Modern Italy Vol. 12, No. 1, February 2007, pp. 17-38 Routledge Taylor & Francis Croup Remaking Italy? Place Configurations and Italian Electoral Politics under the 'Second Republic' John Agnew The Italian Second Republic was meant to have led to a bipolar polity with alternation in national government between conservative and progressive blocs. Such a system it has been claimed would undermine the geographical structure of electoral politics that contributed to party system immobilism in the past. However, in this article I argue that dynamic place configurations are central to how the 'new' Italian politics is being constructed. The dominant emphasis on either television or the emergence of 'politics without territory' has obscured the importance of this geographical restructuring. New dynamic place configurations are apparent particularly in the South which has emerged as a zone of competition between the main party coalitions and a nationally more fragmented geographical pattern of electoral outcomes. These patterns in turn reflect differential trends in support for party positions on governmental centralization and devolution, geographical patterns of local economic development, and the re-emergence of the North—South divide as a focus for ideological and policy differences between parties and social groups across Italy. Introduction One of the high hopes of the early 1990s in Italy was that following the cleansing of the corruption associated with the party regime of the Cold War period, Italy could become a 'normal country' in which bipolar politics of electoral competition between clearly defined coalitions formed before elections, rather than perpetual domination by the political centre, would lead to potential alternation of progressive and conservative forces in national political office and would check the systematic corruption of partitocrazia based on the jockeying for government offices (and associated powers) after elections (Gundle & Parker 1996). But a persisting feature of Italian electoral politics is the continuing lack of electoral bipolarity at other geographical scales, such as the regional and local. Italy remains politically a 'geographical expression' with little evidence of either emerging nationwide swings between party groupings or nationwide opinion voting in which any voter anywhere Correspondence to: John Agnew, Department of Geography, UCLA, Los Angeles, CA 90095-1524 USA. Email: jagnew@geog.ucla.edu 1353-2944 Print/1469-9877 Online/07/010017-22 © 2007 Association for the Study of Modern Italy DOI: 10.1080/13532940601134791 18 /. Agnew is equally likely to vote for any party. There is also, as yet, an absence of institutional bipolarism in the sense of true left and right parties replacing the ad hoc arrangements at work in what remain strange and often ideologically incoherent coalitions. The whole concept of a 'normal country', however, is deeply problematic. It is based on an idealized model of electoral politics in Britain and the United States which presumably lack the geographical and ideological fractures of Italy and as a result effortlessly produce alternation in national office between distinctive left- and right-leaning political forces (Agnew 2002, chapter 4). Of course, Italian politics has many unique features. But geographical variance in support for political parties is not one of them. This is a widespread characteristic of electoral politics around the world (Agnew 1987). Lacking in the study of Italian politics, as well as elsewhere, has been an understanding of why this is the case. Crucial has been the seeming difficulty of thinking geographically about national politics. Michel Foucault (1980, p. 149) has captured most vividly what seems to have happened in conventional thinking about space and time: Space was treated as the dead, the fixed, the undialectical, the immobile. Time, on the contrary, was richness, fecundity, life, dialectic.... The use of spatial terms seems to have the air of an anti-history. If one started to talk in terms of space that meant that one was hostile to time. It meant, as the fools say, that one 'denied history.'... They didn't understand that space... meant the throwing into relief of processes—historical ones, needless to say—of power. As Foucault was suggesting, the devaluation of spatial thinking is a well-established intellectual tradition in its own right. So, it is no surprise that thinking about Italian politics should follow a similar logic. Except, that is, because there can be few countries that would seem to be so ripe for the application of spatial thinking. Not only is Italy obviously divided geographically by significant economic and cultural cleavages, its politics has often been understood in spatial terms, by students of the 'southern question'—the North-South gap in economic development—and of fixed regional political cultures (the red and white zones) as well as by those suggesting more complex typologies of region-based voting processes (exchange or patronage votes in the South, party identification votes in the coloured zones, and opinion voting in the Northwest) and centre-periphery relationships in relative power between central and local governments (Agnew 2002, chapter 2). Yet, time and again, influential commentators have announced the immanent demise of a geographically divisible Italy as votes nationalized around two major parties (as in the 1970s with the Christian Democrats [DC] and the Italian Communist Party [PCI]) or as the media controlled by one man, Silvio Berlusconi, have finally unified the country politically in a nationwide electoral marketplace that is transcending older and now largely residual local and regional mediations (as in recent years). In this article, my primary goal is to challenge the theoretical and empirical adequacy of the vision of a single Italy as the emergent trend of post-1992 Italian electoral politics. My central thesis is that the historical pendulum does not swing from local to national but constantly around these and other geographical scales (through the Modern Italy 19 linkages that tie places together as well as separate them in their particularity) as a result of which, though the balance of importance between them changes, there is never a final victory for one, be it either local, regional or national (Agnew 1987, 2002; Brenner 1999). However fervently it may be desired by nationalizing (or regionalizing) intellectuals, permanent geographical catenaccio (lock-down) at a single scale is not yet a feature of Italian politics. Before turning to consideration of some of the details of post-1992 Italian electoral geography, I want to describe briefly some of the current conventional wisdom about a nationalizing Italy and to say a little about what I mean by 'place configurations and electoral polities'. Envisioning a Single Italy There are two different versions of the 'single-Italy thesis' as applied to contemporary Italian electoral politics—if one emphasizes Berlusconi's putative revolutionary use of mass communication to reduce Italy to a single homogenized 'public opinion', the other argues for the recent homogenization of political opinion around a nationwide menu. The rise of Berlusconi's Forza Italia party, undoubtedly the centrepiece of the centre-right alliance since 1994, is seen as particularly representative of this new national homogenization. In both cases, therefore, national politics is seen as operating increasingly without mediation by places or territory. From these perspectives, the question of 'where?' is ever more irrelevant to understanding the workings of Italian electoral politics. Berlusconi and Television The advent of Silvio Berlusconi, Italy's media baron, to national political office as leader of his own political party, Forza Italia, and twice Prime Minister of centre-right Liberty Pole/House of Freedoms governments, is frequently interpreted as representing the success of a national 'telecratic' model of politics over the old party-based model. In this interpretation, as recounted, for example, by Daniele Zolo (1999), parties no longer call themselves such (they are slogans, as in Forza Italia, or known as a Lega (league), Alleanza (alliance), Polo (pole), Casa (house), Rete (network), or Ulivo (olive tree)) and they relate to the public and their voters 'in ways that are radically different from those in the past' (Zolo 1999, p. 727). Notoriously, political communication now is largely in the hands of one man through his control over most television channels, both private and public. But what is more important is that this man, Berlusconi, has changed the rules of the political game. Other politicians have followed where he led. 