The general elections that took place on 9 and 10 April 2006 marked the ultimate expression of bipolarism in Italy. By ultimate we do not mean that bipolarism is destined to end any time soon, but rather that in the future it is unlikely that it will be represented so effectively, with such duality and such fragmentation. An electorate split exactly in half emerged from the ballot boxes, as the two competing coalitions, the centre-left Union (Unione) and the centre-right House of Freedoms (Casa delle liberta` – CdL), obtained an almost equal number of votes. In the lower chamber (Chamber of Deputies), the Union prevailed by

25,000 votes, a very slim margin of 0.07 per cent of the total valid votes (which is still contested today, with a re-count underway in a sample of electoral areas), yet sufficient to guarantee the Union the majority bonus as provided for by the new electoral law, originally proposed by the CdL, the governing majority at the time, and approved by parliament in autumn 2005. In the Senate the CdL prevailed, according to the number of votes

(but also by a narrow margin: 270,000). However, it was the Union that again won the majority, though only by two seats (which are also contested!). This victory was made possible by the new electoral law, with its differing ways of granting a working majority1 – on a national basis in the Chamber, and on a regional basis in the Senate, where the Union received more votes than the CdL in more regions (10 out of 17). In addition, the parties of the CdL, unlike those of the Union, ran on divided tickets in the constituencies of ‘Italians living abroad’, who were allowed to vote for the first time. This decision inevitably doomed the CdL to defeat. But the meagre and uncertain majority held by the government in the Senate has weakened it, and has given powers of pressure – and blackmail – to all the political groups represented, including the smallest. The new electoral law provides that parties are voted for directly, but

awards a majority bonus to the winning coalition. The law also encourages all parties to form coalitions because it establishes a particularly low threshold for party lists belonging to a coalition to obtain seats (less than 2 per cent). Thus the political space for challenging the major coalitions is reduced to practically nothing. Indeed, the number of votes captured by lists outside of the two principal coalitions was reduced to about 170,000, or 0.5 per cent of valid votes. This was the lowest level since 1994 (Figure 1). The consequence is the bipolar and fragmented structure, not only of the

electorate, but also of political and parliamentary representation in the Italian party system. However, reference by observers and political actors to a ‘country split in half’ is partial, to say the least (cf. Mannheimer and Natale 2006). It is also true that the two dominant coalitions are affected,

in turn, by a plurality of party formations that are all determining and influential, no matter how small (see Bardi 2007). Thus the 2006 general elections presented an exemplary opportunity for

reflecting on the Italian political system and on the parties and coalitions of the centre-right and centre-left. They offered a faithful summary of the main features of electoral orientations and the party system as a whole:

1. a fragmented and unstable bipolarism, which works for winning elections, but becomes a setback when governing or organising the opposition in parliament;

2. a party structure based on a few stable parties, and joined by others that act as free riders;

3. personalisation, meaning the importance assumed by leaders whose image plays a role in offering a unified front to quite divided and unstable alliances; and

4. a tension between unity and competition, between Party (or Federation) and parties.