The following is an archived copy of a message sent to a Discussion List run by the Campaign Against Sanctions on Iraq.

Views expressed in this archived message are those of the author, not of the Campaign Against Sanctions on Iraq.

[Main archive index/search] [List information] [Campaign Against Sanctions on Iraq Homepage]


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[casi] Multi-national Federalism, Federacy, Power-Sharing & the Kurds of Iraq



http://www.kdp.info/pdf/the_future_of_Iraq.pdf

1

Multi-national Federalism, Federacy, Power-Sharing & the Kurds of Iraq

Brendan O'Leary

Lauder Professor of Political Science

Director of the Solomon Asch Center for the Study of Ethnopolitical Conflict

University of Pennsylvania

Address for correspondence:

3819-33 Chestnut Street, Suite 305

St. Leonard's Court Complex

Philadelphia, PA 19104

Or boleary@sas.upenn.edu

Text to accompany keynote address to the conference

'Multi-Nationalism, Power-Sharing and the Kurds in a New Iraq'

Cafritz Foundation Conference Center, George Washington University

September 12, 2003

This paper has benefited from joint work with Professor John McGarry of the
Queen's

University Canada, and from research funds provided by the United States
Institute of

Peace and the Lauder chair. The author is solely responsible for its
contents.

2

The future of Iraq is, and should be, a matter for the peoples of Iraq to
determine,

through their own constitutional deliberations, negotiations and processes
of ratification.

This phrasing emphasizes that outsiders can advise but that insiders should
decide; it also

highlights an insistence that Iraq is not comprised of just one people1,
though the

peoples of Iraq may well agree to share a common state and a common
citizenship. What

political scientists and constitutional lawyers may usefully do as outsiders
is to provide

comparative reflections, useable knowledge from experiences elsewhere, that
may help

insiders clarify their preferences and negotiating strategies. It is our
duty to clarify

feasible proposals rather than to commend our own favored recipes; and it is
important

that we specify where there is no professional confidence over what
constitutional and

institutional designs work best, or simply well. The constitutional
restructuring of Iraq

necessarily involves debates that have already begun over the processes of
constitutionmaking,

as well as the full panoply of issues entailed in institutional design, such
as the

organization of executives, legislatures and courts, electoral and
party-systems, territorial

governance, human-rights protections, fiscal and monetary agencies and
formulae, and

the organization and accountability of military and policing institutions.

It is my task to address emergent ideas about federalizing a renewed Iraq.
And, to do so

through the prism of Kurdish interests, identity and ideas, the prime focus
of the

organizers of this conference, to whom we are all indebted. I will do so by
distinguishing

national from multi-national federations, and evaluating their merits; I
will then elaborate

the merits of 'federacy' as opposed to wholesale federal arrangements -
drawing upon

my own thoughts2, and some unpublished work by David Rezvani3.

*****

Federalism is a political philosophy that commends both shared and
self-government4.

Federal political systems encompass a range of possible political
organizations that

1 On the importance of peoplehood see Smith, Rogers M. 2003. Stories of
Peoplehood: the Politics and Morals of

Political Membership. Cambridge: Cambridge University Press.

2 O'Leary, Brendan. 2003. The Kurds Must Not Be Betrayed Again. Financial
Times, March 24th, 23.

3 Rezvani, David R. 2003. Federacy: The Dynamics of Semi-Sovereign
Territories (unpublished manuscript).

4 See inter alia Elazar, Daniel. 1987. Exploring Federalism. Tuscaloosa:
University of Alabama, King, Preston. 1982.

Federalism and Federation. London: Croom Helm, King, Preston. 2001.
Federalism and Federation. 2nd ed. London:

Frank Cass.

3

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 where I live, Philadelphia5. 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 government6. 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; and it will be a central part of my argument
to suggest that,

for Kurds and Iraqis, following the US model of federation would be
inappropriate. Not

because I harbor any hostility towards my new home and its institutions - to
the contrary,

but rather because it is plain that Iraqis and Kurds should look elsewhere
for inspirations

that suit their circumstances better7. Federations vary, first, in the
extent to which they are

'majoritarian'. All federations place some constraints on the powers of
federation-wide

5 For elaborations on these distinctions see Watts, Ronald A. 1998a.
Federalism, Federal Political Systems, and

Federations. Annual Review of Political Science 1:117-37.