'Italy has evolved, in less than twenty years,' Zolo (1999, p. 728) asserts, 'from a neoclassical democratic model, founded on the competitiveness of the multi-party system, to a post-classical democratic model, that is to say, beyond representation, dependent on the television opinion polls and the soundings of public opinion.' Reaching everywhere in Italy, television has replaced grass-roots organization as the main instrument of political involvement. Thus, 'the new politicians no longer belonged to "parties"—they became elites of electoral entrepreneurs who, competing among themselves through advertising, spoke directly to the mass of citizen consumers offering them their symbolic "products" through 20 /. Agnew the television medium according to precise marketing strategies' (Zolo 1999, p. 735). As a result, 'Not only is political communication almost totally absorbed by television, but so is the whole process of the legitimization of politicians, of the production of consensus and of the definition and negotiation of the issues which have no other location and, so to speak, no other symbolic places except television studies and popular entertainment programmes—to which the stars of the political firmament are often invited' (Zolo 1999, p. 739) (my emphasis). Television in general, and Berlusconi's ability to use it to advantage in particular, have undoubtedly had major effects on Italian electoral politics. Parties do increasingly rely on advertising and polling to push their agendas. Television viewership is relatively higher per capita in Italy (and elsewhere in southern Europe) than in the US and northern Europe (Wise 2005). Certainly, in Italy and elsewhere, celebrity politicians—think of the actor Arnold Schwarzenegger in California, the wrestler Jesse Ventura in Minnesota, the businessman Thaksin Shinawatra in Thailand—have challenged the centrality of professional or machine politicians (Street 2004; Hilder 2005). The personalities of candidates increasingly eclipse the character of parties as major elements in political campaigns (Venturino 2005). Political parties everywhere have also lost much of their capacity to make voters identify strongly with them, not least, perhaps, because governments everywhere have lost their ability to enact what the parties that compose them promise at election time. Specifically, with economies less nationally structured under conditions of globalization, governments believe themselves less able to execute independent economic policies. But whether these trends, particularly that of the centralization of media control, have had the totalizing effects on national politics alleged by Zolo and others is open to question. For one thing, many segments of the population do not rely as heavily on television for entertainment or information as is often alleged. Young people in Italy, for example, are increasingly drawn to radio rather than to television. This is one reason why Berlusconi has recently set his sights on increasing his share of the notoriously fragmented Italian radio business (Taddia 2004) . People also tend to watch the television channels and programmes that already appeal to them and avoid those that do not. In this regard television (as with partisan newspapers) tends to reinforce and mobilize already held opinions rather than convert people to new ones. More importantly, opinions are also still formed in everyday interaction with other people, notwithstanding their joint reliance on increasingly homogenized national sources (Katz & Lazarsfeld 1955; Zuckerman 2005) . People in different social groups and operating in different milieux interpret what they encounter in viewing television in radically different ways. However persuasive television often appears, the best attempts at persuasion often backfire when people bring their own 'common sense' and identities to bear in interpreting what they see (Hall 1980). People are neither as gullible nor as ignorant as pollsters or media critics often make them appear. When they are, they do not require television to encourage or validate them! As Giovanni Sartori (1989) has argued, television can also encourage localism more than nationalization. It takes attention off parties and puts it on politicians and their service to constituencies. At one and the same time television thus moves between the extremes of 'no place' and 'my place'. Any sort of national 'good', as inherited from the nationalism of the French Revolution, is lost in between (Sartori 1989, p. 189). Modern Italy 21 Finally, Forza Italia's success probably owes more to Berlusconi's role as a politician than his role as a media baron. Not only has he been effective as a coalition builder, at the very least politically mobilizing local business elites and Rotarians all over Italy and bringing together various political forces from the political right, but Forza Italia has become much more of a membership organization rather than a simple electoral vehicle operated from the offices of Berlusconi's main business, Mediaset. Even Forza Italia has had to organize itself territorially. It seems to have done so relatively successfully (Poli 2001; Mannheimer 2002).1 Vital to this success has been the image Berlusconi has cultivated of himself as a persecuted outsider crusading for the interests of other 'self-made' people, drawing something perhaps from the US Republican Party strategy of portraying its electoral adversaries as the 'enemies' of ordinary people, which at one and the same time both obfuscates and subtly suggests his own dependence on political connections for his own business and political success. In his very disavowal of insider status his initial dependence on a political mentor, Bettino Craxi, for his business success and his constant run-ins with the judicial system for shady business practices proclaim him as the living symbol of the well-established Italian politics of raccomandazione and crony capitalism (Zinn 2001). At the same time, and hardly unique to Italy, Berlusconi also represents the appeal of a person who has made it in business (and as President of AC Milan, in football, the most popular sport in the country) who constantly draws attention to the fact that his political adversaries 'have never worked'. This appeals to those dismayed and alienated from professional politicians and 'politics as usual'. Even without Berlusconi, therefore, there will be a continuing basis for this aspect of 'Berlusconismo' and the emphasis on a populist rapport between the leader and the population at large. Opportunistic in pursuing themes that appeal to a centre-right electorate (such as the Catholic hierarchy's objections to fertility treatments, etc.), Berlusconi has worked most actively to fuzzy the distinction between state and market beloved of true liberals, not least to preserve his own vast business interests from competitive pressures (Pasquino 2005). Paul Ginsborg (2003), hardly insensitive to the role of Berlusconi's media ownership in recent Italian politics (Ginsborg 2004), argues that too much emphasis on television risks ignoring 'the degree to which other forces were at work in Italian modernity, forces which ran counter to any idea of the facile manipulation of the individual'. Indeed, the old pre-1992 Italy was dominated by two 'churches', the Catholic one (DC) and the Italian Communist Party (PCI). Since their erosion and then disappearance as political agents, a plethora of small groups, non-governmental organizations, and single-issue movements have pluralized the Italian political scene, although unevenly from place to place. Moreover, television has long been overtly politicized in Italy with different parties previously dominating different state channels in both personnel and message. What is perhaps most important about Berlusconi is that his almost total dominance of Italian television in recent years has helped to shape popular tastes in such a way that favour his type of celebrity political candidacy (Ginsborg 2004; Bendicenti 2006). Even so, distracting mass publics and steering public opinion are imperfect arts as Berlusconi's defeats in 1996 and 2006 attest. Active human agents can always react perversely to media 'spin' and often match what they see on television with their own prior experience at the expense of the former (Thompson 2000, pp. 262-263). 