Cleavages

Peter Mair. Handbook of Party Politics. Editor: Richard S Katz & William Crotty. Sage Publications. 2006.

Almost 50 years ago, in what has since become one of the classic texts in political sociology, S.M. Lipset (1960: 220) observed that ‘in every modern democracy conflict among different groups is expressed through political parties which basically represent a “democratic translation of the class struggle.” … On a world scale, the principal generalization which can be made is that parties are primarily based on either the lower classes or the middle and upper classes.’ The notion of the class struggle being democratically translated into politics is compelling, and thanks to Lipset, it has since remained part of the terms of reference of the discipline. But of course this was not the only social struggle that was being translated. As Lipset went on to indicate in Political Man (1960: 221), and as he later extensively elaborated in his path-breaking work with Stein Rokkan on European political development (Lipset and Rokkan, 1967), religious, cultural and regional struggles were also translated into political divides, albeit less evenly and less frequently than the struggle between classes. Moreover, while class conflict proved the most pervasive of the various social conflicts carried through into the political realm, it was not universal, even among the established democracies, and it was translated with differing levels of meaning and intensity. It scarcely figured at the party political level in the Irish Republic, for example, where, despite sometimes pronounced class identities, the various attempts to politicize class opposition were usually drowned out by the overwhelming attention that was paid to nationalist issues (Mair, 1992). In the crucially formative election of 1918 in Ireland, the Labour Party had stood aside to allow the newly expanded electorate a clear run in expressing support for the nationalist movement, and thereafter the party had never proved capable of moving away from the margins of the system. Nor did it translate into the party political realm in the United States, despite various social biases in the distribution of partisan support between Democrats and Republicans. Working-class support may have proved crucial to building and maintaining the postwar Democratic coalition (Hout et al., 1999), but, in contrast to the majority of European states, it never led to the mobilization of a major socialist or social democratic party (Lipset and Marks, 2000). In countries such as the Netherlands and Switzerland, where class conflict was successfully translated into party political alternatives, it never succeeded in developing into the overriding polarity, and was always vulnerable to the challenge posed by the translation of other divides. In France and Italy, by contrast, class conflict not only proved one of the most dominant sources of political opposition, but also one of the most radicalized and intense, with communist parties quickly gaining the upper hand in the contest to represent the interests of the organized working class.

In practice, then, what might be seen as a fairly simple and straightforward process—the translation of social conflict into political and party alternatives—turns out to be quite fraught and complex. There is nothing automatic at work here, and while conflicts are sometimes translated into partisan divides, at other times and in other places they are not, or only partially so. Class divides are usually translated into politics, but, as we have seen, this is not always the case, and not always in the same way. Religious divides are also sometimes translated, but not in all circumstances, and not always with the same intensity. Gender divides, however important at the level of society, have scarcely been translated at all.

What accounts for this variation? To a large degree, as Sartori (1990) has emphasized in an assessment of Lipset and Rokkan’s approach to voter alignments, it depends on the actual translator, or, as he then put it, on ‘the persuader.’ In other words, the social conditions that are eventually translated into party politics should be seen as the necessary facilitating conditions, while the primary agency that is at work is the party—or other organization—that intervenes to politicize those conditions. ‘To put it bluntly.’ argues Sartori (1990: 169), ‘it is not the objective class (class conditions) that creates the party, but the party that creates the “subjective” class (class consciousness) … [W]henever parties reflect social classes, this signifies more about the party end than about the class end of the interaction.’ In part, then, it is supply that makes for demand. Of course the same is also true for other divisions. It was the churches and their affiliated organizations that helped to translate religious divisions into electoral alignments, for example, even though the political parties that grew to prominence during this mobilization process soon developed their own momentum and outpaced the intentions of their original founders. As Kalyvas (1996: 257) has concluded: ‘Confessional parties were not the historically predetermined and automatic reflection of preexisting identities and conflicts, nor were they the emanation of structural, economic, or political modernization. They were instead a contingent outcome of the struggle among various organizations facing a multitude of challenges under tight constraints.’ In short, the shift from society to politics is determined at least in part by the active intervention of political forces in the society.

Much of the writing on cleavages—whether they are seen as social, political, or cultural -has tended to neglect this dynamic perspective. In some cases, cleavages are treated as if they were the more or less natural outgrowth of social stratification. If there are divides in the society, it is these which are seen to explain the presence of parties and politics; and if these divides then change, such that old lines of stratification fade away, and new ones emerge to take their place, this inevitably leads to the eclipse of one set of parties and to the emergence and growth of others. Political change in this sense is to be explained by social change, and in this way we acquire what Sartori (1990) refers to as a deterministic ‘sociology of politics.’ In other and more recent approaches, cleavages are assumed to be about belief systems, with traditional social structural divides such as classs or religion being seen to erode, only to be replaced by something that is built almost exclusively on preferences, mindsets, or ‘values’ (see especially Flanagan, 1987; Inglehart, 1990; Kriesi, 1998), and that has few if any relevant social correlates. In the one approach, norms and beliefs are not seen to be important; in the other, social structure counts for little.

