is governability a constitutional...

“Elecciones México 2009”

111222333


Is Governability A Constitutional Principle?

Italy’s Electoral System

Massimo Cuono*

[email protected]

ABSTRACT

Governability, intended as effectiveness in decision-making, is a key concept in law, legal theory

and social sciences since the late seventies. In contrast to political representation, governability is

often conceived as the standard for evaluating electoral systems. At the crossroad of «the 1st and

2nd republic of Italy» in the early nineties, the debate on governability was framed in terms of

stability. The aim of this paper is (1) to analyse Italian electoral system in historical perspective in

order to (2) assess the relation between governability and representation. After World War II the

constituent assembly opted for a proportional electoral system, considered highly consistent with

the overall parliamentary system. With the remarkable exception of the so-called «legge truffa» in

1953, this proportional system lasted four decades. In 1993 it was replaced by a single-member

plurality system, making Italy the only advanced democracy in world history to pass from a

proportional to a majoritarian electoral system. A few months prior to the general elections of

2006, the right wing majority in Parliament, strongly supported by the Berlusconi administration,

adopted a new electoral rule: Basically proportional, it includes high thresholds and an additional

seats system that needs to be assessed in a comparative perspective.

And therefore when any one of these pantomimic gentlemen, who are so clever that they

can imitate anything, comes to us, and makes a proposal to exhibit himself and his poetry, we will

fall down and worship him as a sweet and holy and wonderful being; but we must also inform him

that in our city such as he are not permitted to exist; the law will not allow them. And so when we

have anointed him with myrrh, and set a garland of wool upon his head, we shall send him away to

another city.

Plato, Rep. 398a

* Research Fellow, Department of Economics, Institutions and Society, University of Sassari (Italy).

mailto:[email protected]

TTTEEEPPPJJJFFF

111222444

The image which recurs in these and other statement is one of the disintegration of civil

order, the breakdown of social discipline, the debility of leaders, and alienation of citizens. Even

what have been thought to be the most civic of industrialized societies have been held to be prey

to these disabilities, as observers speak of the Vietnamization of America and Italianization of

Britain.

Report to the Trilateral Commission 1

1. Introduction

The evolution of the Italian electoral system offers an exceptionally clear case study for testing the

relation between the classical standard of political representation and that of contemporary

governability practices. The Italian debate on stability that foreclosed the rise and the development

of its electoral reforms is a way to critically assess the supposed positive trade-off between

governability and representativeness. My aim is to show how the rhetoric of governability in the

case of Italy’s electoral reforms should be viewed in connection with the phenomenon of

personalisation of politics, leading up to a subtle but substantial variation in the constitutional

arrangement. Governability thus appears to strengthen executive power, consequently weakening

the branch of government archetypically associated with political representation, i.e. the legislative

power of Parliament.

In the second half of the seventies, the discourse on governability entered the scientific and

political scene. The term was borrowed from the hard sciences and, from the analytical

perspective, may be defined as «the quality of being governable, that is, capable of being

controlled or managed.»2 With the publication of the Trilateral Commission Report, drafted by

Crozier, Hungtington and Watanuki, on the risks involved in the loss of governability, the debate on

theory of democracy underwent a radical turn: Focus switched from the lack of democracy to that

of the excess of democracy.

1 M. Crozier, S.P. Hungtington, J. Watanuki, The Crisis of Democracy: Report on the Governability of Democracies to the

Trilateral Commission, New York University Press, NY 1975, p. 3. 2 I. Janin, Governability, in Encyclopedia of Governance, edited by M. Bevir, Sage, Thousand Oaks (CA) 2007, vol. I, p. 363.


111222555


Samuel Hungtington states this in a crystal clear way:

views as to what constitutes the precise desirable balance between power and liberty, authority

and democracy, government and society obviously differ. In fact, the actual balance shifts from

one historical period to another. Some fluctuation in the balance is not only acceptable but may

be essential to the effective functioning of constitutional democracy. At the same time,

excessive swings may produce either too much government or too little authority. The

democratic surge of the 1960s raised once again in dramatic fashion the issue of whether the

pendulum had swung too far in one direction.3

The economic crises of the seventies spurred the debate on the economic and social

changes brought about in the years immediately preceding the oil crisis. These debates evolved

under the warning sign stressing the efficiency and effectiveness setbacks of government action,

eventually resulting in loss of consensus in the electorate.4

This line of study progressively became hegemonic in the realm of social sciences, as well as

in political debates. From here sprung the alleged need in contemporary societies to develop new

solutions to problems facing democratic states overloaded by claims and unable to properly

answer these.5 It has been stressed that such readings paralleled the rise of neo-liberalism

following the election of Margaret Thatcher in the UK and Ronald Reagan in the US.6

This general tendency of political theory translated in Italy into a debate on institutional

reform concerning both the form of government and the electoral rule. This debate generated a

vast consensus in the public eye on the need to abandon parliamentary government and

3 M. Crozier, S.H. Hungtington, J. Watanuki, The Crisis of Democracy cit., p. 63.

4 R. Rose, G.B. Peters, Can Government Go Bankrupt?, Basic Books, New York 1978.

5 The reference of course goes to F. Hayek, Law, Liberty and Legislatio, Routledge, London 1973. See also R. Rose,

Ungovernability: Is There Fire Behind the Smoke?, in «Political studies», 1979, XXVII, pp. 351-370. The vocabulary of new public management and the notion of governance were taken over from these debates: «To put it in Hegelian terms, we can think of governance as being the synthesis that supersedes both the thesis (legitimacy) and the antithesis (governability). Governance seeks, indeed, to combine the demand for participation and inclusion called for by the legitimacy reading of the social crisis with the demand for autonomy and self-regulation called for by the governability reading. However, it is a false synthesis, since it operates entirely within the governability framework» (B. de Sousa Santos, Beyond Neoliberal Governance: the World Social Forum as Subaltern Cosmopolitan Politics and Legality, in Law and Globalization from Below, Cambridge University press, Cambridge 2005, p. 35). 6 See P. Dardot, C. Laval, La nouvelle raison du monde. Essai sur la société néolibérale, La Découverte, Paris 2009, pp. 278

ff.

TTTEEEPPPJJJFFF

111222666

proportional electoral system. Both of these political arrangements had been chosen at the end of

WW2.

Focus here is on the debate that developed around the electoral reforms. The aim of this

study is to show that the rhetoric of governability, at the roots of the electoral engineering

proposals, impacted determinedly on the evolution of Italy’s political system. The first section will

be dedicated to a brief theoretical reconstruction of the concept of governability (§2). In section

two I suggest an assessment of the soundness of the functional argumentation on governability

that we find in contemporary Italian history, from WW2 until today (§3). In section three, I shall

shed light on the electoral system adopted at the end of 2005 for the election of MPs (§4).

The methodology steams from the understanding that we need to keep together and bridge

the outlooks and toolboxes of political theory, political science and constitutional law in order to

provide satisfactory analysis of the concept of governability, and account for the theoretical

problems inherent in the concept and explain the evolution that brought the notion into the

limelight of science and media alike.

2. Governability as a Conceptual Problem

Of course, the Trilateral Report is not the first attempt to protect governability. The noble

forefathers of the theory of governability may be found in the so-called functional theories of

democracy, and even earlier in the considerations of Max Weber on the risks of «leaderless

democracy.»7 These thinkers addressed the issue of measuring and understanding the efficiency of

different forms of government and their electoral systems.

In a brief and selective overview of the (proto) functional arguments of the 20th century,

mention has to be made, first and foremost, of Weber’s critical remarks on «leaderless

democracy» and his strong preference for plebiscitary solutions8 able to contrast the «excesses» of

7 M. Weber, Politics as a Vocation, in Id., The Vocation Lectures, Hackett, Indianapolis 2004.

8 See G. Roth, Politische Herrschaft und persönliche Freiheit, Surhkamp, Frankfurt a.M. 1987; A. Kalyvas, Democracy and

the Politics of the Extraordinary: Max Weber, Carl Schmitt, and Hannah Arendt, Cambridge University Press, Cambridge 2009, chapter 1.


111222777


proportional representation that give rise to «a parliament of closed, philistine minds, in no sense

capable of serving as a place where political leaders are selected.»9

Secondly, we should not forget the considerations of Ferdinand A. Hermens10 on the fall of

the Weimar Republic, inspired by Weber’s ideas: Hermens criticized the inclusive consequences of

the proportional electoral system that paved the way for the KPD’s and the NSDAP’s access to the

Reichstag.11

The major representative of functional democracy, still remains Joseph Schumpeter. In his

Capitalism, Socialism and Democracy from 1942, where theory of democracy meets theory of

elites, democracy is described as a procedural technique for leadership selection. It is often

underlined that his model is radically opposed to that of his contemporary, Hans Kelsen.12

According to the latter, in fact, democratic government amounts to a set of procedures grounded

on the parliamentary form of government, proportional representation, a system of parties so as

to include progressively the hurdled masses into the political life of the nation.

The relationship between representativeness and governability can thus be reformulated in

terms of a contrast between political equality and social order. The link between governability and

order could be explained through the problem of conflict neutralization: the overload of political

demand could create a lack of efficiency of the institutions that prevents the possibility of

answering with effective decisions and brings to social conflict. A political system that is not able to

neutralize conflict puts in jeopardy the effectiveness,13 i.e. the stability and even the survival, of the

state itself. On the other hand, with political equality we can intend the double democratic

meaning of inclusivity (of larger groups of individuals in the political area) and equal weight of

9 M. Weber, The President of the Reicht (1919), in Id., Political Writings, Cambridge University Press, Cambridge 1994, p.

306. 10

See in particular F.A. Hermens, Europe Between Democracy and Anarchy, University of Notre Dame, Notre Dame (Indiana) 1951; Id., The Trojan Horse of Democracy, in «Social Research», November 1938. 11

There is a vast literature dealing with the topic. See, inter alios, G. Leibholz, Die Auflösung der liberalen Demokratie in Deutschland und das autoritäre Staatsbild, Duncker und Humblot, Leipzig 1933; C.J. Friedrich, Constitutional Government and democracy. Theory and Practice in Europe and America, Ginn, Boston 1950, pp. 275 ff.; R. Fenske, Weimar. Der deutsche Liberalismus in der Krise, Surhkamp, Frankfurt a.M. 1957; H. Schulze, Weimar: Deutschland 1917-1933, Severin und Siedler, Berlin 1982. 12

For a reconstruction of Kelsen’s theory of democracy and its links to law, see Lars Vinx, Hans Kelsen's Pure Theory of Law: Legality and Legitimacy, Oxford University Press, Oxford 2007. 13

In this case effectiveness is used in the meaning of Hans Kelsen’s General Theory of Law and State (1945), The Lawbook Exchange, Clark NJ, 2007, pp. 121 ff.

TTTEEEPPPJJJFFF

111222888

every vote (as the French revolution claim of one vote per member, rather than one vote per

estate).14

The classic theory of democratic government defended the instauration of an equalitarian

order (from the point of view of equal participation of individuals in political decision making). It

defended an order of inclusiveness by the means of a progressive extension of franchise. On the

contrary, democratic functionalism advocated the idea of independence of the order guaranteed

by the oligarchies’ constant control over government and democratic equality as a source of

legitimacy spurring the competition among elites.

With the Trilateral Report, the persuasiveness of governability strongly prevailed in the

scientific debate as well as in the political debate. This led to the claim that there are desirable

limits to political democracy. Dardot and Laval, for instance, hold that the three authors of the

Report «se plaignaient de l’‘excès de démocratie’ apparu dans les années 1960, c’est-à-dire, à leurs

yeux, de la montée des revendications égalitaires et du désir de participation politique active des

classes les plus pauvres et le plus marginalisées».15 This is the core meaning of the thesis according

to which there is a trade off16 between governability and representativeness; a thesis endorsed by

Hungtington to whom it is patent that «to assume that there is no conflict between these two

requirements is sheer self-delusion.»17

The trade-off thesis (the growth of political equality entails the decrease of governability,

with risks for the social order18) is today common sense; it denies the classic thesis of democratic

supporters, according to whom the increase of political equality (in the double sense of inclusivity

14

See N. Bobbio, Dall’ideologia democratica agli universali procedurali (1987), in Id., Teoria generale della politica, Einaudi, Torino 1999, p. 381; M. Bovero, La democrazia e le sue condizioni, felstivalfilosofia, Modena 2009, pp. 7-10. 15

P. Dardot, C. Laval, La nouvelle raison du monde, cit., p. 278. 16

For a well-argued overview of the major pros and cons of the different approaches to electoral systems, see the third chapter in P. Norris, Electoral Engineering. Voting Rules and Political Behaviour, Cambridge University Press, Cambridge, 2004. 17

M. Crozier, S.H. Hungtington, J. Watanuki, The Crisis of Democracy, cit., p. 63. 18

On the relationship between governability and order, Crozier stresses in the Trilateral Report that «order and efficiency may be more surprising items to put among the core political beliefs of West Europeans. *…+ Whenever the development of freedom threatens to bring chaos, the demand for order is immediate, even violent. It is not a lost or dwindling part of core political beliefs whatever the possible evolution of its forms in the direction of more tolerance. The special West European form of order, however, has a more social and less juridical connotation that in the United States. Things (and people) have to be put in their proper place for society to operate. Due process is not the cardinal element of this belief. Furthermore, efficiency may be added to it inasmuch as it has a legitimating connotation of a well-functioning society. West Europeans still value the good ‘efficient’ scheme more than the concrete result. Order is the burden of the white man; efficiency may be the demonstration of it in a modern rationalized society» (M. Crozier, S.H. Hungtington, J. Watanuki, The Crisis of Democracy, cit., p. 45).


111222999


and equal weight to every vote) could lead to social stability by allowing participation of the mass

to political decisions.19

This is exactly what Hungtington denies when he claims that consolidation of democratic,

liberal and equalitarian values, during the sixties, leaded US to a dangerous overload of political

demand (especially because of African-American’s integration in politics and welfare policy

increase).20

From then on, just a few years after the clue date of 1975, Claus Offe speaks of a real «neo-

conservative crisis theory.»21 According to Offe, criticizing democracy was a prerogative of

Marxism in the aftermath of WW2 but this new crisis-theory emerged out of the void left by the

great leftwing movements. The buzzwords are no longer class-conflict and uncompleted

democratic inclusiveness, but rather the crisis of mass-democracy. The agenda setting moves from

the issues focused on the clash between political democratic equality and market economy. The

new agenda politicized the lack of proportion between citizens’ claims on western democracies

and the scarce response offered by governments.22 In the wording of the sociologist George

Lakoff, the conservatives imposed the frame23 of their worldview:24 A «too equal» society (or a

society that looks too much to the value of equality) would somehow become impossible to rule.

19

See the documents of the Association réformiste (Genève 1865), Proportional Representation League (New York 1867), Proportional Representation Society (London 1869), Associazione proporzionalista (Roma 1871). 20

M. Crozier, S.H. Hungtington, J. Watanuki, The Crisis of Democracy, cit., pp. 112-113. 21

C. Offe, ’Ungovernability’: the Renaissance of Conservative Theories of Crisis, in Id., Contradictions of the Welfare State, MIT Press, Cambridge (Mass.) 1984. 22

Recently, Boaventura de Sousa Santos claimed that, with the Trilateral, «the crisis of government by consent was thereby transformed into a crisis of government tout court, and the crisis of legitimacy became a crisis of governability» (B. de Sousa Santos, Beyond Neoliberal Governance, cit., p. 33). 23

As Erving Goffman explains in Frame Analysis: An essay on the organization of experience, a frame, in social and communication theory, consists of a schema of interpretation that individuals rely on to understand and respond to events. It relates to the construction and presentation of a fact or issue «framed» from a particular perspective. Framing is an effective heuristic, i.e. mental shortcut or cognitive bias, affecting the outcome of choice problems to the extent that several of the classic axioms of rational choice do not hold. See S. Plous, The Psychology of Judgment and Decision Making, McGraw-Hill, Columbus 1993; A. Tversky, D. Kahneman, The Framing of Decisions and the Psychology of Choice, in «Science», 1981, 211, pp. 453-458. In social sciences a frame defines the packaging of an element of rhetoric in such a way as to encourage certain interpretations and to discourage others: According to some sociologists, the «social construction of collective action frames» involves «public discourse, that is, the interface of media discourse and interpersonal interaction; persuasive communication during mobilization campaigns by movement organizations, their opponents and countermovement organizations; and consciousness raising during episodes of collective action» (B. Klandermans, The Social Psychology of Protest, Blackwell, Oxford 1997, p. 45). 24

G. Lakoff, Don't Think of an Elephant!: Know Your Values and Frame the Debate, White River Junction, Chelsea Green 2004.

TTTEEEPPPJJJFFF

111333000

3. Governability and Electoral Reform in Republican Italian Debate

In Italy – as in the rest of the (Western) world – the end of the seventies and the beginning of the

eighties marked a turn in the imposition of the frame of governability. This became particularly

clear in the debates over institutional and electoral reforms. Nonetheless, the effects impacted

exclusively the field of electoral rule, since the form of government and the institutional

arrangement are regulated under the 1948 Italian constitution including aggravated mechanisms

for amending the constitution (ex art. 138) and the possibility of confirmative popular

referendum.25

The «electoral issue» has for a long time been the favourite playing field of the affirmation of

democratic functionalism. This was facilitated by a certain liberal tradition and the entrenched and

well-rooted tradition of elitism going back to the pre-fascist era; two traditions of thought that had

elaborated the classical themes in the criticisms of proportional representative democracy. It shall

come as no surprise that, while the Germans were discussing the motives behind the fall of the

Weimar Republic using similar arguments, in Italy, numerous studies were published attempting to

establish a connection between the introduction of the proportional electoral system in Italy in

1919 and the consecutive rise of fascism.26

3.1. 1943-1984: Proportional Representation and Parliamentary System

At the end of WW2, a vast consensus is formed in Italy around the parliamentary form of

government and the proportional electoral system within the institutional system as well as in

society at large. In this particular historical period, we could claim that the frame of representation

clearly prevails.

The resistance movement and the post-fascist political organisation in the liberated

territories were dominated by mass parties that had been forced into clandestine existence during

the fascist regime. These parties were organised in the Committee of National Liberation. The

25

The most blatant case is the constitutional reform that was strengthening the powers of the prime minister, elected by the majority of a centre-right dominated parliament in 2005 and then abolished by referendum in June 2006. Contrarily to abrogative referendums (ex art. 75), so-called constitutional referendums (ex art.138) do not require the quorum of 50%+1. 26

G. Maranini, Storia del potere in Italia (1848-1967), Vallecchi, Firenze 1967, p. 283.


111333111


important mass-parties (especially the Christian Democratic Party,27 the Socialist Party28 and the

Communist Party29) gain significant representation in the Constituent Assembly,30 following the

victory of the advocates of the Republic form of State over the monarchists in the referendum of

the 2nd of June 1946. In fact, the Constituent power opted for the proportionality grounded

electoral rule on the basis that it was the most compatible arrangement within the framework of

parliamentary government.31 This was evidently a matter of political preference32 motivated by

the aim to protect and include minorities in government processes. The preference also favoured

the consolidation of a system of mediation between the great mass-parties, i.e. the essential

components of the democratic form of government as depicted in Hans Kelsen in Von Wert und

Wesen des Demokratie (1929).

Even in the Constituent Assembly, however, there were voices endorsing a functionalist

view, especially among the liberals. Suffice to say that the great Italian advocate of liberalism and

first President of the Republic, Luigi Einaudi, defended the introduction of a system similar to that

in use in the UK. His arguments were based on a criticism of the lack of «personalisation of

politics»: in other words, his claim was that focus had to be on single leaders at the vertex of the

decision-making process. He also criticized the growing influence of parties in this process.33

Indeed, the criticism of the mass-party system34 is a Leitmotiv in functionalist discourses

condemning proportional arrangements.35 There are good reasons to believe that these influential

voices within the Constituent Assembly are, in all likelihood, one of the motives behind the fact

that the Assembly did not choose to include the proportional electoral rules in the constitutional

27

From here on DC. 28

From here on PSI. 29

From here on PCI. 30

The sum of the ballots cast in favour of the three parties reached 75% of votes. 31

For a detailed reconstruction of the use of the proportional system in Italy, see C. De Fiores, Rappresentanza politica e sistemi elettorali in Italia, in Id. (ed.), Rappresentanza politica e legge elettorale, Giappichelli, Torino 2007. 32

See M. Luciani, Il voto e la democrazia. La questione delle riforme elettorali in Italia, Editori Riuniti, Roma 1991, especially p. 27. 33

L. Einaudi, Contro la proporzionale (1944), in id., Il buongoverno. Saggi di economia e politica, Laterza, Roma-Bari 2004. On the Italian statesman, see Roger Chartier, Luigi Einaudi – between politics and history, Ithaca, New York 1988. 34

See S. Lupo, Partito e antipartito. Una storia politica della prima Repubblica (1946-78), Donzelli, Roma 2004, where the so-called «Particracy argument» is analysed historically and in relation with the fascist ideological heritage. 35

Besides Einaudi, a significant name in the debates in Italy in those years was Giuseppe Maranini and in particular his work Miti e realtà della democrazia, Comunità, Milano 1958. In the field of political science, in turn, the import of functionalist theories should be credited to Giovanni Sartori that defended many of Schumpeter’s central theses as early as in his 1957 book Democrazia e definizioni. For Sartori, the true power of the electorate consists in choosing who will rule.

TTTEEEPPPJJJFFF

111333222

dictate. Hence, the proportional electoral system was relegated to the realm of technicalities,

disciplined under ordinary legislation.

The political arrangement turned Italy into a case of «double majority» systems. This system

ruled a historical situation marked by great hostility between the Catholics and the Communists, in

the cold war international context of a world divided in two blocs, on the verge of falling back into

war. This led to an extremely stable government, dominated by one and the same party for over

forty-years. This situation was tempered by an even larger majority (the so-called «arco

costituzionale») needed for constitutional amendments that had to be built between the big mass-

parties which never really questioned the basic choices made by the founding father. The

Communists were excluded from the government, but were necessary for changing the

Constitution.

The failed constitutionalization of the proportional electoral system brought on a first

modification in 1953 known as «Legge truffa»,36 the swindle law. Just a few months after the 1953

general elections, the majority, under the leadership of the DC and its undisputed leader Alcide De

Gasperi, approved a law according to which a significant bonus would be conferred to the party (or

coalition) with 50%+1 votes. The electoral reform later became one of the central issues that the

opposition parties (such as the Communist and the Socialist) campaigned on. To the opposition,

the new law had the only effect of strengthening the already influential power of the DC. In the

following general election, no party attained the notorious threshold of the bonus, leading to the

abolition of the law in the next legislature.

However, the «swindle law» does not yet belong to the governability-centred context of

later electoral reforms. Looking at the debates from the 50’s, there is no trace of accentuation of

leadership abilities or plebiscitary rhetoric’s. Rather, we face a mix of the «Weimar syndrome», i.e.

the fear of assisting passively to the growth of anti-system parties such as the PCI, and the desire to

strengthen the centrist power that had until then dominated the political scene of post-WW2

Italy.37

36

Law 148/1953. See A. Renwick, The Politics of Electoral Reform. Changing the Rules of Democracy, Cambridge University Press, 2010, pp. 113 ff. 37

C. De Fiores, Rappresentanza politica e sistemi elettorali in Italia, cit., pp. 35-38.


111333333


3.2. 1976-1994: The Crisis of Democracy?

The mid-seventies38 turned out to be the breeding ground for the winds of change: until then, the

electoral system, rooted in the proportional arrangement, had been shielded by all major political

parties. From this point on, nonetheless, the debate drifted towards the new buzzword:

«governability». Around this date, the term acquired two different meanings. The first meaning is a

particularity of the Italian debate and indicates stability of the executive branch of government.

The second meaning encountered a greater dissemination, making headlines in other countries as

well. This second meaning stresses the loss of representativeness of mass-parties: These become

the target of censure since they are held to be interest groups with the mere goal of partitioning

political power.

