Regimes and peace processes: Democratic (non)development in Armenia
and Azerbaijan and its impact on the Nagorno-Karabakh conﬂict
Nina Caspersen
Department of Politics, Philosophy and Religion, Lancaster University, United Kingdom
a r t i c l e i n f o
Article history:
Available online 7 April 2012
Keywords:
Conﬂict resolution
Democratisation
Outbidding
Fragmentation
Karabakh
a b s t r a c t
The lack of democratisation in Armenia and Azerbaijan is by many observers argued to
constitute a key obstacle to the Nagorno-Karabakh peace process, but other observers
contend that the problem is that the respective leaderships are not sufﬁciently secure to
agree to a difﬁcult compromise. This article, however, ﬁnds that the relationship between
regime types and conﬂicts is a lot more complex than is often argued in the literature, and
that the Karabakh peace process has been undermined by the worst of two worlds: intense
elite competition, but without the restraint and widened participation that democratisation
could provide.
Ó 2012 The Regents of the University of California. Published by Elsevier Ltd. All rights
reserved.
The Armenian enclave of Nagorno-Karabakh gained de facto independence in 1994, after a bloody two-year war with
Azerbaijan, yet it has failed to gain international recognition. Continuous attempts have been made to reach a peaceful
settlement, but all have been fruitless and the conﬂict has for years appeared stuck in an immutable stalemate. One of the
explanations for this lack of progress points to the type of regimes found in Armenia and Azerbaijan. Laurence Broers (2005),
for example, argues that the two regimes have promoted a culture of homogenizing militancy which now leaves them little
room for difﬁcult compromises on the Karabakh issue. The solution to this problem, he contends, is to be found in democratisation:
“only democratic dividends deriving from improved state–society relations may ultimately furnish the necessary
resources to break the current impasse in the peace process”. Several other observers reach similar conclusions. Arzu
Abdullayeva, argues that the leaders have monopolised the peace process and concludes that the lack of democracy in both
Armenia and Azerbaijan constitutes the greatest obstacle to achieving peace.1
The regimes are argued to feel compelled to reify a hard-line stance on the Karabakh issue, since they are not sufﬁciently
strong in their domestic mandates (Broers, 2005). But this could also lead to the conclusion that less, not more, democracy is
needed. It is, for example, often argued that no progress is possible when elections are looming (Babayan and Perch, 2006),
and observers frequently point to the political insecurity especially of the Armenian president as a signiﬁcant obstacle to
a settlement (Cornell, 2011: 137). This has, to a large extent, been the approach adopted by the international mediators,
resulting for example in the muted international criticism of the 2008 Armenian presidential election. It has, in any case, been
questioned if the democratisation that some observers are calling for is even possible in the context of war legacies and
unresolved conﬂict.
The seeming inability of the two regimes to agree to a compromise settlement can therefore be construed both as an
argument for and against further democratisation. This debate has a parallel in the theoretical debates over the effect of
democratisation on conﬂict and conﬂict resolution. After the end of the Cold War, democratisation was seen almost as
1
Author’s interview (Baku, 12 June 2009).
Contents lists available at SciVerse ScienceDirect
Communist and Post-Communist Studies
journal homepage: www.elsevier.com/locate/postcomstud
0967-067X/$ – see front matter Ó 2012 The Regents of the University of California. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.postcomstud.2012.03.005
Communist and Post-Communist Studies 45 (2012) 131–139
a panacea for peace and prosperity, but optimism was quickly replaced by pessimism. Rather than promoting peace and
stability, a number of authors argued that democratisation breeds instability and insecurity (Mansﬁeld and Snyder, 1995;
Paris, 2004). When it comes to conﬂict resolution, many analysts similarly point to the need for a strong leader who is able to
bring followers along, marginalise spoilers and agree to a peace agreement without the risk of outbidding from more extreme
rivals. This would indicate that democratic, and in particular transitional leaders, would represent an obstacle. Other analysts,
however, disagree: if a settlement is to stick, then there is a need for a representative leader who enjoys popular legitimacy.
These questions tend to be under-analysed in existing literature and much of the debate has been based on general
assumptions of ethnic outbidding. Moreover, there is often insigniﬁcant attention paid to the ways in which the regime is
itself affected by the conﬂict and war. The experience of war has, as Shale Horowitz (2003) has demonstrated, a signiﬁcant
impact on subsequent attempts to democratise, and the relationship between regime types and armed conﬂicts is therefore
more complex than what is often assumed in the literature. This article ﬁnds that the Karabakh peace process has suffered
from the worst of two worlds: intense competition, especially in Armenia, but without the restraint and widened participation
that democratisation can provide.
Regime types and conﬂict
Edward Mansﬁeld and Jack Snyder (1995: 5) signiﬁcantly qualiﬁed the democratic peace thesis (Maoz and Abdolali, 1989)
when they found that “in the transitional phase of democratization, countries become more aggressive and war-prone, not less”. In
the transition from one regime to another, they argued, elite competition is likely to be intense and competitors are likely to
“play the ethnic card” in order to mobilise popular opinion. Since institutions are still being established, there is no one to rein
them in (Mansﬁeld and Snyder, 1995). This deceptively simple explanation has gained widespread popularity: democratisation
is dangerous, it risks spreading instability and war (Chua, 2004). It has, however, been challenged by more recent
studies which ﬁnd that democratisation reduces the risk of war, be it intra-state or inter-state, and that only certain forms of
transitions are associated with a higher risk of violence. Michael Ward and Kristian Gleditsch (1998), for example, found that
only rocky or especially rapid transitions or reversals are associated with warfare, while the Political Instability Task Force
found that “partial democracies” are particularly prone to instability (Goldstone et al., 2005). But this more nuanced view has
also been challenged. James Raymond Vreeland (2008) argues that even the relationship between ‘partial democracy’ and
civil war disappears if the experience of political violence is removed from the deﬁnition of the former.
