Inconsistent Irredentism? Political Competition, Ethnic Ties, and the Foreign Policies of Somalia and Serbia

Full text

Since the fall of the Berlin Wall, perhaps the greatest threat to European security has been
irredentism—states seeking to redeem “lost” territories inhabited by ethnic kin. At the heart of
the Yugoslav conflict are dueling irredentisms: Serb and Croatian. While Serbia gets most of the
blame for the conflict, Croatia has also sought to annex neighboring territory in Bosnia inhabited
by Croats. Greece’s foreign policy obsession has been that Macedonia might seek to annex its
northern regions. In addition, Greece may have irredentist designs towards Albania. The
existence of a Hungarian diaspora has fed fears of Hungarian irredentism. Most significantly,
analysts and policymakers are very concerned with Russia’s pre-occupation with the Near Abroad
—the other former Soviet Republics inhabited by more than twenty million Russians. To develop
security policy in post-Cold War Europe, we need to know which states will be irredentist.1
Analyzing political competition may help us understand under what conditions states are
likely to be irredentist. Many analysts have asserted that domestic political competition forces
politicians to follow aggressive foreign policies to divert their domestic audience.2 However,
increased political competition can also cause a state to be cooperative and peaceful.3 Changes in
political competition will influence a state’s foreign policy depending upon how competition
challenges the reigning and competing politicians. We must unpack “political competition” by
considering a) how do politicians compete; b) for whom do they compete; and c) how strong is
the competition they face. Politicians compete with each other by trying to attract supporters.
Without support, politicians lose office. In democracies, politicians who cannot gain enough
votes lose office, and in authoritarian systems, politicians who lose the support of the military or
some other necessary constituency at best lose their position (at worst, they might lose their
heads). To maintain and attract support, politicians must appeal to potential and existing
supporters’ interests. Consequently, the preferences of potential and existing supporters matter a
great deal. The ethnic ties between potential supporters and the ethnic groups in the neighboring
territories crucially influences possible supporters’ preferences. Their desires matter most when
competition is quite strong—when supporters can throw their support to an alternative politician,
and this can change the balance of political power. We should then expect irredentism when
politicians compete for the support of groups who have ethnic ties in neighboring territories.
In making the case for political competition’s influence on foreign policy, I first discuss
why our traditional tools for analyzing international relations do not help explain irredentism.
Second, I present how others have argued how political conflict and competition can cause
politicians to engage in diversionary wars. Third, I develop a relatively simple view of political
competition, taking seriously the preferences of those individuals and groups that politicians need
to gain and maintain influential positions. Fourth, I test these arguments against Somalia and
Serbia, because these two inconsistently irredentist states will facilitate intriguing comparisons.
Finally, I develop some implications this study has for theory and for policy.
1 The Puzzles
This article addresses three related puzzles: political competition’s influence upon foreign
policy, irredentism in general, and inconsistent irredentism in particular. Recently, analysts have
energetically debated whether political competition causes foreign policy to become more
aggressive or acquiescent. The Democratic Peace debate asserts that democracies are less likely
to war with other democracies, and one core logic of this argument is that political competition
restrains leaders.4 The opposing argument, diversionary theories of war, generally argues that
competition compels politicians to active aggressively abroad. Although this paper will only deal
with one side of this debate, diversionary theories of war, it explores a central puzzle for foreign
policy analysis: what are the foreign policy consequences of political competition?
1

Political Competition and Foreign Policy, page 2
The second puzzle is irredentism, defined here as state support5 for annexing neighboring
territories inhabited by ethnic kin,6 which is somewhat problematic for systemic theories of
international relations. For Neo-realism, security concerns are supposed to motivate states, but
irredentist efforts often impair a state’s security by creating enemies and alienating allies.7 Neo-
liberals usually assume states to be maximizing the welfare of their societies, but the quest for
“lost territories” tends to undermine the well being of the state. The cases in this analysis,
Somalia and Serbia, illustrate the self-destructive consequences of irredentist policies. Further,
systemic theories cannot account for the changes in a state’s foreign policy when the systemic
level variables do not vary. Likewise, they have a hard time explaining why a state’s foreign
policy might remain constant despite a changing environment. Consequently, neither Neo-realism
nor Neo-liberalism can account for our third puzzle: inconsistent irredentism.
Inconsistent irredentism serves as our third puzzle. Though frequently viewed as
constantly aggressive, irredentist states are often inconsistent in the strength and direction of their
foreign policies. Irredentist states frequently seek certain territories while ignoring others, and
vary in their aggressiveness over time, suggesting that ethnic identity or nationalism, by
themselves, cannot explain irredentism (a constant cannot explain something that varies). This
inconsistency can be helpful analytically as it allows variation in the dependent variable, providing
leverage for understanding the sources of both aggressive and acquiescent foreign policies.
Of course, the choice of these three puzzles is not random. Variations in political
competition may help us understand when states will be irredentist and when they will not.
Before discussing political competition, I will consider how systemic level theories might
approach this issue, and the limits of such arguments.
2 The Limits of Systemic Approaches
Systemic level arguments are less useful in understanding irredentism for two reasons:
their motivations for the behavior of states do not seem to capture what is going on; and the
variations in state behavior do not coincide with changes in external constraints. Neo-realism
asserts that anarchy motivates states to maximize their security. Given the relative shortage of
resources needed to defend itself, Somalia should have and did rely on the traditional means of
security: alliances. However, during both periods of Somali irredentism, early 1960’s and late
1970’s, Somalia’s foreign policies caused its essential allies to abandon Somalia and even assist
Somalia’s targets. Somalia engaged in irredentism against Kenya in the early 1960’s although it
would and did mean the loss of significant subsidies from the British, and the loss of any possible
protection in the future. At the same time, Somalia engaged in irredentism against Ethiopia when
Ethiopia clearly had Somalia out-gunned. In the latter 1970’s, despite the opposition of its
alliance partner, the Soviet Union, Somalia repeated the same mistake, as its irredentism towards
Ethiopia caused its ally to switch sides.8 Likewise, Serbia’s irredentism placed its most important
ally, Russia, in a difficult position, while alienating those powers that could provide significant
security and economic assistance: the European Union and the United States. While Neo-realist
arguments account for situations where the policies of rational, security-maximizing states might
lead to irrational or sub-optimal outcomes, they cannot account for self-destructive policies if the
states’ choices clearly run against what is best for the country. In an anarchic world, states should
not alienate important alliance partners, but Somalia did this twice. Unless the territories Somalia
and Serbia sought to annex held some special resources,9 it is hard to argue that war was a
rational course of action, given the likely costs.
The other major problem for systemic arguments is when systemic variables do not co-
vary with the behavior of states. In the cases of irredentism studied here, foreign policies varied
despite constant international constraints, and foreign policies remained constant despite changing
international constraints. Barry Posen, using a Neo-realist approach, has argued that increased

Political Competition and Foreign Policy, page 3
external threats causes politicians to use nationalism to mobilize the society in preparation for
war.10 This argument has a key prediction: nationalism ought to vary with the level of external
threats. The problem is that changes in the international climate do not necessarily cause the
degree of nationalism to vary. Somalia chose to fight Ethiopia both when Ethiopia was relatively
stronger than Somalia and when it was relatively weaker. So that the strength of the threat posed
by Ethiopia varied, but the policy did not. Serbia’s foreign policy became more cooperative (i.e.,
less irredentist) as the international community posed a more credible threat.11 This runs counter
to Posen’s prediction that nationalism should have increased in Serbia as the international threat
grew more severe, but that has not been the case.12 Of course, the counter Neo-realist argument is
that Serbia was more constrained by international opposition in the latter period. This reveals the
indeterminacy in Realist arguments. Table 1 indicates the foreign policy of these two states did
not co-vary with external threat.
Insert Table 1 Here
For Neo-liberalism, states facing common problems can overcome collective action
problems and cooperate by creating and maintaining norms and international organizations. On
boundary issues, Neo-liberals have argued that states that are vulnerable to separatism will not
support separatism elsewhere.13 Further, they argue that norms and international organizations,
where they exist, serve to constrain states and encourage them to cooperate. However, both
Somalia and Serbia have been and continue to be vulnerable to secessionism, while their
irredentism has violated norms and defied international organizations. As Table 1 indicates, the
foreign policies of these two countries are puzzling from the Neo-liberal perspective as well. The
point of this article is not to reject either Neo-Realism or Neo-liberalism as valid approaches to
international relations, but to provide an explanation for behavior for which systemic approaches
cannot account.
Before going further, it is important to address another alternative explanation: that
ethnicity and nationalism drive these policies, and states are merely tools of ethnic groups.14 This
approach essentially argues that ethnic ties or ethnic nationalisms are sufficient for explaining
outcomes without regard to domestic political processes. While this article’s argument focuses
on the role played by ethnic ties, it also asserts that ethnic ties are not sufficient by themselves. If
that were the case, the irredentisms of Somalia and Serbia, and probably that of others, would be
much more consistent both over time and among the possible targets of irredentist foreign
policies. Somali nationalism should have resulted in support for all Somali separatist movements,
and should not have varied over time, if the only causal force is nationalism. Likewise, Serbia’s
irredentism varied, which could not be predicted nor explained simply by reference to nationalism
alone. The cases will show that a focus on nationalism by itself is insufficient. States have
frequently not engaged in irredentism despite the existence of ethnic ties. In addition, irredentism
has been selective, with some states trying to annex some territories but not others.15 By focusing
on political competition, this article can explain policy changes and inconsistencies; something
neither systemic nor nationalist approaches can address.
3 Domestic Political Competition and Diversionary Wars
What are the foreign policy consequences of increased or decreased domestic political
competition? Diversionary theories of war suggest that when politicians’ positions are threatened,
they will follow aggressive foreign policies to divert the domestic audience from internal problems
and to increase support by mobilizing nationalism. This debate stemmed from social psychology:
“Conflict with out-groups increases internal cohesion.”16 These studies further suggest that
politicians may intuitively understand this, and use it to their advantage. When their positions are
threatened, politicians act aggressively abroad to create menacing “out-groups,” resulting in

Political Competition and Foreign Policy, page 4
increased domestic cohesiveness and greater support for themselves. Studies of public opinion
verify this social logic.17 Quantitative analyses of this hypothesis have used a variety of
independent variables seeking to capture the impact of political instability, rather than political
competition, upon foreign policy. The findings of these studies are mixed.18 Many suffered from
the same problem: the relationship between domestic instability and international aggression was
too deterministic. These works defined the question as: does internal conflict lead to external
conflict? Since there are many kinds of internal conflicts, and leaders have a variety of responses
to such problems, it is not surprising that the subsequent results were less than clear.
Furthermore, in this approach, when instability or competition compels aggressive foreign
policies, analysts frequently assume that the politician’s supporters respond positively to the
creation of external enemies. Their tests do not consider the enemy’s identity. This oversight
glosses over an essential aspect of any attempted diversion: the constituents, either narrow
interests or the masses, may have some preference about the choice of adversary. For this article,
the preferences of supporters matter, as these preferences not only shape the nature of foreign
policy, aggressive or acquiescent, but also are the likely targets of aggressive foreign policies.
Morgan and Bickers significantly revise the diversionary argument. They take seriously
the possibilities that a state may consist of many groups; that the leaders need not rely on the
support of the entire society to remain in power; and that the group in power may consider other
groups in the society to be out-groups. Thus, when the in-group is under stress, it may aim the
diversion at out-groups that exist within the state, rather than at external enemies.19 Morgan and
Bickers, therefore, assert that leaders should react differently depending on whether opposition or
discontent exists among those whose support is necessary for holding office, the in-group, or
among out-groups within the state. Diversionary foreign policy actions are more likely when
“support among members of the ruling coalition are eroding.”20 Morgan and Bickers find a
relationship between eroding support by the in-group and hostile foreign policy initiation.
However, their indicator of in-group support is the level of support among people expressing
membership in the president’s party. This omits any notion of preferences among supporters,
although their theoretical discussion implies that those preferences matter. Specifically, they
mention that groups vary within the state, contrary to previous applications of Coser’s arguments,
as “some will feel greater kinship with groups in other countries than with other domestic
groups.”21 Morgan and Bickers rightly argue that politicians care most about losing the support of
key constituents, but fail to consider the foreign policy preferences of those constituents who are
likely to leave the ruling coalition. Again, in their analysis, foreign policy activity—the creation of
external enemies—is the solution to eroding support, regardless of the defecting constituents’
concerns.
Edward Mansfield and Jack Snyder have argued that though mature democracies may not
be aggressive, democratization causes war. They argue “threatened elites from the collapsing
autocratic regime, many of whom have parochial interests in war and empire, use nationalist
appeals to compete for mass allies with each other and with new elites.” 22 They assert that there is
an “incentive for weak leaders to resort to prestige strategies in foreign affairs in an attempt to
enhance their authority over diverse constituencies.”23 Because elites in democratizing states are
not fully accountable to the electorate, and because elite interest groups rather than broader
parties mobilize the electorate, politicians care more about maintaining coalitions of narrow
groups. As in Snyder’s Myths of Empire, 24 logrolling among elite groups produces policy, and, at
least a few groups in the ruling coalition have “parochial interests in war, military preparation,
empire, and protectionism.” The ruling coalition then relies upon ideology and nationalism to
mobilize the masses, which apparently do not know better because they lack adequate
information.25

