Occasional Paper: United Nations Development Programme

View metadata, citation and similar papers at core.ac.
uk brought to you by CORE

provided by Research Papers in Economics
United Nations Development Programme
Human Development Report Office
OCCASIONAL PAPER
Background paper for HDR 2004
BUILDING INCLUSIVE STATES
Brendan O’Leary
2004/9
BUILDING INCLUSIVE STATES
Brendan O’Leary
Lauder Professor of Political Science
Director Solomon Asch Center for the Study of Ethnopolitical Conflict
University of Pennsylvania
Commissioned Paper for the United Nations Human Development Report 2004
1. Functioning states are necessary conditions of inclusion; and the major
causes of exclusion.
States matter more than societies in building inclusion. That is because they
define societies, and their possibilities1. Impersonalized institutions that have
some degree of centralized and procedurally governed political decision-making
characterize functioning states. They have coercive capacities: they can regulate
all instruments of potential public violence, and prevent or inhibit their own
agents from being predators. They express authentic legal sovereignty over
persons, property, and their movements, and are recognized as such entities by
their citizens, civil society organizations, and by other states. Through self-help
or alliances they can defend themselves against other states. Lastly, functioning
states are defined by their recognized sovereignty over territory, and its
accompanying prerogatives: control over entry and exit of persons and entities. If
Brendan O'Leary Page 1 first vs. Oct 03, corrected vs. 7/12/2004
1
states lack these capabilities they cannot protect elementary human rights,
promote human development, or be inclusive in any meaningful sense.
Conversely, failing and failed states are personalized: dominated by rulers – a
family, clan or clique - who do not distinguish public and private realms. They
become ‘kleptocracies’, governments of thieves, before or during their collapse.
They lack coherent, institutionalized, rule-governed patterns of centralized
decision-making that inhibit predation. They are predators. They lose their
monopoly on the regulation of coercion, and are challenged by guerillas,
paramilitaries, terrorists, Mafiosi; they may be invaded, looted, and occupied by
other states. They neither make nor enforce law. Those over whom they have
failed to rule despise them as much as they fear them. Functioning states pay lip-
service to failing states, avoiding their ‘officials’ if they can, invading ‘their’ lands
if they deem it necessary.
These properties of failed states should remind us that ‘inclusion’ is only possible
within well-ordered states. ‘Inclusivity’ and human development require more
than the diffusion of the right values; they need the soil of functioning states
because they are unlikely to grow in ‘anarchia’. Order, to put it bluntly, is a
necessary condition of inclusivity. Without it we have Thomas Hobbes’ condition
of ‘Warre’: ‘the time men live without a common Power to keep them all in
awe…in which every man is Enemy to every man; …wherein men live without
2
other security, than what their own strength, and their own invention shall
furnish them…In such condition, there is no place for Industry; because the fruit
thereof is uncertain: and consequently no culture of the Earth, no Navigation, nor
use of the commodities that may be imported by Sea; no commodious Building;
no Instruments of moving, and removing such things as require much force; no
Knowledge of the face of the Earth; no account of Time; no Arts; no Letters; no
Society; and which is worse of all, continuall feare, and danger of violent death;
and the life of man, solitary, poore, nasty, brutish and short’2. Hobbes was right
to emphasize the necessity of order, but wrong to mandate an authoritarian
solution. But we may still draw a lesson from his arguments: without a well-
ordered state there cannot be democracy, functioning markets, human
development or inclusivity.
Unlike Hobbes, we know that states may be lethal, more lethal than the ‘war of
all against all’. States are the most powerful agencies of exclusion, governments
the major killers in human history. Many state-builders and nation-builders have
been, respectively, people-killers and nation-killers. In the twentieth century
Rudolf Rummel has calculated that governments killed nearly 170 million people
within their borders, a figure that exceeds those killed in wars between states3.
Genocide, killing people because of their presumed ascriptive characteristics, has
been more common than most countries’ official histories acknowledge, and
governments have been the major perpetrators - and the list is not confined to
3
Ottoman Turkey, Nazi Germany, and Interahamwe Rwanda. Politicide, killing
those deemed political opponents of regimes, has been equally recurrent in
modernity. The Soviet Union, especially under Stalin, was the major killer regime
of the last century: nearly 62 million are estimated to have perished at its hands;
Maoist China was often as brutal, killing over 35 million people. Democide, the
killing of peoples, is the ultimate form of exclusion, and governments its major
perpetrators. Governments have also organized, encouraged, or not stopped,
the expulsion of whole categories of persons from their land borders or their
shores – people whom they have helped to define as undesirable, non-
indigenous, non-national, or disloyal. The twentieth century can be described as
the century of expulsion, of the ‘cleansing’ of populations, as much as one of
genocide. Governments have also been the prime architects of policies of
discriminatory control: organizing dominant national, ethnic, religious or
communal groups, disorganizing and subordinating other national, ethnic,
religious or cultural groups through policies of systematic discrimination, or
exclusion.
These should be commonplace thoughts. Regrettably, they are not sufficiently
appreciated. States define people’s life-chances, both individually and
collectively. And it is to their practices that we must look when considering how
greater inclusion, as defined in this report by Amartya Sen, may best be
facilitated. Exclusion, in the regulation of national, ethnic and religious
4
differences, is the product of coercive homogenization and of differential
treatment, privileging communities: counting some categories of persons and
communities ‘in’, as members, and others ‘out’. Genocide, expulsion and the
imposed partition of territories are modes of homogenizing peoples that are now
internationally outlawed, even if the relevant laws are frequently not enforced. If
and when international norms against genocide and expulsion are rigorously
enforced, through interventions ranging from diplomatic sanctions, economic
sanctions, military embargos (or lifting of embargos), military occupations, and
international criminalization of genocidal officials, then the worst forms of
exclusion may be halted or inhibited. Success in this endeavor requires both
moral universalism – genocide and expulsion need to be treated as criminally
wrong everywhere – and a re-orientation of the foreign policies of the great
powers, so that both their policy-makers and domestic constituencies see the
prevention and punishment of genocide and expulsion as in their interests4.
The record of recent history is mixed: it need not occasion despair. Sanctions
against exclusionary practices – against discriminatory racist, religiously
intolerant and xenophobic regimes – had major successes in the twentieth
century. The undermining of apartheid in South Africa had both international
and domestic sources, both moral and political causes. Decolonization, de-
segregation and civil rights movements are very important elements in the
affirmative and inclusionary history of the twentieth century. Inclusion, by
5
implication, encompasses sets of policies, institutions and practices that both do
not coercively homogenize peoples, that do not entrench unjustified differential
treatments of peoples, and that are embedded in the practices of well-ordered
states. Elaborating this understanding is the purpose of this paper.
2. Assimilation.
The political and cultural homogenization of people is the underling goal of
many states that promote assimilation. Historically it has been an important
policy of inclusion in two model and hegemonic states, the USA and France5,
both of which have influenced the political development of many member-states
of the UN. We may be tempted to define assimilation by dictionary entries in
English and French. The Oxford English Dictionary tells us that to assimilate
means to
1. Absorb and digest (food etc.) into the body, absorb (information etc.)
into the mind, absorb (people) into a larger group.
2. Make like; cause to resemble.
3. Make a (sound) more like another in the same or next word.
4. Be absorbed into the body, mind, or a larger group (Concise Oxford
English Dictionary 1990) .
The first, second and last of these meanings point towards the term’s political
connotations, and highlight the experiential fears attached to ‘absorption’ and
‘digestion’, as well as the more neutral implications of ‘becoming alike’, or
6
‘coming to resemble’. The Concise Oxford French Dictionary tells us that when
translating French into English that assimiler may mean
1. To assimilate; to liken, to equate;
2. To digest;
3. To integrate; s’~ qch. To be assimilated; to be integrated; s’~ qch., to adopt,
to incorporate (sth. foreign).
This suggests, among other things, that in French the notions of ‘assimilation’
and ‘integration’ are fused, which may be an outcome of French political
practices and beliefs. The same dictionary, in guiding us in translation from
English into French, counsels us that ‘to assimilate’ should be translated by
assimiler; but that ‘to integrate’ should be translated as compléter (when it means
‘to complete’), as unifier (when it means ‘to form into a whole’); and as intégrer
(when it means ‘to assimilate’) (Concise Oxford French Dictionary, 2nd ed. 1980).
These inspections suggest the need to move beyond ordinary usages in both
languages to achieve clarity.
Assimilation, as a strategy of national and ethnic conflict regulation, may be
defined as (the voluntary or coercive) erosion of politically and privately salient
cultural differences between groups, with the intent of achieving national or
ethnic homogenization. It achieves inclusion through making common citizens in
public and private domains who are not culturally differentiated for political
purposes. Acculturation involves assimilation by culturally weaker groups,
7
where, for example the members of group B absorb group A’s culture: A + B = A.
Fusion, by contrast, involves assimilation through a full-scale merger that
produces a novel entity, e.g. three groups A, B and C fuse to form a novel group,
D: A + B + C = D. What differentiates assimilation as a public policy from both
integration and multi-culturalism, it may be suggested, in accordance with much
recent social science literature, is that assimilation eliminates public and private
differences between people’s cultures, whereas integration eliminates public
differences between people’s cultures but not necessarily their private cultural
differences; multiculturalism, by contrast, maintains (some) public and private
differences between people’s cultures.
Coercive assimilation excludes certain identities by compulsion. It is experienced
by its victims as “linguicide”, “ethnocide” or “theocide”, referring to the
destruction of languages, hereditary communities of descent, and religious
institutions and practices respectively. Linguistic standardization is a common
feature of a coercively assimilationist state. One langauge, one grammar, and in
extremis one correct intonation is mandated by the public authorities. One
language has a monopoly of the public domain; other languages (stigmatized as
dialects or patois) may be illegalized in private – as General Franco of Spain
illegalized Euskedi in the Basque country, and as General Attaturk illegalized
Kurdish in Turkey. Schools, universities, armies, parliaments, assemblies and
councils, bureaucracies and legal institutions are used to socialize all the targets
8
of the policy in the mandated language. Assimilation of this kind may extend to
‘namification’: standardizing and regulating ‘private’ naming, limiting the stock
and extension of legitimate surnames and first names of persons, and of public
signage, street-names and toponyms. Saddam Hussein’s Arabization of Kurds
and Turkomen included Arabic namification. Assimilation may also involve
historical standardization: the reconstruction of a common past based more on
amnesia than a representative sample of historical testimonies. It usually
involves the illegalization and repression of nationalist, ethnic and communal
parties and discourses that are opposed to the dominant nation’s assimilationist
program. In this manner Turkish governments have outlawed both overtly
Kurdish and Islamist parties and declared Kurds to be ‘mountain Turks’6.
Coercive assimilation may involve deliberate residential and employment
policies to disperse immigrants and indigenous minorities, weakening their
capacities for biological and cultural reproduction to encourage both
acculturation and fusion. Within formal democracies it will invariably involve
territorial governance and systems of decentralization designed to prevent
national, ethnic, religious and cultural minorities from having the possibility of
forming local political majorities.
Voluntary assimilation, by contrast, works though both market and public policy
incentives. Citizens and immigrants may converge on learning languages that
are useful for communication, their education and the development of their
9
wider human capital. This fact accounts for the success of distinct lingua franca
around the world, including languages like Swahili and former colonial
languages, such as English, Spanish, French, Portuguese and Russian. Citizens
and immigrants may change their religious identities or beliefs in the interests of
status enhancement. They may acculturate to political practices and mores when
they are welcomed when they do so. Shared citizenship and common public
deliberation may lead to a shared identity and the loss of historic cultural ties.
Persons may inter-marry across traditional endogamous barriers between races,
ethnic communities and religions. Voluntary assimilation may be praised as it
has been in immigrant states, such as the USA,7 Canada8, Australia9 and
Argentina.
The declared goals and the motives of state officials who have encouraged
assimilation - or promoted it through coercion - have often been inclusionary and
developmental in the broad senses of these terms. They have portrayed
assimilation as “nation-building”, promoting equality among citizens, and
fraternity (not in an exclusively male sense, but as ‘solidarity’). Through a
common individualism a common, deliberate and energised nation may
develop10. In 1789 in emancipating the Jews it was said before the French
National Assembly that, <<Aux Juifs comme nation nous ne donnons rien; aux
Juifs comme individous nous donnons tout>> - ‘For the Jews as a nation we
grant them nothing; for the Jews as individuals we grant them everything’11.
10
Assimilationists have also promoted nation-building to achieve state
consolidation and military security,12 but usually they have argued that
assimilation promotes civilizational uplift; as John Stuart Mill famously (or
infamously) suggested of “peripheral” Europeans: “Nobody can suppose that it
is not more beneficial to a Breton, or a Basque of French Navarre, to be brought
into the current of the ideas and feelings of a highly civilized and cultivated
people — to be a member of the French nationality, admitted on equal terms to
all the privileges of French citizenship, sharing the advantages of French
protection, and the dignity and prestige of French power — than to sulk on his
own rocks, the half-savage relic of past times, revolving in his own little mental
orbit, without participation or interest in the general movement of the world. The
same remark applies to the Welshman or the Scottish Highlander as members of
the British nation”13. The defence of assimilation as broadening one’s
developmental horizons has gone beyond the claim that it is better not to sit on
rocks. Assimilationists argue that their policies ‘de-parochialize’ people; they
encourage cosmopolitanism; they help break the suffocating loyalties of tribes,
clans, of kin, of what Ernest Gellner called the “tyranny of cousins”14; and, not
least, they claim that assimilation (into a high modern culture) frees children and
women from the stifling chains of patriarchy.
Coercive assimilation, on most criteria, is, however, directly contrary to one of
the international bills of rights. Article 27 of the International Covenant on Civil
11
and Political Rights declares that “In those States in which ethnic, religious or
linguistic minorities exist, persons belonging to such persons shall not be denied
the right, in community with other members of their group, to enjoy their own
culture, to profess and practice their own religion, or to use their own language”.
This clause does not directly protect national minorities, as opposed to the
specified minorities. By their actions coercive assimilationists usually provoke
national, ethnic and cultural revivalist movements, and, in their immediate
wake, secessionist movements – experiences which have occurred recently in
Pakistan, Iraq and Turkey.
That said, voluntary assimilation, as a model of inclusion, might work in certain
conditions. Evidence of success of voluntary assimilation is found in states where
there is large-scale inter-marriage - or child-bearing - across historical national,
ethnic and religious boundaries; in the existence of hybrids, fusions, mulattoes,
mestizos, and creoles; in the free development of political parties and party
platforms and electoral competition over non-national, non-ethnic and non-
religious issues; and in the consensual shunning of policies, programs and
institutions that might politically highlight salient cleavages between the
descendants of historic communities. Believing that is (or should be) its past the
French state often declared that Article 27 of the International Covenant is not
applicable in France, because France has no minorities, and to confess the
existence of minorities is to admit discrimination15.
12
Mutually agreed assimilation projects by definition have excellent prospects of
success – though that is not to say that they are irreversible. There are, however,
severe constraints on voluntary assimilationist projects. One blockage may lie
within strata of the Staatsvolk, the dominant people. Either its elites or its less
advantaged may resist assimilation if they see it as entailing undesirable cultural
transformations or losses of economic or other advantages. They may set limits
to the scale and pace of immigration for these reasons16. They may refuse to fuse
or share space with national minorities who share historic residency in their
territory. Another source of blockage may lie within the weaker communities,
the presumed beneficiaries of acculturation or fusion. They may wish to retain –
and develop - their cultures for both defensive and developmental reasons.
Where the status of the homeland people is contested thoroughgoing voluntary
assimilation is not known in modern times17. Acculturation, especially, appears
to involve a collective surrender of group identity, ethnicity, language and
religiosity. This complaint, well-grounded, is that of the proclaimed indigenous
peoples of the world and their empathizers. Lastly, there is some evidence to
suggest that at least two contemporary planet-wide “mega-trends” inhibit the
prospects of extensive voluntary assimilation. One is the fact of large-scale
migration and refugee movements which inhibit the “purification” of nation-
states within the better-off and more well-ordered states; and the second is the
greater ease with which migrant communities today may retain the culture,
13
especially the language and religion, of their homelands through easy access to
mass and customized media in print and electronic formats.
Voluntary assimilation works best through indirect incentives, and works most
effectively upon voluntary immigrants willing to trade their culture in return for
citizenship. It does not work so easily with forced immigrants (slaves and their
descendants) or with homeland and indigenous peoples. Voluntary assimilation
has many benefits in expanding the repertoire of human capacities for choice and
development but it is not without psychic costs for those who acculturate or
fuse18. But, very often assimilationists are exclusionists, policy-makers and
people who wish to imprint their cultures on others – albeit for their mutual
benefit. Such assimilationists are prime source of collective antagonisms.
“Nationalizing states”19 in both post-colonial and post-communist settings have
reaped what they have sown, i.e. protests, riots, armed secessionism, and
external interventions. “Sinhala only” policies in Sri Lanka, “Hindustani”
policies in India, “protecting Serbia” in the former Yugoslavia, “Arabization” in
Iraq and Islamization in the Sudan register similar stories and counter-stories.
Coercive homogenization generates a predictable backlash among those with the
capacities to lash back. It is of interest that anti-assimilationist and anti-
globalization discourses draw from the same repertoire of arguments.
3. Integration and Multiculturalism
14
The language and practices of “integrationists” are, at first glance, much more
generously inclusionary than those of assimilationists. Integration here is defined
as a strategy that seeks to eliminate national, ethno-national, and cultural
differences from having salience in the public domain. The object is to create a
common unifying citizenship or civic national identity. In the English language it
is rendered under formulae that have the same substantive meanings: “liberal
patriotism”, “civic nationalism”, “constitutional patriotism”, “republicanism”,
“unionism”, “privatization of culture”20, “difference-blind liberalism”,
“individualism”. Under these banners citizens are to be treated as equals,
regardless of their national, ethnic or cultural characteristics or origins.
Integration fosters the removal of ethnic identities from the political arena. It
attempts to remove barriers to access and mobility that might otherwise be
imposed on minority communities and identities. Integration does not, as a
matter of policy, seek the destruction, eradication, or erosion of these identities;
they are simply to be “privatized” and tolerated. Integration has as its nominal
focal concern the human individual. It promotes the primacy of the individual
over that of the group, freedom from group control, or, as it is often put,
“freedom from”, period. Integration rejects unreasonable, i.e. non-merit-based,
discrimination in political representation, the legal system, education and
employment. “Equal citizenship” is its motto.
15
Integration eliminates the relevance of difference in culture (of origin) among
citizens while managing cultural differences through their privatization. Each
may have their own religion provided it is not brought into the legislature or
imposed on others; each may have their own language at home – or in private
schools, as long as no attempt is made to disrupt national unity through bi- or
multi-lingualism. Integration is said to be the ideal for liberal democracies:
creating a common public culture, but leaving to citizens to choose their
communities of belonging in private domains, enabling them to maintain their
cultures without being constrained by them. Integration promotes one public
identity for citizens but allows them as many private identities as are compatible
with the liberties of others.
Public officials, aware that assimilation may provoke negative reactions and
hostility, may pursue integration. So may new governing elites sincerely intent
on overturning a genocidal and coercively homogenizing past. It may be sought
by those who have been excluded or treated as second class citizens, e.g. African
Americans, where the African denotes community of origin, or by those who are
recent arrivals (e.g. migrants, guest-workers, metics or permanent residents who
wish to become citizens). Where integration is sought both from above and
below its inclusionary character is indisputable.
In constitutional norms, laws and public policies integrationists may appear to
16
differ little from voluntary assimilationists. They want no publicly
institutionalized recognition of national, ethnic, linguistic and religious
differences; they promote territorial governance structures that avoid the
ethnicization of units; they affirm the merits of political parties and electoral
systems that encourage parties to compete on principled programs – or indeed
socio-economic and class differences – rather than on cleavages based on
nationality, language or theology; they desire voters to “pool” their votes across
historic communities of origin; they argue for a common public education system
and a common curriculum at all levels of development – the school and the
university will overcome the parochial and discriminatory impulses of the
family, the clan, the tribe, the church, synagogue, temple and mosque; they
promote public and private sector housing policies that “mix” people and avoid
the creation of ghettoes; they encourage an ethic of citizenship, insisting
citizenship rights should not be given to immigrants before at least partial
mastery of their new public culture is evident.
But, integrationists maintain, there is a major difference between themselves and
assimilationists. They are liberals. They are neither coercive homogenizers, nor
homogenizers (except in one domain), and so they see themselves as more
inclusionary. Critics argue that integration is, however, not as neutral as its
exponents pretend. Its inclusions have price-tags. The integrationist state has a
face; the state broadcasts; the state dresses; the state has styles; the state defines
17
what is public and what is private. The state just cannot be as difference-blind
patriotic, constitutionally neutral and civic as some advocate, or pretend. Its
business is conducted in a language, the language of the Staatsvolk. Its neutrality
will benefit one community’s religion: after all, what holidays (holy-days) are
public, and why? Its dominant community will have its culture fully expressed
in both the public and private domains; others will not be so fortunate. That
which is “civic” inevitably has cultural and ethnic content, even if civic
membership is in principle open to all-comers21. Integration, may be a fair
contract for voluntary migrants willing to pay the price of learning a new public
culture, but for others it may be a provocative imposition: a partisan desire to
disorganize historic nationalities or indigenous peoples on their homelands22.
Lastly, critics of integration deny that it can lead to a permanent or stable
settlement. They maintain that integration is an unstable equilibrium, either a
waiting room for assimilation, or for multi-culturalism. If cultures are not
sustained by governmental resources they may wither in the private domain, an
argument that seems especially true of languages – but less so of religions23. If
integrationists cannot ban the intrusion of arguments, interest-groups and parties
that promote agendas different from that of the approved public culture then
democratic practices will inevitably take on a more multicultural character.
Multi-culturalism is one of the most contested and discussed terms of our times
in both political philosophy and comparative politics24. It has been used to
18
describe the demography of states, philosophical relativism, the beliefs of
literature and anthropology professors, fashions, architectures and cuisines. Here
I define it to describe official state constitutions, legal norms and public policies
that institutionalize more than one culture (be it national, ethnic, linguistic,
religious) in the public domain, as well as permitting more than one culture to
flourish in the private domain. Self-styled multi-culturalists generally portray
themselves as more pluralist, authentically choice-enhancing, and more
politically inclusionary than integrationists or assimilationists. Sometimes this is
true; sometimes it is not. For example, Switzerland, for many the model of a
successful multi-cultural democracy – an argument with much merit, historically
excluded women from the comprehensive adult suffrage for longer than all of its
European neighbors, and presently excludes one fifth of its residents, guest-
workers, from having full citizenship rights.
We may differentiate multi-cultural states in four ways that sometimes overlap.
One, by the nature of the cultures that are politically included in the multi-
cultural pact. Are they national, ethnic, linguistic, religious, or some
permutation thereof? Is the state bi- or multi-national, bi- or multi-ethnic, bi- or
multi-lingual, bi- or multi-religious? And, are some (homeland and immigrant)
cultures not included fully in the public domain while others are? Two, by the
categories of culture included. This review cannot consider all the identities and
cultures covered under some discussions of multi-culturalism, such as gendered
19
and sexual orientations, or class and caste divisions; here the focus is on
encompassing cultures, sometimes called “societal cultures”, that could in
principle be deployed for nation-building. Three, by their degree of
institutionalization. Are the relevant cultures merely symbolically recognized on
an equal footing? Are they present in the executive, the legislature, the legal
system, the security forces, public bureaucracies, educational institutions, and
personal law? And four, by their territorial organization: Are the relevant
cultures granted their own territorial domains in which their members have self-
government? Are cultural rights distributed equally or asymmetrically across the
state? Rather than explore these four variations in depth, for simplicity’s sake it
is easier to distinguish two dominant forms of political multi-culturalism within
states25 which organize with respect to encompassing cultures, are intended to be
inclusive, and have distinct styles of institutionalization and territorialization, i.e.
arrangements inspired by consociational and federalist thinking respectively.
These are considered in order.
4. Consociational arrangements
Consociation, from “con, with”, and “societas, community”, is a multi-society
political system with both inclusionary shared rule and autonomous self-rule.
Shared rule is expressed through:
(a) cross-community executive power-sharing, and through the application of
(b) proportionality rules of representation and distribution of benefits to
20
communities in political institutions and the public sector.
Self-rule is expressed through
(c) allowing the communities to govern themselves at least with respect to
key components of their culture – e.g. linguistic, religious or educational
autonomy.
Consociations may have a fourth trait connected to both shared and self-rule:
(d) they may empower community leaders with veto rights over
constitutional or key policy changes.
Consociations may be found in sovereign states, in regions of states; and
elements of consociational arrangements are identical to those that characterize
many confederations of meber-states.
Consociational thinking has a long pedigree. In the western world its lineages
may be traced to the sixteenth century Protestant philosopher Johannes
Althusius (1557-1638), the early twentieth century Austro-Marxists, Karl Renner
and Otto Bauer, and, more recently, the Nobel Laureate, Sir Arthur Lewis. It is
associated in our times with Arend Lijphart, a distinguished past president of the
American Political Science Association, 26 who has generated a large research
program on the subject. Consociational thinking is not restricted to the academy.
Politicians have refined, innovated and reinvented consociational institutions
and practices in Belgium, Canada, the Netherlands, Switzerland, Northern
Ireland, the Lebanon, and Macedonia. On occasions, in the guise of “power-
21
sharing”27, consociation is the prescribed method of conflict regulation of the
international community. This has been evident in the internationally
supported, implemented, and maintained power-sharing agreements in
Afghanistan, Macedonia, Bosnia-Hercegovina, and Northern Ireland, as well as
in prospective power-sharing agreements in Cyprus and Sri Lanka. It may
become evident in post-occupation Iraq. There is, however, no consensus over
consociational theory. Two axes of disagreement are prominent:
consociationalists and their critics differ over the normative merits of
consociation; and over how consociations are established, maintained, or
breakdown. Here we focus on the first axis of disagreement.
A. Arguments Made Against Consociation. Conservatives, liberals, socialists and
feminists attack consociational prescription and explanation. Conservatives
detect a hint of utopianism, and are right to detect “rationalism”, the belief that it
is sometimes possible to engage our reason in benign political engineering.
Conservatives tend to condemn consociational ideas as futile: they say such ideas
will have no (or no long-run) impact on deeply-rooted, zero-sum identity based
conflicts. This is the archetypal conservative riposte; consociations make no
difference; or, they don’t work, ergo they are not a remedy. A more sophisticated
variation of this thesis is that consociations are only likely to work well where
they are not needed, or are redundant (i.e. in moderately rather than deeply
divided societies). Donald Horowitz maintains that consociations “are more
22
likely the product of resolved struggles or of relatively moderate cleavages”, and
that they are “inapt to mitigate conflict in severely divided societies.”28
Critics of consociational ideas are especially prominent among liberals, socialists
and feminists who pride themselves on their universalism, developmentalism,
and democratic dispositions. They often argue that consociations are perverse,
achieving the opposite of their ostensible purposes. Consociation, such critics
reason, reinforces the presumed sources of conflict; it reinforces divisive
traditions; it freezes and privileges (undesirable) collective identities at the
expense of emancipated or more “progressive” identities; it does not resolve
conflict. At best, it organizes and regulates a stalemate around the relevant
collective identities; it encourages a politics of immobilism and gridlock. Paul
Brass is typical when he argues that the elites whose prudence is hailed by
consociationalists are the very ones with vested interests in maintaining
collective antagonisms. Consociation, he claims, reinforces their respective
dominance within their own communities. Consociation’s proponents are said
to operate with the “mistaken assumption that cultural differences among ethnic
groups are “objective” factors”, and to exaggerate the problems associated with
strong collective identities, and to assume that “ethnic divisions are more
inflammatory than other types.”29 Liberal, socialist and feminist critics suggest
consociation is informed by primordial pessimism. They see consociationalists
as conservatives, who take people as they are (or have been made to be) and not
23
as they might be (or long to be). Political integration, the creation of a common
citizenship and public sphere, and the non-recognition of cultural differences in
the public domain, from this perspective, is much preferred. As Brass puts it, it is
best to “keep some possibility for change, internal division [of communities], and
secularization open, for the sake of the ultimate integration of the people in a
common political order and to preserve individual rights and the future
prospects of individual autonomy”30.
Another standard objection to consociation is that it jeopardizes important values,
principles and institutions on the altar of reinforced national, ethnic, linguistic
and religious identities. Encouraging proportional representation, it is said, will
lead to the likely irreversible formation of ethnic, communal or sectarian parties,
thereby breaking with the possibilities afforded by a politics of programs and
interests, as opposed to a more intractable politics based on identities. Quotas or
preferential policies will lead to the weakening of the merit principle – thereby
creating new injustices as well as inefficiencies in resource allocation31.
Recognition of difference in the public domain will progressively lead to respect
for (unjustified) inequalities, the unequal treatment of similarly situated
individuals and groups. Brass speaks for these critics when he asserts that
“Consociational democracy inevitably violates the rights of some groups and the
rights of some individuals”32. Brass also claims consociational politics is
undemocratic; it excludes opposition; it is a loser-takes-all system; “a fully-
24
developed consociational system is inherently undemocratic”; it is elitist and
postpones rather than facilitates the “democratization of multi-ethnic societies”.
Courtney Jung and Ian Shapiro concur: consociation permits “the same
combinations of elites to entrench themselves at the peaks of spoils and
patronage hierarchies more or less continuously”; it emphasizes “participation
and representation to the virtual exclusion of opposition”; the “democratic
benefits that can accrue from “tossing the rascals out” are unavailable”;
consociational systems “do not give powerful parliamentary players incentives
to keep government honest by shining light in dark corners”. “Mutual vetoes can
be expected to lead to mutual logrolling, rather than to political confrontations
among elites, and to promote insider clubism”. The price of consociation it seems
is “abandoning a viable opposition politics”33. The same authors suggest that
consociational systems – as with negotiated agreements or “transplacements” –
do not meet Samuel Huntington’s definitional test of a democracy, two peaceful
turnovers of power following elections34. The charge that consociation is
undemocratic is perhaps the strongest normative objection that is raised35.
Arguments in this vein celebrate the merits of oppositional politics.
Consociation’s opposition to adversarial democratic politics is just wrong-
headed maintains Brass. Adversarial politics, he insists, “have in fact worked to
an extent in non-homogeneous societies such as Great Britain … and in the US”36.
Adversarial politics in Canada, India and Sri Lanka, he maintains, are no worse
than the allegedly consociational experiences of Malaysia, Lebanon, and Cyprus.
25
A last argument sometimes deployed against consociations is the outright denial
of their existence. Sometimes this argument rests on the alleged incoherence of
consociational ideas37. A Marxist variation on the illusory quality of consociation
suggests it is a mirage with consequences: it divides and disorganizes the
working class around false identities38.
This summation of anti-consociational arguments is not a caricature, and not
based on selecting the worst rhetorical excesses39. While one must concede that
critics of consociation have pointed to difficulties in Lijphart’s formulations and
expositions (to which he has made measured and reasoned responses), plainly it
cannot be true that consociation is simultaneously perverse, i.e. reinforces and re-
entrenches ethnic antagonisms, and at the same time jeopardizes all key liberal,
democratic and international values, and, all the while is futile, i.e. makes no
difference. The futility thesis is evidently the weakest of the criticisms of
consociation, and it is fair to say that the weight of the critics’ case rest on
perversity and jeopardy arguments.
B. Political Arguments For Consociation. Consociationalists understand themselves
as realists, counselors of necessary triage, and as inclusive democrats. They are
just as concerned about justice and freedom as their critics. They submit that
consociational settlements are “naturally” recurrent phenomena --- generated
26
through negotiations by politicians after the failure of homogenizing or
exclusionary politics. They argue that certain collective identities, especially
those based on nationality, ethnicity, language and religion, are generally fairly
durable once formed, especially when they have been politicized. To say they
are durable, or are likely to be durable, is not to say that they are either
primordial or immutable, or necessarily throat-cutting. But durable identities, as
opposed to shallow and malleable interests and identities, can be, and often are
mobilized in a politics of antagonism, perhaps especially during the
democratization of political systems. Politicians, parties, and communities
themselves interpret their histories and futures through powerful narratives,
myths and symbols as well as through realistic rather than merely prejudiced
appraisals of past group antagonisms. These narratives, myths and symbols may
have significant resonance and truth-content – without these traits politicians,
ceteris paribus, might be less successful in their manipulative endeavors.
Consociationalists demur at the suggestion that they are utopian: it is
integrationists who are too facile about the capacities of political regimes to
dissolve, transform or transcend inherited collective identities. Consociationalists
operate with a hermeneutics of suspicion. They question the cosmopolitan or
emancipatory protestations of many anti-consociationalists. These protestations,
they think, too often cloak a partisan endorsement of one community’s identity
and interests (into which others are to be “encouraged” to integrate or
assimilate).
27
Consociationalists argue from a standpoint of moral and political necessity: they
do not embrace pluralism for its own sake, or because they want a romantic
Herderian celebration of a thousand different flowers (or weeds). They maintain
that a hard-confrontation with reality in certain conjunctures forces certain
options on decision-makers in (potentially) deeply divided territories; and in
some tough cases, their claim is that the effective choice is between
consociational arrangements and worse alternatives: sustained armed conflict,
genocide, ethnic expulsion, imposed partition, or imposed control. Their view is
that it is best not to have to build democracy after filling graveyards. Negotiated
consociation is better than winner-takes-all, especially where taking all implies
defining the state, and killing, expelling or assimilating the losers.
Consociationalists maintain that democratic versions of socialism and liberalism
may sometime be feasible only within consociational buildings (either as
temporary constructions or as more durable dwellings). They invoke necessity
and realism to challenge the confidence of liberals in majoritarian and
adversarial democracy. J.S. Mill famously warned of the dangers of a tyranny of
the majority, but in his most illustrious texts he failed to emphasize that a
national, or ethnic exclusionary tyranny was feasible within democratic
institutions40. Consociationalists are mostly liberals, but they are cautiously
skeptical about the current celebration of civil society as the or even a vehicle of
28
transformation, peace-making, peace-building, and inclusion. In divided
territories, they observe, there is more than one society and their relations may be
far from civil. A well ordered system of governance is necessary to enable
effective civil societies to flourish. Those who embrace a politics of deliberative
democracy as the prescription for conflict-resolution need reminding that
deliberation takes place in languages, dialects, accents and ethnically toned
voices, and that it is not often possible to create “ideal speech situations”.
Consociationalists respond to socialists in two ways: one, by showing that
consociational ideas have been present in the best of the socialist tradition41; and
two, by observing just how regularly and pervasively working class and popular
unity has historically been rendered hopeless by national, ethnic, religious and
communal divisions that might have been amenable to consociational treatment.
Within consociational arrangements trust may develop that may enable wider
working class or popular unity behind the welfare state or other forms of
progressive distributive politics.
Consociationalists are friends of democracy as well as critics of its palpably
inappropriate versions. Consociationalists want majorities rather than the
majority, or the plurality, to control or influence government. Lijphart modestly
credits the contemporary invention (though not the naming) of consociational
democracy to Sir Arthur Lewis, the Nobel-prize winning economist, who argued
in his Politics of West Africa (1965) that the post-colonial multi-ethnic states of
29
West Africa suffered from the inheritance of British and French majoritarian or
winner-takes-all democratic systems. He reasoned that what they required were
wide and inclusive coalition governments, PR electoral systems, and federations
that would give ethnic communities territorial autonomy. Lewis’s argument was
made by deduction: he did not draw on an extensive empirical comparative
politics of democratic types. But his arguments withstand empirical evaluation.
Majoritarian democracy, especially based on a single-party government rooted
in one community, is, consociationalists say, very likely to provoke serious
communal conflict in territories with two or more significantly sized
communities with durable identities differentiated by nationality, ethnicity,
language and religion. Political bargaining and adjustment among leaders
should be designed to achieve widespread consensus – to prevent the possibility
that democracy will degenerate into a war of communities. Realists should
therefore, in their view, endorse a politics of accommodation, of leaving each
group to their own affairs where that is possible and widely sought – “good
fences make good neighbors”. Consociation should be designed to protect basic
natural rights, of individuals and communities, the most important right being
the right to exist: the international law of Grotius applied within states.
Consociationalists argue positively, not just by pointing to the horrors of the
alternatives. Consociation provides autonomy, and enables sensible shared inter-
community co-operation, toleration and co-existence. Consociationalists claim
30
they have a better and more inclusive model of democracy for certain pluralist
states. Much more of the people than a plurality or a majority may influence or
control the executive. Much more than a majority gets effective “voice”.
Consociation does not eliminate democratic opposition within communities, but
enables such divisions and oppositions as exist to flourish in conditions of
security. Nothing precludes intra-bloc democratic competition, and turnover of
political elites, and shifts of support between parties; and in a liberal consociation
nothing necessarily blocks the dissolution of historic identities if that is what
voters want. Consociationalists do not say that achieving accountability over
political leaders and parties is not a problem. They merely claim that there is no
insuperable problem within consociational arrangements. Indeed it is a fallacy to
suppose that consociation mandates that all governments be wholly
encompassing grand coalitions. Since this fallacy is so widespread it needs to be
carefully rebutted.
C. Consociation, Democratic Inclusion and Opposition. There is variety within
consociations, both in their degree of inclusiveness and of democracy. We may
distinguish between complete, concurrent and weak democratic consociational
executives.
A democratic consociational executive is complete when the political leaders of
all significant communities of an ethnically differentiated territory are included.
31
Imagine there are two ethnic groups, NA and NB, and that all their voters split
their votes between two political parties respectively, giving rise to a four party
political system, A1, A2, supported by community NA, and B1, B2, supported by
community NB. In a complete consociation these four parties will all be
represented in the executive. This hypothetical scenario corresponds to the idea
of “grand coalition”. But not all consociational executives need be “grand”,
implying “total” and “all-encompassing”. What matters is meaningful, cross-
community, joint decision-making within the executive. There may be cases
where political parties choose to exclude their leaders from office by refusing to
sit in a coalition government – and insist on being in opposition by choice.
Imagine that parties A2 and B2 in the above example decide to exclude
themselves. In this case each community has representation in the executive,
from parties A1 and B1, but not all of each communities’ leaders are present.
Likewise, a cross-community coalition may exclude some hard-line ethnic parties
from the executive. This need not be either anti-consociational or anti-
democratic. We may say that a democratic consociational executive is concurrent
when each significant ethnic community has representation in the executive and
that executive has at least majority support from each such significant
community. To clarify these distinctions consider the Nagel spectrum42 that
ranges from rule by one (dictatorship), to rule by all (unanimity or perfect
consensus), i.e. from 1/N to N, where N is the number of voters (See Figure 1.1).
In a homogeneous society points on this spectrum between dictatorship and
32
unanimity can be described on the basis of how many (actual or potential) voters
influence (or control) the government. Where democracy masks rule by a small
number who control the agenda and have their preferences translated into public
policy against the preferences of the overwhelming majority then this is elite
rule. Where a faction of a party controls a cabinet which in turn controls a
legislative party which can determine law and public policy against the
preferences of a significant majority of voters then this is factional rule. Where a
plurality of voters, under “winner takes all” electoral systems, elects a
government that carries out their wishes, we have what Nagel deems
“pluralitarian” democracy. Where over a majority (N/2 + 1) of voters influence
governmental policy then we have majoritarian democracy; where over two
thirds of voters support government policy we have “supermajoritarian”
democracy; and, lastly, where nearly all voters support government policy we
have consensus democracy.
33
QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture.
Figure 1.1. The Nagel Spectrum: The Scope of Popular Support for an Executive
We may now describe both complete and concurrent democratic consociational
executives, and two further categories of weak and authoritarian consociational
executives. In an executive that is complete each community of voters, NA and
NB, and so on, has all its parties in the executive, leading to consensus across all
peoples in the consociation (here there is for the time being no formal prospect of
organized opposition). In the example we have used above the executive would
comprise parties A1A2B1B2. By contrast, a concurrent consociational executive is
34
one in which each significant community, NA and NB, and so on, has over half of
its voters supporting parties in the government and acting on their behalf, but
less than the level of representation that would exist with consensus support in
each community. In the example we have been using the executive would
comprise A1B1, and each party would have majority support within their
respective communities, NA and NB. A democratic executive may be described as
weakly consociational when each significant community, NA and NB, and so on,
has competitively elected political leaders in the executive, but only to the extent
that in at least one community the relevant leadership has just plurality support
among the relevant voters. An executive remains weakly consociational, by this
typification, if one or more community merely gives its plurality assent while
other communities give majority or higher levels of support to the government.
An illustration of a weakly consociational executive from our stylized example
would be a government comprised of leaders of parties A1and B1, each of which
is the largest (but not majority party) in its community, and each of which is
opposed by a range of smaller parties within each community, e.g. A2 A3… An,
and B2, B3, … Ban. The Nagel spectrum can lastly be employed to describe non-
democratic consociations: where unrepresentative factions or elites
(authoritarians) from each community share executive power with one another,
but are not effectively democratically accountable to (their own or any) ethnic
communities. Post-Tito and pre-disintegration Yugoslavia may reasonably be
considered a factional-elite consociation, as may Czechoslovakia under
35
communist rule. Where executives have democratic support within one
community but co-opt an unrepresentative faction or elite from another
community we do not have an authentic democratic consociational system,
because the “joint-ness” across communities is fraudulent: we have a system of
control, albeit with a pseudo-consociational veneer.
This clarification, distinguishing complete, concurrent, weak and authoritarian
consociational executives, is operational. Table 1.1. illustrates classifications on
the basis of our stylized example43. This excursus should resolve the recurrent
and widespread misperception that a democratic consociation requires a
complete, total or all-encompassing grand coalition in the executive (or in other
institutions), and thereby precludes opposition, dissent and other features of a
well-ordered democracy. What a democratic consociation must have is
meaningful cross-community executive power-sharing in which each significant
community is represented in the government with at least plurality levels of
support within its community44. This clarification refutes the insistence that all
consociational practices are inherently undemocratic because they preclude
opposition.
This distinction between complete, concurrent, and weak consociational
executives need not be applied just to executives. It is equally fruitful to apply
these distinctions to legislative procedures, and especially constitutional
36
amendment procedures – not least because these will normally have a bearing
upon the capacities of executives.
QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture.
Figure 1.2. Nagel spectra applied to two ethnic communities, with each spectrum describing
support for the executive within each community.
Community A Community B Type of consociation

