There can be no question of offering a detailed electoral sociology of the social-democratic parties here. My aim is more modest and precise: to present some basic sociological data identifying the main features of social-democratic electoral development; and to furnish, as and where necessary, what I regard as essential elements for an initial explanation of the loosening of the link between the working-class and social-democratic parties, which was complex and ambivalent, but politically vital. To this end, I have chosen to study principally – but not exclusively – six national cases: the British Labour Party, the Danish SD, the Norwegian DNA, the German SPD, the Austrian SPÖ, and the Swedish SAP. The analysis will be rounded off by a study of the French left, which is assumed to be ‘atypical’ given the absence of a ‘classical’ social-democratic party.
None of these scenarios by itself represents the ‘truth’ of the social-electoral mutation of the left in Europe, but each, in its fashion, illuminates an important dimension of that mutation. These cases are, as it were, privileged observation points from which we can ‘catch in full flight’ the characteristics of the relationship between the working class and voting for the left. In Great Britain and Denmark, where the examples are (or are considered) extreme, the upheaval initially assumed the dimensions of a sociological earthquake. In Norway the decline in the working-class vote was gradual, smooth and silent, like a sociological mechanism that starts up gently and produces its effects imperceptibly but inexorably. In Sweden and Austria, two models of working-class social democracy, the social-democratic vote among workers also declined, but the contrasts between these two countries underline the impact of political factors on the evolution of class voting. In Germany the trend was rather towards stability, despite slight disturbances, thus contributing to the ‘steady crisis’ of the SPD. In France, fluctuation in the working-class vote was at the centre of political upheavals, making this country a model case where, more conspicuously than elsewhere, the fluctuation in class voting is conditioned by changes and variations in the interplay of partisan competition.
Social Democracy and the Working-Class Electorate: The Weakening of a Vital Link
Great Britain
(i)The Trends
Analysis of Labour’s penetration of manual strata makes it possible to identify an evolution in four phases: an expansive phase, a downturn in two successive stages, and renewed expansion.
(a)During the first period, 1959–70, the level of Labour’s penetration of manual strata reaches a peak (62 per cent in 1959, 64 per cent in 1964, 69 per cent in 1966).
(b)Retrospectively, in spite of Labour’s two electoral victories in 1974, the period 1970–79 may be considered as the first stage of a major uncoupling. In the context of a general electoral low-water mark, Labour’s penetration of manual workers falls back below the 60 per cent threshold and settles at a level markedly inferior to that of the previous decade (58 per cent in 1970, 57 per cent in 1974, 50 per cent in 1979). Compared with 1959, the party’s constituency among non-manual workers remains fairly stable (22 per cent in 1959, 22 per cent in 1964, 25 per cent in 1970, 23 per cent in 1979). Relative to the 1960s, only Labour’s penetration of manual strata undergoes marked decline, its support among non-manual workers remaining virtually intact (Table 7.1).
(c)In the third phase (the 1980s) the party crosses a new quantitative threshold in its downturn, despite a slight recovery in 1987. Now all social groups turn away from Labour, and manual workers rather more than the others (index of evolution of Labour’s penetration: 65.41 per cent among manual workers; 77.08 per cent among non-manual strata).
(d)At the 1992 election, after three consecutive defeats, the Labour Party lost once again. Its electoral recovery was insufficient to repair the electoral damage caused by its last period in government (1974–79) and its political options in the 1980s. It did, however, improve its influence appreciably among both non-manual and manual strata (Table 7.1). But despite this improvement, working-class and popular support for British Labourism (measured by its penetration of manual workers) was markedly inferior to its influence in the 1960s, and even the first half of the 1970s (index of evolution 1992/1960s: 76.69 per cent). By contrast, Labour’s influence among non-manual workers, despite the preceding fluctuations, was at exactly the same level in 1992 as the average for the 1960s.
Table 7.1 British Labour Party penetration according to the manual/non-manual cleavage (1964–97)
| M | NM | M | NM | ||
| 1964 | 64 | 22 | 1979 | 50 | 23 |
| 1966 | 69 | 26 | 1983 | 42 | 17 |
| 1970 | 58 | 25 | 1987 | 45 | 20 |
| 1974 (Feb.) | 57 | 22 | 1992 | 51 | 24 |
| 1974 (Oct.) | 57 | 25 | 1997 | 58 | 40 |
Source: See David Sanders, ‘Voting and the Electorate’, in Patrick Dunleavy et al., Developments in British Politics 5, Macmillan, London 1997, pp. 55–7.
The 1997 election was a veritable landslide for ‘New Labour’. The party of Tony Blair advanced on all fronts and terrains, regardless of region, type of constituency, or social class. But it made its most spectacular breakthrough in the Cl category (skilled non-manual workers), where it obtained 47 per cent of the vote (+19 points compared with 1992). This result disrupted the ‘classical norms’ of British electoral behaviour inasmuch as traditional Conservative supremacy within this category made Britain a partially ‘atypical’ case among European countries. Similarly, Labour once again became the unquestionable leader (54 per cent) within the C2 category (especially among skilled manual workers, according to the statistical categories used by opinion pollsters) – a category that had swung, according to the dominant approach among British psephologists, in favour of the Conservatives during the previous twenty years (Labour’s advance relative to 1992: +15 points).1 In consolidating and reinforcing its traditional pre-eminence among semi-skilled and unskilled manual workers (61 per cent, or +9 compared with 1992), Labour re-established itself as the unchallenged party of the working class following the painful wilderness years of the 1980s. We should nevertheless emphasize that Labour’s penetration of manual workers remained below the level achieved in the 1960s, despite Blair’s exceptional electoral performance in 1997. Compared with the general election of 1964 (an election precisely comparable to that of 1997 in terms of the national vote obtained by Labour: 44.1 per cent in 1964 as against 43.3 per cent in 1997), Blair’s party obtained 58 per cent of the manual vote (against 64 per cent in 1964) and 40 per cent of the non-manual vote (against 22 per cent in 1964)! So, if the gap in Labour voting between manual and non-manual categories was reduced from 42 per cent in 1964 to the still significant figure of 27 per cent in 1992, it was only 18 per cent in 1997.
Table 7.2 Index of evolution of the Labour vote according to the manual/non-manual cleavage (1964–97)
| (averages by decade; base 100 for the 1960s average) | ||
| Manual | Non-manual | |
| 1970s/1960s | 83.45 | 98.95 |
| 1980s/1960s | 65.41 | 77.08 |
| 1990s/1960s | 81.95 | 133.33 |
Sources: See Table 7.1; my calculation.
(ii)Discussion
According to the dominant approach in British electoral sociology, changes in electoral behaviour in Great Britain since the 1970s have made it one of the best examples of class dealignment. The authors of How Britain Votes,2 however, have challenged this interpretation. Undertaking a redefinition of the schedule of socio-professional categories, which is doubtless closer to the real world of social classes, they have argued that relative class voting is not in decline. What is involved is not a decline in class voting, but ‘trendless fluctuations’, with circumstances and electoral appeal mainly explaining working-class behaviour. Given the importance of the issue, and the fact that it has a direct bearing on the nature of the argument advanced throughout this chapter, we shall take up certain aspects of the history – this other history – of class voting in Great Britain.
Two different but complementary impressions emerge from scrutiny of Table 7.3. The impression for the period 1970–79 – and hence prior to the political trauma of 1983 – is too clear to be open to ambiguity: Labour fell back among workers and foremen/technicians, considered by Heath et al. to be close to the working class, and solely among them. It was precisely in the social zone traditionally identified with Labour that the great electoral migration commenced. By contrast, Labour held its position or advanced, albeit marginally, among all other classes or social groups (see the indices of evolution in Table 7.4).
Table 7.3 Labour’s electoral penetration according to social class (1964–92)
| Salariat | Routine non-manual | Petty bourgeoisie | Foremen/technicians | Working class | |
| 1964 | 19 | 26 | 14 | 46 | 68 |
| 1966 | 25 | 41 | 19 | 61 | 71 |
| 1970 | 29 | 41 | 19 | 56 | 61 |
| 1974 (Feb.) | 22 | 29 | 19 | 40 | 60 |
| 1974 (Oct.) | 23 | 32 | 13 | 52 | 64 |
| 1979 | 22 | 32 | 13 | 43 | 55 |
| 1983 | 13 | 20 | 12 | 28 | 49 |
| 1987 | 15 | 26 | 16 | 36 | 48 |
| 1992 | 20 | 30 | 17 | 45 | 56 |
Sources: 1964–87: Anthony Heath et al., Understanding Political Change, Pergamon Press, London 1991, pp. 68–9; 1992: David Denver, Elections and Voting Behaviour in Britain, Harvester Wheatsheaf, London and New York 1994, p. 73.
Table 7.4 Social classes and index of evolution of the Labour vote (1964–92)
| (averages by decade, base 100: average for the 1960s) | |||||
| Salariat | Routine non-manual | Petty bourgeoisie | Foremen/technicians | Working class | |
| 1970s/1960s | 109.09 | 100 | 96.96 | 89.25 | 86.33 |
| 1980s/1960s | 63.63 | 68.65 | 84.84 | 59.81 | 69.78 |
| 1992/1960s | 90.90 | 89.55 | 103.03 | 84.11 | 80.57 |
Sources: See Table 7.3. Indices of evolution calculated by the author.
In contrast to this first impression, the 1980s reveal a Labour Party suffering general electoral discredit. In the context of an unprecedented drop, Labour lost part of its electorate among all groups in the population. The losses among foremen/technicians – ‘a kind of blue-collar elite’, according to Heath et al.3 – were the most important on average, but did not reach levels unknown among other social groups (see Table 7.4).
Basically, the Labour coalition in its 1960s form was ‘shattered’ – initially by the considerable erosion in working-class support (the main moments of decline being 1970 and, in part, 1979, which ‘capped’ undistinguished governmental records); and then by a generalized decline of British Labourism in nearly all sectors of society. In addition, the 1992 result indicates that, compared with the 1960s, Labour’s working-class influence was fixed at a level proportionately below its influence in other social groups, when likewise compared with that of the 1960s. This difference in workers’ behaviour, however, is not as important as conventional approaches have maintained (indices of evolution of Labour’s penetration 1992/1960s: working class: 80.57 per cent, foremen/technicians: 84.11, routine non-manual: 89.55, salariat: 90.90, petty bourgeoisie: 103.03 – see Table 7.4). The contrast becomes sharper, however, if we compare the election of 1992 with that of 1964: Labour strengthened its penetration of all non-manual groups as well as the petty bourgeoisie; it remained stable among foremen/technicians; it fell back only among the working class.
Various conclusions follow from this analysis:
(a)Since the beginning of the 1970s, a significant section of the broad category of manual workers – or a less extensive but far from negligible part of the working class in the strict sense – has deserted the Labour Party. It is common sense to observe that as regards the electoral behaviour of workers, what is involved is class dealignment, at least ‘in the absolute, as opposed to the relative, sense’.4 Regardless of all other considerations, this decline in absolute class voting among workers is in itself significant: it affects Labour’s class basis; it weakens its positions and place among its ‘natural’ electorate. With its influence in the working class reduced by at least 20 per cent compared with the 1960s, Labour in the 1980s is not the Labour of the 1960s, even if we accept that its relative class anchorage changed only marginally. An ‘unsuccessful class party’ (Heath et al.) is different from a successful class party, precisely in terms of its class anchorage and class identity.