Besides the aforementioned cultural climate dominating international relations,39 the

political life in Italy was irremediably scored by the tragic impact of extremist violence, i.e. left-wing

as well as right-wing terrorism. This explosion of terror culminated in the kidnapping and murder of

the Christian Democrat’s leader Aldo Moro by the Red Brigades in 1978. Moro had been the

advocate of the agreement known as the «historic compromise» between the DC and the PCI, that

first proposed it. After Moro’s kidnapping, Communists aimed to guarantee «national solidarity» in

the worst days of the terror emergency. Some voices in the public opinion, nonetheless,

understood this period in rather different terms: in fact, some saw this as the boosting of political

«consociativism»; a category which would later serve to grasp the entire history of the Republican

era.40

This diffuse perception was reinforced by the successive rise of a new political player in Italy

that would leave an important sign on the 80’s. The socialist leader Bettino Craxi strengthened the

PSI as the Third political party in Italy, making it an alternative to both the major parties: the

Communist Party hitherto the opposition party par excellence allied with the PSI during the Fifties

and the Christian Democrats, hitherto dominating player on the Italian scene, even though, starting

38

See S. Tarrow, Democracy and Disorder: Protest and Politics in Italy, 1965-1975, Clarendon Press, Oxford 1989. 39

It should be stressed that the majoritarian turn started back in 1975 when the important Italian constitutional lawyer and legal theorist Costantino Mortati published a famous commentary on art. 1 of the Italian Constitution where he uncovers all main arguments for abolishing the proportional system that had been elaborated within the functionalist tradition. See C. Mortati, Art. 1, in Commentario della Costituzione, ed. by G. Branca, Zanichelli, Bologna-Roma 1975. 40

For a critical outlook on the issue, see M. Revelli, Politica italiana: le avventure del ‘consociativismo’, in «Teoria politica», 1994, 2, pp. 7-28.

TTTEEEPPPJJJFFF

111333444

in 1963, it shared executive power with the Socialist. Already in his management of the terror

crisis, Craxi emerges as the voice offering an alternative to the two other big parties’ intransigence

in dealing with the terrorists and this soft line will meet with the favour of Moro himself who

thanked Craxi in a letter from his prison.

Moreover, Craxi’s Socialist Party presents itself as a new movement, strongly linked to its

leader and marked by strong advocacy for the functionalist interpretation of democratic

government that had until then been marginalized in the scientific debate. This is the starting point

of the raging debates over the institutional and electoral reforms that will peak with the reforms of

1993. Actually, the first attempt to reform the institutional arrangement goes back to 1984 when

the “Bozzi Commission” tried, unsuccessfully however, to open the public eye to the need to

reform the system depicted as immovably stuck. In its Italian version, the argument for endorsing

the notion of governability hinged on the call for a popular claim to stability.41 Clearly, such a claim

has to be understood in the frame of a political system that frequently changed his government,

even though the majority behind each of this government was constantly leaded by the DC since

1948.

A feedback loop between the public and the scientific debate took hold, due to the

intervention of political scientists, historians and legal scientists in the public debate. These were

the glory days of the Italian political scientist Gianfranco Pasquino, that, on the basis of the

functionalist theory of Giovanni Sartori, claims that the time was ripe for giving back to the citizens

the power to select its own leader while criticizing the partitioning of power and negotiation

practices in use among mass-parties.42 Moreover, we should mention the prestigious Italian

constitutional lawyer Augusto Barbera43 and the catholic historian Pietro Scoppola that will go as

far as to call for «the end of party-based democracy.»44 Even the President of the Republic,

Francesco Cossiga, declared himself sceptical of the party system. These were also the years when

the party system went into a deep crisis following a strain of judiciary inquiries: the so-called

41

In the entry Governabilità written for the Treccani encyclopedia, Gianfranco Pasquino defined governability as «political stability + decision-making efficiency». See G. Sartori, Comparative Constitutional Engineering: An Inquiry into Structures, Incentives and Outcomes, Macmillan, London 1994, ch. VI, § 6. 42

G. Pasquino, Restituire lo scettro al principe: proposte di riforma istituzionale, Laterza, Roma-Bari 1985. 43

A. Barbera, Una riforma per la repubblica, Editori Riuniti, Roma 1991. 44

P. Scoppola, La repubblica dei partiti: evoluzione e crisi di un sistema politico 1945-1996, il Mulino, Bologna 1997, p. 436.


111333555


scandal of tangentopoli. This strain of investigations put high-level officials, including Craxi himself,

on the bench of the accused. Many of these officials had, just like the socialist leader, made a

number of criticisms to the party-based system45, often blamed for the dysfunction of the

proportional electoral system. In this context, two referendums were held on the abrogation of

significant clauses in the electoral system.

The first referendum was held in 199146 and the second was held in 199347, leading to the

new electoral law approved in the same year. The laws n. 276 and n. 277 dating August 4th 1993,

radically impacted the electoral system by introducing a mixed system48 for both chambers, in

which 75% of MPs are elected through a single-member plurality system while only 25% are

elected through a proportional system of lists elaborated previously by the parties without any

possibility for expressing individual preferences for specific names. The system also includes a

threshold of 4%49 for the votes aggregated at the national level.

3.3. 1994-2006: The so-called «2nd Republic»

The 1994 general elections are commonly referred to as the start of the phenomenon that goes

under the French-sounding appellative of «second Republic» that paved the way for the

overturning of the traditional parties following corruption charges and different investigations

(tangentopoli). This is the period in which the frame of governability emerged as the hegemonic

frame for grasping the political life of Italy. In this phase governability appeared to be the key

unlocking the doors of «democracy of alternation»50 and bi-polarism, i.e. a two party (or two

45

On the criticism of the party-system in the Italian socialist movement led by Craxi, see A. Mastropaolo, Antipolitica. All’origine della crisi italiana, l’Ancora, Napoli 2000, pp. 63 ff. 46

The referendum on the provision reducing the number of preferences to a single one in the ballot electing the Chamber obtained 95,6% of the votes. This provision had been defended on the basis of arguments such as the need to combat «consociativism» among parties, i.e. power-sharing arrangements involving guaranteed group representation; and the very Italian phenomenon of «voto di scambio», i.e. the ballot cast regularly by a voter unmotivated by political preferences but rather motivated by the do ut des logic involving promised favours of personal kind. 47

82,7% voted in favour of the referendum introducing the first past the post system for electing the Senate. 48

The Italian political scientist Giovanni Sartori defined this system Mattarellum, a latinization of the surname of its inventor. For a brief synopsis and analysis of this system, see A. Renwick, The Politics of Electoral Reform cit., pp. 169 ff. 49

The systems for electing the two chambers differed as far as the modalities of attributing the proportional quota were concerned and in respect of the dimension of the electoral districts: the Italian Constitution explicitly declares that the Senate is elected on a regional basis (ex art. 57). 50

See M. Bovero, Democracia, alternancia, elecciones, Instituto Federal Electoral, Mexico City, 2000.

TTTEEEPPPJJJFFF

111333666

coalition) political system.51 Bi-polarism was considered to be the only viable option for

«normalizing» Italy’s stormy political relationships and turning it into something «European».52

On the 1994 campaign trail, the Christian Democrats appeared weakened and split, unable

to form a strong centrist grouping. A name change made the split more evident: The party

changed name to Popular Party, hereby recovering its pre-fascist appellative. The Communist Party

changed its name to the Left Democratic Party, letting the traditional symbol of “the scythe and

the hammer” in a tiny corner of their logo. Notwithstanding internal divisions, this latter party was

the only one that maintained stability in the level of consensus within the electorate. The Socialist

Party was basically swept away from politics.

The great novelty of these general elections was the businessman Silvio Berlusconi’s decision

to (in his own wording, with a soccer metaphor) «enter the playing-field» (discesa in campo). He

was to become the undisputed protagonist of the next fifteen years of Italy’s political life.

Berlusconi appeared to be a successful businessman, an outsider of the political establishment,

even though he had been intimately linked to the Socialist leader Craxi. Openly hostile to

traditional parties53, Berlusconi emerged as a tough promoter of an unprecedented approach to

political debate that fluctuated between efficiency-focused rhetoric —characteristic of many so-

called post-ideological approaches following the 1989 fall of the Berlin wall— and a strong thrust

towards an uncompromising friend-enemy opposition to communism through the constant

evocation of its ghost and the rediscovery of rhetorical tools belonging to the cold war period. This

technique gave him the spotlight, transforming him into a charismatic leader of the first big

leadership-based party in Italy.54

Now, the rise of the so much talked about bi-polarism was supposed to offer the key to

efficient government at the prize of representativeness. This led to the definitive abandonment of

51

For a brief synopsis about Italian case, see G. Sartori, Il bipolarismo non si uccide, in «Il Corriere della Sera», 21/11/2007. 52

See S. Ceccanti, S. Vassallo (eds.) Come chiudere la transizione. Cambiamento, apprendimento e adattamento nel sistema politico italiano, il Mulino, Bologna 2004. 53

About anti-party arguments, not only in Italy, see the monographic issue of the «European Journal of Political Research», XXIX, 3, 1996, edited by S. Scarrow and T. Poguntke. 54

M. Calise, Il partito personale. I due corpi del leader, Laterza, Bari-Bari 2010². At the first ballot using this new system (the general elections of March 1994), Berlusconi’s Forza Italia imposes itself as the biggest party on the national level with its 21%. On Berlusconi’s style of government in a comparative perspective with the later French case of Sarkozy, see B. Cousin, T. Vitale, De Porto Rotondo à Wolfeboro. Vertus et faux-semblants de la comparaison Sarkozy-Berlusconi, in «Mouvements», 2007, 52, pp. 105-113.


111333777


the «double majority system». On one hand, this enabled the heirs of the Communist Party to

enter government. On the other hand, it brought anti-system parties —such as the re-styled ex-

fascists (MSI) and the recently founded regionalist movement Lega Nord— into the democratic

game of the nation.

The change of the electoral law occurred in parallel with the perception of charismatic

leadership becoming a source of legitimacy within the democratic context.55 Throughout the entire

post-fascist era, this source of legitimacy had been deemed incompatible with democratic

government. Unsurprisingly, these changes to the political climate had significant consequences on

the form of government. Even though the Constitution had not been amended, the relationship

between the Parliament and the executive took a new turn. The substantial strengthening of

executive power in the name of governability,56 with the consecutive sacrifice of the legislative

branch and its representative function, is not a phenomenon restricted to Italy.57 The process of

legitimizing the leadership58 seems inversely proportional to the loss of credibility of the party-

system. This latter organisation, in effect, was increasingly held to be inefficient and unable to

mediate between society and political institutions even though parties still provided the main

mechanism for selecting candidates for high-official positions.59

Nevertheless, the system of alternation of power between right- and left-wing forces that

was strengthened and promoted by the plurality voting system60 did not have the desired effect of

reducing the number of parties. On the contrary, the number of parties increased, conferring to

55

As early as 1991 an influential Italian intellectual called for a strong leader using arguments taken from Weber’s «plebiscitary democracy». Back then a small minority would endorse such calls. See A. Panebianco, Plebisciti e democrazia, in «il Mulino», 1991, n. 3, pp. 427-435. 56

Bovero recently said that «political life in real democracies seem today just a competition, with few and too little democratic rules, among few people called leaders, who look for investiture of a nearly autocratic power. We are playing another game, this is a different form of government: I suggest to define it electoral autocracy» (M. Bovero, La democrazia e le sue condizioni, cit., p. 16). 57

See T. Poguntke, P. Webb, The Presidentialization of Politics. A Comparative Study of Modern Democracies, Oxford University Press, Oxford 2005. More specifically on the Italian case, see M. Calise, Terza repubblica. Partiti contro presidenti, Laterza, Roma-Bari 2006. 58

One of the technicalities introduced to personalize electoral competitions is the insert of leader names on the parties’ logos. 59

Mastropaolo defined this as a passage from «Kelsenian parties» with strong representative and mediating functions to the «Schumpeterian parties» focused on electoral competition, candidate selection, interest protection. See A. Mastropaolo, Crisi dei partiti o decadimento della democrazia?, in www.costituzionalismo.it accessible on-line at: http://www.costituzionalismo.it/articolo.asp?id=173. 60

Of the three ballots cast with the first-past-the-post system, the centre-right coalition led by Berlusconi won twice (in 1994 and in 2001) and the centre-left coalition led by Romano Prodi won once (1996).

http://www.costituzionalismo.it/

http://www.costituzionalismo.it/articolo.asp?id=173

TTTEEEPPPJJJFFF

111333888

very small aggregations of ballots the power of vetoing important policies. This kind of dysfunction

were explained by several Italian political experts as an output of the proportional quota of 25% by

which some MPs were elected. In an attempt to abolish this proportional quota, two further

referendums were held in 1999 and in 2000 but did not lead to any change since they did not

attain the required quorum.61

4. A Brand New Electoral System

In December 2005, after five years of government and three months before the general elections,

the centre-right majority passed a bill on the electoral system,62 despite the protests of the entire

opposition. In the phrasing of its proponents, this bill would have struck a balance between, on one

hand, democratic representativeness, defended by the smaller parties that would have preferred

to return to the proportional system of the early days, and, on the other hand, governability that

would be guaranteed through the bonus-adjusted system and the threshold.

The new system is a strongly counter-tempered proportional rule. The lists of candidates for

both chambers can be turned into a coalition, so as to increase the likelihood of obtaining the

bonus. As far as the election of the Chamber of Deputies is concerned, the coalition that obtains

the highest number of ballots at the national level automatically obtains 55% of the seats in the

Chamber, regardless of how many votes the coalition as such actually obtained. As far as the

Senate is concerned, the bonus is delivered on a regional basis. This implies that in every region the

winning coalition brings home 55% of the seats reserved in the Senate for the respective region. In

both cases, the lists are drafted by the parties. This means that, once the number of seats

corresponding to each list is established, in every electoral district, the MPs will be selected on the

basis of their position on the lists. The bill also allows plural candidatures: A candidate may present

himself or herself in several electoral districts.

The threshold limits are also of varying nature. As far as the Chamber is concerned, only

parties that have obtained at least 4% of the ballots participate in the distribution of seats. In the

event of a coalition obtaining 10% or more of the votes, all the parties belonging to that coalition

61

In 1999 the quorum was not reached by a very slight margin since 49,6% of the voters participated. In 2000 only 32,2% participated. 62

Law 270 from the 21st of December 2005.


111333999


that have obtained more than 2% will enter Chamber, as well as the first party below the 2%

threshold. The system for electing the Senate is similar but the calculation is made at the regional

level with an overall 8% threshold. For the lists of a coalition totalling 20% or more votes, the

threshold is lowered to 3%.

Finally, the electoral system is not homogenously applied throughout the country: special

clauses are applied to the region of Trentino Alto-Adige, only for the Senate, and for both

chambers to the region of Valle d’Aosta and the collegio estero, i.e. the so-called «foreign district»

granting extraterritorial representation of Italians living abroad.63 In fact, this electoral district was

formed following the constitutional amendments of the 17th of January 2000 (n.1) and of the 23rd

of January 2000 (n.1) that conferred franchise to all Italians who have taken up permanent

residence abroad and are inscribed at the AIRE (Anagrafe per Italiani Residenti all’Estero). These ex

pats are entitled to select by ballot 6 senators and 12 MPs on the basis of macro-districts.

In the general elections held on the 9th of April 2006 the centre-left coalition was granted a

broad majority in Chamber but obtained a much smaller majority in the Senate.64 The executive

led by Romano Prodi – among the most crowded administrations in the history of the Republic65 –

was able to hold on to power for a little less than two years. Thereafter, the general elections

brought Berlusconi once again to the top of the executive. As a matter of fact, the Prodi

administration fell because of the obstructionism of the tiny party Udeur when it became public

domain that Clemente Mastella, leader of the Udeur party and Minister of Justice, was being

subject, together with his wife, of an investigation including corruption among the charges.66 The

Udeur party, that had obtained a mere 1,39% of the votes, enjoyed some 14 MPs and 3 senators in

the Prodi administration.

Moreover, the investigation of Mastella added up to the uncritical plea for the existing

electoral rule that was being challenged by three new referendums. However, none of them will

63

On the electoral mechanisms, campaign trails, voters’ profile etc. for the «foreign district», see G. Tintori, Fardelli d’Italia? Conseguenze nazionali e transnazionali delle politiche di cittadinanza italiane, Carocci, Roma 2009, especially pp. 118 ff. 64

Pursuant to the Italian Constitution, government has to enjoy the trust of both chambers (ex art. 94). 65

In this administration there were 26 ministers, 10 voce-ministers, 66 undersecretaries, totalling 103 members from 11 different parties to which we need to add other 11 parties that supported it without partaking in it. 66 See http://www.lastampa.it/redazione/cmsSezioni/politica/200801articoli/29316girata.asp.

http://www.lastampa.it/redazione/cmsSezioni/politica/200801articoli/29316girata.asp

TTTEEEPPPJJJFFF

111444000

reach up to the quorum of 50%+1 and they were therefore declared null and void.67 Even if we set

aside the particular circumstances of Italy’s political life, what is of interest with the new electoral

law is the discrepancy between the reasons behind it and its consequences that were not only

predictable but also explicitly foreseen.

Here I suggest four different criteria for assessing the law 270/2005 that was defined by its

promoter – the minister Roberto Calderoli68 – an «obscenity» (porcata). This appellative made the

law go under the infamous name of Porcellum; latinization of the Italian swearword «swine»

coined by the Italian political scientist Giovanni Sartori.69

(a) Political suitability

A first observation that we should make is that it is neither appropriate, nor suitable to change the

electoral law just a few months ahead of the general elections. To make matters worse, it is clearly

out of place to pass such a bill in a parliament dominated by a majority fearing to be relegated to

minority. To change the rules of the game while playing, without due agreement between players,

is blatantly unfair, showing lack of respect for the principle of legality as such.

Even from the point of view of the governability frame, this element is incompatible with the

«democracy of alternation» that the advocates of institutional change thought to be the output of

bi-polarism or bi-partisanship.

(b) The rule of law

Following the 2006 general elections, the law known by the name of its main promoter –

Calderoli – triggered a lengthy lawsuit concerning the attribution of some seats in the Senate. In

fact, one of the clauses of the provision can be read in multiple ways: This is relevant for the

67

These three issues, even though they left the law basically untouched, would have eliminated the privileges guaranteed to small parties by standardizing the required thresholds to 4% in the lower house and 8% in the upper house. A few months before the government crisis, Mastella declares at a press conference in Naples: «I am saying it clearly: we will not be there when the referendum comes. If there is the referendum, the risk is a government crisis *…+ once governability is guaranteed the rest is boredom *…+ Before the brutality of those who want a referendum just because things are not working we respond with equally strong determination». He resigned, officially because of lack of solidarity in the government following the scoop of the investigation on him, on the same day the supreme court declared the referendum admissible (16

th of Jan 2008). In this Judgments (15, 16 and 17/2008) the Court censured, as an

advise for the Parliament, several elements of the Law. 68

Source: TV interview Matrix, 15th

of March 2006. 69

G. Sartori, Il Porcellum da eliminare, in «Il Corriere della Sera», 1/11/2006.


111444111


election of senators in a region where the coalition obtains more than 55%. This technical flaw had

been stressed before the law came into force: During the debate in the Commission for

Constitutional Affairs (Commissione affari costituzionali) in the Senate. The majority MPs preferred

not to modify the text but rather to speedily carry on with the reform. The rule of law

(Rechtssicherheit) was thus sacrificed on the altar of speediness.

Clearly the concern for legal security and rule of law is not a priority in political debate and it

is often left to the lawyer to defend the gubernaculum per leges.70 Yet, the problems that arose

because of the technical flaws of the law, drafted in a clumsy and rushed manner, had weighty

consequences on the institutional efficiency, contrarily to what governability advocates had hoped.

(c) Democratic representativeness

The system though which the votes are transposed into seats in the Calderoli Act does not only

result in a deformation of the political preferences expressed by voters at the level of Parliament.

Such a deformation has been portrayed as the prize to pay for efficiency and governability. With

this law, indeed, the logical connection between the distribution of votes and the composition of

the two chambers is lost.

The absence of a minimum amount of votes required accessing at the bonus ends up in the

paradox according to which 55% of the seats will automatically be attributed to the party or

coalition that obtained more votes regardless of the amount of votes in excess. This, in turn, means

that the misrepresentation is far worse than in the previous and highly criticised «Legge Acerbo»

that Mussolini coveted in 1923, and the aforementioned «Legge truffa» from 1953. In effect, the

first of these laws included a provision making the majority party obtaining automatically 2/3 of

the seats in Parliament, in the event that party, however, would win at least 25% of the votes. The

second law granted 65% of the seats to the coalition of the majority in the event it won 50%+1 of

the votes.

In addition to this, the threshold system introduces different thresholds. This system enables

small parties belonging to important coalitions to be granted a distorted weight. In fact, it is

paradoxical that, in the Senate, a party attaining 3% may be enter while another obtaining 7,9%

70

All classical theories of rule of law defended the characteristics of generality and abstraction that aim to guarantee the equality before the law and the possibility for the citizens to foresee the legal consequences of their acts.

TTTEEEPPPJJJFFF

111444222

may not, depending on which coalition it belongs to. A party having obtained merely 410.499 votes

is currently in the Italian Parliament, while another having gained 1.124.298 votes was excluded.

Another issue, perhaps even more worrisome even if less noticed, concerns the

autonomous region of «Valle d’Aosta». In the lower house, this region is entitled to one single

deputy elected with a first-past-the-post voting method. The ballots cast by the citizens, resident in

Aosta Valley, are the only ones that are not counted when it comes to attributing the bonus on the

national level. This is highly problematic because the provision stands in stark contrast to the

constitutional norm guaranteeing equal weight to every vote (ex art. 48).71

Finally, the extraterritorial vote given to Italian residents abroad through the system of an

extraterritorial constituency (collegio estero) deepens the gap between Italians living abroad and

legally resident migrants living in Italy but lacking franchise, i.e. taxpayers subject to a legal and

political system which they have no means of influencing. Guido Tintori, among others, stressed

this aspect: «The combined effect of the Italian laws on citizenship, the electoral law enabling

extraterritorial ballot casting, and the 2006 electoral system have engendered a vicious circle in

legislation and politics from which it will not be easy to escape. The electoral regulation makes it

highly unlikely that those voting from abroad are Italians temporarily living abroad. On the

contrary, holders of an Italian passport, third or fourth generation emigrants are automatically

registered on the lists of voters of the AIRE, so they receive the ballot package directly in the

mailbox. Those elected in the «foreign district», in turn, are often holders of dual nationalities and

the Italian state does not require them to give up the foreign citizenship like elsewhere (...). To

paraphrase: this can be summed up with no-taxation and overrepresentation.»72

These problems tally up to the mechanism of the lists of candidates being drafted by the

parties, a necessary requirement in an arrangement with macro-constituencies. The overall result

is a frustrating stalemate for those that intended to promote governability and spoke up

condemning the mass party system. The current mechanism, indeed, enables very high

predictability of the outcome of the ballot before it is even cast.

71

Another problem of potential unconstitutionality is the introduction of a provision obliging any party to declare his prime minister candidate, even though it is not binding. In the Italian constitution, however, the President of the Republic, with the consecutive approval of both chambers, chooses the prime minister that is thus not elected directly by the people (ex art. 92). 72

G. Tintori, Fardelli d’Italia, cit., pp. 118-119.


111444333


(d) Governability as a form of stability

The arguments presented here would not offer a complete picture if we do not address the

question whether the law guarantees at least the stability of government (which is considered the

most serious flaw of Italian democracy). Since figures do not lie it is often stressed that from 1948

until today Italy has had 58 administrations and 24 prime ministers.

Moreover, it should be noted that the bonus-adjusted system translates into very different

and heterogeneous majority coalitions in the lower and in the upper house. This leads to the risk of

polarization within each chamber without nonetheless guaranteeing that a consensus will be

found on the candidate for the prime minister.

Furthermore, the threshold system was intentioned to safeguard majorities from the

blackmailing ability of small parties. This hoped effect has been cancelled through the

differentiation between the distinct thresholds and the introduction of the bonus-adjusted system.

These two systems have contrasting effects that cancel each others’ potentiality. In fact, the lists

drafted by the important parties have to include top leaders from tiny parties in order not to

disperse the ballots. This was the mechanism through which the centre-left Democratic Party

pulled-in 9 MPs from the Radical Party on their list.