The relationship between regime types and the outbreak of war therefore remains contested, and another aspect of the
effect of regime types on the dynamics of a conﬂict has been left under-analysed: How do different regime types affect the
prospect of conﬂict resolution? This is not about the risk of fragmentation or polarisation as such, but about intra-communal
relations: what is the nature of internal divisions and what kind of relationship between leaders and followers do we ﬁnd?
Effect of regime types on conﬂict resolution
There is a growing acknowledgement in the literature that internal structures or dynamics matter for the development of
intra-state conﬂicts and wars; for their outbreak, conduct and duration (King, 2004: 434). Although the effect of regime types
on the prospect of conﬂict resolution remains largely neglected, it is possible to identify three competing views: democratisation
makes a settlement less likely, more likely, or has no signiﬁcant effect.
Most literature would argue that any form of democratisation will have a negative impact on the prospect for a peaceful
settlement. Kathleen Cunningham et al. (forthcoming), for example, ﬁnd that the more fragmented a separatist group is, the
more violent a conﬂict is likely to be, and such fragmentation is positively associated with higher degrees of democracy. The
negative effect of democratisation is based on internal competition for hegemony, which not only appears to make the conﬂict
more violent but also promotes political radicalisation (Tezcür, 2010) and makes it harder for an incumbent leader to accept
a compromise solution. A leader who chooses to accept a settlement will fear being outbid by more extreme rivals who can
and will play “the ethnic card” and take advantage of popular antagonisms and fears (Rabushka and Shepsle, 1972; Horowitz,
1985; Mitchell, 1995). Spoilers who seek to undermine a settlement (Stedman, 1997) are in a sense empowered by the
existence of free competition; they do not even have to resort to violence to bring it down. Based on this, one could be
tempted to conclude that what is needed for conﬂict resolution to succeed is strong leadership (Nordlinger, 1972), and this
may well come in an authoritarian form.
Outbidding should, however, not automatically be assumed. Not all spoilers are successful, not all populations are
antagonised. Michael Findley and Peter Rudloff (2009) actually ﬁnd that increased fragmentation makes a negotiated solution
more likely: a split allows for moderates to push forward, while radicals may be left marginalised. And Andrew Kydd and
Barbara Walter (2002) argue that democracy, including the holding of referendums, can serve to undermine spoilers. It is
also worth remembering that an authoritarian leader cannot necessarily be assumed to have the necessary leeway to agree to
a difﬁcult compromise. Popular constraints may not be signiﬁcant, but the continued power of an authoritarian leader still
depends on being able to bring followers and supporters, such as the military, along to a new position. It is in any case, one
thing to reach a compromise agreement; it is quite another to make it stick, and this may be very difﬁcult with a narrow
agreement, which is essentially based on the will of a single authoritarian leader. Some authors would therefore argue that
there is a need for broadening the process, for including other levels of society and thereby prepare them for a compromise
solution (Lederach, 1998). Such a strategy is much more likely to be chosen if political reforms have been introduced. The
N. Caspersen / Communist and Post-Communist Studies 45 (2012) 131–139132
effect of democratisation, moreover, goes beyond its impact on leadership security and can have a more indirect effect on
settlement negotiations. Democratisation brings with it a degree of pluralism that signiﬁcantly affects internal power relations;
it has the potential to move the conﬂict parties away from military logics and zero-sum thinking.
The effect of democratisation on the prospect of conﬂict resolution therefore remains unclear. Is the prospect for a peaceful
settlement undermined by intra-communal violence and electoral outbidding or is it improved by a widened process
involving leaders with a clear popular mandate?
Democratisation in the context of war and unresolved conﬂict
Now it may be argued that any attempt to analyse the effect of democratisation on settlement talks is a misnomer, since
democratic procedures are likely to be dispensed with following the outbreak of violence. But we actually ﬁnd considerable
variation: in some conﬂicts, elections are still held and the leaders ﬁnd themselves somewhat constrained by popular opinion.
This is particularly likely in a situation of “no war, no peace” (Walker, 1998) such as that which characterises the conﬂict over
Nagorno-Karabakh; warfare has, in these cases, ceased but the conﬂict remains unresolved and may well revert to violence.
Even if the violence has ceased, or a temporary lull is experienced, the legacy of war and the persistence of an unresolved
conﬂict is, however, likely to signiﬁcantly constrain political competition. It is therefore not enough to look at the internal
dynamics, we must also analyse how these are affected by the ongoing conﬂict and war (Kenny, 2010: 552). Horowitz (2003)
argues that the negative impact of war on political and economic reforms works through three main mechanisms: 1)
distraction from any peacetime political and economic reform agenda; 2) military defeat and disruption and associated
weakening and militarization of state authority and 3) post-war economic isolation. War is likely to have empowered the
military and concentrated power in the hands of the president, while the opposition often remains weak (Curtis and De
Zeeuw, 2010). The experience of war, it is assumed, can radicalise popular attitudes, thereby reducing the inﬂuence of
more moderate alternatives.
Therefore, the relationship between regime types and conﬂict is complex, and case studies rather than, or in combination
with, large N-studies are needed in order to fully capture this complexity. This article analyses how attempts to ﬁnd solutions
to the Nagorno-Karabakh conﬂict – from its outbreak to the present – have been affected by the regime types found in
Armenia and Azerbaijan, and how the regimes have in turn been affected by the ongoing conﬂict. It is based on elite interviews,
conducted by the author, in Yerevan in 2006 and 2008 and in Baku in 2009. Grasping more fully the relationship
between regime types and conﬂict resolution is important, not just because it would ﬁll a gap in the literature, but because it
has important policy implications – for the case of Nagorno-Karabakh and also more generally. Should democratisation be
promoted, despite the context of war, instability and unresolved conﬂicts? Or should mediators aim to legitimise and
strengthen authoritarian leaders?
