Why do people tend to vote for the same political party many years, despite the fact that parties change their electoral programs, their leaders and their candidates in elections? Why do they tend to justify the actions of the party much more easily than others? The answer to those questions is provided by the party identification theory originating in the prominent Michigan school of electoral behaviour research.
While there is some evidence that party identification is decreasing in Western countries, the concept is still widely discussed and used in electoral research. There is a huge amount of literature on the measurement issues, changing level of partisanship in US, Europe and other countries, on the determinants and the effects of party identification. The empirical evidence on the formation of the party identification in new democracies, however, is scarce. The case of Lithuania, therefore, is interesting and informative.
The aim of the article is to examine the factors of the emergence of party identification in Lithuania, trying to reveal the mechanisms of the formation of partisanship in the emergent state of party system formation. The analysis is based on the data from the representative face-to-face post-electoral survey carried out in 13 November – 10 December 2012 (N = 1500) for the Lithuanian National Election Study.
In the first part of the article, the two rival theories of party identification are introduced and hypotheses about the factors of party identification formation are presented. In the second part, the methodological issues of measurement of party identification are discussed and the level of party identification in Lithuania is examined. The third part of the article focuses on the determinants of party identification. Three main hypotheses are tested. The first hypothesis presumes that party identification derives from the political socialisation in family and depends on the party identification of parents. Family socialisation in Lithuania, however, should only have an impact on those who were born after 1972 because they received their political socialisation in democratic regime. The second hypothesis is related to the theory of social cleavages. It states that the probability of having party identification is bigger for those who feel the salient social cleavage in the society and recognize its political relevance. Finally, the third hypothesis, based on the theoretical claim of Philip Converse, is that party identification depends on the length of democratic experience of electorate.
Careful analysis of the level of party identification reveals that no less than a third of Lithuanian electorate could be characterised as party identifiers. Only half of them, however, have an affective relation with their party. The partisanship of the others seems to be based on rational evaluation rather than psychological attachment. The three parties that have most of partisan supporters are the Lithuanian Social Democratic Party, the Labour Party and the Homeland Union – Lithuanian Christian Democrats.
The results of the analysis reveal that consistent party identification of parents has an impact on the party identification of their children. The effect, however, is not different for those who are born before and after 1972. It suggests that this is the effect of social network rather than family socialisation. The probability to have a party identification is higher for those who are most anti-communist as it is the dominant political cleavage in Lithuania. The impact, however, is mediated by age and moderated by interest in politics.
The impact of age on party identification is considerable but not consistent with the hypothesis. The theory presumes that the probability to have a party identification should consistently grow until the age of 40 and afterwards it should remain constant. The results of the logistic regression, however, reveal that the highest probability to have a party identification is in the 50–59 age group rather than 30–39, even after controlling other variables such as interest in politics, political sophistication and salience of political cleavages. One possible explanation of this finding suggests that party identification is strongest for the generation that was most active in the formation of democratic political system in Lithuania in 1988–1991.