6 The expression is Elazar's, op. cit.

7 Whatever one's views on the merits of the Allied intervention and
occupation of Iraq few doubt that to impose mutatis

mutandis the current American constitution on Iraq would feed already extant
anti-American nationalism amongst Arab

Iraqis. The frequent exemplary invocation by some American neo-conservatives
of the occupations and regimetransformations

of (west) Germany and Japan (Italy, Austria and Korea are usually omitted)
is oblivious to the fact

these constitutional transformations were partly indigenous, and did not
involve the simple export of American

constitutional models.

4

majorities, but do so to different degrees8. The USA, Australia and Brazil
for example,

allow equal representation to each of their regions in the Senate,
generating massive overrepresentation

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 singleparty

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, Iraqis and Kurds would be
welladvised

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
even a Shi'ia

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

amongst them in religiosity, and dispositions towards Iran and other
neighboring states9.

The more homogeneously Shi'a mobilize and act then 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

8 Stepan, Alfred. 1999. Federalism and Democracy: Beyond the U.S. Model.
Journal of Democracy 10 (4):19-34.

9 For recent discussions of Shi'ite politics in Iraq see inter alia Gardner,
David. 2003. Time of the Shia. Financial

Times, August 30/31, W1-W2, Jabar, Faleh A. 2003. The Shi'ite Movement in
Iraq.. The perspectives of Muhammed

Baqir al-Sadr's, executed by Saddam Hussein in 1980, are widespread amongst
Shi'a, and diverge from the political

Islam of Khomeini and his successors. Historically communism was strong
amongst the Shi'a 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 Shi'a., and that significant numbers of
religious Shi'a 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.

5

nominations to public office, and over treaty-making. Other second chambers,
such as

those in Canada, India and Belgium, are very weak10. 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 it is the regional
governments

which 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 Lšnder governments, empowering both tiers with distinct
enabling and

blocking powers. In all federations the constitutional division of
competencies (even as

interpreted by the courts) may not always 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

capacities - as in the USA where the federal government's pre-eminence is
now

established11.

Over the distribution of competencies it is not too difficult to foresee the
likely

preferences amongst representative leaders of the major ethnic and religious

communities in Iraq. Given the persistent history of repressive dictatorial
government

from Baghdad, under both Ba'athist and pre-Ba'athist governments, it is
extremely

unlikely that Kurds will want to endorse a strong federal government12.
Kurds will want

maximum feasible domestic autonomy in public policy and law-making. Given
their

persistent partial possession of and incorporation into central governmental
power, Sunni

10 Watts, Ronald L. 1998b. Federalism, Federal Political Systems, and
Federations. Annual Review of Political Science

1:117-37, Watts, Ronald L. 2001. Models of Federal Power-Sharing.
International Social Science Journal (March):23-

32.

11 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 Lijphart, Arend. 1979. Consociation and
Federation: Conceptual and Empirical Links.

Canadian Journal of Political Science xii (3):495-515., p. 505.

12 See inter alia C. Tripp, 2002. A History of Iraq , revised edition.
Cambridge: Cambridge University Press , Natali,

Denise. 2001. Manufacturing Identity and Managing Kurds in Iraq. In
Right-Sizing the State: the Politics of Moving

Borders, edited by B. O'Leary, I. S. Lustick and T. Callaghy. Oxford: Oxford
University Press.

6

Arabs may still retain a political culture that favors a strong central
government, but that

culture will surely be tempered by their current fear that others may be
able to use that

governmental power against them. Some Shi'a, by contrast, may be the most
tempted to

create a strong federal or simply strong central government.

*****

If the makers of a new Iraq decide to create a federation, then whatever the
distribution

of competencies between a future federal government and the future regions,
a critical

choice will be whether that federation is to be mono-national or
multi-national. Mononational

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. We are conferencing in the
capital of the

founding and 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 that 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'13. They maintained federation was necessary to provide
a united

defense and external relations - tasks that confederations and leagues were
less wellequipped

to perform14.

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

13 See Forsyth, Murray, ed. 1989. Federalism and Nationalism. Leicester:
Leicester University Press., p. 4.

14 See Riker, William H. 1964. Federalism: Origin, Operation, Significance.
Boston.

7

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 Protestants15. 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 Pennsylvania16. 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.'17 It would be more precise, however , 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, the famous melting pot of what
Milton

Gordon described as 'Anglo conformity'18. 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'19.