Political Competition and Foreign Policy, page 5
Although Mansfield and Snyder are quite persuasive, their argument has a few flaws
relevant for this article. It only addresses changes in competition through democratization.
However, the nature of political competition can vary even when the regime type remains
constant. Further, they tend to consider nationalism as something that mobilizes the masses,
without taking seriously that groups may have real preferences and may vary in their attachments
to different nationalisms. Thus, Mansfield and Snyder may not capture as much variation in
behavior.
Recent work has focused directly on irredentism as a consequence of increased political
competition. V.P. Gagnon, Jr. explains Serbian aggression as part of a strategy pursued by Serb
elites to maintain their positions as the threat to their positions increased. He argues that
Slobodan Miloševi and other Serbs sought to re-define the interests of their supporters and
opponents in ethnic terms, reshaping the political context so that they would not lose their
positions.26 This strategy had an aggressive component, provoking a sense of threat posed by the
other ethnic groups in Yugoslavia and promising policies to protect the Serbs from those threats.
These policies essentially promised a Greater Serbia, where a single Serb state would govern
Serbs in Croatia and Bosnia.27
While Gagnon provides an interesting argument and much data to support his assertions,
he tends to treat Serb irredentism as relatively consistent, assuming that provoking ethnic conflict
was always the most effective way for Serb elites to remain in power.28 This is problematic for
two reasons. First, Gagnon’s study covers the period before Yugoslavia’s disintegration (up to
and including 1991), but does not address Miloševi’s and Serbia’s varying support for Greater
Serbia since 1991. Second, he argues that the conditions favoring aggressive foreign policies are
the degree of domestic “threat to the existing power structure and the size of the coalition of
those within the power elite threatened by change.”29 Concerning the former, Serb irredentism has
not been consistent as this article demonstrates. Regarding the latter, it is not clear how the
degree of threat to the leadership and the size of the leadership determine a conflictual foreign
policy without incorporating variables accounting for supporters’ preferences. If politicians fear
losing support, then they should engage in policies that their exiting supporters prefer. The
argument to be presented here does not conflict with the heart of Gagnon’s, but takes as variables
what he generally considers to be constants, and, by doing so, can explain better the variations in
Somalia and Serb foreign policies.
By using a different understanding of political competition, I hope to improve upon these
existing approaches to diversionary foreign policies. While irredentism is only one possible way
to divert a domestic audience, a study of irredentism may have implications for other forms of
diversionary war.
4 Politicians, Supporters, and Foreign Policy
If politicians seek to gain and maintain political positions, the prerequisite for almost all
other goals attainable through politics, the question then becomes: how do politicians gain and
retain policy-making positions? In any political system, elites need the support of others. The
type of political system determines the nature of this support and who are these supporters.30 In a
democratic system, these supporters, or constituents, are voters, campaign contributors, and the
like, and they support a politician or her competitor by voting, donating money, etc. In an
authoritarian regime, the leaders’ constituency generally consists of those who control the means
of repression, such as the officer corps of the military, the security apparatus, and/or the one
party.31 Competition exists when these supporters can throw their support behind alternative
elites, and when such a change influences which politicians remain in office.
Incumbent politicians work to prevent these supporters from leaving their coalition, i.e.,
exiting.32 When a politician’s supporters abandon him or her and support a competing politician,

Political Competition and Foreign Policy, page 6
the politician will be less able to maintain his or her political position or enhance that position.
Thus, the possibility of supporters defecting is a crucial constraint for the incumbent and a vital
opportunity for those opposing politicians.
How strongly do politicians compete? This depends upon how the constituents support
the politician. If constituents support the politician with votes, then the question becomes how
much do votes matter? Votes matter the most, and the political system is the most competitive,
when the gain or loss of relatively few votes can greatly shift the balance of political power.33
Votes and voters matter much less when significant gains or losses do not change the balance of
political power. This might be the case in consociational regimes with grand coalitions
incorporating most, if not all, parties regardless of electoral outcomes.34
Likewise, if a politician’s constituency is the army, then the threat of the army removing
its support from the politician is severe, particularly when there is someone else or some other
policy option around which the military can rally. For instance, the Russian military’s threat to
abandon President Boris Yeltsin was quite credible and influential at key points in time since
Soviet Union disintegrated.
In states undergoing transitions in their political system, potential constituents will
probably be a mix of the two types. Incumbent politicians may rely on the support of both voters
and elites from the old regime. It is difficult to tell who really matters in a democratizing state
until the incumbents lose an election. If the incumbents allow electoral victors to take power,
then the voters matter. If not, then the groups that are able to repress the election’s results are
the important constituents. In democratizing states, the best we can do is determine through the
politicians’ domestic policies who are their most important constituents, and then deduce from
that what the foreign policies are likely to be.35 Still, we must pay attention to political institutions
and party structures to see how these may shape competition.
Responses to increased competition depend upon who is deserting the politician, who
might do so, and why they might defect. Policies aimed at attracting potential supporters and
avoiding the alienation of current backers require politicians to identify these individuals or groups
and their interests. The interests of likely defectors will determine which direction a politician
moves when competition increases. If hawkish constituents are threatening to defect, politicians
depending upon them will become more assertive. If, instead, more dovish supporters threaten to
abandon politicians, foreign policy will become more cooperative. Therefore, determining the
likely deserters’ preferences is important for both politicians and foreign policy analysts.
In sum, the primary constraint facing politicians are their supporters; this constraint bites
very hard when their backers can throw their support to competitors and when this changes the
balance of domestic political power. Therefore, the constituents’ preferences matter a great deal.
In this analysis, the most important influence upon supporters’ preferences is whether they have
ethnic ties36 or ethnic enmities37 with the ethnic groups inhabiting the likely targets of an irredentist
foreign policy. As is often argued, individual constituents may not have clear preferences on
complex foreign policy issues.38 However, constituents strongly favor policies assisting those with
whom ethnic ties exist.39 First, ethnic identity inherently creates feelings of loyalty, interest, and
fears of extinction.40 International boundaries do not cause members of ethnic groups to stop
caring about their ethnic kin’s condition. Constituents care most about those with whom they
share ethnic ties, or those with whom a history of ethnic enmity exists. For example, rather than
focusing on arms control, American Jews lobbied Congress in the early 1970s to make U.S. trade
policy with the Soviet Union contingent on its treatment of its Jewish population.41 Second, if the
irredentist foreign policy were actually successful, it would change the state’s ethnic composition,
increasing the political power of the enlarged ethnic groups, though not always equally. Third,
ethnic ties influence foreign policy-making because showing support for ethnic kin abroad can be

Political Competition and Foreign Policy, page 7
a litmus test for a politician’s sincerity on ethnic issues at home. Fourth, politicians can use the
circumstances of ethnic kin to emphasize certain ethnic identities at the expense of other identities
and issues. When constituents focus on economic problems or other troublesome issues, a
politician can use a foreign event to increase the salience of ethnic identity, creating unity at least
for the short term.42 Finally, in many polities it is difficult to determine the preferences of one’s
constituents. However, politicians can use as a cue the ethnic identity of likely or actual
supporters to calculate the preferences of potential and current constituents.
Constituents prefer to support their ethnic kin elsewhere, and they would prefer to do this
without any cost to themselves to situations where they bear costs for their irredentism. If
constituents bear severe costs because of their country’s support for their ethnic kin, they will
become less enthusiastic in their support. The sensitivity of ethnic groups to costs certainly varies
among ethnic groups. The point is simply that groups would prefer to gain the desired “lost
territories” without bearing costs than having to bear costs. It is an assumption of the argument
that at some point the costs of the irredentist policy that the constituents pay will become so large
that they outweigh the benefits, so that the constituents will prefer less irredentism or none.43
High costs cause constituents to allow their politicians to lessen assistance for their ethnic kin. To
be clear, it is only when the constituents themselves, not the country as a whole, feel the costs of
the policies, they will decrease their support for irredentism. According to this argument,
economic sanctions or casualties on the battlefield only influence a state’s foreign policy if they
hurt the reigning politicians’ supporters. Thus, states are likely to stop their irredentist activity
only if the decision-makers’ backers directly feel the pain caused by the aggressive foreign policy
We can determine whether ethnic ties and political competition influence foreign policy by
comparing Somalia’s irredentism with that of Serbia’s. Considering Somalia’s and Serbia’s
irredentism over time and among the potential targets of irredentism allows us to perform
simultaneously most similar and most different systems.44 Most different system analyses compare
highly different cases sharing a similar outcome, which allows us to eliminate as causes all of the
differences and suggest as causes only that which they share in common. Somalia and Serbia are
different in many ways: kind and degree of international intervention; the ethnic identities at stake
—clan versus religion; different levels of economic development;45 and regime type. Somalia has
had both relatively stable (nine years) democratic and relatively stable authoritarian periods while
Serbia has been, at best, under democratic transition throughout the entire period under analysis.
The contrasting regime types are most important because one purpose for this study is to indicate
the general dynamics of political competition regardless of regime type.
Most similar comparisons analyze two cases that are very similar, but have a different
outcome, so that the remaining differences are the likely causes of the outcomes. The most
similar state to Somalia between 1967-69 is Somalia in an earlier period or in a later period. The
same is true for Serbia. Likewise, by comparing the policy of an irredentist state towards one
region with its policy towards another region at the same time, we can hold almost everything
constant, allowing us to focus on the variations that are the probable causes. “Neither
comparison (neither across space nor across time) constitutes a perfectly controlled experiment—
far from it—but the two approaches together may provide much stronger evidence for our
hypotheses than either approach alone.”46
5 The Intriguing Case of Somalia’s Inconsistent Irredentism
Somalia is an interesting case to study because scholars have misunderstood its foreign
policies. Analysts have argued that Somalia is “perhaps the most persistently irredentist state in
the postcolonial world.”47 Persistent—yes, but not consistent: Somalia’s support for Somalis in
Ethiopia, Kenya, and Djibouti has waxed and waned over time, and Somalia’s irredentism has
varied considerably among the different “lost territories.” Somalia has fought wars to redeem the