22 /. Agnew Electoral Homogenization A second account emphasizes rather both the declining role of the regional mediation in electoral orientations alleged to have dominated Italy before 1992 and an end to the seeming rise in importance of localism and local government that took place in the 1990s and was closely associated with the political rise of the Northern League. In this construction, offered most forcefully recently by Ilvo Diamanti (2003a, 2003b), these old patterns are said to be giving way to a nationalized 'electoral market' as 'differences in votes between areas decline, the geographical gateways of single parties, those that were considered subcultures, have all faded, not only the white zone of DC but also the green zone of the League that was built and now has faded, and DS and Rifondazione comunista have lost major support in the central region, above all in the districts where they were most weak in this area' (Diamanti 2003b, pp. 239-240). In particular, 'there no longer exist specific area interests that characterize the politics and the policies of the parties; above all those of the centre-right, the present government. Because the majority, particularly Forza Italia, has an electoral base scattered in different zones' (Diamanti 2003b, p. 239). Forza Italia is characterized as practising a 'politics without territory' (Diamanti, 2003a, p. 85). Institutionally, this trend is said to reflect a rebalancing between centre and periphery, such that, for example, the 2001 election saw a ballot in which Berlusconi and his coalition were formally paired together, and the national government has begun to reassert its authority as a result of EU directives and popular demands to deal with 'national' problems. But this is not a return to the past, even if it is a 'return of the state'. To Diamanti, it is more a reimposition of authority at the centre in the face of a vastly changed country in which the swing of the geographical pendulum to the periphery had gone too far. At the same time, however, Diamanti (2003a, p. 7) claims that territory qua place does not simply imply a backdrop to political processes but is 'a crossroads ... where society, politics and history are joined together and where they become visible.' This is a fundamental tenet of his previous writing on Italian politics in general and the Northern League in particular. But here it has become contingent rather than necessary—present significantly only when a dominant political subject, such as the old regionally hegemonic parties in the white and red zones or the League in the far North, brings it into play. What seems to have happened is that Diamanti has perhaps fused three different conceptions of the role of territory in politics without clearly distinguishing their different consequences for Italian national politics—the role of territorial or jurisdictional claims in a party's discourse (critically that of the Northern League in relation to Padania or northern Italy versus the rest of Italy), the role of territory (or place) in social mediation between people and parties, and the relative autonomy of local politics vis-a-vis central government. The fading of either the first or the third, I would suggest, does not necessitate the fading of the other two, particularly the second. It does seem clear that the old regional subcultures, to the extent that they were ever as powerful in the regions to which they were ascribed as Diamanti alleges, have eroded, although there is also evidence that erosion was well under way before 1992 (Agnew 2002, chapters 5 and 6). A case could be made for the re-emergence of central state authority after a period in which it had weakened, even though big city Modern Italy 23 mayors and regional governors have all acquired powers lacking before 1992, and a watered-down devolution law giving certain health, education, and policing powers to the administrative regions passed Parliament in 2005 and came up for referendum vote (along with a mixed bag of other constitutional matters) in June 2006 (it failed to pass by a large national majority). But do trends towards a breakdown of regional party hegemonies and a reassertion of state authority necessarily signify a collapse of geographical mediation in Italian electoral politics tout court! Indeed, Diamanti's (2003a, 2003b) own exposition suggests anything but. His discussion of electoral trends is entirely in terms of changing geographies of support. What he demonstrates, in fact, is that an idealized regional pattern has given way to a pattern of localities or what he himself calls electoral 'archipelagos' (Diamanti 2003a, p. 105). As the least regionalized party, Forza Italia still has a demonstrable electoral geography which, although distinctive in the precise localities it encompasses as areas of strong support, bears a remarkable geographical resemblance to the split North/South vote of the old Christian Democratic Party at the top of its game in the 1960s. At the same time, other major parties retain or have established even more definitive geographies of popular support, even as some regions have become more competitive between parties than hitherto. Even as parties have lost some of their grip on pools of support in different regions, therefore, voters still seem to exhibit distinctive patterns of electoral choice that are definitely not the same irrespective of where they are in Italy. More recently, Diamanti (2004) has explicitly backtracked from the claim of an emerging 'politics without territory'. Even Forza Italia now is said to have a 'territorial character'. If before 1990 there was a certain geographical stability to Italian electoral politics, since then much has changed: The time of ideological fidelity and undiscussed political identities is finished: the time of eternal passions. But whoever believed that television was enough for repositioning, with marketing experts and opinion polls at the service of closed oligarchies, must reevaluate their belief. And to take up studying again: Society, territory, geography (Diamanti 2004, p. 2). Place Configurations and Electoral Politics But how should we go about doing so? I find it useful to think of explanations of political behaviour as either compositional or contextual in nature (more generally, see Goodin and Tilly 2006). I can offer only relatively brief remarks here about this distinction, which I have described in greater detail elsewhere (Agnew 2002, chapter 2). Compositional explanation characteristically locates behaviour in individual persons and, more particularly, in their associated socio-economic attributes. From this perspective, all that needs to be known about people to understand their voting and other political behaviours is which national census categories they belong to. In other words, support for political parties is best explained by reference to the national-level socio-economic composition of their electorates. Rationally, therefore, people can be presumed to vote in line with the putative interests of the specific 24 /. Agnew census groups to which they belong. Contextual understanding, however, emphasizes on the one hand the mediating role of social and political milieux such as workplaces, residential and other living arrangements, party origins and organization, religious practice, association memberships, information sources, and histories of social conflict between the agency of individual persons, and on the other, making electoral choices (Agnew 1987, 1996; Beck et al. 2002; Taylor 2006). From this point of view, census categories are meaningless unless placed in the contexts of everyday life. Thus, such-and-such a class 'membership' in one context can elicit a very different meaning and, consequently, a different type of vote or party orientation than it might somewhere else, for example in a large as opposed to a small city. To make this point more abstractly and, using Ian Hacking's (2004, p. 281) turn of phrase—'Existence precedes essence.' In other words, 'who you are is determined by your own actions and choices' (Hacking 2004, p. 282), not by a priori membership in a category of a classification scheme. Classifications, however, have feedback