15 See Glazer, Nathan. 1983. Federalism and Ethnicity: The American
Solution. In Ethnic Dilemmas, 1964-82 , edited

by N. Glazer. Cambridge, Mass: Harvard University Press.

16 Gordon, Milton. 1964. Assimilation in American Life: The Role of Race,
Religion and National Origins . New York:

Oxford University Press., p. 133.

17 Glazer, op. cit., p. 276.

18 Gordon, op. cit., passim.

19 Publius [John Jay], in Madison, James, Alexander Hamilton, and John Jay.
1987 (1788). The Federalist Papers,

edited and with an introduction by Isaac Kramnick, Penguin Classics.
Harmondsworth: Penguin., Paper II, p. 91

8

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 - 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 Shi'a 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 Shi'a 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'20; just as he would later engage in
repressive massacres

of largely Shi'ite Marsh Arabs.

It is essential to make these elementary comparisons in 'historical starting
points' to

appreciate the inappropriateness of American national federation as a model
for Iraq's

future. There is no equivalent to a sufficiently homogeneous founding
people, blessed by

Providence or not. The Shi'a have the potential to be a Staatsvolk in Iraq,
that is a

dominant people in control of the state, but fully mobilized they are
unlikely to constitute

much more than sixty per cent of Iraq's electorate21; between them Sunni
Arabs, Kurds

20 McDowall, David. 2000. A Modern History of the Kurds . 2nd and revised
ed. London: I.B. Tauris. , 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.

21 Henry Kissinger, America's foremost 'realist' , as late as January 2002,
managed to write in The Washington Post

('Phase II and Iraq', January 13, Page B07) of 'the Sunni majority that now
dominates Iraq' - an example of 'magical

realism' according to Smyth, Frank. 2003. Saddam's Real Opponents. In The
Iraq War Reader: History, Documents,

Opinions, edited by M. L. Sifry and C. Cerf. New York: Simon & Schuster.,
pp. 565-7. In fact, many recent

Washington administrations have tacitly backed Sunni minority dominance in
Iraq because of their fears that the Shi'a

may be carriers of fundmantalism, Khomeini-style. The Shi'a of Iraq are
mostly Arab, but there are a small number of

Shi'ite Kurds, Turkomans and Arabized Persians (Fuller, Graham E, and Rend
Rahim Francke. 1999. The Arab Shi'a:

9

and others have the power to resist any strongly assimilationist project
that Shi'ites might

attempt, and blocking such a project would unify them; and it is unlikely
that the Allied

occupiers would want to oversee the entirety of Iraq under a Shi'ite
hegemony before

their departure.

But, it is not only the demographic, ethnic and religious differentiation of
Iraq's

population which constrains the ambitions of those who might advocate a
national

federation; its demographic distribution, if data on this matter can be
trusted, also tells

against this idea (see Map 1).

Map 1.

It is, of course, feasible to have many regions in which Shi'a would be the
local majority.

Indeed it is probably not feasible to design contiguous regional boundaries
that would

not make the Shi'a dominant in many of them. But, it would be
extraordinarily difficult,

foolish 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 the addition of significant non-Kurdish population
and territories

into each new unit. Such design principles would inevitably return the Kurds
to armed

conflict with the rest of Iraq. No nationally mobilized people in recent
history has

voluntarily accepted or peacefully acquiesced in the partition of its
homeland - and we

must recall that the Kurds have just fought to regain control over their
homeland. As for

Sunni Arabs, there is little doubt that one reason why the Ba'athists remain
partly

embedded among them is the widespread fear amongst them that Shi'a 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.

The Forgotten Muslims. New York: St. Martin's Press., p. 87). If exiled
Iraqi Shi'a return disproportionately their

demographic weight may increase.

10

and others have the power to resist any strongly assimilationist project
that Shi'ites might

attempt, and blocking such a project would unify them; and it is unlikely
that the Allied

occupiers would want to oversee the entirety of Iraq under a Shi'ite
hegemony before

their departure.

But, it is not only the demographic, ethnic and religious differentiation of
Iraq's

population which constrains the ambitions of those who might advocate a
national

federation; its demographic distribution, if data on this matter can be
trusted, also tells

against this idea (see Map 1).

Map 1.

It is, of course, feasible to have many regions in which Shi'a would be the
local majority.

Indeed it is probably not feasible to design regional boundaries that would
not make the

Shi'a dominant in many of them.22 But, it would be extraordinarily
difficult, foolish 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

the addition of significant non-Kurdish population and territories into each
new unit.