It seems evident to the scholars analyzing the electoral rule that the aim of the law simply

has not been attained.

5. Conclusions

The survey of the Italian electoral system, from its post-war emergence until today’s provisions,

that I have presented in this paper, enables us to underline three aspects that are worthy of further

discussion:

(i) What incidence does dominating political rhetoric have on the rules of the democratic

game if these are framed in the procedural democracy tradition?73

73

As far as the procedural theories of democracy are concerned, besides the aforementioned Hans Kelsen and Joseph Schumpeter, see N. Bobbio, Dall’ideologia democratica agli universali procedurali, cit.; L. Ferrajoli, Principia iuris, Teoria del diritto e della democrazia, Laterza, Roma-Bari 2007, vol. II.

TTTEEEPPPJJJFFF

111444444

(ii) To what extent has governability approaches concurred in the «crisis of democracy» and

the erosion of traditional political parties, opening the door to populist or strongly anti-

political movements?74

(iii) If we consider governability as a quasi-constitutional principle what do we make of the

traditional democratic theories focusing on mass-involvement and citizen-participation in

decision-making processes, the foundation of a civil covenant truly different in nature

from that of autocracies?75

74

See P. Rosanvallon, La contre-démocratie: La politique à l'âge de la défiance, Seuil, Paris 2006; E. Laclau, On Populist Reason, Verso, London 2005; Y. Mény, Y. Surel, Par le peuple, pour le people, Fayard, Paris 2000; A. Mastropaolo, La mucca pazza della democrazia, Bollati Boringhieri, Torino 2005; L. Ferrajoli, Rappresentanza politica e organicismo para-democratico, in «Democrazia e diritto», 2003, n. 3, pp. 57-62. 75

See in particular M. Bovero, Contro il governo dei peggiori. Una grammatica della democrazia, Laterza, Roma-Bari 2000.


111444555


La participación política y la reforma electoral

Alberto Ricardo Dalla Via*

a) Participación política

La reforma constitucional de 1994 incorporó el capítulo II bajo el título “Nuevos derechos

y garantías”, ampliando de esta manera el catálogo clásico de derechos a los derechos de

incidencia colectiva o “nuevos derechos”. Frente a los aportes del constitucionalismo

clásico que limitaba los derechos constitucionales a los “derechos subjetivos”, el nuevo

capítulo incorpora los derechos que no están en cabeza de un sujeto sino que se tienen

por pertenencia a un grupo amplio de personas. De esa manera, a los derechos civiles y

políticos (primera generación) y a los derechos económicos y sociales (segunda

generación), se agregan ahora también los derechos colectivos (tercera generación),

entre los que la Constitución enumera los derechos al ambiente y los derechos de los

usuarios y consumidores.

El capítulo también contiene los derechos electorales y de los partidos políticos

que, si bien se consideran clásicamente de primera generación, se ubican en el nuevo

capítulo, ya que no fueron contemplados expresamente en la Constitución histórica,

aunque la doctrina y la jurisprudencia ya los habían recogido a partir de la norma de

habilitación del art. 33 con un criterio amplio. Luego se ubican los llamados nuevos

derechos de participación política, como la iniciativa popular y la consulta popular. De ese

modo, la Constitución acentúa el tránsito desde un modelo individualista y representativo

a un modelo democrático que también contempla la participación como valor. Los

derechos de primera generación se fundaron en la libertad como valor, los de segunda

generación lo hicieron en la igualdad y los de tercera generación lo hacen en la

solidaridad. Son, en definitiva, derechos de la participación.

A los derechos de participación política propios de la democracia participativa y

deliberativa me referiré a continuación, tratando de resaltar los aspectos más novedosos

* Presidente de la Cámara Nacional Electoral. Profesor Titular de Derecho Constitucional (UBA).

Presidente de la Asociación Argentina de Derecho Constitucional. Académico de Número de la Academia Nacional de Cs. Morales y Políticas.

TTTEEEPPPJJJFFF

111444666

que generan un amplio debate social y político, y de la participación ciudadana en los

procesos electorales, para finalmente describir los lineamientos de la última reforma

electoral.

Los derechos políticos son una categoría no siempre concisa, que abarca los

derechos de asociación y reunión, de peticionar a las autoridades, de participación y

control, así como el derecho de elegir y ser elegido conforme a las leyes.

Éstos se titularizan en sujetos que tienen calidad de ciudadanos o calidad de

entidades políticas reconocidas. Los derechos políticos sólo tienen por finalidad la

política.

Mucho se habló de los derechos políticos, y más en países como el nuestro que

vimos cercenado su ejercicio, por lo que es necesario tener presente cómo se ha ido

extendiendo la participación, tanto en el sufragio activo como pasivo y la ampliación de la

protección judicial.

Desde el retorno de la democracia, uno de los temas relevantes fue la protección

de las mujeres en su derecho de sufragio pasivo, lo que trajo aparejado la sanción de la

ley 24.012, llamada de “cupo femenino” o “cuota de género”, fijando la obligatoriedad de

incluir un mínimo de 30% de mujeres en las listas de candidatos para elecciones

nacionales. La reforma constitucional de 1994 incorporó el artículo 37 de la Constitución

Nacional que en su segundo párrafo señala que el sufragio es universal, igual, secreto y

obligatorio —recogiendo así la tradición en materia electoral que arranca desde la Ley

Sáenz Peña (núm. 8871)— y la última parte del mismo artículo favorece la adopción de

acciones que tiendan progresivamente a la igualdad real de oportunidades entre varones

y mujeres para el acceso a los cargos electivos y partidarios.

La Constitución Nacional manda asegurar mediante la implementación de

“acciones positivas” en los textos de los arts. 37 y 75 inc. 23 la igualdad real de

oportunidades entre varones y mujeres para el acceso de cargos electivos.

En este punto, nuestro país ha seguido los principios consagrados en la Convención

Americana sobre Derechos Humanos, en el Pacto de Derechos Civiles y Políticos y en la

Convención contra toda forma de Discriminación de la Mujer, que en materia electoral y


111444777


de partidos políticos se pronuncian claramente a favor de una participación igualitaria y

sin discriminaciones fundadas en meros prejuicios de género.

La Cámara Nacional Electoral ha asumido de un modo cabal, el rol que se le ha

asignado de garante del cumplimiento de las medidas que procuran la igualdad real de

oportunidades entre mujeres y varones para el acceso a cargos electivos y partidarios (cf.

Fallos CNE 1568/93; 1586/93; 1595/93; 1863/95; 1866/95; 1867/95; 1868/95; 1869/95;

1870/95; 1873/95; 1984/95; 2669/99; 2878/01; 2918/01; 3005/02 y 3780/07) y veló por

su respeto en todas las causas que le fueron sometidas a su conocimiento (cf. Fallos

3005/02 y 3780/07).

En particular, ha sostenido que no basta que las listas estén compuestas por un

mínimo de 30% de mujeres sino que además es necesario que tal integración se concrete

de modo que –con un razonable grado de probabilidad– resulte su acceso a la función

legislativa en la proporción mínima establecida por la ley y aquél sólo puede existir si se

toma como base para el cómputo la cantidad de bancas que el partido renueva (cf. Fallos

CNE 1566/93; 1836/95; 1850/95; 1862/95; 1864/95; 1866/95 y 3507/05).

La reciente reforma electoral, ley 26.571,1 establece el derecho de participación

política de las mujeres al establecer que las agrupaciones políticas para su organización

interna deben respetar la ley 24.012, “cupo femenino” (art. 3º, inc. b). Con ello se

consagra la extensión de los derechos políticos de las mujeres, dado que ahora no sólo se

requiere respetar el porcentaje mínimo por sexo en la conformación de las listas para

cargos públicos electivos sino también para las elecciones internas de los partidos.

Otra modalidad de extensión del sufragio activo es la que establece la ley 24.007,

que permite a los argentinos residentes en el exterior votar en los comicios nacionales. En

este punto, se ha sostenido además que respecto a esos ciudadanos “debe procurar

facilitarse *…+ el ejercicio de otros derechos que, como el de afiliación, integra también el

plexo de los derechos políticos ciudadanos propios de toda democracia” (cf. Fallo CNE

1756/94).

1 Sancionada el 02/12/2009 y promulgada por decreto 2004/2009 del 11 de diciembre de 2009 con la

observación de los artículos 107 y 108 (B.O n° 31.800, 15/12/2009).

TTTEEEPPPJJJFFF

111444888

Otro caso fue la admisión del sufragio activo de los privados de libertad sin

condena, que a partir del “caso Mignone” ejercen su derecho sin restricción en las

penitenciarias donde se encuentran alojados.

En este sentido, el fallo Mignone2 del año 2000 hizo lugar a un amparo promovido

por un organismo no gubernamental, el Centro de Estudios Legales y Sociales (CELS) para

declarar la inconstitucionalidad de un artículo del Código Electoral Nacional que excluía

del sufragio activo a quienes se encontraban privados de libertad sin condena en distintos

establecimientos carcelarios del país. De ese modo, se hizo valer el inciso 2° del artículo

23 de la Convención Americana de Derechos Humanos, que limita la potestad de

reglamentación legal de los derechos de participación política “…exclusivamente por

razones de edad, nacionalidad, residencia, idioma, instrucción, capacidad civil o mental, o

condena por juez competente en proceso penal”.

La misma sentencia reconoció legitimación activa a una asociación representativa

de intereses de incidencia colectiva en general, conforme las califica el artículo 43 de la

Constitución Nacional al consagrar el llamado amparo colectivo.

A su vez, en la causa “Zárate, Marcelo Antonio s/amparo”3 del año 2003, la Cámara

Nacional Electoral se volvió a pronunciar sobre el tema destacando que

la privación del ejercicio del sufragio para los ciudadanos que se encuentren en esta

condición procesal, importa vulnerar el principio de inocencia que se encuentra ínsito en el

artículo 18 de la Constitución Nacional y expresamente previsto en los artículos 8°, párrafo

2° de la Convención Americana sobre Derechos Humanos y párrafo 14, inciso 2° del Pacto

Internacional de Derechos Civiles y Políticos, efectuándose así una discriminación

arbitraria… No cabe sino concluir entonces que la restricción de acceder al acto electoral,

impuesta al recurrente por su condición de procesado, constituye un trato incompatible

con el respeto debido a la dignidad inherente al ser humano.

Finalmente, con anterioridad a la sanción de la ley 25.858, frente al incumplimiento

por parte del Poder Legislativo y Ejecutivo de adoptar las medidas necesarias para hacer

2 Fallos CNE 2807/2000 y Corte Suprema de Justicia Fallos 325:524.

3 Fallos CNE 3142/2003.


111444999


efectivo el derecho a votar de los detenidos no condenados como se les intimó en el caso

Mignone, entendió que "corresponde hacer lugar a la acción deducida comunicando al

Tribunal Oral en lo Criminal Federal N° 1 de La Plata –a cuya disposición se encuentra el

recurrente– que deberá arbitrar los medios a su alcance a fin de hacer efectivo [su

derecho]". En el año 2006 se reglamentó, finalmente, el ejercicio del sufragio activo de los

procesados con prisión preventiva.

Un tema de mucha actualidad, en los últimos tiempos, es el ejercicio del sufragio

activo de las minorías desprotegidas y desamparadas que se encuentran en situaciones

de vulnerabilidad, que hace que determinados “personajes” encuentren un campo fértil

para ejercer lo que se ha denominado clientelismo político.

Las prácticas clientelares —entre las que se encuentra la llamada compra de

votos— conspiran contra la expresión de libre voluntad que constituye un presupuesto

indispensable del ejercicio del sufragio. El concepto general de clientelismo político está

acotado en nuestra sociedad a una mera permuta de favores entre jefes partidarios y

potenciales electores provenientes en su mayoría de clases bajas y desamparadas.

Estas prácticas vulneran el ejercicio de los derechos políticos, ya que cuando los

instrumentos internacionales4 hacen referencia al voto secreto que garantice la libre

expresión de la voluntad de los electores, procuran resguardar al sufragante de toda

intimidación, pues la libertad del voto conlleva inexorablemente el derecho de cada

elector de expresar su voluntad sin ser objeto de presión alguna.

En las últimas elecciones, la Cámara Nacional Electoral, ha dictado un

pronunciamiento, en la causa "Sublemas del Acuerdo Cívico y Social de Formosa

s/protesta”5 a fin de salvaguardar los derechos políticos de las comunidades indígenas

frente a distintas maniobras, como la retención de documentos cívicos, que se habían

venido evidenciado en distintas elecciones, tendientes a distorsionar la libertad cívica de

ciudadanos pertenecientes a comunidades indígenas de distintas localidades de la

provincia de Formosa. Allí, se ha establecido que "los hechos denunciados involucran a un

4 Convención Americana sobre Derechos Humanos, artículo 23, inc. 2º y Pacto Internacional de

Derechos Civiles y Políticos, artículo 25, inc. b y en sentido análogo, Declaración Universal de Derechos Humanos, artículo 21, inc. 3º y Declaración Americana de los Derechos y Deberes del Hombre, artículo XX.

5 Fallo CNE 4283/09.

TTTEEEPPPJJJFFF

111555000

gran número de integrantes de pueblos originarios, quienes enfrentan serias dificultades

que los mantienen en una situación de vulnerabilidad y marginalidad" (cf. Fallo CNE

4283/09). Lo que implica "el establecimiento de determinadas medidas con el fin de

procurar en la práctica la igualdad de oportunidades que permitan corregir las situaciones

que son el resultado de conductas discriminatorias" (cf. Fallo cit.).

No es ocioso recordar al respecto que la Corte Interamericana de Derechos

Humanos ha enfatizado que

[e]n lo que respecta a pueblos indígenas, es indispensable que los Estados otorguen

una protección efectiva que tome en cuenta sus particularidades propias, sus

características económicas y sociales, así como su situación de especial

vulnerabilidad, su derecho consuetudinario, valores, usos y costumbres (cf. caso

Comunidad indígena Yakye Axa Vs. Paraguay, sentencia de 17 de junio de 2005).

Como lo ha establecido la Corte Interamericana de Derechos Humanos

el artículo 23 de la Convención no sólo establece que sus titulares deben gozar de

derechos, sino que agrega el término ‘oportunidades’. Esto último implica la

obligación de garantizar con medidas positivas que toda persona que formalmente

sea titular de derechos políticos tenga la oportunidad real para ejercerlos. [...] [E]s

indispensable que el Estado genere las condiciones y mecanismos óptimos para que

los derechos políticos puedan ser ejercidos de forma efectiva, respetando el

principio de igualdad y no discriminación (caso Castañeda Gutman vs. Estados

Unidos Mexicanos, Sentencia del 6 de agosto de 2008).

Los derechos políticos propician el fortalecimiento de la democracia y el pluralismo

político, ya que "el ejercicio efectivo de los [mismos] constituye un fin en sí mismo y, a la

vez, un medio fundamental que las sociedades democráticas tienen para garantizar los

demás derechos humanos previstos en la Convención" (cf. Corte Interamericana de

Derechos Humanos, caso Casteñeda Gutman, cit.).


111555111


Por ello,

[e]l derecho al voto es uno de los elementos esenciales para la existencia de la

democracia y una de las formas en que los ciudadanos ejercen el derecho a la

participación política. Este derecho implica que los ciudadanos puedan elegir

libremente y en condiciones de igualdad a quienes los representarán (cf. Caso

Yatama vs. Nicaragua, sentencia de 23 de junio de 2005).

Vale aclarar al respecto que cuando en los instrumentos internacionales se hace

referencia a voto secreto que garantice "la libertad del voto" (art. 21 de la Declaración

Universal de Derechos Humanos) o "la libre expresión de la voluntad de los electores"

(art. 25 del Pacto de Derechos Civiles y Políticos y 23 de la Convención Americana de

Derechos Humanos) lo que se procura es poner al sufragante al abrigo de toda

intimidación. El principio de libertad del voto significa que cada elector debe poder

sufragar sin ser objeto de presión alguna (cf. Fallo CNE 2534/99).

Como lo ha señalado Habermas "[e]l problema se plantea en las sociedades

democráticas cuando la cultura mayoritaria políticamente dominante impone su forma

de vida y con ello fracasa la igualdad de derechos efectiva, de ciudadanos con otra

procedencia cultural".6 Y al decir de Dworkin

[l]a teoría constitucional sobre la cual se basa [un] gobierno [democrático] no es una

simple teoría mayoritaria. La Constitución [...] esta destinada a proteger a los

ciudadanos, individualmente y en grupo, contra ciertas decisiones que podría

querer tomar una mayoría de ciudadanos, aun cuando esa mayoría actúe siguiendo

lo que para ella es el interés general o común.7

Tales prescripciones se enmarcan en una concepción progresiva de los derechos

fundamentales que no sólo requieren del Estado una posición de mero garante neutral o

abstencionista, sino que le encomienda remover los obstáculos para hacer

6 Habermas, Jürgen. 1999. La inclusión del otro, estudios de la teoría política, Barcelona: Paidós, ,p. 123.

7 Dworkin, Ronald. 2002. Los derechos en serio. España: Ariel, p. 211.

TTTEEEPPPJJJFFF

111555222

verdaderamente efectiva la realización de tales derechos, en el caso, a la participación

política.

En cuestiones institucionales de relevancia, la Cámara Nacional Electoral, ha

mantenido posturas aperturistas en materia de legitimación cuando se han visto

involucrados derechos fundamentales de participación política. En tal sentido, la

jurisprudencia registra la amplitud de legitimación en los casos de “cuota de género” o

“cupo femenino” receptando su cumplimiento para los cargos electivos al momento de

oficializarse las candidaturas, y en la admisión de la acción de amparo colectivo, como en

su momento fue el caso Mignone.

Es así como el Tribunal le reconoció legitimación activa a una ciudadana con el

fundamento de que "[s]i la lista de un partido no se ajusta a lo que marca la ley, no

solamente la está violando sino que también está restringiendo y vulnerando ese derecho

del sufragante que nace de ella y que tiene por tanto raíz constitucional" (cf. Fallos CNE

1836/95).

En cuanto a los partidos políticos, el artículo 38 de la Constitución Nacional los

considera como instituciones fundamentales del sistema democrático, la jurisprudencia

ya había admitido su inserción constitucional sobre la base de los arts. 1°, 14, 22 y 33 de la

Constitución Nacional. El nuevo artículo incorpora también varios postulados del derecho

político: la necesaria forma de organización y funcionamiento interno con contenidos

democráticos; la representación de las minorías en el gobierno de los mismos; la

“competencia” para proponer candidaturas para la ofertas electorales a cargos públicos

electivos; y el derecho a la difusión de sus ideas a través de los medios junto con el acceso

a la información pública.

La norma constitucional, a su vez, trata lo relativo al financiamiento de los partidos

políticos. La solución alcanzada en la Convención Nacional Reformadora equilibra la

responsabilidad primaria del Estado al sostenimiento económico de sus actividades y a la

formación de sus cuadros y la obligatoriedad de dar publicidad del origen y destino de los

fondos privados recibidos y del patrimonio de los partidos políticos. En este sentido, se ha

sancionado la ley de financiamiento de los partidos políticos (primero la 25.600 y, luego

26.215, modificada recientemente por la ley 26.571) que alienta la participación activa de


111555333


la ciudadanía en el proceso de control de los fondos. Dicha participación no puede

considerarse agotada en la circunstancia de que los terceros puedan tomar conocimiento

sobre la procedencia y el destino de los fondos, a efecto de hacerlos más transparentes

propendiendo a lo que se ha dado en denominar el voto informado del elector, sino

también permitiéndoles colaborar en el proceso, al admitirles las observaciones que

presentaren sobre las posibles anomalías que –a su juicio– detectaren sobre los estados

contables. Tal actitud guarda adecuada coherencia con la concepción más participativa

del sistema democrático, conforme resulta de las distintas normas agregadas a la

Constitución en la reforma de 1994 (arts. 36 a 40 Constitución Nacional).

Otro aspecto de la participación activa de la ciudadanía está dada en el proceso de

oficialización de candidaturas, y esto puede verse en la Acordada CNE Nº 32/09 que

estableció la publicación en el sitio de internet del fuero electoral de las listas de

candidatos a los efectos de que los particulares o el representante del Ministerio Público

Fiscal puedan someter a los magistrados las cuestiones que entiendan relevantes a tal fin.

Recientemente, y ante los últimos precedentes de la Corte Suprema en materia de

legitimación activa, la Cámara en la causa “Barcesat”8 le reconoció legitimación para

promover acción de amparo a un elector de la ciudad de Buenos Aires tendiente a evitar

que participen en los comicios legislativos del 28 de junio de 2009 aquellos que estén

ejerciendo cargos públicos electivos sin renunciar con anterioridad a los mismos. En tal

sentido ha señalado que

más allá de la legitimación que quepa conferir al recurrente a la luz de lo resuelto

recientemente por la Corte Suprema de Justicia de la Nación en la causa H. 270. XLII.

‘Halabi, Ernesto c/P.E.N. - ley 25.873 - dto. 1563/04 s/amparo ley 16.986’, resulta

pertinente recordar que esta Cámara ha expresado que la etapa de registro de

candidatos y oficialización de listas (cf. artículos 60 y 61 del Código Electoral

Nacional) reviste especial trascendencia *…+, pues tiene por objeto comprobar que

los ciudadanos propuestos por las agrupaciones políticas reúnen las calidades

constitucionales y legales requeridas para el cargo que pretenden *…+.

8 Fallos CNE 4156/2009.

TTTEEEPPPJJJFFF

111555444

b) Participación política en la Reforma Electoral

Entre los temas más novedosos, no puedo dejar de mencionar la extensión de los

alcances del derecho de sufragio activo que establece la ley 26.571, recientemente

sancionada, en cuanto instaura la participación de todos los ciudadanos en los procesos

de selección partidaria de las candidaturas –elecciones primarias– que luego van a

competir en la elección de las máximas autoridades públicas nacionales: presidente y

vicepresidente, diputados nacionales y senadores nacionales, así como también

parlamentarios del Mercosur.

Tras la crisis de 2001, el legislador ya había establecido –con la ley 25.611– un

régimen de elecciones internas abiertas, obligatorias y simultáneas de todos los partidos

políticos que actúan en el orden federal. Sus magros resultados, por todos conocidos,

obedecieron en parte a una cadena sucesiva de vetos y normas modificatorias, y otro

tanto al escaso apego de la mayoría de los dirigentes partidarios, como de los afiliados y

los ciudadanos independientes, a este sistema que se aplicó por primera vez en las

elecciones legislativas de 2005, ya que en el año 2003, el Congreso la había suspendido

por el término de un año. En su única aplicación, sólo 23 agrupaciones políticas –

distribuidas en 15 distritos– de las 260 que intervinieron en los comicios generales

llevaron a cabo efectivamente el proceso de elección interna abierta. En las restantes

agrupaciones se proclamó una única lista presentada. Modalidad ésta que expresamente

admitía la reglamentación dispuesta por el Poder Ejecutivo. Así, más de 90% del total de

los partidos políticos eludió el mecanismo de las internas abiertas, que finalmente fue

derogado en diciembre de 2006.9

El nivel de participación ciudadana, que fue de apenas 5% del total del padrón,

tampoco demostró grandes expectativas de los electores en el ejercicio del derecho de

sufragio en el contexto partidario.

Hay que sincerar las cosas: tanto la ley 25.611, de internas abiertas, obligatorias y

simultáneas de los partidos políticos, como la ley 25.600, de control de financiamiento

partidario, fueron dictadas por el Congreso en plena crisis, cuando retumbaban en los

9 Ley 26.191 del 27/12/06.


111555555


pasillos del palacio del Congreso, del Palacio de Justicia y los palacios ministeriales, los

ruidos de las cacerolas. Pero una vez acallados tales ruidos, se notó menor entusiasmo en

los grupos legislativos que sancionaron tales leyes en persistir en su cumplimiento.

La reciente ley 26.571 quizás tenga otros efectos; no por la diferencia de contexto

en el que fue sancionada, sino por los distintos elementos que caracterizan al régimen

que ella regula, y que la distinguen mucho del anterior.

En primer término, se prevé la obligatoriedad del voto como rige en la elección

general, por lo que es de esperar un nivel de participación ciudadana de relevancia,

aunque probablemente no alcance los niveles de una elección general.