or 'looping effects' when people act as if they do belong to particular social (or other) categories. But they do so only on the basis of their own reactions to them in relation to the everyday constraints and opportunities they experience, not because the categories define them. Consequently, 'in any place and time only some possibilities even make sense' (Hacking 2004, p. 287). Contexts of 'place and time' are not best thought of as invariably regional or local (as in Putnam 1993), although they frequently have elements of one or both. Rather, they are best considered as always located somewhere with some contexts more stretched over space (such as means of mass communication and the spatial division of labour) and others more localized (school, workplace, and residential interactions). The balance of influence on political choices between and among the stretched and more local contextual processes can be expected to change over time, giving rise to subsequent shifts in political outlooks and affiliations. So, for example, as foreign companies introduce branch plants, trade unions must negotiate new work practices, which, in turn, erode long-accepted views of the roles of managers and employees. In due course, this configuration of contextual changes can give an opening to a new political party or a redefined old one that upsets established political affiliations (on Italy, see for example, Andreucci & Pescarolo 1989; Castagnoli 2004). But changes must always fit into existing cultural templates and cleavages that often show amazing resilience as well as adaptation (Griswold & Wright 2004; Brooks 2006). Doreen Massey (1999, p. 22) puts the overall point the best when she writes, 'This is a notion of place where specificity (local uniqueness, a sense of place) derives not from some mythical internal roots nor from a history of isolation—now to be disrupted by globalization—but precisely from the absolute particularity of the mixture of influences found together there'. I have used the term place to capture the mediating role of such geographically located milieux. What I mean by this word are the settings in which people find themselves on a regular basis in their daily lives where many contexts come together and with which they may identify. Or, as I have put it previously (Agnew 2002, p. 21), 'places are the cultural settings where localized and geographically wide-ranging socioeconomic processes that condition actions of one sort or another are jointly mediated. Although there must be places, therefore, there need not be this particular place'. So if individual persons are in the end the agents of politics, their Modern Italy 25 agency and the particular forms it takes flow from the social stimuli and imaginations they acquire in the ever-evolving social webs in which they are necessarily enmeshed and which intersect across space in particular places. This is a strongly social-historical view of politics, albeit one that also insists on the central role of spatial separations and differences in defining the concrete impacts of social influences. But such place configurations are not set in stone, so to speak. This perspective thus rejects both the extreme compositional view of electoral politics as best explained in terms of individual persons as carriers of national census attributes and the idea of fixed geographical communities in which people exercise no political agency whatsoever except as perpetuators of so-called traditional sub-cultural behaviour. It also rejects the view that space is best thought of simply in terms of territories or spatial blocs in which regions or other spatial units are hierarchically nested, like Russian dolls, within the state territory. The absence of regional political homogeneity, for example, does not signify the absence of a geography to politics, only that it is constituted by a possibly richer or more variegated pattern of places that are linked across space as well as separated from one another. From this viewpoint, therefore, if there is either a homogenization of votes for all parties across a national territory or an increased association of different parties with specific places, the patterns in question are produced by processes that are never simply 'national' or 'individual' but mediated by geographical existence in networked webs of influence that come together in the agency of people in places. Post-1992 Italian Electoral Geography The Impact of the New Electoral Systems There are perhaps four preliminary points that are particularly important to note as conditioning the post-1992 electoral geography of Italy. One is that the national electorate as a whole is decreasingly linked or attached to particular parties. Even before 1992, Italian electoral mobility was substantially greater than is often alleged (Agnew 2002, p. 79; Pasquino 2002). Since then mobility has increased further, creating an ever more volatile electorate. This seems to have two elements to it—a dramatic realignment of votes across parties (particularly in 1994), at least within the broadly centre-left and centre-right groupings, and an even more dramatic increase in abstentions or non-voting, although this trend had begun in 1979 (Wellhofer 2001, pp. 162-166; see also Chiaramonte 2002; Pappalardo 2002; Natale 2006). In 2006, at about one month before the election, one third of the intended voters remained undecided as to how they would vote, according to one poll (Mannheimer 2006). A corollary of this, and a second point, is that the electoral system used from 1994 to 2001 was increasingly competitive across Italy as a whole, largely as a result of the change in the nature of the electoral system from a proportional representation (PR) to a mixed single-member district (SMD) and PR system (Bartolini et al. 2004; Ferrara 2004). Specifically, the number of SMDs competitive between two candidates for the Chamber of Deputies increased between 1994 and 1996 by 80 per cent (Reed 2001, p. 323). Given the high weighting of these seats in the Chamber, 475 out of 630 (155 elected by PR), the single-member element in the mixed-member majoritarian electoral system undoubtedly pushed Italy towards a bipolar system 26 /. Agnew with two major electoral alliances between parties of the left and the right, respectively, and the real possibility, therefore, of alternation in national government because of the incentive it gave to pre-election compacts or coalitions to reduce multipolar competition at the district level. As a result, Reed (2001) finds that Duverger's Law, positing a causal relationship between single-member districts and the emergence of a two-party system, was 'working in Italy'. But the centralization of candidate selection within the alliances to ensure a 'fair' distribution of SMDs across parties in terms of lost, marginal, and safe seats, limited the possibility of building strong ties between candidates and local constituencies. In this respect, bipolar competitiveness in Italy was bought at the expense of strong territorial representation, one of the purported benefits of SMDs. At the same time, the degree of bipolar competitiveness was not the same everywhere in Italy. As Bartolini et al. (2004) show, the South was consistently the most competitive region across the 1994, 1996 and 2001 elections with the Centre as increasingly bipolar but largely non-competitive and the North with a trend towards competitive bipolarism depending on the Northern League's relative incorporation into the Polo/Casa delle Liberta. Even with the change back to a PR system in 2005 (if one with various peculiar features, such as 'topping up' the number of seats for the victorious coalition), this geographical pattern remained much the same in the 2006 election. Across the period, and this is the third point, two coalitions organized along a basically left-right continuum have increasingly accounted for most of the votes. This undoubtedly has owed much to the political polarizing capacity of Silvio Berlusconi but is also reflected in the increasingly rightward drift of the rhetoric and policy positions of the Northern League and the recruitment into the centre-left coalition of more leftwing parties (such as Rifondazione Comunista [Communist Refoundation]) hitherto outside the formal alliances. The left-right distinction may well have largely disappeared elsewhere in Europe as the overriding political archetype, but