Workers,farmers and Catholicism: A history of political class coalitions and the south-European welfare state regime

Variants of a cleavage

Esping-Andersen’s Three Worlds of Welfare Capitalism became an instant classic when it was published in 1990. Together with Peter Hall and David Soskice’s (2001) Varieties of Capitalism, one cannot think of any book that had a greater impact on the field of comparative welfare state research and on comparative political economy more generally, in the last four decades. In fact, it defined an enormously ambitious research agenda that most of us as researchers in this field – explicitly or implicitly, whether we admit it or not – share and still pursue, since its intellectual potential is far from being exhausted.

As everybody knows, Esping-Andersen proposes to distinguish three welfare state regimes, a social-democratic Scandinavian, a liberal Anglo-Saxon and a conservative continental one (Esping-Andersen, 1990). Elsewhere, I have tried to show that this three-regime typology closely corresponds to distinct patterns of party-political interest representation in the postwar era (Manow, 2009; Van Kersbergen and Manow, 2009) – very much in line with Michael Shalev’s (2007) important contention that the ‘key causal argument of The Three Worlds is that countries cluster on policy because they cluster on politics’ (p. 289). Among these patterns, we find a two-party system in the plurality electoral systems of the Anglo-Saxon world, in which more often than not conservative governments rule and in which subsequently a residual system of social protection and income redistribution directs the middle class to search for market solutions in questions of education (e.g. private schools, high student tuitions) (see Ansell and Gingrich, 2013; Iversen and Stephens, 2008), income maintenance in old age (e.g. life insurance, company pensions, house ownership) or health (private health insurance). The countries with generous welfare states all have proportional electoral systems and therefore party systems with a higher effective number of parties. Here, genuine middle-class parties enter into coalitions with social-democratic ones and tax the rich and share the benefit (see, for the general argument, Iversen and Soskice, 2006).

Within this world of redistributive, generous welfare states, one can distinguish basically two coalition patterns that emerged over the postwar period: Scandinavian red–green coalitions between Social Democracy and agrarian parties and a continental pattern with (implicit or explicit) ‘red–black’ coalitions between Social and Christian Democracy. I argued that a crucial factor for the difference between the Nordic and the continental pattern is the presence of a strong state–church conflict in continental Europe. This conflict was absent in the Nordic countries due to the fact that here Lutheran state churches were not only not opposed but more or less identical with the emerging and expanding nation-state of the late 19th century. In its stead, the conflict between the first and second sector, between city and countryside, gave rise to agrarian parties (Arter, 2001), which, in turn, usually cannot be found in the party systems of continental Europe.

In his 1990 book, Esping-Andersen hints at the causal model behind his three-regime typology, a model he had developed already five years earlier (Esping-Andersen, 1985). He especially highlights the ‘political coalition building in the transition from a rural economy to a middle-class society’ (Esping-Andersen, 1990: 32). In his view, it was crucial whether or not farmers – in almost all European countries an electorally central group at the moment of mass-democratization – could enter into a coalition with workers, an argument nicely captured in his phrase that it is ‘one of history’s many paradoxes … that the rural classes were decisive for the future of socialism’ (Esping-Andersen, 1990: 30). ¹ Esping-Andersen (1990) argues that the potential for such an alliance was greater in countries where farming was capital-intensive than where it relied on cheap labour (p. 30). This echoes the famous Barrington Moore argument about the antidemocratic stance of large landholders (Esping-Andersen, 1985: 36, fn. 14; Moore, 1993 [1966]; Rueschemeyer et al. 1992). We know today, however, that landholding patterns fail to explain the European class coalitions in the first three decades of the 20th century (Ertman, 1998; Luebbert, 1991, see especially pp. 308–10). It, therefore, appears promising to reconsider the nexus between ‘political class coalitions’ and the welfare state with a special focus on the potential for worker/farmer alliances.

In this article, I highlight the split between communist and social-democratic parties in the countries of southern Europe (Italy, Spain and Portugal, but also France). I argue that the split on the left is closely related to the decidedly anti-republican position held by the Catholic church in the mono-denominational Catholic countries of Europe’s South, since the deep divide between a sharp anticlerical labour movement and pious farmers under close tutelage of the church left the political left without allies for a reformist strategy. This furthered its radicalization since – in stark contrast to the situation in Scandinavia – farmers did not help socialist parties out of the ‘working-class ghettos’ (Esping-Andersen, 1985: 9). Therefore the left in these countries could not ‘escape political isolation’ (Esping-Andersen, 1985: 9). My argument thus is that the split on the left is first and foremost the result of a highly polarized conflict between a clerical right and an anticlerical left. It is neither grounded in the conflict between labour and capital as such nor the result of the ideological hegemony of large landholders over the family peasantry (compare Rueschemeyer et al. 1992), nor a consequence of the political mobilization of agricultural workers by socialist parties (Luebbert, 1991).

