THE CRACKED FOUNDATIONS

OF THE RIGHT TO SECEDE

Donald L. Horowitz

Donald L. Horowitz is James B. Duke Professor of Law and Political

Science at Duke University, and author, most recently, of The Deadly
Ethnic Riot (2001). He is currently working on a study of constitutional
design for divided societies. The present essay is based on a paper presented at a conference at Boston University, funded by the Smith
Richardson Foundation, on Partitions as a Solution to Ethnic Conflict.

Many things changed with the end of the Cold War. One of the main
changes is the new enthusiasm for public international law. In a world
freed of an overwhelming rivalry of two great power blocs, but by no
means freed of conflict, it became possible to think of a qualitatively
enhanced role for the rule of law in the regulation of international relations. Nowhere has this thinking been more evident than in the regulation
of conflict and warfare between ethnic groups. Two main prongs of
international legal activity are the establishment of tribunals to punish
genocide and crimes against humanity and the elaboration of various
doctrines of human rights, including a possible right of ethnic groups to
secede from states in which they are located. This asserted right, which
would be a reversal of existing international law, is, in my view, ill
considered.
The newly asserted right to secede is to be held by ethnic groups and
is derived from a reinterpretation of the principle of the self-determination of nations. Theorists display varying degrees of enthusiasm in their
advocacy of such a right, but all of them, whether they would permit
secession generously or only reluctantly after certain conditions have
been fulfilled, see secession as an answer to problems of ethnic conflict
and violence. The position I shall take here is that secession is almost
never an answer to such problems and that it is likely to make them
worse. The proposals, in short, are not informed by any serious understanding of patterns of ethnic conflict or ethnic-group political behavior.
Secession, I shall argue, does not create the homogeneous successor
Journal of Democracy Volume 14, Number 2 April 2003

Journal of Democracy

states its proponents often assume will be created. Nor does secession
reduce conflict, violence, or minority oppression once successor states
are established. Guarantees of minority protection in secessionist regions are likely to be illusory; indeed, many secessionist movements
have as one of their aims the expulsion or subordination of minorities in
the secessionist regions. The very existence of a right to secede, moreover, is likely to dampen efforts at coexistence in the undivided state,
including the adoption of federalism or regional autonomy, which might
alleviate some of the grievances of putatively secessionist minorities.
Since most secessionist movements will be resisted by central governments and most secessionists receive insufficient foreign military
assistance to succeed, propounding a right to secede, without the means
to success, is likely to increase ultimately fruitless secessionist warfare,
at the expense of internal efforts at political accommodation and at the
cost of increased human suffering. Efforts to improve the condition of
minorities ought to be directed at devising institutions to increase their
satisfaction in existing states, rather than encouraging them to think in
terms of exit options. In those rare cases in which separation of antagonists is, at the end of the day, the best course, partition can be
accomplished reluctantly, as a matter of prudence, without recognizing
a right to secede. But neither partition nor secession should be viewed
as generally desirable solutions to the problems of ethnic conflict.

Secession and Self-Determination: A Brief Sketch

The so-called right to secede has its origins in the principle of national self-determination. As is well known, that principle formed an
integral part of Woodrow Wilsons plans for post-World War I Europe.
The establishment of, for example, Rumania as a state for Rumanians
certainly exemplified application of the self-determination principle, but,
even apart from the presence of minorities in such new states, the
Wilsonian policy fell far short of according national or ethnic groups
their own states. Indeed, Wilson may well have envisioned autonomy
rather than independence, and he did not necessarily think in terms of
an ethnic fulfillment for the right to self-determination. Yugoslavia and
Czechoslovakia, for example, were multinational states, and Wilsons
proposal that a right of self-determination be incorporated in the Covenant of the League of Nations was rejected. By the time of the Atlantic
Charter in 1941, self-determination was to be limited to peoples living
under foreign domination.1 Decolonization was thus an exercise of selfdetermination, and it was soon made clear by the United Nations (in
1960) and the Organization of African Unity (in 1964), among others,
that secessionist threats to the territorial integrity of states would not be
regarded as further exercises of self-determination.
With the exception of decolonization and the extraordinary emergence