(b)The general and non-class-specific regression of Labour in 1983, partly confirmed in 1987, was added to an already deteriorating situation whose principal and primordial element – as well as its common denominator – was working-class disaffection. This disaffection is not – and cannot be considered – a merely ‘static’ statistical datum that could just as simply be compensated or cancelled by other, subsequent statistical data. It is not, as Patrick Dunleavy would say, ‘context-free’.5 The dynamic process that the class dealignment of the 1970s generated, or in which it was involved, partially prepared the ground for the subsequent general decline of British Labour among all social groups. In fact, the initial erosion of the links between Labour and manual strata in the 1970s, which contributed to calling the party’s governmental vocation into question, subsequently (in the 1980s) favoured the defection of marginal non-manual voters from Labour, as well as the electoral breakthrough of the Liberal/SDP Alliance, creating ‘an across-the-board reduction in support’.6 A sociological and electoral dynamic was set in train by the phenomenon of the class dealignment of the 1970s, whose effects, although difficult to measure, should not, on that account, be ignored.
(c)Nevertheless, when all is said and done, it remains the case that throughout the period 1964–97 – and even during the 1980s, the dark years of British Labourism – Labour’s overrepresentation within the working class, as defined in the narrow sense of Heath et al., was a constant that was never seriously challenged. Compared with its penetration among the whole electorate, Labour’s penetration of the working class – what we might call Labour’s relative working-class penetration – underwent fluctuations, but was not – over the long term – in decline (Table 7.5). This overrepresentation constitutes a weighty argument for defenders of the thesis of ‘trendless fluctuations’.
(d)New Labour, such as it emerged from the 1997 election, is a party close to the catch-all model, having proved capable of enhancing its influence in novel social territory, while re-establishing itself – and this time in incontestable fashion – as the ‘natural’ party of the working class. Class dealignment through Labour’s advance among non-manual strata, particularly evident from 1959 to 1974 and arrested thereafter, becomes spectacular in 1997 (especially among skilled non-manual voters). This influx of non-manual voters into Labour’s electoral ranks is explained by the corrosive effect of government on the Conservatives and, obviously, by Blair’s modernization strategy. Basically, Blair was able to create a ‘new interpretive order’ concerning not only the identity of Labour (and the perception of it by the public and ‘centres of authority’), but also the whole electoral game in the United Kingdom. Here, the strategy of ‘natural party of government’ – to borrow a term from Ivor Crewe7 – largely explains the interclassist character of the new Labourist coalition of 1997. And – should such a thing be necessary – it demonstrates the role of political factors in the establishment of new class alignments or the consolidation of ‘old’ ones.
(e)In Great Britain, the classical class divide has lost some of its former strength, as is clearly indicated by the increase in cross-class voting, the reinforcement of centrist formations, and greater fluctuation in voting behaviour. However, the factor of ‘social class’ has not lost its preeminence in the British political system. More specifically, in terms of the Labour Party’s ability to represent the working class electorally, analysis of the trends has demonstrated on the one hand that it varies considerably according to political conjunctures, and on the other that it has partially but significantly diminished (something that is indicated by the marked decline in absolute class voting among workers). Labour’s overrepresentation among the working class, as well as the ‘ecological paradox’ of class voting in Great Britain (the so-called ‘area class effect’), unquestionably mean that even during the 1980s, Labour remained the ‘natural’ representative of the working class. In truth, Labour found itself in a paradoxical position. It was a ‘natural’ party of the working class that had in part lost – at least conjuncturally and temporarily (in the 1980s) – the principal attribute attaching to its nature: the capacity to represent the great majority of this class electorally.
Table 7.5 Specificity of Labour’s working-class vote relative to the Labour vote of the total electorate (1964–92)
| Year | 1964 | 1966 | 1970 | 1974(F) | 1974(O) | 1979 | 1983 | 1987 | 1992 |
| Index of specificity | 154 | 148 | 142 | 161 | 163 | 149 | 178 | 152 | 163 |
Source: Labour’s electoral results in Yves Meny, Politique comparée, Montchrestien, Paris 1988, p. 207. For Labour’s working-class vote, see Table 7.3. Indices of originality calculated by the author.
To summarize, observation of the whole period 1964–1997, which is not homogeneous, reveals that Labour’s positions within the working class have been weakened. The party became considerably more vulnerable within its ‘natural’ electorate than previously, while ultimately proving victorious (as 1997 clearly demonstrated) among the most peripheral sections of the ‘natural’ electorate of its Conservative opponent (particularly the lowest group of non-manual workers, those designated Cl).8 Are we dealing with ‘class dealignment’ or ‘trendless fluctuations’? The question is complex, and in some respects too technical to be dealt with here.9 Even so, it is hard to deny that since the beginning of the 1970s the working class, sociologically and organizationally the most structured – and structuring – element in Labour’s electorate, has proved to be politically more fragmented, ideologically more undecided, and electorally more unstable than in the past.
Denmark
Working-class disaffection with the Danish social democrats occurred in two distinct phases.
At the beginning of the 1960s, a section of manual workers turned in fairly enduring fashion to parties to the left of the SD, and in the main to the newly formed Socialist People’s Party (the social democrats’ working-class penetration registered an appreciable reduction, falling from 64 per cent in 1964 to 54 per cent in 1971).
At the beginning of the 1970s, and particularly after the electoral ‘disaster’ of 1973 (37.3 per cent of votes cast in 1971, 25.6 per cent in 1973), another fraction moved to the ‘bourgeois’ parties – in particular, the Progress Party (right-wing populist in allegiance) in 1973 and 1975, and the conservative party in 1984 (social-democratic working-class penetration fell from 54 per cent in 1971 to 41 per cent in 1987). Although it can be regarded as a continuation of the sequence that began in the 1960s, this second phase exhibited a new qualitative threshold on the path of decline, inasmuch as it expressed a certain class dealignment. (In 1984, the influence of the bourgeois parties among manual workers was 36 per cent, as opposed to 22 per cent in 1964; the index of evolution for 1964–84 [1964 base = 100] was 164!).
Certainly, in the 1990s the social democrats, profiting from a long spell in opposition (from 1982 to 1993 the SD experienced a continuous period in opposition for the first time since 1929), recovered some of their working-class losses (56 per cent in 1990, 45 per cent in 1994). But this recovery, which was very relative, was not of such a nature as to challenge the fundamental fact of the 1970s and 1980s: the weakening of their popular entrenchment. For the whole period 1964–94, one of general electoral downturn for the social democrats, losses within the different socio-demographic categories were not evenly distributed. Analysed by decade, the erosion of the SD’s electoral penetration compared with the 1960s is systematically more pronounced among manual than among white-collar workers (Table 7.7). In the 1990s the SD even improved its influence, albeit marginally, among these workers (indices of evolution 1990s/1960s: 86.08 per cent for blue-collar workers, 107.39 for white-collar workers). Thus the differential of the blue-collar vote for the SD, compared with the white-collar vote, went from +31 points in 1964 to +15 points in 1994. In addition, the socialist bloc as a whole (SD + SFP) appears very well entrenched among non-manual salaried strata, essentially because of the attraction exercised over this category by the SFP.10 Thus, if the socialist bloc is much more resilient among non-manual wage-earners than among workers, this serves to confirm the picture of a class dealignment in the Danish political system.
Table 7.6 Electoral penetration of the Danish social democrats according to social class (1964–94)
| Year | 64 | 66 | 68 | 71 | 73 | 75 | 77 | 79 | 81 | 84 | 87 | 88 | 90 | 94 |
| Blue-collar | 64 | 59 | 53 | 54 | 41 | 46 | 52 | 52 | 46 | 47 | 41 | 41 | 56 | 45 |
| White-collar | 33 | 30 | 25 | 25 | 18 | 24 | 34 | 36 | 27 | 24 | 23 | 26 | 33 | 30 |
| Self-employed | 13 | 9 | 6 | 7 | 5 | 6 | 9 | 13 | 7 | 6 | 8 | 11 | 8 | 7 |
Source: See Lars Bille, ‘The Danish Social-Democratic Party’, in Robert Ladrech and Philippe Marlière, eds, Social-Democratic Parties in the European Union, Macmillan, London 1999, p. 49.
Table 7.7 Social classes and index of evolution of the social-democratic vote in Denmark (1964–94)
| (Averages by decade, base 100: average for the 1960s) | |||
| Blue-collar | White-collar | Self-employed | |
| 1970s/1960s | 83.53 | 93.41 | 85.74 |
| 1980s/1960s | 74.58 | 85.23 | 85.74 |
| 1990s/1960s | 86.08 | 107.39 | 80.38 |
Sources: See Table 7.6. Indices of evolution calculated by the author.
The central position of the SD since 1924, making it the focal point of the Danish party system, was due largely to the unfailing loyalty of the working-class electorate. This is certainly no longer the case today. The slight improvement in the working-class penetration of Danish social democracy during the 1990s, which accompanied the strengthening of its overall electoral influence, remains markedly inferior to the level of the 1950s, and even the 1960s.
Norway
For a long time, and well before many European countries, Norwegian society was politically structured by the class cleavage. For several decades after the Second World War, the political domination of the Norwegian Labour Party (DNA) was based largely on the stable support of the great majority of the working class, which approached or surpassed two-thirds of the working-class vote. Indeed, during the trente glorieuses the penetration of ‘bourgeois’ political forces among the working-class segment of the population remained consistently below the 30 per cent level. By contrast, until the end of the 1960s the vote of workers for the ‘socialist bloc’ (DNA + left-wing socialists) was invariably massive: 77 per cent in 1957, 75 per cent in 1965, 74 per cent in 1969. From the beginning of the 1970s, however, the working-class vote for the left fell off, the trend being characterized by a gradual and almost continuous decline (70 per cent in 1973, 68 per cent in 1977, 61 per cent in 1981, 63 per cent in 1985, 58 per cent in 1989). For the whole period 1957 to 1989, the ‘socialist bloc’ lost 19 points among the working class,11 whereas it dropped only 7.3 points among the electorate as a whole (going from 51.7 per cent to 44.4 per cent). Symmetrically, working-class penetration of the ‘bourgeois bloc’ improved appreciably, going from 23 per cent in 1957 to 42 per cent in 198912 in a tangible sign of class political dealignment, at least as far as the electoral behaviour of the working class is concerned.
Table 7.8 Electoral penetration of the ‘socialist bloc’ in Norway according to social class (1965–89)
| Year | 1965 | 1969 | 1973 | 1977 | 1981 | 1985 | 1989 |
| Workers | 75 | 74 | 70 | 68 | 61 | 63 | 58 |
| New middle class | 40 | 40 | 42 | 38 | 35 | 38 | 44 |
| Old middle class | 29 | 33 | 28 | 24 | 24 | 18 | 30 |
Source: Olga Listhaug, ‘The Decline of Class Voting’, in Kaare Strom and Lars Svasand, eds, Challenges to Political Parties, University of Michigan Press, Michigan 1997, p. 81.
This picture is confirmed by the data in Table 7.9. First of all, taking the average for the 1965 and 1969 elections as a base, we should note that the Norwegian left, traditionally stronger among the old middle class than the left in other countries, declined sharply among this segment of the population. Next, we note that the influence of left-wing forces in Norway declined markedly – in stages – among the working class, while it remained fairly stable among the new middle strata. Thus, the differential between the socialist vote of workers and the socialist vote of the new middle strata went from +35 points in 1965 to only +14 points in 1989, a development that can be explained more by working-class disaffection than by the rallying of the new middle strata. This fraying of the bond between the working class and left-wing parties – occurring, as it did, in a slow, uninterrupted fashion – differs from the model of strong fluctuations in the working-class vote observed in countries like Denmark, Great Britain or, in another context, France.
