icon caret Arrow Down Arrow Left Arrow Right Arrow Up Line Camera icon set icon set Ellipsis icon set Facebook Favorite Globe Hamburger List Mail Map Marker Map Microphone Minus PDF Play Print RSS Search Share Trash Crisiswatch Alerts and Trends Box - 1080/761 Copy Twitter Video Camera  copyview Whatsapp Youtube
A demonstrator prays with her hand on a policeman's shield in front of a barricade in Independence Square, Kiev, February 2016. MAGNUM/Larry Towell
Special Report 2 / Global

Seizing the Moment: From Early Warning to Early Action

Wars can be prevented or mitigated by early, clear and well-designed political and diplomatic engagement. Yet policymakers are increasingly stretched by a myriad of global crises. Refocusing on knowledge, relationships, frameworks, strategic communication and pathways to peace is crucial to limiting and resolving the world’s current upsurge in deadly conflict.

Executive Summary

After a period of relative calm, an upsurge of crises is testing the international system, pitting major powers and regional players against one another and highlighting the weaknesses of preventive diplomacy. Governments and international organisations were taken by surprise by the Arab uprisings in 2011 and slow to react to crises in South Sudan and the Central African Republic (CAR) in the years that followed. Global and regional rivalries have weakened diplomacy over Syria, Ukraine and the South China Sea. Policymakers, stretched by the symptoms of this wave of instability, including mass displacement and the spread of transnational terrorism, struggle to focus on conflict prevention. 

Yet, preventive diplomacy is not necessarily dead. The Iranian nuclear deal, progress toward peace in Colombia and the high-level push to avoid election-related chaos in Nigeria in 2015 have been reminders of what intensive international engagement can deliver. If politicians, diplomats and international officials invest in key dimensions of early warning and early action – analysing conflict dynamics closely, building sensitive political relationships in troubled countries and undertaking complex “framework diplomacy” with other powers to create political space for crisis management – they still have a chance to avert or mitigate looming conflicts and ease existing wars.

This report, drawing on Crisis Group’s field-centred analysis and policy recommendations from the past five years, sets out a broad strategic framework for preventive diplomacy. Its primary focus is on conflicts, like those in Ukraine and Syria, which directly involve outside powers. While classical inter-state conflicts remain rare, internationalised civil wars are a leading source of regional and global frictions. Building frameworks to address both the internal and external tensions that shape them is likely to be a recurrent challenge for big powers, regional players and multilateral organisations in the years ahead. 

The first half of this report focuses on the internal drivers of recent and current crises. It argues that while it is exceedingly hard to identify specific triggers of future conflicts, it is possible to identify likely threats to peace and work out how they may play out if left unaddressed. It emphasises the need to understand the political dimensions of conflicts and, especially, the leaders and elites whose choices for or against violence are pivotal. Grasping how such leaders make these decisions is essential for effective early warning, but it must be buttressed by much broader political analysis covering, inter alia, the dynamics of ruling parties, opposition groups and civil society, not just at the national but at all levels of society. 

Building anticipatory relations with all these actors constitutes a bedrock for effective early action by outside partners, once a crisis looks set to break. It is important, too, to grasp the politics and strategies of militaries and internal security forces in cases such as Egypt, or of non-state armed groups in chaotic environments like Libya. The report also highlights the sources of many conflicts in countries’ marginalised peripheral regions. Local rebellions in Yemen, Mali, the Democratic Republic of Congo (DRC), Pakistan and CAR, to name a few, have expanded unexpectedly and exponentially, causing widespread violence and overthrowing a number of governments. 

A focus on the internal players in countries at risk must be complemented by efforts to engage and balance the interests of external actors, while recognising that the distinction between “internal” and “external” actors is moot in many crises. In the Middle East and Central and West Africa, conflicts frequently flow across borders, and regional powers simultaneously fuel conflicts and position themselves as peacemakers. Ethnic groups such as the Kurds in the Middle East straddle multiple countries, while organised criminal networks and transnational extremist groups are not restricted to individual states. This means that experts engaged in early warning and early action must treat regional and wider international factors as integral to their conflict analysis and development of appropriate policy.

The report goes on to look more closely at the varieties of “framework diplomacy” that can facilitate the requisite engagement. In many crises it is necessary to look beyond established multilateral frameworks – important though these can be – and pull together case-specific groupings of states and institutions to manage a problem, or at least minimise frictions. Sometimes neither formal nor ad hoc inter-governmental arrangements will be suitable: back-channel diplomacy led by local “insider mediators”, specialised international NGOs or other actors may be the best initial way to handle looming tensions.

The remainder of the report reviews the means available for directly engaging in conflicts as they escalate or in anticipation of their outbreak. It emphasises the need for inclusive approaches to political dialogue, meaning not only outreach to civil society, women’s groups and other constructive forces, but also marginalised minorities and armed groups – including some highly controversial actors such as Islamist extremists. In addition to mediation and other diplomatic options such as deploying high-level envoys, tools include a range of coercive measures and incentives for peace. Coercive tools include diplomatic “naming and shaming”, threats of international legal action in response to atrocities and the use of sanctions. All have significant limitations and can worsen rather than alleviate crises if not well coordinated and aligned to a broader political strategy.

This report, drawing on Crisis Group’s field-centred analysis and policy recommendations from the past five years, sets out a broad strategic framework for preventive diplomacy.

At least equal caution should be applied to the use of force. As the Arab intervention in Yemen has underlined, like many interventions before it, military action can prove costly and counterproductive. This caution also applies to deployments of military peace operations, which have become a standard part of international crisis management (especially in Africa) and increasingly tend toward more robust forms of peace enforcement. While such missions can and do save lives, they can also become entangled in local conflicts, get bogged down in situations from which they have no exit strategy and become overly aligned with governments that do not always enjoy much popular support.

Whatever direct or indirect means of engagement states use, they should set explicit and limited political goals and communicate these clearly to other actors (including their opponents) to avoid violence spiralling beyond control. While coercion may have a role to play in management of a specific crisis, it should be balanced with clear incentives for leaders, elites and their supporters to follow paths away from violence. These may include aid for post-crisis demobilisation, governance reforms and reconstruction. 

More strategically, the best peace incentives that outsiders may be able to offer are ideas and advice to actors in a crisis on how to structure mutually-beneficial arrangements to share power and resources. In Libya, for example, the interest all sides ultimately have in a functioning energy sector could be a point of consensus even while political disputes create friction.

No one group of analysts and forecasters is consistently right in its early warnings (Crisis Group included), and no early action strategy is foolproof. Tackling conflicts as they emerge and develop is an inherently chancy business, and governments and international organisations that engage in it inevitably risk failure. Nevertheless, early, strategic, well-designed engagement predicated on the discipline of close analysis, development of anticipatory relationships and framework diplomacy may help prevent conflict or limit its escalation. To the extent that their resources permit, governments, regional bodies and international organisations should invest in four key areas:

  • Knowledge and relationships. Policymakers, working directly or through others, should develop the closest possible knowledge of troubled countries’ political systems and cultivate channels for frank discussions with leaders, elites, security forces and civil society over the risks of crisis. “Early warning” should, in sum, rest not only on economic and other indicators of danger (although these are useful), but also on in-depth political links with crucial actors.
  • Framework diplomacy. Given the dangers of international and regional tensions exacerbating a crisis, policymakers should make early and concerted efforts to bring international players to the table to assess their interests, hear their analyses and develop common positions on how to act. This can take place in formal multilateral settings or ad hoc, but it is essential to choose mechanisms that enable real bargaining, resulting in frameworks for handling a conflict, rather than formal exchanges or public recriminations. 
  • Strategic planning and communication. It is easy for policymakers to stumble into crises without a clear grasp of what they aim to achieve. The constant need to make statements, launch initiatives and satisfy calls for action makes strategic thinking and planning difficult. It is crucial that governments and international organisations invest in laying out clear overall goals for engaging in crises and communicate these clearly both to the players involved in a conflict and other international actors with interests at stake. 
  • Creating pathways to peace. The ultimate goal of all this relationship-building, framework diplomacy and strategic planning is not simply to guide early action, but to signal to the parties at the centre of a conflict that they can take paths to peace rather than wade into violence. Outside actors can rarely compel leaders and factions on the brink of conflict to step back. But if they are able to engage in well-informed political and diplomatic work and sketch out ideas for lasting peaceful solutions to a conflict, they may persuade their interlocutors to pause before escalating – and perhaps follow an alternative political route that avoids, or at least limits, all-out violence.

Brussels, 22 June 2016

I. Introduction

Five years ago, the Arab uprisings exposed the weaknesses of existing models of early warning and early action in response to political crises. While many analysts were aware of the political, social and economic factors that led to the uprisings in early 2011, few if any foresaw the wave of disorder that spread across North Africa and the Middle East. Governments and international organisations resorted to a variety of policy tools – ranging from offers of mediation to economic sanctions and threats of international prosecution – that frequently failed to alter the calculations of embattled political elites. In many cases, their efforts backfired badly.

