The Class Basis of Social Democracy at the Beginning of the Twenty-First Century

This chapter is dedicated to an empirical analysis of the class basis of social democracy at the beginning of the twenty-first century. The analyses presented rely on data from the European Social Survey 2010–2015. They show a strong tendency of social democratic parties towards becoming cross-class parties, mobilising no specific social class in particular. Despite the widespread assumption, only a minority of social democratic parties have become new class parties strongly mobilising specific segments of the salaried middle classes. The chapter also demonstrates that the working-class vote has become much more fragmented. In this group, social democracy now competes with radical left parties, populist radical right parties and abstention.

step before analysing their party choice and drawing conclusions on the class profile of social democracy. When abstention is high, the decision to support social democracy is made among a relatively smaller group of citizens. The most important decision for an increasing share of workers in this context simply becomes the choice between participating and not participating.
The analyses presented in this chapter focus on the beginning of the twenty-first century. They are based on data from the European Social Survey (ESS), which offers detailed information about the respondents' occupations. More precisely, the analyses focus on the first five years of the 2010s decade for pragmatic reasons of data availability. 1 For each country, I first list all the elections that took place during this time period (this means the 2012 French election, the 2013 Austrian and German elections, the 2010 and 2012 Dutch elections, the 2011 and 2015 Swiss elections and the 2010 and 2015 British elections) and then select the ESS survey round closest to the election. 2 Of course, the results are to some extent influenced by the specificities of these particular elections and the electoral performance of Social Democrats and their competitors in these elections. However, apart from some specific features, all these elections are representative of a new environment in which social democracy faces new competitors in the party systems and copes with low turnouts. In all these elections, social democracy displays a weak electoral performance (see Appendix)-only the French Socialist Party managed to secure a relatively good result in the 2012 election, before a strong electoral defeat in 2017. The electoral decline of social democracy has been particularly strong since 2008 (see Bremer and Rennwald 2019), accelerating a continual decline since electoral peaks in the 1960s and 1970s (see Rennwald and Pontusson 2020). Table 4.1 displays participation percentages in elections by social class in the period between 2010 and 2015. The electoral participation of workers does not go beyond 76%. This value-for Germany-is the highest in our sample. In contrast, in three countries, the participation of managers is higher than 90%. In four countries, this is also the case for the category of large employers and the self-employed. In the 1970s, the participation of workers reached more than 90% in two countries. At that time, workers were almost as likely to participate in elections as the average citizen in several countries. For example, in Austria, worker  5-6 (2010-2012) for the Netherlands, round 6 (2012) for France, round 7 (2014) for Austria and Germany, rounds 6 (2012) and 8 (2016) for Britain and Switzerland. The data are weighted by design weights Names of classes: Prod = production workers, Serv = service workers, Clerk = clerks, Socio = socio-cultural professionals, Tech = technical professionals, Manag = managers, Lar/self = large employers and self-employed professionals, Small = small business owners participation almost reached the average overall participation rate (a ratio of 0.99). In contrast, in the 2010s, the ratio of worker electoral participation to the overall average in Austria was 0.92. Indeed, in all countries, this ratio became smaller: from 0.92 to 0.86 in Great Britain, from 0.96 to 0.91 in Germany, from 1 to 0.88 in France, from 0.90 to 0.88 in the Netherlands and from 0.89 to 0.82 in Switzerland. In contrast, in many countries, participation by the group of large employers and self-employed professionals increased relatively to the average: from 1 to 1.16 in Austria, from 0.98 to 1.14 in Great Britain, from 1.08 to 1.10 in Germany, from 1.07 to 1.25 in France and from 1.11 to 1.14 in the Netherlands. Inequalities in participation therefore clearly increased over time 3 and social democratic parties were less able to bring workers to the ballot box in the 2010s.