‘Political Catholicism’ developed in countries where the process of nation building provoked a vehement conflict between the state and the Catholic church (Kalyvas, 1996). But political Catholicism appeared in two currents: as an intransigent and reactionary enemy of liberalism and modernity in the mono-confessional countries of southern Europe and in a more moderate, centrist version in the denominationally mixed countries of continental Europe (Martin, 1978). This relates to the thesis proposed here, namely that the conflict between the nation-state and the Catholic church manifested itself not only within the bourgeois political camp in the form of Christian-democratic parties in continental Europe but was also reflected on the political left. There it materialized as a rift between reform-oriented (social democratic) and radical (mainly communist, but sometimes anarcho-syndicalist) wings of the workers’ movement in those countries in which the Church took a decidedly anti-republican stance. In turn, this rift had long-term consequences for postwar government composition and for the way in which social politics and industrial relations developed in the South. It finally gave rise to a welfare state model, or to a political economy more generally, that in Esping-Andersen’s original contribution had not been identified as a distinct regime.

The article is organized as follows: I start by highlighting the institutional distinctiveness of the welfare state model for which the rift between social democrats and communists has been so influential: the southern regime. I then develop my argument about the correspondence between welfare regimes and party systems by looking for the factors distinguishing the European party systems – these factors mainly manifested themselves at the moment of mass-democratization after the First World War (WWI) (Caramani, 2000, 2004; Lipset and Rokkan, 1967; Rokkan, 1970). I subsequently sketch my argument about the origins of the split between the reformist and the revolutionary wings of the workers’ movement and then briefly present basic vote share- and government composition-data for post-Second World War (WWII) Europe in support of my argument. By way of conclusion, I return to Esping-Andersen’s class-coalitional theory and propose a simple genealogy of Western Europe’s four welfare state regimes.

Before I start, however, three qualifications or disclaimers are warranted: first, the following argument is necessarily stylized and simplifying and papers over a lot of historical and case-specific complexities. I fear that many country experts will find fault with my treatment of their pet case. Yet, I decided in favour of – at times radical – ‘reduction of complexity’. In this I feel myself very much in agreement with Esping-Andersen who always declared to prefer analytical parsimony over the exuberance of case- and time-specific detail (compare Esping-Andersen, 1997, 1999). Moreover, anybody disagreeing with my argument will be asked to offer a more suitable explanation for why the split on the left proved so persistent and influential in the south of Europe but nowhere else. In that this split explains much of what is specific about the southern regime, experts seem to be agreed (see, for example, Ferrera, 1996: 30–1). My second disclaimer concerns the time period under investigation. Since I try to develop an argument about the interwar class-coalitional origins for the postwar regime differences, my argument refers mostly to the pre-1945 period. This, of course, does not mean that the postwar period can be conceived as uniform and simply unfolding a dynamic that was already fully implied at its start. Third, and again with reference to Michael Shalev’s (2007) excellent reconstruction of the ‘Three Worlds’-argument, I would like to emphasize that the main theoretical task is not to assign regime-labels to countries, but to investigate how much the ‘proximity or distance of a country’s policy profile from the three ideal-types’ is ‘matched by its political configuration’ (p. 291). The split of the left as a systematic characteristic of the southern European party systems would then also justify distinguishing a southern welfare state regime.