Political Competition and Foreign Policy, page 8
Ogaden region of Ethiopia, less directly assisted the efforts of Kenyan Somalis to separate from
Kenya, and has been much less supportive of the efforts of Somalis in Djibouti. This
inconsistency is analytically intriguing not only because it runs counter to popular beliefs about
Somalia, but because it also provides the analyst with variation in the dependent variable: the
state’s support for irredentism.48 Studying Somalia is also fascinating because it went through
highly competitive periods and less competitive periods in both its democratic form and its
authoritarian period. The variations in competitiveness do not neatly match up with type of
regime, so we can determine how competition can vary in both kinds of political systems and
what the foreign policy impacts might be.
Insert Map 1: Distribution of Somalia’s Clans and Clan-Families
6 Somalia’s Ethnic Politics
According to most of the usual markers of ethnicity—language, religion, and race—
Somalia has Africa’s most homogeneous population. Somalis consider themselves to be of a
common ancestry, they speak a single national language, they follow the same religion, Islam, and
most experts consider the Somalis to be a nation. Because the Somali nation is larger than the
actual state of Somalia, the condition and future of Somalis residing in the neighboring territories
has been an important concern for Somalia even before independence.49 At independence, close to
250,000 Somalis lived in Kenya’s Northern Frontier District [NFD], approximately a million lived
in the Ogaden region of Ethiopia, and between forty and fifty thousand resided in French
Somaliland.50
The politics of both democratic and authoritarian regimes in Somalia are largely the result
of alliance formations between various kinship units and of conflict among various sub-clans,
clans, and clan-families. Though many have argued that support for Greater Somalia “was one
issue that united the vast majority of them [the Somalis],”51 interest in pursuing Greater Somalia
varied, both over time and among the differing clans and clan-families.52 Because of the conflicts
between the different clans and clan-families, and due to the particular distribution of clans
throughout Somalia and the neighboring states, any successful effort towards a Greater Somalia
would have distributive consequences for Somali politics. Especially in a democratic system
where numbers count, strengthening a particular clan or clan-family—the likely consequence of
any successful irredentist effort—would alter the balance of power within Somali politics.
Therefore, despite the initial enthusiasm for a Greater Somalia, and despite the common bonds,
not all Somalis felt an interest in irredentism nor did all Somali politicians faithfully pursue
irredentism.
By considering the various clans’ interests and which clans and clan-families politicians
needed to remain in office, this analysis will show that Somali irredentism varied due to changes in
the clans that mattered and how much they mattered. There are six main Somali clan-families: the
Darod, the Hawiya, the Ishaak, the Dir, the Digil, and the Rahanwein.53 The first three of these
clan-families are the largest and most politically significant. The Darod is the largest clan-family,
consisting of over a million and a half Somalis,54 so its constituent clans influence Somali politics
more than those of the other clan-families. “They were therefore the only clan-family with a real
stake in pan-Somalism.”55 While Darod consists of numerous clans, only four are relevant here:
the Majertain, the Marehan, the Dolbohanta, and the Ogaden. The leaders of the Majertain clan
played a central role early in the democratic regime in the 1960’s. The Somali Youth League
[SYL], the dominant party, consisted of members of Darod and Hawiya clan-families, but was led
primarily by members of the Majertain clan. The Majertain are located on the central coast of
Somalia, and, therefore, are less concerned about Somalis in Ethiopia or Kenya. The Marehan are
a smaller clan, inhabiting the border regions between Kenya, Ethiopia, and Somalia. This clan
played a most important role in Somalia’s politics as Siyad Barre, the ruler of Somalia from 1969

Political Competition and Foreign Policy, page 9
until 1991 was of this clan. The Dolbohanta occupy territory overlapping northern Somalia
(formerly British Somaliland) and southern Somalia (formerly Italian Somaliland). They became
politically significant during the mid-1970's, as the head of the National Security Service at that
time was both a Dolbohanta and one of Siyad’s sons-in-law. Because of these two connections,
the Dolbohanta became one of the clans supporting Siyad’s rule.
Finally, the Ogadeni clan, mostly located in the Ethiopian region named after them, has
played a very important role in Somalia’s politics, domestic and foreign. As it is the clan with the
largest Somali population outside Somalia, it has a very keen interest in Somalia incorporating the
their region of Ethiopia into a Greater Somalia. Since the Ogadeni are a sizable part of the Darod
clan-family, any politician seeking to gain the support of the entire Darod clan-family has had to
appeal to the Ogadeni.56
The Hawiya is the largest clan-family in southern Somalia, numbering more than half
million. They occupy the capital of Mogadishu and territories crossing the borders between
Somalia and Kenya and between Somali and Ethiopia. The Hawiya were politically important
during the democratic period, as each of the major political parties courted its support. During
the period of military rule, the Hawiya were excluded from power, and it has been only recently
that the Hawiya have been able to wield influence over southern Somalia.
The third politically important clan-family in Somalia is the Ishaak, who reside in northern
Somalia. Since the unification of British Somaliland with Italian Somaliland, the Ishaak have
more frequently opposed the Somali government, democratic or authoritarian, than they have
been in support. Ishaak opposition has sometimes taken the form of a secessionist movement,
leading to the unrecognized existence of Somaliland since Siyad Barre’s fall. Some Ishaaks live in
Ethiopia, and are rivals with the Ogadeni who also reside there.57 The Ishaaks who graze in
Ethiopian territory prefer gaining access to, rather than control over, these territories.58 The
Ishaaks have generally viewed the efforts to support nationalist movements in Somalia’s
neighbors to be a waste of precious resources needed for development.59
The Dir clan-family have generally played a less influential role in Somalia’s politics, but
one of their constituent clans, the Issa, inhabit the critical territories of Djibouti and northern
Somalia. Issa interest in unifying with Somalia has varied over time, but has certainly been less
enthusiastic than that of the Ogadeni.
In general, to determine how severe competition is, we need to know whether a single
clan or clan-family abandoning the decision-makers would threaten the elites position. Do their
votes matter? In a military regime, the question is whether a particular clan controls or influences
the institutions or groups upon which the regime depends. By examining the political institutions
and the ethnic composition of parties, the officer corps, and other potentially important
constituencies, we can determine the likely pressures facing politicians, and then their likely
foreign policies.
Table 2 sketches the patterns of political competition and foreign policy in four major
periods of Somalia’s history, and then I discuss each period.
Insert Table 2 here
7 Democracy and Irredentism: 1960-1967
At the outset of Somalia’s independence in 1960, competition between the major Somali
parties was intense. While the Somali Youth League won the first elections, the Greater Somalia
League [GSL] was a viable competitor until after the elections in March 1964. The electoral
system was one of proportional representation with multi-member districts, including districts
where seven candidates could win. This meant that a candidate gaining 15% of the vote could

Political Competition and Foreign Policy, page 10
gain office. Consequently, voters could reasonably threaten the SYL by giving support to either
the GSL or another competing party if their preferences were not satisfied.60 The most important
constituents were the entire Darod and Hawiya clan-families—the largest clan-families. The
Darod, with one of their constituent clans, the Ogadeni, having strong ties to Somalis in Ethiopia,
desired an assertive foreign policy towards Ethiopia. The Hawiya had ethnic kin in Ethiopia and
in Kenya’s Northern Frontier District. The two largest parties, each seeking support from these
two clan-families, focused on the issue of Pan-Somalism, as this was the one issue on which these
two clan-families could agree at the time.61 As the party in opposition, the GSL portrayed the
Somali Youth League as too weak in its support for Somali nationalist movements in Kenya and
Ethiopia.62 Ethnic outbidding over the irredentist issue between the two parties occurred as each
of these two parties feared losing supporters to the other. Politicians within these parties
depended upon the favor of the Darod and Hawiya clan-families and had to compete for that
support, compelling politicians to appeal to the ethnically defined interests of these constituents.
The relative size of the Dir clan-family and its Issa clan made them relatively less important to the
competing politicians. This allowed the Greater Somali tendencies of the two predominant parties
to focus on Ethiopia’s and Kenya’s Somalis while caring less about the plight of Djibouti’s
Somalis.
Somalia began its efforts to annex Kenya’s Northern Frontier District immediately after
Somali independence with its national assembly urging the government to use “all possible
means.”63 After the failure of diplomatic efforts with Great Britain, governor of colonial Kenya,
Somalia broke off relations with Britain in March 1963, despite being somewhat dependent on
British subsidies. The Somali government supplied arms, training, and propaganda assistance to
Kenya’s Somalis. Guerrilla warfare between Kenyan Somalis and the Kenyan government broke
out, though Kenya quickly suppressed the Somalis.
“Accused of not providing sufficiently strong leadership on the N.F.D. issue, and
increasingly unpopular for its management of home affairs, the SYL government searched
anxiously for some dramatic action which might restore its popularity and enhance its position
before the coming election.”64 The SYL directed this drastic action towards Ethiopia, as border
clashes between Somali and Ethiopian regular forces erupted. In October 1963, Somali soldiers
launched attacks on Ethiopian villages along the border. The conflict ended in March 1964, only
after Ethiopia’s superior forces repelled the Somali soldiers.
During the first seven years of Somalia’s independence, Somalia paid considerably less
attention to Djibouti’s Somalis. The Liberation Front of the Somali Coast [FLCS] formed in
Mogadishu in 1963, with the Somali government providing financial and military assistance,
bases, and training.65 The Somalis appealed to both the United Nations and the Organization of
African Unity to pressure France into granting Djibouti independence. Though the Liberation
Front of the Somali Coast did receive support from the Somali government, this aid was not as
significant as the assistance sent to Somalis in Ethiopia or to those in Kenya.
In Somalia’s history, the most intense support for Pan-Somali efforts occurred in late 1963
and early 1964—before the March elections. It seems unlikely that the timing of increased
political competition and of Somalia’s most aggressive irredentism is coincidental. The fighting
stopped due to defeat on the battlefield, but Somalia did not immediately try again because the
1964 election decreased the pressures that decision-makers had felt. As result, though Somalia
gave some assistance to separatist groups in its neighbors, Somalia did nothing more than that.
8 Somali Democracy and Détente: 1967-1969
The March 1964 elections changed the nature of political competition. Incumbents faced
less pressure as their competitors disintegrated. Because the Somali Youth League had won the
previous elections, it took advantage of its incumbency by using state funds to encourage, i.e.,

Political Competition and Foreign Policy, page 11
bribe, other parties’ politicians to join the SYL. After the 1964 election, 21 members of other
parties switched their loyalty to the SYL.66 In the 1967 elections, rather than facing a significant
threat from competing politicians, the SYL faced more than sixty small parties, most seeking the
support of sub-clans,67 rather than appealing to a wider audience. The SYL’s ability to buy the
opposition resulted in the disbanding of all of these smaller parties after the election, with their
representatives joining the SYL. The 1969 election provides further proof of the simultaneous
strengthening of the SYL and the disintegration of the rest of the parties. Out of 123 seats, the
SYL won 73, bought another 36, and its coalition partner, Somali National Congress, controlled
all but three of the remaining seats. The fragmentation of the party system, combined with the
Somali Youth League’s ability to persuade members of smaller parties to join it, meant that
competition constrained ruling politicians less. Rather than worrying about supporters defecting
to other parties, the SYL’s leaders successfully encouraged their rivals’ supporters to abandon the
competing parties instead.
After the 1967 election, the SYL formed a coalition with the Somali National League, the
party of the Ishaaks of northern Somalia. Abdirashid Ali Shermarke, the President, and
Mohammad Egal, the Prime Minister, led this coalition. Shermarke was of the Majertain clan
within the Darod clan-family, and Egal belonged to the Ishaak clan-family. Consequently, the
regime had the support of the Darod clan-family, especially the Majertain, and of the Ishaak clan-
family. Having won office, Egal, whose interests and support continued to come from the North,
began to distinguish himself from previous Somali politicians by seeking accommodation with the
neighboring states.68
These policies did not alienate the regime’s core supporters, the Ishaak and the Majertain,
which would not gain from “reclaiming” the Ogaden or the Northern Frontier District. The
Ishaak did not desire a Greater Somalia, as their primary interest concerning Ethiopia was peace.
Peace would give the nomadic Ishaak access to grazing areas in Ethiopia, and it would allow the
Ishaaks of the north more direct access to the capital in southern Somalia. For the Majertain,
their relative power within the Darod clan-family would weaken if the Ogadeni of Ethiopia re-
united with their kin in Somalia. A successful irredentist policy would only weaken the
Majertain’s influence. Due to the fragmentation of the political system, Egal and Shermarke did
not have to worry about losing Ogadeni or Hawiya support.
Therefore, after the 1967 elections, Somali foreign policy changed direction from the post-
1964 election moderate levels of supporting irredentism to détente with the neighboring states.
Prime Minister Egal announced Somalia’s respect for Kenya’s borders, and stopped the supply of
arms to Somalis in Kenya.69 Egal’s détente policy also applied to relations with Ethiopia, stopping
supplies to Somali secessionist movements in the Ogaden. Egal agreed with Emperor Selassie to
form a joint military commission to oversee the boundary, establishing air and telecommunications
links, and re-opening the roads between Somalia and Ethiopia.70 Détente with Ethiopia was part
of a campaign to improve the ties between Northern and Southern Somalia, as Ethiopia’s territory
divided the shortest routes between the population centers of Somalia’s two regions. Egal made
similar, but less formal, efforts to court the French, as Egal tried to reassure them that Somalia did
not have designs on French Somaliland (Djibouti).71
This period of relatively modest competition, due to the fractured party system, allowed
Egal and Shermarke to engage in very cooperative policies with countries that Somali foreign
policy had sought to divide only a couple of years earlier.
9 The Moderate Military Regime: 1970-1974
Shortly after the 1969 election, the military intervened, justifying its coup by condemning
the corruption and tribalism of the democratic regime. The new military regime, the Supreme
Revolutionary Council [SRC] led by Siyad Barre, did not face any competition at first, as no