Unanimous Unanimous Complete
Supermajoritarian Supermajoritarian Concurrent
Supermajoritarian Majoritarian Concurrent
Supermajoritarian Pluralitarian Weak
Majoritarian Majoritarian Concurrent
Majoritarian Pluralitarian Weak
Majoritarian Factional Co-option verging on control
Pluralitarian Pluralitarian Pluralitarian
Pluralitarian Factional Co-option verging on control
37
Factional Factional Factional Authoritarian
Factional Elite Factional authoritarian and control
Elite Elite Oligarchic
Dictatorship - -
Table 1.1. Variations in Public Support Within Two Communities for Power-Sharing Executives
Having established that consociational executives need not be all-inclusive grand
coalitions we may turn more briefly to other controversies in considering the
democratic character of consociational executives. Consociations may exist which
do not include all communities in government. This situation arises trivially in
any large and complex state or region in which there are numerous small ethnic
minorities and categories of persons (especially as a result of recent immigration)
that are not sufficiently demographically, electorally or politically significant to
be organized into any consociational settlement. Thus the non-representation of
Polish immigrants to the Lebanon in its political arrangements, or of Indian
immigrants in Northern Ireland’s arrangements, does not disprove the existence
of a consociation.
But consociational practices may prevail without the participation of one or more
demographically, electorally and politically significant ethnic community. For
example, a dominant coalition may practice consociation among the partners but
deliberately exclude another community, i.e. consociation for the dominant and
control over the dominated. One extension of Lijphart’s discussions is to treat
Israel when governed under Labour (predominantly supported by Ashkenazi
38
Jews) and Likud (predominantly supported by Mizrahi Jews) coalitions as an
illustration of this scenario: consociation among ethnically differentiated Jews,
but joint control by Israeli Jews over Palestinians in Israel (and in the Palestinian
Authority). It is also possible that a whole community, or a significant majority of
a community, may refuse to participate in consociational arrangements, even
though offered places. This is voluntary self-exclusion. Another way in which a
community, or a party from a community, may be excluded from representation
in an executive is through threshold effects that do not cast any immediate
doubts on the democratic rules of the game or on the motivations of the
constitutional designers. Any electoral system to choose a congress, assembly or
parliament, automatically has some logical or formal threshold that candidates or
parties have to achieve before winning representation45. Likewise consociational
executives may have formal rules that produce thresholds of electoral support
and legislative representation that parties must achieve before winning control
over executive portfolios. In South Africa’s transitional consociational
arrangements agreed in negotiations between the National Party and the African
National Congress in 1992-3 political parties had to obtain 5 % of the vote before
they could be guaranteed places in the cabinet, and 20 % of the vote, if they
wanted to be guaranteed one of the two executive vice-presidential posts. In
Northern Ireland, after its 1998 settlement, ten cabinet positions were available to
political parties in proportion to their voting shares. Allocation took place
according to a specific algorithm, the d’Hondt formula, which effectively
39
excludes parties that fail to achieve significant levels of electoral and legislative
support. Four parties, with 22, 21, 20 and 18 per cent of the first preference vote
respectively, took all of the ten cabinet positions, leaving a fifth of the electorate
without a cabinet member representing their first-preference vote46. South
African and Northern Irish illustrate how proportional allocation rules and
threshold effects may lead to the democratic exclusion of smaller communities or
the parties of smaller communities from access to the executive47.
It is very rare, even in an extremely ethnically, religiously or culturally divided
territory, for nearly all voters to vote cleanly for ethnic parties or candidates
representing “their” communities. Some members of some minorities may vote
for ethnic parties or candidates of other communities as proof of their
integrationist or assimilationist intent. Where a political system deliberately
obliges voters to vote only within their own community for their own ethnic
parties, then, and to that degree, the system should be called corporately
consociational. Separate electoral rolls for each ethnic community with a
requirement that everyone register on one and only one roll illustrates this
phenomenon. The British Raj”s organization of separate electoral rolls before
independence is a well-known example of such a device. To obtain access to the
community councils of independent Cyprus citizens had to opt to be on Greek
Cypriot or Greek Turkish rolls. By contrast, in a liberal consociation all voters are
on a common electoral register, and though they may vote for their own ethnic
40
parties, they are not required to do so. The distinction between corporate and
liberal consociational practice corresponds to Lijphart’s distinction between “pre-
determined” and “self-determined” group identity. Plainly the former
arrangement is rightly criticized from the perspective of human development
and liberal individualism.
In competitive democracies there will, of course, be voters who vote for non-
ethnic, inter-ethnic and cross-ethnic parties. Where they are a minority in each
significant community, and a minority overall, they may oppose but not
successfully challenge a consociational regime. In some cases, such voters create
a new community, a community of “others” who reject the available ethnic and
party identifications, and they may either oppose consociational arrangements in
defense of a new, transcendent, hybrid or non-ethnic identity, or, alternatively,
they may start to bargain for a proportionate stake in the system. In other cases,
voters who back non-ethnic, inter-ethnic and cross-ethnic parties are signaling
integrationist or assimilationist dispositions. If such voters become majorities in
each community then it is very likely that consociational arrangements will
voluntarily dissolve.
“Descriptive” representation of communities within otherwise formally non-
ethnic parties may lead to the creation of weakly consociational executives, i.e.
situations in which a large catch-all governing party may enjoy plurality support
41
from each descriptively represented community. Large catchall, (or ideological
or even confessional) parties that deliberately ensure that they have proportional
quotas of candidates for parliamentary or ministerial office from all significant
ethnic identities have clear intentions. They are trying both to appeal to voters as
ethnic voters, by using consociational devices within their own organizations
(e.g. proportionality and power-sharing), and to insure the party against possible
withdrawals of support on ethnic criteria. Where such parties are successful and
go on to form single party governments they are very unlikely to resemble
complete or concurrent consociational executives, but they may well enjoy
plurality support within each significant community. The federal Canadian
Liberal party has been seen as a weak or descriptively consociational governing
party because it usually alternates its party leadership between French and
English speakers, and allots informal quotas of cabinet seats to Anglophones and
Francophones, and because it sometimes enjoys plurality support among both
Anglophones and Francophones48. Similar interpretations were advanced about
the Indian National Congress party in its heyday, and are beginning to be
suggested about the African National Congress in South Africa49. Such
governing parties may be classified as consociational by four criteria:
(i) the extent to which they draw support from each major community of voters
at plurality levels or above;
(ii) the extent to which they are descriptively representative in the legislature and
the cabinet of the state or region that they govern;
42
(iii) their internal party organizational characteristics (i.e. the extent to which
their internal governance follows consociational traits of executive power-
sharing, autonomy, proportionality and veto rights), and
(iv) the extent to which they follow consociational practices to manage crises that
have national, ethnic, linguistic or religious roots50.
A last word on the democratic caliber of consociational executives. Contrary to
what some suggest, presidencies, including collective presidencies, have
ranked at least as prominently as variants of parliamentary premier-cabinet
executives in “actual functioning consociational systems”. Consider e.g.
Switzerland’s seven-person presidency (see Table 1.2). The snapshot evidence
from table 1.2. suggests an interesting riposte to critics of consociation.
Separation of power systems create formal mechanisms for accountability, and
checks and balances, of a different nature to those in parliamentary systems, and
the critics (and proponents of consociation) have an overly strong and
unexamined bias in favor of parliamentary systems. There is no reason why
separation of powers systems with collective presidencies need preclude either
opposition or accountability. What matters from the perspective of consociation
is not whether a democratic regime is parliamentary or presidential, but whether
it has cross-community power-sharing over executive functions and legislative
agenda-setting (and to what degree: complete, concurrent or weak).
43
Presidential Executives Parliamentary Executives
1. Directly elected president appoints 7. Strong premier appoints virtually
persons from other communities to other representative cabinet (Canada)
key cabinet posts, and must appoint a 8. Constrained premier co-exists with other
consociational cabinet (Lebanon). party elites in a bi-communally
2. Directly elected collective executive representative coalition government
presidency (Bosnia and Herzegovina). (Belgium).
3. Indirectly dual presidency elected on a 9. Dual premiership elected by a concurrent
joint platform (Federation of Bosnia and majority of the Assembly works with a
Herzegovina). cabinet allocated by the d’Hondt rule
4. Indirectly elected rotating single person (Northern Ireland).
presidency (segmental majority +
informal concurrent consent) and two
vice presidents, with a representative.
cabinet (Region of Trentino-Alto Adigo,
Province of Bolzano).
5. Indirectly elected collective presidency
(Switzerland).
6. Indirectly appointed dual collective
executive (European Union’s Council of
Ministers and European Commission).
Table 1.2. Empirical variations in executive designs compatible with consociational principles.
The foregoing discussion has qualified Lijphart’s views and that of his critics.
Rather than requiring a grand coalition government, a democratic consociation
necessarily has an executive in which there is significant cross-community
representation and support, though the forms of representation may range from
complete, through concurrent, to weak and virtual modes of support across the
included communities. Consociations vary:
i. in the extent to which communities are included and
ii. in the degree of opposition to the governing coalition in the executive;
iii. in the degree to which they are liberal or corporate in their popular
and assembly voting systems should distinguish democratic
consociations.
44
Consociational arrangements may co-exist with non-ethnic and inter-ethnic
parties. Lastly, consociational executives are as likely to be presidential as
parliamentary – and consociational advocates need have no necessary bias
against collective as opposed to single-person presidencies.
D. The Scale and Scope of Consociation. Consociational arrangements need not be
comprehensive. They may be confined to distinct constitutional and policy
sectors (in the domain of the politics of identity, recognition and constitutional
change); they may be applied piecemeal where they are deemed necessary; they
may apply to regions rather than entire states. They need not be mechanically
applied. Nor are consociationalists peddlers of a panacea: the practices they
commend are not everywhere likely to be either feasible or desirable.
What of political systems that are partly consociational? At various junctures
Lijphart has distinguished between semi-, quasi- and fully consociational
arrangements, notably in his 1985 review of constitutional reform proposals in
South Africa. The categories of semi- and quasi- as I read Lijphart appear to
mean as follows: semi-consociations have autonomy and proportionality
provisions, but lack executive power-sharing and veto-rights51; while quasi-
consociations, by contrast, combine elements of consociation (such as segmental
participation, autonomy, and proportionality) but are weakened by dominant
elements of control (ethnic monopolization of executive power, outright
45
exclusion of a segment or some segments from participation and autonomy, let
alone proportionality, and de facto and de jure possession of veto-rights by a
dominant minority)52. Even if the latter systems are claimed to be consociational
they are not, and they are not democratic in elementary senses so the “quasi”
label is unhelpful53. What Lijphart is pointing to here may better be captured by
Donald Rothchild’s notion of “hegemonial exchange” – systems which combine
elements of consociational and control methods54. More terminological
exactitude is possible. Political systems are semi-consociational if they lack
weak, concurrent or complete executive power-sharing, but distinctly have the
other three traits (autonomy, proportionality and veto-rights). On this
categorization descriptive consociational representation in the executive is semi-
consociational. Let us call political systems extensively consociational if they
have executive power sharing, proportionality and autonomy, but lack full
panoply of legal group veto rights.
Consociation, its advocates declare, facilitates justice, both procedural and social.
Groups govern themselves in agreed domains. Distributions that follow
proportional allocations may be and may be seen as fair: to each according to
their numbers. In a democratizing world that has an underlying moral appeal.
There is also a correlation between numbers and potential power that makes
such a mode of justice likely to be stable and legitimate. Consociationalists need
not endorse the Hobbesian view that justice is each according to their threat-
46
advantage, but one of their tacit arguments is that in some cases proportional
allocations of public posts and resources will be regarded as fair distributions,
and be more robust as a result.
E. Rules of (i) Proportionality , (ii) Autonomy and (iii) Veto Rights For Inclusion
E. (i) The feasibility of the idea of proportionality is not at issue in disputes over
consociation, unlike the idea of grand coalitions, or as we shall see, the idea of
non-territorial autonomy. In fully-fledged consociations the idea is that each
significantly sized community expects to be represented in political bodies in at
least rough accordance with its demographic, or electoral share of the citizenry
or population. The representation may be either descriptive, mirroring
appropriately shares of groups, or substantive, with persons expected to act for
the interests of their groups. There may, of course, be differences between the
demographic and electoral shares of segments, and this may be a source of
political controversy and fear. Proportionality can be partially applied, just to
formal political institutions, or it can be applied to all common institutions in a
state and indeed within civil society (excluding those in which each community
has legitimate autonomy)55.
Proportionality in political institutions may be achieved by electoral systems that
operate with particular formulae. Whole families of such systems ensure that
47
legislative bodies are composed so that parties are represented in proportion to
their vote-shares56. The most commonly used are closed and open party list
systems, and hybrid or mixed systems (which generally combine winner-take-all
systems at the district level with compensatory proportional allocations of top-
up seats for parties). Table 1.3. lists some of the commonly used variants of
electoral formulae. The list is confined to those formulae that may be deployed
with party list and preferential systems of voting. All achieve proportionality57.
There cannot be any uncontroversial notion of proportionality, given that human
beings do not come in fractions, that voters are very unlikely to divide their votes
in neat easily convertible shares, and that each method for achieving
proportionality “minimizes disproportionality according to the way it defines
disproportionality”58. All proportionality systems necessarily require
mathematical rules to deal with the necessary “rounding off” or “sequencing” of
votes into seats. Each rule will have an explicit or tacit notion of what
minimizing disproportionality involves. These rules or formulae, however, may
be rank-ordered as to whether, when bias is inevitable, they favor larger parties
over smaller ones. From the most favorable for largest parties to the least favored
the rank order is: d’Hondt > STV (Droop) > LR-Droop > modified Sainte-Laguë
> LR-Hare/Sainte-Laguë> equal proportions > Danish > Adams59. It is no
surprise that d’Hondt or STV (Droop) are the commonly used formulae since
larger parties are likely to be the key co-architects of electoral institutions,
48
including within consociations. Table 1.4. illustrates the use of Largest
Remainders-Hare, Table 1.5. the use of d’Hondt, and Table 1.6. the use of the
Droop quota. In each case it is assumed that there are 500 voters with 5 seats to
be awarded. The examples show that when different formulae are applied to the
same vote distribution that d’Hondt will benefit the largest party, whereas in this
example under both Hare and Droop the third party gains a seat.
Largest Remainders Variants on Quota Consociational

Method Formulae cases
Award a seat to each LR-Hare V/s Belgian lower
party for every quota in chamber
its total of votes; to fill LR-Droop v/(s+1)
any unfilled seats
reward the largest
remainders
Highest Averages Variants on Nth Sequence (first five Consociational
Method Formulae divisor divisors) cases
Award seats D’Hondt N 1,2,3,4,5 Netherlands &
sequentially to parties Switzerland
according to the Modified (10n- 1, 2.14, 3.57, 5, 6.43
remaining ‘average’ Sainte- 5)7
each presents for the Laguë 1,3,5,7,9
next seat to be awarded.
Each party’s vote share Sainte- 2n-1 0,1.41,2.45,3.46,4.47
is divided by the nth Laguë
divisor from a √n (n- 1,4,7,10,13
prescribed sequence Equal 1)
where n-1 is the number proportions 0,1,2,3,4
of seats it has already
won Danish 3n-2
Adams n-1
Single Transferable Vote Quota
Method
Candidates with Droop V/(s+1) Northern
quota of votes are Ireland
elected. Unfilled seats
filled by transferring
49
surplus votes from
elected candidates and
eliminated candidates.
Table 1.3. Various methods for proportionally allocating seats from votes.
Source: adapted from Gallagher (1992: 470). 60

Notes: (i) v = votes; s = seats.
(ii) The equal proportions and the Adams methods can be used to determine the number of
representatives for each district on the basis of population shares, but not to award seats to
parties (since the initial divisor of 0 would give a seat to every party with 1 vote).
The merits of using proportional representation systems to achieve proportional
outcomes are obvious. Provided district magnitudes are not too small, and
provided that there has not been significant gerrymandering of electoral
districts61, then PR methods will produce outcomes that are usually seen as
technically fair and consistent, even though each may each have distinct quirks
of their own 62. Using such systems on a common roll also has the merit of
enabling voters to decide whether they want to be represented by ethnic, trans-
ethnic or non-ethnic parties, i.e. voters enjoy self-determination.
Table 1.4. Allocation of seats by Hare Largest remainders. V = 500, Seats = 5.
Party A Party B Party C

Quota = 100 300 140 60
Quotas 3 (3) 1.4 (1) .6
Largest remainders 0 .4 .6 [1]
Total 3 1 1
Notes. The numbers in round brackets are seats awarded because of full quotas (v/s, = 100). The
numbers in square brackets are seats awarded because of largest remainders.
Table 1.5. Allocation of seats by d’Hondt. V = 500, Seats = 5.

Divisor
50
1 300 (1) 140 (3) 60
2 150 (2) 70
3 100 (4)
4 75 (5)
Total 4 1 0
Notes: The numbers in round brackets are seats awarded in sequential order of party presenting
the highest (remaining) average.
Table 1.6. Allocation of seats by Droop quota. V = 500, Seats = 5.