Therefore, workers participate in elections much less than in the past. To what extent can social democratic political parties benefit from the support of this smaller group of workers who effectively participate in elections? Figure 4.1 presents the shares of votes obtained by social democratic parties from different social classes in 2010-2015, with parties' average scores on the vertical axis. 4 In the 2010s, production workers were only the strongest supporters of social democracy/labour in Austria and Great Britain. In absolute terms, social democratic parties received   Table 4.1 for sources and names of classes around 40% of production workers' votes in Austria, Great Britain and France, 30% in Germany and less than a quarter in the Netherlands and Switzerland. Clearly, average support for social democracy/labour generally decreased since the 1970s, but even in relative terms production worker support for social democracy was not greatly different to average support. In relative terms, production worker support for social democracy exceeded average support by a factor of 1.31 in Austria (against 1.28 in the 1970s), 1.20 in Great Britain (1.32), 1.13 in the Netherlands (1.30), 1.11 in Germany (1.36), 1.07 in France (1.13) and 1.05 in Switzerland (1.44). Relative support for social democracy therefore shrank in all countries except Austria. 5 Among those fewer workers who took part in electoral contests, there was still an over-representation of social democratic voting, albeit of much weaker intensity (see also Knutsen 2006). The same is also true for service workers-a category that displayed a consistently high level of support for social democracy in the 1970s. They too became less distinct supporters of social democracy over time. Generally, differences across classes in their degree of sympathy for social democracy were generally smaller in the 2010s than in the 1970s. The case of small business owners illustrates this well. While this group continued to remain the least likely to support social democracy, its aversion to this party family reduced in almost all countries. 6 Already in the 1970s, several classes displayed intermediate support for social democracy. Clerks, socio-cultural professionals and technical professionals continued to present intermediate to above-average support for social democracy in the 2010s, with some variation across countries. Managers, who were in some countries relatively strong opponents of social democracy in the 1970s, displayed intermediate support for social democracy in the 2010s. However, none of the classes presented particularly strong above-average support for social democracy, at least in Austria, Great Britain, Germany and the Netherlands. When the distinctiveness of service and production workers in their support for social democracy weakened, no other class took a prominent role and served as the leading social democrat supporters. Social democratic parties were therefore more likely to have become cross-class parties with an intermediate level of support among various classes, including among the former strongholds of production and service workers.
France and Switzerland represent exceptions to this general pattern. In these countries, a specific segment of the salaried middle classes displayed a strong affinity with the socialist parties in the 2010s. In both countries, socio-cultural professionals were the strongest supporters of socialist parties. The support from this class exceeded average support by a factor of 1.53 in Switzerland and 1.48 in France. They clearly outdistanced production workers, who displayed only average support for the French Socialist Party and the Swiss Social Democratic Party. In these two cases, socialist/social democratic parties therefore came closer to being new class parties.
In the recent period, production workers did not make a massive contribution to social democracy's electoral scores. As for the 1970s, Table 4.2 indicates the class composition of the social democratic electorate. It also reports the class composition of the total electorate, which includes non-voters here. Not only was production workers' support for social democracy not very different to that of the average voter (see above), but the importance of workers in the electorate also dramatically reduced. If we only consider production workers, their contribution to social democracy's electoral results was therefore relatively limited. In five of the six countries, production workers represented less than 20% of the social democracy electorate. In relative terms too, social democratic parties were no more working class than the total electorate. One must not forget that the widespread abstention among workers-which is taken into account in this analysis-contributed to depressing the working-class character of social democracy. However, Austria represents an exception to this pattern, with production workers still constituting a quarter of the social democratic electorate. Notes See the information under Table 4.1 for sources, names of classes and numbers of cases (see also Fig. 4.1). The total electorate includes non-voters and therefore refers to the population entitled to vote If we consider service workers as part of the working class, the working-class contribution to social democracy's electoral score becomes more important. Interestingly, it approximates to the scores found in the 1970s for production workers. In the 2010s, production and service workers represented almost half of the SPÖ electorate, around 40% of the Labour and the Socialist Party electorates, a third of the PvdA and the SPD electorates, but only a quarter of the SPS electorate in Switzerland. Hence, the definition of 'working class' is crucial in any evaluation of the contemporary class character of social democracy. Incorporating the new service proletariat into a broad 'middle-class' category would clearly change the conclusions of such an exercise and tilt social democratic parties towards being middle class. However, it should again be noticed that the importance of the enlarged working-class electorate to social democracy is more the product of the class structure-production and service workers together represent non-negligible segments of the electorate-than the product of a strong over-representation of workers among voters for this party family. We noticed before the prominent level of support by socio-cultural professionals in Switzerland and France. However, their contribution to the electoral score of social democracy remained relatively modest, as is shown in Table 4.2. They represented a quarter of the Swiss Social Democratic Party's electorate and a fifth of the French Socialist Party's. In relative terms, this is already a significant contribution since sociocultural professionals represented only 11% of the French electorate and 15% of the Swiss electorate. This segment of the new middle classes was simply too small to have the same importance in the social democratic electorate as production workers in the past: they never exceeded 16% of the total electorate in all countries. While the analysis above of the distinctiveness of classes pointed in the direction of new class parties in France and Switzerland, this second indicator of class contribution puts this conclusion into perspective.