Donald L. Horowitz

of Bangladesh, territorial boundaries proved to be remarkably stable for

a half century: from just after World War II to the end of the Cold War.2
And then a concatenation of eventsthe reunification of Germany, the
dissolution of the Soviet Union (and various sub-secessions in Georgia,
Moldova, and Azerbaijan), Yugoslavia, and Czechoslovakia, the secession of Eritrea from Ethiopia and of the former Somaliland from Somalia,
and finally the de facto detachment of Kosovo from Serbiacombined
in the course of a decade to render boundaries much less stable and to
encourage territorially separate groups to consider the possibility of secession.
As all this was happening, international lawyers and philosophers
had been rethinking the meaning of self-determination and proceeding
to unsettle the former understandings that had discouraged secession
and international support for it. One practical exercise along these lines
was the ill-considered decision of the Badinter Commission that, asked
to pronounce on the validity of secessions from Yugoslavia, declared
Yugoslavia to be a federation in process of dissolutiona state of affairs that entitled its constituent republics to secede intact, taking with
them their minority areas.3 After this decision, many fewer states than
should consider adopting a federal form of government will do so.
Most of the work of international lawyers has been, however, theoretical. It proceeds along several lines.
For some, self-determination forms an integral part of the right of
people to choose their own political regime and to be free of authoritarian oppression. (For Wilson, too, self-determination was connected to
ideas of popular sovereignty.) It is a building block of what was emerging in the thinking of a few international lawyers as a right to live under
a democratic regime.4 That such a legal right is at best aspirational might
be suggested by the fact that more than half of all regimes in the world
are still not democratic, but the early aftermath of the Cold War was a
time of great optimism.
Equally expansive are justifications of secession that rest on the right
of people in general or people with common group characteristics to
choose those with whom they wish to associate politically. 5 Among
philosophers who have endorsed the right of ethnic groups to secede,
most have begun from the premise that self-determination is to ethnic
groups what moral autonomy is to individuals. As individuals have rights
to political self-expression, so, too, do groups; and for groups, selfexpression means self-government, which, in the era of the territorial
state, implies control of territory.6
The analogy of collective self-determination to individual autonomy
is entirely specious. For one thing, collective identity fluctuates, as individual identity does not. New groups come into being, and old ones merge
or divide. No new political entity can solve the problem of collective
self-expression, as Pakistanis, divided in 1947, redivided in 1971, and

Journal of Democracy

still dividing further, have long since discovered. Moreover, even at the
outset, the minorities problem is not a trivial detail in recognizing secessions; it is, instead, the central reason why the idea of a clean break is a
chimera.
A more cautious version of the right to secede is espoused by those
who view secession principally as a remedial right,7 a last-ditch response to discrimination or oppression by a central government.8 If
interethnic accommodation fails and one portion of a population is unalterably hostile to a group of its fellow citizens, then, it is said, republican
theories may support political divorce on the ground that separation would
produce two states in which republican democracy is viable instead of
one that lacks essential preconditions to its successful practice.9
Despite important differences in scope and reasoning among these
justifications for secession, there is a substratum of assumptions in all
of them. Secession, it is assumed, can produce homogeneous successor
states. In those cases in which heterogeneity remains, it is asserted,
minority rights can nevertheless be guaranteed. Like the Badinter Commission, most writers advocating a right to secede make no provision
for further secessions, except, of course, insofar as infinite regress of
secessionist rights may be implied in their formulations. Secession will
also, it is assumed, result in a diminution of conflict that produced the
secessionist movement. Rarely are these assumptions discussed or even
rendered explicit, but they are essential to the analysis.
If you can think about something which is attached to something else
without thinking about what it is attached to, then you have what is called
a legal mind.10 So pronounced the late constitutional lawyer, Thomas
Reed Powell, three-quarters of a century ago. Most theorists of a right to
secession have, in this caricatured sense, legal minds. They have generally not concerned themselves with the ethnic politics that produces
secessionist claims and that will be affected by new rights to secede. It is
no accident that most people who do study ethnic politics are decidedly
less enthusiastic about secession than are the international lawyers and
philosophers who are the main proponents of a right to secede.