Table 7.9 Social classes and index of evolution of the socialist vote in Norwav (1965–89)
| (Averages by decade, base 100: average for the 1960s) | |||
| Workers | New middle class | Old middle class | |
| 1970s/1960s | 92.61 | 100 | 83.87 |
| 1980s/1960s | 81.43 | 97.5 | 77.41 |
Sources: See Table 7.8. Indices of evolution calculated by the author.
Finally – and somewhat against the trend of what occurred in the great majority of countries with a powerful social democracy – the SPD was able to maintain its overall position within the working class, despite a minor tendential decline. In the late 1960s, German social democrats were certainly distinguished within the universe of European social democracy by the weakness of their working-class penetration (a weakness bound up in part with the importance of the religious divide and in part with the division of the country), and their relatively important influence among salaried middle strata. It is above all to the latter that the SPD owes its strong electoral growth and its constitution as a ‘natural’ party of government, particularly since 1969.
Roughly speaking, in the period 1969 to 1987 the German social democrats proved able to preserve their traditional electoral base (working-class penetration of the SPD: 58 per cent in 1969, 60 per cent in 1976, 62 per cent in 1980, 55 per cent in 1983, 59 per cent in 1987). In the 1990 election, the first to be held after reunification, in the context of a general decline in social-democratic influence (33.5 per cent, or the SPD’s worst electoral result since 1957), we observe a very important contraction in the SPD’s working-class penetration.13 In fact, the reduction of its influence among the working class (45.8 per cent in the western half of the country, as against 59 per cent in 1987) was considerably greater than among the electorate as whole (the SPD’s electoral decline in the territory of the former Federal Republic was only 1.1 per cent). It was linked to the inability of the SPD, and its leader Oskar Lafontaine, to live up to the exceptional historical moment of reunification, which was controlled with a firm hand by Chancellor Kohl. Indeed, this poor election result is principally explained by the centrality of ‘national’ issues and the relegation of socioeconomic questions to the background. In the 1994 election, however, the SPD made its biggest breakthrough in the traditional industrial constituencies of the western part of the country.14 It thus proved capable of re-establishing its working-class influence in the main (the social democrats obtained 55.9 per cent of the working-class vote – a good performance given that a new left-wing party from the east, the PDS, obtained 2.2 per cent of the working-class vote in western Germany).
By contrast, it is among the new middle classes – a rather volatile social sector in Germany – as well as the young that the SPD has lost most ground since 1983. The appearance of a new challenger – the Greens – in the 1980s, positioning themselves on the left of the political spectrum, is not irrelevant in the disaffection of a segment of the new middle strata and the younger generations with the SPD (Table 7.10). The threat presented by this – albeit peripheral – political actor destabilized German social democracy. Indeed, in its specificity the German case underscores the formidable challenge to parties of a social-democratic variety represented by the strong emergence of the materialism/post-materialism cleavage.
Table 7.10 Electoral penetration of the SPD according to social class (1961–94)
| Year | 1961 | 1965 | 1969 | 1972 | 1976 | 1980 | 1983 | 1987 | 1990* | 1994* |
| Workers | 56 | 54 | 58 | 66 | 60 | 62 | 55 | 59 | 45.8 | 55.9 |
| Salaried Employees | 30 | 34 | 46 | 50 | 43 | 50 | 43 | 37 | 33.1 | 39.4 |
* Western Germany
Sources: 1961–72: see Ilias Katsoulis, ‘La nouvelle vision “de classe” de la social-démocratie’, in Helga Grebing, L’histoire du mouvement ouvrier allemand, Papazissi, Athens 1982, p. 413; 1976–87: see Ursula Feist and Hubert Kriegel, ‘Alte und neue Scheidelinien des politisches Verhaltens’, Aus Politik und Zeitgeschichte, B12/87, 21 March 1987, p. 38; 1990: see Russel Dalton and Wilhelm Bürklin, ‘The German Party System and the Future’ in Russell Dalton, ed., The New Germany Votes, Berg, Providence and Oxford 1993, p. 244; 1994: see Russell Dalton and Wilhelm Bürklin, ‘The Two German Electorates’, in Russel Dalton, ed., Germans Divided, Berg, Oxford 1996, p. 187.
Nevertheless, despite the relative preservation of the SPD’s working-class bases and its stagnation (since 1983) among the new salaried strata, the dynamic governing the behaviour of the two main components of the SPD electorate in the long term does not differ fundamentally from that in other European countries. The SPD is more resilient among the new salaried strata than among the working class (Table 7.11). In the 1998 election, when it returned to power, the SPD only slightly improved its vote among workers (between 1 and 4 per cent for the whole of Germany, according to the different polling organizations), and among the unemployed (44 per cent against 42 per cent in 1994), despite the spectacular losses suffered by the CDU among the working class in the east (leading to the end of the ‘inverted social profile’ of class voting there). The SPD advanced more noticeably among the middle classes (particularly public servants, or Beamten) and the self-employed, groups traditionally close to the CDU, where it secured 4 per cent more than in 1994 (respectively 37 per cent and 22 per cent, according to Der Spiegel’s statistics). In addition, it achieved a rather disappointing vote among farmers; at 15 per cent, it progressed only marginally by comparison with 1994.15
Table 7.11 Social classes and indices of evolution of the social-democratic vote in Germany (1961–94)
| (Averages by decade, base 100: average for the 1960s) | ||
| Workers | Salaried Employees | |
| 1970s/1960s | 112.5 | 126.84 |
| 1980s/1960s | 104.7 | 118.20 |
| 1990s/1960s | 90.80 | 98.88 |
Sources: See Table 7.10. Indices of evolution calculated by the author.
Sweden and Austria
If the ‘spell’ between workers and Swedish social democrats has been broken, the drop in their working-class penetration remains fairly modest. The proportion of workers voting SAP fell from 74 per cent in 1956 to 66 per cent in 1976 and 63 per cent in 1994 (Table 7.12). The consequent weakening of the Swedish social-democrat party – party of the working class par excellence – is considerably less marked than elsewhere, but no less real for all that – and all the more so since the SAP has serious difficulties attracting new generations, a trend that is observable in all Scandinavian countries.16 It is important to stress that contrary to the Austrian pattern, most voters who abandon the SAP switch to the Left Party, restricting the sociological groundswell of class dealignment in Sweden, at least as far as the working-class vote is concerned. Great fluctuation in the electoral performance of the SAP – which is likewise struck by the ‘electoral instability syndrome’17 – is certainly accompanied by increased volatility in working-class electoral behaviour (as indicated by the SAP’s working-class vote in 1991, when an important percentage of workers went for non-socialist parties).18 More specifically, in the period 1979 to 1994 the SAP’s decline among workers – particularly private-sector workers – is clearly visible. In the same period, however, it gained ground strongly among white-collar workers, especially in the public sector (Table 7.13).
Table 7.12 Working-class penetration of the SAP (1956–94)
| Year | 1956 | 1960 | 1964 | 1968 | 1970 | 1973 | 1976 | 1979 | 1985 | 1994 |
| Blue-collar | 74 | 77 | 75 | 77 | 70 | 69 | 68 | 66 | 64 | 63 |
Sources: 1956–79: see Gøsta Esping-Andersen, Politics against Markets, Princeton, University Press, Princeton, NJ 1985, p. 126; 1985–94: in Christoph Kunkel and Jonas Pontusson, ‘Corporatism versus Social Democracy’, West European Politics, vol. 21, no. 2, 1998, p. 17; 1991.
Table 7.13 Electoral penetration of the SAP and SPÖ according to social class (1979–95)
| SAP | SPÖ | ||||||
| Year | 1979 | 1985 | 1994 | Change | 1986 | 1995 | Change |
| Total vote | 43 | 45 | 45 | +2 | 43 | 38 | -5 |
| Blue-collar | 66 | 64 | 63 | -3 | 57 | 41 | -16 |
| White-collar* | 33 | 35 | 41 | +8 | 40/49 | 32/48 | -8/-1 |
| Self-employed | 13 | 14 | 17 | +4 | 14 | 18 | +4 |
| Private sector | 46 | 43 | 43 | -3 | 51 | 36 | -15 |
| Public sector | 43 | 43 | 51 | +8 | 51 | 44 | -7 |
* For Austria, the first figure refers to Angestellte, the second to Beamte.
Source: Kunkel and Pontusson, ‘Corporatism versus Social Democracy’, p. 17.
Working-class support for the SPÖ (which habitually secured something over two-thirds of the working-class vote) registered a considerable drop, falling from 64 per cent (average for 1971–75) to 57 per cent in 1983 and 41 per cent in 1995 (Table 7.13). Part of the working-class and popular electorate currently supports the FPÖ, a national-populist party which has established itself in the Austrian system as a right-wing party with strong popular ties and a socio-demographic profile close to that of the SPÖ.19 Thus, this migration to the right of a significant band of Austrian workers, particularly those employed in the private sector, expresses a considerable class dealignment. At the same time – and contrary to the Swedish pattern – from 1986 to 1995 the SPÖ recorded sizeable losses among white-collar workers, with the notable exception of civil servants. These, however, were considerably more modest than the drop in the working-class vote stricto sensu ( – 8 as against – 16).
Historically, the SAP and the SPÖ are characterized by their working-class/popular ties and the institutional role allotted to trade unions as a force representing the interests of wage-earners. The electoral shifts indicated above clearly tend in the direction of a marked weakening in the popular entrenchment of these parties. The fall in the working-class vote in Sweden is all the more significant in that the last three decades have witnessed increased unionization. The considerably more pronounced decline in the working-class vote in Austria, where trade-unionism has been weakened in the same years, demonstrates that while high union density cannot reverse the trend to working-class disaffection, it remains a positive factor in influencing workers to vote for the left.20
France
In France, fluctuations in the working-class vote were central to the political and social upheaval that accompanied the transition from the Fourth to the Fifth Republic. The two great breaches in the continuity of working-class electoral behaviour can be clearly located in 1956 and 1973, which constitute the point of departure for two contrasting developments. The year 1958 marked the beginning of a period of class dealignment, whereas 1973 initiated a phase of new and strong class polarization (Table 7.14). By contrast, the 1990s saw a new decline in class voting among workers, comparable to – but rather more pronounced than – the Gaullist period of the 1960s (Table 7.15).
Table 7.14 Electoral penetration of the French left according to socio-professional group (1956–97)
| Workers | Middle managers/white-collar employees | Workers | Middle managers/white-collar employees | ||
| 1956 | 66 | – | 1981Pb | 72 | 62 |
| 1958 | 55 | – | 1981 | 69 | 63 |
| 1965Pb | 55 | 45 | 1986 | 61 | 52 |
| 1967 | 54 | 44 | 1988Pb | 68 | 61 |
| 1968 | 57 | 43 | 1988 | 60 | 55 |
| 1969Pa | 50 | 36 | 1993 | 46 | 46/43d |
| 1973 | 68 | 44 | 1995Pb | 52 | 44/42d |
| 1974Pb | 73 | 53 | 1997 | 52 | 51/54d |
| 1978 | 70 | 54c |
a: first round; P: presidential election; b: second round; c: salaried non-workers; d: intermediate professions/white-collar
Sources: 1956–73: Gerassimos Moschonas, La Gauche française (1972–1988), doctoral thesis, University of Paris II, 1990, pp. 233, 255; 1978 and 1995: Daniel Boy and Nonna Mayer, eds, L’èlecteur a ses raisons, Presses de Sciences Politiques, Paris 1997, p. 109; 1974, 1981 and 1986: Gérard Le Gall, ‘Mars 1986: des elections de transition?’, Revue Politique et Parlementaire, no. 922, 1986, p. 11; 1988: SOFRES, L’état de l’opinion, Seuil, Paris 1989; 1993 and 1997: Fondation Saint-Simon (collective), Pour une nouvelle république Sociale, Callmann-Lévy, Paris 1997, p. 154.