The southern regime

What are the institutional characteristics of a regime that Esping-Andersen initially did not treat as a distinct model, but which time and again has been identified as a separate institutional setting: the southern welfare state regime (see as excellent studies on the southern regime Bonoli, 1997; Ferrera, 1996, 2010; Leibfried, 1993; León and Guillén, 2011; Matsaganis et al. 2003; Rhodes, 1997; Trifiletti, 1999; see also, for an overview over social policy in Spain and Portugal, Huber and Stephens, 2012: Chapter 7)? The southern regime and political economy appears distinct, among others, because of the high standards of employment protection for a mainly male core work force, militantly defended by radical and fragmented unions. These standards translate into strongly ‘dualized’ labour markets with high youth unemployment and low female labour force participation. Strong dualization, in turn, in combination with an occupationalist ‘Bismarckian’ welfare state very much tailored to defend the interests of the (various) political clientele, lead to marked outsider under-protection and at times obscene forms of insider over-protection (‘unparalleled peaks of generosity’, Ferrera, 2010). The southern countries share conflictive industrial relations with almost no traits of corporatism (Siaroff, 1999), since the fragmented unions compete with and try to overbid each other, whereas on the government side it lacks a social-democratic party able to credibly offer unions a corporatist exchange to induce wage restraint. The political–economic equilibrium is consequently one in which a lack of wage coordination translates into high inflation and low international competitiveness. Governments responded – as long as they could, that is, before the Single European Act and European Monetary Union – with trade restrictions and repeated devaluations, firms with low investments. As a consequence, these southern countries did not follow a model of export-led growth (for this argument, see Eichengreen, 1996, 2007: especially pp. 90, 104 and 114–15).

Moreover, they host a substantial shadow economy and have an over-proportionate share of self-employed and free professions, since it is generally understood and was for a long time tolerated for this important clientele of the bourgeois parties to remain largely exempt from taxation. Related features of the southern model are the strongly ‘gendered’ labour markets, a ‘low fertility equilibrium’ (Esping-Andersen) with long dependence on the parental household – generally high importance of solidarity through family ties due to the very uneven coverage of public social protection and a low degree of state capacity (‘low state penetration in the welfare sphere’; see Trifiletti, 1999). It fits the picture that the ‘women unfriendliness’ of the southern welfare state is not only due to the fact that political competition over the female vote was for a long time religiously contorted (Emmenegger and Manow, 2014; Ignazi and Wellhofer, 2013; Morgan, 2013), but that women received the right to vote quite late (Siaroff, 1994: 96–8). The reluctance to let women vote originated not least in the political left suspecting women’s voting behaviour to be under the church’s ideological influence. Another – prima facie contradictory – feature of the southern regime, namely the ‘Catholic’ dependence on family solidarity in unison with the dominant position of the state in early child care and pre-school, can only be understood against the background of the fierce state–church conflict over education in the late 19th and early 20th centuries (Morgan, 2002, 2003, 2006). This shows that the simple reference to ‘familialism’ does not suffice to explain the distinctiveness of the Southern countries (see Esping-Andersen, 1999: 90).

Furthermore, the insurance schemes’ high level of occupational fragmentation in the South comes with a strong dose of workers’ self-administration (autogestion, democratie sociale), often as an institutional safeguard against the danger of the political enemy misusing welfare payments for clientelistic purposes (Lynch, 2009). Additionally, southern welfare states are generally characterized by an overemphasis of pensions over social assistance or unemployment benefits (‘old age welfare state’), combined, however – in contrast to the other conservative or Bismarckian welfare states – with national health-care systems (Ferrera, 1996). Generally, we find ‘extensive clientelism and patronage machines which distribute cash subsidies to political client groups’ (Rhodes, 1997: 6) and a poor quality in administering the welfare system. This fosters inefficiency within expensive systems plus a ‘poor capacity for reform’ (Rhodes, 1997: 16).

Maurizio Ferrera (2010) – among others – has highlighted the extent to which the typical imbalances of the southern model can be perceived as ‘the by-product of the specific pattern of political competition’ (p. 622). For instance, it was the internal division between a maximalist and a reformist left that rendered all postwar attempts to overcome occupational fragmentation in favour of universal social insurance schemes futile (Ferrera, 1996: 31), while universalism was implemented where no vested worker interests were affected – in health care. One important consequence of the split on the left was that the left was in government fewer times – in Italy due to the lack of a credible contender for power, the Democrazia Cristiana even became hegemonic until 1992. And even if in government, the southern left is usually barred from pursuing a reformist programme due to the competition from an orthodox contender (Hopkin, 2004). The rampant clientelism and patronage often mentioned as an additional and independent factor in the explanation of the southern model must in my view – at least partially – be understood as a consequence of the deep enmity between the political camps and of the division within the left. Sara Watson (2008), for instance, has convincingly shown how much the particularism of the Spanish unemployment insurance is a consequence of the Socialists’ strategy to hold down communist landworkers’ unions. Similarly, the weak state is partly explained by the deep mutual distrust between the polarized political camps, which hinders the state bureaucracy from becoming a neutral, impartial authority.

