![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
Abstract
By making use of an original data-set built based on a codification of all investiture debates of the Italian governments from 1946 to 2014, the paper investigates the main factors that explain the choice of a party to devote its attention to the valence issues of corruption and competence in its legislative speeches. Two classes of hypotheses are tested; the first concentrates on spatial reasons, and the second concentrates on contextual factors. Both sets of factors appear to play a significant role, although no clear temporal trend emerges in party attention over almost seventy years of Italian parliamentary debates.
Disclosure Statement
No potential conflict of interest was reported by the author.
Notes
1. The often discussed article in The Economist (Citation2001) claiming ‘Why Silvio Berlusconi Is Unfit to govern Italy’, published just before the 2001 Italian general election, confirms the significance of the issue of competence and its lasting connection with the issue of honesty/corruption in the years following the demise of the Italian First Republic.
2. All the texts have been downloaded from the Camera dei Deputati homepage: http://legislature.camera.it/
3. The coding of party leader speeches is a procedure also followed by the CMP for those parties for which no other available sources exist for policy declarations (see Budge et al. Citation2001).
4. To offer two examples, the following quasi-sentence included in the speech made by Matteo Matteotti, secretary at that time of the Social Democratic Party (PSDI) in 1955, during the investiture speech of the first cabinet led by Segni (Segni I) is assigned to the COMPETENCE category in our data-set: ‘we need to give to our country a cabinet able to face the urgent problems left by the unsatisfactory legacy of the previous government’. In contrast, the quasi-sentence ‘there has been no attempt to take an action against the corruption of the public bureaucracy’, included in the speech made by Silverio Corvisieri (Proletarian Democracy – DP) in 1976 during the Andreotti III investiture debate is assigned to the CORRUPTION category.
5. More specifically, I compared the parties’ left–right positions from each mass or expert survey relating to a particular election year with the vanilla parties’ scores averaged over the entire legislature following that election. Focusing on just vanilla parties’ scores as they appear from the first vote of an investiture debate of the legislature does not change any of the findings. Estimating the left–right positions of parties by following other methods (such as those suggested by Curini and Martelli [Citation2010] and Lowe et al. [Citation2011]) does not produce any noteworthy differences in the final analysis.
6. My theoretical question is similar to the one adopted by the literature on the Comparative Agendas Project (John et al. Citation2013), in that we are trying to understand the reasons why parties could have an incentive to highlight specific issues during their political confrontation. Conversely, and contrary to the Comparative Agendas Project, I focus explicitly on valence rather than policy issues.
7. Using alternative measures of party system ideological polarisation (see for example Dalton Citation2008; Curini & Hino Citation2012) would not change anything in this example.
8. As can be easily estimated, the sum of the three values of SPATIAL PRESSURE in the first case equals half that in the second case.
9. The operationalisation of the OTHERS EMPHASIS variable is similar to that introduced by Marks and Steenbergen (Citation2004) under the name ‘systemic salience’.
10. I row-standardised the connectivity (weight) matrix, as is commonly done in spatial econometrics research; i.e., the emphases on valence issues of all other parties are weighted equally in the above specification (see Franzese & Hays Citation2008). Row normalisation is not necessarily substantively neutral (see Plümper & Neumayer Citation2010). However, I consider it a reasonable choice in this analysis. In a further test, I introduced into the empirical models an emphasis on the two valence issues as placed only by the adjacent parties of each party to control whether there is a specific interdependence effect affecting the two parties (i.e., whether how party i talks about, for example, CORRUPTION is affected by what its contiguous neighbour parties j and z do, besides and beyond what the entire party system does; see Williams Citation2015). This variable, however, never becomes significant, while the remaining parameters are unaffected by its inclusion.
11. See, however, Schofield and Sened (Citation2006) for a contrasting perspective that expects a positive (rather than negative) relationship between being an extreme party and valence issues.
12. Specifically, a party (or a parliamentary group) that is appearing in Parliament for the first time takes the value of 1 in the NEW PARTY variable for all the speeches that MPs belonging to that party give.
13. Controlling for the effect of the first-level electoral cycle does not affect any of the results reported below. The same occurs if we control for the overall level of party system polarisation occurring in each legislative speech. Given what was discussed above, this last finding is not surprising. Finally, I have also controlled for the relative ideological position of parties. While there is a slight tendency for rightist parties to talk less about valence issues than other parties do, the main results reported above are qualitatively not affected.
14. Replicating the analysis employing a random (multilevel) model does not affect the conclusions reported below. I have also estimated a fractional-logit model with respect to COMPETENCE. (Given the square root transformation applied to CORRUPTION, a fractional-logit model cannot be estimated with respect to this latter variable.) Once again, my conclusions remain intact. See the online Appendix: Table .
15. For example, Kumlin and Esaiasson (Citation2012) record all of the corruption or financial misconduct scandals involving political actors during electoral campaigning in the Western European democracies (including Italy) but only from 1977 to 2007.
16. Using an SAR model rather than a spatial-X model makes sense if the content of the speeches made by MPs during an investiture debate can be considered a process of simultaneous shifts. Given that during an investiture debate speeches are usually given according to a pre-established schedule, this assumption could be debatable. In such a situation, the spatial-X model nevertheless allows one to estimate responses by parties to previous speeches.
Additional information
Notes on contributors
Luigi Curini
Luigi Curini is Associate Professor of Political Science at Università degli Studi di Milano. His research centers on the spatial theory of voting, party competition and social-media analysis. His articles have appeared in, among others, the European Journal of Political Research, Journal of Politics, British Journal of Political Science, West European Politics, Electoral Studies, and Comparative Political Studies. He is the author of several books. The most recent one is Why Policy Representation Matters: The Consequences of Ideological Congruence between Citizens and their Governments (Routledge, 2015).