Armenia and Azerbaijan: war, unresolved conﬂict and democratic reversals
The experience of war and unresolved conﬂict has had a signiﬁcant impact on the political systems in Armenia and
Azerbaijan. Democratic stagnation and reversals have characterised both cases and ‘hybrid regimes’ have resulted. This does
not mean, however, that popular legitimacy is of no importance. Moreover, despite similar experiences of conﬂict and war, the
pattern of development is not identical in the two cases, which illustrates the complex effect of war and conﬂict on regime
types.
Nationalism, interrupted democratisation and war
The period leading up to the Karabakh war was in both cases marked by intense competition and increasing militancy, but
the levels of democracy varied. Armenia started out as a promising candidate for democratic transition. Democratic principles
played a central role in the independence movement (Malkasian, 1996), and the Armenian National Movement (ANM)
became the ﬁrst non-communist group to come to power in a Soviet Republic (De Waal, 2003: 56). Popular mobilisation
driven by the Karabakh issue was crucial in this development; it constituted “the locomotive of democratisation” and “disorder
and democratization advanced apace” (Rutland,1994: 844, 857). By early 1990, the ANM had virtually taken over the running of
the republic from the Communist Party (Ibid. 850), and this position was conﬁrmed in elections a few months later. We
therefore see how a looming ethno-nationalist conﬂict can work as a catalyst for democratisation, and not only a by-product
of democratisation. The resulting democratisation was, however, also distorted by this path of development. Paramilitary
formations sprang up in 1990 and armed clashes became commonplace. In response, a state of emergency was declared and
the paramilitary groups were disarmed – or relocated to Karabakh. The leader of the ANM, Levon Ter-Petrosyan, thereby
established himself as a “ﬁrm and capable leader” (Ibid. 853), but his rule was not unproblematic. The 1991 elections were
surrounded by controversy and the period afterwards saw continued gridlock between president and parliament. Nationalism
and the looming conﬂict “pushed other issues off the political agenda, and strengthened the power of the presidency at the
expense of the parliament and political parties” (Ibid. 839).
Political transition in Azerbaijan was characterised by an even bumpier road and was similarly affected by, and in turn
affected, the outbreak of conﬂict. There was no democratic transition in the pre-war period, but this does not mean that there
was no mobilisation for democratic reform. As in Armenia, the looming inter-ethnic conﬂict led to mass mobilisation and
N. Caspersen / Communist and Post-Communist Studies 45 (2012) 131–139 133
Rasim Musabekov (2005) argues that the conﬂict over Karabakh from the beginning “served as a key impulse mobilising the
population for social and political reforms” and thereby “paved the way for the ﬁrst democratic processes in Azerbaijan” (also
Cornell, 2011: 50). The reformist Popular Front was created in 1988 and organised a wave of protests linked to Karabakh (De
Waal, 2003: 82). This led to a crisis of power; the local Communist Party lost control over the republic and the Popular Front
became increasingly militarised, with the leadership unable to rein in local branches (Cornell, 2011: 52–55). Instead of
democratisation, a military crack-down by the Soviet Army resulted. Although the pre-war period in Azerbaijan was not
a period of democratisation, it was certainly marked by intense competition and in this competition the rivalling groups used
the Karabakh factor to further their personal cause (Musabekov, 2005; Cornell, 2011: 56–57).
In Azerbaijan, the fundamental issue of power had therefore not been resolved, and the wartime period came to be marked
by ﬁerce competition, occasional violence and military coups, but also by a short-lived democratic experience. The
Communist government did not fully control the country’s ﬁghting forces and the risk of a civil war with the nationalist
opposition loomed large (De Waal, 2003: 163). This domestic instability signiﬁcantly affected Azerbaijan’s (lack of) military
fortunes in Karabakh and the war in turn directly impacted on the political scene, as governments in Baku “rose or fell as
a result of developments on the battleﬁeld” (Musabekov, 2005). President Ayaz Mutalibov relied on small internal security
forces, fearing that a professional army might threaten the regime (Horowitz, 2003: 37), but this not only made it more
difﬁcult to win in Karabakh it also added to instability in Baku. In the spring of 1992, the opposition forced Mutalibov to resign
following the Khojali massacre in Nagorno-Karabakh, in which hundreds of Azeri civilians were killed, and amid accusations
that the government had not been protecting the town. New presidential elections were called for three months later (De
Waal, 2003: 172–173). However, the fall of the important town Shusha caused a new political crisis. Communist deputies
tried to restore Mutalibov to ofﬁce and a shoot-out followed when the opposition ousted him anew. The following month, in
elections that have been assessed as “probably as good as what was possible in the chaotic circumstances” (Cornell, 2011: 65),
the leader of the Popular Front, Abulfaz Elchibey, was elected president (De Waal, 2003: 182). During the Popular Front’s rule
the news media and political system, for the most part, remained free (Cornell, 2011: 67) and Elchibey, moreover, managed to
create the semblance of an army which initially enjoyed military successes in Karabakh (Ibid. 72). Elchibey’s rule did, however,
not have staying power (Musabekov, 2005) and the government’s downward slide once again began in Karabakh: it lost
control over forces on the front, then the government itself began to crumble and the state degenerated into free-for-all
corruption (Cornell, 2011: 72–74). Elchibey surrendered power in June 1993 when the dissident commander Suret Husseynow
and his men began moving on Baku. The Popular Front, in desperation, invited Heidar Aliyev, the former Communist
leader, to come to Baku. Aliyev quickly consolidated his power and four months later he was credited with receiving an
improbable 98.8 pct of the vote in the presidential elections. The result was “preordained” (De Waal, 2003: 225) and Azerbaijan’s
Freedom House ranking changed from “partly free” to “non free” (Freedom House, 2011a)2
but Aliyev appeared to
enjoy the popular legitimacy that the more democratic Elchibey had been unable to maintain (Cornell, 2011: 78).