In order to better understand the ‘specific pattern of political competition’ behind many of the institutional peculiarities of the southern regime, I now turn, first, to the historical causes of the rift within the left between a reformist and a radical wing. I will then, second, summarize some of its consequences for parties’ vote share, government composition and the political space of south-European party systems.

Europe’s ‘culture war’ ² of the late 19th and early 20th centuries

The counter-reformation in southern Europe had successfully secured Catholicism’s religious monopoly – protected via a liaison between the Church and the forces of the Ancien Régime, that is, the crown and the ruling classes. In the 19th century, the liberal nation-state building elites therefore always attacked both crown and church: ‘coherent and massive secularism’ was pitted against ‘coherent and massive religiosity’ – this is what David Martin (1978) describes as the ‘Latin pattern’ (pp. 6, 36–41, 244–77 and passim). The Catholic church felt its existence threatened by the liberal state-building elites (Clark and Kaiser, 2009; Gould, 1999), and their legislative programme with respect to confessional schools, the Catholic orders, civil marriage, church property, religious festivals and so on did everything to let these fears appear well founded. The church reacted by rejecting modernity, liberalism and the secular nation-state (as in Syllabus Errorum, The Syllabus of Errors, 1864). It developed what is labelled intransigent Catholicism (Perreau-Saussine, 2012).

When industrialization with the rise of the workers’ movement and finally the Russian Revolution gradually changed the main political conflict lines, intransigent Catholicism directed its animosity chiefly and increasingly at the political left. In turn, the left developed an often aggressive anticlericalism, too. This conflict turned into ‘a spiral of fear and mutual repulsion backed by violence until each side feels its very existence endangered by the other … Once this occurs fear is transmuted into reality and the only practical tactic is war à l’outrance’ (Martin, 1978: 17). The church–state conflict did not develop with the same vehemence in countries where Catholicism was a minority religion – David Martin (1978) therefore distinguishes between the mixed and the Latin pattern. This distinction is relevant for our context, since both patterns differ with respect to the political positioning of the church, either non-conciliatory anti-republican and right or moderate and centrist. This translates into different degrees of conflict intensity between the left and the right in southern and continental Europe.

One direct result of the conflict between church and state, according to the standard line of argument, has been the institutionalization of Christian-democratic parties (Conway, 2004; Frey, 2009; Kaiser and Wohnout, 2004; Kalyvas, 1996; Kalyvas and Van Kersbergen, 2010; Van Hecke and Gerard, 2004). By now we know quite a bit about the particular role these parties have played in the development of the welfare state (Huber et al. 1993; Van Kersbergen, 1995; Van Kersbergen and Manow, 2009). Here I argue that there was another indirect consequence of the coalition of an anti-liberal, anti-modern church with the reactionary forces in countries in which it saw itself existentially challenged. Religion in these countries got an unambiguous political coding: it was decidedly right-wing (Berger, 1987). Accordingly, the political confrontations accompanying the mass-democratization of societies turned much more fundamental – the result being violent, civil-warlike conflicts during the first half of the 20th century in all mono-denominational Catholic countries under study here (Nolte, 1998; Traverso, 2007). An important heritage of this conflict à l’outrance between a clerical right and an anticlerical left was the radicalization of the left due to the lack of a plausible reformist option. As a consequence, postwar politics retained a polarized character. The rift on the left between socialists or Social Democrats, on the one hand, and communists, on the other, persists and does so primarily due to the ‘moral’, not to the ‘material economy’ of these countries.

This admittedly quite stylized account of the development fits Italy, which turned fascist as early as 1922, ³ and where the peasantry since 1860 was considered ‘the reserve army of clerical (papal-Bourbon royalist) reaction’ (Absalom, 2009: 128) and where much of the political development after 1861 is explained by the one fact of ‘utmost importance: the hostility of the Catholic church to the new Italian state, and the hold which it had on popular feeling’ (Lyttelton, 1987: 4). It matches the development in Spain, which after 1936 became the stage for a merciless civil war between opponents and supporters of the republic. Spain witnessed violent episodes before the turn of the century, experienced a dictatorship under de Rivera as of 1923 (Preston, 2006, 2012), and, once the civil war ended, was ruled by the brutal military dictatorship of Franco between 1939 and 1975. Support for the fascists was concentrated in Spain’s heartland, dominated by the Catholic smallholding masses, the Church sided clearly with the Falange, and the Catholic party, the CEDA (Confedaración Española de Derechas Autónomas) embarked upon a violently anti-republican rhetoric in the early 1930s, openly calling for insurrection. Portugal’s history exhibits a similar path towards ‘clerical authoritarianism’ (Tumbletey, 2009). Vichy represents the authoritarian, antidemocratic solution that the French political right, in coalition with the national Church hierarchy, had sought repeatedly during the interwar period and almost put in place in 1934 when France came close to a violent overthrow of the republic (Tumbletey, 2009). One therefore cannot be surprised about ‘the enthusiasm with which the overwhelming majority of French Catholics welcomed the establishment of the Vichy regime in 1940’ (Conway, 2004: 241). ⁴