Table 7.15 Socio-professional groups and indices of evolution of the left-wing vote in France (1956–97) (averages by decade)
| Workers (base 100: 1956) | Workers (base 100: average for 1965–69) | Middle managers/white-collar employees (base 100: average for 1965–69) | ||
| 1960s/1956 | 81.81 | |||
| 1970s/1956 | 106.56 | 1970s/1960s | 130.24 | 119.84 |
| 1980s/1956 | 98.73a | 1980s/1960s | 120.67a | 136.90a |
| 1990s/1956 | 75.75 | 1990s/1960s | 92.59 | 111.90/110.31 |
a: The two elections in 1981 have been taken together, as have those in 1988. To calculate the index of evolution for the 1980s, I have used the average for each of the pairs of elections in question.
Source: See Table 7.14.
The specificity of Gaullism – or, at least, the presidential Gaullism of de Gaulle – derived from its ability to attract an electorate from diverse ideological and social horizons, from the centre-left to the ‘extreme left’ (as it was then called). The high level of Gaullist influence among the working class, the attachment of a number of voters of communist origin to the ‘enlarged’ or ‘maximal’ electorate of Gaullism, the use of referenda as an instrument of partisan dealignment – these are so many indices of the socially and politically ‘disruptive’ role of de Gaulle’s approach. Gaullism was unquestionably the great unifier of the conservative firmament in France. But it was a singular, almost atypical conservatism. Its strong representation among workers, and relative weakness among higher executives, the liberal professions and, at the outset, the rural heartland, are among the factors that emphasize the original character of electoral Gaullism.21
If the Gaullist adventure appears unintelligible without reference to the working-class penetration of Gaullism, the Union of the Left is no less so without reference to its ability to attract an overwhelming majority of the working-class electorate from the time of its formation in 1972. It was precisely by reconquering its ‘natural’ social territory that the French left, in the shape of the Union, broke the vicious circle in which it had been trapped during the initial period of the Fifth Republic. In fact, different social groups rallied to the Union in two phases: the first was marked by an adhesion of working-class strata which was so rapid and massive as to be disconcerting; the second by the gradual adhesion of salaried middle strata (Table 7.14). Thus, at the heart of the left coalition’s expansion, chronologically and qualitatively, was the working-class strata’s turn to the left. Everything cohered around this: the Union’s achievement of competitive status allowed it to present itself as a natural force of political alternation; this triggered a process of adhesion by the salaried middle classes to a pole that was credible and possessed of a governmental vocation.
Thus, strengthened class voting in France during the 1970s was matched only by weakened class voting – or absolute class voting, at any rate – in countries like Great Britain or Denmark in the same years.
In the 1980s and 1990s, in the context of a general decline of the French left, a dual dynamic ensued: on one hand, considerable working-class disaffection in 1986, which was very ‘discreet’ in relative terms (the left actually fell back among all social classes), but very visible in absolute terms, and became massive – not to say spectacular – in 1993; on the other, much greater resilience on the part of the left among intermediate professions and middle management. This trend is accentuated if we take into account the electoral behaviour of higher executives and intellectual professions, where we observe a greater tendency to vote for the left in 1995 and 1997 (43 per cent and 49 per cent, respectively, as opposed to 31 per cent in 1978), despite the worse overall performance of the left in those years.22 In sum, the differential in the left vote between blue-collar and white-collar workers diminishes appreciably, falling from 24 points in 1973 to 10 points in 1995, to approach zero in 1997! Accordingly, if swings of the electoral pendulum describe the electoral behaviour of the working class accurately enough over a long period (1956–97), the traditional opposition between waged workers and non-workers gradually loses its salience because of salaried middle strata’s greater attachment to the left. By contrast, the opposition in terms of electoral behaviour between the self-employed and wage-earners (workers and non-workers) persists, because in their clear majority the self-employed (shopkeepers, artisans, farmers) remain attached to the right.23 What must be stressed is the fact that the most significant reductions in the left’s working-class influence – important losses in 1986, violent contraction in 1993 – follow their inglorious governmental experiences. In addition, there is evidence of a veritable realignment of workers towards the National Front, which obtained 13 per cent of the vote in 1988 (11 per cent of the whole electorate), 18 per cent in 1995 (13 per cent overall), and 24 per cent in 1997 (15 per cent overall). This realignment is all the more significant in that it is the workers most integrated into the working-class universe (that is to say, children and spouses of workers) who support the extreme-right party most strongly. Likewise, Chirac’s strong showing among the French working class in the 1995 presidential election was manifestly linked to the candidate’s decision to conduct a campaign against la pensée unique, and put himself forward as champion of the struggle against social inequalities. Thus – more visibly, perhaps, than elsewhere – class voting by workers is punctuated in France by the political appeal of the opposing parties and the bad governmental record of the left. Politics counts; it even counts a lot – that is the great lesson of the history of the last forty years of class voting in France.
Class Voting and the Impact of Political Action
The Major Trends
An overview of the pattern of class voting clearly indicates that a process of class dealignment has occurred in the great majority of cases considered. More specifically, we can identify in this process two dynamics, which are simultaneously complementary and distinct.
(i)Class political dealignment by social democracy’s downturn among manual strata
The first major trend to be registered is the erosion in working-class penetration by social-democratic parties (with the small but ambiguous exception of the SPD). As a general rule, this erosion benefits the traditional ‘bourgeois’ parties or new right-wing ‘populist’ parties, and in some cases – but only partially – formations to the left of the social democrats (to some extent this is the case in Sweden, Denmark and Norway). The nature, extent and rhythm of working-class defection vary according to country and political-electoral circumstances. But the basic bond of mutual trust between these parties and the working class – what Hobsbawm called a ‘unique advantage’ – risks disintegration.
The second major trend to register is an increased instability in the electoral behaviour of workers. Overall, their support for the left is more fragile, more subject to conjunctural factors and the interplay of competition.
Nevertheless, measured over an extended period (1990s/1960s or by decade), and thus setting aside short-term fluctuations, the social-democratic downturn among the working class is fairly moderate. All in all, the picture that emerges is of social democracy in the 1980s and 1990s establishing itself at a level appreciably below that of the 1960s, with its working-class influence fluctuating around 80 per cent of the total for the 1960s. Relative stabilization at this level confirms the impression of a non-conjunctural class dealignment. However, this decline is neither imposing nor spectacular. Moreover, nothing points to a growing deterioration in the links between socialists and the working class in recent years; nothing indicates a more or less brutal aggravation of the trend. On occasion, working-class disaffection has certainly assumed extraordinary proportions (Denmark in 1973; Great Britain in 1983) – and will probably do so again in the future. This has often created the impression of a cataclysmic change, the definitive ‘end’ of class voting. But such a picture sins by exaggeration. It is unquestionably true that working-class attachment to the left no longer possesses the vigour, power, durability and political substance of yesteryear. But it exists. Social-democratic parties remain the ‘natural’ parties of the working class. Yet their capacity to represent it electorally has definitely diminished significantly.
(ii)Class dealignment by the advance – or greater resilience – of social democrats among the salaried middle strata
Over the last thirty years, increased penetration (in relative terms, but often also in absolute terms) of social democracy among these categories is clearly apparent. The social democracies are increasingly sustained by votes from the salaried middle classes, allowing them to limit the damage done by working-class disaffection and the fall in working-class numbers in the population. This influence is essentially concentrated among the ‘new’ middle classes or the ‘salaried middle strata’ (office workers, intermediate executives and professions, technicians and engineers, teachers, health and social service workers, etc.) – categories that are undergoing strong arithmetical expansion. More particularly, the most cultured band of the new middle classes – the intellectual professions (academics, teachers, and the information, arts and entertainment professions) – are categories that display an increased tendency to vote for the left (social democrats, left socialists, ex-communists, or ecologists). There is nothing to suggest that this is a conjunctural shift. The cultural liberalism of these highly cultured categories (‘moral’ permissiveness, anti-authoritarian/participatory political values, universalist/anti-ethnocentric attitudes) makes them suspicious of the cultural conservatism of much of the traditional right, and unsusceptible to the authoritarian and anti-universalist discourse of the extreme right.24
Thus, the structure of social democracy’s electoral development in the period 1960 to 1997 seems clear enough: in phases of electoral breakthrough it gains rather more ground among the salaried middle strata than among workers; and in phases of downturn it is more resilient among the former than the latter. We should also note that in a majority of the countries considered here, the level of social-democratic influence among the salaried middle strata in the 1990s is either higher than or equal to that of the 1960s, even though the electoral performance of socialist/social-democratic parties is often inferior.
The progress of socialist/social-democratic parties among the salaried middle strata is partially challenged, however, in countries where formations of a ‘new politics’ (or semi-new politics) type emerge strongly. Such formations tend to root themselves primarily in these strata, restricting social-democratic influence there (examples are Germany, Norway, Denmark and Sweden).
We should also underline a marked tendency towards a relative strengthening of social-democratic penetration of public-sector strata, regardless of their manual or non-manual character and hierarchical status. The salaried strata of the public sector – and, more especially, those who work in nationalized enterprises, education, health, research, communications, social services in general (the ‘state service sector’, as Wright and Cho term it)25 – are becoming the new epicentre – alongside the working class – of the left’s electoral strength. The ‘state service middle class’ is characterized by a pronounced attachment to left-wing values, and is significantly further to the left than the middle classes in the private sector. The influence of the sectoral divide on the left-wing vote is strong in the majority of the cases studied in this chapter. It is particularly strong in the Scandinavian countries (notably Denmark, but also – and in descending order – Norway and Sweden), France and Greece. It is visible in Austria and Great Britain. Public-sector salaried employees in part represent the new conquered land of the social democrats, and could prove even more of an acquisition in the future, if social democracy succeeds in neutralizing the influence of parties representing varieties of ‘new politics’.
(iii)The social democrats are the object of a persistent rejection by the traditional middle and petty bourgeoisie, as well as – in a majority of cases – small farmers
Relative decline in traditional commerce and craft industry is accompanied by an expansion in new commercial and artisanal categories of a higher social and cultural level. The relative embourgeoisement of the traditional ‘new’ petty bourgeoisie, as well as the increased salience of the tax issue, explain its rejection of the left and electoral support for the right.26 The small and medium bourgeoisie, attached to the culture of the firm, remain the solid, central electoral bloc of the right – in some cases more so than in the past, in others a little less. With the exception of a very limited number of countries (notably Norway, Italy and Greece), the world of small and medium-sized enterprise acts as a stable bloc with extraordinary political cohesion. Thus, if we take as our point of departure and reference the classical Marxist division between bourgeoisie (petty, medium and large) and working class, which is still valid, we observe that class dealignment is restricted to the political behaviour of the working class. We are dealing with a unilateral class dealignment.27
The Working-Class Electorate and the Importance of Politics
The ‘class’ factor has partially lost its former pre-eminence in European political systems. According to a comparative analysis by Paul Nieuwbeerta, who has studied the relationship between social class and electoral behaviour in several countries (measuring absolute as well as relative class voting), ‘a fairly monotonic decline in the countries’ levels of class voting has occurred, but around this trend fluctuations are visible’.28 This is broadly in keeping with the findings of the preceding analysis. The ‘social psychology’ of class voting has changed,29 although there is nothing to suggest the end of the electoral class struggle.
How, then, are we to interpret the ‘lost’ loyalty of part of the working class, and the variations in its political options?