Political Competition and Foreign Policy, page 12
factions of the military threatened to launch a counter-coup, nor were the masses upset by this
event. Since he was not dependent on any single clan or limited number of clans, no clan or clan-
family could threaten Barre’s position by removing their support. Further, Siyad Barre tried to
change the basis of Somali politics to reduce the likelihood of future competition. Because he
came from a relatively small and less than popular clan, Siyad sought to lessen the role of clan
identities in Somali society through various policies.72 This strategy had implications for foreign as
well as domestic policies.
Disinterest in the Greater Somali project continued during the first years of Siyad Barre’s
rule. Because both Siyad’s clan and his multi-clan family coalition were relatively weak, Siyad
tried to ride the wave of disgust for clan politics.73 A Greater Somalia might appear to have a
nationalist appeal for all Somalis. However, any irredentist efforts would upset some clans, as
some had no kin residing in the neighboring states. Further, unless the irredentist effort aimed
equally at all three neighbors, even those clans and clan-families having narrowly defined
irredentist interests would be dissatisfied.
Barre continued Egal’s détente with Somalia’s neighbors during the first years of the
regime. Indeed, in 1970, the new regime crushed the Western Somalia Liberation Front [WSLF],
reducing the level of conflict within Ethiopia.74 Siyad then went to the Organization of African
Unity meeting in Addis Ababa in 1971, thanked his Ethiopian hosts, and did not mention the
border dispute at all.75 There were some disputes between Ethiopia and Somalia over the Ogaden
region, but both sides tried to prevent these disagreements from escalating. Moreover, the status
of Djibouti’s secessionist Liberation Front of the Somali Coast “underwent a sharp decline,” as
the government of Somali “no longer supported it.”76 Siyad Barre indicated that Somalia would
give up its claims to French Somaliland as long as Ethiopia did not annex it.77 Somalia’s dwindling
support for secession was consistent among the three “lost territories” during the first four years
of Siyad Barre’s rule.
10 Ambivalent Aggression: 1974-1980
The turning point in the nature of Somali political competition, and not surprisingly, the
military regime’s ethnopolitical strategy occurred between 1973 and 1975. Efforts to reduce the
role of clan identities in politics were frustrated by both the enduring clan divisions in Somali
society and the convenience of using clan ties to build support for Siyad Barre’s regime. Siyad
used clan politics to create a robust coalition out of his original position of weakness. He
increasingly relied on ties to three clans within the Darod clan-family. Members of the Marehan
clan, Siyad’s clan, were the only Somalis to be members of the Presidential Guard. The Ogadeni
clan, his mother’s, became over-represented in the officer corps of the military. Finally, the
Dolbohanta, his son-in-law’s, controlled the secret police. Most crucially, the clan homogeneity
of the military’s officer corps increased the coherence of its interests, clarified its preferences, and
made more credible and more powerful its threat to rebel against Siyad. Rather than facing the
desertion of just one brigade or battalion, Siyad faced the possibility of the entire military
withdrawing its support. Consequently, those officers’ demands weighed heavily on Siyad.
Due to their many clan members living within Ethiopia, the Ogadeni had a strong interest
in seizing the eastern portion of Ethiopia. Their increased control over Siyad’s key constituency,
the military, endowed them with the greatest influence among the clans. Ultimately, the Ogadeni
held Siyad held hostage, compelling him to go to war with Ethiopia.78 Siyad’s ruling coalition
included only token members of Hawiya or Dir clan-families, so the irredentist project focused
solely on Ethiopia.
Siyad Barre renewed support for the Western Somalia Liberation Front in 1975, which
immediately began to attack Ethiopian targets. By the summer of 1977, border crossings by
regular Somali army units and the contribution of “volunteers” from the Somali army to the

Political Competition and Foreign Policy, page 13
WSLF had escalated into a full invasion of Ethiopia by regular Somali forces. The contrast
between Somalia’s invasion of Ethiopia and its accommodation of Kenya and Djibouti during this
time is striking, as is the contrast between this outbreak of selective irredentism and the 1963-64
aggression against both Kenya and Ethiopia. As conflict mounted with Ethiopia, Siyad Barre
“sought to reassure the Kenyans that Somalia had no plans to support the dissident forces in the
northeastern province. In regard to the Territory of ‘Afars and ‘Iises [Djibouti], Somali diplomats
were also willing to compromise on their original goal of political unity for all Somali peoples.”79
Even as Somalia went to war to annex the Somali-inhabited areas of Ethiopia, its efforts were
inconsistent, as Siyad Barre did not support the efforts of Somalis in Kenya and Djibouti to unite
with Somali. This varying irredentism suggests that Barre’s interests in maintaining his coalition,
rather than in building or relying upon Somali nationalism, drove the direct of foreign policy.
The attack on Ethiopia ended quite badly for Somalia, as the Soviet Union flew in more
than 15,000 Cuban troops, as well as large amounts of arms. The rout caused Somalis in Ethiopia
to flee to Somalia, increasing tensions within Somalia because several clans less interested in
irredentism began to bear great costs of the refugee flow. “The terrible defeat and refugee
invasion quickly led to widespread public demoralization and to an upsurge in ‘tribalism’ as
different groups sought scapegoats to explain the debacle.”80 Only a month after the end of the
war, Majertain military officers attempted a coup, but failed. One Majertain coup leader escaped
to Ethiopia and formed the first formalized opposition movement to Siyad Barre’s government—
the Somalia Democratic Action Front.81 In April 1981, a new opposition movement, the Somali
National Movement, comprised almost entirely by the Ishaak clan-family, was established.
Although Somalia disintegrated in 1991, the failed war caused Somalia’s current problems,
including the secession of Somaliland by the Ishaaks. One at a time, every clan except Siyad’s
eventually took up arms against his regime, even the Ogadeni, eventually ousting Siyad. These
conflicts did not cause further irredentism because the Ogadeni could no longer threaten Siyad
after they left the regime. Since they were no longer constituents, neither the Ogadeni’s support
nor their interests greater constrained Siyad or his foreign policy.
This case clearly demonstrates the importance of ethnic bonds between one’s constituency
and the irredentist movements in question. As table 2 indicates, Somalia’s foreign policy varied,
depending on the ruling clans’ ties with Somalis in Ethiopia, Kenya, and Djibouti, and as the level
of competition changed. Ethiopia’s Somalis had more influence and more support because they
were relatives of one of the largest clans within Somalia’s most politically important clan-family.
During the democratic period, both major parties sought the Darod clan-family’s support. This
essentially enabled the Darod’s members to set their price for supporting the ruling party: greater
assistance for the Ogadenis seeking to secede from Ethiopia.
While the Ogadeni clan and the Darood clan-family were also able to influence the military
regime, the Hawiya clan-family was only able to influence the democratic governments when
politicians required their participation in coalitions. Since the Hawiya cared about the conditions
of their kin in Kenya, parties seeking the Hawiya’s support consequently pushed for greater
efforts on the behalf of Kenya’s Somalis. Because the Hawiya were excluded from power during
the military regime, and since Kenya’s Somalis were predominantly of the Hawiya clan-family,
Siyad Barre had very little interest in assisting them. Yet, the importance of ethnic ties between
constituencies of the state’s leadership and an irredentist movement is perhaps clearest in the case
of Djibouti. Those Somalis in Djibouti belonged to clans whose support was neither critical for
the democratic regimes nor essential for Siyad Barre’s regime. Consequently, they received little
support compared to Somali separatists in either Kenya or Ethiopia.

Political Competition and Foreign Policy, page 14
Political competition clearly mattered as Somalia’s most energetic irredentist efforts
occurred during periods of heightened competition—before the 1964 election and in the mid-
1970’s when the predominantly Ogadeni officer corps could make credible threats to rebel.
Somali politicians cared less about the irredentist project when they did not fear competition—
after the competing parties fragmented and during the first years of the military regime when no
faction or clan posed a significant threat. Thus, political competition and ethnic ties helps us to
understand Somalia’s inconsistent irredentism. However, to evaluate whether the Somali case is
unique or useful for generalizing, we must consider how other seemingly consistent irredentist
states behave and what shapes their irredentism.
11 Inconsistencies in Serb Irredentism
While there still may be some controversy over which factors caused Yugoslavia’s
disintegration, there is a consensus that Slobodan Miloševi’s quest to gain and maintain his
position largely motivated Serbia’s irredentism.82 It would seem then, that a study of Serbia’s
irredentism that focused on politicians’ domestic political interests in Yugoslavia would be either
redundant or trivial. However, concentrating on Miloševi’s efforts towards a Greater Serbia
ignore the inconsistencies in his policies, and these inconsistencies may point to key variables that
shape both Serbia’s irredentism and the dynamics of irredentism in general.
The important inconsistency is that Serbia’s enthusiasm has waned over time, leading to
less support for irredentism even towards Croatia and Bosnia. Why has Miloševi lessened his
irredentist efforts? The answer may lie with two alterations in the Serb political landscape: a
decline in political competition and increases in the costs Miloševi’s supporters had to bear.
Insert Map 2: The Former Yugoslavia
12 Irresolute Irredentism83
The pursuit of a Greater Serbia was never as consistent as portrayed, and became less
consistent over time. Even at the outset, Serbia did not target all republics inhabited by Serbs.84
Slovenia was allowed to secede after a minor conflict, and Macedonia departed with no violence
at all. In contrast, Serbia’s and the YPA’s aggressively supported separatist Serbs in Croatia and
Bosnia. Serbia’s support for Serb irredentists began before Croatia or Bosnia seceded, as
Belgrade appears to have organized and armed the Serb opposition in both republics.85 The YPA
initially did not take sides in the Croatian conflict. However, it quickly did so in the summer of
1991, with devastating effect. The YPA fought against the Croatian forces, using its artillery
against Croatian towns. Not only did the Serb-dominated federal army assist Croatia’s and
Bosnia’s separatists, but Serbia sent irregular forces to both republics. Vojislav Seselj, leader of
one of the most aggressive paramilitary groups in Croatia and Bosnia and of a major party in
Serbia, admitted, “Serbia did everything for us.”86 When the YPA withdrew from Croatia and then
Bosnia, it left behind both equipment and personnel in the hands of the Serb separatists.
However, even this enthusiasm for irredentism waned over time. Beginning in the middle
of 1993,87 Miloševic started to accept various peace plans offered by the United Nations and
European Union representatives, including the Vance-Owen plan, which the Bosnian Serb
parliament rejected. Miloševi then used his control over the media in an attempt to undermine
Bosnian Serb leader Radovan Karadzic’s influence. In the summer of 1994, Serbia closed its
borders to all traffic to Bosnia, except for food and medical supplies. While analysts and other
states questioned the tightness of this sanction, Miloševi and the Serb government clearly
attempted to pressure the Bosnian Serbs into accepting a peace agreement. The Dayton Accords
process also reflected Miloševi’s interest in ending the Bosnian war, even if the Bosnian Serbs
opposed such efforts.

