Predicting political involvement : the role of social identity and commitment to opinion-based groups

Abstract

This thesis examines the main predictors of political involvement from a social identity perspective, thus the main questions addressed is when people decide to take collective action in relation to shared ideas. It is argued that group selfdefinition should predict intentions to undertake political behaviours. Surprisingly the existing literature does not unequivocally show that the strength of subjective group membership is a good predictor of group behaviour in general, or political behavioural intentions in particular. The thesis proposes four key solutions to this problem. The first is that the relationship will tend to be strong when groups are in conflict: intergroup conflict seems to help organize behaviour into oppositional forms. Secondly, the group membership will be a stronger predictor when the groups are normatively relevant to the predicted behaviour. As social identity theorists have long argued, group behaviour is only predicted by group membership when that behaviour is consistent with a relevant norm for the group. Thirdly, the relationship will tend to be strong when relevant groups are chosen, and the focus here is on opinion-based groups: groups defined on the basis of a shared opinion. Fourthly, the relationship will tend to be strong when the degree or strength of self-categorization is measured appropriately. Here the argument is that measures of certainty of self-definition as a group member capture these constructs best for opinion-based groups. To sum up, self-definition as an opinion-based group member should strongly predict political involvement in conditions of intergroup conflict and when the behaviours involved are highly consistent with the norms of the specific salient group membership. In addition to the main prediction that group self-defmition should increase political group behaviour it was also expected that, in line with self-categorization theory, measures which better captured salience of opinion-based group membership should be stronger and more direct predictors of behavioural intentions than standard social identification measures. Consequently self-definition as an opinion-based group member was assessed using standard identification scales but also some new items which were aimed to capture salience of opinion-based group membership. In particular, it was argued that certainty of self-definition as a group member holding a certain opinion would be the best indicator of salience, but only especially so in opinion-based groups. These ideas were explored in a series of six studies (two surveys and four experiments). The results of the first two experiments using minimal opinion-based groups show that intergroup conflict had some impact on the main variables involved, and suggested the direction for the next studies in which the link between opinionbased group self-definition and political behavioural intentions was directly investigated. Results from Studies 3, 4, and 5 strongly supported the hypothesis that the salience measure employed predicted political behavioural intentions over and above identification. However, Study 5 provided no evidence that conflict enhanced the expected relationship. Finally, the results of Study 6 suggest that the salience measure is highly sensitive to the normative context and is a good predictor of political behavioural intentions but only especially so for the highly normative and validated behaviours. The conclusion reached is that certainty of self-definition as an opinion-based group member is an excellent predictor of normatively relevant political behavioural intentions. The success of this measure stems for the fact that it captures the selfcategorization theory construct of salience more accurately than do standard measures of social identification and this is almost surely because these measures capture the (relatively) enduring aspects of salience that are associated with perceiver readiness. In summary, self-definition as a group member is indeed a good predictor of validated normative political behavioural intentions in opinion-based groups, even where there is not explicit intergroup conflict.