Clearly, the transformation of the societies of advanced capitalism does not favour ‘old-style’ class political polarization. Some basic sociological factors (arithmetical decline of the working class stricto sensu, increased professional fragmentation, unemployment, ethnic diversification of the most marginal sections of popular strata, considerable growth in the number of people with an ‘uncertain status’, the development of ‘flexible’ forms of employment), as well as some specifically political factors (the adoption of a catch-all profile by left-wing parties), lead to a further attenuation of the link between the working class and voting for the left. However, while the Keynesian equation proved effective, neither social changes – which have been going on for a long time – nor catch-all strategies – which date back to the 1950s – resulted in working-class disaffection, except marginally or in circumstantial fashion. ‘Structural’ explanations, the ‘“bottom-up” approaches to the structuring of political divisions’ dominant in political sociology, often ignore the fact that the effects of social change on voting are very uncertain. In addition, they underestimate – just as they always did – the impact of political action, and political actors’ capacity to adapt. ‘The most interesting characteristic of Norway’s party system’, Kaare Strom and Lars Svasand have written, ‘lies in the magnitude of the drama we shall observe: the scale and rapidity of social change and the vigour of the organizational response.’30 Are workers’ political choices independent of the action of political organizations – of the ‘organizational response’? Are they reducible to the social change experienced by the countries of advanced capitalism? Are they principally to be explained by ‘subterranean’ developments in social structure? On this vast subject, which is scarcely new, I shall restrict myself to some past and present examples, in order to bring out the importance of political action in the establishment – and testing – of class political alignments.
The theme of the ‘embourgeoisement’ of the working class found its most fervent adepts in the late 1950s and early 1960s. Following three consecutive defeats for Labour, the ‘affluent worker thesis’ was particularly in vogue in Britain. In France, René Mouriaux took it as an established fact in 1970 that ‘workers increasingly vote for the candidates of the right. This is a transnational phenomenon that has assumed greater magnitude in France than elsewhere and a particular complexion with the presidency of General de Gaulle.’31 Likewise, Andre Philip asserted: ‘we are currently witnessing the decline of working-class parties to a certain extent everywhere. I am not referring to the crisis of the SFIO in France … I am thinking mainly of Germany and Great Britain.… Labour has lost three times, the SPD has lost three times.’32 In the event, I am obliged to observe that after the great debate on the ‘historic decline’ of Labour, it recovered and succeeded in exercising power for eleven out of the next fifteen years, before being plunged into a very serious crisis after the electoral defeat of 1979. Similarly, the SPD was able to establish itself as the main party of government in the FRG throughout the 1970s. Moreover, in 1966 and 1972 respectively, the parties in question achieved one of their best electoral performances among the working class.
As for France, the thesis that the decline in working-class ‘leftism’ from 1959 to 1969, and more generally the rise of Gaullism, was caused by social change and an ‘embourgeoisement’ of the working class (established according to the classical equation between economic prosperity and lower class consciousness), was contradicted by the facts. For the great shift to the left among workers occurred in 1973 – still a time of exceptional prosperity. In reality, it was factors of a political nature that made possible the detachment of a significant – though not central – part of the working class from the left. Gaullism’s ability to make non-economic issues (decolonization, institutional modernization, European integration, the Atlantic Alliance) salient questions in French political life, plus the General’s exceptional charisma, explain the trend in question. Thereafter, it was modification of the ‘electoral alternatives’, conditioned by the events of May 1968, the retirement of the man of 18 June, and especially the formation of the left alliance, which led to the strong resurgence of the ‘class’ factor on the French political scene. Thus, in a period of ‘post-industrial change’ and ‘class dealignment’, the left found itself on the verge of power in 1974, thanks in large part to strengthened class voting. This represented the veritable ‘revenge’ of the electoral class struggle which, after a fifteen-year delay, got the better of the interclassist political groupings that had dominated the initial stages of the Fifth Republic – even if this revenge proved (for good reasons) short-lived.
In fact, it would appear that in France, as in Great Britain and elsewhere, whether in the 1960s or today, basic sociological factors ultimately play only a background role. They can certainly help to account for the ‘intensive’ or ‘diffuse’ type of support given to political parties by workers, or for the emergence of new electoral issues and political divisions; but on their own they cannot explain the major or minor shifts among working-class electorates that have occurred, or are occurring, in European countries. Let us take an example. Sociologically, the united left in France was structured in a similar fashion to ‘classical’ social democracy, at a time (the 1970s) when the latter was commencing its own dynamic of ‘de-structuration’. If the Union of the Left fundamentally changed the realities of electoral competition in 1970s France, and contributed to strengthened class voting, the failure of governmental socialism in terms of social policy encouraged a sharp decline in the left-wing orientation of workers – a decline that was very evident in the 1993 election. Working-class defection followed an administration marked by deflation and austerity. We can see the same decline in class voting in Great Britain at the end of the 1960s and 1970s, when the unpopularity of outgoing Labour governments became obvious, and there was ruthless head-on confrontation with trade-unionism. The hypothesis formulated by Bruno Cautrès and Anthony Heath in their study of the decline of class voting in Great Britain and France would seem to be of more general relevance: ‘class voting … loses its intensity at moments marked by the striking unpopularity of outgoing left-wing governments … the support of working-class voters for left-wing governments unquestionably [being] more pragmatic than ideological, unlike the middle strata’.33
Indeed, the available statistics seem to indicate that disaffection with socialist/social-democratic parties on the part of a far from negligible proportion of the world of labour and disadvantaged sections of the population tends to accelerate after a governmental failure by the left, especially in social policy. Thus, in some countries the downward trend in left-wing voting by workers was marked by brutal ‘ruptures’, and often by reverse – but not commensurate – subsequent swings of the pendulum (Great Britain, Denmark, France in part). In other countries, working-class defection was more linear, although not necessarily less pronounced (Norway), and not without ‘minor’ ruptures (Austria; Sweden to some extent).
The slower, more gradual character of the loosening of the link between the working class and socialist voting in Norway might seem – in this instance, at least – to lend more weight to a sociological than a political explanation – all the more so given that Norway experienced stronger economic growth than other countries during the 1970s and 1980s, largely linked to exploitation of the ‘black gold’. Since the 1980s, Norway has been among world leaders in terms of indicators of prosperity (gross domestic product per capita, wage levels, life expectancy, medical cover).34 When the development of Norwegian political life is examined more closely, however, we see that the political situation and the issues at stake in it – economic circumstances become a dimension of the political conjuncture once they are perceived as a matter for political action35 – did not favour the Norwegian Labour Party. Thus, the slowdown in growth and the considerable deterioration in the trade balance coincided with the DNA’s presence in government (1977–81). The Labour government found itself obliged to pursue a policy restricting total demand and freezing prices and wages (September 1978). Between 1978 and 1981, workers’ real wages fell, and Labour, converted to austerity, was driven from office. The DNA’s return to power in 1986, in the teeth of economic crisis, was again accompanied by a dose of austerity as well as rising unemployment, particularly following the major recession of 1988. As a consequence, the electoral cycle weighed heavily on the fate of the DNA, unable to benefit from the economic boom of 1983–86 because it was confined to opposition.36 In fact, even in the case of Norway the weakening left-wing inclinations of the working-class electorate appear to be largely bound up with the political conjuncture and the performance, particularly in social policy, of Labour governments. Here, too, an explanation in political, and not purely sociological, terms seems more pertinent. Moreover, the marked acceleration in working-class defection during this precise period appears to favour this line of explanation.
The case of the Greek PASOK confirms this analysis a contrario. Its first stint in power (1981–85) – rather than weakening the party’s popular bases, as happened in countries with a socialist government during the recession – strengthened its popular anchorage. This novel development was bound up with the fact that the first genuine attempt to construct a welfare state in Greece coincided with PASOK’s arrival in government. Enshrined in PASOK’s populist logic, the satisfaction of narrow sectional demands, accompanied by the establishment of a clientelistic socialist state, certainly led to the state getting deeply into debt in an adverse economic situation, destabilizing and partially delegitimizing the whole enterprise. But even if it was a ‘bad manager’, and politically cynical, the image of a socially ‘generous’ PASOK allowed the Greek socialists to put up better resistance among working-class and popular strata in their electoral defeats of 1989 and 1990. On the other hand, having adopted a managerial profile, and enjoying an incontestable advantage over its conservative opponent on the terrain of ‘European pragmatism’, PASOK alienated important sections of the socially and culturally disadvantaged in 1996, notwithstanding its electoral victory and breakthrough among well-off sections of the population.37 This alienation was generally confirmed and magnified in the 1999 European elections, and in one sense in the legislative elections of April 2000. Returned to power, PASOK, despite a certain recovery of influence among popular strata, increasingly presents itself as the party of ‘the contented’.38 This is a remarkable development for a party often regarded in the not too distant past as ‘populist’.
Lost Comrades: The Working-Class Vote and the Extreme Right
The breakthrough by the populist extreme right among the working class seems to confirm the thesis of the salience of political factors. Indeed, in some countries the extreme right increasingly begins to appear to a significant percentage of the popular electorate as a ‘natural’ political alternative. Today, it is unquestionably true that the formations of the populist right are (in Dominique Pélassy’s words) ‘expanding by popularizing themselves’.39 In Europe, part of the ‘plebeian public sphere’ increasingly speaks the language of the extreme right, which is establishing itself as an ‘exit option’ for workers and culturally disadvantaged strata.
The class profile of parties of the new populist right (e.g. the FPÖ in Austria, the FN in France, the Progress Parties of Denmark and Norway) exhibits ‘an increasing deviation from conventional bourgeois parties’.40 As late as the 1980s, the FPÖ’s electorate comprised more executives, self-employed, members of the liberal professions and graduates than the average for the population. However, Jörg Haider’s accession to the party leadership (1986) drastically altered the structure of its support: a remarkable increase in its influence among workers (in its social make-up the FPÖ is as ‘popular’ today as the SPÖ); an improvement in its penetration among employees; and a drop among educated sections of the population.41
In France, the ‘popularization’ of the FN is more pronounced still, and an analogous pattern has occurred in Denmark and Norway. The three parties in question are overrepresented among the working class and the traditional petty bourgeoisie (though the latter’s participation in the electorate of the Norwegian Progress Party is similar to its weight in the total population), and significantly underrepresented among the new middle classes. In addition, we should note that the electorate of the populist right is not ideologically extremist on all issues and in every respect. In France, Le Pen’s electorate in 1995 was further to the left in its economic ideology than the electorate of the moderate right. Ideologically, it is genuinely ‘inbetween’ the electorates of the left and the traditional right.42 According to Gérard Grunberg and Étienne Schweisguth, the ideological complexion of the French extreme right ‘can in no way be described as simply being more right-wing than the moderate right … but … owes its characterization as extreme right to its extreme rejection of universalist values’.43 The electorate of the Danish Progress Party is also less hostile towards the welfare state than that of the other non-socialist parties.
Typical ‘anti-tax’, anti-bureaucratic and anti-establishment formations, which initially defended the freedom of the private sector, the Danish and Norwegian parties (particularly the former) have gradually become more moderate on all these points. During the 1980s the Danish Progress Party adopted a noticeably more social profile (supporting increases in pensions and health expenditure) and, at the same time, a more ‘anti-immigrant’ stance.44 More faithful to its customary campaign themes, the Norwegian Party is evolving in the same direction, as was evident in the parliamentary elections of 1997. In Denmark and Norway ‘it is probably more correct to speak of “welfare state chauvinism” than of anti-welfare attitudes: the welfare services should be restricted to “our own”’.45 In France, too, an originally very neoliberal discourse has gradually given way to a socioeconomic platform mixing neoliberalism, state interventionism and socializing measures.46 Furthermore, the FPÖ, after the weakening of its liberal wing following a split in 1993 (and the creation of an authentically liberal party, the Liberal Forum), while advocating a reduced economic role for the state, has put more emphasis on the populist and protest aspects of its policies (violent rejection of the European Union and its domination by multinational companies and ‘unrestricted neoliberalism’; reform of the Austrian political system).