Quota = 83.3 300 140 60
Quotas 3.6 (3) 1.68 (1) .72
Largest remainders .6 .68 .72 [1]
Total 3 1 1
Notes: The numbers in round brackets are seats awarded because of full quotas (v/s+1) = 83.3).
The numbers in square brackets are seats awarded because of largest remainders.
In countries that use non-proportional electoral systems, such as winner-takes-all
in single member districts, it may be possible to achieve proportional
representation of communities, although it is more problematic to engineer63.
Reserved seats may be kept for certain groups, e.g. the Maoris of New Zealand
were reserved seats when New Zealand used plurality rule, at least initially
broadly in line with their share of the citizenry, though it started to veer towards
significant under-representation. The territorial concentration of communities
may also facilitate proportional representation of legislators by ethnic, racial,
religious or linguistic origin, though such results cannot be guaranteed. Thus the
Canadian Liberal party runs Anglophone and Francophone candidates in
appropriate ridings to try to ensure it will be internally representative in the
Ottawa House of Commons. After 1943 seats in the Lebanese chamber of
deputies were divided in the ratio of six Christians to five Muslims, or 54.5: 46.5,
51
which supposedly reflected the shares of the two communities in the 1932
census. Constituencies, of differing magnitude, were mandated to produce
specific numbers of Christian and Muslim deputies to achieve this effect. In fact,
the census of 1932 suggested a Christian: Muslim ratio of residents of 50: 49, and
of 52: 47 among registered citizens. So the fixed ratio of 6: 5 was not very
proportionate. Parity, or a ratio of 9:8, would have been more just. The ratio of 6:
5 was also, by definition, inflexible – unless, of course, it could be agreed to vary
the ratio in accordance with changing census returns. But in Lebanese politics it
gradually became impossible to hold a fresh census. When Muslims, presumed
to be expanding demographically64, demanded a fresh census, Christians
demanded that the Lebanese diaspora, presumed disproportionately Christian,
be included in any count. This led to a stalemate and was one of the grievances
that provoked the outbreak of the Lebanese civil war.
Systems of reserved seats, or systems that rely on informal norms, we may
conclude, are less likely to achieve proportionality smoothly than proportional
electoral formulae. They may also involve pre-empting people’s identities and
preferences, or adopting corporate (pre-determined) rather than liberal (self-
determined) principles of representation. Fixed quotas create obvious difficulties,
as in the case of the Lebanon. That said, much the least controversial quota
allocation is to give a community a guaranteed share of positions in the executive
or legislature that is broadly proportional to its demographic or electoral weight.
52
By contrast, the over-representation of significant minorities automatically
creates serious tensions, especially if the over-represented minority already has
other protections or has historically been privileged65. In one famous example the
Cypriot constitution of 1960 guaranteed 30 per cent of the seats in the Cypriot
parliament, and 3 of the 10 seats in the Council of Ministers to Turkish
Cypriots66, even though they comprised less than 20 per cent of the population.
The following provision of the constitution regarding the executive of the
Federation of Bosnia Herzegovina currently gives rise to predictable objections
from the local majority of Bosniacs: “Section IV. B. Article 5: No fewer than one-
third of Ministerial positions shall be occupied by Croats”.
Parity of representation among demographic unequals creates predictable
objections among majority communities who, correctly, argue that
proportionality is a different principle from parity. The new Belgian constitution
requires that the Belgian federal cabinet be comprised of equal numbers of
Flemish and French speakers, even though Flemish speakers are a clear
demographic and electoral majority. The recent Northern Ireland executive
(1999-2002) had equal numbers of unionist and nationalist ministers, but this was
a result of parties agreeing that there should be ten ministries and the subsequent
result of the application of the d’Hondt formula. With the same number of
ministries and a different distribution of seats won among parties there would
have been no guarantee of parity of representation.
53
Lijphart has described both the over-representation of minorities and parity of
representation of unequals, as “extensions of the proportionality rule”67, even
though he also regards them as methods of “disproportional representation”.
This is unfortunate linguistic acrobatics, but it is plain that his intentions have
been to counsel against over-representation of minorities. Disproportional
representation may be conceived of as a method of minority protection, and may
be demanded by minorities on various grounds, but such representation is
legitimately criticized as departing from the norm of proportionality (to each
according to their numbers) and therefore from consociational logic. The
conjunction of over-representation or parity among unequals with minority veto
rights (see below) creates obvious protests from majorities. Majorities may well
accept veto rights on matters or national, ethnic, cultural or religious significance
for minorities, but they do not appreciate why they additionally have to endure
under-representation or parity when they comprise the largest community.
The idea of proportionality also serves as a standard for the disbursement of
public funds by a regional or sovereign consociation. Indeed departures from
proportionality may occasion dissatisfaction, unless one community is
demonstrably poorer than another. The Constitution of Cyprus of 1960
guaranteed funding for the Turkish communal chamber at ¼1/4 of the sums to
be provided for the Greek communal chamber, even though Turkish Cypriots
54
comprised less than 1/5 of the population. Proportional allocation of public
expenditure is relatively easy to achieve technically among groups that are
sharply differentiated and relatively equal in endowments, though the politics of
such allocation cannot be presumed to be unproblematic.
Lastly, in consociations proportionality is used as a standard for the allocation of
positions throughout the state or region, especially in its public (and sometimes
in its private) sector. In 1958 in the Lebanon President Chehab proclaimed the
introduction of the principle of strict parity between Christian and Muslim
appointees to the civil service. This calmed communal relations though tensions
remained because of Maronite “preserves”68 in the security sector; later, when
Muslims were convinced of their increased numbers and educationally uplifted
they started to demand the end of the quota and the proportionality principle.
Examples of the use of proportionality rules or quotas to allocate administrative
and judicial positions abound in the literature of national and ethnic conflict
regulation: Northern Ireland, Canada, and Belgium are all regions or countries
with rigorous fair employment laws, including the use of affirmative action, and
in some cases quotas, to achieve proportionality in public life. The literature on
affirmative action and preferential policies is now vast – so large that it is
separately addressed in a different paper in this symposium.
55
E. (ii). Consociations share political power on matters agreed to be of common
i.e. cross-community concern. They do so in executives and legislatures, and, in
principle, may do so in judiciaries. In independent sovereign consociations key
policies and positions in security institutions (the armed forces, intelligence
organizations, and central policing services) and in economic institutions (such
as central banks) are, however, very likely integrated69. Equally though,
successful consociations are systems of self-rule which delegate decision-making
or grant autonomy (self-government) to communities on matters that are deemed
appropriate. Tacitly, consociations work with their own distinct principle of
“subsidiarity”, making it inappropriate to seek autonomy over what are matters
of common concern, and equally inappropriate to try to make a matter of
common concern what has been decreed to be within a community’s
autonomous rights. The core idea of autonomy from the perspective of
minorities is “rule by the minority over itself in the area of the minority’s
exclusive concern”70. But the idea applies to all communities, including
majorities that have autonomy. The idea of autonomy is easy to state in principle
but its institutional and regulatory manifestations are complex and varied71.
Autonomy is not the same as independence or sovereignty, and exists within
constitutionalized (or statutory or normative) arrangements that regulate the
level of discretion enjoyed by the authorities with autonomy. A standard
contrast in types of autonomy differentiates territorial and non-territorial forms.
56
Territorial autonomy arrangements can be organized in federations (state- or
system-wide forms of territorial government), federacies (particular federalized
arrangements between at least one governmental unit and the central
government) and devolved governmental systems (within unitary states where
the central government reserves full legal sovereignty). These are discussed at
greater length below but Table 1.7. suggests examples of how territorial
autonomy and consociational arrangements may be combined. The columns
organize political systems by the nature of their executives, legislatures and
constitutional amendment procedures, the rows organize systems by the nature
of their forms of territorial autonomy.
Complete Concurrent Weak/descriptive Non-

consociations consociations consociations consociations
Unitary Cyprus (1960- Estonia(1918- Belgium (before Sweden
3) 40) federation) France
Lebanon (1943- Netherlands (1917- Denmark
75) 67) Iceland
Lebanon (1989-)
Devolutionary Northern Ireland Great Britain
(1974) Spain
Italy
Federacies Northern South Tyrol Puerto Rico
Ireland (1998-)
Federations Switzerland Belgium India Australia
Federation of (intermittent) USA
Bosnia and Canada Germany
Herzegovina South Africa (1996-
(1994-) )
‘Federations’ that Bosnia and European
are de facto and Herzegovina Union (1991-)
mostly de jure (1995-)
confederal
57
Table 1.7: Possible combinations of territorial autonomy and consociational (executive and
legislative) arrangements at the centre.
A federation can be completely or concurrently consociational when all the
significant communities enjoy territorial autonomy and a (proportionate or
equal) share in the running of the federal government, and when the boundaries
of concentrated communities correspond with the territorial units of self-
government in the federation, and when the federation at the centre reflects the
other consociational organizational principles of executive power-sharing,
proportional representation of all segments. A federacy is consociational when
the relevant region has constitutionally entrenched territorial autonomy and is
internally organized along consociational lines. A consociational devolutionary
regime is one in which territorial autonomy is enjoyed by one or more distinct
regions at the revisable discretion of the central government, but the regional
governance is internally organized along consociational lines. A confederation is
consociational when its confederal institutions have either complete or
concurrent executives that represent the member-states, and when the
institutions additionally reflect proportionality and mutual veto practices72.
“Non-territorial” autonomy seems puzzling. Must not all powers or rights be
exercised within a territory? Perhaps the “personality” or the “group” principle
are better descriptions, notions sometimes subsumed under the notion of
functional autonomy. Whatever they are called, these “non-territorial” forms of
58
autonomy are distinctly consociational (because territorial autonomy is often
found in non-consociational systems).
Under group autonomy members of communities may exercise the relevant
rights wherever they reside or work in a state or region – irrespective of the
specific territory within which they live or work. So, for example, they may
publicly profess their religious beliefs and hold religious meetings in public no
matter where they live. Under personal autonomy each person may opt to be
recognized or to receive services in accordance with their group membership. So,
for example, a person living in Brussels, Belgium, may opt to receive information
about public services in either French or Flemish73. A modern example of
autonomy is the provision of separate broadcasting networks for different
linguistic users throughout the entirety of a state, especially if each network is
run by its own community’s political institutions or civil society. Striking
examples of corporate legal autonomy are the separate civil law and personal
status arrangements of the ethno-religious communities of the Lebanon. India’s
separate personal laws for its Muslim and Christian minorities are the world’s
most significant examples of corporate legal autonomy. Less surprising to the
integrationist eye, but no less consociational, are the fully funded and separate
educational systems of the Netherlands, Belgium and Northern Ireland.
59
The 1960 constitution of Cyprus contained very extensive provisions for both
legal and administrative pluralism (see Insert below) that illustrate autonomy in
a strongly consociational regime. It created Greek and Turkish chambers with
extensive powers of self-government over religious, cultural and educational
laws and policies. Reading Articles 86, 87 and 2 may create the impression that
these provisions were utterly corporate in nature, albeit dually corporate, and
strongly patriarchal (in that identity of origin was presumed to follow from the
father). They were, however, tempered by some liberal provisions elsewhere in
the constitution. Citizens were free to de-register and join other communities.
The minor religious communities, the Armenians, Maronites and Latins, were
guaranteed the continued enjoyment of the liberties and status they had under
British rule. Members of these groups, as individuals, were guaranteed human
rights and fundamental freedoms comparable to those set out in the European
Convention for the Protection of Human Rights and Fundamental Freedoms, and
both as individuals and as groups, they enjoyed constitutional protection against
discrimination. The constitution enabled the Armenians, the Maronites and the
Latins, as groups, to choose to belong to either the Greek-Cypriot or the Turkish-
Insert 1. Community Autonomy in the Constitution of Cyprus (1960)74.
Article 86. The Greek and the Turkish Communities respectively shall elect from amongst their own members a
Communal Chamber which shall have the competence expressly reserved for it under the provisions of this Constitution.
Article 87. 1. The Communal Chambers shall, in relation to their respective Community, have competence to exercise
within the limits of this Constitution and subject to paragraph 3 of this Article, legislative power solely with regard to the
following matters:-
(a) all religious matters; (b) all educational, cultural and teaching matters; (c) personal status; (d) the composition and
instances of courts dealing with civil disputes relating to personal status and to religious matters; (e) in matters where the
interests and institutions are of purely communal nature such as charitable and sporting foundations, bodies and
associations created for the purpose of promoting the well-being of their respective Community; (f) imposition of
personal taxes and fees on members of their respective Community in order to provide for their respective needs and for
the needs of bodies and institutions under their control as in Article 88 provided; (g) in matters where subsidiary
60
legislation in the form of regulations or bye-laws within the framework of the laws relating to municipalities will be
necessary to enable a Communal Chamber to promote the aims pursued by municipalities composed solely of members
of its respective Community; (h) in matters relating to the exercise of the authority of control of producers' and
consumers' co-operatives and credit establishments and of supervision in their functions of municipalities consisting
solely of their respective Community, vested in them by this Constitution:
Provided that-
(i) any communal law, regulation, bye-law or decision made or taken by a Communal Chamber under this sub-paragraph
(h) shall directly or indirectly be contrary to or inconsistent with any by which producers' and consumers' co-operatives
and credit establishments are governed or to which the municipalities subject,
(ii) nothing in paragraph (i) of this proviso contained shall be construed as enabling the House of Representatives to
legislate on any matter relating to the exercise of the authority vested in Communal Chamber under this sub-paragraph
(h):
(i) in such other matters as are expressly provided by this Constitution.
2. Nothing in sub-paragraph (f) of paragraph 1 of this Article contained shall be construed as in any way curtailing the
power of the House of Representatives to impose, in accordance with the provisions of this Constitution, any personal
taxes.
3. Any law or decision of a Communal Chamber made or taken in exercise of the power vested in it under paragraph 1 of
this Article shall not in any way contain anything contrary to the interests of the security of the Republic or the
constitutional order or the public safety or the public order or the public health or the public morals or which is against
the fundamental rights and liberties guaranteed by this Constitution to any person.
Article 2
For the purposes of this Constitution:
(1) the Greek Community comprises all citizens of the Republic who are of Greek origin and whose mother tongue is
Greek or who share the Greek cultural traditions or who are members of the Greek-Orthodox Church;
(2) the Turkish Community comprises all citizens of the Republic who are of Turkish origin and whose mother tongue is
Turkish or who share the Turkish cultural traditions or who are Moslems;
(3) citizens of the Republic who do not come within the provisions of paragraph (1) or (2) of this Article shall, within three
months of the date of the coming into operation of this Constitution, opt to belong to either the Greek or the Turkish
Community as individuals, but, if they belong to a religious group, shall so opt as a religious group and upon such option
they shall be deemed to be members of such Community:
Provided that any citizen of the Republic who belongs to such a religious group may choose not to abide by the option of
such group and by a written and signed declaration submitted within one month of the date of such option to the
appropriate officer of the Republic and to the Presidents of the Greek and the Turkish Communal Chambers opt to belong
to the Community other than that to which such group shall be deemed to belong:
Provided further that if an option of such religious group is not accepted on the ground that its members are below the
requisite number any member of such group may within one month of the date of the refusal of acceptance of such option
opt in the aforesaid manner as an individual to which Community he would like to belong.
For the purposes of this paragraph a " religious group " means a group of persons ordinarily resident in Cyprus
professing the same religion and either belonging to the same rite or being subject to the same jurisdiction thereof the
number of whom, on the date of the coming into operation of this Constitution, exceeds one thousand out of which at
least five hundred become on such date citizens of the Republic;
(4) a person who becomes a citizen of the Republic at any time after three months of the date of the coming into operation
of this Constitution shall exercise the option provided in paragraph (3) of this Article within three months of the date of
his so becoming a citizen;
(5) a Greek or a Turkish citizen of the Republic who comes within the provisions of paragraph (1) or (2) of this Article
may cease to belong to the Community of which he is a member and belong to the other Community upon -
(a) a written and signed declaration by such citizen to the effect that he desires such change, submitted to the appropriate
officer of the Republic and to the Presidents of the Greek and the Turkish Communal Chambers;
(b) the approval of the Communal Chamber of such other Community;
(6) any individual or any religious group deemed to belong to either the Greek or the Turkish Community under the
provisions of paragraph (3) of this Article may cease to belong to such Community and be deemed to belong to the other
Community upon -
(a) a written and signed declaration by such individual or religious group to the effect that such change is desired,
submitted to the appropriate officer of the Republic and to the Presidents of the Greek and the Turkish Communal
Chambers;
(b) the approval of the Communal Chamber of such other Community;
(7) (a) a married woman shall belong to the Community to which her husband belongs. (b) a male or female child under
the age of twenty-one who is not married shall belong to the Community to which his or her father belongs, or, if the
father is unknown and he or she has not been adopted, to the Community to which his or her mother belongs.
61
Cypriot community. In the event of option, the members of the group enjoyed
the same benefits as the other members of the relevant community. Any religious
group that opted as a group to belong to one of the two communities had the
right under the constitution to be represented in the communal chamber of the
community for which it opted. Any religious group, in common with other
bodies, had the right of recourse as a group to the supreme constitutional court.
Last, the constitution provided for members of the smaller religious groups to
enjoy no less extensive rights in respect of religious matters than they enjoyed in
law before, and matters of personal status remained under their jurisdiction.
These provisions of the Cypriot constitution demonstrate the feasibility of the
personal or corporate principles of autonomy. Principles like these, of course, are
not without their difficulties. While ethnic, religious and linguistic associational
life is unproblematic, modern states cannot dispense with common territorial
jurisdiction, either within a state, or within a region. In some matters – the usual
examples given are criminal or business law – a single code of behaviour and a
single regime of sanctions will usually be sought and rational in at least some
domains of public and private life, if only to avoid perverse incentives to
manipulate identities. The idea of people changing their ethnic, religious or
linguistic identity just to avoid criminal sanctions is not appealing. Meaningful
personal or corporate autonomy has one legal consequence. It will necessarily
generate a complex jurisprudence in which courts and other authorities will have
62
to regulate potentially conflicting territorial, corporate and personal principles.
This, of course, returns us to the domain of many of the normative arguments
about consociation, but what matters here is to establish the coherence and
feasibility of the idea of non-territorial autonomy, i.e. distinctively consociational
autonomy. Distinctively consociational autonomy provides an administratively
cogent way to provide autonomy to territorially dispersed communities and to
different communities that are extensively geographically mixed, as has been
demonstrated at various junctures in the histories of Belgium, Cyprus and
Lebanon75.
E. (iii). The last element in Lijphart’s definition of consociational institutional
traits, mutual community veto rights, is partly implied by the existence of the
other three. To the extent that there is cross-community executive power-sharing
each segment has at least weak and perhaps vigorous protection against
majoritarian executive discretion and legislative agenda-setting. To the extent
that there is consociational autonomy, territorial or corporate or personal, then
minorities who benefit from such autonomy can stop other minorities or
majorities from exercising executive or legislative dominance over them within
the relevant functions and spaces. Autonomy or delegation, to be meaningful,
entails veto-capacities for the autonomous or delegated authorities. Lastly, to the
extent that there is proportional representation of segments throughout the state
63
then the capacity of each group to block dominance by the others is enhanced –
though not guaranteed.
Nevertheless formal mutual veto rights may be sought and institutionalized to
ensure the guaranteed mutual protection of minorities and majorities. This is a
very strong form of political inclusion. Formal mutual veto rights may exist
within the constitution. Within the executive collective presidents or dual
premiers may share agenda-setting and agenda-blocking powers, and vice-
presidents may be granted veto-rights (Bosnia, Northern Ireland, Cyprus).
Within the legislature the consent of all the affected segments may be required
before constitutional change can take place – as in the treaty of Westphalia’s
provisions for co-existence among Christians. This can take the form of requiring
unanimity within the legislature, a concurrent majority within the legislature, a
weighted or super-majority that ensures concurrent or weak cross-segmental
support, as in the “cross-community consent provisions” of the 1998 agreement
in Northern Ireland. Similar restrictions to standard majority rule in the
legislature may apply to ordinary legislation if minorities have the right to
petition that such matters be deemed of national, ethnic or communal
significance. Within the judicial arrangements of a state or region courts may be
charged with protecting group autonomy in bills of rights and charters that
effectively place constitutional constraints on majorities that are equivalent to
entrenched veto rights; ombudsmen may be given similar quasi-judicial roles.
64
Veto-rights create obvious difficulties for standard majoritarian democratic
theory. Minority “tyrannies” may block change; deadlock, immobilism, or policy
stagnation can flow from the deployment of vetoes by all groups; and minorities
that over-use their veto rights may de-stabilize a consociational settlement76. That
said, these difficulties should not be exaggerated. In consociational systems
formal veto rights tend to apply in the domains of the politics of identity: in
ethnic, religious or national domains, and not in every policy sector. In effect, in
these domains groups have parity of power, rather than proportional power.
They are likely to be attached to those provisions that make the state and the
constitution multi-national, multi-ethnic, multi-lingual, multi-territorial, or
multi-religious.
Lijphart argues that mutual vetoes may be formal or informal77. If they are
formal they are found within constitutions and institutional rules. Though
Lijphart is right to point to informal elements in the political cultures of several
countries that are akin to mutual veto rights language should not be stretched.
Without overt, measurable constitutional impact such elements should not be
counted as veto rights. A political system with cross-community executive
power-sharing, autonomy for the segments, and proportionality rules is
consociational, and no doubt will be buttressed by conventions that take these
practices into account. Veto-rights which have legal parity of power for the
65
segments in certain domains should be regarded as elements of more rigid
consociations, and they should be rights as such: i.e. legally enforceable78.
This extended commentary on the elements of consociation may now be
summarized. It is important to distinguish democratic from authoritarian, and
liberal (self-determined) from corporate (pre-determined) consociations. The
latter are plainly much more compatible with liberal human developmentalism.
In democratic consociations there must be executives, which enjoy cross-
community support: either from all significant communities in the electorate
(complete), majorities within each included segment (concurrent), or from a
plurality of included significant segments (weak or descriptive consociation).
There must be recognizable forms of autonomy in cultural domains: in
educational, linguistic, religious and sometimes personal law and policy-making.
Autonomy may be territorial, corporate or personal. It is distinctively
consociational when it is corporate or personal. There must be systematic use of
proportionality formulae in executives, legislatures, and public appointments.
These three traits (cross-community support for the executive power-holders,
autonomy and proportionality) are necessary features of a fully-fledged
consociation. They may entail, but do not mandate a fourth. In more rigid
consociations there will be parity or veto rights. These are most likely to be found
in constitutional provisions.
66
F. The Naturalness of Consociations and Explaining Them. Consociationalists observe
that consociations occur without their urgings. They think of them as “natural”
and creative political responses to a politics of antagonism, as the outcomes of
negotiated deals. Politicians, Lijphart has observed, invented consociational rules
and institutions in the Netherlands in 1917, in Lebanon in 1943, in Malaysia in
1958, and British politicians did so for Northern Ireland in 1972. We can add that
they were re-invented by American diplomats to end the war in Bosnia
Herzegovina at Dayton in 1995; by Lebanese and Northern Irish politicians with
external prompting in 1989 and 1998 respectively; and by EU diplomats in
promoting the Ohrid agreement between Macedonian Slavs and Macedonian
Albanians. The UN and the EU between them recently sought to mediate a
consociational and confederal settlement in Cyprus, albeit one so heavily
weighted toward the concerns of Turkish Cypriots (and Turkey) that they lost
the possibility of support among a majority of Greek Cypriots. And it is not just
politicians who re-invent consociational ideas. Jurists, constitutional designers
and political theorists constantly do so. As do so-called ordinary people. Jurists
in the Holy Roman Empire proposed consociational ideas; and consociational
propositions were freshly minted by the Austro-Marxists, especially by Karl
Renner. Within academic political theory, without a full appreciation of the
history of their ideas, many contemporary multi-culturalists advance
consociational agendas: inclusivity (cross-community power-sharing); quotas
67
(proportionality), and group rights (autonomy) are often advanced in packages
to remedy the participatory defects in some contemporary democracies79.
Much comparative political science has been devoted to explaining the genesis of
consociation, though no consensus on the matter exists. Table 1.8. identifies a n
alphabet of 26 factors proposed by multiple scholars as independent variables
that might explain the development of consociational systems (some of which are
mutually incompatible). They may be ordered by (a) the number and the divisions
between groups (the demography and historical sociology of group relations); (b)
the domestic political regime (and the relations of its elites and citizens); and (c) the
external relations of the state (or region).
This is not the place to review all these possible explanatory variables, or their
respective salience. But three comments on international elements in the
explanations of consociations are necessary. First, consociational arrangements
are less likely to be found among great military powers, because they are
reluctant to embrace consensual decision-making and favor energetic
discretionary executive power80. Such is now the argument of those who want to
create a vigorous apparatus for the foreign and security policy-making of the
European Union. Second, great powers are willing to impose on small powers
Factors Independent Variables