Heterogeneity Before, Heterogeneity After

There are always ethnic minorities in secessionist regions. There were
Efik and Ijaw, among others, in Biafra; there are Hindus in Kashmir,
Muslims in Tamil areas of Sri Lanka, Javanese in Aceh and Irian Jaya,
Serbs and Roma in Kosovo; and there are minorities in all the rump
states as well. As a matter of fact, it is often the desire of regional majorities to deal with minoritiesand not to deal with them in a democratic
waythat motivates or contributes to the secessionist movement in the
first instance. Proponents of rights to secession assure us that minority
rights must be guaranteed in secessionist states and that secession should

Donald L. Horowitz

be less favored if minority rights are unlikely to be respected,11 but the

verbal facility of this formulation masks the difficulty of achieving any
such results. If, after all, conditions on the exercise of an international
law right to secede can be enforced, why not enforce those conditions in
the undivided state so as to forestall the need to secede? International
law has been notoriously ineffective in assuring longstanding, internationally recognized minority rights, and proponents of secession have
no new ideas to offer on this matter. If the failure to respect minority
rights in the undivided state induced a regional group to consider secession, why should anyone assume that the situation will be different when
that group, a minority in the undivided state, comprises a majority in the
secessionist state? If anything, the treatment of minorities in smaller
states is less visible to outsiders.
The more circumscribed the asserted right to secede, ironically
enough, the more dangerous conditions may become for minorities in
the secessionist region. By the time it is concluded that the majority in
the undivided state is unalterably hostile to minority interests, thus in
some formulations permitting the minority to secede, that group may
have accumulated so many grudges that, in their turn, minorities in the
secessionist region may be particularly vulnerable to the expression of
violent hostility or the settlement of old scores. There are many examples:
the fate of Serbs and Roma in Kosovo, of Biharis in Bangladesh, of
Sikhs and Hindus in Pakistan at the time of partition, and of Muslims in
India at the same time. If the problem of minorities is that they do not
enjoy meaningful political participation12 in the undivided state, there
is no reason to think that minorities will enjoy it in the secessionist state
either. Secession merely proliferates the arenas in which the problem of
intergroup political accommodation must be facedand often more
starkly. Contrast Yugoslavia, with six or seven groups and the complex
alignments they created with Bosnia, in which three groups confront
each other. Secession can hardly be said to solve the problem of intergroup accommodation, except, of course, insofar as it enables the former
minority, now a new majority, to cleanse the secessionist state of its
minoritieswhich it could not do previouslyand induces the rump
state to do the same with members of the secessionist group who find
themselves left on the wrong side of a new international boundary.

The Effect of New International Boundaries

Recognition of a right to secede is thus not likely to be the end of an
old bitterness but the beginning of new bitterness. It is, of course, easy
to question whether a slavish devotion to territorial integrity is still appropriate. There has been a great deal of loose talk about the allegedly
artificial character of many international boundaries and the part played
by colonial convenience in settling them. In point of fact, even in