Compared to Azerbaijan, the political scene in wartime Armenia was positively stable. Levon Ter-Petrosyan stayed in
power throughout the war, and his rule was in fact bolstered by the ongoing war and resulting hardship. Due to economic
blockades from Azerbaijan and Turkey and the crisis in Georgia, Armenia was struggling for its economic survival and, for
example, only recovered a decent power supply in 1996 (De Waal, 2003: 206). Instead of leading to domestic instability, this
hardship fostered a spirit of wartime solidarity and mobilised popular support for the Karabakh war (Ibid.). This does not,
however, mean that internal divisions were entirely absent. Rivalry continued behind the scenes and it was signiﬁcantly
affected by the Karabakh war. The ongoing war and the lack of complete central control over the coercive forces provided an
opportune environment for the shadow economy and the “maﬁa” was a powerful behind-the-scenes actor (Ibid.). Moreover,
important divisions existed between Ter-Petrosyan and the leaders in Karabakh (Rutland, 1994: 856; Caspersen, 2008).
Following pressure from Yerevan and from military forces, the Karabakh parliament in the summer of 1992 introduced
military rule with a state defence committee assuming all executive powers (De Waal, 2003: 196).3
But divisions persisted4
and a serious rift soon re-emerged between Ter-Petrosyan, who supported a diplomatic settlement, and the Karabakh leaders,
who wanted to press their new military advantage. These divisions came into the open during the attack on Kelbajar in March
1993 when Ter-Petrosyan called on the Karabakh president to halt his forces (De Waal, 2003: 212–213). Behind the apparent
political stability, divisions therefore persisted and these signiﬁcantly affected the conduct of war and created an important
legacy for the post-war period.
In both cases, the Karabakh conﬂict played a key role in the initial popular mobilisation which led to the emergence of
democratic processes. This highlights the close relationship between nationalism and democracy (Schnapper, 2002) and
points to the reverse causality of what Mansﬁeld and Snyder (1995) describe: democratisation may lead to conﬂict, but
conﬂict can also lead to democratisation. Democratisation under those circumstances, however, takes a speciﬁc form. The
Karabakh conﬂict was used as a pre-text for concentrating power and growing militancy also followed. The conﬂict is often
seen as a bottom-up conﬂict (Kaufman, 2001), but the restrictions on political competition and popular participation mean
2
The Freedom House ranking is based on an assessment of political rights and civil liberties.
3
Author’s interviews with David Shahnazaryan, former Karabakh negotiator for Armenia (Yerevan, 6 September 2006) and Manvel Sarkisian, political
analyst (Yerevan, 9 September 2006).
4
Author’s interviews with Levon Zourabian, former advisor to Ter-Petrosyan (Yerevan, 4 September 2006) and Manvel Sarkisian, political analyst
(Yerevan, 9 September 2006).
N. Caspersen / Communist and Post-Communist Studies 45 (2012) 131–139134
we cannot explain the descend war without also considering elite dynamics (Melander, 2001). These dynamics became even
more pronounced during the war, even though Azerbaijan did have a short-lived experience of a democratising regime.
Democratisation is clearly a tall order in the context of military conﬂict, but the wartime period also showed that the outbreak
of war does not have a uniform effect on the regime: It can lead to instability and violent inﬁghting, but can also foster
a closing of ranks.
With the signing of the ceaseﬁre agreement in 1994, the situation changed anew. The “men with the guns” were no longer
the only ones in control and other political issues became salient. However, the legacy of the war continues to have an
important impact. Its effects include: a signiﬁcant refugee population, former combatants who form potentially important
pressure groups, and a general homogenisation of society. Moreover, the unresolved conﬂict results in continued militarisation
(Bagdasarian and Yunusov, 2005). This may be one of the reasons why, in the cases of Armenia and Azerbaijan, the
post-war trend has been one of increasing authoritarianism rather than democratisation. This should not be taking to mean
that popular attitudes are of no importance. Both countries have had experience, however short-lived, with more democratic
systems and greater political freedoms and combined with the nationalist mobilisation caused by the Karabakh conﬂict this
has created an expectation of popular sovereignty. “Implicit popular legitimacy” matters (Cornell, 2011: 174) and the leaders
are not unconstrained.
Democratic reversals: impact of war and unresolved conﬂict
Following the signing of the ceaseﬁre agreement, Armenia spent the next few years in “continuous political turbulence” (De
Waal, 2003: 251). Ter-Petrosyan was facing declining popularity and in a bid to retain control he chose to severely restrict
political competition. The main opposition party was banned prior to the 1995 parliamentary elections and signiﬁcant
irregularities marked Ter-Petrosyan’s victory in the 1996 presidential elections when he moved tanks on to the streets of the
capital and had several opposition leaders detained (Ibid. 257). Ter-Petrosyan’s hold on power therefore came to depend on
the power ministries, the Karabakh leaders and the army, rather than the general population. These forces were responsible
for removing him from power in 1998, when they argued that he had adopted a too moderate position on the Karabakh
conﬂict (Croissant, 1998: 122–123. Laitin and Suny, 1999: 155–156). In October 1999, Armenia was plunged into a new and
more serious political crisis when ﬁve gunmen stormed the National Assembly and assassinated the prime minister, the
parliament speaker and several other senior government ofﬁcials. The end of the war was therefore accompanied by
a reduction in political freedoms, but an increase in internal rivalry.
It took more than a year for Ter-Petrosyan’s replacement, Robert Kocharyan, to re-establish his authority after the
parliamentary shootings (De Waal, 2003: 266), but he eventually removed many of the restrictions that Ter-Petrosyan had
imposed on political and press freedom (Horowitz, 2003). Armenia was, however, still very far from a model democracy and
Kocharyan’s reelection in 2003 was widely viewed as ﬂawed (Freedom House, 2011b). Turbulence and restrictions on political
freedoms returned in the violent aftermath of the disputed 2008 presidential elections: 10 people were killed and 200 injured
in violent clashes between the police and demonstrators and a state of emergency was declared by the outgoing president.