It is evident that neither of these two alternatives is wholly satisfactory. The notion of value-free religious mobilization, for example, is clearly a contradiction in terms. The notion of a value-free class conflict is also difficult to conceive. Even the most dyed-in-the-wool workers’ movements that mobilized in the early years of mass politics could hardly be seen in this way. Indeed, whether its demands were couched within a frame that emphasized the rights of workers and the need for social justice, or whether they were seen to prefigure the inauguration of a classless society, the politics of this particular social structural divide was inevitably held together by a strong sense of collective solidarity and by a firm commitment to a more or less shared ideology. On the other side, it is also difficult to think of a cleavage being built exclusively on values. Even if we were to regard the materialist-postmaterialist divide as a real cleavage, for example, we could hardly avoid recognizing that the values on the latter side of that divide have been most commonly espoused by younger, better-educated, and reasonably prosperous citizens: Green parties attract fewer votes in underclass ghettos.

In reality, both social structure and values play a role in all cleavages, even if in one case it is the values which carry the greater weight, and in another the social structure. This is also the conclusion that is reached by Knutsen and Scarbrough (1995: 519) in their authoritative review of both the evidence and the literature: ‘The structural basis of political conflict, rather than being eroded, appears quite resilient … At the same time, we should note that the impact of structural variables is less significant, and the independent impact of value orientations more significant, than is implied by the cleavage model.’ This also depends, of course, on how the different elements are measured, and with what degree of accuracy. Knutsen and Scarbrough (1995: 519) also note, for example, that ‘the significance of value orientations has grown over the period 1973-90’—but, lacking the necessary instruments, neither they nor any one else can measure the real weight of value orientations when mass politics was first mobilized at the beginning of the twentieth century, or even when it became consolidated in the 1950s and 1960s. The fact that it is only social structure that can be measured in earlier periods does not mean that it is only social structure which then mattered.

But however important social structure and values may be for our understanding of cleavages, there is also something extra involved, and that is organization. Divisions may exist within the society, and these, in turn, may be associated with particular values or identities, but this does not necessarily mean that they will all become politically relevant. This is the key point which Sartori (1990) underlines in his reference to the importance of translation and persuasion, and it is also the point which Schattschneider (1983: 69) makes when he speaks of some issues—and we might well add some identities or some values—being organized into politics while others are organized out. In other words, the shift from society to politics occurs when a particular social divide becomes associated with a particular set of values or identities, and when this is then brought into the political world, and made politically relevant, by means of an organized party or group. But although these three elements can easily be distinguished from one another at the analytic level, in practice they are heavily interdependent. It is the shared social experiences that allow for the emergence of a collective sense of identity and a common value system; and it is the effect of organizational intervention, or persuasion, that helps to consolidate that identity and make it relevant to politics. In some cases, a formal political organization is scarcely required, since the social group is already very cohesive, and is bound together by a network of other, non-political organizations; in other cases, the identity is scarcely expressed until prompted by a group of entrepreneurial political leaders.

In other words, and as was first outlined in an earlier analysis of the stabilization of European electorates (see Bartolini and Mair, 1990: 212-49; see also Bartolini, 2000: 15-24, and Gallagher et al., 2005: 264-72), cleavages have three distinct characteristics. In the first place, a cleavage involves a social division that distinguishes between groups of people on the basis of key social-structural characteristics such as status, religion, or ethnicity. A cleavage is therefore grounded in a distinct social reality. Second, there must be a clear sense of collective identity involved, in the sense that the groups on which the cleavage is grounded must be aware of their shared identity and interest as farmers, workers, Catholics, or whatever. Among women, for example, it was the long-term absence of such a collective identity that constituted one of the major obstacles to the successful political mobilization of a gender cleavage. Third, a cleavage must find organizational expression, whether through a political party, a trade union, a church, or some other body. Each of these elements is an essential part of a cleavage, and it is here that the approach developed by Bartolini and Mair (1990: 211-20) differs from much of the other traditional work in this field. In many treatments, for example, the notion of cleavage is qualified, such that reference is made to ‘political cleavages.’ to ‘social cleavages’ or to ‘value cleavages.’ and so on, in a way that suggests that the different components can be separated out from one another and used to define different types of cleavage. In fact, such efforts at disaggregation simply lead to conceptual confusion, for there is almost nothing in a so-called political cleavage, for example, that is different from a political conflict or divide, and hence nothing that demands the use of the term ‘cleavage.’ It is equally impossible to distinguish the notion of a so-called ‘social cleavage’ from the notion of ‘social stratification.’ and hence here too there is no real added value in the term ‘cleavage.’ By contrast, when the concept of cleavage is restricted to those phenomena in which social reality, identity and organization combine and interact with one another, we then bring it back to a consideration of those fundamental divides that have shaped the parties and the party systems of contemporary Europe, and that have been so ably theorized by Lipset and Rokkan (1967; see also Rokkan, 1970, 1999).