in Italy it lives on powerfully in political rhetoric (e.g. '(ex-) Communists still eat babies and Berlusconi is the new Mussolini') if less in terms of actual policy positions (Bobbio 1996). Finally, as a result of these trends, focusing exclusively on the PR element in the vote for the 1994-2001 elections is particularly problematic as a singular guide to contemporary Italy's electoral geography, even though it allows for continued comparison with the pre-1993 electoral system. Yet, as Bartolini et al. (2004, p. 17) also point out, the new coalitions show few signs of turning into true parties in their own right. Seemingly, existing party identities are still too strong for that to happen. So, as long as a PR element exists—and it made a major comeback in the electoral system in force in 2006—the parties, including those marginal to the main coalitions, will continue to try to attract votes by running candidates, even when running them in their own right is a lost cause as far as acquiring seats in either PR or SMD contests is concerned. After all, in 2001 around 4 million voters voted for neither of the two main electoral alliances in the single-member contests. A better strategy for smaller parties under the mixed electoral system was to merge into the coalitions and manoeuvre for representation that way. It is clear that in this context small parties have 'a large marginal value' in that they can expand the vote for coalitions (Bartolini et al. 2004, p. 12). But in return they could negotiate to place their candidates in SMDs and were compensated for allying by receiving more PR seats Modern Italy 27 when they did not win SMD ones thus perpetuating themselves in an alliance system (Cox & Schoppa 2002, p. 1050). This brings me back to the first point. When all parties, and particularly coalitions, have weaker linkages with the electorate, bipolar competitiveness cannot be guaranteed. Voters can always defect to smaller parties or abstain. Although, it is important to note, the electoral system in force for the 2006 election required a very high threshold (four per cent of total votes) for the parties to cross in order to acquire seats in the Chamber of Deputies if they stayed out of the two coalitions (but only a two per cent threshold if they remained inside). As a result far fewer voters strayed outside the two coalitions in 2006 than previously (Istituto Cattaneo 2006). Even so, and notwithstanding that much of it involves flows of votes within electoral blocs on the left and the right and into and out of the electorate than switching between coalitions, electoral volatility—both nationally and at other geographical scales-will not disappear any time soon. As a result, the party system is both more competitive nationally and more likely to produce governing alliances than under the First Republic but still favourable to the continued existence of small parties, if often within two loose coalitions. The Geography of the National Election Results, 1994-2006 Bearing these caveats in mind, contemporary Italian electoral geography is best examined in terms of the two 'parts' of the 1994-2001 national elections: SMD and PR components and the same indicators emanating from the PR system adopted for 2006. With respect to the SMD component and seats/votes for the two major coalitions, the trends from 1994 to 1996, 2001 and 2006 show a fairly clear geography when, for expository purposes, the administrative regions are adopted as the units of electoral aggregation. In terms of seats won, the northern and southern regions are much the more volatile ones overall, with the central ones much the more stable. Yet, in the North and South, Lombardy, the Veneto, Friuli, Basilicata, and Sicily are exceptions to this rule in showing rather consistent affiliations to one or other coalition across all five elections, even if the Veneto and Lombardy 'defected' in 1996 when the League stood separately from the centre-right Polo/Casa (Figure 1). As in other majoritarian systems, it was the 'swing' regions with the most marginal seats that determine the national outcome. In this regard, Piedmont, the southern peninsula regions, and Sardinia are much the most crucial regions. Overall, though, it is important to note that the electoral boost of Lombardy gave the Polo something of a head start there in accumulating SMDs even as the centre-left Ulivo (Olive Tree coalition) had the most consistently solid base with its hold on the central regions. With respect to votes cast, rather than looking at trends by administrative region over elections it is more useful geographically to examine trends by election over regions (Figure 2). This provides a visual perspective by way of a profile or transects running roughly from north (left) to south (right). For the two main coalitions, the North and Centre now appear as mirror images of one another with the southern regions (except Sicily) as more similar and hence competitive in terms of vote percentages. These graphs also reveal that across all regions votes for the two coalitions stabilized between 1996 and 2006 relative to 1994—in other words, even 28 /. Agnew Figure 1. The two major electoral alliances, single-member districts (1992-2001) and Italian administrative regions, by overall majority of seats won in Chamber of Deputies: (a) 1994, (b) 1996, (c) 2001, (d) 2006. Source: Author. Modern Italy 29 10- B7 c o E 0- o I — I ro I to I en I ra s: = c c — c CO I CO I CO I CO I CO CTJ o CO o If! ,o ZS CO N ZJ < CD CO c cO Q. E CO O CCi Z> CO o CO CO CD o o CO CO c en Q> ■o CO CO E LU Figure 2. Geographical transects by administrative region (from Piemonte on the left to Sardegna on the right) of percent votes cast for (a) Center-Right alliance and (b) Center-Left alliance for 1994, 1996, 2001 and 2006 in Chamber of Deputies elections (excludes Valle d'Aosta and Trentino-Alto Adige). Source: Author. when non-competitive, contests are increasingly bipolar. The transects for both sides are also increasingly shallow over time as together they have taken a larger share of total votes relative to all parties but there are still obvious north/centre/south differences between them. In terms of electoral outcomes, therefore, it seems clear that the centre-left has much the same problem that the PCI used to have in becoming competitive away from 'home' in central Italy, whereas the Casa/Polo shares the old DC problem of having to cater to such distinct constituencies as the Veneto and Lombardy on the one hand, and Sicily on the other. Elections are still won locally, even if candidates are also still parachuted in from outside. 30 /. Agnew It is not surprising, therefore, that levels of support for different coalitions and swings between elections to the benefit or detriment of either differ significantly across regions. Italy is not yet socially, economically, or politically a giant pinhead with no or even limited spatial variation in its electoral geography. Geographical variance in support for the two coalitions is extremely significant in determining which wins nationally. The enhanced role of the South in determining national electoral outcomes because of greater electoral volatility is brought into focus by the results of the 2004 European, 2005 regional and 2006 national elections when, as Mannheimer (2004) relates for 2004, 'The decisions of southern citizens are producing results ever more relevant to the political equilibrium of the entire country'. New Place Configurations and the Election Results But post-1992 Italian electoral politics is not simply a recapitulation of the old PCI/ DC dualism, however much the SMD component of the first three elections and the overall results at the regional level in 2006 might make it seem that way. Beneath the surface, the tectonic plates of electoral competition have been moving, particularly in the North and the South, to create different place configurations that represent a further breakdown of the regional patterns of party affiliations that had generally prevailed in the 1960s and 1970s but that were already under stress in the 1980s. In fact, 1992 is not geographically the watershed that it is often characterized as being by those of a purely historical cast of mind. The localization of Italian electoral politics was underway as early as 1979, prior to the electoral emergence of the Northern League and the collapse of the main old parties of government, the