Such design principles would inevitably return the Kurds to armed conflict
with the rest

of Iraq. No nationally mobilized people in recent history has voluntarily
accepted or

peacefully acquiesced in the partition of its homeland - and we must recall
that the Kurds

have just fought to regain control over their homeland. As for Sunni Arabs,
there is little

doubt that one reason why the Ba'athists remain partly embedded among them
is the

widespread fear amongst them that Shi'a will create a majoritarian
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.

22 Non-contiguous regions might have to be designed or there might be too
few regions for a system-wide federation

(see below).

11

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 Iraq23.

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 control24. This thinking derives from a venerable
tradition that goes

back to James Madison and in our times is articulated by Donald L. Horowitz.
The

underlying belief is that a federation should be built on balance of power
principles -

proliferating the points of power away from a focal center, encouraging
intra-ethnic or

intra-religious competition and creating incentives for inter-group
co-operation - by

designing regions without ascriptive majorities25. There is nothing wrong in
principle

with advocating this design, but it has no prospect of success in Iraq. To
design or redraw

regional borders along these lines would require the services of the armed
forces of

the Allied occupiers or future UN forces. This thinking is a non-starter
with Kurds,

because Iraq is not one nation. 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.

****

Advocates of multi-national federations have a different goal: they seek 'to
unite people

who seek the advantages of a common political unit, but differ markedly in
descent,

language and culture'26. They seek to recognize, express and
institutionalize at least two

23 The US's most influential foreign policy journal published a scenario for
re-building Iraq in the summer of 2003,

Dawisha, Adeed, and Karen Dawisha. 2003. How to Build a Democratic Iraq.
Foreign Affairs 82 (3):36-50. 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.

24 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.

25 Horowitz, Donald L. 1985. Ethnic Groups in Conflict. Berkeley: University
of California Press., chapters 14 & 15.

26 Forsyth, op. cit., p. 4.

12

national cultures, on a durable, and often on a permanent basis.
Multi-national federations

involve the maintenance of two or more nations, and reject the strongly
integrationist and

assimilationist dispositions of national federalists. Multi-national
federalists believe that

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. 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 been advocated within both liberal and Marxist
traditions,

and has a significant following within the Anglophone academy27, 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'28.
Multi-national and multiethnic

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 Caribbean29. In the Arabic
world the

27 See inter alia Hechter, Michael. 2000. Containing Nationalism. Oxford:
Oxford University Press, Keating, Michael.

2001. Managing the Multinational State: Constitutional Settlement in the
United Kingdom, edited by T. Salmon and M.

Keating, Linz, Juan. 1997. Democracy, Multinationalism and Federalism. Paper
read at Juan March Institute, Moore,

Margaret. 2001. The Ethics of Nationalism. Oxford: Oxford University Press,
Stepan, Alfred. 1998. Modern

Multinational Democracies: Transcending a Gellnerian Oxymoron. In The State
of the Nation : Ernest Gellner and the

Theory of Nationalism, edited by J. A. Hall. New York: Cambridge University
Press, Stepan, Alfred. 1999. Federalism

and Democracy: Beyond the U.S. Model. Journal of Democracy 10 (4):19-34.

28 Snyder, Jack. 2000. From Voting to Violence: Democratization and
Nationalist Conflict . New York: W.W. Norton. ,

p. 327.

29 O'Leary, Brendan. 2001. 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

(3):273-96.. For other discussions see Franck, Thomas M. 1968. Why
Federations Fail: An Inquiry into the Requisites

for Successful Federation. New York: New York University Press, Hicks,
Ursula K. 1978. Federalism, Failure and

Success: A Comparative Study. London: Macmillan, Watts, Ronald L. 1971. The
Survival or Disintegration of

Federations. In Options for a New Canada, edited by R. Watts and D. M.
Brown. Toronto: University of Toronto Press,

Watts, Ronald L. 1996. Comparing Federal Systems in the 1990s. Kingston,
Ontario: Institute of Intergovernmental

Relations/Queen's University.

13

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

would highlight five key elements that have facilitated the breakdown of
multi-national

federations30:

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.

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 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 -for the Kurds that means that they must have a referendum
in their

own unit to endorse any freely negotiated constitution. A foundational act
of co-operation

30 McGarry, John, and Brendan O'Leary. 2003. Federalism, Conflict-Regulation
and National and Ethnic Power-

Sharing. Paper read at Annual Meeting of the American Political Science
Association, August 28-31 2003, at

Philadelphia., August 28-31..

