What Class, What Vote? Post-Fordist Social Groups, Class Coalitions and the Mainstream Left

Abstract

This article investigates the electoral behavior of post-Fordist social classes between the beginning of the 2000s and the end of the 2010s in eight Western countries, focusing particularly on the vote for the mainstream Left. More specifically, the work answers three research questions. How has social stratification changed in the countries analyzed? How have the post-Fordist social groups voted in the last two decades, and how has their support for the Left changed? What kinds of class coalitions supporting Left parties have now emerged, and how stable are they? The empirical analysis shows that the electoral behavior of post-Fordist social groups varies considerably according to the welfare regime considered. The comparative analysis also points out that Left-wing parties have taken different paths in response to the change of social stratification.

Keywords:

• Social Stratification
• post-fordist social groups
• class politics
• left parties

Notes

1 In contrast with Oesch’s original scheme, we included higher grade managers within the traditional bourgeoisie group. Despite their work logic being managerial, over time this group has followed an independent work logic, given its broad room for manoeuvre at the company level.

2 Regarding the other social classes, Oesch and Rennwald (2018) suggest that: a) the consensus of the technical (semi-) professionals and clerks is contested by all poles; b) associate managers keep voting for the centre-right; c) as regards the petty bourgeoisie, the deterioration of its status could convince this group to vote for the radical right.

3 For the USA, to make the sample more robust we aggregated ISSP 2002, 2003 and 2004 datasets - to analyse social class voting in the early 2000s - and ISSP 2014, 2015 and 2016 datasets - to detect social class voting in the late 2010s

4 For example, the first round of the ESS does not allow to reproduce the Oesch’s class scheme for France. For this reason, we decided to use the ISSP. Furthermore, at the time of writing, ESS did not provide updated data concerning electoral behavior for Denmark, Sweden, and Spain at the end of the 2010s, and therefore we opted to use the 2017 ISSP dataset.

5 Considering the early 2000s, for France and the USA this variable is not included in the ISSP. Therefore, we were obliged to rely on the variable “Party Affiliation”.

6 For France, the 2002 ISSP dataset does not include the “VOTE_LE” variable, and therefore French electoral turnout in the early 2000s cannot be shown.

7 We refer to survey data, which, in general tend to overestimate the electoral participation. The values therefore need to be treated with caution.