heading of ’Gemeentebelangen’, yet others have more regular names such as ‘Burger Belangen Beek’ in Beek, Limburg, or ‘Sterk Meppel’ in Meppel, Drenthe. In the case of Sweden, there has been a similar pattern of rising support for local lists since the 1970s, although at a more modest level.
Votes for municipal political parties in the 2002 and 2010 Dutch and Swedish local elections (per cent).
Sources: http://www.electionresources.org/western.europe.html#NL. Gemeenteraadsverkiezingen 2002, 2010, http://www.nlverkiezingen.com/GR_GR.html (26 September 2012); Statistics Sweden. Allmänna valen 2002, III, Table 5; Allmänna valen 2010, III, Table 8.
Part of the reported increase of support for local lists in Sweden (Erlingsson, 2005: 50–1; Wörlund, 1999: 54), however, includes national-level parties mainly supported by specific regions or localities (in effect ‘local party branches’, see Boogers and Voerman, 2010: 78), such as the Sweden Democrats (‘Sverigedemokraterna’), an extreme right-wing party that eventually entered parliament after the 2010 general elections. If this type of party is included, the statistics yield an increase from 1.6 per cent in the 1976 local elections to 6.3 per cent in the 1991 elections. In the following elections they have hovered around 4 per cent. 3 Even if we include local lists only (possible at an aggregate level from the 2002 elections and onwards), there is still an increase in voter support – from 2.2 per cent in the 2002 elections to 2.6 per cent in the 2010 elections (Table 1).
Before we focus on Sweden, the figures in Table 1 require three considerations. Firstly, pertaining to the problem of dealignment – a feature suggested most notably by the Dutch pattern – the findings of Mair (1997: 28–31) nevertheless indicate relative block stability of voting in the Netherlands. More recently, these results have partly been confirmed by van der Meer et al. (2012), whose study, like Mair’s analysis, concerns the national rather than the local level. By comparison, Boogers and Voerman suggest that local lists perform a function by establishing new ‘dividing lines between local and supra-local interests’ (2010: 88). They conclude that local lists are prototypes ‘of a modernized party organisation’ (Boogers, 2008: 166) and, hence, serve as new intermediaries ‘between politics and society’ (Boogers and Voerman, 2010: 88)., Logically, however, the emergence of new ‘dividing lines’ between local and national levels does not automatically translate into local lists having a new mediating role in the vertical dimension. Rather, it may well be the other way around. Local lists operate according to a different logic, according to which party ideology and block identities on the national level become less of an issue as in the case of ‘Sterk Meppel’ in the Netherlands, or ‘Åsele kommunlista’ (formerly ‘Åselepartiet’) in Sweden. According to their programmes, they are concerned with any local issues, left, right, or otherwise, of interest to the enhancement of the municipality. ‘Åsele kommunlista’ also stresses the importance of cross-party cooperation. 4 Cooperation to resolve practical issues is more important than positioning itself on a left–right scale. This does not mean that local politics lacks ideological dimensions. ‘Localism’ per se may be considered as an ideological concept (see below.) But even so, a localistic outlook on politics does not necessarily require mediating functions with respect to ‘supra-local interests’.
Secondly, there is the question whether increased support for local lists at the expense of national-level parties and their local branches relate to decreasing levels of trust in politicians more generally. No such pattern is detectable, either in the Netherlands, or in Sweden, according to the European Social Surveys (Table 2). In both countries the proportion of citizens with above average levels of trust in politicians (categories 6–9 and ‘complete trust’ (Table 2)) increased between 2002 and 2010 (cf. the study by van der Brug and van Praag, 2007, which reports an increase rather than decrease in political trust, including institutions such as parliament and the government).
‘How much do you personally trust politicians?’ The Netherlands and Sweden, 2002 and 2010, per cent.
Sources: European Social Survey. N = 1290 (2002) and 1350 (2010, The Netherlands); 715 (2002) and 765 (2010, Sweden).