10

Journal of Democracy

Africa, where this charge is most frequently encountered, boundaries

were not settled as disrespectfully of ethnic patterns as is frequently
asserted.13 In any event, patterns of settlement are such that virtually
any boundaries would have a large element of arbitrariness to them.
Secession would not be a way of rectifying boundaries, because there
are no truly natural boundaries.
If it does not solve boundary problems, secession does do something
else. A secession or partition converts a domestic ethnic dispute into a
more dangerous international one. And since states are able to procure
arms with few of the restraints that periodically bedevil insurgents, the
international dispute often involves escalating weapons and the prospect of international warfare. Consider the nuclear armaments possessed
by India and Pakistan and the recurrent warfare between those states.
One reason for the greater danger that often follows secession is the
activation of irredentist claims. For reasons I have explicated elsewhere,14
the serious pursuit of irredentasmovements to retrieve kindred people
and their territory across international boundarieshas been relatively
rare in the post-World War II world. But successful secession or partition is likely to change this benign state of affairs. Either the rump state
or the secessionist state will desire to retrieve minorities stranded on
the wrong side of the border. There are examples readily at hand: Kashmir, Serb claims in Bosnia and in the Krajina region of Croatia, warfare
between Ethiopia and Eritrea. And when irredentism gets going, it usually involves ethnic cleansing, so as to eliminate troublesome minorities
in the region to be retrieved. A recent quantitative study of the effects
of partition finds that partition does not prevent further warfare between
ethnic antagonists, and it has only a negligible (and easily reversed)
positive effect on low-grade violence short of war.15
The recurrent temptation to create a multitude of homogeneous ministates, even if it could be realized, might well increase the sum total of
warfare, rather than reduce it. The right direction for international boundaries is upward, not downward, so that states are so heterogeneous that
no one group can plausibly dominate others.16 Although this degree of
benign ethnic complexity is exceedingly difficult to achieve, it is still
true that India, with its many groups, is a better model than Kosovo or
Rwanda, with just two or three.

Secession Rights and Internal Accommodation

Articulating a right to secede will undermine attempts to achieve interethnic accommodation within states. As things now stand, the principal
reason that states are reluctant to devolve power to territorially concentrated minorities, either by means of regional autonomy or federalism,
is their fear that it will lead to secession. That fear is usually unfounded,
unless the conflict has already dragged on for a long time and the central

Donald L. Horowitz

11

government has been utterly ungenerous. Nevertheless, central governments are risk averse about devolution. The best way to dry up devolution
as a tool of interethnic accommodationand a promising tool it isis
to establish a right, recognized in international law, for territorially concentrated minorities to secede. If there is a well-recognized right to
secede, the first stirrings of territorially based ethnic discontent will be
likely to be met with repression. The possibility that federalism or regional autonomy can lawfully ripen into secession will make any such
experiment too costly to entertain. It has been difficult to persuade central decision makers in Indonesia and Sri Lanka to devolve power to
regions. A right to secession would easily dissuade them.
One reason central governments are so reluctant to countenance the
possibility of secession, even for troublesome regions that some central
decision makers might wish to be free of, is that the secession of one
region upsets ethnic balances and forces groups in other regions to think
afresh about whether they wish to remain in the truncated state with its
new ethnic balances. This was clearly visible in Yugoslavia after the
Slovene and Croat decisions to secede, when others had to decide in
turn whether the relative expansion in Serb power in the rump state was
in their interest. Yoruba narrowly decided to stay in Nigeria, despite the
relative increase in Hausa power when Ibo decided to leave the state in
1967, and the departure of East Bengal (Bangladesh) from Pakistan
destabilized relations among the groups that remained in the rump state.
Quite often the fears of central authorities about secession are derided
as unsubstantiated apprehensions of domino effects. But domino effects
are usually conceived as action based merely on a successful example
in another location, whereas what is involved in the first secession is
action that affects directly, rather than just by example, the relative positions of other groups remaining in the state.
The creation of a right to secede could not be more untimely. More
and more states have been designing internal political arrangements,
including devolution, to reduce the incidence of ethnic conflict. That is
where the emphasis needs to be, not on making exit strategies more
plausible. More about this shortly.
A right to secede effectively advantages militant members of ethnic
groups at the expense of conciliators. Since most central governments
will not recognize the right to secede, those who wish to pursue such a
course will need to resort to arms. Those who are willing to resort to
arms are by no means simply latter-day versions of the politicians of
their own group whom they seek to displace. Contrast Hashim Thaci of
the Kosovo Liberation Army with the Kosovar political leader Ibrahim
Rugova; Prabhakaran of the Tamil Tigers with Amirthalingham of the
Tamil United Liberation Front, whom he had assassinated; the Southern Peoples Liberation Army in the Sudan with the old Liberal Party
that preceded it. Violence disproportionately attracts people with