DC and PSI. The beginnings of localization were heralded in particular by the breakdown of the DC hegemony in the North-east and the southernization of the PSI (Agnew 2002, chapter 5). It is the PR element of the post-1993 elections that provides the best window into the balance between old and new place configurations.3 As Diamanti (2003a, p. 105) vividly describes it, this is the 'geography of an archipelago'. He provides a 'political map of the second republic' to make his case (Figure 3). Of course, Diamanti is claiming that, because the zona azzurra (blue zone) or area of strong support for Forza Italia is geographically fragmented, the map as a whole represents 'politics without territory'. Presumably this is because Berlusconi (and Forza Italia) appeals to diverse constituencies that are not geographically concentrated in a single contiguous region. Putting this rather curious use of the term 'territory' to one side, his general interpretive map does offer a valuable perspective on what has changed electorally since 1992 and what has stayed the same, even if one might quibble about some of the details. The logic of this approach, but with a threshold defined relative to party performance over the entire period 1994-2006, can also be applied to the 2001 and 2006 results (Figure 4) with some interesting similarities and differences. An expanded zona rossa is one of the starkest features of the maps in Figures 3 and 4, though its extent and configuration elicits little or no comment from Diamanti (2003a) given his focus largely on the 'archipelago' of Forza Italia. The successor parties to the PCI, particularly the Democratic Left (DS) and Rifondazione Modern Italy 31 Figure 3. The Italian "Archipelago" according to Diamanti: PR votes cast in Chamber elections, 1994-2001. Note particularly the "zone azzura" or blue zone. The "zones" are defined as sets of provinces where a party or in the case of the red zone, party grouping, (1) received its largest share (top quartile) of the vote in 2001 and (2) finished first in at least one of the three elections and second in at least one other. Source: Diamanti 2003a, 106 (2006). Comunista, have succeeded in keeping a strong regional hold even as their traditional hegemony within civil society has undoubtedly dissolved (Ramella 2005). This probably reflects the strength and competence of the parties in local government, national alliance with the Catholic-left (Margherita-Daisy list), the continuing importance of local government economic regulation to local economic development, and scepticism about the expected performance of central government (Agnew 2002, chapter 6, Vandelli 2002, Messina 2001). What Robert Leonardi (2003) calls 'the denationalization of policy making', particularly the rising importance of elected mayors, may be especially important in underpinning the resilience of the zona rossa under new circumstances (Rampulla 1997; Magnier 2004). Not everything has changed in Italy's political configuration of places. Yet, there is something new about the zona rossa of the 'Second Republic', as suggested in both Figures 3 and 4. This is its constitution in two segments—a still ggg LN and Fl rmiLN FI p~| FI and Left ~J Zona rossa Left and AN LSI AN and FI | | Neutral LN and Fl KOln FI I FI and Left ^| Zona rossa EZ2 Left and AN ■ an AN and FI | | Neutral Figure 4. Emerging place configurations. Chamber elections, 2001 and 2006. The zones are defined as sets of provinces where parties or party grouping exSed a threshold defined as the top quartile of votes pooled across all elections (1994-2006) for the 2001 and 2006 elections. This is a more relevant definition of the relative persistence of zones than that adopted by Diamanti (2003a) and also shows provincial changes between recent elections to illustrate the dynamism of Italy's contemporary electoral geography. Source: Author. Modern Italy 33 strong but attenuating central Italian one based around Tuscany and an emerging southern one based in Campania, Basilicata and Calabria. It is in central Italy that the economic model of small firms in industrial districts has entered into greatest crisis leading perhaps to a revaluation by some of the established connection between the left-wing parties and local economic health (Agnew et al. 2005; Shin et al. 2006). Some of the poorest parts of the South in 2006 voted in proportionately greater numbers for the centre-left than those in Emilia-Romagna, signifying a distinctive shift in geographical constituency as well as probably a diversification in the identities and interests of the centre-left's nationwide voter base. The increasingly integrated alliance of the traditional left with the ex-DC of the Margherita is probably important in this southern extension into areas formerly strongly supportive of the Christian Democrats. Of course, Forza Italia and, by 2006, a seemingly fading League have between them successfully colonized large parts of the North. This trend represents the continuing collapse of the old DC hegemony but also the rise of the class of entrepreneurs and their industrial districts that have been the source of much northern economic growth over the past 30 years (Agnew 2002, chapter 7; Golden 2004). The centre-left has not figured out how to appeal to this electorate in national elections because it remains inattentive to the appeals for better infrastructure and lower taxes that are central to regional political discourse (Illy 2006, Panebianco 2006).4 At the same time, however, it is a mistake to draw too bold a line between Forza Italia and League areas in the North, except perhaps that Forza Italia has exhibited greater support in cities, particularly in Milan, but has recently also extended its appeal into the previous strongholds of the League. Only a populist message, however, is likely to have much appeal in the latter areas, and this does not necessarily go down well with either the business orientation that Forza Italia takes in Milan or its appeal to consumerism everywhere else (Agnew et al. 2002; Pasquino 2003). The call of Umberto Bossi, the League's leader, during the 2006 election campaign to introduce protectionist measures to defend small northern businesses against Chinese competition, for example, is unlikely to be well received by the more internationalist segments of Milan's business elite (Girardin 2006). Increasingly, however, the League has turned itself into the harder populist edge of Forza Italia, with its government ministers over the period 2001-2006 often leading the charge for its other more quiescent electoral allies. Thus, in practice, if not perhaps in the minds of the hard core of League voters, the Northern League and Forza Italia have become largely bonded together (Albertazzi & McDonnell 2005). In the South all of the parties, with the obvious exception of the Northern League (apart that is from 2006, when the Lega allied itself with a local notable in the province of Catania, Sicily, Raffaelle Lombardo) are now differentially in play. This may not be the case in Sicily, where Forza Italia and the Union of the Democratic Centre (UDC) have a stranglehold throughout much of the island. But elsewhere, the geographical pattern is fragmented. There is now no single electoral zona meridionale. This perhaps reflects the fact that although the South as a whole continues to experience serious economic disadvantage no single party promises easy resolution of its problems (Cannari & D'Alessio 2003; Cannari & Chiri 2004). Given that all major parties (and some minor ones) are now potentially parties of government, the historic bias of southern voters towards favouring parties of 34 /. Agnew national government now benefits none in particular. Furthermore, the different trajectories of political patronage and opinion voting in different places in the South, exposed most graphically by Simona Piattoni (1999), suggest that different 'souths' will find different parties more or less attractive depending on what is on offer. The South is also less and less a single region economically, with Abruzzo and some metropolitan regions, particularly Palermo and Naples, growing at the same time that other areas, such as much of Calabria, have stalled (Guerrieri & Iammarino 2006). The zona azzurra as a singular 'region' self-destructs under closer inspection. But this does not make it an example of 'politics without territory'. Diamanti is a victim