The fundamental character of the political conflict reveals the explanatory limits of an argument based solely on socio-economic analysis. In these conflicts, religion becomes relevant, first, in the explanation of the totalitarian episodes of the southern countries – since religion renders coalitions between workers and peasants impossible and thereby fosters the fascist path. These totalitarian episodes are then, second, an important explanatory factor for the persistence of political polarization in the postwar period, inter alia manifested in the fragmentation of the left in their party systems. Yet, rarely has research taken this into further consideration. For example, the denominational dimension of conflict is almost completely missing in Luebbert’s (1991) ⁵ study of Europe in the interwar period as well as in Geoffrey Eley’s (2002) history of the European left. The same can be said of Sheri Berman’s (1998) study on interwar Social Democracy. Stefano Bartolini (2000) treats the religious cleavage as a contextual factor for the mobilization of the left, but not as a cause for the rift between reformist and revolutionary wings of the labour movement. In Esping-Andersen’s (1985) account of the ‘social democratic road to power’, this dimension is lacking as well, probably due to the fact that he develops his theory from an exclusive treatment of the Scandinavian cases – but one cannot grasp the specificity of the Nordic pattern by looking exclusively at the Nordic countries, where indeed the state/church cleavage had been largely absent. Esping-Andersen is in good company, though, sharing his selection bias with a large, heavily Nordic-tilted welfare state literature. But if one wants to understand what enabled northern farmers and workers to enter political alliances – a coalition responsible for the early build-up of the Nordic postwar welfare state (see especially Baldwin, 1990) – one needs to compare the Nordic with other cases, and a comparison with the economically, but not politically similar southern pattern seems particularly fruitful. I would like to briefly elaborate this point.

Coalition options in the interwar period

The southern – like the northern – countries were relatively late to industrialize; that is, they were still very much agrarian societies at the time of mass-democratization, that is, around 1920 (see Figure 1).

OPEN IN VIEWER

Figure 1.

Employment share in agriculture in West-Europe, 1910–1940.

For Europe’s South we might sketch the basic constellation of social forces in the interwar period as follows: a coalition between workers and smallholding (Catholic) farmers was unthinkable due to the former’s militant anticlericalism. Farmers rather tended to ally with the established, reactionary forces – the rural elite, the entrepreneurial class, the military, but also with the urban petit bourgeoisie – against the political left. This coalition became increasingly likely the more the Catholic church felt threatened by the liberal elite during the creation of the nation-state. In other words, the fiercer the conflict between church and state once was, the fiercer the conflict between the political left and the Catholic church became. This severe conflict then caused small farmers, the family peasantry, to recoil from a coalition with the labour movement and its doctrinaire Marxism – which treated the rural proprietor as a doomed class anyway (Eley, 2002; Judt, 1979). It was therefore in Italy’s and Spain’s North among the smallholders – not in the South where the latifondista/latifundista dominates – that fascism made its most successful inroads into the countryside (Bosworth, 2009; Corner, 1975; Ertman, 1998; Farneti, 1978; Luebbert, 1991; Lyttelton, 1987; Preston, 2006). In turn, the radicalization of the left became all the more probable when coalitions between workers and farmers became wholly unlikely. The resulting polarization eventually turned violent in almost all of the countries studied here, often completely unrelated to the onset of the Great Depression in the late 1920s, as in Italy in 1922 or in Spain in 1936.