12

Journal of Democracy

an interest in aggression. The people willing to take up arms for secession are those who are willing to be brutal with their ethnic enemies and
with their own rivals as well. As their advantage grows, new bouts of
ethnic cleansing can be expected.
In some formulations, secession is said to be an exceptional right
that comes into play if it is the only way that a defined population
can exercise its right of internal self-determination.17 But the facts do
not support the assumption that secession is ever the only way. Are
the Kurds in Iraq secessionist or autonomist? They have gone back and
forth. Are Philippine Muslims? They, too, have gone back and forth.
Ibo tried unsuccessfully to secede and then reintegrated into Nigerian
politics. In such cases, it looks to outsiders at any given moment as if
secession is the only way minorities can participate in determining
their own future, but there is more fluidity to ethnic politics than those
who write about populations that are unalterably hostile18 to each other
have sensed.
Moreover, the seemingly moderate position of some proponents of a
right to secession that secession is justified only if others are unalterably opposed or minorities have been victimized is not likely to work
out moderately in practice, for it is an incentive to ethnic polarization.
If independence can only be won legitimately after matters have been
carried to extremes, then, by all means, there are people willing to carry
them to extremes. In the 1980s and early 1990s, Sikh separatists in the
Indian Punjab were willing to attack Hindus in order to precipitate attacks on Sikhs elsewhere in India. There is no shortage of methods to
satisfy tough standards of victimization or oppression. A right to secede could indeed contribute to the sense that secession is the only way.
There may be times when it is felt best to part peoples. The British
believed such a time had come in India in 1947, and the United Nations
believed such a time had come a year later in Palestine. When it is prudent, parting can be done by consent, as in the former Soviet Union and
in Czechoslovakia, or occasionally by international action. To do this
requires the creation of no rights.
Consider the pernicious effect on the balance of intragroup opinion of
a right to secede in a concrete case: Sri Lanka. Will the Sri Lankan Tamils
return as readily as they would otherwise to a thoroughly reconstructed
but undivided Sri Lanka if they discover that the secession to which they
turned so reluctantly was merely an exercise of their rights under international law? It is always hard for antagonistic groups to accommodate
each other in a single state. A right to secede will make it harder.

The Case for Humility in Public International Law

As I indicated at the outset, rights to secession are only in some formulations grounded in the extreme oppression of minorities. In others,

Donald L. Horowitz

13

rights to secede derive from the alleged commitment of international

law to democracy. In fact, international law has not had a particularly
deep or longstanding commitment to democracy. Only in 1992 did Thomas Franck purport to discover an emerging entitlement to be governed
democratically.19 International law has always placed great emphasis
on the writings of jurists, but traditionally for their syntheses of legal
rules and for their influence on state practice that could then change the
law, not for the direct creation of new rights by cobbling together new
formulations. Such a practice is particularly doubtful when it is recalled
that international law is a field in which judicial and bureaucratic institutions far outrun representative ones.
In 1998, the British jurist James Crawford produced a long and exhaustive survey of state practice and the international law pertaining to
secession.20 Crawfords conclusion was simply stated:
. . . State practice since 1945 shows very clearly the extreme reluctance
of States to recognize or accept unilateral secession outside the colonial
context. That practice has not changed since 1989, despite the emergence
during that period of 22 new States. On the contrary, the practice has
been powerfully reinforced.21