The new President, Serzh Sargsyan, is generally viewed as a weak leader without sufﬁcient clout to agree to difﬁcult
compromises (Cornell, 2011: 137), and the issue of Karabakh remains salient in the continued political rivalry.
In contrast to Armenia, the post-war period in Azerbaijan has been characterised by increased stability. The signing of the
ceaseﬁre agreement allowed Aliyev to tighten his control; he gradually cleared the ﬁeld of actual and potential opponents,
beginning with the army and rebellious militias. Moreover, Aliyev was able to use increasing oil revenues to stabilise the
country (De Waal, 2003: 251–252). This increased control enabled him to abolish the state of emergency (Musabekov, 2005),
and in 1998 he was re-elected president “with a predictably vast majority” (De Waal, 2003: 252). Gradually, however, Aliyev’s
grip on power began to lessen, in part due to concerns about his ill health and lack of an obvious successor (Ibid.; Cornell,
2011: 102).
In 2003 he was replaced by his son, Ilham Aliyev, who was elected president in an election marked by irregularities. Initial
predictions were for the younger Aliyev to be unable to preserve his father’s iron grip on power (ICG, 2010: 2), but he has
deﬁed these predictions and established himself as a strong leader. The resulting stability has, however, come at the price of
“the consolidation of authoritarian rule, greater suppression of freedoms and an increased reliance by elites on corruption and
patronage networks to dominate virtually all aspects of public life” (Ibid. i). Throughout this period, both Aliyevs have used the
unresolved conﬂict over Karabakh to justify repressive measures. They invoked the need for stability arguing that Azerbaijan’s
defeat had been due to the domestic turmoil that characterised the pre-war and wartime periods (Musabekov, 2005).
Both in Armenia and in Azerbaijan we therefore see democratic reversals in the post-war period. Armenia’s political rights
went from 4 to 5 in 2005 and to 6 in 2009 (1 signiﬁes the highest level of rights and 7 the lowest) and the country is classed as
“partly free” (Freedom House, 2011a). Azerbaijan’s political rights remain at 6 but a downward trend is recorded and the
country is classed as “non free” due to its worse record on civil liberties (Freedom House, 2011c). Laura Bagdasarian and Arif
Yunusov (2005) describe the situation as a ‘no war, no peace syndrome’, which is marked by militarisation, stalled democratisation
and the internalisation of identities of victor and victim.
The nature of the regimes nevertheless differs: whereas wartime Armenia was characterised by relative stability and
Azerbaijan was embroiled in turmoil, the roles were largely reversed in the post-war period. The question is therefore if the
impact of the war and the unresolved conﬂict is really that signiﬁcant. Hratch Tchilingirian (2005) has, for example, argued
that the lack of democracy is not due to the Karabakh conﬂict; rather it is due to a combination of regime-induced and
N. Caspersen / Communist and Post-Communist Studies 45 (2012) 131–139 135
inherited systemic problems. Conﬂict and war do not always undermine democratisation; democratisation can co-exist with,
and even be spurred by, ongoing conﬂict (Horowitz, 2003: 26). But the above analysis suggests that regimes that manage to
instrumentalise the conﬂict tend towards becoming ‘hybrid’ or ‘semi-authoritarian’ regimes; that is pluralistic regimes that
combine a rhetorical acceptance of liberal democracy, and the existence of some democratic institutions, with essentially
illiberal or even authoritarian traits (Cornell, 2011: 93; Diamond, 2002). How do such hybrid regimes affect attempts at
conﬂict resolution?
Too much or too little democracy for conﬂict resolution
In order to assess the effect of hybrid regimes, it is ﬁrstly important to recognise the variations found within this category
of regimes. The level of competition and the regime’s stability varies in particular. Moreover, a hybrid regime does not entail
complete autonomy for the leadership; they are often signiﬁcantly constrained by rivalling elites and by the general population.
Hybrid regimes have by a number of authors been linked to the risk of instability (Ward and Gleditsch, 1998), while
others have argued that it makes difﬁcult compromises less likely (Horowitz, 2003). The Armenian and Azerbaijani political
systems certainly appear ill-suited for negotiating a peaceful settlement, and raise the question: is the problem one of “too
much or too little” democratisation? Or: is the deciding factor the degree of internal competition, rather than the nature of the
regime?
During the war, the Armenian president was having signiﬁcant difﬁculty controlling the Karabakh leaders and this appears
to have led to a delay in reaching a ceaseﬁre agreement. It is, however, noteworthy that Ter-Petrosyan was constrained by
actors with whom he was not engaged in direct political competition, but was nevertheless unable to use Armenia’s political
and military leverage to get them to toe the line. He feared that such pressure would result in a domestic backlash (Caspersen,
2008). The worry was not so much about democratic competition but rather about his lack of a stable powerbase and his
insufﬁcient control over the power ministries and the army. These were precisely the forces that managed to oust TerPetrosyan
after the war, when he agreed to a phased solution to the Karabakh conﬂict (Ibid.).
In the case of Azerbaijan, one could speculate that the return of Aliyev in 1993 and the calming down of the political
situation facilitated the acceptance of the May 1994 ceaseﬁre agreement. Aliyev did indeed start negotiating as soon as he
took over power, but he was also tempted to try to retake Azerbaijan’s lost territories (De Waal, 2003: 227), and when the
Bishkek protocol was ﬁnally signed it was notably not signed by Aliyev personally. Instead, the agreement was signed by the
deputy speaker of parliament. The reason was apparently that Aliyev feared a domestic backlash and lack of agreement within
the governing ranks (Ibid. 238). At this point Aliyev had not yet established a monopoly over the use of force, (Cornell, 2011:
86–87) and while the lack of stability clearly affected the conﬂict, it cannot therefore be concluded that the signing of the
ceaseﬁre agreement was due to the political stability brought on by a more authoritarian leadership.