The church’s massive anti-republican agitation cast substantial doubt on the political loyalty of the rural classes – which opened the void for the political violence of the extreme right. At the same time, given the verbal radicalism, revolutionary maximalism and aggressive anticlericalism of the left, the fascists appeared like promising enforcers of order, family, property and religion, and as guarantee of an anti-Bolshevik bulwark in a period of civil war-like conflicts. Whether or not the Catholic church and Catholicism actively participated in a conflict stylized to be the ‘historic battle of resistance to bolshevism’ (quoted in Nolte, 1965: 19), or acted as a critical structural factor rendering a coalition of workers and farmers highly unlikely, the religious dimension was critical for the basic coalition options during the interwar period. The importance of religion was quickly understood by the fascist movements: the initial anticlericalism of the urban fascists swiftly receded as the movements gain massive support in the countryside among the small property owners (Pollard, 2009). Later followed the Lateran treaties in Italy or the official recognition of the Franco regime by Pope Pius XII.

Admittedly, this is a very broad-brush depiction of the southern European model (including France). The history of each country should certainly be told in a far more nuanced manner. In doing so, the overlapping time frames of national developments and their mutual influence on one another would have to be taken into consideration – as the impact of the violently anticlerical Mexican revolution on the political position of the Vatican in the late 19th century, or as the impact of the Nazi’s advent to power on the strategy change of French communists who were now ordered to defend the republic together with the socialists against the fascist threat. One would also need to differentiate between various actors: national church hierarchies, local priests, Christian-democratic parties, ‘Rome’, and so on. Moreover, church and state relations prove to be far more complex and changing than could be described here. To the picture belongs the final distancing of the Vatican vis-a-vis the Catholic Action Française, or Rome’s (late) critique of the Franco regime as well as the volatile relationship between the Vatican and Mussolini (Kelikian, 2002; Webster, 1960: chapter 7). It also would be foolish to deny the important differences between, say, the French case in which a united left succeeded in holding the extreme right under control in the 1930s, and the Italian case which succumbed to fascism already in the early 1920s. Still, as one expert recently summarized,

It would be no exaggeration to say that Catholic support for fascism was a major consequence of the ‘culture wars’ between Catholicism and liberalism that had raged in Europe and in parts of Latin America, since the early nineteenth century, and that the Spanish Civil War of 1936–1939, in which Catholic saw the hand of anticlerical liberalism, Freemasonry, and bolshevism, was the last and greatest of Europe’s ‘culture wars’ and had, accordingly, a massive impact on the attitude of European Catholics to fascism. (Pollard, 2009: 176)

As a consequence, the pious rural classes in southern Europe could not bring themselves to enter into coalitions with the Marxist workers’ movement, which forestalled the reformist option in response to the social and economic turmoil of the interwar years. The religious barriers to such a coalition fostered the radicalization of an isolated left, and the questionable stance of the Catholic rural classes vis-à-vis the republic opened the political space for the violence of the extreme right.

In my view, the creation of powerful communist parties – and unions – needs to be recognized as an important heritage of this polarized constellation in ‘Latin Europe’. And the split on the left continued to shape the southern party systems after the Second World War (if these countries turned at all democratic), as did strong communist unions, which had an impact on their industrial relations (Ebbinghaus and Visser, 2000). It is the political violence and the fundamental character of political conflict before 1945 which explains the persistence of the polarized electoral milieus of the clerical right and the anticlerical left after the Second World War. These milieus remained for a long time basically unaltered by the profound socio-economic change over the postwar period, unimpressed by the slow rapprochement among the political elites of either side (in Italy culminating in the compromesso storico in the late 1970s) and also not substantially affected by secularization, that is, the weakening strength of religious sentiments and values.

How much the northern pattern differs from this southern one is relatively well known, not the least thanks to Esping-Andersen’s (1985) detailed reconstruction of how worker–farmer alliances formed the Scandinavian political economy. It is true that the Nordic labour movement was split between social-democratic and communist parties as well (see below). And like their counterparts in southern Europe, the latter were quite successful in the immediate post-WWII period and mobilized not only the core industrial areas but also in marginalized, precarious agricultural regions (Tarrow, 1967a, 1967b). But except for the case of Finland, in which the civil war and the subsequent forceful repression of communists is to be understood in connection with the Finnish war of independence, the radicalization of the labour movement in the Scandinavian countries is not the expression of a cultural conflict over the fundamental issue of affiliation with the nation-state and to basic values connected to faith, property and family. The lack of such a conflict – and not the labour movements’ existing or non-existing ‘instinctive antipathy to the countryside’ (Judt, 2006: 405) – changes the basic political coalition options for Social Democracy. Also, it was not Social Democracy’s simple ‘inability or unwillingness to reach out to farmers’ (Berman, 1998: 204) which explains why outside of Scandinavia workers and farmers proved unable to ally. Similarly, ‘whether or not socialist movements had become engaged in class conflicts within the countryside’ (Luebbert, 1991: 11) does not explain the formation or non-formation of such a class coalition. It is instead the absence or presence of the religious cleavage which explains why a coalition between farmers and workers, that would have made sense in economic terms and that was feasible in northern Europe, proved impossible in the south due to non-economic reasons. ⁶