Despite their originally neoliberal profile and strangely archaic conception of the role of public power (fewer economic and social functions, more police), the parties of the extreme right are today typically in the process of adopting – albeit slowly, and in a highly equivocal fashion – a more composite ideological-programmatic profile, less marked by neoliberal pro-capitalism.
There can obviously be no question here of analysing the distinctive features of a very heterogeneous political family. But given our aim, one observation is necessary: the parties of the new extreme right, whose doctrine is inegalitarian and anti-universalist, at a less conspicuous but increasingly explicit level produce an egalitarian ‘ambiance’, which is not (or not yet) an articulated discourse. In attacking the ‘powerful’, the ‘establishment’ and the ‘coalition of elites’ in the name of ordinary people and the ‘small man’, in timorously adopting a more ‘social’ discourse, they are becoming the perverted expression of a certain anti-egalitarian egalitarianism.47 They are becoming the political instruments of a populist ‘egalitarianism’ – selective, diffuse and badly articulated – which coexists with a liberal-capitalist anti-egalitarianism (in marked retreat) and an aggressive, domineering ethnocentric anti-egalitarianism, the strongest and most distinctive element in the ideological/programmatic profile of the new extreme right.
In fact, in every instance the extreme right’s strongest niche is the issue of immigration and ethnocentrism. The parties of the populist extreme right are unquestionably protest formations that garner the protest vote against the traditional parties. But above all, they are organizations that ‘capitalize on the neglected issue’.48 This applies even to the parties in Scandinavia, where immigration became a salient issue ‘only after the Progress Parties were established’.49 It is on immigration, ‘law and order’ and, more generally, attitudes towards universalist values that these parties find the most fertile ground for establishing and consolidating themselves, before finally blurring the division between left and right. Within a system comprising ‘authoritarianism, ethnic particularism and market liberalism’,50 immigration and ethnocentrism (‘anti-universalism’ according to Grunberg and Schweisguth) play a major role in differentiating the extreme right from left and moderate right alike.
But why does a far from insignificant percentage of the working-class and popular strata turn to the new extreme right? Can this shift be explained by the notorious ‘cultural traditionalism’ of workers, or the historically greater diffusion of ‘anti-universalist’ values among them? Is it a product of ‘working-class fascism’?
Historically, social democracy, and the left in general, established itself as a central, strategic pole within the popular milieu – as an ‘attractor’ (to borrow a term from Luc Boltanski). It was an ‘attractor’ that was certainly political, but also cultural, contributing to the formation of the collective identity of the subaltern classes. Historically, the left was not simply a force for the representation of socioeconomic interests but also a force for civilization, bearer of an ‘ethical’ project and universalist, egalitarian values. Thus, if in the past there was a certain discrepancy between left-wing values and ‘spontaneous working-class consciousness’ (the famous – very controversial – ‘working-class authoritarianism’), the effectiveness of ‘economic leftism’, as well as the ideological efforts of left-wing organizations, played a decisive role in the attraction, unification and attachment of the working class to left-wing parties that ‘aimed at broader and anti-authoritarian goals’.51 Class interests took precedence over value orientations – or better: left-wing parties (and trade unions), possessing a remarkable capacity for the ideological and cultural orientation of their class, were in a good position to influence – almost to fashion – value orientations in popular milieux. Today, by contrast, the economic and social policy of the social-democratic left no longer polarizes social groups sufficiently, because it is no longer perceived as sufficiently distinct from its neoliberal opponents. The left ‘no longer succeeds as readily in inflecting demands of an economic kind to the left ideologically’.52 Because of its ‘failure’ in social policy, it does not succeed in carrying workers in its ideological/cultural wake either. With the ideological, cultural and organizational decline of the political and trade-union left, and its traditional networks of influence, a political and cultural space has opened up. A new polarization has been created, which is most conspicuous in France: the ideological centre of gravity of popular strata (even those who do not vote FN) has partially shifted towards the values of the right and extreme right (xenophobia, ethnocentrism, anti-universalism), while the cultivated upper classes maintain universalist positions that verge on a certain left-wing humanism.53
Basically, the absence of a clearly perceptible left-wing differentia specifica in economic and social policy distances a section of workers not only from the left, but also from the ‘universalist’ values it represents. Moreover, the increased weight in the social-democratic parties of the new ‘cultural class’ – liberal, often neoliberal, in its value orientation – shatters the solidarity between the ‘people’ and social-democratic elites, between apex and base of the social-democratic pyramid. Now, in the absence of a strong and strongly shared left-wing ideological framework today, this gulf between social democratic ‘elites’ and ‘masses’, between those ‘above’ and those ‘below’, is a supplementary factor in the defection of the workers. Basically, a vicious circle is initiated, the cultural effects of political options being compounded by the political effects of cultural ‘predispositions’. When the impact of the traditional socioeconomic divide on voting decreases, the libertarian versus authoritarian political dimension assumes greater salience and importance. The result of socioeconomic disappointment among workers is thus added on to the result of popular authoritarianism, and it is not always easy to disentangle the two.
Voting for the populist extreme right by popular strata is largely the indirect result of what is today perceived as an ‘illusory choice’: it springs from great disillusionment among a section of the popular classes.54 Thus, at the geometrical centre of several reasons for the consolidation of the popular bases of the new extreme right, at the core of its dynamic, is the major failure of its principal opponent: the egalitarian project of the left. In this respect there is no better gauge of the social defeat of the left in government than the extreme right’s breakthrough in working-class milieux.55
The Breakdown of the Keynesian Equation and Working-Class Defection
We cannot understand the electoral link between social democracy and the working class, its nature and political significance, without considering a large number of parameters from which this link derived. Consequently, the hypothesis foregrounding political factors requires much more corroboration. Nevertheless, it seems clear that it was with the ‘breakdown’ of the Keynesian equation that social democracy’s position was significantly weakened in the world of labour. For the period that runs from the beginning of the 1970s to the present, considered from a macroscopic angle, the trend is pronounced and unambiguous. Basically, the decreasing propensity of workers and disadvantaged strata to vote for the left exposed the social-democratic parties’ inability to manage the ‘social question’ effectively. Dependent on the very difficult general conjuncture, which was unfavourable to parties in power, socialists in government paid a high price. The corrosive effect of being in power was nothing new. What was new was that it appeared to exact a higher social cost than in the past (see Chapter 6). In this respect, the minor and partial German exception is not really such an exception. Out of power from 1982, the SPD was largely able to maintain its positions within the world of labour. This will not necessarily be the case in the future.
In reality, if workers are less susceptible than before to the social-democratic political appeal, it is partly because this political appeal is less sensitive to working-class demands.56 Depending on social circumstances, the success or failure of a governmental experience, the structure and dynamic of the conservative ‘opponent’ and ‘opposition bloc’ – in short, the conjuncture of electoral competition – the working-class influence of social-democratic parties can decline or recover. But in every instance, today as in the 1950s and 1960s, the political appeal and governmental performance of left-wing parties appears to play a very important role.
So, at a time when working-class attachment to the left is proving less strong and stable, less total and intense, more episodic and conditional – in short, ‘instrumental’ – largely owing to the social and cultural recomposition analysed by supporters of the ‘structural’ thesis, the ‘political strategy’, ‘electoral appeal’, and especially ‘governmental performance’ of the left become central factors in accounting for the electoral behaviour of workers and subaltern classes.
It is absurd to oppose a ‘vote on the issues’ whereby an individualist worker, free of all social/cultural constraints, makes his or her choice according to the questions of the hour and the programme on offer, and a ‘sociological’ vote to which each elector is compelled in so far as she or he is constrained by the weight of class, religion and family tradition.57 The ‘sociological’ vote does indeed exist; without it, the structure of party systems in Europe would be utterly unintelligible. Accordingly, foregrounding the political and conjunctural factors that determine working-class voting does not entail that we can thereby clearly demarcate ‘conjunctural’ from ‘structural’ determinants. A specific political conjuncture does not constitute ‘a short-term political force’ with only temporary effects.58 When it comes to partisan identification, the ‘conjunctural’ is often the vector of the ‘structural’; it can result in stable electoral behaviour, especially if a ‘conjuncture’ lasts a long time, as was the case after 1973. Moreover, political factors are all the more important when they affect a social class’s ‘potential for mobilization’. Political action and the effects of political action are decisive in the construction of class identities. In this sense, the attenuation of class voting by workers as a result of political factors is primarily a political phenomenon with social consequences. But it can become a social phenomenon with political consequences.59
Class politics is in retreat. It is in retreat because of the transformation (not the decline) of classes, but also because of the politics of the left. The analysis above does not prove that the electoral class struggle is at an end. Instead, it shows that the mediating role of political factors – and actors – between class structure and the act of voting is very important – in all likelihood, more important than it was in the past.60
The Social-Democratic Electorate:
The Internal Sociological Dialectic
The Social Profile of the Social-Democratic Electorate
The social profile or social specificity of an electorate describes its class cohesion. It thus represents a useful indicator for identifying the social character and, indirectly, the cultural character of a party, as well as its ability to adopt a coherent class politics without clashing with whole sections of its own social base.
Moreover, focusing not on the electoral behaviour of classes, but on the class composition of electorates, not on the political specificity of social classes but on the class specificity of electorates, makes it possible to understand the internal sociological dialectic of the social-democratic coalition, by identifying the social groups within it which are in the ascendant or in decline.
If we look at the data in Table 7.16, we see that the class cohesion of electoral social democracy has been strongly affected since the 1960s, confirming the hypothesis of a significant recomposition of the social-democratic electorate.
Table 7.16 Social composition of the social-democratic electorate
| Denmark | Austria* | Germany | ||||
| 1966 | 1987 | 1969 | 1995 | 1967 | 1994** | |
| Workers | 72 | 50 | 68 | 39 | 61 | 43 |
| Salaried middle strata | 23 | 46 | 25 | 50 | 32 | 46 |
| Self-employed | 5 | 4 | 7 | 4 | 8 | 5 |
| Norway | Sweden | Great Britain*** | ||||
| 1965 | 1989 | 1960 | 1994 | 1966 | 1992 | |
| Workers | 64 | 48 | 76 | 52 | 81 | 63 |
| Salaried middle strata | 22 | 37 | 18 | 43 | 19 | 37 |
| Self-employed | 14 | 15 | 4 | 4 | ||
* In reaching these figures, we have excluded housewives and retired people from our calculations.
** West Germany.
*** The class composition of the Labour Party is established according to the manual/non-manual distinction.
Sources: For the Danish Social Democrats, see Gøsta Esping-Andersen, Politics against Markets, Princeton University Press, Princeton, NJ 1985, p. 128 for 1966, and Jorgen Goul Andersen, ‘Denmark: Environmental Conflict and the “Greening” of the Labour Movement’, Scandinavian Political Studies, vol. 13, no. 2, 1990, for 1987. For the SPD, see Anton Pelinka, Social Democratic Parties in Europe, Praeger, New York 1983, p. 46 for 1967, and Russell Dalton and Wilhelm Bürklin, ‘The Two German Electorates’ in Russell Dalton, ed., Germans Divided, Berg, Oxford 1996, p. 199 for 1994. For the SPÖ, see Anton Pelinka and Fritz Plasser, The Austrian Party System, Westview Press, Boulder, CO 1989, p. 75 for 1967, and Kurt Richard Luther, ‘The Social-Democratic Party of Austria’, in Robert Ladrech and Philippe Marlière, Social-Democratic Parties in the European Union, Macmillan, London 1999, p. 25. For the Norwegian Labour Party, see Jorgen Goul Andersen and Tom Bjorklund, ‘Structural Changes and New Cleavages’, Acta Sociologica, vol. 33, no. 3, 1990, pp. 208–9. For the SAP, see Nicholas Aylott, ‘The Swedish Social-Democratic Party’, in Ladrech and Marlière, eds, Social-Democratic Parties in the European Union, p. 196. For the British Labour Party, see Bo Salvik and Ivor Crewe, Decade of Dealignment, Cambridge University Press, Cambridge 1983, p. 89 for 1966, and David Denver, Elections and Voting Behaviour in Britain, Harvester Wheatsheaf, London and New York 1994, p. 11 for 1992.