Despite the democratic reversals, competition remained intense in both cases and the actual or potential outbidding
reduced the leeway enjoyed by the leaders. The decisive audience for this competition was, however, often not the general
population. For example, the coups and attempted coups in Azerbaijan were staged by hardliners unhappy with military
losses, or with their own lack of power. Similarly, when Ter-Petrosyan was ousted from the Armenian presidency in 1998, the
decisive actors were found within his own ranks; not in the opposition and not in the general population. It is noteworthy that
his successor, the former president of Nagorno-Karabakh, Robert Kocharyan, had a bumpier accession to power than
anticipated and had to rely on support from the state media, and allegedly vote fraud, to ensure his victory in the 1998
presidential elections (De Waal, 2003: 262).
To this day, the leaders of both countries face signiﬁcant competition both from the opposition and from within the ruling
elite; in the case of Azerbaijani politics, the latter even appear to be the more important division (Cornell, 2011: 119). Any
attempt to agree to difﬁcult compromises on the issue of Karabakh is therefore likely to be constrained. In Armenia, power is
ﬁercely contested and the “Karabakh card” is used by the opposition to denounce the ruling regime (Tchilingirian, 2005). The
current Armenian President Serzh Sargsyan is widely viewed as a weak leader, but even his more powerful predecessor
Robert Kocharyan, was under pressure when negotiating a settlement for Karabakh. There were speculations that the 1999
parliament shootings were linked to the Karabakh talks, and although this has not been substantiated (De Waal, 2003: 266), it
is suggestive of an atmosphere in which a compromise agreement would carry signiﬁcant risks. Such risks do not necessarily
come in the shape of a gun, but could also be of a political nature. Thus, during the 2001 Key West talks the Armenian
parliament clearly rejected the proposals that Kocharyan brought back for consultations (Ibid. 267). After the disputed 2008
elections, international pressure for an agreement on Karabakh increased and the opposition announced that it was suspending
its mass actions for two months (Grigoryan, 2008). Ofﬁcially this was to enable the president to negotiate in earnest
and reach an acceptable agreement,5
but there were also concerns that the ongoing talks would otherwise be used as an
excuse to crack down on the opposition.6
In Azerbaijan, Ilham Aliyev is usually seen as a stronger leader and Cornell (2011: 153) assesses that he would be able to
carry a settlement. However, Aliyev Senior’s attempts to agree to a settlement illustrate that not even an authoritarian leader
5
Author’s interview with David Shahnazaryan, Director of Concord (Yerevan, 20 October 2008).
6
Author’s interviews with Haroutin Khachatrian, journalist and political analyst (Yerevan 19 October 2008) and Levon Zourabian, coordinator of the
Armenian National Congress (Yerevan, 9 November 2008).
N. Caspersen / Communist and Post-Communist Studies 45 (2012) 131–139136
can expect unconditional support from within his own ranks. Thus, when Heydar Aliyev in 1999 considered a settlement
which involved a territorial exchange, it led his closest circle of advisors to resign (De Waal, 2003: 264. Cornell, 2011: 144).7
In
1997–1998 Aliyev had, in an attempt to immunise himself from opposition criticism, published all the hitherto secret
proposals on Karabakh. This strategy succeeded in demonstrating that the opposition lacked alternatives, but it did not give
Aliyev the necessary leeway to compromise. Similarly, when he consulted his domestic base during the Key West talks, he was
not even able to secure the approval of his inner circle (Musabekov, 2005). Since Ilham Aliyev came to power there has been
a certain hardening of position on the Karabakh issue and the use of more belligerent rhetoric. This may have been necessary
for Aliyev to consolidate his power, and is primarily meant for an internal audience,8
but it constrains his ability to agree to
a settlement.
Is the problem therefore the risk of a backlash from the general population? It does not appear so: the talks have so far
failed either due to a lack of willingness to compromise or due to the inability of the leaders to ensure the support of their own
ranks. But popular attitudes are not without importance. Cornell (2011: 174) argues that although voters have not been
permitted to determine the outcome of an election in Azerbaijan or to remove an incumbent from ofﬁce, the general population
have at important turning points inﬂuenced outcomes. For example, when Aliyev was facing a coup attempt in 1994
and appealed for popular support, people showed up in their thousands (Ibid.), and the rumours of a ‘territorial swap’ in 1999
led to large demonstration which increased fears that a settlement would endanger Aliyev’s rule (Ibid. 145–146). Similarly,
when Ter-Petrosyan was ousted by his three closest allies in 1998, it mattered that his popular legitimacy had been eroded
through electoral fraud and post-election violence. The lack of popular support weakened him in the internal competition.