In a few countries, managers made up a sizeable proportion of the social democratic electorate in the 2010s (up to one in five social democratic voters). The reason is that in almost all the countries they represented an important group in the total electorate, even larger than the group of socio-cultural professionals in most of the countries. However, this class has no particular affinity with social democracy. Its sizeable contribution to social democracy's electoral score was therefore more a product of the class structure. Hence, the social democratic electorate was to some extent a mirror of the composition of the total electorate. This again suggests a proximity to a cross-class type.

The New Fragmentation of the Working-Class Vote
Were there other parties which competed for the working-class vote? As the previous chapter showed, in the 1970s production workers constituted the party preserve of social democracy in most countries. Hence, social democracy enjoyed a sort of monopoly over the representation of production workers. The dominance of the socialist/Labour Party was only contested in France and the Netherlands, by the Communist Party in France and by the Christian Democratic Party in the Netherlands. In the 2010s, competition for the working-class vote was fiercer, since several parties received above-average electoral support from production workers. Table 4.3 lists the parties which were in competition with social democracy for the working-class vote in the 2010s. It only includes parties that received above-average support from production workers. The table indicates the parties' vote shares among production workers next to the parties' names. It is important to remember here that radical right voters are generally under-represented in surveys, and the ESS is no exception to this. Interestingly, only parties from the radical right and the radical left 7 enjoyed above-average electoral support from production workers, after social democracy. This was never the case for mainstream right parties or Green parties. In three of the countries (Austria, Great Britain and Switzerland), social democratic parties were in competition with the radical right. In the other three countries (Germany, France and the Netherlands), social democratic parties competed with both the radical right and the radical left. Therefore, in some instances, the working class was the contested stronghold of social democracy and the radical right, while in others there was open competition among three parties for the working-class vote. Social democracy's monopoly over the working-class vote had therefore clearly ended. Relative electoral support by production workers was generally more pronounced in the case of radical right parties than in the case of radical left parties. This means that production workers differed more from the average voter in their support for radical right parties than in their support for radical left parties. The ratios are particularly high for the National Front (FN) in France and the Party for Freedom (PVV) in the Netherlands. Production workers' support for these parties was twice as strong as that of the total electorate. In the elections analysed in this table, the Alternative for Germany (AfD) was only at the beginning of its electoral progression (it later gathered 12.6% of the votes in 2017). However, it already received more than average support from production workers.
Of course, one should remember that, being large parties, mainstream right parties often gather a sizeable share of production workers' votes in absolute terms. By extension, they are also in competition for the working-class vote. This was especially the case in Germany, where in this period the Conservatives (CDU) obtained 37% of production workers' votes, against an average of 38% from the overall electorate. The results are less impressive in the other countries, but the British Conservatives still received 28% of production workers' votes (against 37% on average) and the Austrian People's Party (ÖVP) obtained 20% (against 27% on average). In the other cases, the share of production workers' votes going to mainstream parties of the right was more limited (between 10 and 16%). However, in none of the countries studied did mainstream right parties received above-average electoral support from production workers, not even in the Netherlands, where the Christian Democrats (CDA) competed most clearly for the working-class vote in the 1970s.