Finally, we should note the spatial differentiation of voting for local lists (Tables 3 and 4). This differentiation is obvious in the Netherlands and noticeable in Sweden. In the Netherlands, certain provinces, particularly Drenthe, Limburg and Noord-Brabant, stand out. Historically speaking, the latter two form part of cross-border regions. Although the support for local lists dropped somewhat in Limburg and Noord-Brabant between 2002 and 2010 (Table 3), the latter province in particular was the scene of extensive territorial adjustments in terms of mergers of municipalities in the late 1990s. Likewise, local lists tend to be more frequent in certain areas of Sweden. These areas represent three types of municipalities: one group of large but sparsely populated municipalities in the north; small and densely populated municipalities in the Stockholm metropolitan area; and rural municipalities in the south and south-west of the country (Table 4). Unlike continental Europe, Sweden lacks regions based on distinct linguistic and ethnic identities. In addition, Wörlund’s (1999, 2008) findings suggest that local parties seem to have emerged with no apparent connection to spatial or socioeconomic differentiation. Yet, in light of the development from the late 1990s onwards there is reason to return to the possible role of these factors, but we must begin with a brief review of existing explanatory models.
Votes for municipal political parties by regions in the 2002 and 2010 Dutch local elections (per cent).
Sources: http://www.electionresources.org/western.europe.html#NL. Gemeenteraadsverkiezingen 2002, 2010, http://www.nlverkiezingen.com/GR_GR.html (26 February 2012).
Votes for municipal political parties by regions in the 2002 and 2010 Swedish local elections (per cent).
Sources: Statistics Sweden. Allmänna valen 2002 (Stockholm, 2003), III, Table 5; Allmänna valen 2010 (Stockholm, 2011), III, Table 8.
Types of explanations
Following Erlingsson (2005: 53–60, 85–109), four basic types of explanations can be applied to explain the dynamics of party systems, viz. institutional, organizational, structural and actor-oriented models. To a varying extent they incorporate traits of Downsian (Downs, 1957) ideas of the rationality of politics. Usually, political scientists draw on several of these models simultaneously, such as in the case of Duverger’s (1967 [1954]) analysis of European party systems, or Katz and Mair’s (1995) discovery of cartel parties. Duverger stressed the importance of institutions (e.g. electoral systems and election laws), but also attached significance to the dynamics of individual party organizations. Katz and Mair focused on the relations between party organizations, civil society and, in particular, the state. More current examples include a majority of the contributions in Reiser and Holtmann (2008) on the organizational features and institutional role of local lists; van Biezen and Rashkova’s (2012) study of the interrelation between state regulation and the organizational dynamics of party systems; Poguntke’s (2012) analysis of German catch-all parties; and the work of Lowery et al. (2011), which combines institutional and organizational ecology perspectives on the Dutch single national district structure. The ‘rules of the game’, then, is quite different compared to Germany, for instance, where constitutional path dependencies differ between the Länder and, presumably, produce different outcomes with respect to political organization (cf. Freitag and Vatter, 2009). Nevertheless, the issue agenda, Lowery et al. conclude, and ‘the space it provides for parties to build independent identities’, become crucial factors behind changes in the density of the party system (2011: 20).
Organizational factors are stressed by, for instance, Bolleyer (2012) in her analysis of party typologies in Western Europe. Drawing, among others, on Panebianco (1988) and Katz and Mair (1995), the relation between party hierarchy and stratarchy is analysed. However, Gunther and Diamond (2001), has also pointed out that a focus on typologies must also account for the functions that different parties execute, which, regardless of typology, are dynamic and differ across time. Arguably, the same holds for local lists since they include a wide range of organizations including anything from clear cut ‘parties’ to loosely integrated interest groups (cf. Gidlund, 1978).