Luego, el sistema es de elección abierta plena. Es decir, todos los ciudadanos, sin

distinción entre afiliados e independientes, podrán votar a cualquiera de los

precandidatos de cualquiera de los partidos contendientes.

Podría decirse que este esquema permite calificar a las nuevas internas abiertas

como unas elecciones verdaderamente primarias, en el sentido de que el mismo cuerpo

electoral llamado a votar a sus representantes acude primero a las urnas para

preseleccionar a los candidatos partidarios.

A este respecto, la ley dispone incluso que "para las elecciones primarias se utilizará

el mismo padrón que para la elección general en el que constarán las personas que

cumplan dieciocho (18) años de edad a partir del día de la elección general". Como se

advierte –más allá de la deficiente redacción– los jóvenes que tengan 17 años el día de la

elección primaria podrán votar si cumplen la mayoría de edad antes o el mismo día de

la elección general.

También se modifica el criterio del sistema anterior, que autorizaba a las

agrupaciones a proclamar una “lista única” –generalmente acordada en la cúpula

partidaria– y evadir, así, su participación en el acto electoral.

En el nuevo sistema, los partidos, confederaciones o alianzas que tengan

únicamente una precandidatura deben no obstante conseguir que sea votada por una

cantidad mínima de ciudadanos (1.5% del total de votos válidos emitidos en el distrito)

para que pueda participar en los comicios generales.

TTTEEEPPPJJJFFF

111555666

Como se adelantó, este nuevo régimen se diferencia en mucho del aplicado en

2005. Veremos, en las elecciones del año 2011, si sus efectos se distinguen en la misma

medida de los que dejó aquella experiencia.

c) Otros rasgos de la Reforma Electoral

Como ya lo señalara la ley 26.571, denominada “Ley de Democratización de la

Representación Política, la Transparencia y la Equidad Electoral, instauró el sistema de

elecciones primarias, abiertas, simultáneas y obligatorias, pero también contiene

modificaciones a la Ley Orgánica de Partidos Políticos (núm. 23.298), a la Ley de

Financiamiento de los Partidos Políticos (núm. 26.215) y al Código Electoral Nacional.

Previamente a considerar los aspectos técnicos de la reforma es inevitable advertir

que el diseño institucional que propone, en materia de administración electoral,

profundiza la injerencia del Poder Ejecutivo en áreas verdaderamente sensibles. Esto es

cuanto menos paradójico, si se tiene en cuenta que en el continente se ha venido dando

el proceso inverso –excluyendo por completo al Ejecutivo de la administración de los

procesos electorales–, con la ley propuesta no sólo no se reducen sus atribuciones sino

que, por el contrario, se le reconocen nuevas.

Es así como, en la ley se resalta el rol de la Dirección Nacional Electoral del

Ministerio del Interior, ya que las escasas menciones a esa dirección que contenía la

normativa electoral (arts. 18 y 73 de la Ley de Financiamiento –n° 26.215– y art. 105 del

Código Electoral Nacional) en el nuevo texto se incrementa (vgr. arts. 24, 25, 40 y 75 de

las modificaciones al Código Electoral Nacional; arts. 35, 40, 43, 43 bis, 43 ter, 43 septies,

67 y 71 bis de las modificaciones a la Ley de Financiamiento; y en materia de Elecciones

Primarias, arts. 19, 32, 35, 37, 43 y 104).

La reforma, entre otras disposiciones, no contiene la creación del cargo de fiscal de

Cámara siendo necesario por la especificidad del fuero contar con dicho cargo como

todas las Cámaras.


111555777


Agrupaciones políticas

Respecto a los partidos políticos establece un reconocimiento provisorio (partidos en

formación) a los que se le solicitaba los mismos requisitos ya receptados en la Ley

Orgánica de Partidos políticos, núm. 23.298 (acta de fundación y constitución que

acrediten 4% de adhesiones; nombre; declaración de principios y programa o bases de

acción política; carta orgánica; acta de designación de autoridades promotoras; domicilio

partidario y designación de apoderados); mientras dure esta situación no podrán

presentar candidaturas para elecciones primarias ni nacionales, así como tampoco

podrán recibir aportes (art. 7).

A los efectos de obtener la personería definitiva tendrán que presentar a los 150

días (5 meses) el 4% de afiliados del registro electoral del distrito correspondiente –hasta

el máximo de 1,000,000– con copia de los documentos10 y realizar a los 180 días (6

meses) la elección de las autoridades definitivas (cf. art. 7 bis).11 En este aspecto, la

cláusula transitoria del art. 108 difería la vigencia de esta norma al 31 de diciembre de

2011, sin embargo, la misma fue observada por el decreto de promulgación 2004/2009.

En el caso de los partidos nacionales, la ley no reflejó el proyecto del Poder

Ejecutivo que establecía como requisito para la obtención de la personería jurídico-

política la acreditación de la suma total de afiliados en todos los distritos donde tenga

reconocimiento de 1% del total de los inscriptos en el Registro Nacional de Electores (cf.

art. 3, propuesta de modificación al art. 8). Tal propuesta estaba en colisión con nuestro

sistema federal, ya que en aquellas provincias donde el número de electores es inferior

no podrían constituir los partidos reconocidos en éstas una agrupación nacional.

En el caso de las alianzas electorales se permite que los partidos de distrito que no

formen parte de un partido nacional puedan integrar una alianza con al menos un partido

político nacional (cf. art. 10). En este punto, se contradice con la jurisprudencia de la

10

Cf. Fallo CNE 3997/08 que estableció el requisito de la fotocopia del documento a fin de hacer efectiva la fiscalización por parte del juez contenida en el art. 23 inc. “b” de la presente ley que establece que la justicia electoral deberá “comprobar la identidad con la libreta de enrolamiento, libreta cívica o documento nacional de identidad” ya que “la pureza del padrón partidario constituye una de las más relevantes garantías de la indispensable adecuación de la organización interna de las agrupaciones políticas al sistema democrático pues asegura el efectivo ejercicio de los derechos de elegir y ser elegido que ostentan los afiliados”. 11

El proyecto del Poder Ejecutivo requería para constituirse en partido de distrito el 5% de afiliados al momento de su solicitud de reconocimiento (art. 1, modificación al art. 7 inc. a).

TTTEEEPPPJJJFFF

111555888

Cámara (Fallo 3110/03 y otros) que estableció que únicamente pueden constituir alianzas

nacionales los partidos de orden nacional y alianzas de distrito los partidos distritales,

fundándose en que los partidos integrantes de una alianza tienen que tener la misma

aptitud para postular los cargos que postula por medio de la alianza. Se incorpora,

también, que al momento de solicitar su reconocimiento (60 días antes de la elección

primaria) deben acompañar el “acuerdo del que surja la forma en que se distribuirán los

aportes correspondientes al fondo partidario permanente” (cf. art. 10). En este sentido, el

decreto reglamentario 93/2010 señala la forma en que se realizará el mismo “refiriéndola

exclusivamente” a la cantidad de afiliados reconocidos ante la Justicia Federal con

competencia electoral de cada partido al momento de celebrarse el acuerdo o un

porcentaje igual o diferente para cada uno de los partidos que la integran (cf. art. 9°).

Y para el caso que quieran continuar funcionando, luego de la elección general, en

forma conjunta, los partidos que integran la alianza deberán conformar una

confederación (cf. art. cit.).

En el supuesto de confederaciones, la modificación a la Ley Orgánica de Partidos

Políticos establece que dos o más partidos pueden constituir confederaciones de distrito

o nacionales, y para participar en las elecciones generales como confederación deberán

haber solicitado su reconocimiento ante el juez federal con competencia electoral

competente hasta 60 días antes del plazo previsto para las elecciones primarias

respectivas (cf. art. 10 bis). El Fallo CNE 3858/07 estableció que "para constituir una

confederación de orden nacional, se deben reunir cinco partidos reconocidos en distintos

distritos *…+. … *Ello+ fundado en la representatividad de los partidos políticos como

condición de su existencia legal".

Los partidos reconocidos pueden fusionarse con uno o varios partidos políticos y

además –del acuerdo de fusión, el acta donde surja la voluntad de la fusión; requisitos

establecidos en los inc. b) a g) del art. 7, constancia de la publicación del acuerdo de

fusión– el juzgado verificará que la suma de los afiliados de los partidos que se fusionan

alcance el mínimo establecido de 4% de los electores inscriptos en el padrón electoral del

distrito respectivo. A su vez, se podrá formular oposición dentro de los 20 días de la

publicación. El partido político resultante de la fusión gozará de personería jurídico-


111555999


política desde su reconocimiento por el juez federal electoral competente, y se constituirá

a todo efecto legal como sucesor de los partidos fusionados, tanto en sus derechos como

en obligaciones patrimoniales, sin perjuicio de subsistir la responsabilidad personal que

les corresponda a las autoridades y otros responsables de los partidos fusionados por

actos o hechos anteriores a la fusión. Se considerarán afiliados al nuevo partido político,

todos los electores que a la fecha de la resolución judicial que reconoce la fusión, lo

hubiesen sido de cualquiera de los partidos políticos fusionados, salvo que hubieren

manifestado oposición (art. 10 ter).

En cuanto a las afiliaciones, se elimina la renuncia automática, ya que es condición

para la afiliación a un partido la renuncia previa expresa a toda otra afiliación anterior

(art. 25 ter), por ello se incorpora la posibilidad de manera gratuita ya sea por telegrama o

personalmente ante la secretaría electoral del distrito que corresponda (cf. art. 25

quáter). Por su parte, el decreto reglamentario 93/2010 establece la posibilidad de

renuncia sin determinar el partido político (art. 12). En efecto, ante la presunción de

afiliación o no querer manifestar expresamente a que partido se encuentra afiliado, se

admite la renuncia sin identificar el partido político.

Se incorporan también dos incisos al artículo 33 sobre inelegibilidades para ser

candidato y la consecuente prohibición de los partidos políticos de registrar tales

candidaturas, como causal de caducidad: "f) La personas con auto de procesamiento por

genocidio, crímenes de lesa humanidad o crímenes de guerra, hechos de represión ilegal

constitutivos de graves violaciones de derechos humanos, torturas, desaparición forzada

de personas, apropiación de niños y otras violaciones graves de derechos humanos o

cuyas conductas criminales se encuentren prescriptas en el Estatuto de Roma como

crímenes de competencia de la Corte Penal Internacional, por hechos acaecidos entre el

24 de marzo de 1976 y el 10 de diciembre de 1983".

Y “g) Las personas condenadas por los crímenes descriptos en el inciso anterior aún

cuando la resolución judicial no fuere susceptible de ejecución” (cf. art. 33).

En este sentido, la incorporación del inciso f) recepta el precedente de la Cámara

Nacional Electoral en la causa “Muñiz Barreto, Juana María y otros s/ impugna

TTTEEEPPPJJJFFF

111666000

candidatura a diputado nacional” sobre la candidatura de Patti para los pasados comicios

legislativos (Fallo CNE 4195/09).

La ley recientemente sancionada requiere para conservar la personería jurídico-

política que los partidos de distrito mantengan el número mínimo de afiliados requerido

para el reconocimiento y para los partidos de orden nacional, el número mínimo de

distritos necesarios para el reconocimiento. Será el Ministerio Público Fiscal, de oficio o a

instancia del juez, quien verificará su cumplimiento en el segundo mes de cada año, e

impulsará la caducidad. Previo a la declaración de caducidad el juez deberá intimar por 90

días (plazo improrrogable) al partido para que de cumplimiento (arts. 7ter, 8 y 50 incs. e

y f).

En estos dos últimos supuestos, el decreto de promulgación 2004/2008 observó las

disposiciones transitorias contenidas en los artículos 107 y 108 de la ley 26.571 que

establecían un plazo de gracia para los partidos con personería vigente, hasta el 31 de

diciembre de 2011, para reunir el mínimo de afiliados (en el caso de partidos de distrito) y

de distritos (en el caso de partidos nacionales) y así mantener la personería jurídico-

política. Asimismo, extinguía las acciones de caducidad que se encuentren en curso de los

incs. a), b), c), e) y f). El veto parcial motivó que se presentaran en la justicia nacional

electoral sendos amparos con el objeto de que se declare su inconstitucionalidad.

Proceso Electoral

La ley 26.571 respecto a la oficialización de candidaturas en el caso de las elecciones

primarias le otorga esta potestad a las agrupaciones políticas, por intermedio de las

juntas electorales partidarias, con apelación a la justicia nacional electoral. En cuanto a

las elecciones generales establece que las agrupaciones políticas registrarán a los

candidatos proclamados en las primarias ante el juez federal con competencia electoral

correspondiente.

La campaña electoral, para las primarias se inicia treinta (30) días antes de la fecha

del comicio y la publicidad electoral audiovisual puede realizarse desde los veinte (20)

días anteriores a la fecha de las elecciones primarias (cf. art. 31, ley 26.571). Mientras que

en las elecciones generales la campaña electoral se inicia 35 días antes de los comicios


111666111


generales (cf. art. 64 bis, CEN) y la publicidad audiovisual, a los 25 días previos (art. 64 ter,

CEN). A su vez, en relación con la publicidad de actos de gobierno, el nuevo régimen

incrementó de 7 a 15 días anteriores a los comicios la prohibición, tanto para las

primarias como para las generales (art. 64 quater, CEN).

El escrutinio definitivo de las primarias lo realizarán los juzgados federales con

competencia electoral de cada distrito. Éstos comunicarán los resultados a la Cámara

Nacional Electoral, en el caso de la categoría presidente y vicepresidente de la Nación, a

fin de que ésta proceda a hacer la sumatoria de los votos obtenidos en todo el territorio

nacional por los precandidatos de cada una de las agrupaciones políticas; y a las juntas

electorales partidarias, en el caso de senadores y diputados nacionales para que

conformen la lista ganadora.

La elección de los candidatos a presidente y vicepresidente de la Nación de cada

agrupación se hará mediante fórmula en forma directa y a simple pluralidad de sufragios.

Las candidaturas a senadores se elegirán por lista completa a simple pluralidad de votos.

En la elección de diputados nacionales, y parlamentarios del Mercosur, cada

agrupación política para integrar la lista definitiva aplicará el sistema de distribución de

cargos que establezca cada carta orgánica partidaria o el reglamento de la alianza

partidaria. Las juntas electorales partidarias efectuarán la proclamación de los candidatos

electos (cf. art. 44, ley 26.571).

Luego de las elecciones primarias no se pueden constituir alianzas o

confederaciones para participar en los comicios generales.

En las elecciones generales participarán las agrupaciones políticas que para la

elección de senadores, diputados de la Nación y parlamentarios del Mercosur, hayan

obtenido como mínimo un total de votos, considerando los de todas sus listas internas,

igual o superior al uno y medio por ciento (1.5 %) de los votos válidamente emitidos en el

distrito de que se trate para la respectiva categoría. Para la categoría de presidente y

vicepresidente, el uno y medio por ciento (1.5 %) de los votos válidamente emitidos en

todo el territorio nacional (cf. art. 45, ley 26.571).

TTTEEEPPPJJJFFF

111666222

Pero quizás el dato más relevante es la prohibición de contratar publicidad en

emisoras de radiodifusión televisiva o sonora abierta o por suscripción, ya que la pauta

publicitaria será solventada por el Estado y distribuida por la DINE.

La DINE determinará el horario del mensaje de campaña (entre 7 a 1 hs.), la

distribución entre las agrupaciones políticas (50% de manera igualitaria para todos y 50%

proporcional a la cantidad de votos en la elección general anterior de diputados). Los

gastos de producción de los mensajes estarán a cargo de cada agrupación política.

En relación con el financiamiento de campaña, es de resaltar que el nuevo régimen

prohíbe los aportes de personas jurídicas (art. 44 bis, ley 26.215 y modif.) y modifica los

parámetros de distribución del aporte público.

Para las elecciones presidenciales: 50% será distribuido en forma igualitaria y el

otro 50% se distribuirá entre los 24 distritos, en proporción al total de electores de cada

uno y luego a cada agrupación política en proporción a la cantidad de votos que hubieran

obtenido en la elección general anterior para la misma categoría.

En el caso de la elección de senadores: el total de los aportes públicos se distribuirá

entre los ocho distritos que renuevan bancas en proporción al número de electores de

cada uno. Del monto resultante para cada distrito, 50% se distribuirá en forma igualitaria

y el otro, en forma proporcional a la cantidad de votos para la misma categoría. Para la

elección de diputados, los aportes se distribuirán entre los 24 distritos en proporción al

total de electores de cada uno, y luego se utilizará el mismo criterio que para senadores,

50% en forma igualitaria y 50% proporcional a la cantidad de votos para la misma

categoría en la última elección general.

Respecto al aporte para la impresión de boletas, en el caso de las primarias se

otorgará a las agrupaciones políticas el equivalente a una (1) boleta por elector (art. 32,

ley 26.571). En el caso de elecciones generales el aporte será el equivalente a una boleta

y media (1.5) por elector registrado en cada distrito (art. 35, ley 26.215).

Por otra parte, la ley crea en la Cámara Nacional Electoral el Registro de Empresas

de Encuestas y Sondeos de Opinión, y ocho días antes de las elecciones generales no

podrán darse a conocer encuestas o sondeos de opinión, o pronósticos electorales ni

referirse a sus datos.


111666333


Acciones procesales

La reforma electoral prevé, en el caso de elecciones primarias, la apelación ante la

Cámara Nacional Electoral de todas las decisiones de los jueces federales con

competencia electoral que se dicten al respecto (arts. 19 y 28, ley 26.571).

Asimismo, se incorpora un recurso directo ante la Cámara Nacional Electoral de las

resoluciones de la DINE sobre distribución o asignación a las agrupaciones políticas de

aportes públicos o espacios de publicidad electoral. El que debe interponerse dentro

de las cuarenta y ocho (48) horas debidamente fundado ante la Dirección Nacional

Electoral del Ministerio del Interior que lo remitirá al tribunal dentro de las setenta y dos

(72) horas, con el expediente en el que se haya dictado la decisión recurrida y una

contestación al memorial del apelante. La Cámara podrá ordenar la incorporación de

otros elementos de prueba y solicitar a la Dirección Nacional Electoral del Ministerio del

Interior aclaraciones o precisiones adicionales. Luego de ello, y previa intervención fiscal,

se resolverá.

Administración electoral

En materia de administración electoral se establece la informatización del Registro

Nacional de Electores, pero en cuanto al Registro Nacional de Afiliados a los Partidos

Políticos no ha habido modificación al respecto.

Las modificaciones introducidas por la ley 26.571 incorpora las nuevas tecnologías

en cuanto al acceso de información. Garantiza a las provincias y a la Ciudad Autónoma de

Buenos Aires el acceso libre y permanente a la información contenida en el Registro

Nacional de Electores, a los efectos electorales (cf. art. 17, CEN).

Por otra parte, establece que la Cámara Nacional Electoral publicará la nómina de

electores fallecidos, por el plazo que determine, en el sitio de internet de la Justicia

Nacional Electoral al menos una (1) vez al año (cf. art. 22, CEN).

Respecto de los padrones provisorios dispone su publicación en el sitio web como

así también de los definitivos (cf. arts. 26 y 30, CEN). En cuanto al registro de afiliados,

establece que éste es público y que la Cámara Nacional Electoral deberá arbitrar un

TTTEEEPPPJJJFFF

111666444

mecanismo, restringiendo el acceso de terceros, para que el elector pueda conocer la

situación respecto de su afiliación (art. 26, Ley Orgánica de Partidos Políticos).

A su vez, establece el concepto de geografía electoral para la delimitación de

circuitos y ubicación de las mesas electorales.

Por último, la reforma incorporó que

la autoridad de aplicación adoptará las medidas pertinentes a fin de garantizar la

accesibilidad, confidencialidad e intimidad para el ejercicio de los derechos políticos

de las personas con discapacidad. Para ello se adecuarán los procedimientos,

instalaciones y material electoral de modo que las personas con discapacidad

puedan ejercer sus derechos sin discriminación y en igualdad de condiciones con los

demás, tanto para ser electores como para ser candidatos (art. 105, ley 26.571).


111666555


Managing the European Plurinational States: Consolidating Consociationalism through the Electoral Systems

Elias Dinas* Nikos Skoutaris†

ABSTRACT

Lijphart has argued that ‘for divided societies, ensuring the election of a broadly representative

legislature should be the crucial consideration, and P[roportionate] R[epresentation] is

undoubtedly the optimal way of doing so.’1 Despite this assertion, European plurinational States

have used a number of different electoral systems in order to effectively accommodate ethno-

linguistic and religious cleavages in a consociational manner. The scope of the proposed paper is

twofold. On the one hand, it aims at mapping how consociationalism is echoed in the electoral

systems of certain European plurinational States, where power-sharing arrangements are in

place. To achieve this goal, the paper describes the consociational variant in the constitutional

structures of three European divided societies and how the consensus model of democracy has

been translated in their electoral systems. On the other hand, it consists of an effort to assess

whether and, if so, the extent to which the relevant electoral systems have been successful at

consolidating the consociational principle by focusing on the elections that have taken place in

those political systems. By looking closely at the interplay between public opinion and political

actors’ issue stances in those societies we are able to test whether the chosen electoral systems

in the aforementioned power-sharing arrangements have favoured parties with a more

consociational agenda over time or whether parties with a more ‘separatist’ one have managed

to dominate the political arena. Overall, the present paper questions the success of

consociationalism to bridge the ethno-linguistic and/or religious cleavages in the divided

societies of Europe.

* Prize Research Fellow, Nuffield College, Oxford University, [email protected].

† Assistant Professor, International and European Law Department, Maastricht University,

[email protected]. 1 A Lijphart, Thinking about democracy (London, Routledge, 2008) at 78.



TTTEEEPPPJJJFFF

111666666

1. Introduction

Since the Peace of Westphalia, we have been witnessing in Europe the gradual dissolution of the

plurinational Empires and the genesis of the sovereign nation-States. This historical and political

trend that has favoured the building of mono-national over plurinational States in the old

continent reached its peak in the aftermath of the fall of the ‘Iron Curtain’ and the subsequent

dissolution of Soviet Union, Yugoslavia and Czechoslovakia. Despite this, there is still a significant

number of multinational European States such as Belgium, Bosnia-Herzegovina (BiH), Spain,

Switzerland, the United Kingdom. In all those political and constitutional systems -however

successful and functional they have proved to be- the question how to address effectively

national diversity entailing ethno-linguistic and/or religious cleavages remains of cardinal

importance.

In order to respond convincingly to the needs created by the ethno-linguistic and/or

religious cleavages, most of the aforementioned political and constitutional systems have opted

for power-sharing arrangements. Despite the obvious differences in the historical and political

conditions that have led to the adoption of consociational elements in the constitutional

designing of certain States in Europe and the World, all the power-sharing arrangements

challenge the majoritarian interpretation of the basic definition of democracy which entails

‘government by the majority of the people’.2 The consociational model of democracy3 aims at

addressing the issue of exclusion of minority groups from participation in decision-making

especially in deeply divided societies such as Belgium, Bosnia-Herzegovina and Northern Ireland.

According to that model, every significant group should proportionately participate in the

government of the country while at the same time it retains a high degree of autonomy and the

possibility to veto decisions of the majority in order to protect its vital interests.

2 A Lijphart, Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries, (New Haven and

London, Yale University Press, 1999) 31. 3 See in general SM Halpern, ‘The Disorderly Universe of Consociational Democracy’ (1986) 9 West European Politics

181; A Lijphart, Democracy in Plural Societies (New Haven and London, Yale University Press, 1977); A Lijphart, Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries, (New Haven and London, Yale University Press, 1999); A Lijphart, ‘The Power-Sharing Approach’ in JV Montville, JV, Conflict and Peacemaking in Multiethnic Societies (Lexington MA, Lexington Books 1991) 491; KD McRae (ed), Consociational Democracy (Toronto, McClelland and Stewart Limited, 1974).