Political Competition and Foreign Policy, page 15
Perhaps the most surprising development in 1995 was not Miloševi’s support for the
Dayton Accords, but that Serbian armed forces did not defend the Serb-controlled regions of
Croatia from Croatia’s re-conquest in August. Having used force to control large pieces of
Croatian territory, Serbia could have employed its army to retain those territories, but chose not
to do so. Still, Miloševi has not rolled over completely, as he has not served up Radovan
Karadzic or Ratko Mladic to the war crimes tribunal. Further, Miloševi was able to negotiate a
deal giving Serbia much of what it wanted at the Dayton talks.
Insert Table 3 Here
13 Varying Competitive Pressures and Changes in Preferences
Overall, Serbia’s “defense” of Serbs in the other former Yugoslav republics has varied
considerably both over time and between different republics. The competitive pressures within
Serbia and the preferences of those constituents for whom Serb politicians compete provide the
answers to this empirical puzzle.
Most analyses of Yugoslavia’s disintegration start with the rise of Slobodan Miloševi and
his use of Serb nationalism to gain and maintain power. Due to a severe economic crisis,88 and
communism’s declining legitimacy, Yugoslavia’s politicians had to develop new strategies to
remain in power or improve their positions. Miloševi used Serb nationalism to remove his
mentor, Ivan Stambolic, and others who opposed the nationalist strategy, winning the party’s
presidency in 1987. Due to Yugoslavia’s highly federalized political institutions, Miloševi’s main
concern was winning the support of essential constituencies within Serbia, rather than appealing
to a broader audience within Yugoslavia. Within Serbia, he sought and received political support
from the Communist party (renamed the Socialist Party of Serbia [SPS] as elections approached),
farmers, the elderly, the elites who ran state-owned factories, unskilled and semi-skilled workers,
and junior Serb army officers.89 Real political and economic reform would harm each of these
groups, reducing their political power and exposing them to the vagaries of the marketplace.
Miloševi log-rolled ( la Snyder) and created a coalition of these conservatives, who feared losing
power and benefits if Serbia and Yugoslavia truly reformed economically or politically, with
nationalists, who focused on the condition of Serbs within the federation. Both groups sought
recentralization of Yugoslavia under Serb leadership, as this would both protect Serbs and delay
economic and political reforms.
Given the army’s importance in both prosecuting the wars and supporting Miloševi’s
position, we need to know what the army wanted. As one of the few Yugoslav institutions left,
the army strongly desired a unified Yugoslavia. The country’s potential breakup threatened the
officers’ positions. The Yugoslav military focused its attention on Slovenian and Croatian efforts
to arm themselves, increasing the hostility between the army and the two seceding republics.
Once unity was no longer a possibility, other interests came into play. The majority of the army’s
officer corps was Serb,90 and the influence of the Serbs within the army increased as Croatian and
Slovene officers and conscripts began to defect in larger and larger numbers. Consequently, the
army increasingly preferred to defend the Serbs outside of Serbia in the name of Yugoslavian
unity.91
Pressures for elections developed within Yugoslavia, with the first ones held in 1990.
Serbia was among the last Yugoslav republics to hold elections. Voters could vote for
Miloševi’s opponents, so electoral competition existed, although it is still not clear whether
Miloševi would have stepped down had he lost. Miloševi’s Socialist Party of Serbia competed
with the Serbian Renewal Movement [SPO], and the Democratic Party, as well as a coalition of
moderate parties. Vuk Draskovic led the SPO, which had begun as a nationalist, anti-communist
party. During the campaign, Miloševi took relatively moderate stands on nationalist issues.92

Political Competition and Foreign Policy, page 16
This stance, as a moderate nationalist, is problematic. Ethnic politics would predict that
competition would push Miloševi to take strong stands in defense of ethnic kin abroad, yet he
accused Draskovic of being too extreme on this issue.
So, how can we make sense of Miloševi‘s moderate stance? First, Miloševi’s moderate
stance did incur some costs, as more extreme nationalists did move to the SPO, which was the
second most successful party in the 1990 election. Second, being a moderate nationalist in Serbia
in 1990 does not make one an egalitarian or a supporter of multiculturalism. Miloševi was not
ignoring the “plight” of the Serbs in Croatia and Bosnia-Hercegovina. Rather, he was arguing
that he could defend them and their interests without going to war—that his nationalist agenda
was less costly than Draskovic’s. Given a choice between costly nationalist policies and
inexpensive ones, voters are likely to select the latter.
Third, the elections, though relatively free, did not threaten Miloševi’s position as much
as one might think. The exit threat was not as high as it would appear due to the nature of the
electoral system, the advantaged position of Miloševi’s party, and his control of the media. The
electoral system was one of majority districts, meaning that a candidate had to win not just a
plurality, but a majority of votes within a district to gain a seat in parliament. In Yugoslavia’s
elections, this electoral rule gave disproportionately large numbers of seats to the strongest, best-
organized parties.93 Since the SPS inherited the organization of Serbia’s Communist party, they
were able to compete throughout Serbia and win many districts against divided competitors.
Furthermore, Miloševi controlled the media in Serbia, giving him a greater ability to set the
political agenda and to define his opponents as he wished. Finally, Miloševi’s incumbency
enabled him to plunder the Yugoslav and Serb treasuries to give money to pensioners and other
important constituents before the election.
Consequently, Miloševi and the SPS got 45.8% of the vote in the first round, resulting in
nearly 78% of the seats in the parliament. However, the elections did not settle Serbia’s political
conflicts as protests developed in Belgrade in March 1991. Serb elites considered using force at
this time against the protestors, and while there were some casualties, the leaders did not declare
an emergency. At this time, violence in Croatia escalated as Serb paramilitary forces led by
Vojislav Šešelj started to attack elements of the Croatian police. Two months later, Miloševi and
Serbia blocked the normal rotation of the federal presidency, leading to the final breakdown of
Yugoslavia.
The war’s initial stages polarized the military and Serbian public opinion. Not only did
Slovenes and Croats defect from the military, but officers and conscripts from other ethnic groups
also deserted, leaving the military even more dominated by Serbs.94 The SPS began to lose
support, mostly to Šešelj’s more extreme Serbian Radical Party but also to the more moderate
Democratic Party. The SPS, in fact, lost support from all sectors of its coalition. The moderate
competing coalition disappeared, the SPO lost support, and the Serbian Radical Party now
became a major player.95 The next major election in Serbia pitted Miloševi against Milan Panic,
an émigré executive returning from the United States, presenting a clear choice between
nationalist conflict and a more peaceful approach. Since the previous election, Serbia had
aggressively and successfully supported Croatia’s and Bosnia’s Serbs. Panic had pushed for more
accommodative foreign policies. On Dec. 20, 1992, after a year and a half of conflict, the Serbian
people chose conflict over moderation, as Miloševi won over Panic, but lost votes to Vojislav
Šešelj’s ultranationalist Serbian Radical Party.96
However, neither Serbian public opinion nor the nature of competition has remained
constant over time. The economic sanctions levied against Serbia and Montenegro (the new
Yugoslavia) in May 1992 have hurt the already damaged Serb economy. “Although the embargo
took a serious toll on the civilian population, it did have positive effects… it made Miloševic

Political Competition and Foreign Policy, page 17
readier to make compromises.”97 In December of 1993, both Vuk Draskovic’s Serbian Renewal
Movement and the more moderate Democratic Party criticized Miloševi’s efforts to compromise.
However, in the summer of 1994, after sanctions started to damage the Serb economy, the major
opposition parties approved of Miloševi’s support for a peace plan that Bosnian Serbs opposed.
Miloševi even went so far as to impose his own sanctions on the Bosnian Serbs, which led to the
lifting of some international sanctions.
Political competition has also changed over time, declining over the course of crisis,
though it intensified again during the fall 1996. It is difficult to determine how and whether
elections matter in political systems in the process of transition. However, Miloševi’s behavior
from 1990 to 1997 does suggest that he would prefer to win elections than lose them, though his
behavior also suggests that a lost election may only pose difficulties,98 rather than remove him
from power.99 Miloševi has been successful in past elections, but he also has been quite willing to
undermine political institutions and popular regimes in Kosovo, Vojvodina, and Montenegro.
Given his resistance in the fall of 1996 to allowing his opponents to take positions won at the
local level, it is hard to believe that Miloševi will give up his position easily. Indeed, since the
1992 election, Miloševi has built an alternative basis of support—a police force that is larger than
the army. Miloševi has paid special attention to Serbia’s police, greatly increasing its size, trying
to attract potential defectors within the military with higher salaries. On the other hand, the
military, which had lost troops and material when it retreated from Slovenia, Croatia, and Bosnia
faced budget cuts in 1992.100 Is competition a threat to Miloševi now as it was in the late 1980’s
or at the outset of the conflicts with Croatia and Bosnia in 1991? The existence of his police
force indicates that Miloševi may not care as much about losing the support of the army or, at
least, that he sought to reduce his dependence upon the military. Although Miloševi gave ground
to the winners of the 1996 local elections and to the street protests, the police supported
Miloševi and tried to crush the resistance in numerous instances. Controlling a very large police
force may not be sufficient to rule (i.e., it is not a minimum winning coalition by itself), it certainly
weakens the threats made by other constituents.
Given the relatively few Serbs in Slovenia and Macedonia, we would expect Serbia to be
relatively less supportive of Serbs in those areas. This prediction is borne out by Serbia’s policies,
as Slovenia met with weak resistance when it seceded and Macedonia met with no opposition.
The higher numbers of Serbs in Croatia and Bosnia should cause Serbia to give more support to
Serbs in these two republics, and this expectation was correct as Serbia very aggressively
irredentist towards those two republics. If politicians did not face as much competition or their
supporters were feeling the costs of aggressive policies, then the state give less support to
irredentist movements. Again, this expectation was largely borne out, though Serbian inaction in
the face of Croatia’s attack upon Serb-held parts of Croatia is still somewhat surprising.
To evaluate the requisites of political competition in Serbia today, more research is
required. Due to the difficulty of determining the relevance of their elections, there is a greater
level of uncertainty about the causes and consequences of Serbia’s domestic and foreign policies
than about Somalia’s.101 Still, this case does demonstrate that Serbia’s irredentism has not been as
single-minded as conventionally thought. This study suggests that when political competition was
high, Serbia supported efforts to annex or separate pieces of those republics inhabited by
“enough” Serbs.
The presence of Serbs in Slovenia and Macedonia raises an important question: how large
does an ethnic group have to be in a neighboring territory before the state becomes irredentist?
Given that Serbs were twelve percent of Croatia’s population and thirty-two percent of Bosnia’s,
there is apparently some threshold in the Serb case between two and twelve percent (or between
40,000 and 500,000).102 When that threshold is crossed, then irredentism becomes significant.

Political Competition and Foreign Policy, page 18
Another explanation is that the history of the region and the lack of Serb population centers in
Slovenia and Macedonia make irredentism towards these two states unlikely. While the absence
of historical ties matters, it is still surprising that Serb politicians would not seriously consider the
condition of Serbs in these seceding regions. Yet, another variable may have played a role. These
two seceding regions had the best human rights record and treatment of minorities policies,
according to the European Community’s commission evaluating the various applications for
recognition in the winter of 1991-1992. In sum, the lack of support for Macedonia’s and
Slovenia’s Serbs may be over-determined. More work is required to evaluate the competing
explanations.103
This study also suggests that as competition declined, and as Serbs within Serbia felt the
pain of war, Serbia’s enthusiasm for irredentism also declined. Serbia began to push the Bosnian
Serbs to compromise, and Serbia did not come to the defense of Croatia’s Serbs during Croatia’s
1995 offensive. While other factors were certainly at work, including the increased credibility of
external actors, Miloševi’s creation of an army-sized police force altered some of the constraints
he faced. He could worry less about opposition from the military. Further, this case suggests that
there are limits as to how much people care about ethnic ties—the more costly it is to support
one’s ethnic kin, the less enthusiastic one is. As the costs of the war itself and the resulting
sanctions mounted, Serbs became less interested in the war effort. Overall, though this case
provides less certainty than the Somalia case, Serbia’s inconsistent irredentism suggests that
Miloševi’s motivations for his irredentist policies were to attract the constituents of other
politicians and to keep his supporters from defecting.
14 Theoretical Findings and Policy Relevance
15 Implications for Theory
Given that the existing literature on diversionary theories of war largely focuses on how
internal conflict leads to external conflict, re-aligning the debate to consider more seriously
political competition should provide some interesting findings. Examining political competition
requires us to analyze the preferences of politicians’ potential and current supporters. Those
supporters’ preferences are one of the crucial sources of variation in foreign policy, and analyzing
them provides some insight into which states are likely to be the targets of diversionary conflicts.
The existing literature assumes that any external target improves a politician’s standing, but,
realistically, the choice of target should matter for domestic politics. By taking political
competition more seriously, we can make more determinate predictions, particularly the likely
targets of diversionary attacks, than previous approaches could. Further study, both qualitative
and quantitative, on this larger debate should therefore consider the interests of politicians’
supporters to be a significant variable.
By taking into account the preferences of politicians’ supporters, we can address a second
theoretical puzzle: why states engage in seemingly irrational foreign policies such as irredentism.
Critics argue that Neo-Realism fails to explain the poor choices states frequently make. While
Neo-realism accounts for why rational policies may produce irrational outcomes, the cases
discussed here as well as others suggest that states chose policies that are unlikely to provide
significant gains in a country’s security and are likely to backfire.104 I do not argue that policy-
makers are irrational, but that they are maximizing their own interests, and not those of the state.
Slobodan Miloševi has been quite successful, though Serbia has not been. Miloševi is still in
power despite the failure to achieve a Greater Serbia, despite the economic costs and despite
alienating much of the world. Siyad Barre’s efforts to stay in power resulted in the demise of
Somalia as a state and as a functioning society, but he did manage to stay in power much longer
than any other Somalia politician. A rational choice approach focusing on how politicians