For other variants of the Catholic pattern, we need to note that political Catholicism is moderate where Catholics are a minority, like in Germany, the Netherlands and Switzerland (Martin, 1978: 51, 57 and passim), and that Catholicism, where it took part in nation building in confrontation with an ‘alien’ and denominationally different country (as in Ireland vs Britain; Poland vs Germany and Russia; Belgium vs Netherlands, but also vs laïcist France), became a unifying factor, and subsequently either did not lead to the formation of a Christian Democratic party (Poland, Ireland) or rendered this party politically moderate (Belgium). Consequently, in these countries, religion did not turn into a fundamentally contested issue between left and right (compare Martin, 1978: 37, 42–5 and passim). ⁷

The next section briefly addresses the cleavage on the left as a significant source of systematic variation between West European party systems.

Party systems, party families and the ‘political space’ in southern Europe

We have ample empirical evidence for the uniqueness of southern Europe with respect to the vote share of the different party families in European comparison, the dimensionality of the party system and with respect to government composition.

In order to compare the electoral strengths of party families over the entire postwar period and to chart the development of the European party systems, I use data of the Comparative Manifesto Project (Budge et al. 2001; Volkens et al. 2012) and the ParlGov dataset (Döring and Manow, 2012). For the two dimensions of interest here, I employ the basic right–left position of party and a dimension that shows the significance of the church/state, specifically the pro-/anticlericalism cleavage. ⁸

Figure 2 shows that after initial electoral success in the immediate postwar period, the vote shares of Scandinavia’s communist parties flatten out very consistently around 10 percent. Contrary to this, the French and Italian communist parties, the Parti Communiste Français (PCF) and Partito Comunista Italiano (PCI), enjoy vote shares more than twice as high until the 1970s and remain strong during the trente glorieuses of the welfare state. Scandinavian and southern communist vote shares do not converge before the late 1990s. Divergence and convergence are due to two trends: sectoral change in the North had essentially resolved the ‘agrarian question’ by the 1960s, which weakened the ability of communist parties to mobilize voters in rural regions. In the South, however, the conflicts over confessional schools, divorce, abortion, contraception, civil marriage, and the like, repeatedly renewed and reinforced the cultural cleavage. Convergence was also due to the fact that the communist parties of Spain and Portugal after the transition to democracy were less strong than initially expected (see Linz, 1967). This was partly due to the fact that, in light of the long-term dictatorship and wary of the fragile transition process, voters and party elites preferred moderate positions (compare Hopkin, 1999). Consequently, we see that Juan Linz erred in predicting that in Spain, as in Italy, Christian Democrats and Communists would become the two strongest parties as soon as the country became democratic (compare Linz and Montero, 1999). It was also important that membership in the European Union (EU) now represented the reformist option which the labour movement had lacked for so long. In Portugal, for instance, the EU was the ‘alternative to the project of Portuguese socialism pushed by the Communist Party (PSP) during and after the revolution’ (Huber and Stephens, 2012: 219). Nonetheless, the split between socialists and communists continued to affect electoral competition and industrial relations, and thereby social and economic policies.

OPEN IN VIEWER

Figure 2.

Vote shares of communist parties in Scandinavia and southern Europe.

When we examine the position of the parties in the political sphere, it becomes evident that southern European party systems are particularly polarized in the left/right and in the clerical/anticlerical dimension. Communist parties with large vote shares occupy the anticlerical pole (see Figures 3–5). Electoral geography confirms for France and Italy that communists are often ‘heir to an anti-clerical tradition’ (Taylor and Johnson, 1979: 188, my emphasis).

OPEN IN VIEWER

Figure 3.

The party systems of continental Europe, 1946–2012, left/right and pro-/anticlerical.

OPEN IN VIEWER

Figure 4.

The party systems of southern Europe, 1946–2012, left/right and pro-/anticlerical.

OPEN IN VIEWER

Figure 5.

The party systems of northern Europe, 1946–2012, left/right and pro-/anticlerical.