111666777


The present paper focuses exactly on party competition dynamics of the segmented

societies of Belgium, Bosnia-Herzegovina and Northern Ireland by mapping how

consociationalism is echoed in their electoral systems. Although the case studies we examine do

not provide for an exhaustive list of all the power-sharing arrangements in Europe, they do

represent three models of managing the plurinational States. First, Belgium amended its own

constitution in 1994 to adopt its current federal structure in order to address demands

expressed by both ethno-linguistic segments for more decentralisation. Second, Bosnia’s

constitution is part of the wider Dayton Peace Agreement4 that ended three and a half years of

fierce fighting. Third, in Northern Ireland -albeit not a sovereign State itself- a very sophisticated

power-sharing arrangement has been adopted within the frameworks of the Good Friday

Agreement on the one hand and the UK asymmetric devolution on the other. Thus, the

differences in the historical and political conditions that have led to the adoption of the

respective variant of the consociational model allow us to offer an important comparative insight

to the choices made with regard to the electoral systems of European divided societies. Overall,

in parts 2 and 3 of this paper we focus exactly on how the consociational principle has been

applied in the constitutional structure of those power-sharing arrangements and its influence on

the respective electoral system.

However, consociationalism more than a constitutional principle that influences the

electoral system of a given society, it should mainly be a political modus vivendi advanced -

among else- by electoral engineering. This is the reason, why in part 4 of the paper, we focus on

the elections that have taken place in those political systems during the last decades. In so doing,

we are able to test whether the chosen electoral engineering in the aforementioned power-

sharing arrangements have favoured parties with a more consociational agenda over time or

whether parties with a more ‘separatist’ one have managed to dominate the political arena.

Unfortunately, a systematic empirical examination of these processes questions the success of

consociationalism to bridge the ethno-linguistic and/or religious cleavages in the divided

societies of Europe.

4 General Framework Agreement for Peace in Bosnia and Herzegovina (“the Dayton Peace Agreement”), initialled at

Dayton on 21 November 1995 and signed in Paris on 14 December 1995.

TTTEEEPPPJJJFFF

111666888

2. The Consensus Model of Democracy

Consociation derives from the Latin consociatio, which means ‘the action or fact of associating

together’ or ‘union in fellowship’.5 The term appears as early as 1603 in Althusius’s Politica

Methodice Digesta, partly as an attempt to analyse the process of new polity creation in the

early 17th century Low Countries, ‘without either a strong governmental apparatus or an

articulate national identity’,6 and partly as a response to Bodin’s Les six livres de la republique of

1576 and his novel conception of sovereignty.7 The first modern exponent of consociationalism

was Apter who, in his study of bureaucratic nationalism, defined this form of political

organisation as ‘a joining together of constituent units which do not lose their identity when

merging in some form of union’.8

But it was Arend Lijphart, reflecting on the paradoxical nature of the Dutch polity, in

combining political stability, religious differences and social fragmentation,9 who was the first to

stress the stabilising effects of the consensus model of democracy in plural societies and offered

a general model of it. Lijphart’s model of consociational democracy consists of the following four

defining properties: participation of the representatives of all significant segments of the plural

society; high degree of autonomy for each group to run its own internal affairs; proportionality

as the principal standard of political representation, civil service appointments and allocation of

public funds; and the mutual veto or ‘concurrent majority’ rule, which serves as an additional

protection of vital minority interests. What follows is an analysis of those four main

characteristics and how they are embodied in the constitutional architecture of Belgium,10 BiH11

and Northern Ireland.12

5 B Barry, ‘Political Accomodation and Consociational Democracy’ (1975) 5 British Journal of Political Science, 478.

6 H Daalder, ‘On Building Consociational Nations: the cases of The Netherlands and Switzerland’ (1971) 23

International Social Science Journal, 358.7 JH Franklin, On Sovereignty: Four Chapters from the Six Books of the Commonwealth. (Cambridge, Cambridge

University Press, 1992).8 D Apter, The Political Kingdom in Uganda: A Study in Bureaucratic Nationalism (Princeton, NJ, Princeton University

Press, 1996) 24.9 A Lijphart, The Politics of Accommodation: Pluralism and Democracy in The Netherlands (Berkeley, CA, University of

California Press, 1968).10

See in general the Belgian Constitution; http://www.fedparl.beconstitution_uk.html.11

See in general the Constitution of Bosnia-Herzegovina; http://www.servat.unibe.ch/icl/bk00000_.html.12

See in general the Northern Ireland Act 1998 [1998] c. 47; http://www.legislation.gov.uk/ukpga/1998/47/contents.

http://www.fed-parl.be/constitution_uk.html

http://www.servat.unibe.ch/icl/bk00000_.html

http://www.legislation.gov.uk/ukpga/1998/47/contents


111666999


2.1 Power-Sharing

The primary characteristic of the consensus model of democracy is the joint exercise of

governmental, executive power. This may take various institutional forms. The most straight-

forward is that of a grand coalition cabinet in a parliamentary system. For instance, the Belgian

constitution contains a formal requirement that the executive includes representatives of the

large linguistic groups. Article 99 of the 1994 federal constitution stipulates that with ‘the

possible exception of the Prime Minister, the Council of Ministers [cabinet] includes as many

French-speaking as Dutch-speaking members’.13

Moreover, in Northern Ireland, the posts of the First Minister and Deputy are tied

together by section 16 of the Northern Ireland Act 1998. According to the same section ‘the

largest political party of the largest political designation shall nominate *...+ the First Minister’

while ‘the largest political party of the second largest political designation shall nominate [...] the

deputy First Minister’. In other words, this provision ensures that the First Minister will be from a

unionist party while the Deputy from a nationalist or republican one. More interestingly, the

ministers are not chosen by this dyarchy. Instead, section 18 of the Northern Ireland Act provides

that the ministerial posts are allocated to all of those parties with significant representation in

the Assembly. The number of posts to which each party is entitled, is determined according to

the d’Hondt method of proportional representation.14 The actual posts are chosen by the parties

in the order that the seats were awarded. This does not mean that apart from the two largest

parties, the other parties are required to enter the Executive. They can choose to go into

opposition if they wish. However, until now all Northern Irish cabinets have been comprised by

at least four parties.

13

A Alen, and R Ergec, Federal Belgium After the Fourth State Reform of 1993 (Brussels, Ministry of Foreign Affairs, 1994).14

This method favours larger parties and operates through successive rounds of voting. In the first round the number of votes for each party (in this case, the number of assembly seats) is counted and the highest receives the first post. In the second round the party which won the first executive post has its representation halved. The party with the highest number of seats by this formula now gains an executive post. The rounds continue, with each party having its divisor increased by one of each of the executive posts it receives.

TTTEEEPPPJJJFFF

111777000

On the other hand, it has been observed that it is less common to accommodate a

power-sharing arrangement in presidential systems.15 Practice suggests that one possible way is

to distribute the presidency and other high offices among the different groups. The Constitution

of Bosnia-Herzegovina provides for an example. According to Article V, the Chair of the BiH

Presidency rotates among three members: one Bosniak and one Croat, each directly elected

from the territory of the Federation, and one Serb directly elected from the territory of the

Republika Srpska. Each of them is elected as the Chair for an eight-month term within their four-

year term as a member. The Chair of the Council of Ministers is nominated by the Presidency

and approved by the House of Representatives. S/he is then responsible for appointing a Foreign

Minister, Minister of Foreign Trade and others as appropriate. ‘Together the Chair and the

Ministers [...] constitute the Council of Ministers, with responsibility for carrying out the policies

and decisions of Bosnia and Herzegovina.’16

2.2 Group Autonomy

A second deviation of the consensus model of democracy from majority rule is segmental

autonomy, which entails minority rule: rule by the minority over itself in the area of the

minority’s exclusive concern. It complements the principle of joint rule. On all issues of common

concern, decisions should be made jointly by all of the segments together with roughly

proportional degrees of influence; on all other issues, decisions should be left to be made by and

for each separate group. If the groups have a clear territorial concentration, group autonomy

may be institutionalised in the form of federalism. If the groups are intermixed, autonomy will

have to take a non-territorial form or a combination of territorial and non-territorial forms.

In Belgium, territorial federalism as a form of segmental autonomy has been particularly

important since 1970. However, the form of federalism adopted in 1994 is unique since it

consists of three geographically defined regions17 –Flanders, Wallonia, and the bilingual capital

of Brussels– and three non-geographically defined cultural communities18 –the large Flemish-

15

See in general A Lijphart, Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries, (New Haven and London, Yale University Press, 1999).16

Article V, Paragraph 4 of the BiH Constitution.17

Article 3 of the Belgian Constitution.18

Ibid at Article 2.

http://en.wikipedia.org/wiki/Council_of_Ministers_of_Bosnia_and_Herzegovina


111777111


speaking and French-speaking communities and the much smaller German-speaking community.

The segmental autonomy of the constituent units of the Belgian federation is also guaranteed by

the list of competences that are assigned to them by virtue of Articles 127-130 of the

Constitution and a series of constitutional laws passed under Article 134 using the procedure laid

down in Article 4. In general, the Regions are responsible primarily in the fields of economic and

social affairs while the Communities exercise authority in the areas of culture, social policy,

health and education.

Being a ‘classic example of consociational settlement’19 post-Dayton Bosnia-Herzegovina

is a (con)federal State comprised by two ethnically defined entities: a Serb one and a Bosniak-

Croat one. The former is a unitary mono-national Republic as its name Republika Srpska (RS)

itself defines and the latter a bi-national federation of ten autonomous cantons with the rather

neutral name of the Federation of Bosnia-Herzegovina (FBiH). The two entities enjoy all the

competences that are ‘not expressly assigned in this Constitution to the institutions of Bosnia

and Herzegovina’.20 Given that the list of the enumerated competences of the federal State is

very limited,21 we can safely conclude that the two entities enjoy wide-ranging powers of self-

government and thus autonomy.

The situation in Northern Ireland concerning segmental autonomy is somewhat different

from the ones in Belgium and in BiH. The Northern Ireland Act 1998 does not transfer

competences to the two ‘political designations’ separately. Instead, within the framework of the

UK asymmetric devolution, the Westminster Parliament transferred competences to the

Northern Ireland Assembly and its executive. The Act recognises three distinct varieties of policy:

the ‘excepted powers’, the ‘reserved powers’ and the ‘transferred’. The ‘excepted’ powers

cannot be transferred to the Assembly without amendment of the Act itself22 while the

‘reserved’ powers can be transferred if cross-party support within the Assembly is evident.23

Powers that are not within those lists are described as ‘transferred’. So, the two ethno-religious

19

Sumantra Bose, Bosnia After Dayton: Nationalist Partition and International Intervention, (Oxford, Oxford University Press, 2002) 216.20

Article III, Paragraph 3(a) of the BiH Constitution.21

Ibid, Paragraph 1. 22

Schedule 2 of the Northern Ireland Act 1998.23

Ibid, Schedule 3.

TTTEEEPPPJJJFFF

111777222

segments may not enjoy autonomy from each other but the UK devolution has ensured

Northern Ireland’s legislative autonomy as a whole.

2.3 Proportionality

Proportionality, the third characteristic of the consensus model of democracy, serves as the basic

standard of political representation, public service appointments and allocation of public funds.

Its great advantage is that it is widely recognised as the most obvious standard of fair

distribution. In addition, it facilitates the process of decision-making because it is a ready-made

method that makes it unnecessary to spend a great deal of time on the consideration of

alternative methods of distribution. With regard to political representation, proportionality is

especially important as a guarantee for the fair representation of ethnic minorities. This is the

main reason that Lijphart has argued that ‘for divided societies, ensuring the election of a

broadly representative legislature should be the crucial consideration, and P[roportionate]

R*epresentation+ is undoubtedly the optimal way of doing so.’24 Indeed, as we shall see in the

third section of this paper in all three case studies variants of PR have been adopted as their

electoral system.

But, the proportional composition of decision-making bodies does not solve the problem

of how to achieve proportional influence when the nature of the decision is basically

dichotomous. In such case, unless there is spontaneous unanimity, there will be winners and

losers. Ultimately, the use of either majority rule or minority veto cannot be avoided. Hence, the

constitutional designing of some States embody a variant of the principle of proportionality that

entails even greater deviation from majority rule: the deliberate overrepresentation of small

segments and in some cases parity of representation. Examples of paritarian bodies in which

minorities are overrepresented is the Belgian cabinet which consists of equal numbers of

Flemish-speaking and French-speaking ministers, the Northern Irish diarchy of the First Minister

and the Deputy and the rotating Chair of the BiH Presidency.

24

A Lijphart, Thinking about democracy (London, Routledge, 2008) at 78.


111777333


2.4 Minority veto

The fourth characteristic of the consociational democratic theory is the minority veto, which

consists of the ultimate weapon that minorities need to protect their vital interests. Even when a

minority participates in a power-sharing executive, it may well be out-voted or overruled by the

majority. This may not present a problem when only minor matters are being decided, but when

a minority’s vital interest are at stake, the veto provides essential protection. The veto power

clearly contains the danger that the entire system can be undermined if one or more minorities

overuse or abuse their veto power. It works best when it is not used too often and only with

regard to issues of fundamental importance. A clear example of minority veto is the right vested

to each member of the BiH Presidency to veto decisions that might violate the ‘vital interests’ of

their Entity according to their view.

3. Electoral Systems and Divided Societies

Building an institutional framework to accommodate the needs of a segmented society is a

daunting task. The challenge becomes even bigger in those occasions where a society tries to

heal the wounds of an inter-community conflict as it has been the case in Northern Ireland and

the BiH. In such deeply divided societies

the issues of proportionality (which has implications for appropriate representation of minorities) and of party discipline (important in facilitating the capacity of elites to broker inter-group deals) acquire a particular priority, rendering proportional representation – and especially the closed list system – singularly attractive.25

This is the reason why the electoral design of all the consociational arrangements we

examine embrace proportional representation systems to effect power sharing and minority

protection as we have already mentioned.26

What follows in the next three sections is a brief description of the electoral systems

chosen in the segmented societies under review. Such analysis will not just show how the

25

J Coackley, ‘The Political Consequences of the Electoral System in Northern Ireland’, (2009) 24 Irish Political Studies, 253, 254.26

See above section 2.3 of the paper.

TTTEEEPPPJJJFFF

111777444

consociational principle has influenced the electoral engineering of the political systems but it

will also shed light to an interesting academic debate concerning electoral systems for divided

societies. Interestingly enough, all three of them have chosen different variants of the PR. First,

Belgium has opted for an open list PR system. According to Lijphart, this is the most effective

mechanism for facilitating inter-group deals in consociational democracies not just because this

system secures the equitable representation of groups – particularly minority ones – within

parliament, but in particular because it may facilitate elite control of parties, thus freeing the

party leadership to engage in power-sharing deals.27 Second, the election of the tripartite BiH

Presidency is based on an alternative vote system which according to some analysts offers

significant incentives to compromise.28 Third, the elections for the Northern Ireland Assembly

are held according to the Single Transferable Vote system (STV). The influential Brendan O’Leary

and John McGarry have argued that STV is not just proportional in its effect, but that it promotes

moderation by encouraging parties to seek lower preferences by tempering their own policy

positions.29

3.1 Belgium

In 1899,30 Belgium became the first country in the world to introduce PR for national legislative

elections to its lower chamber. The 1994 federal Constitution of the Belgian State has not

questioned the use of the PR as the electoral system. According to Article 62 of the Constitution

the elections of the 150 members of the Belgian Chamber of Representatives ‘are carried out by

the system of proportional representation that the law determines.’

However, the linguistic borders -which largely correspond to the regional ones- have

greatly affected the electoral system. Currently, there are five mono-lingual Flemish districts, five

mono-lingual Wallon districts and the bilingual district of Brussels-Halle-Vilvoorde, which is the

27

A Lijphart, Democracy in Plural Societies: A Comparative Exploration (New Haven, CT, Yale University Press, 1977).28

Palley pointed to the role of this system in giving an advantage to the more moderate of two competing parties within the same ethnic group; C Palley, Constitutional Law and Minorities (London, Minority Rights Group, Report No. 36, (1978)), 16-17. Horowitz made a similar point about the potential of the alternative vote to promote compromise, especially in the context of single-vacancy contests DL Horowitz, Ethnic Groups in Conflict (Berkeley, University of California Press (1985))639–643.29

J McGarry and B O’Leary, (2009) ‘Power shared after the deaths of thousands’ in R Taylor (ed) Consociational Theory: McGarry and O’Leary and the Northern Ireland Conflict (London, Routledge, 2009) 15.30

This section of the paper builds on the work of Dr. Lars Hoffmann.


111777555


only bilingual electoral district in Belgium. The seat allocation in the ten mono-lingual regions is

allocated according to the d’Hondt system based on the basic parameters set in Article 63 of the

Belgian Constitution. Given that the parties do not run nationally, the 5 per cent threshold for a

party to be represented in the Parliament is accordingly not applied nationally but among the

linguistic communities. This has made possible for a larger number of parties to be elected to the

federal Parliament and thus the two communities to be represented more proportionately. At

the same time, the fact that the Belgian parties do not run nationally means that it is almost

impossible for members of one community residing across the linguistic border to vote for

representatives belonging to their ethno-linguistic group.31 In other words, although the

electoral system in Belgium has allowed the two main ethno-linguistic communities to be

adequately represented, its integrationist effect to the political system is rather doubtful given

the limitations that the linguistic borders set. This will become even more apparent in the next

section where we analyse the electoral results.

3.2 Bosnia-Herzegovina32

In the first post-Dayton elections and up to the 2000 ones, the system of proportional

representation with closed party lists was adopted for the legislative bodies in the Bosniak-Croat

FBiH, the Republika Srpska and at state/federal level. Each party fixed the order of candidates

elected and voters were unable to express a preference for a particular candidate. The system

aimed at ensuring the inclusion and representation of all groups. However, it facilitated the

victory of the main nationalist parties.

Since 2000, open lists and multi-member constituencies have been introduced. By

allowing the voters to indicate both their favoured party and favoured candidate this moderate

change to the electoral system aimed at forging a link of accountability between elected

31

On the right to stand as a candidate in the Belgian elections see European Court of Human Rights, Case of Mathieu Mohin and Clerfayt v Belgium, (Application No. 9/1985/95/143) (judgment of 2 February 1987). On the same issue, the BiH Constitutional Court has explained that ‘*t+he Belgian system does not preclude per se the right to stand as a candidate solely on the ground of language. Every citizen can stand as a candidate, but has – upon his election – to decide whether he will take oath in French or Flemish . . . whereas provision of the Constitution of the Federation of BiH provide for a priori ethnically defined Bosniak and Croat delegates, caucus and veto powers for them’. Partial Decision of the Constitutional Court, 1 July 2000, para 120.32

For a comprehensive account of electoral engineering in BiH, see R Belloni, 11(2004) 'Peacebuilding and consociational electoral engineering in Bosnia and Herzegovina', (2004) International Peacekeeping, 334.

TTTEEEPPPJJJFFF

111777666

representatives and the electorate as well as to the party. Open lists, though, might increase

accountability but do not necessarily favour moderation and multi-ethnic parties as we shall later

observe.

More importantly, the Permanent Election Law33 introduced a preferential/alternative

voting system for the election of the tripartite Chair of the BiH Presidency. It was hoped that this

change, in Horowitz fashion, would moderate Bosnian politics. As peacebuilders explained:

‘When more than just the first preference votes are taken into consideration, the moderate

candidates stand a better chance of winning the elections, as they will have support from a large

cross-section of the electorate. Extreme or radical candidates have less chance to win the

elections.’34

In the next section we will see whether this contention survived the reality test of the

elections.

Irrespective of whether this amendment has been proved successful to moderate the

politics of a post-conflict society, it is critical to point out that it was not accompanied with

changing the electoral basis. Instead, the law maintained that in order to be eligible to stand for

election to the Presidency or the House of Peoples of Bosnia and Herzegovina, one has to

declare affiliation with a ‘constituent people’. In a groundbreaking decision the European Court

of Human Rights decided that this provision violates the non-discrimination principle contained

in Article 14 and the right to vote as provided by Article 3 of Protocol No 1 of the Convention of

Human Rights.35

3.3 Northern Ireland36

According to Article 33 of the Northern Ireland Act 1998 the Assembly's 108 members are

elected from 18 six-member constituencies on the basis of universal adult suffrage. The

33

The Election Act 2001 (published in Official Gazette of Bosnia and Herzegovina no. 23/01 of 19 September 2001, amendments published in Official Gazette nos. 7/02 of 10 April 2002, 9/02 of 3 May 2002, 20/02 of 3 August 2002, 25/02 of 10 September 2002, 4/04 of 3 March 2004, 20/04 of 17 May 2004, 25/05 of 26 April 2005, 52/05 of 2 August 2005, 65/05 of 20 September 2005, 77/05 of 7 November 2005, 11/06 of 20 February 2006, 24/06 of 3 April 2006, 32/07 of 30 April 2007, 33/08 of 22 April 2008 and 37/08 of 7 May 2008) entered into force on 27 September 2001.34

Association of Election Officials of BiH, ‘Technical Series No. 1/2001’, 8.35

See European Court of Human Rights, Case of Sejdić and Finci v Bosnia and Herzegovina (Applications Nos 27996/06 and 34836/06) (Judgment of 22 December 2009).36

For a comprehensive account of the electoral engineering in Northern Ireland, see J Coackley, ‘The Political Consequences of the Electoral System in Northern Ireland’, above n 26.


111777777


constituencies used are the same as those used for elections to the Westminster Parliament. The

subsequent Article of the Act provides that the electoral system to be used is the single

transferable vote (SVT).37 The choice has been hardly surprising and largely uncontested given its

long history in Northern Irish politics.38 As already mentioned specialists on the Northern Ireland

conflict like Brendan O’Leary and John McGarry have long defended its suitability for the politics

of this segmented society. Its success will be analysed in the next section.

4. The political implications of constitutional designs

The purpose of this section is to shed some light on the way in which the three constitutional

structures analysed above and their respective electoral systems shape party competition. We

will focus primarily on the interplay between parties’ policy positions and how these issue

stances are evaluated by public opinion. In the last part of this section, we will also delve into the

long-standing imprint of consociationalism on voters’ decision making mechanisms. Due to lack

of individual-level data for BiH, this part is mostly devoted to the cases of Northern Ireland and

Belgium. That said, aggregate-level evidence will be also presented for Bosnia, indicating a

largely common pattern to that found for our two other cases.

37

Article 34 (3) of the Northern Ireland Act 1998 provides that: A single transferable vote is a vote— (a)capable of being given so as to indicate the voter’s order of preference for the candidates for election as members for the constituency; and (b)capable of being transferred to the next choice when the vote is not needed to give a prior choice the necessary quota of votes or when a prior choice is eliminated from the list of candidates because of a deficiency in the number of votes given for him.38

J Coackley, ‘The Political Consequences of the Electoral System in Northern Ireland’, above n 25, 256-258.

TTTEEEPPPJJJFFF

111777888

4.1 Consociationalism and satisfaction with the political system

Figure 1: Average level of satisfaction with Democracy in Belgium,

in Wallonia and in Flanders, 1973-1999.

Note: Source, Eurobarometer Trend File.

As a way to motivate the discussion, Figure 1 depicts the moving average of people’s

satisfaction with the way democracy functions in Belgium over a 25-year period that goes up to

1999. This individual-level information comes at an almost annual basis from the Eurobarometer

surveys that have been held in Belgium since 1973. Given that the Belgian political system was

gradually transformed into two ethno-linguistically-defined subsystems, corresponding to the

Flemish and the French-speaking communities that largely comprise the country, this

progressive but evidently important development should be captured by citizens’ evaluations of

their political system. As is shown in the Figure, however, no clear pattern, that could be


111777999


convincingly attributed to the process of political decentralisation characterising Belgian politics,

is found. This seems to be the case both in Wallonia and in Flanders. The only interesting change

seems to be the gradual convergence of the two communities in their levels of satisfaction with

their political system. Importantly, this increased similarity appears to be the outcome of a

gradual decline in the level of satisfaction with democracy among the Dutch-speaking

community. Moreover, although one could refer to the sharp drop in the average level of

confidence to the Belgian political system almost right after the 1994 constitutional reform, the

shock lasts only a couple of years, since by 1998 both communities regress to their prior mean

levels of support.39 Interestingly, during this period and despite initial level-differences the two

groups seem to move in a parallel fashion, denoting similar reactions both in terms of direction

and, most of the times, in terms of magnitude to the contextual stimuli generated by their

immediate political environment. On the whole, Figure 1 reveals that the gradual process of

decentralisation has not improved people’s evaluations about the way their political system

operates. By the same token, it seems that even if the constitutional change might have come as

a result of a public vital demand, it does not seem to have helped the members of both

communities to reorientate their beliefs about the way Belgian Parliamentary Democracy works.