While the United Nations (UN) Security Council mandated military action in Libya to protect civilians in March 2011, the uprising against Muammar Qadhafi resulted in a fractured state that slid into chaos while outside powers focused elsewhere. In Yemen, an initially successful UN mediation ran out of steam, paving the way for the Saudi-led intervention in 2015. Arguments over these crises also fuelled geopolitical confrontations, variously involving the West, Russia, China, and Arab and African powers, that have severely complicated later attempts at conflict management. Doubts about international crisis response have since been compounded, as conflicts have escalated from South Sudan to eastern Ukraine. In some cases, such as Mali and the Central African Republic (CAR), analysts and officials saw crises escalate but did not react promptly or decisively. In others, as in Ukraine, the pace of events appeared to take outside actors by surprise.

By 2014, the Uppsala Conflict Data Program (UCDP) has calculated, there were some 40 conflicts worldwide, eleven involving over 1,000 battle deaths a year: “the highest number of conflicts since 1999”.[fn]Therése Pettersson and Peter Wallensteen, “Armed Conflicts, 1946-2014”, Journal of Peace Research, vol. 52 (4), p. 536. International Institute for Strategic Studies (IISS), with a different methodology, contends that conflicts worldwide have in fact declined since 2010, but there has been a “steady increase in lethality” (which levelled off overall in 2015, despite increased deaths in Afghanistan, Nigeria, Somalia and Yemen) and a long-term rise in conflict-related displacement. Anastasia Voronkova, “Editor’s Introduction,” in Voronkova (ed.), Armed Conflict Survey 2016 (IISS, 2016), p. 5.Hide Footnote  Many governments and international organisations focus on managing the fallout from these conflicts. Donors have had to repurpose funds to handle the record numbers of refugees and internally displaced persons (IDPs). International coalitions are trying to contain and rollback violent Islamist extremist groups in the Middle East and North Africa with a mix of military aid to both state and non-state actors, covert operations and airstrikes. These measures crowd out discussion of long-term conflict prevention and resolution.

This is short-sighted. As UN officials have recently emphasised, the key to stemming the humanitarian crisis in the Middle East is not only to increase funding to aid agencies, but also to resolve the conflicts there. Crisis Group has argued that vital to countering the influence of groups like the Islamic State (IS) and al-Qaeda is to ratchet down regional confrontations, in particular the rivalry between Saudi Arabia and Iran; redouble efforts to contain (if not immediately resolve) the conflicts these groups exploit; and work toward local solutions based on the inclusion, rather than alienation of vulnerable communities. Panels convened by the UN and Organization for Security and Cooperation in Europe (OSCE) have underscored the need to focus on politics and diplomacy in addressing conflicts.[fn]“One Humanity: Shared Responsibility; Report of the Secretary-General for the World Humanitarian Summit”, UN document A/70/709, 2 February 2016, pp. 6-12. Crisis Group Special Report, Exploiting Disorder: al-Qaeda and the Islamic State, 14 March 2016. “Our Shared Responsibility”, Report of the High-Level Independent Panel on Peace Operations, (UN) A/70/95 – S/2015/446, 17 June 2015; “Back to Diplomacy”, Final Report and Recommendations of the Panel of Eminent Persons on European Security as a Common Project, OSCE, 3 December 2015.Hide Footnote

The strategic case for effective early warning tools and early action mechanisms to avert potential conflicts, or at least stop them from escalating and spreading into broader confrontations, seems clear enough.

Even if governments pay lip service to such notions, many have resorted to covert or overt military actions to manage crises: examples range from Russia in Ukraine and Saudi Arabia and its allies in Yemen to Uganda and Sudan in South Sudan. While the bulk of current conflicts are intra-state wars, at least a third are internationalised – with foreign forces from one or more other countries in the fight – exacerbating regional and wider international tensions and rendering conflict resolution significantly more complex.[fn]Pettersson and Wallensteen, op. cit., p. 537. A recent summary of the political economy of armed conflicts underlines that most “do not fit neatly” into the categories of intra- and inter-state wars, due to the complex nature of their violence and divisions. Achim Wennmann, “The Political Economy of Violent Conflict”, in Armed Conflict Survey, op. cit., p. 20.Hide Footnote  This report thus pays most attention to internationalised intra-state conflicts, but also draws lessons from other flashpoints, like the South China Sea.

The strategic case for effective early warning tools and early action mechanisms to avert potential conflicts, or at least stop them from escalating and spreading into broader confrontations, seems clear enough. Yet, there is a daunting mix of obstacles to effective early international response. These range from understanding the implications of political frictions in peripheral areas of weak states, such as Mali, to the diplomatic challenges of forging international frameworks to handle cases like Syria. Few if any of these challenges are unprecedented – for examples of the problems of volatile peripheral areas, one can go back to the demise of the Austro-Hungarian Empire a century ago – but addressing them requires sharp political insight, judgment and action that still often elude policymakers.

In parallel with the deterioration of particular conflicts, the norms that have underpinned much post-Cold War thinking on conflict prevention and resolution are in flux. Russia, China and other non-Western powers argue that NATO abused the Responsibility to Protect (R2P) during the Libyan crisis in 2011. Even if political pluralism and representative government still offer the best hope of stability in most countries, the difficulties of democratisation are obvious. Conflict prevention specialists are now as likely to emphasise the dangers associated with elections in fragile states as their advantages. The post-Cold War trend toward strengthening international justice, symbolised by the International Criminal Court, is also encountering increasing pushback. Yet, the last quarter century’s ideals still have some purchase. In Africa in particular, the African Union (AU) and sub-regional bodies repeatedly, if inconsistently, cite human security, prevention of mass atrocities and defence of legitimate governments to justify interventions.[fn]See, for example, Crisis Group Africa Report N°234, Implementing Peace and Security Architecture (III): West Africa, 14 April 2016, p. 9.Hide Footnote  Rather than rising or declining linearly, these norms ebb and flow on a case-by-case basis.

This report maps out how governments and multilateral organisations can best respond to looming crises in this uneasy international environment. It begins by asking how relevant existing thinking about early warning and early action is today. It then explores recent lessons about drivers of conflict, including elite decision-making, localised violence and regional political factors. Finally, it turns to the diplomatic tools, coercive measures and incentives typically available to policymakers trying to address crises and the strategic and diplomatic frameworks needed to put these tools to use. It is necessary to be realistic about the chances of halting fast-moving crises, but effective and rapid action is often possible.

II. Early Warning and Early Action: In Search of Political Strategies

“Early warning” and “early action” are phrases open to multiple definitions. This report concentrates on early warnings of violent conflict and strategies of early action that external actors may take to address those risks. There is a perennial debate about what “early” means: should it include long-range indicators of instability and long-term actions, such as economic assistance, that may alleviate them? While acknowledging the value of long-term warnings, this paper takes a narrower view and focuses on medium- and short-term warnings and responses to political dynamics that have a clear potential to lead to violence. This encompasses imminent threats and risks that may require some years to come to fruition. The precise timeline is less important than the presence of signs that leaders, political factions or other armed groups are taking steps that could ultimately lead to conflict.

This focus on looming conflicts requires three qualifications. The first is that poli­tical analysis should identify not only threats, but also actors who favour peace and unexpected opportunities for settling disputes. As noted in Section III below, officials and analysts should build relationships with political figures, civil society members and others who can promote non-violent solutions to a crisis.

The second qualification is that, while this report largely discusses emerging and escalating crises, it is essential to keep watch for unexpected developments in active and ongoing conflicts. Events such as the rise of IS in Syria in 2014 or the upsurge of violence in Ukraine in early 2015 can fundamentally transform the dynamics of an existing war. The detailed political and security analysis promoted below can and must continue even after a conflict explodes.

Thirdly, it is necessary to ask who is best-placed to conduct this analysis and direct early action. This report does not focus on the early warning and early action mech­anisms of any one government or international organisation. The lessons and advice it offers are designed to apply to a wide range of governments and other bodies concerned with international security; generic references to “outsiders” and “policymakers” are deliberately non-specific.

As Section IV emphasises, who does early warning and early action is increasingly complicated and contentious. When Crisis Group launched in the mid-1990s, the U.S., its allies and the UN appeared to dominate the field. Today, a diverse array of often mutually mistrustful states, organisations and non-governmental groups are engaged. The exact mix of relevant players differs from case to case. Complex “framework diplomacy” – painstaking efforts to establish case-specific diplomatic mechanisms for analysing, managing and mediating conflicts – is frequently required to bring them together, or at least limit friction. By taking a broad view of who can deal with crises and how, this paper points to some principles for such cooperation.