As it became a smaller party, the CDA increased its 'bourgeois' character. By contrast, in Austria, Great Britain and Germany, production workers became less reluctant to support the mainstream right in the 2010s compared to the 1970s. The most important change concerns the CDU, whose level of support among production workers was almost as strong as that among the entire electorate. 8 The new working-class profile of radical right parties has attracted much attention in recent years. Therefore, our results can now be complemented by turning to the class composition of radical right parties. It is not an entire surprise that workers make up a sizeable proportion of the radical right electorate. If we again count production and service workers together, they represented almost half (48%) of the PVV's electorate in the Netherlands (but only 34 and 37% of those of the Dutch Labour Party and the Socialist Party, respectively), over half (57%) of the FN's electorate in France (but only 39% of the French Socialists' and 38% of the Front de Gauche's) and much more than half (62%) of the Freedom Party of Austria (FPÖ) (but only 49% of the Austrian Social Democrats'). The share of the working class in the United Kingdom Independence Party (UKIP) electorate was slightly lower (44% against 41% of the Labour Party's). This was also the case of AfD (42% against 33% of the German Social Democrats' and 30% of Die Linke's). The Swiss People's Party's (SVP) share (33% against 22% for the Swiss Social Democrats) was clearly smaller in Switzerland, where farmers and small business owners-the traditional constituency of the former agrarian party-still constitute an important electoral base (e.g. Rennwald 2014). Undeniably, at the aggregate level, social democracy and radical right parties share commonalities in the class profile of their supporters. However, recognising this does not mean that there were direct transfers of votes between the radical right and social democracy. Studies show that the numbers of direct transfers were relatively limited. In an article analysing the electoral decline of social democracy, Rennwald and Pontusson (2020) show that only 8% of former social democratic party voters switched to the radical right in the next election in the period 2001-2015 in sixteen industrialised countries. The most common option was to choose mainstream parties of the right (45%), followed by the radical left (16%), abstention (16%) and the Greens (15%). In a detailed study of voter transitions in the British context, Evans and Mellon (2016) find that in 2015 UKIP primarily mobilised 2010 Conservative voters, obtaining its strongest result in all general elections with 12.6% of the vote. However, these same conservative voters were former non-voters or 2005 Labour voters. In a time of high electoral volatility, indirect transitions from social democracy to the radical right, transitioning via the centre-right or abstention, are well plausible.
Interestingly, many studies suggest that social democratic parties and radical right parties recruit from different social milieus. Radical right parties draw strong support from non-unionised voters, while social democracy still performs better among unionised voters (Mosimann et al. 2019). Workers who vote for social democracy do so based on economic considerations, while those supporting radical right parties do so for cultural reasons (Bornschier and Kriesi 2013;Oesch and Rennwald 2018). Instead of emphasising (indirect) transitions from social democracy to the radical right, one can therefore adopt a different focus and emphasise the existence of different working-class milieus (see also Vester 2001). From this perspective, social democracy fails to mobilise a more 'leftist' and 'conscious' working-class milieu (which prefers abstention or radical left parties), while radical right parties succeed in mobilising a non-organised and more 'rightist' working-class milieu. This analysis challenges a widespread perception of the presence of a common pool of supporters who switch from the left to the radical right.
Finally, it should also be emphasised that changes in workers' voting patterns are slow in coming and take place over several elections (e.g. Rennwald and Evans 2014). The workers who nowadays vote for the radical right are simply not the same persons as those who voted for parties of the left a few decades ago. Therefore, generational replacement may be another and complementary way to interpret the new competition from the radical right (Gougou and Mayer 2013;Gougou 2012).
It is not possible to give a definitive response about transitions of the working-class vote at the individual level in the framework of this book. However, we can deliver a precise account of the fragmentation of the working-class vote at the aggregate level. In order to get a global view of workers' votes today, it is possible to combine information on abstention and party choice. 9 Fig. 4.2 presents the complete set of options available, first for production workers only and second for the average citizen (indicated as 'total' in the figure). The information about party choice is restricted to the largest parties competing in the given country. These parties are classified in the usual European party families, while smaller parties are included in the 'other' category. Again, the vote shares  for populist radical right parties are smaller than their actual proportion according to official election results. Figure 4.2 reminds us of the importance of not voting among the options available to production workers. While we analysed not voting separately at the beginning of this chapter, this direct comparison shows the scope of workers not voting very well. Abstention went from a quarter of production workers (in Germany) to a small half (in Switzerland). With the exception of Germany, it was production workers' first choice. Among the options available to those who effectively voted, social democracy was still the most common one in many countries-although it never exceeded a third of all options. It should be noticed that among those who effectively went to the ballot box in Germany the first option for production workers was a conservative vote (even before abstention), while in Switzerland it was a vote for the Swiss People's Party. These results suggest that in all the countries, social democracy competed for the workers' vote on many fronts. Choosing to vote for social democracy therefore represented just one possible option among many. Social democracy's monopoly over production workers' votes is definitely a story of the past.