Political Competition and Foreign Policy, page 19
strategically use foreign policy to stay in power may provide insight into other cases where states
choose costly foreign policies despite the existence of less costly alternatives.
Scholars have made much progress the past few years comparing the impact of regime
type on foreign policy with the Democratic Peace debate. This article suggests that the politics of
democracies and authoritarian regimes may not be as dissimilar as frequently argued. The threat
to leaders’ positions can cause diversionary conflicts regardless of the regime type of the initiating
state. Somalia started wars with Ethiopia to gain “lost territories” when it was a highly
competitive democracy (just before national elections that mattered), and when it was an
authoritarian regime. This finding is not an indictment of the Democratic Peace argument since
Somalia’s target was not a democracy. Further, the democratic credentials of all actors within
Yugoslavia, particularly Serbia’s and Croatia’s, were shaky at best. Still, we need to take more
seriously how states with different types of political systems may have similar political dynamics.
Likewise, we need to study how states with similar political systems may develop very
different foreign policies as competition can vary within a single form of government: democracy
or authoritarian regimes. Thus, we need to do more work to develop criteria for determining the
competitiveness of political systems. This article suggests that foreign policy analysts ought to
consider discussions of electoral systems, party systems, coalition dynamics and the like within the
Comparative Politics literature. These works may help us understand when politicians in
democracies are threatened and by whom. Likewise, considering the existing work on
authoritarian regimes may help us develop better criteria for the relative competitiveness of
different regimes. Also, there is a burgeoning literature on liberalizing regimes that might be of
use.
Regardless of institutional design, we need to determine the credibility of exit threats.
How do we know ex ante that a particular group poses a credible threat to leave a politician’s
constituency? For this article’s purposes, it was sufficient to show that particular groups did
make credible threats to defect, such as in the case of the Somali military in 1975, that the
political institutions made the desertion of supporters matter, as in the Somali elections of 1964,
or that new leaders arise after attracting supporters from incumbents, as in the case of Miloševi
in 1986-87. For future work, particularly in authoritarian and democratizing cases and for large
“n” studies, clearer coding criteria for the competition variable would need to be established. One
method to consider is to use the homogeneity of supporters and of their preferences as an
indicator, since the threat to leave is probably more credible and more intimidating when the
potential defectors agree on policy direction.
The relevance of ethnic ties is also an important finding. While it seems obvious and
perhaps trivial to argue that ethnic ties influence irredentism, this study indicates that we need to
treat this variable more seriously to understand under what conditions states will be irredentist.
Much of the debate on ethnic conflict has focused on the most simplistic notions of ethnicity:
either ancient hatreds or the creation of identity by politicians. To comprehend Somalia’s
irredentism, focusing solely on Somali identity clearly is not sufficient, and can lead to faulty
explanations and inaccurate predictions. In this case, the important identity is not religion,
language, nor race, but clan. Clan identity was crucial for identifying the interests of Somali
politicians and the likely targets of Somalia’s irredentism. In most countries, multiple ethnic
identities exist, so we need to consider why some identities matter more than others. Why do
leaders and followers care about particular identities and not others?
To understand Serbia’s irredentism, it is not enough to ask whether Serbs exist in a
particular territory, as there were Serbs in each Yugoslav republic. Perhaps a threshold exists at
which a Serb minority elsewhere is significant in Serbia’s domestic politics. This may be true for
Croats as well since Croats are 2.7% of Slovenia’s population and 4.8% of Vojvodina’s, as

Political Competition and Foreign Policy, page 20
compared to the pre-war Croat population of Bosnia—17.3%.105 A different approach is to
consider the political and economic conditions of the potentially irredentist regions. The
treatment of ethnic groups in those states should affect both the desires of such groups to join the
irredentist state and their salience in the irredentist state’s political system. Perhaps because
Macedonia and Slovenia threatened their Serb populations less, the plight of these Serbs did not
resonate in Serbia’s domestic politics. By studying more irredentist states and states refraining
from irredentism, we may determine better whether the variance in irredentism can be explained
by thresholds, the condition of the ethnic kin abroad, both or neither.
16 Implications for Policy
This study is also relevant for the foreign policy of the United States and other states as
they deal with a major source of conflict in the Post-Cold War Era—irredentism. For preventing
irredentism, the first questions are who are likely irredentists and who are likely targets? What
should the targets do to avoid war? The second set of questions considers how other actors
should respond, and the third set of implications would be recommendations to leaders seeking to
remain in power about the best targets of potential diversionary foreign policies.
Who are the irredentists of the future? We should focus our attention on states
(democratic or authoritarian) where political competition is increasing, and where the relevant
constituencies have significant ethnic ties to endangered ethnic groups in other states. Within the
Balkans, there is a significant threat of more irredentism, though necessarily not by Serbia. Since
Albanians make up 21% of Macedonia’s population, and 82% of Kosovo’s,106 there should be
much international concern about potential conflicts between Albania and Macedonia and between
Albania and Serbia over Kosovo.107 Given the economic and political dislocations in Albania
following the collapse of the many pyramid schemes, and given the growth of an armed
movement within Kosovo,108 a shift towards a more aggressive foreign policy would not be
surprising.
Likewise, it was quite reasonable for the United States and the rest of the world to worry
about Russia’s elections, as there are a large number of Russians in the other former Soviet
republics. Political competition has frequently pushed Yeltsin to take a more nationalist course.
Yet, Russia has not been actively irredentist thus far, though it has threatened countries in its
“Near Abroad” if they treat their Russian populations badly. An examination of Russia’s
domestic politics might indicate why Yeltsin’s regime has been less irredentist than expected.
Specifically, studying Yeltsin’s key constituents may suggest that some of his supporters may
have other interests, such as economic ties to the West, that constrain their interest in defending
Russians in the Near Abroad. Successors to Yeltsin may not be similarly restrained, so Russia’s
foreign policy may become irredentist in the not too distant future.
Hungary also poses a potential threat, as there is a large Hungarian minority in Romania
(Transylvania), and smaller Hungarian minorities in Slovakia and Serbia’s Vojvodina region.
Instead of attacking its neighbors, Hungary has aggressively sought agreements with Slovakia and
Romania to guarantee the security and rights of Hungarians in those countries. This may be a
pre-emptive strategy to take the irredentist issue out of the hands of potential challengers to the
ruling elites. Further study of both Russia and Hungary is required to explain more definitively
their current policies.109 This study is just a first step towards evaluating the likelihood of
irredentist conflict in these regions.
What should the potential targets of irredentist states do? First, they can try to take the
issue off of the domestic political agendas of the potentially aggressive state. While it may
difficult for domestic politics to favor or protect particular minorities, states fearing irredentism
ought to do so. While more work is needed to determine what makes a potential group a more
favored target of irredentism, improving the treatment of minorities may decrease the salience of

Political Competition and Foreign Policy, page 21
your country’s group of Hungarians, Russians, Albanians or the like. This may cause the
irredentist state to attack some other state, which may not be desirable from the standpoint of
international peace stability, but is desirable from your own perspective. Second, a potential
target of irredentism should prepare for war by arming and gaining allies, including stealing the
allies of your adversary. Ethiopia was able to rebuff Somalia’s efforts through the support of its
allies with the Soviet airlift of arms and Cuban troops in 1977-78.
What can the U.S. and others do, either beforehand to deter irredentism, or during a
conflict to end such efforts? This article suggests that domestic political concerns, rather than
international threats, motivate politicians, and the cases provide evidence for this. Somali
politicians were relatively insensitive to international costs, compared with the domestic political
costs of not pursuing selectively the Greater Somalia project. Miloševi and Serbia have also
seemed insensitive to international disapproval although economic sanctions almost certainly
caused Miloševi to compromise on his Greater Serbia project due to their domestic political
consequences. Arguing that the economic sanctions had some effect in moderating Serbian
foreign policy suggests that the West could have done more at the outset of the conflict,
preventing much of the horror that was to follow. However, this is not necessarily true, as the
threat of sanctions is probably less influential than the accumulated impact of actual sanctions.
The external threat of sanctions may have strengthened the nationalists at the outset, as it justified
their paranoid claims about being persecuted. Consequently, the increased sense of encirclement
may have caused Serbs to support more strongly the irredentist cause. Serbia’s population may
have changed their preferences over time, but only as they bore the economic costs quite directly.
Thus, while threats may backfire, actually changing the living conditions of a politician’s
constituency may lead to the desired policy changes. Of course, this raises an essential problem
for economic sanctions—targeting them to hurt the decision-maker’s supporters, rather than the
population at large.110
If political competition for members of individual ethnic groups influence states’ foreign
policies, then the most logical policy recommendations are: (1) reduce the level of political
competition; (2) force the politicians to rely on the support of multiple ethnic groups; or (3) alter
the preferences of the relevant constituents. The first two options represent competing strands of
a debate in comparative politics. Do you develop institutions that ensure that most major ethnic
groups share power, and therefore do not have to compete for votes?111 Alternatively, can
institutions compel politicians to appeal to more than one ethnic group, making ethnic appeals a
less successful strategy?112 The latter strategy is problematic as many potentially irredentist states,
such as Russia and Hungary, are predominantly of a single ethnic group, so it would be quite
difficult to force politicians to appeal to multiple ethnic groups. These two approaches both
suggest changes in constitutions and electoral laws. This implies that American foreign policy
should aim at reforming countries’ political systems so that ethnic politics within such states
produces more peaceful politics and less aggressively nationalist foreign policies. However, the
time for this approach may have passed as the constitutions and political institutions of the states
that were formerly part of the Soviet empire are no longer being designed, but have been put into
place.
The third option, altering the preferences of the relevant constituencies, is not a very
straightforward proposition.113 Attempts to manipulate or influence public opinion can backfire, as
the Serbs have not used the information available from the Voice of America and other external
sources to evaluate critically Miloševi’s claims. Threatening to sanction Serbia would not have
been sufficient because politicians can use external threats to justify their policies and divert
blame; and there were other aggressive actors, such as Croatia, stoking the flames of ethnic
conflict.

Political Competition and Foreign Policy, page 22
One implication of this approach is that we need to consider more seriously whether and
how external actors can influence the preferences of a politician’s supporters before a conflict
starts. Western policy towards the Baltic Republics has focused on making access to western
markets, particularly the European Union, contingent on treating their Russian populations well.
Consequently, Russian voters and other important supporters care less about the plight of
Russians in the Baltics, keeping this nationalist issue off the Russian political agenda and out of
the hands of extremist politicians. The US and the European Union are also trying to change the
preference ordering of the supporters of politicians in the former Soviet space (East Europe and
the former Soviet Union) by improving the economic situations in these countries. The idea is
that if the relevant constituents in these countries are happy with their current situation, they will
be less likely to risk war and the loss of markets for the defense of their ethnic kin. Hungary’s
accommodation with its neighbors suggests that this might work, but it is too soon to tell.
If the U.S. and others cannot stop states from engaging in aggression because such states
are motivated by domestic politics, the only remaining recourse is to arm and perhaps defend the
targets of irredentism. If states use war to annex territories inhabited by ethnic kin, then outsiders
can try to prevent such efforts from succeeding.
This article also suggests that politicians seeking to remain in power ought to consider the
ethnic ties of their most important supporters, and if needing a boost in support, should support
aggressive actions against countries inhabited by the constituents’ ethnic kin.114 Rather than
randomly choosing targets, politicians ought to take care when choosing the targets of
diversionary wars.
The relationships between identity, politics, and foreign policy are complex, as the New
World Disorder has proved. By focusing on ethnic ties and competition among elites for the
support of particular ethnic groups, this article has provided some explanations of the self-
destructive foreign policies of Somalia and Serbia and the possibility of fruitful future directions
for research into the issues of political competition’s impact on foreign policy and of irredentism.