Bangladesh, Crawford noted, was best viewed not as exemplar of a recognized right to secede but rather as a fait accompli achieved as a result
of foreign military assistance in special circumstances.22 Other cases,
such as Eritrea and the Baltic states, involved mutual consent. Where
central governments oppose unilateral secession, Crawford found, the
secessionists gain little or no international recognition. This certainly
has been the case in northern Somalia and Transdniestria, among others. And, finally, there is no recognition of a unilateral right to secede
based merely on a majority vote of the population of a given subdivision or territory. In principle, self-determination for peoples or groups
within the State is to be achieved by participation in its constitutional
system, and on the basis of respect for its territorial integrity. 23
Crawford might have, but did not, note the contrary but truly exceptional position of the government of the United Kingdom, which has
stated, in ways meant to be binding, that a majority of the people of
Northern Ireland might vote to dissolve their union with Britain and to
join the Irish Republic instead. But this has not been the position of other
states, and it is not the position of the United Kingdom with respect to
Scotland or Wales.
If a right to secede is a by-product of an emerging right to democratic
governance, there is not much evidence of it. Subtract the Franck formulation and a somewhat earlier one by Cassese,24 and the dedication of
international law to democratic governance becomes much thinner. What
little there is mainly is confined to writers, rather than custom, state practice, treaties, or court decisions.

Journal of Democracy

14

Why the connection of international law to democratic governance

is so thin should be very clear. However much we may favor the worldwide spread of democracyand I have done my time trying to spread
democratic institutions to reduce ethnic conflictvalorizing a right to
democratic governance would imperil the universality of international
law. Despite the developments of the postCold War period, there are
almost as many authoritarian as democratic states. If international law
is to enhance its influence on state behavior, which it needs to do in
order to reduce and regulate interstate conflict and to facilitate interstate transactions, it cannot simultaneously undercut the governing
arrangements of nearly half the states that are to be subjected to the
rules laid down. 25
There is, of course, an even more obvious reason why international
law ought to be exceedingly restrained in its enthusiasm for secession.
Secession is an anti-state movement, and an international law that forgets that states are its main subjects risks its own survival.
There is always a tendency of law to preempt social complexity with
rules, and there are many temptations to the promiscuous creation of
rights. But law does best when it is informed by what Karl Llewellyn
called situation sense,26 a sound idea of the type of phenomenon it
seeks to govern. If self-determination is a phrase simply loaded with
dynamite, 27 in the words of Wilsons secretary of state, there are international lawyers who are playing with this dynamite.

Living With Heterogeneity

I said earlier that emphasis needs to be on fostering interethnic accommodation within states. The choice between secession or partition,
on the one hand, or murderous conflict, on the other, is a false choice.
Institutions can mitigate conflict. This is much too large a topic to discuss at length here, but, since I have been so critical of those who have
readily endorsed the right of ethnic groups to leave states, it is incumbent
on me at least to sketch briefly what some alternatives might look like.
Most states are ethnically heterogeneous, and many are severely divided. Many groups seek to treat the state as an ethnic patrimony, as if
it were homogeneous or as if they had a prior claim to legitimacy and
others were there merely on sufferance. Why this is so is a complicated
story. Despite these depressing tendencieswhich are highly variable
rather than universalit is long past the time when ethnic kinship could
form the foundation for homogeneous communities. Territorial proximity is now an inescapable basis for political community.
In general terms, there are two competing prescriptions for solving
the problem of ethnic conflict in a democratic framework. Each has its
proponents and detractors, its strong and weak points.
The first prescription goes by the name of consociational democracy:

Donald L. Horowitz

15

a formula for government by grand coalition of all groups, for minority

vetoes on important policy issues, for ethnic proportionality in cabinet
positions, civil-service posts, and financial allocations, and for cultural
autonomy for all groups as well.28 Many criticisms have been made of
consociational theory, for its neglect of democratic opposition (if everyone is in a grand coalition, where will opposition come from?) and
its alleged propensity for excessively limited government and immobilism. My own criticism is that grand coalitions are impossible where
divisions are severe, because the very formation of such a coalition produces opposition based on the accusation that group interests have been
sold out. Moreover, consociation is essentially a system of guarantees,
and so is attractive to minorities, but not to majorities, who prefer majority rule. As a result, there are consociational features adopted
occasionally by states, but few full-blown consociational regimes. Grand
coalitions and minority vetoes are particularly scarce.
My own preferred course involves the use of political incentives to
encourage interethnic moderation. There are various institutions, particularly electoral systems, that are capable of inducing moderate
behavior on the part of politicians. If election depends, at the margin,
on the ability to gain some votes from members of groups other than
ones own, then political leaders will behave in an ethnically conciliatory fashion for that purpose. 29 One thing we know is that politicians
like being elected and reelected. If consociational theory provides no
motive for compromise behavior, incentive theory, by definition, does
not share that defect.
Yet there are obstacles to the adoption of incentives schemes as well.
Although one can find incentive-based devices adopted by states, a full
ensemble of institutions containing incentives to foster conciliation is
not easy to find.30 There is evidence that these devices work, but what
are the incentives to adopt the incentives?
Often, processes of bargaining over institutions produce compromises
that dilute the effect that could have been expected, had there been a
more thoroughgoing and consistent set of institutional changes. Still,
partial adoptions can have some positive effect on conciliation.
The upshot of the problem of adoption is that most severely divided
societies will not soon become dramatically more harmonious. Over
time, some have, and others will, but not necessarily wholly as a result
of political engineering. Political engineering will play its part, a greater
part at certain unusually propitious times (and such moments should be
seized), but the difficulty of wholesale adoption means that many societies will muddle along, sometimes severely conflicted, sometimes better
able to achieve compromise if their partial adoption of conciliatory devices is well considered. But, for present purposes, what needs to be
emphasized is that efforts at conciliation will not be helped by providing either a liberal or a constrained right to secede. There is an inevitable

Journal of Democracy

16

tradeoff between encouraging participation in the undivided state and

legitimating exit from it. The former will inevitably produce imperfect
results, but the latter is downright dangerous.
NOTES
1. See the excellent entry by Michael Banton, Self-Determination, in Athena
Leoussi, ed., Encyclopaedia of Nationalism (New Brunswick, N.J.: Transaction Publishers, 2001), 27173.
2. The so-called secession of Singapore was actually an expulsion of the city-state
from Malaysia and was portrayed in legal instruments as occurring by mutual consent.
3. Opinions of the Arbitration Committee, reported by Alain Pellet, The Opinions
of the Badinter Arbitration Committee: A Second Breath for the Self-Determination of
Peoples, European Journal of International Law 3 (1992): 178, Appendix, at 183.
4. See Thomas M. Franck, The Emerging Right to Democratic Governance, American Journal of International Law 86 (January 1992): 4691. See also Antonio Cassese,
Political Self-Determination: Old Concepts and New Developments, in Antonio
Cassese, ed., UN Law/Fundamental Rights: Two Topics in International Law
(Alphanaanden Rijn, Netherlands: Sijthoff and Noordhoff, 1979), 13765; and Morton
H. Halperin and Kristen Lomasney, Toward a Global Guarantee Clause, Journal of
Democracy 4 (July 1993): 6069.
5. See for example Daniel Philpott, Self-Determination in Practice, in Margaret
Moore, ed., National Self-Determination and Secession (Oxford: Oxford University Press,
1998), 79102.
6. For surveys of such theories, see Donald L. Horowitz, Self-Determination: Politics, Philosophy, and Law, in Margaret Moore, ed., National Self-Determination and
Secession, 197200; and Allen Buchanan and David Golove, The Philosophy of International Law, in Jules Coleman and Scott Shapiro, eds., The Oxford Handbook of
Jurisprudence and Philosophy of Law (Oxford: Oxford University Press, 2002).
7. See Allen Buchanan and David Golove, The Philosophy of International Law.
See also Allen Buchanan, Secession: The Morality of Political Divorce from Fort Sumter
to Lithuania and Quebec (Boulder, Colo.: Westview, 1991); and Diane F. Orentlicher,
International Responses to Separatist Claims, paper presented at the Annual Meeting
of the American Society for Political and Legal Philosophy, San Francisco, 23 January
2001.
8. Robert McCorquodale, Human Rights and Self-Determination, in Mortimer
Sellers, ed., The New World Order: Sovereignty, Human Rights, and the Self-Determination of Peoples (Oxford: Berg, 1996), 2425; and Gerry J. Simpson, The Diffusion
of Sovereignty: Self-Determinations in the Post-Colonial Age, in Mortimer Sellers,
ed., The New World Order, 48, 5456.
9. Diane F. Orentlicher, International Responses to Separatist Claims, 1718.
10. Quoted by Thurman Arnold, Criminal AttemptsThe Rise and Fall of an Abstraction, Yale Law Journal 40 (193031): 58.
11. See for example Daniel Philpott, Self-Determination in Practice, 92; and Diane
F. Orentlicher, International Responses to Separatist Claims.
12. Hurst Hannum, Territorial Autonomy: Permanent Solution or Step Toward
Secession? paper presented at the conference on Facing Ethnic Conflicts: Perspec-