A. What Can “Early Warning” Do?

A focus on political actors and analysis is in line with earlier studies that argued early engagement in crises must rest on an understanding of political dynamics. The Carnegie Commission on Preventing Deadly Conflict concluded in 1997 that “mass violence invariably results from the deliberately violent response of determined leaders and their groups to a wide range of social, economic and political conditions that … usually do not independently spawn violence”.[fn]“Preventing Deadly Conflict”, Carnegie Corporation of New York, December 1997, p. 29.Hide Footnote  Although political scientists and forecasters search for statistically verifiable causes of conflict, such as economic factors, many are giving more weight to leaders and political factors as creators of violence.[fn]See Michael C. Horowitz, Allan C. Stam and Cali M. Ellis, Why Leaders Fight (Cambridge, 2015).Hide Footnote  Whereas researchers once claimed that ethnic cleavages were an innate cause of conflicts, analysts now emphasise that leaders’ deliberate use of ethnically-loaded rhetoric plays a crucial part in dividing and radicalising communities.[fn]Wennmann, “The Political Economy of Violent Conflict”, op. cit., p. 22; Crisis Group Africa Report N°235, Burundi: A Dangerous Third Term, 20 May 2016, pp. 5-7.Hide Footnote  Recent work on climate change and environmental degradation, for example, indicates that these “only trigger violence if the social and political context of a country are particularly disadvantageous”.[fn]Gerald Schneider, Nils Petter Gleditsch and Sabine C. Carey, “Exploring the Past, Anticipating the Future: A Symposium”, International Studies Review, vol. 12 (1), 2010, p. 5.Hide Footnote

This basic assumption about the importance of political factors has long underpinned policy thinking on early crisis response, with a focus on developing strategies to shape crucial elites’ decision-making. If diplomats or international officials want to engage in a country on the verge of conflict, they need not only to develop a sense of its underlying problems, but also to have a working knowledge of the interests and political calculations of the leaders, parties and factions involved.

Experts on early warning are often rightly wary of quantifying these issues: “the exact degree to which elites are in harmony or conflict, to which opposition movements have popular support, or to which the ruler is supported by neighbouring or foreign states, is not always easy to pin down”.[fn]Jack A. Goldstone, “Using Quantitative and Qualitative Models to Forecast Instability”, United States Institute of Peace, 1 March 2008, p. 6.Hide Footnote  It is arguably even harder to anticipate the precise trains of events that lead to specific acts of violence. Some potential flashpoints, such as divisive elections or the death of an authoritarian leader, may have a high chance of engendering instability. Nonetheless, “long term trends (‘causes’) are often clear enough, but not the proximate causes, or triggers…. What precipitates a conflict may be a sudden, unforeseen event: an accident, misreading or miscalculation, or a temperamental leader’s flash of hubris”.[fn]Joost Hiltermann, “Chemical Wonders”, London Review of Books vol. 38 (3), 4 February 2016, p. 3. Hiltermann is Crisis Group’s Middle East and North Africa Program Director.Hide Footnote

We can, however, still identify and assess the political factors that make a crisis more or less likely and explore how that crisis might play out. Through detailed information gathering and analysis, it is possible to show how the policies and strategies of leaders and other power-brokers are liable to raise tensions, destabilise societies and initiate conflicts. It is feasible to foresee the political dividing lines that might emerge at trigger moments in the future. In some cases, this information also allows observers to estimate how the ensuing confrontations could unfold, if often only roughly, offering a spectrum of possible developments. A review of Crisis Group reports demonstrates the potential of such analysis to flag looming risks, even if it cannot identify precise triggers.

A good example of a medium-term warning comes from Crisis Group’s work on Iraq. In August 2013, it published a report highlighting sectarian tensions and that “Prime Minister al-Maliki has implemented a divide-and-conquer strategy that has neutered any credible Sunni Arab leadership”.[fn]Crisis Group Middle East Report N°144, Make or Break: Iraq’s Sunnis and the State, 14 August 2013, p. i.Hide Footnote  The report warned that many Sunni Iraqis now felt that their “only realistic option is a violent conflict increasingly framed in confessional terms”. While noting the increased activism of the “newly minted” IS, it did not foresee the speed and scale of the group’s offensive in northern Iraq the following year. But in identifying the poisonous effects of Maliki’s political strategy, its impact on Sunni opinion and the probability of armed conflict, it did diagnose many of the drivers of the 2014 crisis. At a time when the U.S. was keen to put the Iraq war behind it, and its allies had disengaged, however, the warning went largely unheeded.

Crisis Group similarly laid out well in advance the dynamics that led to the recent crisis in Burundi, highlighting President Nkurunziza’s strategy of centralising as much power in his hands as possible. Crisis Group insisted in 2012 that the country was “regressing” toward a “one party system characterised by the end of dialogue between the opposition and the ruling party, the government’s authoritarian drift and the resumption of political violence”. Nonetheless, international actors with a stake in Burundi’s stability largely attempted to work with the government rather than confront it: the EU increased aid, while the UN cut back its political presence, even as warning signs grew stronger.[fn]Crisis Group Africa Report N°192, Burundi: Bye-Bye Arusha, 25 October 2012, p. i. Thierry Vircoulon, “Burundi: How to Deconstruct Peace”, International Peace Institute Global Observatory, 24 November 2015. Vircoulon was then Crisis Group’s Central Africa Project Director.Hide Footnote

Crisis Group was, for example, one of the first to recognise how the growing power of the Huthis in Yemen could upset the country’s fragile peace.

These two cases show how focusing on the strategies and behaviour of leaders can help identify impending crises, even if those strategies and behaviours are informed by deeper contextual factors. Where conflicts intensify, analysts may also be able to identify how short-term political developments may create instability and potentially reshape dynamics. Crisis Group was, for example, one of the first to recognise how the growing power of the Huthis in Yemen could upset the country’s fragile peace. The group did not initially appear to be a major spoiler, but military success turned it into a significant and ambitious political force in 2013-2014. In February 2014, a Crisis Group Conflict Alert raised the possibility it would try to take the capital, Sanaa.[fn]Crisis Group Middle East Report N°154, The Huthis: From Saada to Sanaa, 10 June 2014; and “Yemen: Conflict Alert”, 26 February 2014.Hide Footnote  Yet, many outsiders, keen to see Yemen as a success story, focused on UN-led efforts to consolidate a new political settlement and played down the threat until the Huthis did indeed enter Sanaa that September.

Elsewhere, early warnings have more successfully led to early action. In late 2014, Crisis Group was among organisations that emphasised signs Nigeria’s 2015 presidential elections could lead to large-scale violence. Signals included increasing low-level sectarian attacks and local politicians arming followers in anticipation of worse to come. Crisis Group advocated a high-level international push to persuade President Goodluck Jonathan and his opponent, Muhammadu Buhari, to renounce violence. Following intensive personal diplomacy by luminaries such as U.S. Secretary of State John Kerry and former UN Secretary-General Kofi Annan, coupled with pressure by domestic powerbrokers, Jonathan accepted his eventual defeat gracefully. Research in the Niger Delta, a centre of his support, suggests that local leaders had been ready for violence.[fn]Crisis Group Africa Reports N°s 220, Nigeria’s Dangerous 2015 Elections: Limiting the Violence, 21 November 2014; and 231, Curbing Violence in Nigeria (III): Revisiting the Niger Delta, 29 September 2015, p. 14.Hide Footnote

Analysts with a good grasp of the political dynamics can also make credible (if inher­ently probabilistic) assessments of how events will unfold. In late 2011, for example, Crisis Group assessed the increasingly chaotic security picture in Syria and identified factors that have since come to characterise the conflict, including the prevalence of “sectarian retribution and criminal activity” and the mounting risks of “foreign intrusion”.[fn]Crisis Group Middle East Briefing N°31, Uncharted Waters: Thinking Through Syria’s Dynamics, 24 November 2011, pp. 5-6.Hide Footnote

Crisis Group certainly does not always foresee future developments accurately: it did no better than others in forecasting the 2011 Arab uprisings and gave no advance alert of South Sudan’s collapse or the Ukrainian crisis.[fn]Like governments and other actors, Crisis Group’s analysis is affected by resource constraints: it did not have staff in Ukraine in 2013 (this has since been remedied).Hide Footnote  At best, analysts work with partial information and have to make judgment calls about which risks are most pressing. If policymakers are sometimes inclined to discount warning signs, there is a parallel danger of “over warning”, of perceiving every fresh political twist as a harbinger of inevitable conflict.Nonetheless, good analysis-based early warning can identify not only the underlying risks of future conflicts, but also (i) how political actors are exacerbating the dangers of a crisis through their medium-term strategies; (ii) how shorter-term tactical developments may accelerate tensions; and (iii) what possible paths a conflict could take if not controlled. The goal of early action is then to determine how to persuade or push actors to pursue alternative courses that avert or minimise violence, or, where the internationalisation of a conflict is a risk, at least contain it.

B. The Complexities of Early Action

Early action “tools” fall into three broad categories: (i) facilitative (high-level diplomacy, mediation and confidence-building measures); (ii) coercive (diplomatic penalties, sanctions, threats of international justice and, ultimately, use of force); and (iii) incentives (such as financial aid, security guarantees and institutional support for new power- and resource-sharing arrangements). While it is important to understand the strengths and weaknesses of specific tools, they can only rarely be utilised in isolation from each other.