Political Competition and Foreign Policy, page 23
Table 1: Systemic Theories, Varying Irredentism
Case External
Threat:
Vulnerability to
Separatism:
Specification of
Norms, Level of
International
Organization
Actual Policy:
Somalia,
1960-67 High High Increasing
Strong Support for
Separatists in Kenya,
Ethiopia
Somalia,
1967-69 High Low High No Support
Somalia,
1970-74 Low Low High No Support
Somalia,
1975-80 Low High High
Strong Support for
Separatists in
Ethiopia
Serbia,
1991-93
Increasing High Increasing Strong Support for
Serbs in Croatia,
Bosnia
Serbia,
1994-97
High High High No Support of
Serbs in Croatia,
weak Support for
Serbs in Bosnia

Political Competition and Foreign Policy, page 24
Table 2: Somalia’s Inconsistent Irredentism
Time
Period
Constituents Having
Ethnic Ties To Somalis
In (Politically Relevant
Clans):
Competition: Predictions: Actual Policy:
1960-67
Kenya, Ethiopia
(Darood including
Ogaden; Hawiya)
High
Strong support for
Separatists in
Kenya, Ethiopia
Strong Support for
Separatists in Kenya,
Ethiopia
1967-69 Ethiopia (Ishaak,
Majertain) Low No Support No Support
1970-74 Ethiopia, Kenya,
Djibouti (all major clans) Low No Support No Support
1975-80 Ethiopia (Marehan,
Dolbohanta, Ogaden) High
Strong Support
Separatists in
Ethiopia
Strong Support for
Separatists in
Ethiopia

Political Competition and Foreign Policy, page 25
Table 3: Serbia’s Inconsistent Irredentism, 1991-1995
Seceding
Republic Ethnic Ties: Competition: Predicted Policy: Actual Policy:
Slovenia Weak High, then Low Weak support
changing to none
Weak support
changing to none
Croatia Strong High, then Low Strong support
changing to weak
Strong support
changing to none
Bosnia Strong High, then Low Strong support
changing to weak
Strong support
changing to weak
Macedonia Weak High, then Low Weak support
changing to none No support for Serbs

Political Competition and Foreign Policy, page 26

1 One of the few discussions of this topic can be found in Naomi Chazan, ed., Irredentism and
International Politics (Boulder: Lynne Rienner, 1991). David Carment and Patrick James have
recently analyzed irredentism, taking seriously some of the same variables that are treated here, but
asking somewhat different questions and reaching different conclusions, “Internal Constraints and
Interstate Ethnic Conflict: Toward a Crisis-Based Assessment of Irredentism,” Journal of Conflict
Resolution 39, no. 1 (March 1995): 82-109.
2 For the most recent work on diversionary theories of war, see T. Clifton Morgan and Kenneth N.
Bickers, “Domestic Discontent and the External Use of Force,” Journal of Conflict Resolution 36,
no. 1 (March 1992): 25-52; and Ross A. Miller, “Domestic Structures and the Diversionary Use of
Force,” American Journal of Political Science 39, no. 3 (August 1995): 760-785.
3 James Lee Ray, Democracy and International Conflict: An Evaluation of the Democratic Peace
Proposition (Columbia: University of South Carolina, 1995).
4 Ray, Democracy and International Conflict.
5 Irredentism can also refer to efforts by political movements aiming at uniting a “lost territory,” but
the purpose of this paper is to explain the policies of states, not of sub-state or non-state actors.
State support can range from giving arms and logistical support to separatist groups in one’s
neighbors to invasions to conquer territory inhabited by the ethnic brethren. Irredentism can also
refer to efforts by political movements aiming at uniting a “lost territory,” but the purpose of this
paper is to explain the policies of states, not of sub-state or non-state actors.
6 Irredentism refers only to expansion of a state to include territories to which a country has ethnic
and historical claims. Once a state expands the conquest beyond such territories, the policy is not
longer one of irredentism. For instance, Hitler’s foreign policy was irredentist as long as it focused
on territories inhabited by German-speaking peoples such as the Sudetenland, but became something
else when he invaded the rest of Czechoslovakia.
7 Jack Snyder does in his work what I hope to accomplish here—explain behavior that generates
threats that causes other states to respond. Neo-realism and balance of threat theory excel at
explaining why states balance rather than bandwagon to security threats, but fail to explain why

states engage in behavior that produces counter-balancing coalitions. See Jack Snyder, Myths of
Empire: Domestic Politics and International Ambition (Ithaca: Cornell University Press, 1991).
8 For more on Soviet foreign policy in this region, see Steven R. David, Choosing Sides: Alignment
and Realignment in the Third World (Baltimore: Johns Hopkins University Press, 1991).
9 I am not denying that there were some security interests in expansion (shortening the routes
between northern and southern Somalia; access to many of Yugoslavia’s arms factories which were
in Bosnia), but that decision-makers should not have expected war to be the best way to achieve
these interests. As discussed below, peace with Ethiopia allowed Somalia to have easier access
between its northern and southern parts.
10 Barry Posen, “Nationalism, the Mass Army, and Military Power,” International Security 18, no. 2
(Fall 1993): 80-124.
11 Further, Serbia’s foreign policy was most consistent during the fall of 1991 despite the most
significant change in the international distribution of power—the disintegration of the Soviet Union,
Serbia’s most likely ally.
12 In another article, Barry Posen argues that security dilemmas within failing empires cause ethnic
groups to act much like states in the international system, engaging in pre-emptive wars, “The
Security Dilemma and Ethnic Conflict,” Survival 35, no. 1 (Spring 1993): 27-47. This approach may
well explain the breakup of Yugoslavia and the behavior of groups within Croatia and Bosnia as each
ethnic group was threatened by others, causing all to consider pre-emptive actions. However,
Posen’s argument cannot explain Serbia’s foreign policy once Croatia and later Bosnia became
independent. Neither Croatia nor Bosnia posed severe threats to Serbia’s security because the two
new states lacked the offensive capabilities that Posen considers to be important. For Serbia’s
foreign policy to be motivated by insecurities of Croatian and Bosnian Serbs, Posen has to argue that
the security of ethnic groups is indivisible—that all Serbs are threatened if some are. Posen does not
make that step, nor would that be consistent with his other work. Further, his argument cannot
explain Somalia’s foreign policy behavior, Somalia’s internal anarchy occurred long after its
irredentism. Still, Posen’s security dilemma might be able to account for the clan conflicts of the late

1980’s and 1990’s. For a somewhat different approach to ethnic security dilemmas, see author.
13 For the best expression of this argument, see Jeffrey Herbst, “The Creation and Maintenance of
National Boundaries in Africa,” International Organization 43, no. 4 (Autumn 1989): 673-692. See
author as well.
14 My thanks to an anonymous reviewer for pushing me on this point.
15 This also suggests that Posen’s argument cannot explain irredentism since nationalism is his
intervening variable between external threats and foreign policies, “Nationalism, the Mass Army, and
Military Power.”
16 Lewis Coser, The Functions of Social Conflict (Glencoe, IL: Free Press, 1956), 87.
17 John Mueller, Wars, Presidents, and Public Opinion (New York: Wiley, 1973).
18 See Jack Levy’s review of this literature, “The Diversionary Theory of War: A Critique,” in
Handbook of War Studies, ed. Manus I. Midlarsky (Boston: Unwin Hyman, 1989).
19 Morgan and Bickers, “Domestic Discontent,” 28.
20 Ibid., 33. Actually, Morgan and Bickers leave open the possibility that leaders can either divert or
mollify those within the ruling coalition.
21 Ibid., 32.
22 Edward D. Mansfield and Jack Snyder, “Democratization and the Danger of War,” International
Security 20, no. 1 (Summer 1995), 20.
23 Ibid., 20.
24 Snyder, Myths of Empire.
25 Jack Snyder and Karen Ballentine, “Nationalism and the Marketplace of Ideas,” International
Security 21, no. 2 (Fall 1996): 5-40.
26 V.P. Gagnon, Jr., “Ethnic Nationalism and International Conflict: The Case of Serbia,”
International Security 19, 3 (Winter, 1994/95): 135-137.
27 Ibid., 162-164.
28 Ibid., 135.
29 Ibid., 140.

30 T. Clifton Morgan and Sally Howard Campbell, “Domestic Structure, Decisional Constraints, and
War: So Why Kant Democracies Fight?” Journal of Conflict Resolution 35, no. 2 (June 1991), p.
191.
31 Alfred Stepan, Rethinking Military Politics: Brazil and the Southern Cone (Princeton: Princeton
University Press, 1988).
32 Albert O. Hirschman, Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations,
and States (Cambridge: Harvard University Press, 1970).
33 Donald Horowitz, “Incentives and Behaviour in the Ethnic Politics of Sri Lanka and Malaysia,”
Third World Quarterly 10, no. 4 (1989), pp. 18-35.
34 Arend Lijphart, Democracy in Plural Societies (New Haven: Yale University Press, 1977).
35 This avoids the circularity problem of determining the politician’s key constituents by considering
the foreign policy of the state, and then using this finding to “explain” the state’s foreign policy.
36 Because ethnicity is perceptual, it is sometimes difficult to say when ethnic ties exist or are
perceived. Focusing on the attributes of race, language, religion and kinship helps.
37 It is also the case that politicians will tend to oppose those groups internationally that share a
history of ethnic enmity with their constituents. This plays out in the case of Greece’s policies
towards Macedonia.
38 For an excellent discussion of the literature on public opinion and foreign policy, see Ole R. Holsti,
“Public Opinion and Foreign Policy: Challenges to the Almond-Lippman Consensus,” International
Studies Quarterly 36, no. 4 (December 1992), pp. 439-466.
39 In reality, individuals within any ethnic group vary widely in how strongly they feel about those
with whom they share ethnic ties. For the purposes of this article, I assume that members of an
ethnic group care enough about the condition of their ethnic kin to cause them to change whom they
support politically. Future research can relax this assumption and perhaps evaluate it through
surveys.
40 Donald Horowitz, Ethnic Groups in Conflict (Berkeley: University of California Press, 1985),
especially chapter 4.

41 For an historical discussion of the impact of ethnic ties upon U.S. foreign policy, see Alexander
DeConde, Ethnicity, Race, and American Foreign Policy: A History (Boston: Northeastern
University Press, 1992).
42 See Gagnon, “Ethnic Nationalism and International Conflict,” for a discussion of Slobodan
Miloševi’s efforts to use events in Kosovo and other parts of Yugoslavia to set the political agenda
away from economic problems and political reform.
43 In the future, we can relax or examine this assumption.
44 See John Stuart Mill, Philosophy of Scientific Method, edited by Ernest Nagel (New York:
Hafner, 1950), 211-233; and Theodore W. Meckstroth, “‘Most Different Systems’ and ‘Most
Similar Systems’: A Study in the Logic of Comparative Inquiry,” Comparative Political Studies 8,
no. 2 (July 1975): 132-157.
45 Given the destructiveness of the war and the damage inflicted by Miloševi’s economic policies, it
is hard to remember that Yugoslavia was once considered one of the most advanced and reformed
economies of the socialist world.
46 Gary King, Robert O. Keohane, Sidney Verba, Designing Social Inquiry (Princeton: Princeton
University Press, 1994), 202.
47 Donald L. Horowitz, “Irredentas and Secessions: Adjacent Phenomena, Neglected Connections,”
in Irredentism and International Politics, 14.
48 Somalia is also a challenging test of the ethnic politics approach because Somalia’s irredentism
varied although Somali elites had ethnic ties with Somalis in all three of the neighboring states
throughout the period being studied, including shared language and shared religion.
49 See Map 1: Distribution of Somalia’s Clans and Clan-Families. The subtleties of Somalia’s clan
politics are not entirely captured by Map 1, which indicates that the Ogadeni clan would be
interested in Kenya’s Northern Frontier District. However, not all diya-paying groups or sub-clans
have strong ties to the clans and clan-families to which they belong. David D. Laitin and Said S.
Samatar, Somalia: Nation in Search of a State (Boulder: Westview Press, 1987), 135-136.