Even in hybrid or semi-authoritarian regimes, it is therefore not just about unconstrained elite competition in institutions
that are nominally democratic but lacking in substance. This is symptomatic of the problems often identiﬁed in the Karabakh
peace process. The two societies have been militarised and radicalised through a discourse of ‘othering’, and the Armenian
and Azerbaijani leaders are now worried that they will not be able to persuade them to compromise; Bagdasarian and
Yunosov (2005) argue that there is therefore a perception that it is the societies and not the political elites who are not
ready for resolution of the conﬂict. The chairs of the Minsk Group certainly supported this view when they, after the
breakdown of the Key West talks, declared that the two presidents were “ahead of their people” (Cornell, 2011: 147). International
mediators have generally drawn the conclusion that the Armenian and Azerbaijani leaders therefore need to be
insulated from popular pressure. Thus, in connection with the fraudulent 2000 Azerbaijani elections Aliyev reportedly created
expectations in the West that an agreement was imminent and the Western embassies, as a result, pressured the opposition
to stop their protests and rallies, arguing that internal stability was needed.9
The same thing appears to have happened
following the 2008 Armenian elections. Sargsyan was needed for a compromise agreement and instead of criticising the
elections and the violence that followed, the international mediators sought to bolster his position.10
The most signiﬁcant
threat to compromise arguably comes from within the ruling elite and the leaders can therefore be said to have used the
ongoing conﬂict to avoid international pressure for reform. It could, however, also be hypothesised that authoritarian leaders
tend to mistrust the population and therefore really do fear involving them in the process.11
Democratisation is frequently seen as a threat to conﬂict resolution, but the above analysis has clearly demonstrated that
the leaders of semi-authoritarian regimes are not immune to political insecurity and also frequently lack the necessary leeway
to agree to difﬁcult compromises. Moreover, if these leaders do decide to compromise they cannot count on the possible
positive effects of democratisation: they lack the genuine popular mandate that could bolster them against more radical rival
when trying to make concessions. This does not mean that democratisation will necessarily have a positive effect on conﬂict
resolution, but popular attitudes cannot be ignored anyway and if the spoilers, who are trying to bring down an agreement,
are unrepresentative, then democratisation offers a way of strengthening more moderate forces. There is little to indicate that
the Armenian and Azerbaijani leaders are currently unrepresentative when it comes to the issue of Karabakh, but the
secretive nature of the talks has hardly been conducive to the creation of more moderate views. Perhaps mindful of possible
backlashes, Ilham Aliyev has begun to involve civil society organisations in the peace process, but little is done to prepare the
general population for a compromise.12
What we get therefore appears to be the worst of both worlds: intense political
competition, often behind the scenes, without the accountability and gradual development of more positive-sum attitudes
that democratisation could result in.
Conclusion
It has been variously argued that the lack of progress in the Nagorno-Karabakh peace process is due to either too much or
too little democratisation in Armenia and Azerbaijan. In a sense, both are right. The preceding analysis has tried to
demonstrate that the process suffers from signiﬁcant intra-communal divisions, but without the positive effects that genuine
democratisation could provide. So which direction would be more promising? Even completely authoritarian leaders may, as
7
Author’s interview with Eldar Namazov, political analyst and a former chief advisor to Heydar Aliyev (Baku, 16 June 2009).
8
Author’s interview with Fariz Ismailzade, political analyst (Baku, 11 June 2009).
9
Author’s interview with Ilgar Mammadov, political analyst (Baku, 19 June 2009).
10
Author’s interview with Levon Zourabian, coordinator of the Armenian National Congress (Yerevan, 9 November 2008).
11
Author’s interview with Avaz Hasanov, political analyst (Baku, 11 June 2009).
12
Author’s interview with Eldar Namazov, political analyst and a former chief advisor to Heydar Aliyev (Baku, 16 June 2009).
N. Caspersen / Communist and Post-Communist Studies 45 (2012) 131–139 137
argued above, face a signiﬁcant backlash if they agree to far-reaching concessions, and studies have shown that consolidated
democracies are more stable than consolidated autocracies (Hegre et al., 2001). Even if democratisation does not necessarily
lead to instability and continued radicalisation, it is also not a panacea for conﬂict resolution either. Its positive effect depends
on the spoilers of peace being unrepresentative, and any move towards further democratisation in Armenia and Azerbaijan
would therefore need to address popular attitudes and broaden the peace process. This could be a by-product of democratisation
but it is not guaranteed.
The effect of regime types on the prospect for conﬂict resolution has until now been under-analysed. Much of the debate
has been based on general assumptions of outbidding, or borrowed from the much-contested argument that democratisation
breads instability. But although electoral outbidding can indeed be observed in some cases, the assumption of almostautomatic
outbidding is often misleading: damaging internal rivalry does not depend on democratisation and often has
little to down with ‘playing the ethnic card’ and appealing to an antagonised public. This article has argued that the greatest
obstacle to conﬂict resolution appears to come from regimes that have only seen very partial democratisation; political
systems where limited democratic institutions and often intense elite competition co-exists with weak institutional checks
and unclear popular mandates. Partial democracy can therefore be said to be associated with instability, as suggested in the
literature, but it is not because of spoilers outbidding their rivals by appealing to extreme popular views. Popular attitudes
can, on the other hand, not be ignored. Some level of popular legitimacy is important for a settlement to be sustainable, even
though the most signiﬁcant threat to peace often comes from within the ruling elite.
These ﬁndings have important policy implications, but they are complicated by the fact that the regime type is itself
affected by the experience of war and unresolved conﬂict. This article has argued that these conditions tend to lead to exactly
the kind of regime that is not conducive to conﬂict resolution. So is genuine democratisation even possible? Although there
might be a tendency towards hybrid regimes, the analysis also pointed out that an intensifying conﬂict, along national lines,
can act as an impetus for democratisation. Similarly, the legacy of war, or even the persistence of an unresolved conﬂict, does
not mean that democratisation is impossible. Other cases in the Caucasus have, for example, progressed further when it
comes to democratic reforms, and a case such as Serbia demonstrates that democratisation can indeed take place in the
context of war legacies and an unresolved conﬂict. Although signiﬁcant, the obstacle posed is not insurmountable, but in
order to progress it is imperative to gain a fuller understanding of the complex relationship between regime types and conﬂict
resolution.
Acknowledgements
Research for this article was funded by a grant from the UK’s Economic and Social Research Council (RES-000-22-2728)
References
Babayan, A., Perch, J.-C., 2006. Nagorno-Karabakh: Russia Calls for Mutually Acceptable Solution. RFE/RL, 6 April.
Bagdasarian, L., Yunusov, A., 2005. War, social change and “no war, no peace” syndromes. In: Broers, L. (Ed.), The Limits of Leadership: Elites and Societies in
the Nagorny Karabakh Peace Process. Conciliation Resources, London.
Broers, L., 2005. Introduction. In: Broers, L. (Ed.), The Limits of Leadership: Elites and Societies in the Nagorny-Karabakh Peace Process. Conciliation
Resources, London.
Caspersen, N., 2008. Between puppets and independent actors: Kin-state involvement in the conﬂicts in Bosnia, Croatia and Nagorno Karabakh. Ethnopolitics
7 (4), 357–372.