The Divisions 1. Numbers a. No dominant community
b. A dual balance of power
c. Multiple and equally sized communities
68
d. Not too many communities
e. Demographic equilibrium
2. History f. Elite traditions of accommodation
g. Past violence and its costs promotes accommodative
politics
h. Non-colonial relationships among communities
i. Absence of special homeland claims
j. Over-arching unifying identities, shared rather than rival
nationalisms
3. Sociology k. Internal cohesion among communities
l. Cross-cutting cleavages that weaken central axis of
antagonism
m. Geographically heterogeneous groups
n. Distinct lines of cleavage (that increase community
security) including geographical isolation
o. Approximate socio-economic equality of the communities
The Political 4. Political culture p. Publics disposed toward accommodation and power-
Regime sharing
q. Elites that know one another (size – see below under v)
and negotiate without too much constituency pressure
r. Adequate articulation of community interests
s. Secure and autonomous elites - popular legitimacy of the
governing elites
5.Political t. Institutions facilitate complete, concurrent, weak or
institutions descriptively consociational executives
u. Systems of electoral and party law that create elite
predominance or secure and accountable elites
The external 6. External position v. Small size – Elites that know one another
environment of the state or regime w Small size – Easier governance
x. Shared external threats
y. Relatively low foreign policy load
z. External pressure and international norms
Table 1. 9. Alphabet of variables hypothesized to increase the likelihood of consociation