It remains true, as argued in a 2000 essay on early action, that “preventive diplomacy will usually require ‘mixed strategies’ combining coercive elements capable of posing a credible deterrent, and inducements and other reassurances that provide positive incentives for cooperation”. Optimally, such strategies should include a concept of a peaceful end-state to a crisis that all major players can buy into. “In even the most terrible of civil wars, for some there is always a threat more terrifying than the war itself”, noted a former UN official, namely, “the wrong peace”.[fn]Bruce W. Jentleson, Opportunities Missed, Opportunities Seized: Preventive Diplomacy in the Post-Cold War World (Carnegie, New York, 2000), p. 13. Tom Hill, “Fear of the Peace: Why Assad is Not the Main Obstacle to a Deal on Syria”, The Daily Telegraph, 3 October 2015.Hide Footnote  Parties to a conflict may ignore both coercion and incentives if they believe they will lose fundamentally from a final political settlement. This long-term view must be factored into early action where possible, even if there is inevitably always a short-term focus on averting an immediate crisis.

This is all hard. Devising and applying “mixed strategies” to manage emerging or ongoing crises typically involves bringing together not only the tools available to a single government or international organisation, but also pooling the influence and resources of multiple actors, who often have very different short- and long-term perspectives. Even where concerned international actors have roughly similar strategic goals, it can be hard to match up their strategies. In 2015, for example, Crisis Group warned that the Security Council risked undermining regional peace-making in South Sudan by threatening sanctions on six generals who actually favoured a settlement (two were sanctioned, duly creating resentment).[fn]Crisis Group Statement, “South Sudan: No Sanctions Without Strategy”, 29 June 2015. This case underlines the advantage of close political analysis. The Council targeted the generals largely due to their positions; Crisis Group argued they were comparatively moderate due to their views.Hide Footnote

More daunting still, it is sometimes necessary to coordinate strategies with actors with deeply opposed views, as in recent efforts involving regional powers, Russia and the West in Syria. In such cases, it is impossible to distinguish neatly between mediators and parties to the conflict and strategic competitors and diplomatic collaborators. As Crisis Group President Jean-Marie Guéhenno has argued, such situations demand a “multi-layered” response with wider international, regional, national and local diplomatic efforts continuing in parallel.[fn]“The World’s Fragmenting Conflicts”, Crisis Group “Future of Conflict series, 26 October 2015.Hide Footnote

III. Identifying Dangerous Political Dynamics

If diplomats, analysts and international officials want to recognise warning signs of political dynamics that are liable to lead to violence, what should they look for? It is necessary to nod to the cliché that all politics is local, and no two crises play out precisely the same way. Nevertheless, Crisis Group reports highlight four recurrent sets of warning signs: (i) evidence that leaders and elites are adopting political strategies conducive to conflict, or signs of breakdowns in the bargains that hold leaders and elites together; (ii) evidence of discontent or political radicalisation among militaries and security forces; (iii) violence in “peripheral” areas with potentially broader implications; and (iv) signals that outside actors are engaging in an “internal” conflict, or spillover effects from such a conflict. This section concludes with thoughts on how policymakers can use knowledge of such warning signs to build “anticipatory relationships” and take very early preventive action.

A. Leaders and Elites

Some observers argue that there is an “end of leadership” globally, as transnational communications and organisations gain influence at the expense of national figures. Where a leader such as Nigeria’s President Jonathan is willing to release his grip on power, however, underlying political and social tensions can be eased, while a recalcitrant chief can have the reverse effect. As Crisis Group observed in 2011, the initial protests in Syria, having created an “unprecedented sense of awareness, solidarity and responsibility among large segments of the population”, had the potential to engender peaceful change, but President Bashar al-Assad guaranteed wider violence by whipping up the fears of his base, especially in the Alawite community, and signalling his intention to “go down fighting”.[fn]Moisés Naim, The End of Power (New York, 2014). Crisis Group Middle East Briefing N°31, Uncharted Waters: Thinking Through Syria’s Dynamics, 24 November 2011, p. 3.
Hide Footnote

It is essential to understand not only individual leaders but also the political currents around them. As Guéhenno has argued, “political leaders have lost some of their capacity to control outcomes, and multiple actors, from the bottom up, need to be influenced”. This involves understanding the political organisations, factions and elites that underpin any leader, as well as the strength and strategies of opposition groups and the wider constellation of local leaders, armed groups and other secondary players who might exploit a crisis.[fn]Guéhenno, “The World’s Fragmenting Conflicts”, op. cit. Wennmann, “The Political Economy of Violence”, op. cit., pp. 23-26.Hide Footnote

As Guéhenno has argued, “political leaders have lost some of their capacity to control outcomes, and multiple actors, from the bottom up, need to be influenced”

The need to understand such secondary political actors was made clear in South Sudan in 2013. For much of that year, there were signs of an “unravelling” of the ruling Sudan People’s Liberation Movement (SPLM), which had papered over serious internal divisions on gaining independence from Sudan in 2011. Civilian and military factions now jockeyed for control of the party, creating discontent in the army and threatening President Salva Kiir's grip on power. While symptoms of this struggle became increasingly public, UN and Western diplomats focused on working with Kiir. They arguably missed opportunities to engage with a wider range of actors and were caught badly off-guard when the country collapsed into war that December. A host of armed groups and ethnic militias joined in, fighting grew exponentially, and “communal mobilisation and spiralling violence quickly led to appalling levels of brutality against civilians, including deliberate killings inside churches and hospitals”.[fn]Crisis Group Africa Report N°217, South Sudan: A Civil War by Any Other Name, 10 April 2014, pp. 3-5, p. i.Hide Footnote

Egypt has also highlighted the importance of tracking opposition and other interest group dynamics. From their 2012 election, President Mohamed Morsi and his Freedom and Justice Party had a confrontational relationship with the bureaucracy, which went “on an informal strike”. He and his foes adopted polarising policies, culminating in emergence of the Tamarod opposition movement, supported by a mix of “activists, political parties and establishment figures” and later businessmen and religious leaders.[fn]Crisis Group Middle East/North Africa Briefing N°35, Marching in Circles: Egypt’s Dangerous Second Transition, 7 August 2013, p. 3, fn. 6 (interview, senior National Salvation Front member, Cairo, 9 June 2013); p. 4.Hide Footnote  The security services and military manipulated these groups and exploited Morsi’s intransigence to legitimise his overthrow.

In more propitious circumstances, civil society and economic interest groups can act as restraints on violence. The National Dialogue Quartet (a coalition of civil society groups with a strong popular base) helped to avert a similar breakdown in Tunisia in 2014. “In a region where civil-society groups often face repression and are marginalised”, Crisis Group’s North Africa Project Director noted, “the Tunisian example shows the value of having actors outside formal politics play a role in moments of crisis”.[fn]Issandr el Amrani, “Tunisia’s National Dialogue Quartet Set a Powerful Example”, Crisis Group, 10 October 2015.Hide Footnote  In West Africa, Guinea has avoided the full-scale wars that affected many neighbours in part thanks to the role of civil society groups as “powerful balancing mechanisms” against violence.[fn]Crisis Group Africa Report N°178, Guinea: Putting the Transition Back on Track, 23 September 2011, p. 29.Hide Footnote

Where there is high political polarisation and few domestic constraints on violence, leaders can easily initiate civil conflict by design or accident.

“Civil society” is, of course, an amorphous phrase that covers very different types of entities with variable levels of leverage in different societies. These are most likely to have a positive effect when and where they have a solid popular base, and key poli­tical factions have some willingness to compromise. In Tunisia, the Quartet was able to sustain peace in part because the Islamist government chose to relinquish power voluntarily, due both to its leaders’ greater inclination toward compromise and their fear of suffering Morsi’s fate. Where political factions are intent on violence, civil society may only be able to mitigate the resulting conflict. National Christian and Muslim leaders have, for example, called for peace throughout the CAR crisis but could not stop the deterioration in 2013. Some lower-level religious figures actually incited sectarian violence.[fn]Crisis Group Africa Briefing N°96, Central African Republic: Better Late Than Never, 2 December 2013, p. 4.Hide Footnote

This brief cross-section of cases shows that the best way to predict how crises may evolve is to have a clear picture of the politicians and factions at the centre of decision-making and that political drivers of violence must be analysed from a range of angles. First, it is necessary to recognise when a leader is willing to address threats to his/her rule through long- or short-term strategies of polarisation and radicalisation. It can be helpful to focus on inflection points in political processes, such as elections or the date of a constitutionally-set term limit, which are likely to be polarising moments. Of course, analysts should not concentrate solely on such risky moments lest they miss other tensions and flashpoints.[fn]On “inflection points”, see Richard Gowan, Bruce D. Jones, Sara Batmanglich and Andrew Hart, “Back to Basics: The UN and Crisis Diplomacy in an Age of Strategic Uncertainty”, NYU Center on International Cooperation, pp. 12-15. In 2012, for example, Crisis Group warned that violence around Libya’s first post-war election risked “undermining an already fragile transition”. It was relatively successful but arguably lulled many outsiders into a false security sense, so they did not track the ensuing deterioration closely enough. Crisis Group Alert, “Libya’s Elections under Threat”, 3 July 2012. President Jonathan’s concession in Nigeria’s 2015 election may have averted serious violence in the Niger Delta, but that region remains tense. Nnamdi Obasi, “Buhari’s Nigeria: Boko Haram Off Balance, But Other Troubles Surge”, Crisis Group, 30 May 2016.Hide Footnote Secondly, it is important to understand the coalition of political actors that support – or aim to undermine – a leader in his/her party, such as Kiir’s opponents in the SPLM. Thirdly, how opposition forces and civil society may fuel, defuse or mitigate a political crisis must be assessed. Where there is high political polarisation and few domestic constraints on violence, leaders can easily initiate civil conflict by design or accident. Conversely, they may take advantage of conflict abroad to strengthen their position at home, as Russian President Vladimir Putin appears to have done in Ukraine.