Mobilising the Working Class and Allied Classes
Let us finally make a systematic comparison of the mobilisation of production workers and allied classes at the two time points that we have analysed in this and previous chapters. For this purpose, we can use the indicators of distinctiveness (whether workers voted more than the average for social democracy) and the absolute contribution (the proportion of workers in the social democratic electorate). It is possible to group the values of these indicators by distinguishing between strong, medium and weak mobilisations (for more details, see the note below Table 4.4). Of course, the thresholds are always somewhat arbitrary but they allow a synthetic view of changes over time. Notes The first column indicates the distinctiveness of support by production workers, the second column shows the contribution of production workers to the party's electoral score and the third provides the number of parties with whom social democracy competes for the production workers' vote. For distinctiveness, the values are grouped in four categories: > 1.30 = strong, 1.15-1.29 = medium, 1-1.14 = weak, < 0.99 = absent. For the contribution, > 45% = strong, 25-44% = medium, 0-24% = weak. The 'strong' category is highlighted in bold As Table 4.4 shows, the distinctiveness and contribution indicators show a relatively large mobilisation of production workers in the 1970s. This helped social democracy to beat the competition from other parties-the third column indicates an absence of competition in many countries. Social democracy could therefore enjoy a kind of monopoly over the working-class vote. Due to the importance of the size of the working class among the population eligible to vote, achieving some degree of mobilisation among production workers was the key to relative success for social democracy at this time.
In the more recent period, social democracy was less able to mobilise production workers' votes, as is indicated by the distinctiveness and contribution indicators. It therefore faced greater competition for workers' votes. With the exceptions of Austria and Great Britain, there were simultaneously weak distinctiveness and a weak contribution of production workers.
Allied classes are also important for social democracy. In the 1970s, mobilisation among the lower white-collar classes contributed strongly to social democracy's electoral results. In the 2010s, socio-cultural professionals had a significant new role for social democracy in some countries. It is therefore interesting to understand the relationship of social democracy to allied classes by comparing the mobilisation of these two classes. Table 4.5 shows the mobilisation of service workers in the 1970s and that of socio-cultural professionals in the 2010s.
In terms of contribution, the positions of service workers in the 1970s and socio-cultural professionals in the 2010s are relatively similar. Their contribution was weak in most cases-it did not go beyond a 20% threshold. The only exception is Switzerland in the 2010s, where the contribution of socio-cultural professionals reached a medium level. For Notes For explanations and legends, see Table 4.4 the indicator of distinctiveness, there was greater divergence across countries in the 2010s than in the 1970s. The mobilisation of service workers was weak to medium in most of the countries in the 1970s. In the 2010s, the mobilisation of socio-cultural professionals was either absent or weak in three countries and strong in two other countries. Overall, in the 2010s, some social democratic parties combined a weak mobilisation of production workers with a strong mobilisation of socio-cultural professionals. This was the case of the Swiss Social Democratic Party and the French Socialist Party. In these cases, the direction of modernisation adopted by these parties was towards being new class parties. To some extent, this was also the road of the Dutch Labour Party, which showed a medium mobilisation of socio-cultural professionals. In the other cases, there was a combination of weak to medium mobilisation of production workers and no mobilisation of socio-cultural professionals. This suggests that social democracy took the direction of becoming a cross-class party. Perhaps one should add an unsuccessful cross-class party. The original intuition of enlarging its support to various classes included the idea that this would be a new 'winning formula' for social democracy.

Summing Up
This chapter has shown a growing distance between production workers and social democratic parties in the recent period. Not only were production and service workers less likely to participate in elections in the 2010s but when they did they were also less likely to support social democracy than in the past. It is striking to observe a rapprochement between production workers and other classes in their levels of support for social democracy.
In the 2010s, social democratic parties more strictly represented the class composition of the population eligible to vote than in the past. Certainly, given the importance of the (new) middle classes in social structures, it is logical for social democratic parties to also become more middle class. However, there is not a particular affinity, as expressed by the indicator of distinctiveness, between the new middle classes and social democracy, although Switzerland and France are exceptions. It is therefore more accurate to think of social democratic parties as evolving in the direction of cross-class parties than as simply middle-class parties.
In the 2010s, social democracy was in competition with many parties for the working-class vote. While the focus is often on the radical right in the public discourse, the new voting choices of the working class are much more complicated. Not only radical right parties but also radical left parties received above-average support from the working class, while abstention was also a very prominent choice among workers. Last but not least, parties from the mainstream right also captured a non-negligible share of the working-class vote, even if it was below average. All observations point in the same direction: the dominance of social democracy over the working-class vote is definitively over.
Notes 9. The number of cases in each country is slightly smaller than in Table 4.1.
Some individuals declare they participated in the election (in the information presented in Table 4.1) but then do not give any information on the party they chose (refusing to answer or responding 'don't know').
Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/ by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made. The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.