Oct 3, 2008

Political preferences in Italy

A new research paper by Alesina and Holden and Vox sheds some light on the ambiguity of the political platform of US Presidential candidates. The authors present a model where parties are pulled in different directions by two forces: on the one hand, the median voter theorem (MVT) induces candidates to present the political program preferred by the median voter; on the other hand, highly politicized campaign contributors pull candidates away from the median voter. This results in a polarized equilibrium and in ambiguity of political messages. That's cool, because it recognizes that the distribution of political preferences may not be a bell-shaped normal, but rather have the traits of a bimodal one (polarized political spectrum).
Could this line of thinking be applied to the sick political (and economic) man of Europe, Italy? As you may know, the Italian political system is highly polarized between those who worship Mr. Berlusconi and those who detest him, without necessarily liking his political "opponents" (guess where I stand). What happened in the last elections is that Mr. Berlusconi won easily, the newly created "Democratic Party" of Mr. Veltroni lost miserably, and the left-wing parties (ex-communists, ex-socialists who had not sided with Mr. Berlusconi and the Green Party, all reunited under the colorful coalition called "Sinistra Arcobaleno") were totally wiped out like Napoleon's army at Waterloo. Historically, Italy was governed by essentially right-wing "Democrazia Cristiana" (basically, the Vatican's political arm) and by Craxi's Socialist Party (the masters of corruption) for fifty years, then by Mr. Berlusconi one-man-party for six years or so. Mr. Prodi and friends (center-left unstable coalitions) governed for seven years or so.
In the light of this evidence, it is quite safe to assume that the Italian median voter is mildly right-wing (ID: catholic, low-to-medium educational attainment, over 50). If this is true, then the distribution of political preferences may be represented by a bimodal distribution which is skewed to the right:

Figure 1. Distribution of political preferences in Italy

Now, even if the distribution is like this, the MVT still holds. Each party will try to tailor the program to best suit the catholic, 50-or more year old and with low-to-medium educational attainment median voter. Say, for illustrative purposes that the median is at 1 in the figure. If the left-wing party proposes a platform equal to 1, however, all the people around -5 will be really pissed off, because the distance between their preferences and the proposed platform is huge. There may even be a threshold distance that makes the voter indifferent between voting for the left-wing candidate and dropping from the electorate. If this is the case, it is easy to see that the left-wing candidate finds himself (there are no significant "she's" in Italian politics) in an impossible situation. If he stays at 1 he loses all voters below the threshold, he moves away from 1 he leaves more that 50% of the electorate to the right-wing candidate. Really, his best chance is to send mixed messages, ambiguous enough not to alienate too many voters on either side, and to hope that the distaste for the other candidate, coupled with the ambiguity of the platform he proposes, will be enough to get the votes of everybody on the left of the median. Very tough job indeed.
This simple conjecture explains quite a few facts of Italian politics. To start with, it is consistent with the real distribution of political preferences. Second, it rationalizes the high levels of ambiguity contained in Mr. Veltroni's platform (Italian satirical commentators made successful jokes about this). Third, it is consistent with the high levels of cohesion among the right-wing parties around the charismatic figure of the boss - Mr. Berlusconi. Finally, it explains the debacle of the left-wing parties, which lost Parliamentary representation.
Another sad but realistic implication is that - given the progressive ageing of Italian society, the total lack of empowerment of the young and the weaknesses of the educational system - the skewed distribution of political preferences is not likely to change any time soon, making it easy for Mr. Berlusconi and friends to dominate the political arena. Very sad message indeed.

1 comment:

Dany said...

but you are note considering the political system. In Gringolandia a presidential system with election in proportion to very different states. In Italy a parliamentary system proportional to the parties votes... who is the median voter? the media is just bimodal or multimodal and proportional to states, parties or whatever cluster of election?