B. The Security Sector and the Military

Egypt also highlights the need to monitor national security forces and militaries as sources of instability. A strong army’s capacity to threaten constitutional government is obvious. Yet, it is also necessary to recognise the dangers associated with security forces that have lost status and self-confidence (post-2011 revolutionary Tunisia), lack cohesion to ward off internal and external enemies (Iraq, 2014), are not rooted as an established institution (Libya) or are linked to only part of society (Syria). While outsiders often invest heavily in training and equipping militaries and security forces, their political dynamics tend to be poorly understood.

It is not enough to ask to what degree civilians formally control the military and security structures. In many states, relations between uniformed and civilian authorities are a matter of constant manoeuvre. In the run-up to the 2012 coup that combined disastrously with secessionist violence in the north to push Mali to the brink of collapse, Bamako was “buzzing with accusations of dangerous liaisons between political and military elites and major drug and hostage traffickers and rumours of plots by junior officers angry about the way the president pampered senior officers”. In Tunisia, tensions are growing between the army, Internal Security Forces (ISF), poli­tical parties and the public. ISF “isolation” from the public is a potential source of fresh political friction. There is evidence of breakdowns in the ISF chain of command and “emergence of mutually exclusive clans” in units that limit their ability to fight dangerous Islamist extremists.[fn]Crisis Group Africa Report N°189, Mali: Avoiding Escalation, 18 July 2012, p. 18; Middle East and North Africa Report N°161, Reform and Security Strategy in Tunisia, 23 July 2015, pp. 15-16.Hide Footnote

Outsiders concerned by such security dynamics need to assess (i) whether nation­al security forces have the political cohesion to threaten a government in their own right; (ii) whether their divisions could lead to in-fighting or create security vacuums; and (iii) if, where states face external threats or internal disorder, security forces have the capacity to provide an adequate defence and the discipline and professionalism to maintain public trust.

Troubling examples include the “shambolic” nature of Nigerian law enforcement in areas Boko Haram threatens and the Kabul government’s use of the “cheap and dangerous” Afghan Local Police. The Pakistan army’s “poorly conceived counter-insurgency strategies, heavy-handed methods and failure to restore responsive and accountable civil administration and policing” complicate efforts to oust Islamist extremists from tribal areas, creating or exacerbating more problems than they resolve. Elsewhere, security forces may act as forces for restraint in volatile situations: there are indications Venezuela’s military has played a positive if opaque role in lowering tensions after potentially explosive 2015 elections.[fn]Crisis Group Reports, Africa N°s 216, Curbing Violence in Nigeria (II): The Boko Haram Insurgency, 3 April 2014, p. 32; 237, Nigeria: The Challenge of Military Reform, 6 June 2016; Asia N°s 268, The Future of the Afghan Local Police, 4 June 2015, p. i; 242, Pakistan: Countering Militancy in PATA, 15 January 2013, p. i; and Latin America Briefing N°34, The End of Hegemony: What Next for Venezuela?, 21 December 2015, p. 7.Hide Footnote

The difficulties of assessing the intentions and capacities of formal military and security forces are often compounded by the proliferation of militias and informal armed groups with uncertain affiliations. In the wake of the Minsk II agreement to halt fighting in Ukraine in February 2015, for example, Crisis Group warned that both Moscow and Kyiv needed to be ready for a “mass collapse of discipline” among the militias that had sprung up in the east (including such oddities as “an Orthodox Christian unit, now in schism”). In the event, Russia has kept a firm grip over these groups, but it is sometimes necessary to treat irregular and semi-regular armed groups as serious political actors, not marginalise them. Crisis Group has thus criticised the UN-led political process for not including “a concerted effort to bring [Libya’s] security actors together in support of [a national] government”.[fn]Crisis Group Europe Report N°235, Eastern Ukraine: A Dangerous Winter, 18 December 2014, p. 13. Testimony by Claudia Gazzini, Crisis Group senior analyst, U.S. Senate Committee on Foreign Affairs hearing on “Libya: The Path Forward”, p. 4.Hide Footnote  While outsiders may be tempted to write off “warlords”, it remains necessary to assess – and potentially engage – them as real political actors.

C. Peripheral Conflicts

Studies of armed groups can link to another, often-overlooked challenge to weak states: disruptive political dynamics and trends in violence in peripheral regions, where central authorities have poor relations or little control or oversight.[fn]“Peripheral” refers here not only to geographically remote regions (though in many cases, such as northern CAR, volatile areas are far removed from national power centres), but also to those that are marginalised politically and/or economically, or cut off from state institutions. A more detailed study would also look at the emergence of “peripheral” areas within cities, where services and rule of law are absent. Crisis Group recently chronicled Mexico’s efforts to address social and economic alienation in Ciudad Juaréz as part of its campaign against drug cartels. Latin America Report N°54, Back from the Brink: Saving Ciudad Juaréz, 25 February 2015. Robert Muggah, “Visualizing Urban Fragility”, UN University Centre for Policy Research Blog, 10 February 2016.Hide Footnote  Outside observers sometimes assume that widespread disorder in outlying regions is either insignificant or normal.[fn]National elites can easily fall into the same trap. Addressing the Syrian regime’s instability in 2011, Crisis Group experts Peter Harling and Robert Malley noted that “today’s ruling elite has forgotten its roots. It has inherited power rather than fought for it, grown up in Damascus, mingled with and mimicked the ways of the urban upper class and led a process of economic liberalization that has benefited large cities at the provinces’ expense. The state abandoned vast areas of the nation, increasingly handling them through corrupt and arrogant security forces”. “How the Syrian Regime is Ensuring its Demise”, The Washington Post, 1 July 2011.Hide Footnote  This happened at the start of the CAR crisis, when the Seleka rebel group that overthrew the government in 2013 was initially dismissed as a “heterogeneous consortium of malcontents” from the perennially unstable north east.[fn]Crisis Group Africa Report N°203, Central African Republic: The Priorities of the Transition, 11 June 2013, p. 6.Hide Footnote  Yet, it morphed into a serious threat, as it seized territory, realised it could overthrow the government and became more radical.

Similar threats have emerged in the peripheries of other weak states with highly destabilising results: recent major crises have often been tied to regions where minority groups feel cut off from, or threatened by, national political dynamics, such as the Tuaregs in northern Mali, ethnic Russians in eastern Ukraine and the Huthis in northern Yemen. In 2015, Nepal’s gradual recovery from civil war was severely set back when its ruling parties rushed through a constitutional statute that alienated minorities, including those in the southern plains, where mass protests contributed to a five-month blockade on goods entering from India.[fn]Crisis Group Asia Report N°276, Nepal’s Divisive New Constitution: An Existential Crisis, 4 April 2016.Hide Footnote  Even where there is no immediate trigger, disputes over ethnic issues, language rights or religion in peripheral regions can combine with economic grievances to create fertile conditions for political tensions and violence.[fn]For a further case study by Crisis Group of a little-known region facing this mix of problems, see Europe Briefing N°63, Georgia: The Javakheti Region’s Integration Challenges, 23 May 2011.Hide Footnote

Watching how authorities handle a specific area’s problems can illuminate dangers affecting the state as a whole. Crisis Group recently explored heavy-handed army tactics in Arsal, a Lebanese border town host to many Syrian refugees, as a case-study of a much wider “self-reinforcing loop in which the measures the government takes to compensate for its shortcomings make matters worse”.[fn]Crisis Group Middle East Briefing N°46, Arsal in the Crosshairs: The Predicament of a Small Lebanese Border Town, 23 February 2016, pp. 2-3. For other recent examples of analysis of peripheral towns and regions see Africa Reports N°s 198, Sudan’s Spreading Conflict (i): War in South Kordofan, 14 February 2013; 204, Sudan’s Spreading Conflict (ii): War in Blue Nile, 18 June 2013; and 209, Sudan: Preserving Peace in the East, 26 November 2013.Hide Footnote