14

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 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, a
subject that

John McGarry will address in the next panel. 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.

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. 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

15

but judging by the record of the twentieth century the presence of one of
these

conditions is a necessary condition of enduring federations31.

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 may
serve to

check systematic transgressions against such communities. Federations that

protect collective identities help make the respective communities feel
secure -

and in consequence may facilitate the emergence of inter-ethnic and
interreligious

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, Shi'a 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
amongst

them; they have intra-religious cleavages; and they have class differences.
Vigorous Shi'a

majoritarianism would guarantee a prolonged Sunni Arab resistance that would
not just

31 O'Leary, 2001, op.cit.

16

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'ia can comprise a
Staatsvolk our

argument suggests that power-sharing at the 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 what political scientists call 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 communities32. This
evidence strongly

suggests that Iraq needs an executive that is cross-community and
cross-regional in

character. Unlike some, I take the nine-member collective presidency of the
Governing

Council as 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
Shi'a

dominated regions - would necessarily have a cross-regional and
cross-community

character - and would not require any 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. The collective presidency might have responsibility for
a

circumscribed set of affairs - principally defense, foreign affairs, federal
economic

management, and the tasks of the head of state might be rotated and divided
amongst two

members. Swiss thinking, it seems to me, is to be commended rather than the
restoration

32 Adeney, Katharine. 2002. Constitutional Centring: Nation Formation and
Consociational Federalism in India and

Pakistan. Journal of Commonwealth & Comparative Politics 40 (3):8-33.

17

of the monarchy, as suggested by the Dawishas33. The collective presidency
should be

indirectly elected, like the Swiss, but to emphasize its federal character
it should be

elected by each of the delegations of regional senators, perhaps using the
alternative vote

within each delegation. It would be sensible to require a collective
presidency to appoint

ministers from parties in proportion to their strengths in the federal lower
chamber, using

a device such as the d'Hondt rule for the proportional allocation of
portfolios: this would

ensure a proportional and inclusive executive, removing the need for
protracted

bargaining over the distribution of posts in coalition governments. Under
this model the

largest party would have an entitlement to the first choice of ministerial
portfolio, the

next largest the second choice and so on. The ministers would be held to
account both by

the collective presidency and by the federal lower chamber. Measures to
ensure that

federal bureaucrats, 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
willingness of

the Bush administration's national security advisor, Condoleezza Rice, to
encourage the

deployment of Turkish troops in Iraq suggests insensitivity on these
matters: it has

already opened tensions between Iraq's new foreign minister, Hoshyar Zebari,
and

Ahmad Chalabi, this month's President of the new Governing Council34.

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. It
would be my

judgment, which I shall not elaborate here, that the lower chamber of the
federal

33 Dawisha and Dawisha, op. cit., p. 42. Their advocacy of the restoration
of the Hashemites as reassurance for the

Sunni seems naÔve; it forgets that what is reassurance for the Sunni is not
reassurance for the Shi'a, and ignores the

secular republican and republican Islamist dispositions of both Kurds and
Shi'ia. While republics in the Arab world

have proven undemocratic so far, it remains the case that monarchies are
nowhere limited and properly

constitutionalized among the eight kingdoms (Jordan, Saudia Arabia, Morocco,
Oman, Kuwait and the Gulf emirates).

For an elegant essay on monarchies in the Arab world see Halliday, Fred.
2000. The Fates of Monarchy in the Middle

East. In Nation and Religion in the Middle East. Colorado: Lynne Reinner.

34 Brian Knowlton, 'U.S. aid offer prods Turkey on troops for Iraq',
International Herald Tribune , September 10,

2003, p. 3.

18

parliament should be elected by party-list proportional representation, with
each region

having as many members of parliament as their population warrants, and that
the second

chamber should be elected by the single-transferable vote, with each region
electing

twenty senators on a party basis. It would be my judgment, also not
elaborated here, that

it would be symbolically appropriate to distribute the head offices of the
executive,

legislature and courts in different regions, and likewise the offices of
federal ministries.