Variables (a) & (n) are the most favorable conditions cf. Lijphart; (t) & (z) are my additions.
consociational arrangements e.g. the USA and European powers used vigorous
coercion and inducements to promote consociational settlements in Bosnia
Herzegovina and Macedonia, and in the last century the European powers
intervened to create autonomy and rights packages for Christian minorities
within many of the provinces of the Ottoman empire – ones that they did not
always apply to their own religious minorities. Similarly, the centers of sovereign
unitary states may be willing to induce local elites to agree consociational
69
autonomy settlements in small regions without re-engineering their core states,
e.g. Great Britain and Northern Ireland, and Italy and South Tyrol. Third, the
direct and indirect effects of international norms matter in the genesis of
consociations. The received understanding of the Westphalian system is that
sovereign states were to leave one another alone in their domestic cultural
zones. Their sovereignty gave them the right coercively to assimilate or integrate
minorities within their borders (on some interpretations sovereignty even
included the right to commit domestic genocide).
This standard reading of the Westphalian system has never been without
challengers. The historic treaty protected some religious power-sharing. In the
1920s (after the collapse of the Habsburg, Ottoman and Czarist empires) new
European states recognized at Versailles signed minority rights treaties that in
principle could be regulated by the League of Nations. The treaties bound them
not to abuse their minorities and in some cases required them to develop semi-
consociational practices (notably in religious, educational and linguistic
autonomy). The story was hardly a success, and the United Nations was partly
constructed in San Francisco in a deliberate rejection of these experiments. But
the post de-colonization international law of self-determination, and the politics
of recognition of the post-communist successor states, have seen a revival of
efforts to lock new states into systems of minority protection. In turn this has
70
provided some external shield for minorities that advance consociational
demands.
Other indirect effects of international norms and interventions are also evident
on consociational prospects. International proscriptions exist against genocide
and expulsion. Norms of some significance reward states that are democratic –
and make discriminatory regimes potential pariahs (though these now often
parade themselves as allies in the global war on terror). There are also
proscriptions against coercive assimilation. There are strong dispositions in the
international system against secession and hostile to partitions. The conjunction
of all these norms leave international organizations and great powers when they
intervene in national, ethnic, and communal conflicts confined to promoting one
of three repertoires of conflict regulation:
(i) territorial autonomy and/or federacies and federations;
(ii) (ii) integration; and
(iii) (iii) consociation.
In some scenarios to prescribe integration --- Bosnia Herzegovina, Macedonia,
Northern Ireland, Cyprus, Iraq --- is to prescribe the partisan victory of one
community over another. In short, the normative prohibition, if not factual
exclusion, of certain options sometimes create leverage in favor of consociational
arrangements, perhaps especially in small political systems or weakened states.
71
Emergent Cases. There are at least three current experiments in “complex
consociations”, Northern Ireland, Bosnia Herzegovina, and Macedonia, that
illustrate these tentative suggestions. (Cyprus might have been another had
Turkey not over-negotiated and had the United Nations and the European Union
managed the local politics with greater perspicacity). All three involve executive
power-sharing, proportionality, autonomy and veto-rights; but they also involve
international efforts to resolve national self-determination disputes, mediation,
negotiation, arbitration and implementation of peace settlements, and cross-
border or confederal relationships (and sometimes institutions) for national
minorities with their kin in other states. They are somewhat less frequent than
the international promotion of standard territorial autonomy settlements but
their presence in our times is suggestive. They specifically attempt to address
national self-determination disputes between communities, i.e. where there is an
opposition between at least some secessionists and some unionists (or
federalists). The settlements institutionally recognize more than one people,
nation or society, and attempt to provide constitutional architecture within
which they can co-exist. They may involve defining the state as multi-national,
recognizing national minorities as well as majorities, referendums to ratify such
settlements in more than one jurisdiction, or mechanisms to trigger referendums.
They involve peace processes – mechanisms, confidence-building measures and
institutional and policy transformations to halt conflict and future violent
recurrences. They involve both the reform and the restructuring of security
72
systems, and measures intended to end secessionist (and anti-secessionist)
paramilitarism, as well as new human rights protections. These involve at least
one other conflict-regulating strategy or principle in their design. These three
consociations are combined with territorial autonomy. In Northern Ireland the
region has autonomy from Westminster. The Federation of Bosnia Herzegovina
has internal territorial autonomy for Bosniacs and Croats. Macedonia will have
de facto territorial autonomy for Macedonian Albanians. These consociations
have arbitration mechanisms for resolving disputes --- impartial courts,
commissions, international judges, international commissions,
intergovernmental conferences. They may have elements of integration ---
citizenship equality provisions --- and, they may have mechanisms that enable
the secession of the relevant unit of consociational governance, or, alternatively,
a procedure for enabling the central state to “down-size”. Northern Ireland
illustrates the point; by contrast, there is no such provision in the Dayton or
Ohrid Agreements. A last element of complexity is international involvement in
the making, ratification and maintenance of the relevant consociational or
autonomy settlements. This may involve neighboring states, regional powers,
great powers, regional organizations (e.g. the OSCE or the EU), or international
organizations (e.g. the UN), in organizing and monitoring cease-fires, in
providing good offices for the making of settlements, in designing
implementation arrangements, and providing default mechanisms to arbitrate
disputes. Levels of international institutionalization may vary. Domestic
73
incorporation of international human and minority rights standards does not
necessarily challenge the sovereignty of the state. Institutionalized cross-border
co-operation and the formation of bodies with executive powers in more than
one formally sovereign jurisdiction, by contrast, entail at least a pooling if not
necessarily a diminution of sovereignty. High commissioners appointed by great
powers may be indistinguishable from the prefects of protectorates (or those of
colonial overlords).
Summary. The rival moral and political evaluations of consociation are unlikely
to be immediately resolved; and are not amenable to decisive confirmation or
falsification by empirical evidence. They reflect different philosophies and
factual appraisals. Anti-consociationalists fear that consociation will bring back
racism, fundamentalism and patriarchy, whereas consociationalists fear that
integrationists will provoke avoidable wars and are biased towards the interests,
identities and ideas of dominant communities. The intensity of this debate
attests to the influence of consociational thought. The test of concepts,
taxonomies and theories in the social and legal sciences is twofold: i.e. whether
they serve worthwhile explanatory or normative purposes. The claim here is
that exponents of consociation, when their case is put carefully, can rebut the
wilder charges made against their moral and political positions. Consociations,
simple or complex, are certainly difficult to love and celebrate – even if their
makers often fully merit intellectual, moral and political admiration. They are,
74
after all, usually cold bargains, even if they may be tempered by political
imagination. Ardent secessionists and uncompromising unionists should be
counseled that their efforts may land them with these systems if they fail to win
on the battlefield. In explaining consociations significant preliminary work has
been done, but a comprehensive comparative historical analysis of consociational
settlements and their outcomes remains to be completed.
5. Federal Arrangements
Federal principles of inclusion, like those of consociation, combine both shared
rule and self-rule81. Federal political systems encompass a range of possible
political organizations that reflect this philosophy: including confederations,
federations, certain kinds of unions, federacies, associated states, leagues and
cross-border functional authorities. Federations are a very specific federal
political system, arguably first unambiguously invented in 1787 in the city of
Philadelphia82. And, like consociations federations can be built around the idea
of including multiple peoples, creating a multi-society political system83, though
it is vitally important to distinguish such multi-society or multi-national or pluri-
national federations from national or integrationist federations. The arguments
below distinguish national from multi-national federations, and evaluate their
merits for achieving inclusion. Lastly they elaborate briefly on the less well-
known notion of “federacy” as opposed to wholesale federal arrangements –
75
drawing upon my own thoughts84, and some unpublished work by David
Rezvani85.
In a genuine federation at least two governmental units, the federal and the
regional, enjoy constitutionally separate competencies – although they may have
concurrent or shared powers. Both the federal and regional governmental units
are empowered directly to deal with their citizens – which differentiates most
confederations from federations – and in a democratic federation citizens directly
elect at least some components of both the federal and regional governments.
Federations are “covenantal”, the authority of each government is derived from
the constitution, and not from another government86. In authentic federations the
federal government cannot unilaterally alter the horizontal division of powers:
constitutional change affecting the division of competencies requires consent
from both tiers of government. Therefore federations automatically require a
written, codified constitution, and normally require a federal supreme court,
charged with upholding the constitution and umpiring differences between the
governmental tiers. They also usually involve a bicameral legislature – a chamber
of the citizens as a whole, and a second chamber that represents the regions.
Federations vary extensively; first, in the extent to which they are “majoritarian”.
All federations place some constraints on the powers of federation-wide
majorities, but do so to different degrees87. The USA, Australia and Brazil for
76
example, allow equal representation to each of their regions in the Senate,
generating massive over-representation for small units such as Rhode Island or
Tasmania. A majoritarian federation concentrates political power at the federal
level, and facilitates executive and legislative dominance either by a popularly
endorsed executive president, or by a single-party prime minister and cabinet. A
federation is not majoritarian to the extent that it has inclusive executive power-
sharing arrangements in the federal tier of government; institutionalizes
proportional principles of representation and allocation of public posts and
resources; and has mechanisms, such as the separation of powers, bills of rights,
monetary institutions and courts, that are insulated from the immediate power
of a federal governing majority. On this design choice, Arabs and Kurds in
contemporary Iraq would be well-advised to avoid a strongly majoritarian
federation. Kurds have been an enduring minority in Iraq and, judging by the
historical record, would be long-run losers from the creation of a strongly
majoritarian federation – in which either an Arab or a Shià majority might
threaten their national, linguistic and cultural identities, as well as their regional
and economic interests. Sunni Iraqis too have an interest in constraining the
power of a potential federal majority that might be inimical to their religious and
other interests. Shi‘a may be the most tempted by a majoritarian political system,
but they may be less homogeneous than some of them hope and others fear –
given differences among them in religiosity, and dispositions towards Iran and
other neighboring states88. The more homogeneously Shià mobilize and act then
77
the greater the likelihood that they will generate a coalition of minorities against
them.
Federations, secondly, vary significantly in the distribution of powers within the
federal government. Some create very powerful second chambers. The US Senate
is arguably more powerful than the House of Representatives because of its
special powers over nominations to public office, and over treaty-making. Other
second chambers, such as those in Canada, India and Belgium, are very weak89.
Some have separately elected executives; some have executives chosen by the
federal first chamber; and there are both single person and collective executives.
Thirdly, federations differ in the distribution of competencies between the
federal and regional governments. In some federations the powers of the federal
government are constitutionally circumscribed and delimited; in others the
regional governments have their capacities specified and delimited. In the
German model the federal government makes broad policy and law while
administration and implementation are in the hands of Länder governments,
empowering both tiers with distinct enabling and blocking powers (the
European Union appears to be federalizing along German lines). In all
federations the constitutional division of competencies (even as interpreted by
the courts) may not be an accurate guide to the policy-making autonomy and
discretion held by the separate tiers. The superior financial and political
resources of one tier (usually the federal) may allow it to weaken the other tier’s
78
capacities – as in the USA where the federal government’s pre-eminence is now
established90.
But whatever the distribution of competencies between federal government and
and its regions, the critical choice on political inclusion is whether the
federation is to be mono-national (or integrationist) or multi-national. Mono-
national or national federations aspire to national homogeneity: to eliminate
internal national - and perhaps also, ethnic - differences from lasting political
salience. The goal of national federations is nation-building. The USA is the
paradigmatic example of a national federation. The Latin American federations
of Mexico, Argentina, Brazil and Venezuela, at various junctures in their history
have adopted this US model. Germany, Austria, Australia, Malaysia and the
United Arab Emirates are also national federations. National federalists think
one nation and one federation can be combined successfully. The earliest-
federalists in what became the Netherlands, the German-speaking Swiss lands,
the USA and the Second German Reich were stepping-stone nationalists: the
prime function of federation was “to unite people living in different political
units, who nevertheless shared a common language and culture”91. They
maintained federation was necessary to provide a united defense and external
relations – tasks that confederations and leagues were less well-equipped to
perform92.
79
American and American-educated intellectuals, political scientists and
constitutional lawyers, often propose national federations to manage
heterogeneous post-colonial and post-communist societies. Indeed, they have a
distinct animus against multi-national federations, which they regard as divisive
and likely to collapse through secession. As the USA expanded southwestwards
from its original largely homogenous citizenry of the 13 founding colonies – a
citizenry which, of course, excluded African slaves and native Americans – no
new territory received statehood unless minorities were outnumbered by White
Anglo-Saxon Protestants93. Sometimes the technique deployed was to
gerrymander state boundaries to ensure that Hispanic or Indians were
outnumbered, as in Florida. At other times statehood was delayed until the
region’s long-standing residents could be swamped. America’s nation-builders
were even cautious about immigrant groups concentrating too much in given
territorial locales, lest this give rise to ethnically based demands for self-
government. Grants of public land were denied to ethnic groups per se to
promote their dispersal: William Penn dissuaded Welsh immigrants from setting
up their own self-governing barony in Pennsylvania94. This is why the US
federation, in the words of one of its most distinguished analysts, shows “little
coincidence between ethnic groups and state boundaries.”95 It would be more
precise to say that the sole coincidence is between white majorities and state
boundaries, and that that is no coincidence. National federation, as a strategy of
growth and incorporation, aided the homogenization and assimilation of whites,
80
the famous melting pot of what Milton Gordon described as “Anglo
conformity”96. Celebration of the homogeneity of the founding people of the
federation was evident in the now sacramental The Federalist Papers. In the words
of John Jay: “Providence has been pleased to give this one connected country to
one united people – a people descended from the same ancestors, speaking the
same language, professing the same religion, attached to the same principles of
government, very similar in their manners and their customs, and who, by their
joint counsels, arms and efforts, fighting side by side throughout a long and
bloody war, have nobly established liberty and independence”97.
This quick assessment of the “historical starting point” and development of the
USA suggests the inappropriateness of American national federation as a
model for many multi-national, multi-religious and multi-lingual territories,
where there is no sufficiently homogeneous founding people, blessed by
Providence or not. In such cases to create regions in which one culture or one
people is pre-eminent, the historical practice of the USA, is a recipe for conflict,
especially in a non-immigrant state. We may take contemporary Iraq as an
example for analysis. It takes little historical knowledge to argue that no one
could plausibly advance John Jay’s arguments during the making of Iraq’s new
constitution. Iraq may be contiguously connected on maps, but it has not had a
united people, i.e. a people who think of themselves as descended from the same
ancestors, who speak the same language, or who profess the same religion –
81
Islam has, after all, divided them as much as it has united them. They neither
flow from a common stock, nor are they united by a common immigrant or
assimilationist experience. They have not “by their joint counsels, arms and
efforts” just fought a combined war of national liberation. To the contrary: only
the Kurds fought with the Allies; the Shià were reluctant to rise given their
previous abandonment by the 1991 coalition to Saddam Hussein’s mercies; and
some Sunni Ba’athists to this day are fighting the Allied occupation. It is true that
many Iraqi Arabs, be they Sunni or Shi’ite, fought side by side in Saddam
Hussein’s long and bloody war with Iran, and that that war proved that for most
of them ethnicity trumped religiosity, but some Shià did enroll with Iran, and
most were at the front through conscription rather than by choice. Notoriously
during that war Saddam organized genocidal massacres of Kurds, who were not
in any sense “his own people”98; just as he would later engage in repressive
massacres of largely Shi’ite Marsh Arabs.
It is, of course, feasible to have many regions in Ireaq in which Shià would be
the local majority. Indeed it is probably not feasible to design contiguous
regional boundaries that would not make the Shià dominant in many of them.
But, it would be extraordinarily difficult, and divisive to devise regional
boundaries to prevent Kurdish or Sunni communities from becoming regional
majorities anywhere in Iraq. In the case of the Kurds such a strategy would
require the partition of the existing regional government’s de facto jurisdiction and
82
the addition of significant non-Kurdish population and territories into each new
unit. Such design principles would return the Kurds to armed conflict with the
rest of Iraq. No nationally mobilized people in recent history has peacefully
acquiesced in the partition of its homeland. As for Sunni Arabs, one reason why
the Ba’athists remain partly embedded among them is the widespread fear
among them that Shià will create a majoritarian democracy that they will see as
a dictatorship. In short, no better plan for provocative conflict could be devised
than designing the territorial boundaries of the new Iraqi federation to prevent
either Kurds or Sunnis from having regions in which they are the
demographically and electorally dominant group. Regrettably, these elementary
considerations are overlooked by those who argue that a new Iraqi federation
should be built around the eighteen provinces of Ba’athist Iraq99. One American
political scientist has argued that the regional boundaries should be drawn to
prevent any of the three major communities, Kurds, Sunni Arabs or Shi’ite Arabs
from having local majority control100. There is nothing wrong in principle with
advocating this design, but it has no prospect of success in Iraq. To design or re-
draw regional borders along these lines would require the services of the armed
forces of the Allied occupiers or future UN forces. It is also difficult to see how
this thinking could even be regarded as feasible before a reliable new census; and
if it were known that the census would inform the drawing of such new borders
that in turn might create perverse incentives to expel exposed minorities.
83
Advocates of multi-national federations have a different conception of how to
achieve inclusion. They eschew assimilation, and embrace deeper pluralist
thought. Like consociationalists they seek “to unite people who seek the
advantages of a common political unit, but differ markedly in descent, language
and culture”101. They seek to recognize, express and institutionalize at least two
national cultures, on a durable, and often on a permanent basis. Multi-national
federations involve the maintenance of two or more nations. Their exponents
believe it is possible for the citizens of such federations to have dual or multiple
loyalties, e.g. a patriotic attachment to the federation and a nationalist
attachment to their regional homeland (see Professor Stepan’s contribution to
this symposium). They believe it is wrong to assume a priori either that multi-
national federations will lead to the abuse of the rights, interests and identities of
regional minorities, or that they will necessarily make secessionists victorious.
Multi-national federalism has long been advocated within both liberal and
Marxist traditions, and has a significant following within the Anglophone
academy102, including both those who see federations as devices to hold peoples
together as well as those who emphasize the merits of territorial autonomy for
historic national minorities. Multi-national federations are workable. Switzerland
and Canada are among the world’s oldest states – they have lasted in
recognizably similar forms since 1848 and 1867 respectively. But, while multi-
national federations have their enthusiasts, no one can deny that in the twentieth
century that they have had “a terrible track record”103. Multi-national and multi-
84
ethnic federations have broken down or have failed to remain democratic
throughout the communist and post-communist world (Yugoslavia,
Czechoslovakia and the USSR; and Ethiopia “lost” Eritrea); and they have also
broken down in much of the postcolonial world, in sub-Saharan Africa, South
Asia and the Caribbean104. In the Arabic world the United Arab Emirates is the
sole surviving postcolonial federation – and it is a national federation, and
hardly a model democracy.
The breakdowns of these federations do, however, have elements in common
which the architects of the new Iraq would be well advised to bear in mind. John
McGarry and I highlight five key elements that have facilitated the breakdown of
multi-national federations105:
1. Coercion: They were usually forced together rather than being the outcome
of voluntary agreements, e.g. the constituent republics of the Soviet
Union.
2. Authoritarianism: They were not democratic for much of their histories,
and when many such federations democratized that created the
institutions and opportunities for secessions to occur, e.g. Bangladesh’s
secession from Pakistan; e.g. Slovenia and Croatia’s secessions from
Yugoslavia.
85
3. Maltreatment of smaller nations: They failed to resolve tensions between the
largest or the historically dominant nation and smaller nations, e.g.
between Malays and Chinese.
4. Distributive conflicts: They failed to develop or maintain economic
distributive and redistributive formulae regarding economic policy,
taxation, revenue-sharing, and public expenditures, that were widely
regarded as fair, e.g. Czechoslovakia.
5. Centralizing coups, putsches or maneuvers: Breakdown was often preceded
by authoritarian attempts to centralize the federations, e.g. the conduct of
Serbian politicians in Yugoslavia.
The implications for Iraq (and other possible postcolonial and postconflict sites)
of this rapid inspection of the failure of multi-national federations are
straightforward. The conditions for a successful federation include the following.
One: The federation must be a voluntary pact, and not regarded as an American
or UN imposition. The federation must be ratified by its respective and
prospective units –e.g Kurds must have a referendum in their own unit to
endorse any freely negotiated constitution. A foundational act of co-operation is
more likely to promote future traditions of accommodation. Two: the federation
must be democratic, with the full repertoire of liberal democratic institutions,
universal adult suffrage, competitive elections, freedom for political parties and
interest groups to mobilize, a constitution with the rule of law, human rights
86
protections, and a free media. Three: constructive relations based on mutual
recognition must be built between the three largest national and religious
communities, Kurds, Sunni Arabs, Shi’ite Arabs, as well as the smaller minorities
of Turkomen, Christians and others. Four: robust and adequate agreements have
to be built over the sharing of Iraq’s natural resources. Lastly, there must be
significant constitutional checks – and preferably some international arbitration
mechanisms – that would inhibit future efforts to centralize the federation, e.g.
there needs to be significant default mechanisms to protect Kurds should a
governing coalition in the rest of Iraq in the future try to undermine Kurdistan’s
newly won constitutional status.
Inspecting the failures of twentieth century multi-national federations is not, of
course, the only way to think about these matters. The major surviving federal
multi-national democracies, notably Belgium, Canada, India and Switzerland,
have had histories, institutions and practices that may separately or jointly
explain their relative robustness:
1. Multi-national federations may well benefit from having one large group,
a Staatsvolk. All other things being equal a Staatsvolk can feel secure and
live with what it will regard as the price of multi-national federation. It
has both the practical power to resist secession, and the capacity to be
generous to discourage secessionism.
87
2. Conversely, multi-national federations that lack a Staatsvolk, if they are to
survive as democratic and durable entities, must have cross-community
power-sharing practices in the federal government, i.e. consociational
practices. These practices must minimally encompass the interests of all
the national, ethnic and ethno-religious communities with the capacity to
breakaway. Neither the presence of a Staatsvolk, condition one, nor cross-
community power-sharing practices in the federal government, condition
two, are sufficient to ensure the survival of a democratic multi-national
federation but judging by the record of the twentieth century the presence
of one of these conditions is a necessary condition of enduring
federations106.
3. Federations are more likely to be stabilized if they have non-
interventionist neighbors who do not seek to play major roles in the lives
of their cross-border co-ethnics or co-religionists.
4. An authentic multi-national federation will be democratic. Democratic
arrangements allow the representatives of national, ethnic and ethno-