Donald L. Horowitz

17

tives from Research and Policy Making, University of Bonn, 1416 December 2000,
4.
13. See Saadia Touval, The Boundary Politics of Independent Africa (Cambridge:
Harvard University Press, 1972).
14. See Donald L. Horowitz, Ethnic Groups in Conflict (Berkeley: University of
California Press, 2000), 28188; and Donald L. Horowitz, Irredentas and Secessions:
Adjacent Phenomena, Neglected Connections, in Naomi Chazan, ed., Irredentism and
International Politics (Studies in International Politics) (Boulder, Colo.: Lynne Rienner, 1991), 922.
15. Nicholas Sambanis, Partition as a Solution to Ethnic War: An Empirical Critique of the Theoretical Literature, World Politics 52 (July 2000): 43783.
16. See Nicholas Sambanis, Partition as a Solution to Ethnic War, 479; and Donald
L. Horowitz, Ethnic Groups in Conflict, 59296.
17. Diane F. Orentlicher, International Responses to Separatist Claims, 14 (emphasis omitted).
18. Diane F. Orentlicher, International Responses to Separatist Claims, 17.
19. See Thomas M. Franck, The Emerging Right to Democratic Governance.
20. James Crawford, State Practice and International Law in Relation to Secession, British Yearbook of International Law 1998 (Oxford: Clarendon Press, 1999),
85117.
21. James Crawford, State Practice and International Law in Relation to Secession, 114.
22. James Crawford, State Practice and International Law in Relation to Secession, 115.
23. James Crawford, State Practice and International Law in Relation to Secession, 116.
24. Antonio Cassese, Political Self-Determination.
25. The International Court of Justice has itself recognized periodically that customary international law comprises a limited set of norms to assure coexistence and
cooperation among states. See for example Case Concerning Delimitation of the Maritime Boundary in the Gulf of Maine Area, 1984 I.C.J. 246.
26. Karl N. Llewellyn, The Common Law Tradition: Deciding Appeals (Boston: Little,
Brown, 1960).
27. Robert Lansing, quoted in David Fromkin, Kosovo Crossing: American Ideals
Meet Reality on the Balkan Battlefields (New York: Free Press, 1999), 127.
28. See Arend Lijphart, Democracy in Plural Societies: A Comparative Exploration
(New Haven: Yale University Press, 1977).
29. I have written about such devices in various places. See for example Donald L.
Horowitz, A Democratic South Africa? Constitutional Engineering in a Divided Society
(Berkeley: University of California Press, 1991).
30. See Donald L. Horowitz, Constitutional Design: An Oxymoron, NOMOS 42
(2000): 25384.