Political learning in adolescence: the development of party preferences in a multiparty setting.

Although contemporary studies on voting behavior generally treat citizens as individual actors, the reality of political choice is a inevitably a social reality. Party preferences are not formed in isolation, and the immediate social network of the parental family has traditionally been the primary setting in which these preferences are expected to be developed, during the pre-political phase of adolescence. This was at least the general expectation of the first generation of US based scholars developing a first model of political learning in the 1960s, 70s and 80s. Political learning in adolescence in a dealigned multiparty setting The times have changed since then, and the political behavior of citizens has undergone important changes as well. Over the past decades, we have observed a process of dealignment in which strong and affective ties between voters and political parties have loosened. This is illustrated by a rise in electoral volatility and a drop in party membership. This evolution is most clearly visible in European multiparty settings such as Belgium, the focus of this study. It is with this evolution in mind that I aim to reexamine the models of early political learning in the family within this Belgian multiparty context. This study fits in a broader trend of renewed attention for the revisited ‘social logic of politics’ in the past decade, that put the family back to the forefront as an important determinant of political preferences. An important advance in recent studies is that they have demonstrated the importance of acknowledging the individual sense of agency of adolescents. They are no longer treated as passive receivers of parental political stimuli. I attempt to build further on these recent studies and analyze both the continuing relevance of parental socialization, and the individual determinants of adolescents’ party preferences. Empirical approach The most important empirical source is the Parent-Child Socialization Study (2012-2013), a two-wave panel study conducted among more than 3,000 15- to 16-year-old adolescents and both their parents in Belgium. In a methodological note in this dissertation, we demonstrate the importance of surveying the family members separately. We simulate how some of the central analyses in this dissertation would have led us to different or even opposite conclusions if our data would be gathered among the adolescents only. To what extent do the traditional political learning mechanisms still apply? The first research goal was analyzing to what extent the traditional political learning mechanisms, (still) apply in a contemporary European multiparty setting. This research goal provided the central framework for the first four empirical chapters (Chapter 2-5). What is demonstrated in a first set of analyses is that the role of parents in the development of party preferences among adolescents is (still) very important. Even in this fragmented multiparty context, we observe an elevated level of parent-child correspondence in party preferences. On top of this high observed correspondence, adolescents strongly overestimate these similarities. It seems that they try to resemble their parents even more than they already do, underscoring the importance they attach to the political views of their parents. Political discussion in the family is obviously an important mechanism in explaining parent-child similarities, and particularly discussions with the father seem to increase parent-child congruence. Talking about politics with one’s father is found to have a stronger effect on father-child correspondence than discussing politics with one’s mother. In an attempt to map the influence that family members exert on one another – explaining these elevated similarities – I analyzed the different reciprocal (top-down, between parents, and trickle-up) patterns of influence within the family. The results of this analysis strongly point in the direction of a process of shared parental and mostly top-down influence. Mothers and fathers tend to contribute equally to the development of congruent party preferences among their children. Finally, parental socialization is not only found to influence support towards political parties, but also affects the development of explicit negative orientations towards other political parties. This is not only the case for polarizing parties, such as the radical right party, but also for more centrist, mainstream parties such as the Liberals or the Christian democrats. I conclude that in this way, parents play an important role in the development of a ‘consideration set’ of potential parties to choose from, determined by eliminating those who are out of bounds. The individual decision: consistency between attitudes and party preferences In a second part I focus on the extent to which the adolescents’ party preferences are in line with their other political and social orientations. If we want to acknowledge their individual agency, it is also useful to explore whether they can indeed come to a ‘logical decision’, based on their own preferences. A first study is based on longitudinal data following young Belgians from the age of 16 to the age of 21, and focuses on the rising success of the New Flemish Alliance (N-VA) among these young citizens. Results show that those who initially had a stronger Flemish identity indeed moved on to an intention to vote for this Flemish nationalist party later on. The young citizens displayed a strong consistency between their own attitudes and their party preference, and in a sense followed their own predispositions to come to this preference. In a second study, we find that this ability of adolescents to draw a substantive link between their own attitudes and a party preference is significantly enhanced by their level of political sophistication. Furthermore, it is important to mention that the level of this consistency among adolescents is still significantly lower than the same substantive links among the more experienced, adult voters in our study. The long-term stability of early political learning Finally, I addressed the stability of these initial preferences. The results are again in line with the traditional socialization models: preferences acquired through parental socialization tend to be more stable over time, and this parental correspondence even has a stronger effect on the stability of these preferences than the substantive consistency with the adolescents’ ideological predispositions. In sum, parental socialization remains an important starting point for the adolescents’ first encounter with party politics, but we need to be careful in underestimating the adolescents’ agency in the whole process. Yes, parents still matter and to a large extent influence the adolescents’ likelihood to support or (strongly) not support parties. The findings are quite robust, yet we do need to be careful in interpreting parent-child resemblance as purely a matter of socialization, as our results indicate that it could well be a matter of inspiration as well. In this interpretation, adolescents are not passive receivers of parental stimuli, but active learners trying to find their way in the political world, making use of the cues provided by their parents as a guideline to come to a party preference that ‘makes sense’. Only a small minority of the adolescents explicitly indicates that they would vote for a certain party simply because their parents do so, and that is of course an important qualification to these results. I conclude this dissertation with a call to action to incorporate the immediate social surrounding more widely in studies on electoral behavior. Just as we include socioeconomic status, religion, gender or ideology in our models explaining vote choice, we expect that one’s family background remains an inherent part of the explanation as well.nrpages: 304status: publishe