religious communities to engage in dialogue and open bargaining, which
facilitates the development of political co-operation. Liberal democratic
arrangements that protect individual rights and collective organization in
civil society serve to check systematic transgressions against such
communities. Federations that protect collective identities help make the
88
respective communities feel secure – and in consequence may facilitate the
emergence of inter-ethnic and inter-religious co-operation.
5. Prosperous and fair federations are more likely to endure than those that
are not. One should not exaggerate the power of materialism in politics. It
would be wrong, for example, to insist that prosperity is a necessary
starting condition of the success of multi-national federation –
Switzerland, Canada and India did not start rich, and India is far from
being rich. But, federations that over time facilitate increasing per capita
prosperity ceteris paribus have better prospects of success.
The application of these arguments to the future of Iraq may now be briefly
sketched. First, Iraq has a potential Staatsvolk, Shià Arabs, who might be
demographically reinforced by the return of deportees, exiles, and refugees. But,
several factors tell against the materialization of this prospect. They have not
been the historically dominant people in the state; and it is unlikely that they will
be politically homogeneous – provided they get a fair stake in the new order.
They have religious and secular cleavages among them; they have intra-religious
cleavages; and they have class differences. Vigorous Shià majoritarianism would
guarantee a prolonged Sunni Arab resistance that would not just be political.
And Sunni Arabs, by virtue of their past dominance, have greater resources than
their potential rivals. Second, if there is no compelling evidence that the Shi‘a
can comprise a Staatsvolk our argument suggests that power-sharing at the
89
centre as well as autonomy within the regions will be necessary to preserve the
federation. Federalism, after all, involves both “shared rule” as well as “self-
rule”. The exclusion of national, ethnic or religious communities from
representation and power at the center is a sure recipe for conflict and
secessionism. Durably democratic multi-national federations, Canada,
Switzerland, Belgium, have had consociational or power-sharing practices in
their federal governments: cross-community executive power-sharing,
proportional representation of groups throughout the state sector (including the
military, police and judiciary) and formal or informal minority veto-rights. And,
it has been argued that India has been at its most stable when its executive has
been descriptively inclusive of that state’s diverse religions and linguistic
communities107. This evidence strongly suggests that Iraq needs an executive that
is cross-community and cross-regional in character. The nine-member collective
presidency of the Governing Council is a good portent of sensible future
compromises on the construction of a future federal executive. A five-person
collective presidency comprised of representatives from five regions – a
Kurdistan region, a Sunni dominated region, Baghdad and two Shià dominated
regions – would necessarily have a cross-regional and cross-community
character – and would not require formal “set-asides”, the bugbear of many
western constitutionalists. Given many Iraqis’ interest in avoiding too powerful a
central government, a collective presidency commends itself as the best means
to create widespread security. Measures to ensure that federal bureaucrats,
90
military, police and judges are representative of Iraq’s diversity would cement
the necessary political accommodation. Third, the external conditions for the
success of federation in Iraq are not difficult to spell out: Turkey, Syria, Iran and
Saudi Arabia will have to keep out of their neighbor’s territory, and avoid
sponsoring paramilitary organizations of any kind. The fourth and fifth
conditions of long-run success in multi-national federations, democratization and
economic prosperity cannot be assured in advance, but nothing in Iraq’s cultures
or communities’ talents need necessarily prevent them.
This extended treatment of one possible example stresses the possibility of a
democratic federation in Iraq, but if, and only if, that future Iraqi federation is
bi-national, multi-ethnic, tolerantly multi-religious, and multi-regional. Bi-national,
because there are two nationally mobilized and linguistically distinctive
collective communities, Kurds and Arabs. Multi-ethnic, because there are a
range of other ethnic communities, notably Turkomen, who will need to have
institutional recognition and protections, both at the federal and regional levels.
Multi-religious, both to manage the Shià and Sunni divide, their internal
divisions, and the non-Muslim religions, as well as those who have no religion.
This will require collective compromises on personal law, and a separation of
the state from any distinctive religion, though it need not preclude the
constitution from recognizing Islam as the major religion of the peoples of Iraq –
a policy that would avoid establishing any clerisy. Regional and proportional
91
funding of education might also resolve many possible religious sources of
conflict.
We may also treat the notion of ‘federacy’, also raised in Professor Stepan’s
contribution through the same example. Many Kurds have been
programmatically committed, for a long time, to a confederal or federal Iraq.
Kurds have a long tradition of seeking territorial autonomy and having
autonomy arrangements, territorial or cultural, betrayed by governments in
Baghdad (or London or in other capitals in the world). Kurds cannot, of course,
impose a federation on their prospective negotiating partners. They can only
negotiate with those willing to make a deal with them. Kurds, however, have
three immediate political priorities:
1. To promote a bi-national, multi-ethnic, tolerantly multi-religious and
multi-regional Iraqi federation with a significantly sized Kurdistan as one
of its units, and within that unit they should deepen and extend their own
evolving democratic institutions and, as they intend, provide cultural
rights for Turkomen that Turks have not given to Kurds in Turkey. Kurds
seek a whole Kurdistan as a region of an Iraqi federation, and power-
sharing in the federal government, and full cultural rights for Kurds living
outside Kurdistan.
2. To insist that any negotiated constitution be ratified by the people of
Kurdistan, as well as the rest of Iraq.
92
3. To insist on default mechanisms that would protect Kurdistan in the
event of breaches of any new Iraqi constitution.
But Kurds have to consider their options if the rest of Iraq chooses not to accept
any mutually agreeable model of a bi-national, multi-regional federation. One
option would be for Kurds to insist on a distinctive ‘federacy’ agreement. They
can say that they will accept the rest of Iraq choosing to be unitary, or indeed
choosing to be a centralized US-style national federation, provided that Kurds
themselves have a ‘federacy’. A federacy is a federal arrangement that is not a
part of a system-wide federation; it creates a semi-sovereign territory different in
its institutions and constitutional competencies from the rest of the state; it
creates a division of powers between the federacy and the central government
that is constitutionally entrenched, that cannot be unilaterally altered by either
side, and which has established arbitration mechanisms, domestic or
international, to deal with difficulties that might arise between the federacy and
the central government. Federacy is autonomy that is not devolution; it is not a
revocable gift from the central government; it is domestically constitutionally
entrenched so that the federacy can veto any changes in its status or powers; and,
ideally, its status and powers are internationally protected in a treaty. In short,
while Kurds have no right to impose a federation on the rest of Iraq, they have
every right to insist on federacy arrangements for Kurdistan as one means
through which they can exercise national self-determination. In this scenario
93
they would probably seek looser power-sharing arrangements in the central
government, especially in foreign relations, while seeking to protect the cultural
and human rights of Kurds outside Kurdistan.
Summary. Multi-national federations, federacies and consociations, and
permutations of all three systems, are ways of achieving inclusive but non-
homogenizing political institutions capable of managing national, ethnic,
linguistic and religious diversity. There are working examples of each. The
model of the homogenizing nation-state not only need not be followed, it is often
a recipe for protracted conflict. This is not to say that maintaining a state must be
the sine qua non of political practice: sometimes national self-determination
exercised through secession is the best remedy for historic ill-treatment and a
fresh start. But unionists and federalists determined to maintain their states
would be well advised to look at multi-national federations, federacies and
consociations as feasible and principled alternatives to assimilationist or
integrationist nation-states.
94
Appendix 1. Electoral Arrangements and Inclusion.
Electoral systems achieve greater inclusivity and assist human development
more significantly the greater the ease with which voter-registration can be
accomplished, the greater the transparency, uniformity and integrity of electoral
administration and the counting process, the greater the personal security of
voters, the greater the degree of impartial or judicial as opposed to party-political
determination of district boundaries, the greater the degree of openness and
competitiveness of media outlets, and the greater the transparency and equality
of funding of parties and candidates. These propositions apply to all electoral
arrangements.
The discussion above has identified four broad political strategies that encourage
specific models of inclusion: assimilation, integration, consociation, and
federalized models of territorial autonomy. The first two are attached to nation-
state models of homogenization; the latter two attempt to build and expand dual
identifications among the citizens of the relevant peoples. Permutations of these
strategies are, of course, possible, but are not treated here. Each of the four
strategies fits better with certain electoral designs – though this comment
requires more extensive qualification that can be provided here.
95
Assimilationist and integrationist democrats prefer state-wide political parties,
differentiated by programmatic, left-right divisions, or, relatedly, appeals based
on divisible, tradable, and material interests. For these reasons they are disposed
toward electoral systems that promote two-party or two bloc systems, and that
are likely, in parliamentary systems to generate a single-party – or at least single
bloc – government, facing a single-party (or at least single bloc) opposition. They
believe the best systems generate pressures for parties to support median or
moderate voters. They likewise favor the direct election of presidents.
Assimilationists unambiguously support winner-take all electoral systems – in
the form of plurality rule in single-member districts (the historic British and
American norm) or double-ballot systems (the historic French norm). They may
support preferential-majority systems in single-member districts (the alternative
vote), that is used to elect the lower chamber of the Australian federal
parliament. National federalists are generally integrationists with assimilationist
ambitions, and for these reasons they too tend to favor majoritarian electoral
systems. Multi-national federalists, by contrast, may favor majoritarian systems
within regions – enabling each region to have self-government – but may prefer
more proportional (in the house) or parity (in the senate) systems at the level of
the federal legislature.
Winner-takes all and majoritarian systems have obvious drawbacks in deeply
divided territories, or in territories where ascriptive identities are strongly held.
96
Such systems can reinforce existing divisions creating tyrannies of the majority
(or of the plurality). Conflict in Sri Lanka stems from the 1950s where
competition within the dominant Sinhala bloc encouraged linguistic (Sinhala-
only) extremism, and where Tamil votes did not count in the way they might
have (partly because plantation Tamils at that time were disenfranchised). In
Northern Ireland between 1920 and 1972 the dominant Ulster Unionist Party was
able to make all region-wide elections referendums on the national question,
thereby unifying the unionist bloc against Irish nationalists. In the deep South of
the USA, plurality rule, combined with franchise restrictions and
gerrymandering, led to one-party governments dominated by southern whites
for a century. In fairness, however, it has to be said that these are not necessary
results of these electoral systems (and be it noted that changes of such electoral
systems towards more proportional ones may not immediately improve inter-
group relations). Throughout their democratic histories Canada and India, both
multi-lingual federations, have had plurality rule in single-member districts for
federal and regional elections, and have produced federation-wide political
parties capable of appealing across ethnic and linguistic blocs and being broadly
representative. But, their systems are vulnerable to take-over by plurality-
supported hard-liners from the dominant Anglophone and Hindu communities
respectively.
97
Certain integrationists commend the merits of two means of making the votes of
ethnic, religious and linguistic minorities count. One recommends distributive
requirements which require successful candidates for office – or parties - to
achieve minimum levels of support across districts (or a proportion of districts)
over an entire state or region. The logic is to ensure that candidates and parties
have incentives to win broad-based support. The hope is that will ensure that
successful candidates and parties have incentives to appeal to minority voters.
Such systems, however, have certain drawbacks: they may not specify what
happens if no parties or candidates achieve the requisite requirements (i.e. they
fail guarantee clear outcomes); and a candidate might defeat another candidate
even though the “winner” has a lower share of the overall vote. The second is to
commend the alternative vote – preferential voting in single-member districts.
The logic here is that if minorities are present in enough heterogeneous electoral
districts then larger parties and their candidates will seek to win lower-order
preferences from minority voters – and thereby moderate tendencies towards
ethnic, religious or linguistic extremism. This is sometimes called “vote-pooling”.
This possible outcome, however, is heavily dependent upon the possibility of
having enough such heterogeneous districts, and forgets that majorities within
districts can be constructed by extremists as well as by moderates, especially
when the former can represent the latter as betraying the relevant community.
Moreover, the alternative vote, because it requires successful candidates to
obtain 50 per cent of the vote, plus one (before or after lower-order preferences
98
have been counted), makes it very unlikely that small parties will fare well, and
that means minority parties will be disproportionally under-represented. So,
even if minorities may get to influence the behavior of the candidates of larger
parties under the alternative vote, it may well be at the price of achieving the
separate representation that they might win under proportional representation.
Consociational logic, and the management of multi-national or multi-ethnic
federations at the center, point, by contrast, to inclusivity and proportional
representation as the preferred goal and method of the electoral system. Other
mechanisms can be used. Communal rolls have been used in consociational and
semi-consociational systems, e.g. in pre-independence India and Burma, and
independent Lebanon and Cyprus, but for reasons discussed in the text these are
corporate and illiberal in nature. Reserved seats for minorities have been used to
guarantee minority representation in parliaments in numerous polities, both
consociational and non-consociational (e.g. in Croatia for Hungarians, Italians,
Germans and others, and in Western Samoa, for non-indigenous minority
peoples). Over-representation of regions in parliaments may achieve similar
effects. So may ‘affirmative gerrymandering’ within winner-takes-all systems -
where ‘spiders’, ‘lollipops’ and generally contorted districts are drawn to ensure
e.g. African-American or Latino majority districts in the USA. Such systems may
be useful, especially for groups that might otherwise fail to win any elected
officials; but they may also be condemned as tokenistic ‘set-asides’.
99
Liberal consociationalists, however, reject reserved seats (except for very small
minorities); and they reject communal rolls of voters (though they object less to
designated rolls of parliamentarians where the elected officials opt to be part of
one community). For liberal consociationalists the effective choice is between list
proportional representation systems (the formulae that can be used with such
systems are cited in the main text) and the preferential voting system known as
the single transferable vote. List proportional representation systems enable
voters to cast one vote for a party that has presented and ranked-ordered its list
of candidates. The number of candidates elected is determined by the party’s
share of the vote, the formula used (see above), and the district magnitude (the
number of seats to be filled). These systems enable voters to choose that party
with which they identify. They may pick a national, ethnic, religious or linguistic
party, or not, the choice is theirs. Such systems normally enable party leaders to
have control over the rank-order of candidates on the list (though open-list
systems are also possible), and they are sometimes defended because they fortify
the autonomy of political leaders willing to sustain a politics of cross-community
accommodation. By contrast, preferential proportional representation enables
voters to vote for either party or individual candidates, and to express
preferences in numerical order across both parties and candidates. This system
prevents voters from ‘wasting’ their votes - if they express first-preference
support for a candidate or a party that is unpopular the may rest assured that
100
their lower-order preferences are likely to count in determining which
candidates reach the quotas required for election to office. This system has the
advantage of enabling voters to reward the moderate candidates or moderate
parties from other communities by giving them lower-order transfers on their
ballot-papers. It enables both ‘proportionality’ and ‘vote-pooling’.
No political scientist would claim that electoral systems on their own can achieve
successful “inclusion”, especially of minorities, but all would insist that an
inappropriate electoral system may well be deeply damaging for the prospects of
assimilation, integration, national federation, multi-national federation or
consociation.
101
Endnotes
1 This assertion takes issue with the emphasis on ‘social capital’, inspired recently by Robert D.
Putnam’s Making Democracy Work, 1993 – and see his "Tuning in, Tuning Out: The Strange
Disappearance of Social Capital in America." PS: Political Science and Politics xxviii, no. 4 (1995):
664-83. The ‘social capital’ school focuses on societies’ capacities to explain developmental
differences between regions and states, including in political participation, citizens’ conceptions
of their efficacy and democratic development. The notion of ‘social capital’ has merits – and
difficulties - but it is states that determine the trajectories of their societies, including their levels
of social capital, for good or ill.
2 Hobbes, Thomas. Leviathan (or the Matter, Forme & Power of a Common-Wealth Ecclesiastical and
Civill), edited by Richard Tuck, Cambridge Texts in the History of Political Thought. Cambridge:
Cambridge University Press, 1991 (1651), pp. 88-9.
3 Rummel, Rudolph J. Death by Government, with a Foreword by Irving Louis Horowitz. London and
New York: Transaction Publishers, 1997.
4 Power, Samantha. A Problem from Hell: America and the Age of Genocide. New York: Basic Books,
2002.
5 See inter alia Horowitz, Donald L, and Gérard Noiriel, eds. Immigrants in Two Democracies:
French and American Experiences. New York: New York University Press, 1992, and Noiriel,
Gérard. Le Creuset Française. Paris: Sueil, 1988.
6 ‘The consistent policy of the state of Turkey from its inception has been the destruction of
Kurdish culture and the forced assimilation of Kurds into a purely Turkish society’, Hannum,
Hurst. Autonomy, Sovereignty, and Self-Determination: The Accommodation of Conflicting Rights.
Philadelphia: University of Pennsylvania Press, 1990, p 188. See also US Helsinki Watch
Committee. "Destroying Ethnic Identity, The Kurds of Turkey." New York and Washington D.C.,
1988.
7 For contrasting perspectives on some of the US literature see Alba, Richard. "Immigration and
the American Realities of Assimilation and Multiculturalism." Sociological Forum 14, no. 1 (1999):
3-25; Brubaker, Rogers. "The Return of Assimilation? Changing Perspectives on Immigration and
Its Sequels in France, Germany, and the United States." Ethnic and Racial Studies 24, no. 4 (July)
(2001): 531-48; Foner, Nancy, Rubén G Rumbaut, and Steven J Gold, eds. Immigration Research for
a New Century. New York: Sage, 2003; Gans, Herbert J. "Toward a Reconciliation of "Assimilation"
and "Pluralism": The Interplay of Acculturation and Ethnic Retention." International Migration
102
Review 31, no. 4 (1997): 875-92; Glazer, Nathan. "Is Assimilation Dead?" Annals of the American
Academy of Political and Social Science 530 (1993): 122-36, and We Are All Multiculturalists Now.
Cambridge, Mass: Harvard University Press, 1997; Gordon, Milton M. Assimilation in American
Life: The Role of Race, Religion and National Origins. New York: Oxford University Press, 1964;
Horowitz, Donald L, and Gérard Noiriel, eds. Immigrants in Two Democracies: French and American
Experiences. New York: New York University Press, 1992; Kazal, Russell A. "Revisiting
Assimilation: The Rise, Fall and Reappraisal of a Concept in American Ethnic History." The
American Historical Review 100, no. 2 (1995); Portes, Alejandro, and Rubén G Rumbaut. Immigrant
America: A Portrait. Berkeley: University of California Press, 1990; Rumbaut, Rubén G.
"Assimilation and Its Discontents: Between Rhetoric and Reality." International Migration Review
31, no. 4 (1997): 923-60; Sowell, Thomas. Ethnic America: A History. New York: Basic Books, 1981;
Waters, Mary C. Black Identities: West Indian Immigrant Dreams and American Realities. Cambridge,
Mass: Harvard University Press, 2001.
8 For some largely comparative reflections on Canadian assimilationist policies towards the ‘first
nations’ see, inter alia, Armitage, Andrew. Comparing the Policy of Aboriginal Assimilation: Australia,
Canada and New Zealand. Vancouver: University of British Columbia Press, 1995; Cairns, Alan C.
"Dream Versus Reality in "Our" Constitutional Future: How Many Communities." In
Reconfigurations: Canadian Citizenship and Constitutional Change: Selected Essays by Alan Cairns,
edited by D. Williams, 315-47. Toronto: McClelland and Stewart, 1995; Noel, Sid J.R. "Canadian
Responses to Ethnic Conflict: Consociationalism, Federalism and Control." In The Politics of Ethnic
Conflict-Regulation: Case Studies of Protracted Ethnic Conflicts, edited by John McGarry and Brendan
O'Leary, 41-61. London: Routledge, 1993; Werther, Guntram F.A., ed. Self-Determination in
Western Democracies: Aboriginal Politics in a Comparative Perspective. Westport: Greenwood Press,
1992.
9 For some discussions of historic coercive assimilationist policies towards Australasian
aboriginals see inter alia Armitage, Andrew. Comparing the Policy of Aboriginal Assimilation:
Australia, Canada and New Zealand. Vancouver: University of British Columbia Press, 1995; Birch,
Anthony. Nationalism and National Integration. London: Unwin Hyman, 1989; Brett, Judith. "Every
Morning as the Sun Came Up: The Enduring Pain of the "Stolen Generation"." In Times Literary
Supplement, 4-5. London, 1997; Human Rights and Equal Opportunity Commission. "Bringing
Them Home: Report of the National Inquiry into the Separation of Aboriginal and Torres Strait
Islander Children from Their Families." 689. Sydney NSW Australia, 1997; Hutchinson, John. "Re-
Interpreting Cultural Nationalism." Australian Journal of Politics and History 45, no. 3 (1999): 393-
407; Reynolds, Henry. Aboriginal Sovereignty: Reflections on Race, State, and Nation. St. Leonards,
103
NSW: Allen & Unwin, 1996.
10 For an interesting comparison of ‘nations’ to ‘batteries’ see Canovan, Margaret. Nationhood and
Political Theory. Cheltenham: Edward Elgar, 1996.
11 Cited in Tessler, Mark. A History of the Israeli-Palestinian Conflict. Bloomington: Indiana
University Press, 1994, p. 25.
12 See inter alia Finer, Samuel E. "State and Nation Building in Europe: The Role of the Military."
In The Formation of National States in Western Europe, edited by Charles Tilly. Princeton: Princeton
University Press, 1975; Tilly, Charles, ed. The Formation of National States in Western Europe.
Princeton, NJ: Princeton University Press, 1975, passim; Mann, Michael. "The Emergence of
Modern European Nationalism." In Transition to Modernity: Essays on Power, Wealth and Belief,
edited by John A. Hall and Ian C. Harvie, 137-63. Cambridge: Cambridge University Press, 1992;
and Anthony W. Marx [fill in reference]
13 Mill, John Stuart. "Considerations on Representative Government." In Utilitarianism, on Liberty
and Considerations on Representative Government, edited by H.B. Acton. London: Dent, 1988, ch. 16.
14 Gellner, Ernest. Conditions of Liberty: Civil Society and Its Rivals. London: Hamish Hamilton,
1994.
15 Thornberry, Patrick. "The UN Declaration on the Rights of Persons Belonging to National, or
Ethnic, Religious and Linguistic Minorities: Background, Analysis, Observations, and an
Update." In Universal Minority Rights, edited by Alan Phillips and Allan Rosas, 13-76. Abo and
London: Institute of Human Rights Abo Akademi University and Minority Rights Group
International, 1995, p. 21.
16 Rowthorn, Bob. "Migration Limits." Prospect 2003, 24-31.
17 For some discussions of homeland psychology see inter alia Connor, Walker. "The Impact of
Homelands Upon Diasporas." In Modern Diasporas in International Politics, edited by Gabriel
Sheffer, pp. 16-46. London: Croom Helm, 1985; Esman, Milton. "Two Dimensions of Ethnic
Politics: Defence of Homelands, Immigrant Rights." Ethnic and Racial Studies 8, no. July (1985);
and Yiftachel, Oren. "The Homeland and Nationalism." In Encyclopaedia of Nationalism, edited by
Alexander Motyl, pp. 359-83. New York: Academic Press, 2001.
18 For a celebration of cosmopolitanism that tacitly rests on the merits of assimilation see
Waldron, Jeremy. "Minority Cultures and the Cosmopolitan Alternative." In The Rights of Minority
Cultures, edited by Will Kymlicka, 93-122. Oxford: Oxford University Press, 1995.
19 Brubaker, Rogers. "National Minorities, Nationalizing States, and External National Homelands
in the New Europe." In Nationalism Reframed: Nationhood and the National Question in the New
Europe, 55-76. Cambridge: Cambridge University Press, 1996.
104
20 For a consistent polemic on behalf of integration and against multiculturalism see Barry, Brian.
Culture & Equality: An Egalitarian Critique of Multiculturalism. Oxford: Polity, 2001.
21 See Smith, Rogers M. Civic Ideals: Conflicting Visions of Citizenship in U.S. History. New Haven:
Yale University Press, 1997, and Stories of Peoplehood: The Politics and Morals of Political Membership.
Cambridge: Cambridge University Press, 2003.
22 See the chapter by Kymlicka in this collection.
23 In the case of the USA contrast the fears of Schlesinger, A.M. The Disuniting of America:
Reflections on a Multicultural Society. New York: W.W.Norton, 1992 with the careful empirical
arguments of Hall, John A, and Charles Lindholm. Is America Breaking Apart? Princeton, NJ:
Princeton University Press, 1999.
24 For a mere sample of the huge, cacophonous albeit Anglophone literature see inter alia Alba,
Richard. "Immigration and the American Realities of Assimilation and Multiculturalism."
Sociological Forum 14, no. 1 (1999): 3-25; Appiah, K. Anthony. "Culture, Subculture,
Multiculturalism: Educational Options." In Public Education in a Multicultural Society: Policy,
Theory, and Critique, edited by Robert K. Fullinwider: Cambridge University Press, 1996; Barber,
Benjamin R. "Multiculturalism between Individuality and Community: Chasm or Bridge?" In
Liberal Modernism and Democratic Individuality; George Kateb and the Practices of Politics, edited by
Austin Sarat and Dana R. Villa. Princeton: Princeton University Press, 1996; Barry, Brian. Culture
& Equality: An Egalitarian Critique of Multiculturalism. Oxford: Polity, 2001; Baubock, Rainer.