However, the above models are also important since they allow for including institutional and other macro-level changes, such as the impact of the EU on political organization (Herrschel and Tallberg, 2011; Keating and Loughlin, 1997; MacLeod, 2001). The same holds for structural models of explanations, including perspectives that put stress on a presumed shift in the Western world towards a more individualized society based on post material values (cf. Inglehart, 1997). Political cultures, though, differ not only between countries, but also on sub-national level, analogous to Steyvers et al. (2008) review of Belgian local lists, Freitag and Vatter’s (2009) study of Germany, or Vergani’s (2012) analysis of the Democratic Party in Italy. Such features should not be confused with the impact of the neighbourhood effect as illustrated by political geographers (David and van Hamme, 2011). In Vergani’s case, regional differentiation draws on culture and is tangible enough to produce variation in the local party organization and in a way that partly supports Putnam’s (1992) observations of the patronage culture in southern Italy.
Actor-oriented models of explanation, finally, incorporate the role of individuals in party systems formation and change. These include Aldrich’s (1995) analysis of the US, and Erlingsson’s (2005) study of Sweden. Aldrich (1995: 4) maintains that any number of personal reasons, including pure ambition and office-seeking, but also altruistic motives can move actors to form parties. In addition, actor-oriented models have the potential to include local context and hence also locally specific modes of rationality (Agnew, 1996). Yet, when put to the test, Erlingsson, for instance, faces problems in accounting for the role of the local level to party formation. The approach shows a tendency to draw on generalized psychological assumptions about the behaviour of individual actors: ‘hot feelings’ and frustration among ‘political entrepreneurs’, Erlingsson claims, are the catalysts behind party formation (2005: 240–55). But this does not explain why these hot feelings occur in the first place.
Consequently, it appears that organizational explanatory models (which in this case put stress on the capacity of established parties to integrate new grass root impulses and ideas), discarded by Erlingsson, in fact play an important part in explaining individual behaviour at the micro level, because discontent with established parties provides the political entrepreneurs with motives (Erlingsson, 2005: 170–99). We are left with the same type of explanation as suggested by the representatives of ‘Åsele kommunlista’ in connection with the 2010 election campaign: the voters were fed up with the prospects of yet another round of confrontational ‘bloc politics’. 5
Importantly, the above type of explanation does not include the consideration that local lists are determined by a different type of rationality compared to nation-state parties. A more detailed explanation of why ‘hot feelings’ and disappointment with established parties erupt to begin with is lacking. Local lists certainly compete on ‘electoral markets’, but they do so on the basis of a strictly localistic outlook. Discontent need not relate to the same type of ideological cleavages as on the national level (cf. Edelman, 1988). But this interpretation implies a narrow definition of ideology as a clearly formulated and coherent set of ideas about society. However, if we consider ideology as a latent and discursive feature of social and political power relations (Eagleton, 1991; Fairclough, 1995; Thompson, 1984; van Dijk, 1997), localism is definitely situated in the realm of ideology. Accordingly, local politics becomes something more than just the daily running of the mill. In the words of the programme of the ‘Lidingöpartiet’ near Stockholm, it is a means to defend the idiosyncrasies and traditions of a small, coastal community from its expansive, metropolitan neighbour and its minions – the established political parties. 6 The claims of local lists, such as ‘Lidingöpartiet’, or ‘Åsele kommunlista’ in northern Sweden 7 – that they occupy the middle ground of politics – is in fact an ideological thrust aimed at the failure of the established parties to manage issues relevant to the local community. ‘Hot feelings’ and disappointment are only the symptoms of something more fundamental, an alternative way of looking at society and politics compared to the national level. From a localistic perspective, the demographic and socioeconomic make-up of the electorate becomes less important although there are exceptions in the form of local senior parties, for instance, in the Netherlands. But all in all, the business of municipal parties is the locality.