This also serves to emphasize one additional property of cleavages: they are deep structural divides that persist through time and through generations. They persist for a variety of reasons. In the first place, they persist because the interests that are involved remain relevant, and the groups that are involved retain their sense of collective identity. Second, they persist because alternative political identities are only likely to be mobilized when large numbers of new votes become incorporated into the political system, and this process came to an end with universal suffrage. Third, the rules of the game—the form of electoral system, the structure of the parliamentary system, the institutional set-up more generally—tend to favour the persistence of the parties that devised the rules in the first place, and hence also favour the persistence of the cleavages on which these parties have been built. Finally, they persist because they continue to be organized into politics by parties that seek to survive by controlling the terms of reference of political conflict and by narrowing down their electoral markets: as Schattschneider (1983: 66) once noted, ‘the definition of the alternatives is the supreme instrument of power.’

In this context it is also revealing to recall that in Lipset and Rokkan’s comparative analysis of four centuries of political developments in Europe they note the existence of just four core cleavages—the Church-State cleavage, the center-periphery cleavage, the primary-secondary economy cleavage, and the owner-worker cleavage. To be sure, this long period was also marked by innumerable other divides between political actors and parties; but it was not marked by innumerable cleavages. On the contrary, as Flora (1999: 7) notes in his evaluation of Rokkan’s general theory, these are ‘fundamental oppositions … which stand out from the multiplicity of [other] conflicts rooted in social structure.’ When talking about cleavages, therefore, we are talking about the elementary building blocks of modern democratic development, and there are plenty of different terms that can then be used for the many other more tangential, peripheral, or short-term divides that play a role in contemporary politics.

Even building blocks can decay however, and precisely because a cleavage is constituted by three components, it can come under stress for a variety of different reasons. In one case, social change can result in the gradual erosion of the social reality underpinning the cleavage, such as when modernization and urbanization ate away at the traditional social base of the Scandinavian agrarian parties, forcing them to become more catch-all centre parties in an effort to maintain their positions within their respective party systems. In another case, the cleavage can decay when the sense of collective identity begins to fragment, and when interests are no longer seen to be shared. In yet other cases, the cleavage can decay because the organizations which shape it develop an ambition for more expansive political strategies, and no longer foster a reliance on traditional heartlands. What rarely seems to happen in contemporary politics, however, is the wholesale substitution of a cleavage, such that one fading alignment is replaced by another emerging divide. At most, the evidence seems to point towards dealignment, in which weakening cleavages give way to non-structured electorates, and in which the powerful and stabilizing combination of social stratification, collective identity and organized expression yields to the emergence of more volatile individualized or particularized sets of preferences. Alternatively, where a new politics does look like it is gaining ground, as might be seen to be the case with the mobilization of Green parties, on the one hand, or right-wing populist parties, on the other, the effect seems often short-lived and the challenge seems capable of adaptation. Finally, it is also striking to note how new issues and new concerns can sometimes breathe life into what had been an otherwise dormant cleavage, as was the case, for example, when the European issue seemed to reinvigorate the once highly salient centre-periphery cleavage in Norway.

When cleavages do decay, this is usually manifest in increasing levels of aggregate electoral instability—voters become more inclined to shift between parties and to cross cleavage boundaries, and their behaviour may therefore become less predictable and more random. As has been argued elsewhere, however (Bartolini and Mair, 1990; Mair, 2001), electoral volatility as such does not necessarily indicate cleavage decline. Rather, it is important to note the type of volatility that is involved, and its location within the party system. Shifts of votes between friends, or between cleavage allies, are in this sense less important than shifts of votes between enemies, for while both may indicate a weakening of the hold of individual party organizations, it is shifts between enemies which are more likely to indicate a weakening of the hold of cleavages. Shifts in aggregate electoral support from social democratic parties to communist parties in the early years of the twentieth century, or from communist parties back to social democratic parties at the century’s close are clearly of relevance to our understanding of party systems and how they change, but they tell us little about the degree of cleavage closure. Similarly, shifts from one Protestant party to another in the Netherlands, or from one bourgeois party to another in France, may impact little on the overall cleavage structure, but they may well have a major impact on how the parties compete with one another. For this reason it is important to distinguish between overall levels of instability or volatility in the system, on the one hand, and levels of inter-area or inter-block volatility as given by the different cleavages structures, on the other (Mair, 1983: 408-14; Bartolini and Mair, 1990: 41-6). As more than a century of mass politics has shown, voters are much more willing to cross the boundaries separating individual political parties than they are to cross the lines of cleavage. Hence, while parties have come and gone, cleavages have tended to persist. This is also why cleavages are important.