But these are details I do not have time to address here. What I insist on
is the possibility

of a democratic federation in Iraq, but if, and only if, that future Iraqi
federation is binational,

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

Shi'a and Sunni divide, their internal divisions, and the non-Muslim
religions, as well as

those who have no religion. This, in my judgment, 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

funding of education might also resolve many possible religious sources of
conflict. But,

it is the multi-regional nature of a successful federation on which I wish
to focus some

discussion.

What should be the number of units in an Iraqi federation? That is, of
course, up to the

peoples of Iraq, and their negotiators. But three distinct regions have to
exist on the logic

of the foregoing arguments: Kurdistan, and at least one region dominated by
Sunni

Arabs. Those are the first two. Greater Baghdad, where up to a fifth of the
population of

Iraq is believed to exist, must also constitute a region, as it would be
very difficult to

carve it up amongst other regions. And, it is so large that it would be
grossly

undemocratic to have political impotence imposed on it in the manner of
Washington DC

- which has no vote in either the House of Representatives or the Senate. It
follows that if

Baghdad is not to be too large in relation to other entities in the
federation, and thereby

19

wield too much power, that an Iraqi federation should have five regions if
they are to be

of roughly equally-sized in population. A re-configuration of the South,
West and

Baghdad into three regions, in which the Shi'ia would likely be
preponderant, would

make the construction of a federal collective presidency from the total of
five regions

easy, ensure proportionate weight for each major and minor community -
without

requiring each community to be organized as one bloc or bloc of parties, and
still provide

incentives for parties to compete - and co-operate - throughout the
federation. Two- and

three-unit federations have a poor track-record, and that is another reason
why to this

outsider a five-unit federation seems plausible. Canada and Australia with
populations in

the same range as Iraq have ten and six unit federations respectively - and
both have

some very small units, such as Tasmania and Prince Edward Island. Nothing,
of course,

should preclude each of the five regions from having extensive local
governments. This

type of regional design, it seems to me, is one that might flow relatively
easily from the

known preferences and dispositions of the likely negotiators of the
constitution, but that

remains to be seen.

*****

Lastly, let me take up a question of special interest to Kurds, the notion
of 'federacy'.

Many Kurds have been programmatically committed, for a long time, to a
confederal or

federal Iraq. Kurds equally 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
multiregional

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

20

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.

3. To insist on default mechanisms that would protect Kurdistan in the event
of

breaches of any new Iraqi constitution.

But Kurds will also 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' agreement35.
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 they would probably seek
looser powersharing

arrangements in the central government, especially in foreign relations,
while

seeking to protect the cultural and human rights of Kurds outside Kurdistan.

The Allied Coalition's Chief Administrator in Iraq has said that the writing
of Iraq's new

constitution will be the fourth step in Iraq's return to sovereignty. He has
indicated that

35 See footnote 3 above.

21

cannot be done in 'days or weeks'36. That is so. The preparatory committee
on how the

process of making the constitution should be written is supposed to report
by the end of

this month. We all await its proposals with interest. Thank you for
listening to me.

References in the Footnotes

Adeney, Katharine. 2002. Constitutional Centring: Nation Formation and
Consociational

Federalism in India and Pakistan. Journal of Commonwealth & Comparative

Politics 40 (3):8-33.

Dawisha, Adeed, and Karen Dawisha. 2003. How to Build a Democratic Iraq.
Foreign

Affairs 82 (3):36-50.

Elazar, Daniel. 1987. Exploring Federalism. Tuscaloosa: University of
Alabama.

Forsyth, Murray, ed. 1989. Federalism and Nationalism. Leicester: Leicester
University

Press.

Franck, Thomas M. 1968. Why Federations Fail: An Inquiry into the Requisites
for

Successful Federation. New York: New York University Press.

Fuller, Graham E, and Rend Rahim Francke. 1999. The Arab Shi'a: The
Forgotten

Muslims. New York: St. Martin's Press.

Gardner, David. 2003. Time of the Shia. Financial Times, August 30/31,
W1-W2.

Glazer, Nathan. 1983. Federalism and Ethnicity: The American Solution. In
Ethnic

Dilemmas, 1964-82, edited by N. Glazer. Cambridge, Mass: Harvard University

Press.

Gordon, Milton. 1964. Assimilation in American Life: The Role of Race,
Religion and

National Origins. New York: Oxford University Press.

Halliday, Fred. 2000. The Fates of Monarchy in the Middle East. In Nation
and Religion

in the Middle East. Colorado: Lynne Reinner.

Hechter, Michael. 2000. Containing Nationalism. Oxford: Oxford University
Press.

