Changing Models of Electoral Choice in Italy

Abstract

Scholars argue that the realignment of the electorate which took place with the transition to the so-called Italian Second Republic followed mainly a traditional partisan pattern, with electors of the former centre ruling parties (the Christian Democrats and the Socialists) turning to vote for the new centre and right parties (Forza Italia and the National Alliance), while left-wing voters continued to hold their traditional allegiance. Behind this apparent electoral turmoil there would appear to be little in the way of voter mobility. Such a reading implies continuity in the motivations of voters who behaved according to their previous 'personal electoral history' and in accordance with their sub-cultural political identification. Here an alternative interpretation is proposed in which it is argued that as a result of the 1994 realignment elections voters who deserted the centre (the heirs of the Christian Democrats) did so also according to their class interest and in response to the policy proposal of the centre-right. Since that time a ‘valence model’ of electoral behaviour has begun to emerge where 'reasoning voters' react to the performance of the incumbent, and voting also depends on an assessment of leadership, policy performance and issues.

Acknowledgements

A first version of this article was presented at the annual conference of the Association for the Study of Modern Italy (London, 26–27 November 2004). The author wishes to thank participants at the conference and the two anonymous reviewers for their most useful comments and suggestions. Any responsibility for errors or omissions remains with the author. The analysis was carried out on the basis of the work of the Italian National Election Study (ITANES) survey series.

Notes

Notes

[1]  ITANES is the Italian National Election Study research programme launched at the beginning of the 1990s by the Istituto Cattaneo and later joined by several Italian universities, which carries out surveys on the occasion of national elections. Information on the studies, participants and data can be obtained at www.cattaneo.org.

[2]  Sub-cultural belonging is defined as a compound index based on the following variables measuring individual attitudes and behaviour in five surveys conducted between 1968 and 2001: (1) Catholic sub-culture: Church attendance, membership in Catholic organizations, trust in the Catholic church; (2) Socialist sub-culture: Union (Cgil) membership, class identification (variable missing in 1996 and 2001), trust in unions, unions best represent my opinions. Source of surveys: 1968, S. Barnes study; 1975: Political Actions Study; 1985: Four Nation Study; 1996–2001: Italian National Election Study (ITANES). The data-sets were released under the auspices of ICPSR and the Istituto Cattaneo.

[3]  It must however be noted that in the last 2006 election there was a slight increase in the Catholic vote for the centre-right coalition compared to previous elections. Segatti (2006) attributes this to the bioethical issues raised during the campaign, arguing that the Catholic vote is therefore subject to political competition, though it no longer represents a stable political cleavage.

[4]  We coded Left a vote for all parties on the left–right continuum up to the Socialist Party; Centre the Christian Democratic Party and its heirs after 1994, plus the minor centre parties; Right the Liberal party, MSI-AN, the Lega Nord and Forza Italia after 1994. The sources of the surveys, in addition to those cited in note 2 were: 1972, S. Barnes and G. Sani study; 1979: Eurobarometer 11; 1983: Eurobarometer 20; 1990, 1992, 1994, 1996, 2001, 2006: Italian National Election Study (ITANES). The data-sets were released by ICPSR, Eurobarometer, ZA, University of Cologne, and the Istituto Cattaneo.

[5]  ‘Stable voters’ are those who in the 1992–94 period and after voted for parties directly originating (through change, split, or fusion) from earlier groups (e.g., voted for the DC in 1992 and for the PPI or CCD in 1994).

[6]  A fuller discussion of the model is in Bellucci (2006). See Clarke et al. (2004) on ‘valence politics’ theory, and for a thorough, theoretical and empirical appraisal of rival models of voting behaviour.

[7]  Competence is computed in three steps: first, by totalling for each respondent across nine policy domains (immigration, health care, taxation, education, unemployment, judiciary, sate administration, criminality, Europe) the times each coalition is perceived as being best at handling it. We obtain one variable for every coalition, each varying from 0 (no issues are attributed to the coalition) to 9 (every issue is best handled by the coalition). Then we subtract the centre-left competence from centre-right competence, coming up with the coalition differential, an interval variable ranging from –9 and +9. This has been re-coded into three values: prevailing competence of centre-left (a differential value between –9 and –1: this is 27 per cent of the sample); same competence (each coalition receives the same score, so the differential value is 0: it comprises 25 per cent of the sample); prevailing competence of centre-right (values between 1 and 9: 48 per cent of the sample).

[8]  We cannot go into great detail here as to why there was a lack of popular enthusiasm for the competence of the Unione, except to suggest that this was due mainly to the policy heterogeneity of the coalition, the difficult agreement reached by the parties over the election manifesto, the complexity of the few slogans employed in the campaign, and a tarnished image of the leader.

[9]  These rates refer only to direct vote transfer between the centre-left and centre-right coalitions, a component of the larger total swing-vote which amounted to 10.6 per cent in 1996, 17.5 per cent in 2001, and 10.8 per cent in 2006 (ITANES 2001; De Sio 2006).