Chua, A., 2004. Our Most Dangerous Export. The Guardian. 28 February.
Cornell, S.E., 2011. Azerbaijan Since Independence. M.E. Sharpe, Armonk.
Croissant, M.P., 1998. The Armenia-Azerbaijan Conﬂict: Causes and Implications. Praeger, Westport.
Cunningham, K.G, Bakke, K.M., Seymour, L., forthcoming. Shirts today, skins tomorrow: the effects of fragmentation on conﬂict processes in selfdetermination
disputes. Journal of Conﬂict Resolution.
Curtis, D., De Zeeuw, J., 2010. Rebel Movements and Political Party Development in Post-Conﬂict Societies. Program on States and Security. Ralph Bunche
Institute for International Studies.
De Waal, T., 2003. Black Garden: Armenia and Azerbaijan through Peace and War. New York University Press, New York.
Diamond, L.J., 2002. Thinking about hybrid regimes. Journal of Democracy 13 (2), 21–35.
Findley, M., Rudloff, P., 2009. Combatant Fragmentation and The Dynamics Of Civil Wars. Paper Presented at the Annual Meeting of the American Political
Science Association, Toronto, 3–6 September.
Freedom House, 2011a. Freedom in the World: Country Ratings and Status, 1972–2011. http://www.freedomhouse.org/report-types/freedom-world/.
Freedom House, 2011b. Freedom in the World: Armenia. http://www.freedomhouse.org/report/freedom-world/2011/armenia.
Freedom House, 2011c. Freedom in the World: Azerbaijan. http://www.freedomhouse.org/report/freedom-world/2011/azerbaijan.
Goldstone, J.A, Bates, R.H, Gurr, T.R., Lustik, M., Marshall, M.G., Ulfelder, J., Woodward, M., 2005. A Global Forecasting Model of Political Instability. Paper
presented at the Annual Meeting of the American Political Science Association, Washington DC, 1–4 September.
Grigoryan, M., 2008. Armenia: The Opposition Takes a Break from Yerevan Protests. Eurasia Insight. 19 October.
Hegre, H., Ellingson, T., Gates, S., Gleditsch, N.P., 2001. Toward a democratic civil peace? American Political Science Review 95, 33–48.
Horowitz, D., 1985. Ethnic Groups in Conﬂict. University of California Press, Berkeley.
Horowitz, S., 2003. War after Communism. Journal of Peace Research 40 (1), 25–48.
International Crisis Group, 2010. Azerbaijan: Vulnerable Stability. Europe Report N
207, 3 September.
Kaufman, S., 2001. Modern Hatreds: The Symbolic Politics of Ethnic War. Cornell University Press, Ithaca.
Kenny, P., 2010. Structural integrity and cohesion in insurgent organizations. International Studies Review 12, 533–555.
King, C., 2004. The Micropolitics of social violence. World Politics 56 (April), 431–455.
Kydd, A., Walter, B., 2002. Sabotaging the peace: the politics of extremist violence. International Organization 56 (2), 263–296.
Laitin, D., Suny, R., 1999. Armenia and Azerbaijan: thinking a way out of Karabakh. Middle East Policy 7 (1), 145–176.
Lederach, J.P., 1998. Building Peace: Sustainable Reconciliation in Divided Societies. Washington DC: United States Institute for Peace.
Malkasian, M., 1996. Gha-Ra-Bagh: The Emergence of the National Democratic Movement in Armenia. Wayne State University Press, Detroit.
N. Caspersen / Communist and Post-Communist Studies 45 (2012) 131–139138
Mansﬁeld, E.D., Snyder, J., 1995. Democratization and the danger of war. International Security 20 (1), 5–38. 5.
Maoz, Z., Abdolali, N., 1989. Regime types and international conﬂict, 1815–1976. Journal of Conﬂict Resolution 33 (March), 3–35.
Melander, E., 2001. The Nagorno-Karabakh conﬂict revisited. Journal of Cold War Studies 3 (2), 48–75.
Mitchell, P., 1995. Competition in an ethnic dual party system. Ethnic and Racial Studies 18 (4), 773–796.
Musabekov, R., 2005. The Karabakh conﬂict and democratization in Azerbaijan. In: Broers, L. (Ed.), The Limits of Leadership: Elites and Societies in the
Nagorny Karabakh Peace Process. Conciliation Resources, London.
Nordlinger, E., 1972. Conﬂict Regulation in Divided Societies. Center for International Affairs, Harvard University, Cambridge.
Paris, R., 2004. At War’s End: Building Peace after Civil Conﬂict. Cambridge University Press, Cambridge.
Rabushka, A., Shepsle, K., 1972. Politics in Plural Societies: A Theory of Democratic Instability. Merrill, Columbus.
Rutland, P., 1994. Democracy and nationalism in Armenia. Europe-Asia Studies 46 (5), 839–861.
Schnapper, D., 2002. Citizenship and national identity in Europe. Nations and Nationalism (1), 8.
Stedman, S.J., 1997. Spoiler problems in peace processes. International Security 22 (2), 5–53.
Tchilingirian, H., 2005. New structures, old foundations: state capacities for peace. In: Broers, L. (Ed.), The Limits of Leadership: Elites and Societies in the
Nagorny Karabakh Peace Process. Conciliation Resources, London.
Tezcür, G.M., 2010. When democratization radicalizes. Journal of Peace Research 47 (6), 775–789.
Vreeland, J.R., 2008. The effect of political regime on civil war. Journal of Conﬂict Resolution 52 (3), 401–425.
Walker, E. 1998. No peace, no war in the Caucasus. Occasional Paper, Belfer Center for Science and International Affairs.
Ward, M.D., Gleditsch, K.S., 1998. Democratizing for peace. American Political Science Review 92 (1), 51–61.
N. Caspersen / Communist and Post-Communist Studies 45 (2012) 131–139 139