Hicks, Ursula K. 1978. Federalism, Failure and Success: A Comparative Study.
London:

Macmillan.

Horowitz, Donald L. 1985. Ethnic Groups in Conflict. Berkeley: University of
California

Press.

Jabar, Faleh A. 2003. The Shi'ite Movement in Iraq.

Keating, Michael. 2001. Managing the Multinational State: Constitutional
Settlement in

the United Kingdom, edited by T. Salmon and M. Keating.

King, Preston. 1982. Federalism and Federation. London: Croom Helm.

36 L. Paul Bremer III, 'Iraq's Path to Sovereignty', Washington Post,
September 8th 2003, p. A21.

22

King, Preston. 2001. Federalism and Federation. 2nd ed. London: Frank Cass.

Lijphart, Arend. 1979. Consociation and Federation: Conceptual and Empirical
Links.

Canadian Journal of Political Science xii (3):495-515.

Linz, Juan. 1997. Democracy, Multinationalism and Federalism. Paper read at
Juan

March Institute.

Madison, James, Alexander Hamilton, and John Jay. 1987 (1788). The
Federalist Papers,

edited and with an introduction by Isaac Kramnick, Penguin Classics.

Harmondsworth: Penguin.

McDowall, David. 2000. A Modern History of the Kurds. 2nd and revised ed.
London:

I.B. Tauris.

McGarry, John, and Brendan O'Leary. 2003. Federalism, Conflict-Regulation
and

National and Ethnic Power-Sharing. Paper read at Annual Meeting of the

American Political Science Association, August 28-31 2003, at Philadelphia.

Moore, Margaret. 2001. The Ethics of Nationalism. Oxford: Oxford University
Press.

Natali, Denise. 2001. Manufacturing Identity and Managing Kurds in Iraq. In
Right-

Sizing the State: the Politics of Moving Borders, edited by B. O'Leary, I.
S.

Lustick and T. Callaghy. Oxford: Oxford University Press.

O'Leary, Brendan. 2001. 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 (3):273-96.

O'Leary, Brendan. 2003. The Kurds Must Not Be Betrayed Again. Financial
Times,

March 24th, 23.

Rezvani, David R. 2003. Federacy: The Dynamics of Semi-Sovereign Territories

(unpublished manuscript).

Riker, William H. 1964. Federalism: Origin, Operation, Significance. Boston.

Smith, Rogers M. 2003. Stories of Peoplehood: the Politics and Morals of
Political

Membership. Cambridge: Cambridge University Press.

Smyth, Frank. 2003. Saddam's Real Opponents. In The Iraq War Reader:
History,

Documents, Opinions, edited by M. L. Sifry and C. Cerf. New York: Simon &

Schuster.

Snyder, Jack. 2000. From Voting to Violence: Democratization and Nationalist
Conflict.

New York: W.W. Norton.

Stepan, Alfred. 1998. Modern Multinational Democracies: Transcending a
Gellnerian

Oxymoron. In The State of the Nation : Ernest Gellner and the Theory of

Nationalism, edited by J. A. Hall. New York: Cambridge University Press.

Stepan, Alfred. 1999. Federalism and Democracy: Beyond the U.S. Model.
Journal of

Democracy 10 (4):19-34.

Watts, Ronald A. 1998a. Federalism, Federal Political Systems, and
Federations. Annual

Review of Political Science 1:117-37.

Watts, Ronald L. 1971. The Survival or Disintegration of Federations. In
Options for a

New Canada, edited by R. Watts and D. M. Brown. Toronto: University of

Toronto Press.

Watts, Ronald L. 1996. Comparing Federal Systems in the 1990s. Kingston,
Ontario:

Institute of Intergovernmental Relations/Queen's University.

Watts, Ronald L. 1998b. Federalism, Federal Political Systems, and
Federations. Annual

Review of Political Science 1:117-37.

23

Watts, Ronald L. 2001. Models of Federal Power-Sharing. International Social
Science

Journal (March):23-32.



_______________________________________________
Sent via the discussion list of the Campaign Against Sanctions on Iraq.
To unsubscribe, visit http://lists.casi.org.uk/mailman/listinfo/casi-discuss
To contact the list manager, email casi-discuss-admin@lists.casi.org.uk
All postings are archived on CASI's website: http://www.casi.org.uk


[Campaign Against Sanctions on Iraq Homepage]