It is often hard for diplomats and international officials in capitals (or further afield) to get a clear picture of developments in peripheral regions. If violence is widespread, they may be banned from travelling, leaving them reliant on other sources, such as humanitarian workers who resent being turned into “spies with food”. Central governments are often happier to relay “news” that is not always reliable: in Russia, “a powerful propaganda machine promotes the ‘success story’ of today’s Chechnya”, despite its continued “intimidation, humiliation and violence”.[fn]Crisis Group Europe Report N°236, Chechnya: The Inner Abroad, 30 June 2015, pp. i-ii.Hide Footnote  Observers should look past such misinformation to ask how direct security threats may emerge from peripheral regions and how arguments over ways to engage these regions may feed back into central political tensions in capitals.[fn]Not all conflicts emerge in peripheral regions. Disorder in Cairo’s Tahrir Square and Kyiv’s Maidan can be equally or more threatening. Politically and culturally symbolic sites are potential flashpoints: the Holy Esplanade in Jerusalem (the Temple Mount and Noble Sanctuary to Jews and Muslims respectively) has become “a microcosm of the Israeli Palestinian-conflict. It sees repeated violent upsurges that never decisively end, only fade; as a final status issue it is in a stalemated peace process [and] its disposition remains unclear”.[8] Crisis Group Middle East Report N°159, The Status of the Status Quo at Jerusalem’s Holy Esplanade, 30 June 2015, p. i.Hide Footnote

D. External Drivers of Conflict

A focus on national leaders, political factions, security politics and peripheral conflicts can give outside observers a clearer understanding of the chains of events that may destabilise a government or create conditions for violence. But, as noted above, it is also necessary to evaluate how these internal factors are tied to external political pressures and outside actors’ interests, and how international actors may assess each other’s engagement. One country’s peripheral conflict may be another’s bid for security or influence: there is evidence that the emergence of Seleka as a threat in CAR was at least abetted by neighbouring Chad. Moscow used ethnic Russian concerns in Ukraine to legitimise its incursions in 2014. Saudi Arabia interpreted the Huthis’ rise in Yemen as proof of Iranian meddling in its backyard, though it may have overestimated Tehran’s original involvement and pushed the Huthis closer to its regional rival by intervening.[fn]Crisis Group Reports, Central African Republic: The Priorities, op. cit., p. 8; Middle East N°167, Yemen: Is Peace Possible?, 9 February 2016, pp. 10-12.Hide Footnote

Strains within security forces and between uniformed and civilian leaders in a fragile country may also be exacerbated by external threats. In 2014, Crisis Group tied growing rifts in the security apparatus to broader anxiety arising from Tunisia’s insecure neighbourhood: “an increase in violence along the Algerian border; the chaotic situation in Libya; the advance of radical Islamism in the Middle East – all made all the more acute by an alarmist anti-terrorist discourse”.[fn]Crisis Group Middle East and North Africa Briefing N°41, Tunisia’s Borders (II): Terrorism and Regional Polarisation, 21 October 2014, p. 1.Hide Footnote  It added that arms and drugs traffickers had become increasingly active along the borders. Transnational criminal networks frequently exacerbate instability in other vulnerable regions. Crisis Group’s Latin American experts, for example, regularly balance political analyses with research on parallel dynamics in the drug trade.[fn]Just as analyses of the Libyan and Ukrainian conflicts must pay attention to the outlooks and capacities of armed groups, it is necessary to see how shifting balances of power inside criminal networks affect their trajectories. “The capture of … local drugs lords has shaken once powerful organisations”, Crisis Group observed regarding Central American trafficking, “allowing a new generation of sometimes more violent leaders to emerge”. Latin America Report N°52, Corridor of Violence: The Guatemala-Honduras Border, 4 June 2014, p. i.Hide Footnote

The activities of cross-border political movements, bound together by ethnicity, faith or strategic calculations, can also easily result in spillover conflicts. This pattern is all too familiar from past Balkans cases and is currently a matter of urgency in the Middle East, where the rise of Syria’s Kurds has contributed to Turkey’s repressive approach toward its own Kurdish population.[fn]Crisis Group Middle East Briefing N°49, Steps Toward Stabilising Syria’s Northern Border, 8 April 2016.Hide Footnote

However, policymakers now tend to prioritise two facets of the internationalisation of conflict: the spread of violent jihadist groups, primarily al-Qaeda and IS, in many troubled states in the Middle East, North Africa and Central Asia; and regional and wider international powers’ roles in proxy wars, including in Ukraine, Syria, Yemen and South Sudan. Given the prominence of these trends in diplomatic discourse, it is worth testing their importance.

There is no doubt that jihadist groups have played a brutal part in recent conflicts in the Arab world, in addition to instigating and inspiring terrorist acts globally. Yet “jihadists’ growing prominence over the past few years is more a product of instability than its primary driver”.[fn]Crisis Group Report, Exploiting Disorder, op. cit., p. 8.Hide Footnote  The Iraqi government’s prolonged marginalisation of Sunnis, noted above, fuelled the rise of IS, which then benefitted from President Assad’s vicious, radicalising response to the Syrian uprising. In Mali, Libya and Yemen, IS, al-Qaeda or other extremist movements have taken advantage of existing instability to seize territory.

The activities of cross-border political movements, bound together by ethnicity, faith or strategic calculations, can also easily result in spillover conflicts.

It would be foolish to argue these groups are not a serious threat in many regions. Their presence vastly complicates efforts to end conflicts, given the increasing military potency of some of them and that their aspirations and ideology are hard to envisage as part of a political settlement; in any case, few show much interest in peace processes.[fn]But Crisis Group believes it is as necessary to understand the political goals of jihadists as of other actors, since “what they want, particularly related to the state system, their openness to sharing power and tolerance toward other sects or religious groups, bears on policy” (ibid, p. 29).Hide Footnote  Overall, though, these groups prey on existing crises and wars more than they start new ones. There are risks Western policymakers will see Syria, Libya, Yemen and the rest of the Muslim world solely through the prism of a renewed “war on terror” (or “countering violent extremism”), targeting jihadist movements but not addressing other, deeper stresses. Applying a counter-terrorism lens to such cases risks stigmatising members of disadvantaged communities as potential extremists, reducing the chance to solve their underlying grievances.

Equally, there is nothing new about outside powers engaging in proxy warfare, subversion and direct intervention in long-suffering states such as Yemen. As Stephen John Stedman underlined on the basis of a monumental study of civil wars in 2001, the greatest threats to peace agreements are “spoilers – factions or leaders who oppose the peace agreement and use violence to undermine it – and neighbouring states that oppose the peace agreement and assist the spoilers”.[fn]“Implementing Peace Agreements in Civil Wars: Lessons and Recommendations for Policymakers”, International Peace Academy, May 2001, p. 2; Stedman, Donald Rothchild and Elizabeth M. Cousens (eds.), Ending Civil Wars: The Implementation of Peace Agreements (Boulder, 2002).Hide Footnote

Nonetheless, the tense international politics that surround many of today’s conflicts – and that so many powers have overt or covert military roles in other states’ wars – fundamentally complicates efforts to analyse and respond to existing and looming crises. In South Sudan, Uganda’s decision to send troops to back President Kiir in the 2013 crisis, coupled with Sudan’s support to his foes, threatened to turn the conflict into a proxy war. In eastern Ukraine, the supposed leaders of the separatist groups know they are “expendable” and that “all major political and military decisions are taken in Moscow, and their implementation is overseen by Russian officials on the ground”.[fn]Crisis Group Africa Report N°223, Sudan and South Sudan’s Merging Conflicts, 29 January 2015; Europe and Central Asia Briefing N°79, Russia and the Separatists in Eastern Ukraine, 5 February 2016, p. 7.Hide Footnote

The Libyan conflict has been fuelled by arms supplies and other forms of military aid from countries including Chad, Egypt, Qatar, Sudan, Turkey and the United Arab Emirates; in some cases this is motivated by security concerns, in others by “ideology and regional rifts, notably over what role Islamist movements such as the Muslim Brotherhood will play in Libya and whether they will use its wealth to support like-minded movements elsewhere”.[fn]Crisis Group Middle East and North Africa Report N°157, Libya: Getting Geneva Right, 26 February 2015, p. 11.Hide Footnote  The regional animosities and great power tensions that have grown up around the Syrian civil war are even more complex and destructive, if it is possible to quantify such things.

This report considers how governments and international organisations may be able to manage such complicated tensions around future conflicts. Yet, the divisions that have sprung up around these cases are not simply the product of chance or bad policy. They represent deeper shifts in the international context for early warning/early action. Much thinking on these issues dates from the first ten to fifteen post-Cold War years, when Western analysts presumed (sometimes optimistically) that the U.S. and its allies could line up sufficient states behind specific conflict management and resolution strategies if only they tried. While Washington retains far more power to play a guiding role in managing conflicts than any other state, the geopolitical context is shifting: lining up political actors for early action is becoming more difficult, a dilemma considered in greater depth below.

E. Beyond Analysis: Anticipatory Relationships and Actions

The preceding pages have laid out a series of issues that should interest analysts and policymakers looking for signs of looming crises. These include: (i) evidence of leaders promoting political polarisation or radicalising their bases, and signs of political elites and parties breaking up and/or interest groups mobilising against leaders; (ii) political discontent and divisions among security forces and military actors; (iii) emerging threats from violent groups in peripheral regions of weak states; and (iv) signs of external actors fuelling conflicts through military engagement, supplying weapons or political and diplomatic means.