39

To further explore the potential impact of the 1994 reform in the trend observed after this year in Figure 1, we regressed the mean level of satisfaction with a dummy (i.e. a binary indicator that takes only values 0 and 1 to denote the realisation of the measured attribute of interest) denoting all years after 1994. Its effect was effectively zero (b-coefficient: .-.075, bootstrapped 95% Confidence Intervals (C.I.): [-.012 .024]).

TTTEEEPPPJJJFFF

111888000

Figure 2: Average level of satisfaction with the way democracy

works in Northern Ireland, 1989, 1994 and 2004.

Note: Dots present average estimates, vertical lines correspond to the 95% confidence intervals, associated with each

sample estimate, source, European Election Studies.

Figure 2 provides equivalent, albeit less detailed, evidence for the case of Northern

Ireland. Here, we use information from the European Election Studies (EES), which take place

right after the elections for the European Parliament since 1989. We use the same question,

which however was not available in the 1999 EES. Importantly, until 2009 the EES treated

Northern Ireland as a different political context, rather than as a region of Great Britain. This

means that the resulting datasets provide a separate large-N sample of Northern Ireland, which

permits us to infer changes over time with lower levels of uncertainty. As is shown in Figure 2,

the Northern Ireland Act 1998 does not seem to have improved people’s level of satisfaction

with the political system. Rather, things seem to have deteriorated. The vertical lines passing

through the small dots of the graph denote the 95% confidence intervals, associated with each


111888111


estimate. 40 The results denote that although the small difference between 1989 and 1994 might

have simply come from a true null (from a zero population difference), the drastic decline of the

average level of satisfaction by 2004 seems to denote a statistically significant pattern: there is

no overlap in the confidence bands between 2004 and either 1989 or 1994. Clearly, the

asymmetric devolution put forward during Blair’s first term in office does not seem to have

advanced people’s confidence in the political system of Northern Ireland.41

The evidence from BiH is quite similar.42 90 per cent of those living in Bosniak majority

areas believe that the current political situation will be deteriorated in the near future. The

percentages in the Serb and the Croat majority areas are slightly lower but still relatively gloomy

(65 and 70 per cent respectively). Minorities living in each of these three communities are

equally pessimistic. Moreover, the Composite Political Stability Index, which constitutes an

encompassing scale including various measures about the way the political system and party

competition works in BiH, seems to have gradually declined over the last years.43 By 2009, 83 per

cent of the Bosniak sample demonstrated high levels of dissatisfaction with the governing parties

and 90 per cent took the view that they do not deserve to remain in power.44 Although clearly

less critical about their own parties, both Croat and Serb public opinion were still largely negative

in their evaluations of their parties’ government record. More importantly, in a question about

vote choice in the coming election, half of the sample did not choose any of the existing parties.

Clearly, there are still important steps that need to be taken in order to improve the existing

levels of political representation among all ethnic communities. The most pertinent problem,

40

Since each estimate comes from a different sample, parametric inference is problematic. Accordingly, we use bootstrapped measures of uncertainty, by resampling with replacement (100 simulations) each of the three original samples. See B Efron, ‘Bootstrap confidence intervals for a class of parametric problems.’ (1985) 72 Biometrika, 45.41

A rather obvious critique to this finding could be that there might be important community-based heterogeneity in this pattern. The Northern Ireland Act has been evaluated in a different way among Catholics and among Protestants. Yet, in this case it is important to mention that the results hold for each of the two groups to an approximately equal extent. When each year’s sample is divided into the two religious groups the findings are very similar, although the confidence bands are now substantially higher (as could be expected since half of the respondents is used in each analysis). That said the gap between 2004 and 1994 remains statistically distinguishable from zero. 42

Due to the lack of individual-level information we resort to a summary of the findings from the Early Warning System, a UN global development project that monitors public opinion in key areas of politics, social security and ethnic relations in the country during the last ten years. All results reported here can be found in their annual reports, accesible in the world wide web: http://www.undp.ba/index.aspx? PID=3&RID=54.43

2009 Early Warning System Report (EWS), 15.44

Ibid, 21.

http://www.undp.ba/index.aspx?PID=3&RID=54

TTTEEEPPPJJJFFF

111888222

however, seems to be that despite EU’s efforts to strengthen BiH central government, these

attempts have been strikingly unsuccessful. As Johan Galtung has put it:

The BiH elections show a heavily divided non-state: Srpska wants out, the US-imposed federation of Croats and Bosniaks is divided with the voters voting largely for their own kind. And that is what it is all about: they want to be governed by their own kind, and outside forces deny them even the right of self-determination in a referendum.45

Undoubtedly, the prospects are not very positive and this complex and unique design

that was employed under the consensus model of democracy does not seem to be a sustainable

solution for the three ethnic groups residing in BiH.

4.2 Party polarisation and electoral success

With this essentially illustrative evidence serving as a departure point, we now move to a more

systematic exploration of the political implications of consociationalism-driven constitutional

designs. Given that we lack individual-level data for BiH, the empirical analysis from now on will

focus only on Northern Ireland and Belgium. The main question addressed is whether those

different solutions adopted in Belgium and in Northern Ireland have had any positive effect in

gradually reducing the weight of the long-standing ethno-linguistic and/or religious cleavages

that have shaped political competition in each of these areas.46

45

Transcend Media Service, Editorial, 4 October 2010 http://www.transcend.org/tms/2010/10/a- yugoslav-community-for-a-yugosphere/.46

Unavoidably, it would be rather unrealistic to ask for such signs in the case of BiH, where the war ended only fifteen years ago.

http://www.transcend.org/tms/2010/10/a-yugoslav-community-for-a-yugosphere/

http://www.transcend.org/tms/2010/10/a-yugoslav-community-for-a-yugosphere/


111888333


Figure 3: The relationship between parties’ postures in the issue of administrative

decentralisation and their electoral success in Belgium, both Wallonia and Flanders.

Note: The solid curve presents a locally weighted regression curve (bandwidth: .75) fitted into the scatterplot of

change in vote share from 2003 to 2007 in Belgium.

To shed some light on this question, we first need data about parties’ policy stances in

these matters. For this reason, we use information from expert surveys conducted by Benoit and

Laver within the Party Policy in Modern Democracies project.47 In this survey, that took place in

2004, experts have been called to place the Belgian parties in a dimension ranging from 1 to 20,

where 1 stands for ‘high decentralisation of public administration and decision making’ and 20

stands for ‘low decentralisation of decision making’. Thus, we have information about parties’

actual stands on that issue. What Figure 3 shows is the relationship between their stances on

that issue and their electoral success. In other words, the first (leftmost) panel of Figure 3

addresses the question of whether it actually still pays to advance the issue of further

decentralisation (implying a step forward towards a confederal state) in a country in which

47

K Benoit and M Laver, Party Policy in Modern Democracies (London, Routledge, 2006). During the last decades a rather voluminous literature on the measurement of parties’ policy and ideological positions has developed. Within this framework, expert surveys, ie surveys with well-structured self-administered questionnaires completed by a random sample of experts (political scientists whose research focuses on a given country) about a variety of questions regarding parties’ issue stances, have been regarded as the safest way of mapping parties’ placements in the multi-dimensional policy space. See indicatively M Laver,(ed) Estimating the Policy Positions of Political Actors, (New York, Routledge, 2001).

TTTEEEPPPJJJFFF

111888444

consociationalism has already created a rather segmented political setting. The answer is that, if

not anything else, the incentives to play the ethno-linguistic card have not ceased to exist. What

we see is the change in the vote share of each party from 2003 to 2007. These figures are plotted

against the level of decentralisation promoted by each of the salient parties taking place in each

election.

Although Flanders and Wallonia operate as two distinct party systems, we have included

all main parties of the two regions together because differences in the observed patterns

between them are only infinitesimal. To be sure, even when combining the two communities,

we are still in front of a small-N problem, since we only have nine observations available. This

means that rough-and-ready parametric regression analysis, based on distributional assumptions

that are difficult to be satisfied with such a small number of observations, ceases to give

unbiased results. Therefore, we choose a more indirect but probably more informative

alternative. A local regression curve (loess) is fitted into the scatterplot, providing a visualisation

of the relationship between a party’s position on decentralisation and its electoral success. As

with all non-parametric regression methods, the basic idea behind the loess curve is to trace the

salient features of the mean response making only minimal assumptions about its distribution.48

Thus, a loess curve showing a downward monotonic pattern can be considered as a good

indication that choosing sides in this issue still entails significant electoral benefits. This is exactly

what Figure 3 shows. Moreover, experts were asked to locate parties in the same scale but in

terms of the importance given to each issue dimension, again ranging from 1 (low salience) to 20

(high salience). Here the pattern is reversed, exactly as we would expect if the political context

favored polarised views on the allocation of powers among the different ethno-linguistic

communities. Parties have still clear incentives to give priority on this issue and moreover they

are even better off by adopting extreme pro-decentralisation stances.

48

See indicatively, WG Jacoby, ‘Loess: A Nonparametric Graphical Tool for Depicting Relationships Between Variables.’ (2000) 19 Electoral Studies, 577; G Fitzmaurice, NM Laird and JH Ware, Applied Longitudinal Analysis, (New Jersey, Wiley, 2004).


111888555


Figure 4: Polarisation in parties’ positions in the issue of British presence

in Northern Ireland and electoral success, 1998, 2003 and 2007.

Note: The solid line presents the locally weighted regression curve fitted into the scatterplot of absolute distance from

status quo (upper panel)/issue salience (lower panel) and parties’ vote share in each of the three elections.

The case for Northern Ireland is even more straightforward. The expert survey included a

scale that was designed only for this region: (1) stands for ‘*party+ opposes permanent British

presence in Northern Ireland; (20) stands for ‘*party+ defends permanent British presence in

Northern Ireland.’ Figure 4 presents the results. Again, a loess curve is fitted into a scatterplot

but this time the actual vote share of each party, rather than the change from one election to

the other, is presented. Moreover, to show how polarisation is still the main engine driving party

competition, instead of locating parties in the horizontal axis in terms of their actual position, we

place them in terms of their distance from the median, neutral position. Looking at the three

graphs from 1998 to 2007, we see that extreme positions to either of the two poles of the

dimension are still important predictors of electoral success. More interestingly, this is as true in

TTTEEEPPPJJJFFF

111888666

2007, if not even more, as it was in 1998. The lower panel of Figure 4 shows the equivalent

findings, using the 1 to 20 scale measuring the importance given to this issue by each party.

Figure 5: Change in parties vote share from 1998 to 2003 and 2007 against their postures on the

issue of British presence in Northern Ireland.

Note: The graphs are similar to the previous Figures with the only difference being that instead of the actual vote

share of each party, its change from 1998 to 2003 and 2007 is used for the plot.

A potential criticism to this finding could be that parties’ stances on this issue are by now

quite irrelevant with their electoral success but the relationship seems to hold because parties’

stances and their electoral share are moving slowly through time. Thus, although big parties in

Northern Ireland might hold very clear and opposing stances on this issue, Figure 4 might

provide a misleading pattern about whether this specific issue is still equally salient or whether

this association is driven by path dependence. To address this argument, Figure 5 shows the


111888777


change in parties’ vote share from 1998 to 2007 and from 2003 to 2007. Although here we do

have the important exception of the Alliance, a relatively moderate party that has shown to be

remarkably stable in its electoral performance during all these years, the pattern is otherwise

similar: parties that embrace extreme positions on this issue or (as shown in the lower panel of

the figure) parties that attach important weight on the question of the involvement of Great

Britain in the Northern Ireland -the exact question that the Northern Ireland Act 1998 purported

to solve- are likely to increase their vote share. Undoubtedly, consociationalism is probably the

most effective solution to stop intra-state conflict among different religious or ethno-linguistic

communities. However, it does not seem to perform well in gradually advancing the integration

of these communities by helping in the progressive evaporation of the traditional cleavages that

led to this solution at first place.

4.3 Public Opinion Polarisation in Northern Ireland

We now move to a more close inspection of the attitudinal implications of these designs for

public opinion. We start this time with Northern Ireland and examine the same argument but

from a different angle. Has the 1998 Act brought voters closer to the two main opposing parties

or do we still observe the typical ‘communicating vessels’ profile, whereby being close to Sinn

Fein automatically means being far from the Democratic Union Party (DUP)? Figure 6 shows that

the latter is more likely than the former. Again, this is probably more the case now than it was

some years ago. The data are from the 1989, 1994 and 2004 EES. The horizontal axis locates

people in terms of their feelings about DUP, whereas the vertical axis places respondents with

respect to their sympathy towards Sinn Fein. As a measure of party evaluations, we use a survey

item that has been explicitly designed for the measurement of party preferences.49 The question

resembles a hypothetical (probability) question and goes as follows: ‘In a scale where 0 denotes

‘not likely at all’ and 10 denotes ‘very likely’ how likely is it that you would ever vote for Party X?’

X typically captures all significant parties in a given political context. Previous research has

identified various aspects of this question that make it the most suitable indicator of party

preferences at least in a European setting.50 Indicatively, it needs to be mentioned that the use

49

J Tillie, Party Utility and Voting Behavior, (Amsterdam, Het Spinhuis, 1995).50

C van der Eijk and M Marsh, ‘Comparing the Validity of Non-Ipsative Measures of

TTTEEEPPPJJJFFF

111888888

of ‘ever’ breaks the link in respondents’ replies with the political setting within which the

question is formulated. To put it differently, although people are interviewed right after the

elections for the European Parliament, this question does not capture current electoral choice

but rather maps people’s more general party preferences.51

Figure 6: Probability to Ever Vote Sinn Fein plotted against respondents’

preference of DUP, 1989, 1994, 2004.

Note: A locally weighted regression curve is fitted in the scatterplot showing the negative correlation between

peopler’s preferences for Sinn Fein and their attitudes towards DUP.

Party Support in CSES and EES.’ Paper Presented at the Elections, Public Opinion and Parties Annual Conference, Manchester, UK, September, 2008.51

Moreover, the fact that people do not think in probability terms means that they do not put a score to each party so that the total sums one. This is important because it has been shown that preferences are non-ipsative, contrary to what has been assumed by binary vote choice indicators. See C van der Eijk, W Van der Brug, M Kroh and M Franklin. ‘Rethinking the dependent variable in voting behavior: On the measurement and analysis of electoral utilities.’ (2006) 25 Electoral Studies, 424.


111888999


The loess curve fitted through the scatterplot is clearly monotonic and indicates a rather

unsurprising pattern: the more people prefer Sinn Fein the less likely they are to attach a high

value to DUP and vice versa. Interestingly, however, this negative relationship seems to be

stronger in 2004 than either in 1989 or in 1994. The slope of the local regression curve is steeper

after the 1998 constitutional reform than before.52 It is probably safe to conclude that the

Northern Ireland Act 1998 does not seem to have mediated the polarised context within which

party competition develops in this region.53

4.4 Lost in Translation: The ideological dimension in an ethno-linguistically divided society

As a last step before we conclude, we test a more long-standing implication of consociationalism

in the formation and the evolution of the political system and party competition. People tend to

use the Left-Right dimension as a helpful shortcut that enables them to communicate their

attitudes and make their voting decisions, coping with the particularities and the complexities of

the political world. In almost every European country, these terms have meaningful

connotations creating the basis upon which parties try to differentiate themselves in order to

achieve votes. It is important to clarify that this is not a normative argument. However, previous

evidence has clearly suggested that a good indicator of the level of maturation of a given political

system is the extent to which the primary and most salient issues can be effectively incorporated

within a single encompassing and rather generic multiple-issues dimension. Almost invariably, it

is the classical ideological scale that serves in this role.54 Moreover, having such a summarising

spatial analogy that helps to introduce various social and cultural divisions facilitates partisan

discourse and creates incentives for electoral participation.55 Importantly, precisely as a result of

accommodating various otherwise bipolar issues, the left-right scale enables the visualisation of

52

That this is the case is also confirmed by a parametric estimation of the correlation in preferences between Sinn Fein and DUP. The ordinary least squares slope of the 1989 and the 1994 panels of Figure 6 is -.324 and -.331 respectively. The equivalent figure for 2004 is -.481.53

Since people in Belgium can only opt among the parties of their ethno-linguistic community, respondents of this country were only asked these question for the parties of their region. Thus, this analysis cannot be replicated with the simultaneous use of French-speaking and Flemish parties. 54

See H Thorisdottir, JT Jost, I Liviatan and PE Schrout, ‘Psychological Needs and Values Underlying Left-Right Political Orientation: Cross-National Evidence from Eastern and Western Europe.’ (2007) 71 Public Opinion Quarterly, 175. 55

S Verba, NH Nie and J Kim, Participation and Political Equality: A Seven-Nation Comparison (Chicago, Chicago University Press, 1987).

TTTEEEPPPJJJFFF

111999000

differences between parties and permits a more normal distribution of parties’ ideal points. This

simply means less polarisation and higher levels of political stability.

Let us see now whether consociationalism facilitates the emergence of such a salient

encompassing dimension in divided societies. For data availability reasons we can only test the

case of Belgium. The reason for this is that we use the 2009 -thus the most recent comparative

study that is available- EES, which does not offer a separate sample for Northern Ireland. That

said, this is the case that would make it more difficult to find a difference in a comparative

setting. Both Northern Ireland and, to an even higher degree, BiH have still recent memories of

conflict and violence. It is much harder to expect significant changes as a result of the

constitutional design among these countries. In Belgium, however, decentralisation has taken

place in a gradual and non-conflictual fashion. Moreover, Belgium is the case in which the time

that has elapsed since these institutional designs were first put forward is much longer, making it

more likely to expect that the two regional political systems have come to the point to resemble

their European counterparts. Moreover, the fact that vote can only vary within each ethno-

linguistic segment implies that there are more incentives to create inter-party differentiations in

ideological rather than nationalistic terms.

To examine how the typical left-right scale helps to map politics in the two regions we

proceed as follows. We first use a party-specific measure of the extent to which people agree in

their perceptions about the party’s location in the ideological scale. When this scale is relevant in

the political context, people may differ in their exact placement of a given party but they tend to

be closer in their perceptions than when this scale is not helpful to distinguish parties in political

terms. Imagine, for example, a scale ranging from 0 (left) to 10 (right) in which people have to

locate the German Christian-Democrats (CDA) or the Danish social-democratic party. Not

everyone is going to locate the party in the same position. Actually, it is likely to find people

locating a given party in the exact opposite pole of the dimension. However such cases are

probably going to be only a trivial portion of the sample. If this is the case, then one could argue

that people tend to agree in how they perceive a given party.

To measure the level of perceived agreement in parties’ ideological stances, we use a

formula that gives room to what is known as van der Eijk’s coefficient of agreement (A), a

measure that ranges from -1 (which means complete disagreement to 1 (complete


111999111


agreement).56 If everyone locates the SPD at point 5 (or any other point) this coefficient is equal

to 1. If half people locate SPD in 0 and half of them in 10 the mean is going to be the same but A

is now -1. Last but not least, if an equal portion of the sample locates the party on each one of

the points of the scale, A becomes 0. Evidently, for this last case, LR is not a good predictor of the

party’s issue positions since there is a complete lack of understanding about where the party

stands. Now, if this pattern is observed for most parties of the political system, left-right ceases

to be a good encompassing dimension to map politics in this context.

The problem is that A is a measure that refers to each party scale. In most party systems

we have more than one parties. What we need here is an encompassing measure of agreement.

To take such a measure we calculate the party-specific coefficients and then we take the

weighted average to come up with a single A for each country. Parties are weighted according to

their vote share in the last national election. In this way, rather than assuming that all parties are

equally influential we accept that parties that take more votes are more likely to influence

regional (in Belgium) or national (in most other countries) political affairs.

Table 1: Average coefficient of perceived agreement

in parties’ ideological positions, EES 2009.

Agreement coefficient LR

Flanders 0.297

Wallonia 0.445

(Brussels) 0.468

France 0.633

Netherlands 0.501

Germany 0.543

56

C van der Eijk, ‘Measuring Agreement in Ordered Rating Scales.’ (2001) 35 Quality and Quantity, 325.

TTTEEEPPPJJJFFF

111999222

Agreement coefficient LR

Denmark 0.536

Greece 0.507

Spain 0.547

Sweden 0.592

Italy 0.62

Austria 0.499

Portugal 0.581

Basque Country 0.315

Catalonia 0.464

Note: Entries are weighted averages (according to the vote share of each party in the last national election).

Table 1 shows the results from Flanders, Wallonia and the region of Brussels. To be able

to evaluate these figures, we have also calculated the same coefficient for a number of European

countries. Apparently, the level of agreement about where parties stand in the left-right

dimension is almost half in Flanders than in an average European country. The figures for

Wallonia and Brussels are slightly higher but again the lowest among all other countries. The

pattern is rather unequivocal. Actually, as is seen in the last two rows of the Table, the level of

agreement among the Dutch-speaking and the French-speaking is to be only compared with that

found in the two regions of Spain with salient regional parties, namely Catalonia and the Basque

Country. Clearly, the party system in the Dutch-speaking as well as in the French-speaking region

of Belgium have not managed yet to surpass the ethno-linguistic cleavage so as to normalize


111999333


their political system, thus shifting to typical economic and social debates that can be easily

captured within the left-right dimension.57

Table 2: A comparative look at the importance of ideological

proximity across different European political contexts.

LR-proximity EU-proximity LR-direction EU-direction

Flanders

N: 437 (3933)

-.039

(-.044 -.034)

-.022

(-.026 -.016)

.086

(.073 .099)

.053

(.041 .065)

Wallonia -.043

(-.051 -.036)

-.018

(-.026 -.011)

.090

(.071 .110)

.028

(.010 .045)

(Brussels)

N: 198 (1821)

-.046

(-.059 -.033)

-.046

(-.061 -.032)

.130

(.094 .165)

.077

(.039 .114)

France

N: 1007 (7620)

-.059

(-.063 -.056)

-.009

(-.013 -.006)

.144

(.135 .153)

.017

(.010 .024)

Netherlands

N=1003 (9047)

-.058

(-.061 -.055)

-.026

(-.030 -.023)

.151

(.142 .156)

.062

(.053 .071)

Germany

n=862 (4691)

-.068

(-.073 -.062)

-.020

(-.026 -.015)

.159

(.145 .172)

.044

(.032 .055)

Denmark

N=911 (3979)

-.060

(-.064 -.057)

-.025

(-.029 -.021)

.173

(.164 .182)

.064

(.055 .074)

Greece

n=822 (5119)

-.064

(-.073 -.055)

.0001

(-.008 .009)

.174

(.152 .195)

.022

(.003 .042)

57

Since these figures come from different samples, we have engaged once again in bootstrapping in order to obtain levels of uncertainty associated with each of these estimates. We have used 100 bootstrapped samples and with these we have constructed a distribution of A’s across each country. Although we do not present the confidence intervals here, in none of these cases are these differences even close to come from a true null.

TTTEEEPPPJJJFFF

111999444

LR-proximity EU-proximity LR-direction EU-direction

Spain

N=1002 (7994)

-.057

(-.062 -.052)

-.019

(-.025 -.013)

.158

(.146 .171)

.052

(.039 .066)

Sweden

N=1003 (7002)

-.062

(-.065 -.059)

-.021

(-.024 -.018)

.170

(.163 .178)

.054

(.046 .062)

Italy

N=791 (5341)

-.054

(-.058 -.050)

-.013

(-.017 -.009)

.143

(.134 .152)

.034

(.026 .043)

Austria

N=802 (3447)

-.057

(-.061 -.053)

-.015

(-.018 -.011)

.140

(.129 .151)

.029

(.021 .037)

Portugal

N=891 (6336)

-.059

(-.064 -.053)

-.013

(-.018 -.008)

.145

(.131 .158)

.018

(.008 .029)

Basque Country

N=211 (1221)

-.009

(-.012 -.005)

-.012

(-.015 -.008)

.025

(.014 .035)

.024

(.015 .034)

Catalonia

N=356 (1562)

-.022

(-.029 -.015)

-.015

(-.021 -.009)

.056

(.038 .074)

.041

(.025 .058)

Note: Entries are individual fixed-effects regression coefficients with bootstrapped confidence intervals in parenthesis

(bootstrapped samples have been taken independently within each country). Dependent variable is PTV, within the

stacked data matrix. EU unification is used to facilitate the evaluation of between country differences in the LR

dimension. N denotes Second Level units (individuals) nested within First level units shown in parentheses

(individual*party combinations). Indidivudal-level intercepts are included in all models.