of his 'either there is a region or there is no geography' approach to the understanding of place. His Forza Italia zone is truly a congeries of places with apparently little in common save a faith in at least one of the images portrayed by Berlusconi and Forza Italia. At most, what it might suggest is that in the zona azzurra opinion voting (as supposedly long dominant in parts of the North) and patronage voting (held to be characteristic of the South) have found common cause in the imagery of self-serving materialism offered by Forza Italia (Piattoni 2005). The centre-right government had successive crises in 2004-5 largely because its components, Forza Italia and the Northern League (the so-called northern axis), on one side, and Alleanza Nazionale and the Christian Democrat UDC, on the other, have different local and regional bases (the former in the extreme North and the North-east, the latter in the South) and, consequently, different views about national economic and social policies (Diamanti & Lello 2005). In particular, the more economically developed extreme North typically favours less redistributive economic policies and less intrusive social policies, whereas in the South government is a major employer and there is much greater reliance on government redistributive programmes. But even within these larger regions, for example between the Centre and the North-west, there are fundamentally different public rankings of the importance of issues such as the standard of living, public services, and criminality (Ricolfi 2005, chapter 2). The policy differences between the parties are only at the tip of the remarkable local and regional differences in popular political attitudes across Italy. Conclusion The unification of Italy is still far off, Pasquale Saraceno (1988), the famous economist, averred in 1988 when discussing the economics of the southern question—why southern Italy has systematically lagged behind the north in economic development. It is still true today, electorally as well as economically and not just in terms of a North-South division. In this Italy is not anomalous, even if its two post-1992 electoral systems may be. My main conclusion is that 'nationalization' as a concept confuses vote parallels between districts and areas in how well parties perform with causal similarity with respect to how votes are arrived at in different places. Now, nationalization can be used to describe a trend towards national homogenization of votes across places over a period of time. I have used it this way myself (Agnew 1987, 2002). But nationalization as deterritorialization, in the sense of politics without territory or place in its formulation, would be something Modern Italy 35 else again. I hope I have shown for the Italian national elections since 1992 that 'politics without territory' or place—as suggested both by those invoking the powers of television and celebrity and those claiming the emergence of a model median national voter without anything much of a geographical standard deviation—is a slogan devoid of empirical or theoretical meaning. An Italy there may still well be, in the sense of a state apparatus with a defined territorial extent, but its electoral-geographical divisibility in novel as well as persisting ways remains far from exhausted. Of course, this is one of the things that most fascinates many of us about Italy, notwithstanding the best efforts of those scholars and commentators who would turn it into simply a model case of failed national modernity. Acknowledgements An earlier version of this article was presented as a keynote lecture at the Annual Conference of the Association for the Study of Modern Italy at the Italian Cultural Centre, London, November 2004. Thanks to Simon Parker for inviting me and to Simon Parker, Gianfranco Pasquino, Mark Donovan, and two anonymous reviewers for helpful comments. They are not responsible for what I have made of them. Notes [1] See also Gianfranco Pasquino's article in this issue on the coalition and party building skills of Berlusconi. [2] Notwithstanding much talk about mergers between DS and Margherita on the centre-left and Forza Italia and Alleanza Nazionale on the centre-right not much has actually come of this yet except for the decision to run candidates of the former two in Chamber contests in 2006 under the sign of L'Ulivo, formerly the name of the centre-left coalition as a whole. [3] Although there was a strong correlation between votes in the SMDs and the PR contests, in that there was not a large 'leakage' of votes between one coalition in the former and affiliated parties of the other in the latter, the centre-left tended to perform better than its constituent parties and vice versa for the Polo. In 2001 only 400,000 votes separated the coalitions nationally in the SMDs, while 3.3 million more voters preferred the Polo/Casa parties in the PR ballot (Parker & Natale 2002, p. 669). Much of this 'gap' was due either to more voting by centre-left voters for non-coalition parties in the PR part or to higher rates of abstention on their part in PR contests than shifting across coalitions between the two parts of the election. The 'gap' between the two parts provided much of the incentive for Berlusconi and his allies to change the electoral system in 2005 towards a more proportional system to blunt the centre-left's tendency to do better in majoritarian contests. [4] In his careful empirical analysis of the outcomes of Berlusconi's 2001 'Contract with the Italians,' Ricolfi (2006) shows that the centre-right did not have much justification in campaigning in 2006 as if it had achieved much success in the areas of tax reform and public works during its five years of government. From this viewpoint, voting for the centre-right in 2006 could hardly be construed as validating retrospective voting along economic lines! Indeed, it could even be said that the contract itself was more leftwing, in its emphasis on pensions and job growth, than it was rightwing (in a neo-liberal sense), ignoring completely such matters as privatization of state assets, liberalization of labour markets, and reducing the government deficit (Cazzola 2005). References Agnew, J. (1987) Place and Politics: The Geographical Mediation of State and Society, Allen & Unwin, London. Agnew, J. (1996) 'Mapping politics: how context counts in electoral geography', Political Geography, vol. 15, no. 2, pp. 129—46. Agnew, J. (2002) Place and Politics in Modern Italy, University of Chicago Press, Chicago. 36 /. Agnew Agnew, J., Shin, M., & Bettoni, G. (2002) 'City versus Metropolis: the Northern League in the Milan Metropolitan Area', International Journal of Urban and RegionalResearch, vol. 26, no. 2, pp. 266-83. Agnew, J., Shin, M., & Richardson, P. (2005) 'The saga of the "Second Industrial Divide" and the history of the "Third Italy": evidence from export data', Scottish Geographical Journal, vol. 121, no. 1, pp. 83-101. Albertazzi, D. & McDonnell, D. (2005) 'The Lega Nord in the second Berlusconi government: in a League of its own', West European Politics, vol. 28, no. 5, pp. 952-72. Andreucci, F. & Pescarolo, A. (1989) 'La formazione delle regioni "rossi" in Italia: il caso di Toscana', in Gli spazi di potere. Aree, regioni, Stati: le coordinate territoriali della storia contemporanea, eds F. Andreucci & A. Pescarolo, Usher, Florence, pp. 93-111. Bartolini, S., Chiaramonte, A., & D'Alimonte, R. (2004) 'The Italian party system between parties and coalitions', West European Politics, vol. 27, no. 1, pp. 1-19. Beck, P. A., Dalton, R.J., Greene, S., & Huckfeldt, R. (2002) 'The social calculus of voting: interpersonal, media, and organizational influence on presidential choices', American Political Science Review, vol. 96, pp. 57-73. Bendicenti, D. (2006) 'La televisione e la campagna elettorale permanente', // Mulino, no. 423, pp. 56-63. Bobbio, N. (1996) Left & Right: The Significance of a Political Distinction, Polity, Cambridge. Brenner, N. (1999) 'Beyond state-centrism: space, territoriality, and geographical scale in globalization studies', Theory and Society, vol. 28, no. 1, pp. 39-78. Brooks, C. (2006) 'Voters, satisficing, and policymaking: recent directions in the study of electoral polities', Annual Review of Sociology, vol. 32, pp. 1-21. Cannari, L. & Chiri, S. (2004) 'La bilancia dei pagamenti di parte corrente Nord-Sud (1998-2000)', Temi di Discussione, Banca