A sizeable displacement in the sociological centre of gravity of this electorate towards the salaried middle strata has occurred, such that two social poles of almost equal importance now seem to coexist within the social-democratic space. This being so, the arithmetically massive and dominant presence of workers, which was constitutive of social democracy, faces significant challenge. As a consequence, the class character and social cohesion of this political force are considerably diluted.
This development seems to be in line with the gradual modification in the class equilibrium of modern societies over the last thirty years, and the strategic issue it raises: establishing an ‘alliance’ between workers and the expanding new middle strata, which, in the long run, would mean the sociological conversion of social democracy from a working-class/popular party into a party of salaried strata relatively close to the catch-all model.
The increased interclassism of contemporary social democracy is unquestionably a phenomenon of adaptation, but it also results from workers’ reduced propensity to vote social-democratic, and the electoral weakening that ensues. Considered from this viewpoint, it is – albeit in the second instance – a retreat. But at the same time, in that it signals social democracy’s implantation on new social terrain, it is primarily – in a paradox that is merely apparent – an interclassism of conquest, of expansion. This new, profoundly interclassist profile is the infallible sign of a mutation in social democracy’s electoral space. The composition of this space, its mould, has changed considerably.
Conclusion: The New Diversity of Electoral Social Democracy
Following the data we have just presented, some reflections on the ‘new’ face of social-democracy-in-the-electorate suggest themselves:
1.As in the past, social democracy is essentially constituted socially around two pillars: the working class and the salaried middle strata. But the architectonic and balance of the edifice have been modified.
(a)The relative weight of the traditionally central bloc of the social-democratic electorate, constituted by the working class, has diminished markedly. This is due to two factors: (i) an important reduction, both absolute and relative, in the number of workers in the population; (ii) the proportionately greater dissatisfaction of workers with parties of a social-democratic type. Given that working-class defection is fairly moderate, the first is more important.61
(b)A second grouping – the salaried middle strata – has gradually asserted itself alongside the working class, which remains an important pole of support, but whose centrality has diminished significantly. Social-democratic influence within this large and very heterogeneous category is concentrated mainly among two subsets: the first comprises salaried employees in the public and para-public sector; the second, only partially coinciding with the first, comprises the intellectual professions that increasingly occupy a privileged position in the social base of the left, on account of their cultural weight.62
2.The logic of this evolution tends to attach the salaried middle strata to the centre of the social-democratic coalition, although the great diversity of national situations dictates prudence on this score. The lines of force of the new social penetration by socialists confirm the coexistence within the same coalition of voters from two major social groups, which are internally heterogeneous. Because the value priorities of these two central components of the social-democratic coalition differ sharply (the well-educated middle strata evince a strong adhesion to post-materialist values and cultural liberalism, workers a rather weak adhesion, if not a certain cultural conservatism), an evident dissymmetry opens up at the heart of the social-democratic electorate. Potential tension between these two universes, which are culturally distinct and electorally ‘allied’, is an important source of weakness for parties of a social-democratic type.
3.The emergence and consolidation of the divide between materialism and post-materialism, or (according to Walter Korpi) ‘ecology’ and ‘technology’, contributes to reducing the salaried middle strata’s propensity to vote social-democratic. The influence of the ‘green’ parties or certain left-wing socialist parties (Denmark, Norway) tends to check social-democratic penetration among these categories. Everything indicates that this new line of division did split the social-democratic coalition of the 1950s and 1960s, making it possible to detach from it a politically significant proportion of educated sections of the population and youth. As a result of this divide, social democracy does not always fully benefit from the culturally post-materialist segment of the electorate, whose epicentre is often located in the public sector, a segment essentially composed of youth and white-collar employees.
4.Meanwhile, notwithstanding marked national disparities, a number of indicators reveal that social democracy tends to become, par excellence, the representative of social forces most attached to the state, and especially the welfare state, despite differences of emphasis and interest. These forces are obviously workers, who remain orientated towards the state and its redistributive functions, and salaried public-sector employees, whose interests are inextricably bound up with the fate of the state. Accordingly, socialists find themselves confronting a paradoxical and quasi-schizophrenic situation: at a time when they are adopting policies of neoliberal inspiration for reducing the role of the state, they simultaneously present themselves as the political actor representing the social forces connected with the state or demanding its intervention. This relation and contradiction – a source of both socialist strength and weakness – renders adoption of a ‘pure’ neoliberal strategy too costly in electoral terms.
5.The working class is no longer the privileged sociological marker of social-democratic electorates. Their internal socio-demographic balance has shifted in the direction of a relative equalization of the working-class and ‘salaried-middle-strata’ components. As a result, the class specificity of these electorates, which historically confirmed the predominance of the working class, is being dissolved in a socially heterogeneous body. This drastically diminishes the social democrats’ ability to adopt coherent class policies, since it inevitably increases the potential electoral cost of such policies. It also enables us to appreciate how much more difficult dealing with political options becomes with such diverse clienteles.
6.Compared with their own past, contemporary social-democratic parties are constructed on the basis of a profoundly interclassist format, by far the most interclassist in the whole history of social democracy. Their ‘natural’ sociological inclination is their definitive mutation into ‘leftist catch-all parties’. At the present time, this is confirmed in terms both of the social composition of their electoral base and of their ideological/programmatic orientation. Moreover, such a pattern partially invalidates Przeworski and Sprague’s hypothesis according to which social-democratic parties tend to abandon interclassist strategies when they lose votes among the working class.63 To be sure, ‘perfect’ catch-all parties, large formations that are able to attract ‘everything’ in the strict sense of the term, are virtually unknown in Europe. The sociology of the great left-wing parties remains ‘a preferential sociology’.64 The social-democratic parties cover a broad social spectrum, but none of them ‘catches all’ the positions in this space in some undifferentiated, homogeneous fashion. None has been able to penetrate in depth the central core of the opposing electorate (large bourgeoisie, traditional petty bourgeoisie, liberal professions, in part the rural world), just as no conservative party has been able profoundly and enduringly to penetrate the central kernel of the left-wing electorate. In this sense, contemporary social-democratic interclassism is incomplete and, most probably, incapable of being completed. During the initial postwar period, however, social-democratic interclassism was considerably more limited, and to a certain extent was even a ‘projected’, rather than an ‘accomplished’, interclassism (see Chapter 3). The discrepancy between the ‘programmatic’ and the ‘sociological’ is significantly less marked today. In a dual, complementary dynamic, the decline in class voting among workers and the increasingly left-wing orientation of salaried non-workers lead to this result.
7.The relative disaffection of workers and popular strata with social democracy is directly linked to its inability to ‘pilot’ the socioeconomic system and implement its commitment to the principle of ‘rectifying’ inequalities. Social-democratic formations unquestionably remain popular electoral forces, but their capacity to represent the working-class and deprived sections of the population has diminished significantly. As a result, the first group of social-democratic-type parties identified in this study (see Chapter 3) – those with maximum working-class penetration (systematically approaching or surpassing two-thirds of the working-class vote) – is in the process of disappearing from the electoral map. The SAP, party of the working class par excellence, certainly remains very strong among manual workers, despite some falling off. Yet its constituency has manifestly been weakened, losing the stability and imposing density of the past. Today, the majority of social-democratic formations belong to an intermediate group whose main feature is the fluctuation of their working-class influence around the symbolic threshold of 50 per cent of the working-class vote (between 40 per cent and 60 per cent). The SPÖ, the SPD, the DNA, and the British Labour Party now belong to this category. The Danish SD, formerly very strong, appears to be doomed to languish below the 50 per cent bar. From the perspective of electoral sociology, it is no longer a valid generalization to say that social-democratic parties are parties of the working class.
8.The major phenomenon of the last thirty years is the advance – or greater resilience – of the left among non-manual salaried employees. ‘Pools of attraction’ have been constituted around the great socialist/social-democratic parties containing social categories that are less and less the ‘natural’ social categories of the left. The advance is considerable enough to compensate for the retreat of the left among workers, as well as the electoral effects of the reduction of the working-class group in the population. The scale of the consequences of this historically unprecedented development could be fully appreciated only towards the end of the 1990s when – despite the decline of the traditional working class – socialist parties dominated the European political scene.
9.‘Labour is nowadays the “natural” party of the minority social class in Britain,’ Anthony Heath et al. have written, ‘but that was the cross the Conservatives bore until well into the 1960s and it never stopped them from regularly winning elections.’65 This is true not only of the British Labour Party, but of all the parties in the European socialist/social-democratic family. It is true on one condition: that we avoid the error of imagining that the ‘nature’ of a political formation can be constituted independently of the social bases that give it electoral ‘materiality’, dynamism and health. If the overall economy of the social bases of social-democratic parties is in the process of changing, then – even had social-democratic working-class penetration remained unaltered (which is not the case) – the ‘nature’ of these parties, regarded as the ‘natural’ parties of the working class, cannot fail to be affected. Even during the 1990s, parties of a social-democratic type were able to preserve their privileged relationship with the working class (as is indicated by their consistent overrepresentation within this category). But the lines of force of their overall social penetration have been significantly – one could say structurally – disrupted. Neither class parties nor, at times, parties of the class, social-democratic formations have changed. From enlarged coalitions of the working class they have been transformed into interclassist coalitions with a strong working-class influence. And the difference is not insignificant.
Compared with the beginning of the twentieth century and the interwar period, social democracy in the 1950s and 1960s changed considerably in adopting a catch-all ideological and programmatic profile. But despite increased penetration among the middle classes, the sociological architectonic of its electoral space still remained massively structured by the working class. Significant as it was, the social-democratic mutation of the 1950s and 1960s was not a metamorphosis. Today, this architectonic finds itself altered and, with it, the very mould of social-democracy-in-the-electorate. Whether ‘left-’ or ‘right-wing’, the new social democracy is resolutely different from ‘old-style’ social democracy in terms of electoral sociology.
Notes
1.Colin Railings and Michael Thrasher, ‘Old election certainties buried in the avalanche’, The Sunday Times, 4 May 1997.
2.Anthony Heath, Roger Jowell and John Curtice, How Britain Votes, Pergamon Press, London 1985.
3.Ibid., pp. 15–16.
4.Anthony Heath, Roger Jowell and John Curtice, ‘Trendless Fluctuation: A Reply to Crewe’, Political Studies, vol. XXXV, no. 2, 1987, p. 257.
5.Patrick Dunleavy, ‘Class Dealignment in Britain Revisited’, West European Politics, vol. 10, no. 3, 1987, p. 415.
6.See ibid., pp. 415–17.
7.Ivor Crewe, ‘On the Death and Resurrection of Class Voting: Some Comments on How Britain Votes’, Political Studies, vol. XXXIV, no. 4, 1986, p. 637.