But how does this difference in agreeing about where the parties are located in LR terms

manifest itself when it comes to voting behaviour? Table 2 shows an important implication of

this pattern. Every row denotes two different regression estimates. The first comes from the

inclusion of the the first two predictors shown in columns 1 and 2 of the Table. The second


111999555


comes from the inclusion of the last two predictors, as shown in columns 3 and 4 of the Table. If

ideology is an important constraint that shapes preferences, ideological proximity (the extent to

which the respondent locates her views close to the point in which she places the party in the

same 0 to 10 scale) should help predict vote choice. To see whether this is the case, we use as

our dependent variable the Probability to Vote (PTVs) questions that were described earlier. To

avoid connecting coefficients with particular parties, we have ‘stacked’ the original dataset so

that the unit of analysis becomes the combination of individual with the party. Each individual

has as many observations as the number of PTV questions to which she has given a valid

answer.58 The resulting variable ranges also from 0 to 10 and refers to a generic party

preference, or what makes a party attractive -without suffering from problems of party label,

history or size. The main predictor of interest is the measure of ideological proximity, a measure

that counts the quadratic distance (analogous to the absolute value, i.e. (3-5)2=(5-3)2=4, with the

only difference that extreme values are overrepresented) between respondent and party and

ranges from 0 (absolute congruence) to 100 (absolute divergence, individual is in point 0 and

party is in point 10 of the scale, or vice versa). Clearly, since the closer the respondent finds

herself to the party the more likely it is to like this party, we expect this coefficient to be

negative. The hypothesis tested here is that since in Belgium people do not have a good

understanding of where parties are located in this scale, LR should be a weaker predictor of party

preference in this case than in the rest of Europe. To better evaluate differences, we also add an

equivalent proximity measure but this time measuring people’s similarity in their preferences

about the process of European unification (0: has gone too far/10: needs to go further) with the

parties. As is clearly seen in the first two columns of the Table, the difference between the two

coefficients (the one measuring the effect of LR-proximity and the one measuring the effect of

EU-proximity) is much smaller for Flanders (non-existent for Brussels) than for all other European

countries. On average the coefficient of LR is about the triple of its counterpart related to EU. In

Flanders, it less than double. This is not because the EU-proximity indicator is particularly high. It

is mainly because the LR-coefficient is exceptionally small. Moreover, the fact that the EU-

58

See for a detailed description of this method W van der Brug, ‘Perceptions, Opinions and Party Preferences in the Face of a Real World Event: Chernobyl as a Natural Experiment in Political Psychology’, (2001) 13 Journal of Theoretical Politics, 53.

TTTEEEPPPJJJFFF

111999666

proximity might appear slightly higher in the Belgian areas is mainly because the LR cannot

absorb this effect (i.e. cannot represent these differences in the encompassing ideological scale)

as effectively as in most other countries. Once again, the only examples that bear resemblance

to this pattern are Catalonia and the Basque country, where regionalism cross-cuts the existing

traditional cleavages.

The last two columns of the Table present similar evidence but instead of using LR

proximity employ an alternative measure of similarity between parties and voters which is

generally known as the directional measure.59 The choice of this alternative coding of

preferences is based on previous findings that have advocated that more polarised contexts are

more likely to encourage a directional thinking about parties.60 If this is what drives our results

here, then the difference between the three Belgian areas and all other cases should be

vanished when the directional indicators are used. However, as we see in the last two columns

of Table 2 this difference is actually exacerbated. Whereas in Flanders the difference in the

magnitude between the two scales is approximately 20 per cent, in all other countries, the LR

measure is on average about four times higher than the equivalent measure of EU unification.

Thus, as we see, the pursuit of consociational principles comes with its own price: the resulting

political contexts have important difficulties absorbing the traditional cleavages that created

these institutional mechanisms. In other words, by protecting these principles consociationalism

keeps them pertinent and makes it difficult for a given political context to form a more

standardised and less polarised basis of party competition.

59

The directional model of party competition does not see the scale as an indicator of distance but rather as an indicator of direction and strength of preference. According to this theory, people who are indifferent locate themselves in the middle of the scale. If they care about the issue, they choose the side (the direction of change from the status quo) they prefer and denote their strength of preferences according to whether they are located near to the extreme or not. This idea is captured in empirical terms by the product between the party term and voter’s term. Imagine the scale both for the party and for the voter is centered, the midpoint, 5, becomes zero, and thus 0 is now -5 and 10 is 5. The sign of this product signals convergence in directions, i.e. if it is positive it means that both the voter and the party advocate policy change towards the same direction; if it is negative it means that they advocate opposing policy change. The magnitude of this product denotes the level of commitment on behalf of the party; the voter; or both. This is the indicator used to measure LR-direction and EU-direction in columns 3 and 4 respectively.60

S Pardos-Prado, and E Dinas, ‘Systemic Polarization and Spatial Voting.’ (2010) 49 European Journal of Political Research, 759.


111999777


5. Conclusion

One of the main scopes of the present paper has been to depict how the

main model of democracy used in Europe an divided societies influences the

choice of the electoral system. As we have seen in parts 2 and 3, following

Lijphart’s contention, all three consociational constitutional structures

analysed before have opted for variants of PR. However, when one looks

closer to the electoral results in those polit ical systems, the integrationist

effect of the chosen polit ical and electoral system is highly questionable.

In fact, Benjamin Rei l ly has argued that PR allows for

the development of hard-line nationalist political parties, who can achieve electoral success by making narrow, sectarian appeals to their core ethno-political bases . . . the surest route to electoral victory under PR is to play the ethnic card – with disastrous consequences for the longer-term process of democratization.61

It is exactly those l imits to the integrationist effects of

consociationalism that have led even Elazar to argue that ‘consociational

regimes retain their consociational character for two generations, no

more.’ 62

However, one should be careful not to overplay the signif icance of

the constitutional architecture and the electoral systems to the failure of

those societies to overcome their ethno -linguistic and/or religious

cleavages. In fact, i t is exactly those consociational stru ctures that have

allowed the Northern Irish and the BiH cit izens to enjoy a more peaceful

polit ical l ife. So, more than a failure of consociationalism, the domination

of the polit ical scene from parties with a ‘separat ist’ agenda is a result of

the deep societal cleavages. In any case, the question that remains to be

answered is how the consensus model of democracy can lead to the

61

B Reilly, ‘Elections Post-Conflicts: Constraints and Dangers’, (2002) 9 International Peacekeeping, 132.62

DJ Elazar, ‘Federalism and Peace-making’; http://www.jcpa.org/dje/articlesfed-peace.htm.

http://www.jcpa.org/dje/articles/

TTTEEEPPPJJJFFF

111999888

progressive evaporation of the traditional cleavages that have led to the

adoption of this model in the f irst place. Under this perspective, the take-

away point from this analysis should not be that it is solely

consociationalism that drives polarisation. Although it might be one of the

factors that contribute in this process, it is applied in societies where the

ethical , l inguist ic or religious trauma is already present. What however

comes rather eloquently from the empirical examination of Belgium and

Northern Ireland is that although consociationalism is probably the only

way to keep the broken limbs together with the rest of th e body, it is not

actually a remedy that heals these wounds.. .


111999999


AV OR NOT AV? Lessons from Australia

Graeme Orr and K D Ewing*

I. Introduction

The United Kingdom (UK), long a bastion of first-past-the-post elections, is considering a

move to the alternative (AV) or preferential vote used in Australia. A referendum is

planned for 5 May 2011 to consider adopting AV for elections to the Westminster

Parliament.1 The referendum is the outcome of coalition negotiations between the

Conservative and Liberal Democrat parties. Ironically, the AV system is not the first

‘preference’ of either party. Conservative policy has been to maintain first-past-the-post

electoral laws; Liberal Democrat policy has until now been to push for proportional

representation. Interest in the AV system in the UK has, historically, been more associated

with elements of the Labour Party. For instance Labour proposed a reform bill including AV

as long ago as 1930,2 and pledged in its 2010 manifesto to hold a referendum on the topic.

The AV system has attracted interest in other parts of the world, including the United

States where it is known as ‘instant run-off voting’.

In this paper we consider: (i) the nature of the alternative vote (what is it?); and (ii)

its implications for voters, parties and candidates, and regulators (how will AV operate in

practice?). The practical implications - not of AV but of the run-off or second ballot for

which AV is a substitute – were once excoriated by Ramsay MacDonald for giving rise to

‘degrading bargaining, bribing, and other ways of cadging for majorities which would

follow the announcement of the figures of the first election’.3 We can consider whether

concerns of this nature are legitimate or justified, in view of Australia’s rich experience

with AV. Australia has long used AV not only for its triennial national elections, but also for

* Respectively Associate Professor, Law School, University of Queensland and Professor, Law School, King’s College London. 1 Parliamentary Voting System and Constituency Bill 2010 (UK). See further, House of Commons, Political and

Constitutional Reform Committee, 3rd

Report of Session 2010-11. 2 Representation of the People (No 2) Bill 1930 (UK).

3 J Ramsay MacDonald, Socialism and Government, vol 1 (1909), p 140.

TTTEEEPPPJJJFFF

222000000

state elections, all but one of which involve AV.4 Over 73 percent of Australia’s

parliamentarians are elected by AV; the rest, mostly in upper houses, are also elected by

preferential voting, but proportionally through the single-transferable vote or STV.5

Whilst preferential voting may liberate electoral choice, it comes at a cost, in

particular of disenfranchising electors who make mistakes. The Australian experience also

teaches that the competition for preferences motivates party apparatchiks to barter over

and campaign for preferences —which have become a new kind of currency— in ways that

bear out some of Ramsay McDonald’s concerns. This paper discusses these costs and

effects and the regulatory measures laws implementing AV may require to address

unintended consequences and attempts to game the system.

II. The Alternative Vote

The ‘alternative vote’, or AV for short, is a political science term for a form of voting in

which electors rank candidates in order of preference. AV goes by various names, notably

‘instant runoff voting’ in the US and ‘preferential voting’ in Australia. Each elector has a

single vote, but that vote can be transferred between candidates, during the count,

according to the elector’s ranking of preferences. The invention of preferential voting from

the 1850s, in the form of multi-member electorates using proportional representation

under STV, is most associated with Thomas Hare. It was particularly promoted by John

Stuart Mill. Then, in the 1870s, an MIT Professor of Architecture William Robert Ware

devised AV as a majoritarian, single-member simplification of STV.6 Thus British and

American ingenuity combined to give birth to the concept of AV. Australia is often credited

with originating electoral ideas when it is better described as an early adopter and refiner:

Australia ‘does indeed deserve her reputation as a seed-bed of experiment’ but the seeds

4 The small state of Tasmania does not use AV for its lower house.

5 M Mackerras, ‘Single Transferable Vote Systems in Australia’ (1996) 34 Representation 62 at 62. Papua New

Guinea, a former Australian territory, is the only other country with significant experience with AV: B Reilly, ‘Preferential Voting and Political Engineering: a Comparative Study’ (1997) 35 Journal of Commonwealth and Comparative Politics 1. Outside Australia, STV is used most notably in Ireland and Malta: M Mackerras and W Maley, ‘Preferential Voting in Australia, Malta and Ireland’ (1998) 7 Griffith Law Review 225. 6 See B Reilly, ‘Preferential Voting and its Political Consequences’ in M Sawer (ed) Elections: Full, Free and Fair

(1997), p 78 at 83.


222000111


often ‘have come from outside’.7 Curiously then, whilst AV is today pitched as a borrowing

from Australia, far from being a purely Australian discovery, ‘British debates over

preferential electoral systems had the most influence on events in Australia’.8 AV is most

obviously contrasted with first-past-the-post voting or FPP for short. FPP is the simple,

traditional form of voting for Westminster elections. It derived from the simple method of

early polling where electors merely had to publicly declare the name of their favoured

candidate: it matured in the ballot era into a method where electors just have to place a

cross in the box beside their favoured candidate’s name. (To illustrate the additional

complexity of choice under AV, a sample Australian House of Representatives ballot paper

is appendix 1.)

How does AV work once the ballot boxes are opened? Initially, invalid or informal

ballots are put aside. Then the ‘primary’ (or first preference) votes are counted. If no

candidate has 50 percent of the primary votes, the count proceeds by eliminating the least

popular candidates, in turn, and transferring each of their ballots to the next preferred

candidate remaining in the count, in accordance with each elector’s intention. This

continues until one candidate has a majority.9 In contrast, FPP often fails to produce a

winner with a majority of the vote, particularly in three-way contests or where there are

numerous minor parties on the ballot. (Candidates who top FPP polling with under 50

percent of the vote are known in the jargon as ‘plurality’ winners.) The chief institutional

attractiveness of AV is the extra legitimacy it may give successful MPs and in the formation

of government. It has a second, democratic attraction of giving electors more choices.

(Indeed it may encourage parties to run candidates in electorates they cannot win).

Similarly, under AV electors do not need to vote strategically or vote-barter,10 and parties

7

W ‘Keith’ Hancock, Politics in Pitcairn: And Other Essays (1947) 81, quoted in D Farrell and I McAlister, ‘1902 and the Origins of Preferential Electoral Systems in Australia’ (2005) 51 Australian Journal of Politics and History 155 at 158. Compare D Butler, ‘Australia’s Contribution to Political Institutions’ (2001) 7 The New Federalist 47 at 47: ‘Australia has been the most fertile source of constitutional experimentation, both at the federal and the State level.’ 8

Farrell and McAlister, ibid at 158. 9

The official count continues, for indicative purposes, to produce a final ‘two-party preferred’ or ‘two-candidate preferred’ result. This helps estimate swings needed to win elections. In a largely two-party system like Australia this can be done nationwide as well as constituency-by-constituency: a nationwide figure is less sensible in genuinely multi-party systems. 10

As to which see B Watt, UK Election Law: a Critical Examination (2006), pp. 76-82.

TTTEEEPPPJJJFFF

222000222

need not withdraw candidates strategically. Supporters of parties that run third or worse

do not feel their vote is wasted. It achieves these benefits by avoiding vote-splitting. That

is the effect, familiar in both the UK and US, where candidates with similar ideological or

policy attractions divide the vote, permitting a rival whose positions do not command

majority support to win. For example, Ralph Nader’s progressive Presidential candidacy in

2000 perhaps deprived Democrat Al Gore of victory over Republican George W Bush.

Similarly, during some periods of Conservative rule in the UK, there has effectively been a

Lib-Lab opposition with majority electoral support, hindered by FPP yet unwilling to form

an electoral coalition to minimise vote-splitting.

The purpose of AV is to ensure a majoritarian outcome in each electorate. However,

AV does not guarantee that a governing party will have earned a majority even of the two-

party preferred vote (that is, of votes finally tallied as if all races had resolved into a battle

between the two dominant parties). This is not a flaw unique to AV: it can be seen even

more starkly in FPP results, and is a result of the ‘winner-takes-all’ nature of single member

electorates. Such distortions of proportionality tend to be exacerbated by quirks in

electoral boundaries and in a party’s support being concentrated in a few safe seats or,

conversely, spread too thinly across the country. In short, whilst the family of STV systems

employed in multi-member electorates achieves proportional representation, its AV child,

when employed in single member electorates, is majoritarian rather than proportional.

Unless minor party support is concentrated, they will likely be under-represented.

Nonetheless outcomes tend to mirror the ‘two-party preferred’ vote. For example in

Australia in 2010, the Labor Party secured 50.1% of that vote to 49.9% for the Conservative

opposition: each of these major parties won 72 seats.

That said, AV comes with several potentially significant costs. One is that voting is

rendered more complex. For the first few elections it is employed, the number of

unintentional informal votes will inevitably rise. Such wasted votes will then remain higher

than they would be under FPP. Second and more profoundly, whilst removing some of the

discouragement to supporters of third and minor parties, AV reconfigures choices for both

electors and campaigners. Under AV, electors are asked not just to vote for the party or

candidate they most like, but also to rank parties for whom they may have little


222000333


enthusiasm. For some, this may transform elections into a choice of the least disliked,

rather than the most liked. Parties in turn are encouraged to fight for preferences as much

as primary votes, especially in key marginal electorates.

III. Implications of AV for Voters

Variations of AV

If we think of AV from the viewpoint of democratic considerations, it is right to focus first

on its effects on electors although, as we shall see, the effects on political parties may be

just as great. At this point, it is necessary to understand that there are variations of AV,

and that much will depend on the variation adopted. The law can mandate that electors

express a full set of preferences. Or it can permit electors to rank as many preferences as

they wish: that is the optional preferential vote or OPV.11 Full preferences are mandated in

most Australian jurisdictions, and have been since 1918 for national elections, but in

recent years two States have employed OPV.12 Full preferences make life easier for the

parties by limiting the fear of vote-splitting. But mandating full preferences puts a heavier

onus on electors. If electors wish to vote formally, they must rank all candidates on offer.

(This is subject only to ‘savings’ provisions: an elector who mistakenly repeats a number on

their ballot might be treated as having voted validly up to that point).13 Mandating full

preferences also may appear to guarantee legitimacy, by forcing electors to choose

between the likely governing parties, say Conservative and Labour, or at least between the

two most competitive parties in any particular constituency. But it does so even of electors

who might see that ultimate choice as one between Tweedledum and Tweedledee.

The Australian High Court has rejected challenges to laws mandating full preferences

on several occasions, both before and after the discovery of an implied freedom of political

11

Or indeed as further variations it may mandate electors to make two preferences; or permit electors to make two preferences, as in the case of the Representation of the People (No 2) Bill 1930 (UK). 12

In New South Wales, OPV is constitutionally entrenched so that a referendum is required to alter it: Constitution Act 1902 (NSW) s 7B and Sch 7. See also Electoral Act 1992 (Qld) s 113. An intermediate option is to require a minimum number of preferences other than one: but this makes little sense in single-member, as opposed to multi-member electorates. 13

Compare Commonwealth Electoral Act 1918 (Cth) s 270 (saving certain Senate ballots).

TTTEEEPPPJJJFFF

222000444

communication in the Australian Constitution.14 At first glance, the case for full strength AV

may seem illogical. How can an elector be forced to have a sincere or real preference

amongst candidates about which he or she is equally ignorant or uninterested? The Court’s

response has been that the question is ultimately for parliamentary sovereignty: the

design of voting systems is for the legislature not the courts.15 AV (in its different forms)

has long been accepted by political scientists as one reasonable option in a smorgasbord of

voting systems. While it may be hard on electors to compel them to rank all the options to

vote formally, the High Court has effectively said that it is not irrational to do so. Like it or

not, electors must have politicians to represent and govern them, so why not require

electors to choose, ultimately, between the parties of government? Parliament is entitled

to hope that, over time, even the most apathetic electors may, having been required to

express preferences, come to think of party politics not as a set of binary oppositions or to

plump for the party they have always plumped for, but to see the parties on a policy

continuum and to develop genuine preferences between them.

In democratic theory, OPV is preferable than mandating full preferences. It obviates

the problem of forcing electors to find some point of differentiation between all the

candidates. As former Australian Prime Minister Whitlam observed, OPV is ‘perhaps the

only electoral procedure in the world that permits electors to express their indifference to

candidates’.16 An elector can even number all squares, but leave the last one blank, in a

symbolic snub of an unloved party such as the BNP (on which more below). Yet although

OPV may be better from the point of view of personal freedom, it is not without its

problems. One institutional concern with OPV is that, in a worst case, it can collapse back

into FPP, because it permits electors to just plump for a single party. Whether because

they lack confidence in the electoral system (see below) or do not care to embrace

multiple parties or candidates, some electors may see the extra choice offered by AV as

nugatory. A dominant party with plurality support may encourage electors to plump by

running a ‘Just Vote 1’ advertising campaign. The Labor Party in Queensland employed just

14

Judd v McKeon (1926) 38 CLR 380, Faderson v Bridger (1971) 126 CLR 271 and Langer v Commonwealth (1996) 186 CLR 302. 15

Except in extreme circumstances of a blatantly undemocratic or partisan system. 16

EG Whitlam, The Whitlam Government 1972-1975 (1985), p 679 (emphasis added).


222000555


this strategy when its conservative rivals were divided in the 1990s-early 2000s. It was

pitched at encouraging conservative supporters to just vote for the party of their primary

choice, splitting the vote by limiting the flow of conservative preferences.

The latter kind of mischief-making can be dealt with two ways. One is by ensuring

that electors understand how the OPV system works, and that the choices are always up to

them. In a country with multiple voting methods such as the UK, such official education

will always be a challenge. The other would be to mandate full preferential voting. The

model to be considered at the 2011 UK referendum involves OPV. It is not merely that the

notion of electors being led by law to be good citizens sounds illiberal to British ears, a

concern hardly assuaged by the fact that (unlike in Australia) electors are free not to vote,

and that if full preferences were mandated electors might prefer to spoil their ballot rather

than express preferences between Tweedledum and Tweedledee.17 Under a system

mandating full preferences an elector would have to choose not to vote to avoid having to

express a preference between, say, the BNP and some other extremist party. In the end,

mandating full preferences involves requiring voters to express a preference between

parties which may be obnoxious to them. Yet without a measure of compulsion, AV is likely

to be diminished if a few voters submit a fully completed list of candidates, more submit a

partially completed list, and most vote only for a single candidate.

Complexity and Wasted Votes

Under FPP the problem of ‘wasted’ ballots is the problem of votes cast for a party that has

no chance of affecting the outcome. AV obviates that problem, but creates a different type

of wasted vote. These are informal votes: ballots rejected during the count for not

embodying a valid vote. These are typically cast by electors who either do not fully

understand the system or who make mistakes in completing their ballot. FPP is the

simplest of systems because it elicits the least information. It asks electors to plump for

their favourite party or candidate. AV elicits more information. If full preferences are

17

Faderson v Bridger (1971) 126 CLR 271 at 272. As it is, fewer than two percent of Australians deliberately spoil their ballots. Indeed, if too many people voted informally, electoral legitimacy could be undermined more than if electors did not turnout, as deliberate informality is a clearer protest than failure to turnout which can represent contentment, uninterest or protest.

TTTEEEPPPJJJFFF

222000666

required, it may, subject to any savings provisions, require an elector to rank order a dozen

or more candidates. The complexity problem with AV may be particularly acute at by-

elections, which often attract such large fields of candidates that they are compared to the

fields facing punters in the biggest horse races.18 Some of these candidates are attention

seekers, especially if the by-election attracts national interest. The problem of excessively

long ballot papers could be addressed by increasing the level of deposit required to

nominate, or raising the bar to registration of political parties.19 But such ‘solutions’ are

themselves problematic for democratic participation, and may run counter to one purpose

of AV, which is to enhance choice. More generally, the very shift to a new electoral system

- particularly a more complex one – will cause disenfranchisement by confusing electors,

especially in the initial, transition period.

Education is vital, but can only go so far. Under compulsory, preferential voting,

Australia has high rates of informal voting by Western standards: although it ran as low as

three percent in the 1990s, it exceeded five percent in 2004. An official survey ranked

Australia 46th out of 146 nations for informality, with an average of 3.7 percent. The UK,

under FPP, ranked the best of all: 146th with an average informal rate of just 0.2 percent.20

Concern over inadvertent informal voting is such that the Australian Electoral Commission

regularly publishes studies of all informal ballot papers. These suggest higher rates of

informality amongst electors from non-English speaking backgrounds, in the form of a

strong correlation between electorates with above average foreign language density and

informal voting.21 Low formal education is a secondary but ‘valuable predictor’ of informal

voting,22 and there is anecdotal evidence of electors with shaky handwriting misnumbering

18

Hence the phrase ‘a Melbourne Cup field’, applied to long ballots in Australia: ‘an Epsom Derby field’ might be the UK equivalent. 19

In Australian national elections, the deposit for a lower house nomination is approximately £250; half the amount for Commons’ elections. It is somewhat easier to register a new party in the UK: a fee of £150 and an intent to stand candidates. In Australia the fee is about £250 plus 500 members. 20

Australian Electoral Commission, Informal Vote Survey: House of Representatives 2001 Election (AEC, Research Report No 1, 2003) at 6-7 and Table 1. <http://www.aec.gov.au/pdf/research/papers/paper1/res_rep_01.pdf> 21

Ibid, pp 14-15. 22

Ibid, pp 15-16.