50 Umorzurike Oji Umozurike, Self-Determination in International Law (London: Archon Press,
1972), 226.
51 Keith Somerville, Foreign Military Intervention in Africa (London: Pinter Publishers, 1990), 46.
52 Somalia’s society is characterized by segmental lineages. The population is divided into clan-
families, each consisting of several clans; in turn, each clan contains many sub-clans or lineages, and,
each of these is composed of many diya-paying groups—the fundamental kinship unit in Somalia.
See I. M. Lewis, The Modern History of Somalia: Nation and State in the Horn of Africa (Boulder:
Westview, 1965), 11.
53 The transliterations used here are consistent with the Map 1.
54 Lewis, Modern History, 6. These population figures and those below are estimates made during
the early 1960’s.
55 David D. Laitin, Politics, Language, and Thought: The Somali Experience (Chicago: University
of Chicago Press, 1977), 91.
56 Saadia Touval, “The Sources of Status Quo and Irredentist Policies,” in African Boundary
Problems, ed.Carl Gosta Widstrand (Uppsala: Scandinavian Institute of African Studies, 1969), 116.
57 John Markakis, National and Class Conflict in the Horn of Africa (Cambridge: Cambridge
University Press, 1987), 233.
58 Lewis, Modern History, 6.
59 Markakis, National and Class Conflict, 181.
60 Laitin and Samatar, Somalia, 76. My understanding of Somalia’s electoral system benefited from
conversations with Mathew Shuggart.
61 Attempts by both major parties to support irredentism towards both Kenya and Ethiopia is similar
to the dynamics of log-rolling and overexpansion in Snyder, Myths of Empire.
62 Saadia Touval, Somali Nationalism: International Politics and the Drive for Unity in the Horn of
Africa (Cambridge: Harvard University Press, 1963), 98.
63 Lewis, Modern History, 186.
64 Lewis, Modern History, 200.

65 Fantu Agonafer, Djibouti's Three-Front Struggle for Independence: 1967-77 (Ph.D. Dissertation,
University of Denver, 1979), 101-103.
66 Ahmed I. Samatar, Socialist Somalia: Rhetoric and Reality (London: Zed Books, 1988), 67.
67 The SYL’s access to government funding had a perverse effect on the politics of Somalia. It
forced opposition candidates to rely even more on their sub-clans for the finances required to win
office. The winning candidates, in debt to their families, were then quite willing to cross over to the
SYL in exchange for money to pay off their debts. Colin Legum and John Drysdale, eds., Africa
Contemporary Record: Annual Survey and Documents, 1969-1970 (Exeter, UK: Africa Research,
1970): B176.
68 For a discussion of Egal’s new foreign policy, see Raymond Thurston, “Détente in the Horn,”
Africa Report 14, no. 2 (April 1969): 6-13.
69 Africa Confidential 8, no. 22 (November 3, 1967): 2.
70 Tom J. Farer, War Clouds on the Horn of Africa: The Widening Storm (New York: Carnegie
Endowment for International Peace, 1979), 108.
71 Robert Tholomier, Djibouti: Pawn of the Horn of Africa, trans., Virginia Thompson and Richard
Adloff (Metuchen, NJ: Scarecrow, 1981), 83.
72 I. M. Lewis, “Kim Il-Sung in Somalia: The End of Tribalism,” in William A. Shack and Percy S.
Cohen, eds., Leadership: A Comparative Perspective (Oxford, UK: Clarendon Press, 1979), 16.
73 Richard Greenfield, “Siad's Sad Legacy,” Africa Report 36, no. 2 (April 1991): 14. Somalis were
frustrated with clan politics, but continued to be loyal to their clan. This is similar to the
phenomenon in the United States where people hate Congress and want to get rid of the incumbents,
but have continued (at least until 1994) to vote for their own incumbent time after time.
74 Bereket Habte Selassie, Conflict and Intervention in the Horn of Africa (New York: Monthly
Review, 1980), 108.
75 Siyad Barre, My Country and My People: Selected Speeches of Jaalle Major-General Mohamed
Siyad Barre (Mogadishu: Somali Ministry of Information and Guidance, 1974).
76 Tholomier, Djibouti, 85.

77 Colin Legum, ed., Africa Contemporary Record: Annual Survey and Documents, 1970-1971
(Exeter, UK: Africa Research, 1971), B163.
78 Laitin and Samatar, Somalia, 140. See also David D. Laitin, “The War in the Ogaden:
Implications for Siyaad’s Role in Somali History,” Journal of Modern African Studies 17, no. 1
(March 1979): 112.
79 Laitin and Samatar, Somalia, 140.
80 I. M. Lewis, "The Ogaden and the Fragility of Somali Segmentary Nationalism," African Affairs
88, no. 315 (1989): 575.
81 Samatar, Somalia: A Nation in Turmoil, 18.
82 For instance, see Christopher Bennett, Yugoslavia’s Bloody Collapse: Causes, Course and
Consequences (New York: New York University Press, 1995); and Ivo H. Daalder, “Fear and
Loathing in the Former Yugoslavia,” in The International Dimensions of Internal Conflict, ed.
Michael E. Brown, (Cambridge: MIT Press, 1996).
83 Before the de facto secession of Croatia and Slovenia, Serb efforts to annex their territories is not
formally irredentism, but the secessions caused what was inter-republican policy to become foreign
policy.
84 In 1985 (and leaked in September 1986), the Serbian Academy of Sciences drew up a plan of a
Serb-dominated Yugoslavia that would include Serbia, Montenegro, much of Bosnia-Hercegovina,
the predominantly Serb areas of Croatia, but would exclude Macedonia, Slovenia, most of Croatia,
and would cede part of Bosnia-Hercegovina to Croatia as compensation. Bennett, Bloody Collapse,
81.
85 Patrick Moore, “Commentaries,” 16.
86 “Yugoslavia: Death of a Nation; III. The Collapse of Unity,” Discovery Journal Television
Broadcast (December 28, 1995).
87 It is important to note that Miloševi’s enthusiasm for irredentism declined long before NATO
engaged in its bombing campaign in 1995.

88 Susan L. Woodward, Socialist Unemployment: The Political Economy of Yugoslavia, 1945-1990
(Princeton: Princeton University Press, 1995); and Milica Zarkovic Bookman, Economic Decline
and Nationalism in the Balkans (New York: St. Martin’s Press, 1994).
89 See Miroslav Pecujlic, “University zacoaronom krugu politike,” (“In the Spelled Circle of
Politics”) in Radanje Javnog Mnjenna I Politickih Stranaka- Analiza empirijskih istrazivanja u Srbiji
1990-1991 (Emergence of Public Opinion and Political Parties: An Analysis of Empirical Surveys in
Serbia, 1990-1991), (Belgrade: Institute for Political Studies, 1992), 75-79; as well as Susan L.
Woodward, Balkan tragedy: Chaos and Dissolution After The Cold War (Washington, D.C.: The
Brookings Institution,1995), 93.
90 James Gow, Legitimacy and the Military: The Yugoslav Crisis, (New York: St. Martin’s Press,
1992), 142.
91 Lenard J. Cohen, Broken Bonds: Yugoslavia’s Disintegration and Balkan Politics in Transition,
2nd ed., (Boulder: Westview Press, 1995), 181-188.
92 Stuart J. Kaufman, “The Irresistible Force and the Imperceptible Object: The Yugoslav Breakup
and Western Policy,” Security Studies 4, no. 2 (Winter 1994/95), 290.
93 This electoral system produced disproportionate outcomes in Croatia as well. Vladimir Goati,
“Visepartijski mozaik Srbije,” (Multiparty Mosaic of Serbia), in Radanje Javnog Mnjenna I Politickih
Stranaka (Emergence of Public Opinion and Political Parties), 171; and Cohen, Broken Bonds, 157-
158.
94 Cohen, Broken Bonds, 230-31, 266. Laura Silber and Allan Little, Yugoslavia: Death of a Nation
(TV Books, Distributed by Penguin United States, 1996), 167, 242.
95 Vladimir Milic, “Socijalni lik politickog javnog mnjenja,” (Social Face of Political Public Opinion”
in in Radanje Javnog Mnjenna I Politickih Stranaka (Emergence of Public Opinion and Political
Parties), 132.
96 Milan Andrejevich, “The Radicalization of Serbian Politics,” RFE/RL Research Report, 2, no. 13
(March 26, 1993): 24. Also, see Nicholas J. Miller, “Serbia Chooses Aggression,” Orbis 38, no. 1
(Winter 1994): 59-66.

97 Aleksa Djilas, “Fear Thy Neighbor: The Breakup of Yugoslavia,” in Nationalism and
Nationalities in the New Europe, ed. Charles A. Kupchan (Ithaca, NY: Cornell University, 1995),
100-101.
98 In regimes under transition, politicians prefer to win elections, as the costs of remaining in power
after a lost election increase. Thus, losing an election may be acceptable, but is certainly not
desirable. Remaining in power after a lost election may require more resources and certainly raises
the level of uncertainty and risk for the incumbent. It may cause the politician to become more
dependent on those key constituents who allow the election results to be thrown out, such as the
police. Politicians, like Miloševi, will try to win elections and their foreign policies are shaped by
the desire to win elections, but if a politician can survive an electoral loss, their foreign policy should
also reflect this. The preferences of voters should matter less, and the preferences of those who keep
the politician in power should matter more. Therefore, the fact that Miloševi and others are not
entirely bound by elections does not mean that political competition does not shape foreign policy,
but that political competition can vary.
99 Vesna Pusic, “Dictatorships with Democratic Legitimacy: Democracy Versus Nation,” East
European Politics and Society, 8 (Fall 1994): 396, 399.
100 Woodward, Balkan Tragedy, 27, 254, 262, 293.
101 King, et al., Designing Political Research, recommend that scholars indicate the relative
uncertainty of their findings.
102 It is not clear whether it is relative size of the irredentist population or its absolute size that might
matter.
103 I am grateful to both Ellen Comisso and Stuart Kaufman for challenging me on this point.
104 While it might be rational for a state to start a war if it sees the status quo as being worse than
fighting and losing a war, the situations described in this paper do not fit into that category.
105 Ruza Petrovic, “The National Composition of Yugoslavia’s Population,” Yugoslav Survey 33, no.
1 (1991): 4-12.
106 Petrovic, “The National Composition of Yugoslavia’s Population,” 4-12.

107 Misha Glenny, “Heading Off War in the Southern Balkans,” Foreign Affairs 74, no. 3 (May/June
1995): 98-108.
108 Misha Glenny, “Bosnia II?” New York Times 9 December 1997, A29.
109 For one such study, see Patrice McMahon, The Quest for Unity: Divided Nations and Irredentist
Ambitions, Ph.D Dissertation, Columbia University, 1998.
110 For recent discussions of sanctions success or failure, see Jonathan Kirshner “The
Microfoundations of Economic Sanctions” and Daniel Drezner, “Economic Coercion and Russion
Foreign Economic Policy” in Security Studies 6, no. 3 (Spring 1997); and Robert Pape, “Why
Economic Sanctions Do Not Work,” International Security 22, no. 2 (Fall 1997): 90-136
111 This is the consociational approach, best represented by the works of Arend Lijphart, including
Democracy in Plural Societies.
112 This is the vote pooling approach, advocated by Donald Horowitz, Ethnic Groups in Conflict,
and A Democratic South Africa? Constitutional Engineering in a Divided Society (Berkeley:
University of California Press, 1991).
113 Even arguments that recommend inducing cooperation through incentives are skeptical about their
efficacy. See Stuart J. Kaufman, “Spiraling to Ethnic War: Elites, Masses, and Moscow in
Moldova’s Civil War,” International Security 21, no. 2 (Fall 1996): 108-138.
114 While I personally do not advocate war as strategy for doing well in domestic politics, it is a
logical implication of the argument presented here.