"Liberal Justifications for Ethnic Group Rights." In Multicultural Questions, edited by Christian
Joppke and Steven Lukes. Oxford: Oxford University Press, 1999; Bennett, Fred. "The Face of the
State." Political Studies xlvii, no. 4 (1999): 677-90; Birnbaum, Pierre. "From Multiculturalism to
Nationalism." Political Theory 24, no. 1 (Feb. '96) (1996): 33-45; Eide, A. "A Review and Analysis of
Constructive Approaches to Group Accommodation and Minority Protection in Divided or
Multicultural Societies." Dublin: Forum for Peace and Reconciliation, 1996; Fish, Stanley, The
Trouble with Principle. Cambridge. Mass: Harvard University Press, 1999; Glazer, N.
"Multiculturalism and Public Policy." In Values and Public Polciy, edited by Henry J. Aaron,
Thomas E. Mann and Timothy Taylor, 113-45. Washington, D.C.: Brookings Institute, 1993;
Glazer, Nathan. We Are All Multiculturalists Now. Cambridge, Mass: Harvard University Press,
1997; Guibernau, Montserrat, and John Rex, eds. The Ethnicity Reader: Nationalism,
Multiculturalism and Migration. Oxford: Polity Press, 1997; Habermas, Jurgen. "Struggles for
Recognition in Constitutional States." European Journal of Philosophy 1, no. 2 (1993): 128-55; Harles,
John. "Integration before Assimilation: Immigration, Multiculturalism and the Canadian Polity."
Canadian Journal of Political Science XXX, no. 4 (1997): 711-36; Hell, Judit, and Ferenc L Lendvai.
105
"Multiculturalism and Constitutional Patriotism in Eastern Europe." In Beyond Nationalism.
Soverignty and Citizenship, edited by Fred Dallmayr and Jose Rosales. Lanham, Maryland:
Lexington Books, 2001; Hughes, Robert. The Culture of Complaint: The Fraying of America. Oxford:
Oxford University Press, 1993; Joppke, Christian, and Steven Lukes, eds. Multicultural Questions.
Oxford: Oxford University Press, 2000; Kymlicka, Will. Multicultural Citizenship: A Liberal Theory
of Minority Rights. Oxford: Oxford University Press, 1995; Kymlicka, Will, and Wayne Norman,
eds. Citizenship in Diverse Societies. Oxford: Oxford University Press, 2000; Lehning, Percy.
"Towards a Multicultural Civil Society: Social Capital and Democratic Citizenship." Government
and Opposition (1998); Linder, Wolf. Swiss Democracy: Possible Solutions to Conflict in Multicultural
Societies. New York: St. Martin's Press, 1994; Lukes, Steven. Liberals and Cannibals: The Implications
of Diversity. London: Verso, 2003; Modood, Tariq. "Multiculturalism, Secularism and the State."
Critical Review of International Social & Political Philosophy 1, no. 3 (1998): 79-97; Okin, Susan
Moller. "Feminism and Multiculturalism; Some Tensions." Ethics 108, no. July (-1998): 661-84;
Parekh, Bhikhu. "Dilemmas of a Multicultural Theory of Citizenship." Constellations 4, no. 1
(1997): 54-62; Raz, Joseph. "Multiculturalism." Dissent, no. Winter (1994): 67-79; Shapiro, Ian, Will
Kymlicka, Ethnicity and Group Rights Edited by Ian Shapiro and Will Kymlicka, Nomos ; 39. New
York: New York University Press, 1997; Shweder, Richard A, Martha Minow, and Hazel Rose
Markus, eds. Engaging Cultural Differences: The Multicultural Challenge in Liberal Democracies. New
York: Russell Sage Foundation, 200#; Takaki, Ronald. A Different Mirror: A History of Multicultural
America. Boston: Little Brown, 1993; Taylor, C. "The Politics of Recognition." In Multiculturalism
and 'the Politics of Recognition', edited by C. Taylor and A. Gutman. Princeton: Princeton
University Press, 1992; Taylor, Charles. Multiculturalism and the Politics of Recognition. Princeton
NJ: Princeton University Press, 1992; Tempelman, Sasja. "Constructions of Cultural Identity:
Multiculturalism and Exclusion." Political Studies XLVII (1999): 17-31; van den Berghe, Pierre.
2001; Waldron, Jeremy. "Multiculturalism and Melange." In Public Education in a Multicultural
Society: Policy, Theory and Critique, edited by Robert K. Fullinwider, 90-118: Cambridge Univeristy
Press, 1996; Ward, Cynthia. "The Limits of "Liberal Republicanism": Why Group-Based Remedies
and Republican Citizenship Don't Mix." Columbia Law Review 91, no. 3 (1991): 581-607; Young, Iris
Marion. "A Multicultural Continuum: A Critique of Will Kymlicka's Ethnic-Nation Dichotomy."
Constellations 4, no. 1 (1997).
25 Inter-state modes of multi-culturalism, confederations, such as the European Union cannot be
considered here, though the discussion of federal and consociational principles is in many
respects equally pertinent at the inter-state level within alliances, leagues and confederations.
26 See inter alia Lijphart, Arend, ed. Conflict and Coexistence in Belgium: The Dynamics of a Culturally
106
Divided Society, Research Series, No 46. Berkeley: Institute of International Studies University Of
California, 1981; "Consociation and Federation: Conceptual and Empirical Links." Canadian
Journal of Political Science xii, no. 3 (1979): 495-515; "Consociational Democracy." World Politics 21,
no. 2 (1969): 207-25; Democracy in Plural Societies: A Comparative Exploration. New Haven, London:
Yale University Press, 1977; The Politics of Accommodation: Pluralism and Democracy in the
Netherlands. Berkeley: University of California Press, 1968; Power-Sharing in South Africa. Berkeley:
University of California Press, 1985; "Self-Determination Versus Pre-Determination of Ethnic
Minorities in Power-Sharing Systems." In The Rights of Minority Cultures, edited by Will Kymlicka.
Oxford: Oxford University Press, 1995; and "Typologies of Democratic Systems." Comparative
Political Studies 1, no. 1 (1968): 3-44.
27 ‘Power-sharing’ is not a synonym for consociation because there are other than consociational
ways to share power: e.g. through federation, intermittent and temporary coalitions, alternating
governments, the separation of powers, and through generally ‘collegial’ institutions, cf. Collins,
Randall. "Democratization from the Outside In: A Geopolitical Theory of Collegial Power." In
Macro History: Essays in Sociology of the Long Run, 110-51. Stanford, CA: Stanford University Press,
1999. Each of the types of power-sharing just listed can be deployed in consociational formats.
What makes consociational power-sharing distinctive is that it mandates power-sharing across
communities through formulae of proportionality and autonomy.
28 Horowitz, Donald L. Ethnic Groups in Conflict. Berkeley: University of California Press, 1985, p.
256.
29 Brass, Paul R. "Ethnic Conflict in Multiethnic Societies: The Consociational Solution and Its
Critics." In Ethnicity and Nationalism: Theory and Comparison, 333-48. New Delhi: Sage, 1991, pp.
333, 338.
30 Brass, op. cit., p. 346, n.11.
31 For a typical libertarian objection to affirmative action see Sowell, Thomas. Preferential Policies:
An International Perspective. New York: Quill, 1990.
32 Brass, p. 334, my emphasis.
33 Jung, Courtney, and Ian Shapiro. "South Africa's Negotiated Transition: Democracy,
Opposition, and the New Constitutional Order." Politics & Society 23, no. 3 (1995): 269-308, cf. 274,
275.
34 Ibid. , p. 293.
35 Originally posed by Brian Barry the claim is now endlessly recycled, Barry, Brian. "The
Consociational Model and Its Dangers." European Journal of Political Research 3 (1975): 393-413,
"Review Article: Political Accommodation and Consociational Democracy." British Journal of
107
Political Science 5 (1975): 477-505. Observing that Lijphart accepts Robert Dahl’s definition of
polyarchy, one critic maintains that Dahl’s emphasis on competitive politics implies that
consociational democracy is a contradiction in terms, Van Schendelen, M.P.C.M. "The Views of
Arend Lijphart and Collected Criticism." Acta Politica 1 (1984): 19-55, p. 32.
36 Brass, op. cit., p. 340. Since he uses Great Britain for his argument rather than the United
Kingdom one must assume that his exclusion of Northern Ireland from his endorsement of its
adversarial politics is deliberate (or else he fallaciously equates Great Britain with the United
Kingdom). If Northern Ireland is considered part of the United Kingdom political system, which
it has been, the merits of adversarial politics are much less obvious. As for the USA the claims for
the integrative effects of adversarial politics with respect to native Americans and the
descendants of slaves have been unconvincing to successive cohorts of foreign observers of the
USA from Tocqueville to Myrdal. Their stories are an integral part of the critical histories of
American political development.
37 Halpern, Sue. "The Disorderly Universe of Consociational Democracy." West European Politics 9,
no. 2 (1986): 181-97.
38 Kieve, Ronald. "Pillars of Sand: A Marxist Critique of Consociational Democracy in the
Netherlands." Comparative Politics 13, no. April (1981).
39 These objections to consociation are common to distinguished liberal political philosophers
like Brian Barry and Ian Shapiro and distinguished liberal political scientists such as Ian S
Lustick and Paul Brass, as well as more conservative distinguished political scientists. It is of
interest that one can classify three of these standard political and ethical arguments against
consociation according to the tropes of Albert Hirschman’s The Rhetoric of Reaction: Perversity,
Futility, Jeopardy.
40 In On Liberty and Considerations on Representative Government Mill, like Tocqueville, was much
more concerned about the tyranny of the ignorant, the conformist, the puritanical, and the
unlettered masses. In Considerations, however, he did decree the unlikelihood of multi-national
or bi-national states, but largely because he seemed to consider one nationality and one language
necessary for one public opinion.
41 Bauer, Otto. The Question of Nationalities and Social Democracy, Edited by Ephraim Nimni, Trans.
Joseph O'Donnell. Minneapolis: University of Minnesota Press, 2000.
42 Nagel, Jack H. "Expanding the Spectrum of Democracies: Reflections on Proportional
Representation in New Zealand." In Democracy and Institutions: The Life Work of Arend Lijphart,
edited by Markus M.L. Crepaz, Thomas A Koelbe and David Wilsford, 113-27. Ann Arbor: The
108
University of Michigan Press, 2000. This spectrum was introduced in a criticism of some of
Lijphart’s formulations, and originates in the work of Buchanan and Tullock.
43 This idea had its origins in thinking about how to classify cross-community levels of supports
for elements of the unfolding peace process in Northern Ireland.
44 Lijphart recognizes this by describing joint decision making as the key characteristic of power-
sharing. Our analysis suggests some objections to Lijphart’s classifications may be dismissed, e.g.
Horowitz, op. cit., p. 575, complains none of the four developing countries identified by Lijphart
as having followed consociational practices, viz. Lebanon, Malaysia, Surinam and the
Netherlands Antilles, had grand coalitions because each group was represented by more than
one set of leaders.
45 See Taagepera, Rein, and Matthew Soberg Shugart. Seats and Votes: The Effects and Determinants
of Electoral Systems. New Haven and London: Yale University Press, 1989, pp. 273-5, and Lijphart,
Arend, ed. Electoral Systems and Party Systems: A Study of Twenty-Seven Democracies, Comparative
European Politics. Oxford: Oxford University Press, 1994., pp. 25-30.
46 See O'Leary, Brendan. "The Nature of the Agreement." Fordham Journal of International Law 22,
no. 4 (1999): 1628-67.
47 Another issue here is the question of exclusion and inclusion of communities over time. If a
pivotal or dominant party alternates in its choice of community parties (as occurred in the
Netherlands before the 1970s) in sharing executive power how may such a case be classified?
One simple answer is that the executive is consociational as regards the included communities
(and variable in degree: i.e. the executive may enjoy complete, concurrent or weak support from
within the included communities). But, it is not consociational as regards the excluded. Where
one community stays constantly in the executive, and alternates its partners from other
communities it may seem plausible to label this phenomenon a ‘diachronic grand coalition’, but
this looks too much like conceptual stretching. My thanks to Matthijs Bogaards for obliging me to
address this matter (personal communication).
48 S.J.R. Noel cites historian Frank Underhill as saying that the great Canadian invention of the
nineteenth century was the ‘composite bi-racial, bi-cultural party, uniting both French and
English voters’, "Canadian Responses to Ethnic Conflict: Consociationalism, Federalism and
Control." In The Politics of Ethnic Conflict-Regulation: Case Studies of Protracted Ethnic Conflicts,
edited by John McGarry and Brendan O'Leary, 41-61. London: Routledge, 1993, p. 49.
49 This discussion has benefited from discussions with Dr. Matthijs Bogaards of the University of
Southampton.
50 The second and third criteria are Bogaards’s; the first and last suggestions are mine.
109
51 See Lijphart’s comments on the Study Project on Christianity in Apartheid Society, Power-
Sharing in South Africa p. 48-51.
52 See Lijphart’s commentary on South Africa’s 1993 Constitution, Power-Sharing in South Africa,
pp. 52-64, especially p. 61.
53 Lijphart meant to use quasi to describe ‘having some resemblance, but false’ (personal
communication).
54 Donald Rothchild, ‘Hegemonial Exchange: An Alternative Model for Managing Conflict in
Middle Africa.’ In Ethnicity, Politics and Development (eds.) Dennis L Thompson and Dov Ronen
(Boulder, Co: Lynne Rienner, 1986); Managing Ethnic Conflict in Africa (Washington, D.C.:
Brookings Institution Press, 1997); and ‘Middle Africa: Hegemonial Exchange and Resource
Allocation, ‘ in Comparative Resource Allocation, in (eds.) Alexander Groth and Larry Wade (New
York: Sage Publications, 1984).
55 For example, the appointment of Protestants to Catholic school boards would be against the
spirit of consociational accommodation if education is supposed is a domain of community
autonomy.
56 The literature on electoral systems and their consequences is huge. For useful samples see
Barry Ames, ‘Electoral Strategy under Open-List Proportional Representation.’ American Journal
of Political Science 39, no. 2 (1995): 406-34; Andre Blais, ‘The Classification of Electoral Systems.’
European Journal of Political Research 16 (1988): 99-110, Vernon Bogdanor and David Butler (eds.)
Democracy and Elections: Electoral Systems and Their Political Consequences (Cambridge: Cambridge
University Press, 1983), Andrew M. Carstairs, A Short History of Electoral Systems in Western
Europe (London: Allen and Unwin, 1980), Gary Cox, Making Votes Count: Strategic Coordination in
the World's Electoral Systems (Cambridge: Cambridge University Press, 1997), K.R. de Silva,
‘Electoral Systems for Multi-Ethnic Societies.’ (Geneva: UNRISD, 1995), David M. Farrell,
Comparing Electoral Systems (Hemel Hempstead: Prentice-Hall, 1997), Bernard Grofman and
Arend Lijphart (eds.) Electoral Laws and Their Political Consequences (New York: Agathon Press,
1986), Bernard Grofman and Andrew Reynolds, ‘Electoral Systems and the Art of Constitutional
Engineering: An Inventory of the Main Findings’, in Rules and Reason: Perspectives on
Constitutional Political Economy (eds.) R. Mudambi, P. Navarra and G. Sobbrio (Cambridge:
Cambridge University Press, 2000), Jennifer Hart, Proportional Representation. Critics of the British
Electoral System 1820-1945 (Oxford and New York: Oxford University Press, 1992), Jean A
Laponce, ‘The Protection of Minorities by the Electoral System.’ Western Political Quarterly X
(1957): 318-39, Arend Lijphart (ed.) Electoral Systems and Party Systems: A Study of Twenty-Seven
Democracies, Comparative European Politics (Oxford: Oxford University Press, 1994), A. Lijphart
110
and B. Grofman (eds.) Choosing an Electoral System: Issues and Alternatives (New York: Praeger,
1984), Dieter Nohlen, Elections and Electoral Systems (New Delhi: Macmillan, 1996), Benjamin
Reilly and Andrew Reynolds, Electoral Systems and Conflict in Divided Societies, Papers on
International Conflict 2 (Washington DC: National Academy Press, 1999), and The International Idea
Handbook of Electoral Design (Stockholm: International Institute for Democracy and Electoral
Assistance, 1997), W. Rule and JF Zimmerman (eds.) Electoral Systems in Comparative Perspective:
Their Impact on Women and Minorities (Westport, CT: Greenwood Press, 1994), Giovanni Sartori,
‘The Influence of Electoral Systems: Faulty Laws or Faulty Methods?’, in Electoral Laws and Their
Political Consequences, (ed.) B. Grofman and A. Lijphart (New York: Agathon Press, 1986), pp. 43-
68, Rein Taagepera and Matthew Soberg Shugart, Seats and Votes: The Effects and Determinants of
Electoral Systems (New Haven and London: Yale University Press, 1989).
57 The list includes formulae that have not yet been used in functioning consociational systems.
58 Michael Gallagher, ‘Proportionality, Disproportionality and Electoral Systems.’ Electoral Studies
10, no. 1 (1991), p. 49.
59 I follow Michael Gallagher, ‘Comparing Proportional Representation Electoral Systems:
Quotas, Thresholds, Paradoxes and Majorities’, British Journal of Political Science 22 (1992): 469-96,
and exclude imperiali quotas or divisors because they are not generally proportional as Gallagher
has demonstrated.
60 Gallagher, ‘Comparing Proportional Representation Electoral Systems’.
61 Judges or independent commissions working to establish criteria are incomparably than
politicians in ensuring that electoral gerrymandering is minimized. Indeed, to achieve the
equality of voters, a key measure of inclusiveness, this is perhaps the simplest lesson of
democratic design.
62 Gallagher, ‘Comparing Proportional Representation Electoral Systems’, pp. 491-2, has
demonstrated that ‘highest averages’ methods are free from the paradoxes that can arise with
‘largest remainders’ methods (where parties growing relative to other parties might lose a seat, or
an expansion in the number of seats in a parliament may lead to a region having its
representation reduced even though its population remains the same).
63 For an extended discussion see Arend Lijphart, ‘Proportionality by Non-PR Methods: Ethnic
Representation in Belgium, Cyprus, Lebanon, New Zealand, West Germany and Zimbabwe.’ In
Choosing an Electoral System: Issues and Alternatives, (eds.) A. Lijphart and B. Grofman (New
York: Praeger, 1984), pp. 111-23.
64 The presumed Muslim demographic expansion had two causes: a higher birth-rate, and
Muslim over-representation among the Palestinian refugee population - from the 1948 war
111
accompanying the formation of Israel, and from the ‘Black September’ massacre in Jordan in
1970.
65 This situation should be distinguished from one in which positive preferential or affirmative
action is required to rectify historical imbalances (that is usually the result of historic
discrimination). Positive preferential or affirmative action policies may well temporarily over-
represent the under-represented in new cohorts of appointees. But these policies are intended to
achieve overall proportionality not to deliver net over-representation.
66 Constitution of Cyprus 1960, Article 46 [accessed on October 21 2002].
http://www.pio.gov.cy/cygov/constitution/appendix_d_part3.htm
67 For example see Arend Lijphart, ‘Multiethnic Democracy’, in The Encyclopaedia of Democracy
(ed.) Seymour Martin Lipset, (London: Routledge, 1995: 857).
68 Hanf, ‘Coexistence in Prewar Lebanon’, p. 95.
69 One source of tension in the Lebanon before its civil war began in 1975 was that Maronites
regarded certain key offices as their ‘preserves’: ‘the command of the army, the directorates of
military intelligence and state security, the governorship of the Central Bank and the
chairmanship of the Conséil d’État’, Hanf, Coexistence in Wartime Lebanon, p. 95.
70 Lijphart, Democracy in Plural Societies, p. 41.
71 See inter alia Hurst Hannum, Autonomy, Sovereignty, and Self-determination: The Accommodation
of Conflicting Rights. rev. ed (Philadelphia: University of Pennsylvania Press, 1996) and Ruth E.
Lapidoth, Autonomy: Flexible solutions to ethnic conflicts (Washington, DC: United States Institute
of Peace Press, 1997). See the discussion of territorial autonomy below, and the paper by
Professor Stepan in this symposium.
72 A confederation in which the member-states did not have complete or concurrent
representation in the executive would no longer be a confederation: it would be well on the way
to being a federation (cf. the discussions over the new constitution of the European Union).
73 Kenneth D. McRae, ‘The Principle of Territoriality and the Principle of Personality in
Multilingual States.’ International Journal of the Sociology of Language 4 (1975): 33-54.
74 Available at http://www.pio.gov.cy/cygov/constitution/appendix_d_part5.htm and
http://www.pio.gov.cy/cygov/constitution/appendix_d_part1.htm [Accessed October 21
2002]).
75 Oren Yiftachel, ‘'Right-Sizing' or 'Right-Shaping?' Politics, Ethnicity and Territory in Plural
States’, in Right-Sizing the State: The Politics of Moving Borders, (eds.) B. O'Leary, I. S. Lustick and
T. Callaghy, (Oxford, Oxford University Press: 358-87).
112
76 John C. Calhoun’s defense of the concurrent majority principle is cited by Lijphart, Democracy
in Plural Societies, p. 37, drawing on G. Kateb, ‘The Majority Principle: Calhoun and His
Antecedents’, Political Science Quarterly, 84, 4: 583-605. Calhoun argued that the threats of
blockages would promote consensual and mutual policy-changes. That he used his arguments in
defense of a commitment to slaveholding institutions explains why his formal arguments are not
often recalled.
77 Lijphart, Democracy in Plural Societies, p. 38
78 Section IV. B. Article 6 (1) and (2) of the constitution of the Federation of Bosnia Herzegovina
provides an example of a strong veto-right within the cabinet: ‘(1) Decisions of the Cabinet that
concern the vital interest of any of the constituent peoples shall require consensus. This provision
may be invoked by one-third of the Ministers, excluding the Prime Minister and Deputy Prime
Minister, unless otherwise determined by the Constitutional Court….(2) when the Prime Minister
concludes that the Government cannot reach consensus described in paragraph (1), he shall refer
the pending matter to the President or Vice-President, whoever is not from the same constituent
people as is the Prime Minister, for a decision without delay’.
79 See Professor Kymlicka’s contribution to this symposium.
80 Max Weber went further suggesting that the Machtstaat cannot realize democracy, cf. Weber,
Max. "Between Two Laws." In Political Writings, edited by Peter Lassman and Ronald Speirs, 75-
80. Cambridge: Cambridge University Press, 1994.
81 See inter alia Daniel Elazar, Exploring Federalism (Tuscaloosa: University of Alabama, 1987),
Preston King, Federalism and Federation (London: Croom Helm, 1982), Preston King, Federalism and
Federation, 2nd ed. (London: Frank Cass, 2001).
82 For elaborations on these distinctions see Ronald A Watts, "Federalism, Federal Political
Systems, and Federations," Annual Review of Political Science 1 (1998).
83 On the importance of peoplehood see Rogers M Smith, Stories of Peoplehood: The Politics and
Morals of Political Membership (Cambridge: Cambridge University Press, 2003).
84 Brendan O'Leary, "The Kurds Must Not Be Betrayed Again," Financial Times, March 24th 2003.
85 David R Rezvani, Federacy: The Dynamics of Semi-Sovereign Territories (Unpublished Manuscript)
(2003).
86 The expression is Elazar’s, op. cit.
87 Alfred Stepan, "Federalism and Democracy: Beyond the U.S. Model," Journal of Democracy 10,
no. 4 (1999).
88 For recent discussions of Shi’ite politics in Iraq see inter alia David Gardner, "Time of the Shia,"
Financial Times, August 30/31 2003, Faleh A Jabar, The Shi'ite Movement in Iraq (2003). The
113
perspectives of Muhammed Baqir al-Sadr’s, executed by Saddam Hussein in 1980, are
widespread among Shià, and diverge from the political Islam of Khomeini and his successors.
Historically communism was strong among the Shià of Iraq, but whether the Iraqi Communist
Party will experience a significant revival in the 2000s remains to be seen.. What is plain is that
there are significant numbers of secular Shià, and that significant numbers of religious Shià do
not agree with political leadership being monopolized by clerics. Two authorities remark on a
‘dangerous tendency in the West to equate secular with Sunni and Islamist with Shi’ite’, Fuller
and Francke, op. cit., p. 108.
89 Ronald L Watts, "Federalism, Federal Political Systems, and Federations," Annual Review of
Political Science 1 (1998), Ronald L Watts, "Models of Federal Power-Sharing," International Social
Science Journal, no. March (2001).
90 The proportion of public expenditure allocated by regions as opposed to federal governments
may well be a better guide to their autonomy and power than the text of the constitution --- for
discussions of such measurements see Watts, op. cit. 2001, p. 29, and Arend Lijphart,
"Consociation and Federation: Conceptual and Empirical Links," Canadian Journal of Political
Science xii, no. 3 (1979)., p. 505.
91 See Murray Forsyth, ed., Federalism and Nationalism (Leicester: Leicester University Press, 1989).,
p. 4.
92 See William H. Riker, Federalism: Origin, Operation, Significance (Boston: 1964).
93 See Nathan Glazer, "Federalism and Ethnicity: The American Solution," in Ethnic Dilemmas,
1964-82, ed. N. Glazer (Cambridge, Mass: Harvard University Press, 1983).
94 Milton Gordon, Assimilation in American Life: The Role of Race, Religion and National Origins (New
York: Oxford University Press, 1964)., p. 133.
95 Glazer, op. cit., p. 276.
96 Gordon, op. cit., passim.
97 Publius [John Jay], in James Madison, Alexander Hamilton, and John Jay, The Federalist Papers,
Edited and with an Introduction by Isaac Kramnick, Penguin Classics (Harmondsworth: Penguin, 1987
(1788))., Paper II, p. 91
98 David McDowall, A Modern History of the Kurds, 2nd and revised ed. (London: I.B. Tauris,
2000)., pp. 343-67. It was a standard trope in the English language media that favored
intervention in Iraq that Saddam Hussein had ‘gassed his own people’. That he gassed people is
not in question. That he shared a common peoplehood with his victims most certainly is.
99 The US’s most influential foreign policy journal published a scenario for re-building Iraq in the
summer of 2003, Adeed Dawisha and Karen Dawisha, "How to Build a Democratic Iraq," Foreign
114
Affairs 82, no. 3 (2003). The Dawishas advocate maintaining ‘Iraq’s present administrative
structure, under which the country is divided into 18 units’, p. 39, while on the same page insist
that the Kurds should have their own territorial ‘unit in the federal structure’. This is
contradictory, and whatever it means, the argument is outmoded by the de facto boundaries of
Iraqi Kurdistan, which cut across prior provincial jurisdictions. For similar advocacy of the
eighteen provinces see Rachel Bronson, et. al., Guiding Principles for U.S. Post-conflict Policy in Iraq,
New York: Council on Foreign Relations, 2003.
100 Brancati D. 2004 in press. Is Federalism a Panacea for Post-Saddam Iraq? The Washington
Quarterly 27. Brancati’s case is based on her doctoral research on parties, decentralization and
ethnic conflict – much of which I agree with. It is the erroneous objection of blueprint thinking to
Iraq to which I object.
101 Forsyth, op. cit., p. 4.
102 See inter alia Michael Hechter, Containing Nationalism (Oxford: Oxford University Press, 2000),
Michael Keating, "Managing the Multinational State: Constitutional Settlement in the United
Kingdom," ed. Trevor Salmon and Michael Keating (2001), Juan Linz, "Democracy,
Multinationalism and Federalism" (paper presented at the Juan March Institute, 1997), Margaret
Moore, The Ethics of Nationalism (Oxford: Oxford University Press, 2001), Stepan, "Federalism and
Democracy: Beyond the U.S. Model.", Alfred Stepan, "Modern Multinational Democracies:
Transcending a Gellnerian Oxymoron," in The State of the Nation : Ernest Gellner and the Theory of
Nationalism, ed. John A. Hall (New York: Cambridge University Press, 1998).
103 Jack Snyder, From Voting to Violence: Democratization and Nationalist Conflict (New York: W.W.
Norton, 2000)., p. 327.
104 Brendan O'Leary, "An Iron Law of Federations? A (Neo-Diceyian) Theory of the Necessity of a
Federal Staatsvolk, and of Consociational Rescue. The 5th Ernest Gellner Memorial Lecture,"
Nations and Nationalism 7, no. 3 (2001).. For other discussions see Thomas M Franck, Why
Federations Fail: An Inquiry into the Requisites for Successful Federation (New York: New York
University Press, 1968), Ursula K Hicks, Federalism, Failure and Success: A Comparative Study
(London: Macmillan, 1978), Ronald L Watts, Comparing Federal Systems in the 1990s (Kingston,
Ontario: Institute of Intergovernmental Relations/Queen's University, 1996), Ronald L Watts,
"The Survival or Disintegration of Federations," in Options for a New Canada, ed. Ronald Watts
and Douglas M Brown (Toronto: University of Toronto Press, 1971).
105 John McGarry and Brendan O'Leary, "Federalism, Conflict-Regulation and National and
Ethnic Power-Sharing" (paper presented at the Annual Meeting of the American Political Science
Association, Philadelphia, August 28-31 2003 2003)., August 28-31..
115
106 O’Leary, 2001, op.cit.
107 Katharine Adeney, "Constitutional Centring: Nation Formation and Consociational Federalism
in India and Pakistan," Journal of Commonwealth & Comparative Politics 40, no. 3 (2002).
116