222000777


ballots.23 Only a little over one percent of the votes appear to be deliberately spoilt. On

average, then, upwards of three percent of turnout in Australia is wasted on accidental

informality, and the causes can be summarised as a mixed interaction between

compulsory voting, a high immigrant population and the complexity of the electoral

system.24 One argument in favour of OPV is that by not mandating full preferences, fewer

electors make errors in numbering a long ballot.

Any wasted vote is a shame, not least as the very point of electoral democracy is to

equally value each electoral voice. (Inadvertent informality is particularly ironic in

Australia, where every vote is so sacred that electors are fined for not bothering to vote).

More significant from a systemic viewpoint is whether increased informality – which will

surely occur in the UK - will have partisan effects capable of affecting very close races. The

Australian experience points to a mild effect on the Australian Labor Party, which enjoys

electoral advantage in both immigrant communities and poorer communities with less

formal education. If that concern mapped onto the UK system, it would be a particular

problem for the British Labour Party, since absent compulsory voting, such communities

already have relatively lower turnout than less marginalised communities. The effect might

parallel a problem which attracted international attention in the botched 2000 US

Presidential election, where poorer (and blacker) communities often had the oldest or

least efficient voting technology, and hence the highest rates of wasted votes.25 In a worst

case scenario, especially under AV mandating full preferences, activists might even load

ballot papers in targeted seats with dummy candidates, to create mischief. (This occurred

in a 1975 Australian Senate race).26 Having OPV rather than full preferences would make

23

Ibid, p 16. Some older electors thus inadvertently vote informally, despite having long experience with AV. (The Australian Electoral Commission has not found a significant correlation between electorates with above average populations of octogenarians and informality, but admits the elderly are ‘vulnerable’ to informality.) 24

I McAllister and T Makai, ‘Institutions or Protest? Explaining Invalid Votes in Australian Elections’ (1993) 12 Electoral Studies 23 at 23 and following. This thesis is supported by L Hill and S Young, ‘Protest or Error? Informal Voting and Compulsory Voting’ (2007) 42 Australian Journal of Political Science 515. 25

S Ansolabehere, ‘Voting Machines, Race, and Equal Protection’ (2002) 1 Election Law Journal 61 at 62-63: finding a ‘statistically significant association between race and residual votes’ , that is ‘unmarked, uncounted, and spoiled ballots’. 26

When conservative activists allegedly flooded the Senate ballot in the state of New South Wales. More recently, nine Christian Democratic Party candidates were nominated for the Bradfield by-election of 2009: with 22 candidates, the informal vote spiked to nine percent.

TTTEEEPPPJJJFFF

222000888

such a strategy less attractive. Computerised voting could also radically eliminate

accidentally informal votes, but at the risk of erecting other barriers to voting in poorer

and older communities.

For all these fears, however, AV is not inherently confusing. Albeit as consumers,

electors make choices everyday by ranking often similar products. A cola drinker might like

Dr Pepper best of all, but finding it not available at their local grocer have to give a ‘second

preference’ to Coca-Cola or Pepsi. Such choices are not unknown under FPP elections,

when an elector finds that their favourite party does not run a candidate in their

constituency. But while AV on its own may not be terribly complex, it remains the case that

the risk of informality can only increase when electors are asked to use two or more

different systems on election day (as the 2008 Scottish Parliamentary elections

demonstrated). In Australia, electors are at a heightened risk of informality because

Senate voting permits a ‘tick-a-box’ for a party list, whereas the House of Representatives

requires full preferences under AV. As a result, Senate informality is low; but at a cost to

formality in the House of Representatives:27 a double dilemma since governments are

formed in the latter chamber. A similar effect could be expected if the UK adopts AV for

the Commons but a different voting system is adopted for its upper house. Given the

complexity of the existing British electoral system these considerations tend to reinforce

the case for OPV, so that if some electors mistakenly think they can only vote ‘1’ for an AV

election, their primary vote at least will still be counted. (Whether such deemed formality

is consistent with the integrity of the process is another matter altogether.)

IV. Impact of AV on Political Parties

Reinforcing Party Cartels?

What then of the impact of AV the political parties? Every voting system has different

partisan effects, although these will vary depending on the shape of the party system. The

27

McAllister and Makkai, above n 25 at 34. The position is even worse for New South Wales and Queensland voters, who switch, sometimes in the same year, between the simpler OPV system for State elections to the more exacting full preferential requirement for national elections.


222000999


introduction of AV in Australia was driven by partisan concerns to limit vote-splitting when

one ‘side’ of politics was less united than its rivals. In the UK, AV is also driven by partisan

concerns (Liberal Democrat interests) and is likely to have partisan consequences

(probably Liberal Democrat interests). In many seats where Liberal Democrats run a

respectable second place, they may be elected on the preferences of supporters of the

third-placed party. Their parliamentary representation will be more proportionate to their

actual vote share and they will no longer be also-rans in terms of parliamentary power.

Their seat share may ensure them the balance of power, requiring Labour or the

Conservatives to negotiate coalition government or a memorandum of understanding to

support a minority government. The introduction of AV may thus make coalition or

minority government a typical rather than an atypical feature of Westminster politics,

moving the Liberal Democrats from a party of perpetual opposition to a party of perpetual

government.

However AV is far from an invitation to Italianate chaos. In the Australian experience,

AV has exhibited the tendencies to stability and majority government that are a hallmark

of FPP.28 Compared to proportional voting systems, AV does not present any positive

incentive to the fracturing of the party system (subject to opportunistic exceptions

discussed below). A splinter party or small party is unlikely to win any seats under AV, at

least until it builds strong primary support, though there may be circumstances where

strong preference flows or deals could help small parties such as the Greens to make a

bigger breakthrough than under FPP, especially if the two parties in a coalition government

were deeply unpopular. For the most part, the Australian experience is that AV has tended

to reinforce the large party cartel which operates in many systems with FPP. Indeed the

Australian system has been the most durable and rigid two party system of any

parliamentary democracy in the world. Races for the House of Representatives typically

reduce to a contest between the two major forces, Australian Labor and the Liberal-

National conservative coalition. In the British context, AV is likely to reinforce the cartel

operating between the main parties. It is unlikely to break open that cartel, though it may

28

C Bean, ‘Australia’s Experience with the Alternative Vote’ (1996) 34 Representation 103 at 104. This is why it is the second-choice, behind proportional representation, for many Liberal Democrats.

TTTEEEPPPJJJFFF

222111000

facilitate a process under way in which the two party cartel is replaced by a three party

cartel.

To reinforce this point, it is salutary to look at the experience of Australian national

elections since AV was adopted in 1918. Aside from the effects of two temporary splits in

the Labor party, only one minor party candidate has won a seat at a general election to the

Australian House of Representatives since 1919.29 The exception is where a popular local

independent emerges: in that situation, preferences can be crucial in electing that

independent or having his or her supporters decide the race.30 (Analogues of this of course

have occurred in UK even under FPP: for instance Wyre Forest’s election of an

independent on local hospital issues in 2001 and 2005.) This is not to say that small parties

are shut out completely from representation in Australia. Rather, the parliamentary role of

minor parties such as the Australian Democrats, Greens and One Nation in recent decades

has arisen because of the proportional, STV system for Australia’s Senate, not AV. There is

a parallel here in the UK, where the list system employed for European Parliamentary

elections has permitted parties such as the UKIP, Greens and BNP to gain a foothold. If

rainbow representativeness were the goal, proportional representation (say for a second

chamber) is a better avenue: it will not be achieved through AV. Indeed, where parties can

garner significant regional support, FPP does not prevent parliamentary representation, as

the records of the various nationalist or unionist parties in Scotland, Wales and Northern

Island demonstrates.31

An Expanded Role for Minor Parties?

If AV is likely to strengthen the electoral voice of the large parties, how will it affect their

behaviour during elections? Australian experience suggests that the main parties will seek

to do ‘deals’ with other parties, a practice that may be even more intense in the UK’s three

29

A Greens MP in 2010. See D Jaensch and D Mathieson, A Plague on Both Your Houses: Minor Parties in Australia (1998), pp 210-212. (Labor splits occurred during the Depression and the Cold War). The record at by-elections is almost as paltry, although The Greens briefly took a Labor seat in 2002, with a primary vote of just 23 percent. On the institutional factors involved see Jaensch and Mathieson, pp 173-182. 30

In recent decades independents have had success in Australian regional electorates, as rural politics has become more localised and fragmented in the face of drought and population decline. 31

Although a federation, there are no equivalent sub-national parties in Australia.


222111111


party system than in Australia’s two party system. ‘Preference deals’ are arrangements

created by horse-trading over preference recommendations. Preference recommendations

are where one party advocates or guides its members and supporters to allocate their

preferences in particular ways. Obviously no party can guarantee that members and

supporters will behave as requested.32 But where there is both loyalty to a party and

uncertainty about the other candidates (or the electoral system itself), this guidance can

be very influential. In Australia it is provided through the ubiquitous ‘how-to-vote’ card.

This is a flyer, usually distributed on the doorstep of the polling booth, but sometimes also

advertised in newspapers. In it, a party will replicate an image of the official ballot paper

and seek to instruct their supporters on how to allocate their second and later

preferences. Such guidance has most effect on the staunch supporters of the major

parties: people who habitually vote Labor or Liberal for example, and may not think much

about how to rank the minor parties. But it can also be very influential on supporters of

minor parties, who may otherwise be exercising a ‘protest’ vote. (Sample cards are at

appendix 2).

The UK context is particularly complicated and fluid. In many safe seats, one of the

three parties romps home with at least 50 percent of the vote: that party should continue

to win 50 percent of the primary vote under AV. In others, one party is dominant and the

other parties are so weak that preferences are unlikely to affect the result. But in many

seats, the vote is split three ways such that the preferences of the weakest of the large

parties will decide the outcome. Further, AV may encourage more candidates to run in

many UK constituencies, by lessening concerns about vote-splitting and ‘wasted’ votes. In

the 2007 Australian election, for example, the average number of candidates per

constituency was just over seven. AV can deal parties currently on the margins into the

political process: such as the Greens on the left and the BNP and the UKIP on the right.

(Conversely, the three main British parties could tacitly agree to not do any deals with the

BNP and to symbolically place it last on their how-to-vote recommendations).33 AV will

32

Unless electors are given an option of just voting ‘1’ for a party, and the party channelling those preferences according to a pre-determined flow, as occurs for Australia’s Senate elections. 33

The mainstream parties in Australia adopted this approach to ‘starve’ the racialist, nationalist One Nation Party in the early 2000s; the One Nation Party saw this as evidence of a cartel at work.

TTTEEEPPPJJJFFF

222111222

likely increase the collective vote share of minor parties, which is suppressed by FPP. At

the 2005 Commons election, for instance, the minor party share of the vote was just 5.6

percent. In the 2010 Australian House of Representatives, in contrast, it was over 18

percent.34 Whilst AV would put an end to tactical voting in the UK,35 in its stead it may

erect a market in preferences, which in Australia have become a form of ‘political

currency’.36 Even if preferences are irrelevant to the outcome in a majority of safe and

semi-safe seats, the fates of governments are rarely decided in such electorates. Under AV,

‘second preference’ strategies are crucial to the campaign strategies of every party that

has a realistic chance of winning seats.

Preference deals take several forms. They can be across the board and cover all

seats; or they may only cover particular key seats, with a party making no

recommendations to its supporters in other seats. Parties can do deals for quid pro quo

preferences: a party like the Greens might preference centre left parties in the lower

house in return for preference support in any upper house election. Aside from such

preference ‘swaps’, preference deals offer two benefits for smaller parties. The first is

policy influence, with preference deals providing an opportunity to exert some policy

leverage over the major party suitor. For example, the UKIP could seek a more euro-sceptic

manifesto form the Conservatives.37 In that scenario, a party otherwise marginalised under

FPP is dealt into the sphere of influence.38 It may become a kind of pre-electoral partner of

the major party, to the extent that it has an interest in the major party it preferences

winning seats and government. It will wish to demonstrate that is preference

recommendations make a difference in the outcome in key seats. Although such deals are

not binding contracts,39 a major party in government will only renege on such policy

34

D Kavanagh and D Butler, The British General Election of 2005 (2006), p 204. Compare Australian Electoral Commission <http://results.aec.gov.au/13745/Website/HouseStateFirstPrefsByParty-13745-NAT.htm> 35

On which, see J Curtice et al in Kavanagh and Butler, ibid, pp 238-245. 36

G Orr, ‘‘The Australian Experience of Electoral Bribery: Dealing in Electoral Support’ (2010) 56 Australian Journal of Politics and History 230. 37

Or the Greens could seek an environmentally focused manifesto from centre-left parties. 38

This is not to say a minor party has no influence under FPP. It can threaten to run and split the vote. 39

G Orr, ‘A Politician’s Word: the Legal (Un)enforceability of Political Deals’ (2002) 5 Constitutional Law and Policy Review 1.

http://results.aec.gov.au/13745/Website/HouseStateFirstPrefsByParty-13745-NAT.htm


222111333


commitments at a cost to its credibility. For smaller parties, preferences under AV thus

offer a form of influence without representation.

The other benefit a minor party may gain from a preference deal is tangible support.

If the major party to the deal has spare resources (financial or otherwise), it might agree to

assist the minor party in its campaign. A typical way this occurs in Australia is through

lending volunteers to distribute the minor party’s literature, especially on polling day. At

first glance it seems perverse that one party, in competition with another, would assist the

other with scarce campaign resources. But it is in the major party’s interest that the minor

party’s preference recommendations are publicised. As we shall soon discuss, such

tangible support raises the almost forgotten problem of electoral bribery, invoking the

spectre of Ramsay McDonald’s fears cited at the start of this paper. Of course elements of

such calculations and influence are not unknown to FPP politics. The threat of minor

parties running candidates and bleeding or splitting the major party vote may already

prompt the major parties to make policy appeals to electors who might otherwise support

those minor parties. But preferential voting under AV adds a new dimension to FPP

politics, as preferences become a currency and minor parties are directly dealt into the

system.

V. Impact of AV on Regulation

Bribery

Finally, what about the implications of AV for the regulation of elections? British elections

are governed by the Representation of the People Act 1983 (RPA), the contents of which

generations of British public lawyers have been blissfully ignorant. (British public law is

more concerned with restraining the powers of representative institutions than

understanding their composition, purpose, or operation. Part II of the RPA deals with the

election campaign, and addresses a wide range of practices. Much of this was put in place

by the Corrupt and Illegal Practices Act 1883 (UK) in response to the then endemic custom

TTTEEEPPPJJJFFF

222111444

of crude vote-buying.40 Since then the concerns addressed by the Act have grown, as new

technology and new inventions (such as the motor car) contrived to change the landscape

on which elections were fought. The Australian experience suggests that the adoption of

AV and the likely trade in preferences will mean that some of the provisions of the 1983

Act may need to be adapted and expanded (remembering it is possible that Australian

practices will not be reproduced, but that different problems giving rise to different needs

may emerge).41 The first issue, however, is that preference deals under AV create the

spectre of electoral bribery, an offence largely forgotten in most liberal democracies.

The current definition of electoral bribery in the UK is narrow and dated. It is only

directed at bribes to influence anyone to ‘vote or refrain from voting’.42 If AV were

adopted this would need to be extended in two ways. First, to clarify that bribing a voter to

influence any of his or her preferences is an offence; and second to clarify that bribing

another party to influence its preference recommendations is an offence. Both types of

bribery are covered in the Australian legislation.43 However such clarification does not

settle the question of what influences will be deemed to be corrupt. If preferences are

necessarily a form of political currency under AV, then certain deals in them must be

acceptable. Thus, log-rolling on policy, to secure a minor party’s preferences is

unexceptionable: indeed it is part of the point of AV that it recognises the electoral worth

of minor parties and the votes they attract. Similarly, it seems unobjectionable for two

simpatico parties to simply agree to swap preferences. But what of a major party

promising resources -and especially money- to encourage a minor party to make a

favourable preference recommendation? Any hint of money or money’s worth influencing

preference recommendations should be unlawful for the same reasons that electoral

bribery was originally outlawed: to allow wealth to purchase electoral support offends

democratic egalitarianism and ramps up the cost of participating in elections.

40

G Orr, ‘Suppressing Vote-Buying: the ‘War’ on Electoral Bribery’ (2006) 27 Journal of Legal History 289. 41

The regime for regulatory adaptation is recognised in Representation of the People (No 2) Bill, cl 1(2). 42

Representation of the People Act 1983 (UK), ss 113-114. 43

Commonwealth Electoral Act 1918 (Cth), s 326(2)(a) (‘any vote’) and 326(2)(d) (preference recommendations).


222111555


The use of bribery law to deal with the latter type of case is not, however, without

difficulty. A payment or provision of resources may have multiple purposes. One purpose,

which most would consider corrupt, would be to reward the minor party for buying its

support. Another purpose, which would not be corrupt, is to help a less well resourced

party in a joint endeavour, by promoting its profile and preference recommendations.

These difficulties were highlighted in Australia in two high profile cases, one involving

Liberal Party Minister Vic Garland in 1976, and the other a prominent Labor front–bencher,

and now Minister for the Treasury, Wayne Swan, in 1993. Both men were forced to stand

aside during investigations after they admitted giving modest sums of money to help a

minor party campaign, during preference negotiations. In the earlier case, a Magistrate

dismissed the charge; in the later case the Federal Police did not recommend

prosecution.44 It helped in Swan’s case that the ultimate decision as to preference

recommendations was made in the minor party’s national office, whereas his payment was

at constituency level.

Unethical political deals or arrangements are rarely made public, and even when

they are, they are difficult to prove on any criminal standard of proof. The Australian

experience has been that although prosecutions are rare, the law of corrupt practices

(especially bribery, treating and undue influence) at least provide a backdrop against which

the ethics of campaign practices can be discussed and excesses inhibited. Aside from

ensuring that the existing UK law covering electoral bribery potentially applies to

preference voting and recommendations, it is not immediately clear what institutional or

legislative measures could be taken to guard against corrupt deals without stifling political

practice. For instance, centralising decision-making about preferences in the hands of

national-office, party cadre may help professionalise politics in the hands of those more

likely to be risk averse and take legal advice. But such centralisation creates its own deficits

for participatory democracy. A simple legislative provision could be drafted be to ban

payments or transfers of resources between otherwise rival parties or candidates, but this

would breach the freedom of association of parties that are simpatico or in informal

coalition.

44

See Orr, above n 37 at 230-233.

TTTEEEPPPJJJFFF

222111666

How-to-Vote Material

The emergence of horse trading in preferences under AV thus raises questions about the

adequacy of the existing law, particularly relating to bribery. Preferential voting will affect

campaigning more broadly, in ways that may also require regulatory attention. If

Australian experience is any guide, an inevitable development will be the emergence of

how-to-vote material, through which parties seek to influence how electors complete their

ballot. The ubiquitous how-to-vote cards, distributed outside polling stations, form the

most visible sign of the crucial nature of preferences under AV. To some, this is a perfectly

legitimate practice, and a cherished election-day activity through which activists can

support their party. To other observers, however, these cards are a menace: they may lull

some electors into thinking that to vote Labour, for instance, one must follow the Labour

how-to-vote card. Such a perception dulls one of the democratic purposes of AV, namely

to ask electors to decide for themselves which of the major parties they wish to win in

their electorate. In between, others see how-to-vote cards as mostly wasted paper.

Provided each candidate has a party name or ‘independent’ alongside their name, any

informed elector should be capable of casting their ballot without the assistance of such a

card.45

In response to some of these concerns, two Australian jurisdictions have legislated a

one hundred metre cordon sanitaire around polling booths, within which no electoral

material can be distributed.46 But although this prevents the distribution of the material

where it will have the greatest impact, it does not prevent the distribution of how-to-vote

material by other means, such as via letter-boxing or newspaper advertisements. Electors

are of course free to bring such material with them to the polling station, even in these

jurisdictions. A less restrictive alternative would be to ban the distribution of how-to-vote

cards in the vicinity of polling stations and instead require the electoral officials to post

samples of each party’s card near each polling compartment for those electors who wish

45

In Australia, the preferences of Greens supporters tend to be invariant to their party’s recommendations, probably because such electors tend to have strong ideological commitments. 46

Electoral Act 2004 (Tas) s 177; Electoral Act 1995 (ACT) s 303. These rules only apply to sub-national elections.


222111777


to consult them. This solution cuts down on the environmental cost and treats every

party, however big or small, as equals. It has not, however, been trialled as an option in

Australia, presumably because it deprives the major parties of their natural advantage of

being able to mobilise activists around every polling station.

A more serious concern with how-to-vote cards is not so much the fact of their

distribution but the manner and content. Problems routinely arise in Australia when major

parties circulate ‘second preference’ how-to-vote cards. These are not distributed to guide

the major party’s supporters, but to appeal to the supporters of a particular minor party.

The temptation has been to disguise the true source of the card: the aim is to harvest

preferences by luring minor party supporters into assuming the cards represent the

recommendation of that minor party. Subterfuges include authorising the card under the

name of an obscure major party official without clear mention of that party and adopting a

different get-up to the major party’s established brand. To translate a common Australian

example into the UK context, the Conservatives might publish a card titled ‘Think of Voting

1 for the Greens? Give your 2nd Preference to the Conservatives’. Whilst that text may

seem innocent, it would be less so if it were published on green, recycled paper. In two

egregious Australian cases, Australian Labor activists tried to mislead electors orally and in

the manner of their dress. In one case, they handed out visually neutral cards whilst saying

to electors ‘Voting One Nation?’ (implying that the card had come from the One Nation

Party). In another case, Labor activists donned light blue t-shirts with the slogan ‘Put Your

Family First’, whilst handing out cards designed to appeal to supporters of the religious,

Family First Party.

Such mischievous material and practices in Australia have given rise to ex tempore

injunction actions on polling day,47 and even to an election petition.48 In Australia it is an

illegal electoral practice to mislead an elector in the casting of their vote.49 Judicial

intervention however is an expensive stitch out of time. Election day applications are

47

Goss v Swan *1994+ 1 Qd R 40 (adopting test of ‘gullible’ elector). 48

Carroll v ECQ (No 1) [2001] 1 Qd R 117. 49

Commonwealth Electoral Act 1918 (Cth) s 329. The Australian High Court in Evans v Crichton-Browne (1981) 147 CLR 169 held that this offence applies only to misleading voters as to how to translate their intentions into a valid vote: this includes misleading an elector who wishes to follow a particular party’s preference recommendations.

TTTEEEPPPJJJFFF

222111888

hurriedly argued and polling may be almost over before an injunction can be served.

Worse, an election petition is not only post-facto, but a waste of money unless the winning

margin is extremely narrow. As a result, some Australian jurisdictions have adopted a

regulatory salve, and require all how-to-vote material to be registered well before polling

day. This gives electoral authorities and rival parties time to object to potentially

misleading material.50 For Australian national elections, new laws are mandating that

parties publishing ‘second preference’ cards display their name in legible font.51 Such

micro-regulation is unlikely to stop all mischief and malpractice, but if AV were adopted

elsewhere, it would be wise to legislate such measures. This is particularly so as the UK has

no culture of using the courts on election day, and post-election petitions, which provided

the traditional way of policing electoral cheats, have fallen into disuse.

50

Eg Electoral Act 1992 (Qld) Pt 9 Div 2. 51

Electoral and Referendum Amendment (How-to-Vote Cards and Other Measures) Act 2010 (Cth).

TTTEEEPPPJJJFFF

222222000

Appendix 2: Sample ‘How-to-Vote Cards’

Full Preferential Voting


222222111


Card under OPV

(Labor Party opting for a ‘vote 1’ only strategy)

TTTEEEPPPJJJFFF

222222222

‘Second Preference’ Card

(Issued by Labor Party, appealing to Family First Party supporters, although Family First

Party policy recommended not preferencing Labor)

is governability a constitutional...

Documents