The Decline of Class Voting in Sweden 1968–2014: Reconsiderations, Explanations and the Role of the New Middle Class

Sammanfattning: Class voting has been one of the most wide-spread and persistent patterns of voting behavior in Western democracies throughout the 20th century. Understanding its decline is an important part of understanding the current political situation. This thesis contributes to this with four studies of class and politics at the mass level, employing the surveys of the Swedish National Election Studies program. The trend of decline in class voting has been interpreted in many different ways. Two major strands of interpretation can be summarized with the terms dealignment and realignment. The first holds that the ties between the socio-economic structure and the political parties are on their way towards general dissolution. The second holds that a new order is coming to replace the old one. The first two studies look into the central predictions or assumptions of a specific variant of the realignment interpretation, namely the new class schema by Swiss scholar Daniel Oesch, centered in a typology of work logics. The results show that the historical predictions of increasing explanatory power are not borne out. Indeed, the development in class voting according to the Oesch schema, is remarkably similar to the one we see in a traditional schema. In addition, the causal effects of work logics claimed by Oesch and others do not turn up when tested with panel data and a stronger identification strategy than earlier studies. Individuals that change occupation do not change their political attitudes in accordance with the theoretical expectations. The later studies engage in debates within the dealignment literature. The major argument in these studies is that the last generation of class research has put a considerable emphasis on the contemporary situation in society, and forgotten about generational replacement as a possible mechanism for political change. One of the studies shows that generational replacement better explains the historical decline of class voting in Sweden than does the often-assumed ideological convergence between the parties. The last article studies swing voters (i.e. voters that consider voting for either of the the major political blocs) and finds, contrary to several different theoretical expectations, that this group does not consist of voters from the middle class to any disproportionate extent. The increase in swing voters is thus not a function of an increasing share of middle class voters without clear political alignments, but rather something that happens throughout the class structure.