Policymakers and analysts need to combine tracking these issues with other indicators, such as economic trends, to strengthen their understanding of potential risks. They should also use knowledge of developments in countries at risk as the basis for early, direct, low-key political action; useful analysis should identify not only how key players are behaving, but also chances to nudge them away from dangerous stra­tegies. Diplomats and international officials can build on early warning by creating a network of “anticipatory relationships” with relevant actors.[fn]“Back to Basics”, op. cit., p. 4.Hide Footnote

Optimally, this should include channels for frank communication with leaders on choices and strategies. Persuading presidents like Burundi’s Pierre Nkurunziza or the DRC’s Joseph Kabila to respect term limits should be a long-term project, not a hasty gamble: “There comes a point when leaders are so entrenched that there seems little point in challenging their right to hold office. Calling it early is therefore vital to avoid a position where all the options look bad”.[fn]Richard Moncrieff (Crisis Group’s Central Africa Project Director), “The reluctance of Joseph Kabila to cede power could push Congo to the brink”, The Guardian, 2 May 2016.Hide Footnote

Outsiders who build close ties with authoritarian leaders can, however, become over-entangled with them. As an ex-UN official put it, “we pick or create a leader who is capable of dealing with the international community, but forget to engage with the rest of the society and political sphere”.[fn]Crisis Group communication, 10 May 2016. Richard Gowan, “The Peacekeeping Quagmire”, Georgetown Journal of International Affairs, vol. 16 (2), pp. 39-46. There is, of course, a parallel danger of choosing “good” rebels and opposition figures who prove to be misleading partners, as in the U.S. administration’s collaboration with Iraqi exile politician Ahmad Chalabi prior to Saddam Hussein’s overthrow. See Loveday Morris and Brian Murphy, “Ahmed Chalabi, Iraqi Exile Who Helped Spur U.S. Invasion, Dies of Heart Attack”The Washington Post, 3 November 2015.Hide Footnote  Diplomats should aim to reach beyond the relatively narrow range of officials, contacts and polite society in a capital to include more opposition figures, security officials and representatives of marginalised communities. As noted, international actors’ lack of insight into South Sudan’s politics in 2013 meant they struggled to engage with its breakdown. It may be possible to cultivate potential “insider mediators” (figures from civil society or official circles in a country at risk who may be able to guide crisis talks better than outsiders) or work via international NGOs with more leeway to engage key opposition actors and non-state armed groups than formal diplomats.[fn]Simon Mason and Oliver Wils, “Insider Mediators: Their Key Role in Informal Peace Processes”, Berghof Foundation for Peace Support, 2009.Hide Footnote

Building anticipatory relationships may go hand in hand with “anticipatory actions”: steps to resolve structural dangers in troubled societies, such as misuse of justice, before they fuel worse trouble. Crisis Group recently highlighted that Bangladesh’s “dysfunctional criminal justice system” has potential to fuel wider conflict by “provoking violent counter responses, benefitting violent party wings and extremist groups alike”. It argued that donors should tie some aid to government efforts to improve this. Western countries often prioritise institutional reforms and capacity-building, as these may offer a path to lasting stability, or at least give fragile states “the tools to deal constructively with the violent potential of future conflicts”. However, it is important to recognise the stakes many actors have in stymying reforms and potential political repercussions. In Bangladesh as elsewhere, “years of partisan recruitment, promotions and postings have polarised… institutions to the point that officials no longer conceal their allegiances”.[fn]Crisis Group Asia Report N°277, Political Conflict, Extremism and Criminal Justice in Bangladesh, 11 April 2016, p. i. Sophie-Charlotte Brune, Anne Kovacs, Anaïs Reding and Maryse Penny, “Crisis and Conflict Prevention Strategies: An International Comparison”, RAND, 2015, p. 4.Hide Footnote

In some cases, outsiders may be better advised to focus on supporting civil society groups and other unofficial actors who may help constrain violence, but doing so requires considerable time, and may meet high-level political opposition. Given the mixed chances of success of such preventive actions, it is necessary to consider the tools that external actors can bring to bear on crises as they escalate.

IV. Dilemmas of Early Action

Successful early action consists of steps – including efforts to facilitate a political process, coerce key actors or create incentives for peace – that may open paths to a sustainable settlement of a crisis. A sustainable settlement may range from tweaking the status quo in an unstable country to make it acceptable to all sides, through steps such as limited political reforms, to a large-scale rebalancing of power, including constitutional changes and leadership transitions. Where conflicts are internationalised, territorial compromises and/or the creation of new regional security arrangements may be necessary, albeit difficult steps.

Outsiders must tread carefully when pursuing these goals. All early action involves engaging in fluid political environments. There is a high chance of political friction, with misunderstandings and miscalculations derailing plans. No form of crisis response is neutral. Domestic actors will always perceive outsiders as biased. In some cases they will still welcome engagement as a means to secure their own goals, resolve complex policy issues or minimise violence; in others, they may decide to misuse such help, for example by extending political talks indefinitely.

Understanding domestic political actors’ intentions and interests, which as argued, is at the centre of early warning, is thus also crucial to effective early action. Pathways outsiders want to help devise to avoid or curtail violence must be based on appreciation of what local factors will accept. In country-focused Crisis Group reports in the first third of 2016, 61 per cent of recommendations were aimed at governments or domestic political actors. External actors often appear unable to do more than encourage contacts to behave responsibly. When it comes to complex steps needed to unravel many crises – reducing political influence over institutions, for example, or reining in security services – even the best-placed outsider usually lacks the insights or contacts to do more than nudge national leaders to act.

There are also constraints on external actors in most cases. Policymakers who consider engaging in an escalating conflict assess whether it is in their own interests to expend the resources and take the risks. Internal political issues and competing bureaucratic priorities may militate against acting, even when good policy options are available. This report does not reflect at length on these problems, but it is essential to keep in mind that even when decision-makers want to launch early action to end a crisis abroad, they do not have infinite resources.

In the current context of internationalised conflicts, policymakers face a further layer of dilemmas: how to balance, align or corral other international actors to follow a more-or-less coherent strategy. The trend toward states acting as both combatants and peacemakers (Saudi Arabia in Yemen, Russia and the U.S. in Syria) has been noted but is only one facet of the growing complexity of conflict management. There is a broader diffusion of conflict prevention and peacemaking responsibilities, with new powers, ambitious regional organisations and non-governmental actors taking roles that might once have been filled by the U.S., its allies or the UN. New actors may vary as markedly in strength and style as China and Chad.

China often appears tentative in crises outside its immediate Asia-Pacific area, even if it has been increasingly prominent in Afghan affairs since NATO drew down its main force there in 2014. By contrast, Chad has tried to downplay its internal weaknesses and develop regional leverage in Africa by “pursuing a strategy of military diplomacy, hoping to lead the fight against terrorism in the region”, including operations in Mali and CAR and the fight against Boko Haram.[fn]Crisis Group Asia Report N°244, China’s Central Asia Problem, 27 February 2013, p. i; Africa Report N°233, Chad: Between Ambition and Fragility, 30 March 2016, p. i.Hide Footnote

Regional and sub-regional organisations have, meanwhile, increasingly attempted to take primary responsibility for conflict issues in their own areas. The best known examples are in Europe and Africa, but others include the Association of Southeast Asian Nations (ASEAN) and, in Colombia, the Union of South American Nations (UNASUR). These actors often enjoy clear advantages of legitimacy and local contacts, but internal political divisions and capacity gaps can hold them back. The Economic Community of West African States (ECOWAS), for instance, is candid about “the lack of coordination and cooperation between [its] different departments and slow implementation of decisions”.[fn]Crisis Group Report, Implementing Peace and Security Architecture (III), op. cit., p. 23.Hide Footnote

In the current context of internationalised conflicts, policymakers face a further layer of dilemmas: how to balance, align or corral other international actors to follow a more-or-less coherent strategy.

Multiple local, regional and other international actors often pile into efforts to resolve new crises and create frictions between themselves. ECOWAS “believes the AU disregards it and tends to take over its role at the first opportunity”. It is not hard to find AU officials equally critical of African sub-regional bodies or convinced the UN treats them with disdain; UN officials grumble that the AU sometimes overreaches.[fn]Ibid, p. 22. On sub-regional organisations’ weaknesses, see Crisis Group Africa Report N°181, Implementing Peace and Security Architecture (I): Central Africa, 7 November 2011.Hide Footnote  The tensions are almost endless and perhaps inevitable.

“Framework diplomacy” is thus an essential element of handling any crisis: working out which international actors should (i) set strategies; (ii) handle direct contacts with key political actors; and (iii) manage information exchange and other practicalities.[fn]“Framework diplomacy” is taken from Jack F. Matlock, Superpower Illusions: How Myths and False Ideologies Led America Astray – And How to Return to Reality (New Haven, 2010), pp. 31-56, referring to U.S.-USSR negotiations over the global framework to end the Cold War.Hide Footnote  Since the Cold War, diplomats have created multiple frameworks for individual conflicts with mixed results, including the Contact Group for Bosnia and later Kosovo, the Quartet for the Israeli-Palestinian peace process and multiple “friends groups” at the UN.[fn]See Teresa Whitfield, Friends Indeed? The United Nations, Groups of Friends, and the Resolution of Conflict (USIP Press, 2007).Hide Footnote  Recent cases include successful, low-key cooperation by Cuba, Chile, Norway and Venezuela to assist Colombia’s peace talks (see below) and the higher-profile, larger and troubled International Syria Support Group (ISSG) Russia and the U.S. formed under UN auspices in 2015.