DTtalia, Rome. Cannari, L. & D'Alessio, G. (2003) 'La distribuzione del reddito e della richezza nelle regioni italiane', Temi di Discussione, Banca DTtalia, Rome. Castagnoli, A. (ed.) (2004) Culture politiche e territorio in Italia, 1945-2000, Franco Angeli, Milan. Cazzola, G. (2005) 'Governo di destra, politiche di sinistra', // Riformista, 21 September. Chiaramonte, A. (2002) 'II voto proporzionale: verso la nazionalizzazione della competizione?' in Maggioritario finalmente? La transizione 1994-2001, eds R. D'Alimonte & S. Bartolini, II Mulino, Bologna, pp. 82-111. Cox, K. E. & Schoppa, L. J. (2002) 'Interaction effects in mixed-member electoral systems: theory and evidence from Germany, Japan, and Italy', Comparative Political Studies, vol. 35, no. 9, pp. 1027-53. Diamanti, I. (2003a) Bianco, rosso, verde... e azzurro: Mappe e colori dell'Italia politica, II Mulino, Bologna. Diamanti, I. (2003b) 'II ritorno dello stato: l'ltalia, dal regionalismo al neocentralismo', in Sistemi locali e spazio europeo, ed. P. Messina, Carocci, Rome, pp. 53-81. Diamanti, I. (2004) 'Tramonta il partito 'azzurro': Forza Italia non unisce piu', La Repubblica, 29 June. Diamanti, I. & Lello, E. (2005) 'The Casa delle Liberta: a house of cards?', Modern Italy, vol. 10, no. 1, pp. 9-35. Ferrara, F. (2004) 'Electoral coordination and the strategic desertion of strong parties in compensatory mixed systems with negative vote transfers', Electoral Studies, vol. 23, no. 3, pp. 391-413. Foucault, M. (1980) Power/Knowledge: Selected Interviews and Other Writings, Harvester, Brighton. Ginsborg, P. (2003) Italy and its Discontents: Family, Civil Society, State, 1980-2001, Penguin, London. Ginsborg, P. (2004) Silvio Berlusconi: Television, Power and Patrimony, London, Verso. Girardin, S. (2006) 'II Senatur a Varese difende le piccole imprese', La Padania, 1 April, p. 1. Golden, M. (2004) 'International economic sources of regime change: how European integration undermined Italy's postwar party system', Comparative Political Studies, vol. 37, no. 10, pp. 1238-74. Goodin, R. E. & Tilly, C. (eds) (2006) The Oxford Handbook of Contextual Political Analysis, Oxford University Press, Oxford. Griswold, W. & Wright, N. (2004) 'Cowbirds, locals, and the dynamic endurance of Regionalism', American Journal of Sociology, vol. 109, no. 6, pp. 1411-51. Guerrieri, P. & Iammarino, S. (2006) 'The rise of "many Mezzogiorni": an empirical assessment of the internal differentiation of Italian southern regions', European Urban and Regional Studies, vol. 13, no. 2, pp. 167-178. Gundle, S. & Parker, S. (eds) (1996) The New Italian Republic: from the fall of the Berlin Wall to Berlusconi, Routledge, London. Hacking, I. (2004) 'Between Michel Foucault and Erving Goffman: between discourse in the abstract and face-to-face interaction', Economy and Society, vol. 33, no. 3, pp. 277-302. Modern Italy 37 Hall, S. (1980) 'Encoding/decoding', in Culture, Media, Language, eds S. Hall, et al., Hutchinson, London. Hilder, P. (2005) 'Open parties? A map of 21st century democracy', Open Democracy, www. opendemocracy.net/articles/viewpopuparticle.jsp?id=3&articleid=2312. Illy, R. (2006) 'Riccardo Illy: piü strade e meno tasse. Solo cosi Prodi puö prendere il Nord', Corriere delta Sera, 13 April. Istituto Cattaneo (2006) Elezioni Politiche 2006. www.cattaneo.org. Katz, E. & Lazarsfeld, P. F. (1955) Personal Influence: The Part Played by People in the Flow of Mass Communications, Free Press, New York. Leonardi, R. (2003) 'The denationalization of Italian politics and policy making', in Italy: Politics and Policy, eds R. Leonardi and M. Fedele, Vol. 2, Ashgate, Aldershot. Magnier, A. (2004) 'Between institutional learning and re-legitimization: Italian mayors in the unending reform', International Journal of Urban and Regional Research, vol. 28, no. 1, pp. 166-82. Mannheimer, R. (2002) 'Le elezioni del 2001 e la "mobilitazione drammatizante'", in Dall'Ulivo al governo Berlusconi. Le elezioni del 13 maggio e il sistema politico italiano, ed. G. Pasquino, II Mulino, Bologna. Mannheimer, R. (2004) 'I flussi elettorali', Corriere delta Sera, 20 June. Mannheimer, R. (2006) Tndeciso un terzo degli elettori', Corriere delta Sera, 14 March. Massey, D. (1999) Power-Geometries and the Politics of Space-Time, University of Heidelberg, Department of Geography, Hettner Lectures 2, Heidelberg. Messina, P. (ed.) (2001) Regolazione politico dello sviluppo locale: Veneto ed Emilia Romagna a confronto, UTET, Turin. Natale, P. (2006) 'Tradimenti e conferme come e cambiato il consenso degli Italiani rispetto alle politiche del 2001', La Stampa, 12 April. Panebianco, A. (2006) 'II Nord lontana dai vincitori', Corriere delta Sera, 13 April. Pappalardo, A. (2002) 'II sistema italiano fra bipolarismo e destrutturazione', in Dall'Ulivo al governo Berlusconi. Le elezioni del 13 maggio 2001 e il sistema politico italiano, ed. G. Pasquino, II Mulino, Bologna, pp. 199-237. Parker, S. & Natale, P. (2002) 'Parliamentary elections in Italy, May 2002', Electoral Studies, vol. 21, no. 4, 665-73 [(2003) Erratum: Parliamentary elections in Italy, May 2001, Electoral Studies, vol. 22, no. 4, p. 763]. Pasquino, G. (2002) // sistema politico italiano. Autoritä, istituzioni, societä, Bononia University Press, Bologna. Pasquino, G. (2003) The government, the opposition and the President of the Republic under Berlusconi, Journal of Modern Italian Studies, vol. 8, no. 4, pp. 485-500. Pasquino, G. (2005) 'Berlusconismo senza Berlusconi?', La Rivista dei Libri, 9 September, www. larivistadeilibri.it/2005/09/pasquino.html. Piattoni, S. (1999). 'Politico, locale e sviluppo economico nel Mezzogiorno', vol. 55, no. 2, pp. 117-49. Piattoni, S. (2005) Review of I. Diamanti, 'Bianco, rosso, verde... e azzuxxo' Rivista Italiana di Scienza Politico, vol. 35, no. 2, pp. 165-67. Poli, E. (2001) Forza Italia. Strutture, leadership e radicamento territoriale, II Mulino, Bologna. Putnam, R. D. (with Leonardi, R. & Nanetti, R.) (1993) Making Democracy Work: Civic Traditions in Modern Italy, Princeton University Press, Princeton. Ramella, F. (2005) Cuore rosso? Viaggio politico nell'Italia di mezzo, Donzelli, Rome. Rampulla, F. C. (1997) 'La legge Bassanini e le autonomic locali', // Politico, vol. 62, no. 4, pp. 131^-2. Reed, S. R. (2001) 'Duverger's Law is working in Italy', Comparative Political Studies, vol. 34, no. 3, pp. 313-27. Ricolfi, L. (2005) Dossier Italia: A che punto e il 'Contratto con gli italiani'?, II Mulino, Bologna. Ricolfi, L. (2006) Tempo scaduto. II 'Contratto con gli italiani' alia prova dei fatti, II Mulino, Bologna. Saraceno, P. (1988) L'unificazione economico italiano e ancora lontana, II Mulino, Bologna. Sartori, G. (1989) 'Videopolitica', Rivista Italiana di Scienza Politico, vol. 19, no. 1, pp. 185-98. Shin, M. E., Agnew, J., Breau, S., & Richardson, P. (2006) 'Place and the geography of Italian export performance', European Urban and Regional Studies, vol. 13, no. 3, pp. 257-70. Street, J. (2004) 'Celebrity politicians: popular culture and political representation', British Journal of Politics and International Relations, vol. 6, no. 4, pp. 435-52. Taddia, M. (2004) 'Italy's broadcast news. Silvio Berlusconi's media empire is extending to radio', Financial Times, 19 October, p. 11. Taylor, M. (2006) Rationality and the Ideology of Disconnection, Cambridge University Press, Cambridge. Thompson, J. B. (2000) Political Scandal. Power and Visibility in the Media Age, Polity Press, Cambridge. Vandelli, L. (2002) Devolution e altre storie. Paradossi, ambiguitä e rischi di un progetto politico, II Mulino, Bologna. 38 /. Agnew Venturino, F. (ed.) (2005) Elezioni e personalizzazione della politica, ARACNE, Rome. Wellhofer, E. S. (2001) 'Party realignment and voter transition in Italy, 1987-1996', Comparative Political Studies, vol. 34, no. 2, pp. 156-86. Wise, P. (2005) 'Southern Europe's airwave: higher television viewerships and lower costs are spurring on growth', Financial Times, 19 April, p. 13. Zinn, D. L. (2001) La raccommandazione. Clientelismo vecchio e nuovo, Donzelli, Rome. Zolo, D. (1999) From 'historic compromise' to 'telecratic compromise': notes for a history of political communication between the First and Second Republics, Media, Culture & Society, vol. 21, no. 6, pp. 727-41. Zuckerman, A. S. (ed.) (2005) The Social Logic of Politics: Personal Networks as Contexts for Political Behavior, Temple University Press, Philadelphia.