Occasional Paper: United Nations Development Programme

Uploaded by

Copyright:

Available Formats

Occasional Paper: United Nations Development Programme

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Occasional Paper: United Nations Development Programme

Uploaded by

Copyright:

Available Formats

View metadata, citation and similar papers at core.ac.

uk brought to you by CORE

United Nations Development Programme

Human Development Report Office

BUILDING INCLUSIVE STATES

Lauder Professor of Political Science

Director Solomon Asch Center for the Study of Ethnopolitical Conflict

1. Functioning states are necessary conditions of inclusion; and the major

define societies, and their possibilities1. Impersonalized institutions that have

some degree of centralized and procedurally governed political decision-making

promote human development, or be inclusive in any meaningful sense.

Conversely, failing and failed states are personalized: dominated by rulers – a

become ‘kleptocracies’, governments of thieves, before or during their collapse.

They lack coherent, institutionalized, rule-governed patterns of centralized

monopoly on the regulation of coercion, and are challenged by guerillas,

paramilitaries, terrorists, Mafiosi; they may be invaded, looted, and occupied by

if they deem it necessary.

within well-ordered states. ‘Inclusivity’ and human development require more

because they are unlikely to grow in ‘anarchia’. Order, to put it bluntly, is a

necessary condition of inclusivity. Without it we have Thomas Hobbes’ condition

thereof is uncertain: and consequently no culture of the Earth, no Navigation, nor

use of the commodities that may be imported by Sea; no commodious Building;

no Instruments of moving, and removing such things as require much force; no

Knowledge of the face of the Earth; no account of Time; no Arts; no Letters; no

to emphasize the necessity of order, but wrong to mandate an authoritarian

ordered state there cannot be democracy, functioning markets, human

been, respectively, people-killers and nation-killers. In the twentieth century

Genocide, killing people because of their presumed ascriptive characteristics, has

those deemed political opponents of regimes, has been equally recurrent in

perpetrators. Governments have also organized, encouraged, or not stopped,

shores – people whom they have helped to define as undesirable, non-

indigenous, non-national, or disloyal. The twentieth century can be described as

the century of expulsion, of the ‘cleansing’ of populations, as much as one of

genocide. Governments have also been the prime architects of policies of

discriminatory control: organizing dominant national, ethnic, religious or

communal groups, disorganizing and subordinating other national, ethnic,

religious or cultural groups through policies of systematic discrimination, or

These should be commonplace thoughts. Regrettably, they are not sufficiently

appreciated. States define people’s life-chances, both individually and

greater inclusion, as defined in this report by Amartya Sen, may best be

facilitated. Exclusion, in the regulation of national, ethnic and religious

treatment, privileging communities: counting some categories of persons and

enforced, through interventions ranging from diplomatic sanctions, economic

sanctions, military embargos (or lifting of embargos), military occupations, and

international criminalization of genocidal officials, then the worst forms of

exclusion may be halted or inhibited. Success in this endeavor requires both

moral universalism – genocide and expulsion need to be treated as criminally

wrong everywhere – and a re-orientation of the foreign policies of the great

prevention and punishment of genocide and expulsion as in their interests4.

against exclusionary practices – against discriminatory racist, religiously

intolerant and xenophobic regimes – had major successes in the twentieth

century. The undermining of apartheid in South Africa had both international

affirmative and inclusionary history of the twentieth century. Inclusion, by

not coercively homogenize peoples, that do not entrench unjustified differential

treatments of peoples, and that are embedded in the practices of well-ordered

states. Elaborating this understanding is the purpose of this paper.

The political and cultural homogenization of people is the underling goal of

many states that promote assimilation. Historically it has been an important

both of which have influenced the political development of many member-states