Local lists: A historical perspective
The rationality typical of local lists has emerged on the basis of altered local/nation-state relations. This process is most likely related to the welfare state challenges of the past 40 years, which – as far as Sweden is concerned – has occurred against the background of demographic, administrative and socioeconomic changes. A recent report from the Swedish Association of Public Housing Companies (SABO), summarizes the main trends: declining population in many rural municipalities and an ageing population, as well as a shift from secondary to tertiary production and, hence, increased unemployment leading to an eroding tax base, make it difficult to maintain the level of service demanded by the citizens (Tillåtet stöd…, 2013: 6). According to Judt (2011 [2010]), this is a pattern typical of Europe as a whole, especially with respect to its causes: de-regulation, privatization, cuts in government spending and a gradual withdrawal of the nation-state from the local level. Additional factors that have speeded up and, in part, legitimized this development are EU regionalization policies. ‘Subsidiarity’ has become an appealing argument for altered local/nation-state relations (cf. Herrschel and Tallberg, 2011). But in connection with the downsizing of the welfare state, the outcome may well be that weak municipalities are left to fend for themselves. This is a new scenario at odds with traditional perspectives on local–central relations, particularly in the Nordic countries. Although Judt’s analysis may be considered as a plaidoyer for the good old days of the extensive welfare state, the weakening of national–local relations indicates a dilemma for national parties with respect to their legitimacy on the local level. Although not tested extensively here, our hypothesis seems to reflect real-life opinions voiced locally, such as the programme of the ‘Åsele kommunlista’: the economy of the municipality has come under increased strain in recent years ‘because of decreasing population and government policies’. 8
So far, our explanation provides necessary but not sufficient grounds for explaining local lists. For one thing, municipalities in the dynamic Stockholm region seemingly deviate from the pattern described here. Yet, the programme of the ‘Lidingöpartiet’, for example, resembles that of ‘Åsele kommunlista’ in certain key aspects: in both cases the core issue is the preservation of the locality. The closing down of secondary schools is an issue, but so is the ambition to avoid the construction of a new ‘Manhattan’ at Lidingö, in light of the population boom in Stockholm and the increased pressure on housing. 9 Nimbyism is, however, only part of the picture. As for the relation between socioeconomic variation and the geographic diffusion of new local lists, previous research has so far not detected any strong connection (Erlingsson, 2005; Wörlund, 1999, 2008). What is lacking in these models is an important intermediate historical ‘variable’. In the Dutch case, provinces such as Noord-Brabant and Drenthe with high levels of support for local lists (Table 3) have also recently experienced extensive mergers of municipalities. In a historical perspective, there is a similar pattern in Sweden. New local lists are frequent in municipalities in the south and south-west, 10 where historical research has identified traditions of local dissent and resistance to previous administrative reforms (Wångmar, 2003: 213, 222–225, Appendix 2.1, 2.2). Although further testing is required, election results indicate that votes for local lists in this type of municipality correspond geographically exactly to those older municipalities that were incorporated into larger units in the 1950s and 1970s. 11 In turn, this might signal an historical foundation for localism as ideology. Thus, in cases where localistic sub-cultures of self-reliance and independence have some durability across time, we suggest that they provide the foundations of political mobilization and help trigger party formation when socioeconomic differentiation and discontent with the welfare state reach a critical level.
Conclusion
Although tentative, our hypothesis suggests that the rationality of local lists does not relate primarily to institutional or organizational changes at the party systems level. Neither do these local lists appear as historical artefacts that have outlived the consolidation of nation-state-level parties. Rather, they are a novel phenomenon as a result of macro-level institutional and structural factors in combination with historical legacies. The Swedish example shows, as we suggest, that historically rooted, localistic sub-cultures may trigger political activism in stages when local/nation-state relations, as presently, are altered and re-negotiated. In this process local lists become an expression of dissent in relation to national-level parties. However, in addition to the examples and material used here – party programmes and election results – systematic testing of our hypothesis will require more extensive data, especially local-level interviews, in order to gauge the importance of localism as a distinct ideological dimension and platform for political mobilization.
Footnotes
Funding
This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.