Figures 3–5 depict the two-dimensional political space for the average continental, southern and northern party system, generated by averaging vote shares and positions over the 12 country cases.

While we have to be cautious to give this spatial representation of national party systems too much interpretation, it certainly improves upon previously identified ‘strategic configurations of parties’ (Kitschelt, 2000), obtained in a rather ad hoc manner. Important differences between the three types of party systems are the salience of the religious cleavage, the electoral strength of a left anti-system party and the presence or absence of agrarian parties.

A particular upshot of the southern party constellation is the dominance of centre-right governments in Italy and France, since the rift on the left increases the probability for Christian-democratic or conservative parties to form governments. This also holds true for Portugal and Spain after 1976/1977 (see Figure 6). On a left/right scale ranging from 0 (left) to 10 (right), Figure 6 shows the average government position for the western European countries over the postwar period. ⁹ A geographic variance between North and South is apparent and mirrors party system differences (Manow, 2009). Although left parties gain quite similar vote shares in the North and in the South (compare Bartolini, 2000: 64 and Table 2.1, p. 55 and passim), they do differ quite profoundly with respect to their years in government. And if in government, the southern left is usually barred from pursuing a reformist programme due to the competition from an orthodox left (Hopkin, 2004).

OPEN IN VIEWER

Figure 6.

The average position of postwar governments in western Europe on the left–right spectrum.

Four worlds of welfare capitalism

Esping-Andersen’s three-world typology has proven extremely helpful in analysing and comparing western welfare states. In fact, Esping-Andersen’s book revolutionized the entire field of comparative welfare state research and comparative political economy and laid the basic conceptual groundwork for a research programme that is still ongoing and that reaches far beyond studies of the welfare state. It is on this foundation that today the contours of a ‘unified theory’ in comparative political economy become visible – in particular when we pursue the obvious linkages between Esping-Andersen’s contribution and the ‘Varieties of Capitalism’ approach (see the article by Torben Iversen and David Soskice, 2015).

Here I have followed the author’s suggestion to understand the development of the western European welfare state or the European political economies more generally as a ‘history of political class-coalitions’. In my version of that history, electoral rules, social cleavage structures and the resulting party systems play a central role. I focused on the one welfare state type which Esping-Andersen initially had not recognized as a model of its own, but which has been identified time and again as a distinct regime: the southern European one.

At the centre of my argument stood the distinctive trait of the southern European party- and industrial relations-systems: strong communist parties and unions. I have explained their strength with the intensity of the one conflict that proved to be so decisive for the countries of continental Europe, the state/church conflict. In the mono-denominational Catholic countries this cleavage evolved with particular intensity in the liberal era of nation-state building. Later, in the era of mass-democratization, it turned into a conflict between an anti-republican and clerical right and the republican left. As I argued, its upshots were the decidedly reactionary role of the church and the radicalization on the left in the mono-denominational countries of Europe’s South. In line with Rokkan’s thesis that the conflict between labour and capital was a homogenizing factor for West Europe’s party systems, I argue that the radicalization of the left in the South is due to a political polarization which is rooted in the moral, not in the material economy of these countries. This, in my view, explains why the rift persisted only in the mono-confessional Catholic countries of southern Europe. Time and again, fundamental moral questions helped stabilizing an anticlerical milieu – issues like confessional schools, civil (or today same sex) marriage, abortion, divorce and so on regularly renewed the conflict line between these milieus. The communist electoral strongholds in the core industrial zones or in the rural periphery, however, were affected by sectoral change and a secular increase in wealth and welfare. This renders exclusively economic arguments implausible, leaving only the absence or presence of these profound moral controversies to explain the different developments of the Scandinavian and the southern European communisms in the postwar period. Figure 7 represents the argument about the interplay between electoral rules and cleavage structures.

OPEN IN VIEWER

Figure 7.

Electoral rules, social cleavages and welfares state regimes.

I could only briefly touch upon some of the important consequences that ensue whenever a communist party or a communist union is present in the party system or in the industrial relations of a country, respectively. But many distinctive characteristics of the southern European welfare state model are in a very obvious way linked to the distinct strategic configuration among parties in the South. Not only is the left much less likely to rule if split, if it rules it will also be much more constrained by the presence of a more orthodox contender for basically the same electoral clientele. A more detailed account of the nexus between party competition and government participation on the one side and the political economy equilibrium in the South on the other, however, will have to be provided at another occasion.