8.Perhaps here we should examine the pertinence of the notion, firmly established among specialist political scientists and sociologists, that junior white-collar workers form part of the ‘natural’ electorate of conservatism. While this group is very clearly distinguished from the working class in the strict sense, it is yet more distinct from the world of senior management and the traditional large, medium and petty bourgeoisie. Thus, to consider the political orientation of a rather ‘undecided’ group, over which left and right fight in a majority of countries, as ‘naturally’ conservative, is to remain ‘ensnared’ in an Anglocentric vision of class alignment. (In Great Britain, the lower ranks of non-manuals have traditionally remained on the side of the Conservatives.) It is also to remain ‘ensnared’ in an outmoded conception of class voting that is at odds with reality, judging the present on the basis of criteria fashioned in and for the past.
9.Obviously, in this work I have tackled only one aspect of the approach of Heath and his colleagues, neglecting other, highly innovative aspects of their work. As a general rule, academic discussion of class voting is focused on overall levels of class voting, and the electoral behaviour of each class (class-specific trends) is often not studied separately. This results in a number of misunderstandings.
10.Gerassimos Moschonas, La Gauche française (1972–1988) à la lumière du paradigme social-démocrate. Partis de coalition et coalitions de partis dans la compétition électorale, doctoral thesis, University of Paris II, 1990, pp. 162–4.
11.Ola Listhaug, ‘The Decline of Class Voting’, in Kaare Strom and Lars Svasand, eds, Challenges to Political Parties: The Case of Norway, University of Michigan Press, Michigan 1997, p. 81.
12.Ibid.
13.Russell Dalton and Wilhelm Bürklin, ‘The German Party System and the Future’, in Dalton, ed., The New Germany Votes, Berg, Oxford and Providence, RI 1993, p. 244.
14.Russell Dalton and Wilhelm Bürklin, ‘The Two German Electorates’, in Dalton and Bürklin, eds, Germans Divided, Berg, Oxford 1996, pp. 186–7.
15.These estimates concern the whole of Germany (east and west). See Der Spiegel, October 1998, p. 34; Die Zeit, 1 October 1998, no. 41.
16.Marit Hoel and Oddbjorn Knutsen, ‘Social Class, Gender, and Sector Employment as Political Cleavages in Scandinavia’, Acta Sociologica, vol. 32, no. 2, 1989.
17.David Arter, ‘Sweden: A Mild Case of “Electoral Instability Syndrome”?’, in David Broughton and Mark Donovan, eds, Changing Party Systems in Western Europe, Pinter, London and New York 1999, p. 143.
18.See ibid., p. 153.
19.Richard Kurt Luther, ‘Austria: From Moderate to Polarized Pluralism?’, in Broughton and Donovan, eds, Changing Party Systems in Western Europe, p. 133.
20.Christoph Kunkel and Jonas Pontusson, ‘Corporatism versus Social Democracy: Divergent Fortunes of the Austrian and Swedish Labour Movements’, West European Politics, vol. 21, no. 2, 1998, pp. 20–28.
21.Moschonas, La Gauche française …, pp. 215–18.
22.Daniel Boy and Nonna Mayer, ‘Que reste-t-il des variables lourdes?’, in Boy and Mayer, eds, L’électeur a ses raisons, Presses de Sciences Politiques, Paris 1997, p. 113.
23.Ibid., p. 114.
24.See Gérard Grunberg and Étienne Schweisguth, ‘Recompositions idéologiques’ and ‘Vers une tripartition de l’espace politique’, both in Boy and Mayer, eds, L’électeur a ses raisons.
25.Erik Olin Wright and Donmoon Cho, ‘State Employment, Class Location and Ideological Orientation: A Comparative Analysis of the United States and Sweden’, Politics and Society, vol. 20, no. 2, 1992.
26.Daniel Boy and Nonna Mayer, ‘Secteur public contre secteur privé: un nouveau conflit de classe?’, in Mayer, ed., Les Modèles explicatifs du vote, L’Harmattan, Paris 1997, p. 117.
27.Like all my other conclusions, this derives from calculation of the electoral statistics. However, the accumulation of corroborating indices reveals that a percentage of the upper middle classes and large capital is drawing ever closer to contemporary social democracy. There is no reason to suppose that these indices are insignificant, but they cannot be dealt with here.
28.Paul Nieuwbeerta, ‘The Democratic Class Struggle in Postwar Societies: Class Voting in Twenty Countries, 1945–1990’, Acta Sociologica, vol. 39, no. 4, 1996, p. 368.
29.David Sanders, ‘Voting and the Electorate’, in Patrick Dunleavy et al., eds, Developments in British Politics 5, Macmillan, London 1997, p. 56.
30.Kaare Strom and Lars Svasand, ‘Political Parties in Norway: Facing the Challenges of a New Society’, in Strom and Svasand, eds, Challenges to Political Parties, p. 23.
31.Quoted in Robert Ponceyri, Gaullisme électoral et Ve République, Doctorat d’État, IEP, Paris 1984, p. 539.
32.André Philip, ‘La pensée politique des partis ouvriers’, in Léo Hamon, ed., Les nouveaux comportements politiques de la classe ouvrière, Presses Universitaires de France, Paris 1959, pp. 205–6.
33.Bruno Cautrès and Anthony Heath, ‘Déclin du “vote de classe”? Une analyse comparative en France et en Grande-Bretagne’, Revue Internationale de Politique Comparée, vol. 3, no. 3, 1996, pp. 566–8.
34.Strom and Svasand, ‘Political Parties in Norway’, p. 24.
35.Alain Garrigou, ‘Conjoncture politique et vote’, in Daniel Gaxie, ed., Explication du vote, Presses de la Fondation Nationale des Sciences Politiques, Paris 1989, p. 359.
36.Lars Mjoset et al., ‘Norway: Changing the Model’, in Perry Anderson and Patrick Camiller, eds, Mapping the West European Left, Verso, London and New York 1994, pp. 62–70.
37.Gerassimos Moschonas, ‘The Panhellenic Socialist Movement’, in Robert Ladrech and Philippe Marlière, eds, Social-Democratic Parties in the European Union, Macmillan, London 1999.
38.Gerassimos Moschonas, ‘PASOK: From the “Non-Privileged” to the “Contentment Society”’, Epochi, 21 May 2000 (in Greek).
39.Dominique Pélassy, Sans foi ni loi? Essai sur le bouleversement des valeurs, Fayard, Paris 1995, p. 218.
40.Jorgen Goul Andersen and Tom Bjorklund, ‘Structural Changes and New Cleavages: The Progress Parties in Denmark and Norway’, Acta Sociologica, vol. 33, no. 3, 1990, p. 213.
41.Daniel-Louis Seiler and Romain Meltz, ‘Autriche’, in Guy Hermet et al., eds, Les Partis politiques en Europe de l’Ouest, Economica, Paris 1998, p. 65.
42.Pascal Perrineau, Le Symptôme Le Pen, Fayard, Paris 1997, p. 115.
43.Grunberg and Schweisguth, ‘Vers une tripartition de l’espace politique’, p. 187.
44.Lars Bille, ‘Danemark’, in Hermet et al., eds, Les Partis politiques en Europe de l’Ouest, p. 124.
45.Andersen and Bjorklund, ‘Structural Changes and New Cleavages’, pp. 214, 212.
46.Perrineau, Le Symptôme Le Pen, p. 225.
47.Emmanuel Todd, L’illusion économique. Essai sur la stagnation des sociétés développées, Gallimard, Paris 1998, pp. 283–5.
48.Rudy Andeweg, ‘Elite-Mass Linkages in Europe: Legitimacy Crisis or Party Crisis?’, in Jack Hayward, ed., Elitism, Populism and European Politics, Clarendon Press, Oxford 1996, p. 152.
49.Christopher Anderson, ‘Economics, Politics and Foreigners: Populist Party Support in Denmark and Norway’, Electoral Studies, vol. 15, no. 4, 1996, p. 501.
50.Herbert Kitschelt, The Radical Right in Western Europe, University of Michigan Press, Ann Arbor 1995, p. 277.
51.Andersen and Bjorklund, ‘Structural Changes and New Cleavages’, p. 214.
52.Grunberg and Schweisguth, ‘Recompositions idéologiques’, p. 148.
53.Ibid., p. 177. In France, the lack of a degree explains anti-universalism better than it explains voting for Le Pen (people with a low level of education are characterized by their anti-universalist values, even if they do not vote FN). On the other hand, unemployment explains voting for Le Pen better than it explains anti-universalism (Grunberg and Schweisguth, ‘Vers une tripartition de l’espace politique’, p. 205).
54.The split in the Front National in France in 1999 will doubtless test Jean-Marie Le Pen’s implantation in the working-class electorate. The Front National now risks appearing to voters as a party ‘like the others’, which will reduce its ability to attract the protest vote as well as the impact of its networks of cultural influence. A directly political factor, the split is of a sort to sign away the working-class future – perhaps the future tout court – of the French extreme right.
55.In France, as indicated above, the frequency of the working-class vote for the extreme right increases with the number of working-class attributes. Workers who vote most for Le Pen are better inserted into the working-class milieu and, as a result, have the same social profile as those who formerly voted for the PCF (Boy and Mayer, ‘Que reste-t-il des variables lourdes?’, p. 112). This being so, can the Le Pen vote of what is certainly a minority of workers be regarded as an expression of the decline in class voting? Or, perhaps, as a new – and surprising – form of class voting: as (in a sense) an inverted class vote? If a working-class right-wing vote has always existed, the ‘intense’ (as opposed to ‘diffuse’) support currently given by workers to the extreme right is probably the portent of a rare phenomenon in electoral history: a working-class right-wing vote that is not ‘deviant’, but a perverted class vote, a class vote that does not conceive itself as such.
56.Moschonas, ‘Quo vadis social-démocratie?’, in Michel Vakaloulis and Jean-Marie Vincent, eds, Marx après les marxismes II: Marx aufutur, L’Harmattan, Paris 1997, p. 310.
57.Ronald Cayrol, ‘L’électeur face aux enjeux économiques, sociaux et européens’, in Pascal Perrineau and Ysmal Colette, eds, Le Vote surprise: les élections législatives des 25 mai et ler juin 1997, Presses de Sciences Politiques, Paris 1998, p. 97.
58.Garrigou, ‘Conjoncture politique et vote’, p. 358.
59.See also Crewe, ‘On the Death and Resurrection of Class Voting’, p. 622.
60.See also Kitschelt, The Radical Right in Western Europe, p. 279.
61.Wolfgang Merkel, ‘Between Class and Catch-all: Is There an Electoral Dilemma for Social Democratic Parties in Western Europe?’, in Merkel et al., Socialist Parties in Europe II: Class, Popular, Catch-all?, ICPS, Barcelona 1992, p. 25.
62.Some examples can illustrate the political heterogeneity of the salaried middle strata. Professions located in the ‘repressive’ apparatus of the state (army, police) typically vote for the right, despite belonging to the public sector. The category of ‘welfare and creative professionals’ inclines markedly to the left, despite the fact that a proportion of this category belongs to the private sector. In addition, teachers and lecturers are, in their large majority, completely behind the left in France, whereas teachers in Great Britain vote Conservative and higher education lecturers vote Labour in greater numbers (Boy and Mayer, ‘Secteur public contre secteur privé’; Anthony Heath and Mike Savage, ‘Middle-class Politics’, in Roger Jowell et al., eds, British Social Attitudes: The Eleventh Report, Dartmouth, Aldershot 1994).
63.Merkel, ‘Between Class and Catch-all’, p. 26.
64.Alain Bergounioux, in Bergounioux and Marc Lazar, La Social-démocratie dans l’Union européenne, debate in Les Notes de la Fondation Jean-Jaurès, no. 6, Paris 1997, p. 10.
65.Anthony Heath, Roger Jowell and John Curtice, ‘Can Labour Win?’, in Heath et al., eds, Labour’s Last Chance, Dartmouth, Aldershot 1994, p. 284.