The South Sudan case shows the complexity of framework diplomacy. After the country’s collapse, regional leaders – including some of those overtly and covertly fuelling the conflict – initially attempted to mediate a solution under the Inter-Gov­ernmental Development Authority (IGAD). After more than a year of failure, the AU, China, Britain, Norway, the U.S. and others joined an “IGAD-PLUS” format as a “bridge between an ‘African solution’ approach and concerted high-level, wider international engagement” that forged a peace deal in August 2015.[fn]Crisis Group Africa Report N°228, South Sudan: Keeping Faith with the IGAD Peace Process, 27 July 2015.Hide Footnote

IGAD-PLUS’s complexity is not unique. In an attempt to resolve the long-running insurgency on Mindanao in the southern Philippines in 2009, Japan, Saudi Arabia, Turkey and the UK formed a “hybrid” International Contact Group with four international NGOs. In the view of one of the latter, this was useful, as “states provide a level of diplomatic leverage that NGOs do not have”, but NGOs “provide specific expertise in conflict transformation, which embassies do not necessarily have”.[fn]Kristian Herbolzheimer and Emma Leslie, “Innovation in Mediation Support: The International Contact Group in Mindanao”, Conciliation Resources, 2013, p. 3.Hide Footnote  The group helped midwife a peace deal in 2012 that has since faltered.

Framework diplomacy can extend to mandating parallel mediators and, in a few, generally difficult, cases such as Syria or Darfur, deploying joint mediation teams and peace operations. Transaction costs are high, but the alternative is often fragmentation of international efforts, as in Libya in 2011, where the AU insisted on mediation, while NATO and the Arab League engaged in military action.[fn]Alischa Kügel, “Three’s a Crowd? Inter-organizational Cooperation in Conflict Mediation”, Global Peace Operation Review, 19 November 2015. Emily O’Brien and Andrew Sinclair, “The Libyan War: A Diplomatic History, February-August 2011”, NYU Center on International Cooperation, 2011, p. 14.Hide Footnote

Multiple local, regional and other international actors often pile into efforts to re-solve new crises and create frictions between themselves.

Concrete interests and trade-offs lie beneath disputes about which international actors should “own” a peace process. Policymakers must balance their approach to one crisis with their stakes in others. European officials cannot help viewing the Middle East through the prism of the refugee issue; the U.S. seeks to complement implementation of the 2015 nuclear deal with efforts to contain Iran’s strategy of “forward defence” in Iraq, Syria and Lebanon; Iranian policymakers try to reconcile implementation of the nuclear deal with fears that the U.S. seeks regime change. Efforts to resolve Burundi’s crisis are complicated by the fact that its troops play significant roles in Somalia and CAR.[fn]Crisis Group Middle East Report N°166, Iran After the Nuclear Deal, 15 December 2015, p. 20; Statement, “Burundi: Time for Tough Messages”, 24 February 2016.Hide Footnote

Once again, these difficulties are not new. The “herding cats” problem in conflict management has persisted since the end of the Cold War; many mediation difficulties in cases such as Burundi echo 1990s Balkans dilemmas.[fn]See Chester A. Crocker, Fen Osler Hampson and Pamela A. Aall (eds.), Herding Cats: Multiparty Mediation in a Complex World, United States Institute of Peace, 1999.Hide Footnote  A sub-set of today’s crises, however, presents especially acute coordination challenges: those that pit the U.S., Russia and China directly or indirectly against one another.

Escalation risks and obstacles to framework diplomacy are especially great in cases such as Syria, Ukraine and the South China Sea. Even if China has often been “tentative” outside its immediate sphere, its “foreign policy decision-making and implementation skew toward stridency” when its core interests are at stake. A similar logic has guided Moscow in recent years and can still gain traction in Washington.[fn]Crisis Group Asia Report N°267, Stirring up the South China Sea (III): A Fleeting Opportunity for Calm, 7 May 2015, p. ii; Europe Report N°231, Ukraine: Running out of Time, 14 May 2014, pp. 18-20.Hide Footnote  Cooperation over a serious crisis is not impossible for the main powers, as the P5+1 (E3+3) process with Iran demonstrated, but it often relies on fissiparous negotiating mechanisms, such as the ISSG and the “Normandy format” for Ukraine.[fn]The P5+1 were the five permanent Security Council members (China, France, Russia, the UK and U.S.) and Germany; E3+3 refers to the same states in a Europe/non-Europe configuration.Hide Footnote  In many cases, leader-to-leader contacts are necessary, which can leave regional allies alienated, risking new tensions: Saudi Arabia, Turkey and Iran have all pushed back against U.S.-Russian efforts to find an accommodation over Syria.Under these circumstances, framing strategies for early action in looming crises typically involves acting on at least three levels: (i) preparing a response to the immediate circumstances; (ii) assessing and addressing regional political dynamics; and (iii) where necessary, engaging with international powers. Understanding and balancing the external actors’ competing priorities is difficult and time-consuming. A key dimension of any early action strategy should be rapid, multi-level diplomacy to bring on board as many actors as possible. This also involves understanding which policy tools may affect a crisis.

A. Facilitation

Facilitative tools make the most straightforward contribution to creating short-term pathways to avoid violence, if they have political credibility. These include (i) deployment of senior officials for leader-to-leader talks; (ii) mediation; and (iii) confidence-building, including military or civilian monitoring missions. Governments and international organisations have invested heavily in mediation in recent years, often setting up special units to assist high-level envoys.[fn]Multilateral examples include the UN Mediation Support Unit, EU Mediation Support Team and Commonwealth Good Offices Section.Hide Footnote  Even so, “the institutional capacity to provide effective support has not caught up with the collective aspiration to offer it”. Senior envoys “resist the idea of support outside their trusted staff, grounded in the confidence that they have been engaged for their lifetime experience and authority, and no further expertise or training is required”.[fn]Teresa Whitfield (senior adviser to Crisis Group’s president), “Support Mechanisms; Multilateral, Multi-Level and Mushrooming”, Global Peace Operations Review, 17 December 2015.Hide Footnote

While this confidence is not always justified, senior figures remain in demand as crisis managers. Crisis Group frequently notes that outsiders’ best entry-point for dealing with a crisis is leader-to-leader contacts, as shown by the contributions of envoys such as U.S. Secretary of State John Kerry in persuading Nigeria’s President Jonathan to avoid post-electoral violence in 2015. It is now the norm in many African crises for serving or ex-senior politicians to mediate rapidly developing conflicts. South Africa’s Thabo Mbeki has had an extensive post-presidential career in cases such as Côte d’Ivoire and the Sudans.[fn]See African officials’ comments on the uses of high-level mediators in Nora Gordon, “Meet the Envoys”, Global Peace Operations Review, 7 December 2015.Hide Footnote  Nor is this solely an African phenomenon: German Chancellor Angela Merkel and French President François Hollande have been essential interlocutors with President Putin; Kerry bases much of his Syria strategy on ties with Russian Foreign Minister Sergei Lavrov.

Senior political engagement is often tricky to initiate and maintain. Even when willing to engage, top politicians are busy. A good mediator does not mean a particular crisis is actually amenable to resolution or containment: Kofi Annan ended violence in Kenya in 2008 but had little leverage as UN envoy in Syria in 2012. Overreliance on very senior figures can hamper the work of middle- and lower-level officials on details of a political process. Crucial time was lost when IGAD’s leaders took responsibility on South Sudan: “Without them, no one was empowered to advance the process, and often little was done for weeks, and the parties were left to refocus on the war rather than the peace process”.[fn]Crisis Group Report, South Sudan: Keeping Faith, op. cit., p. 16.Hide Footnote

Alternatives to big-name mediators include quiet diplomacy and using NGOs, such as the Centre for Humanitarian Dialogue and Crisis Management Initiative, or local civil society groups to undertake back-channel talks. The value of discreet discussions has been clear in the peace process between the Colombian government and the Revolutionary Armed Forces of Colombia (FARC). While not an early action example, it emerged from a year of secret contacts Cuba, Venezuela and Norway facilitated. That process also confirmed the importance of senior leaders: Venezuelan President Hugo Chávez helped enable it via confidential contacts with both sides that required a high degree of secrecy.[fn]Crisis Group Latin America Report N°45, Colombia: Peace at Last?, 25 September 2012, pp. 1, 16. The conflict, dating back to FARC’s formation in the early 1960s, is one of the world’s oldest.Hide Footnote

Crisis Group often advocates such quiet diplomacy and that there should be no taboo on talking to non-state armed groups, despite the difficulty: “Opportunities to open discreet lines of communication to at least try to define whether groups have demands that could be used as the basis for talks and can be moved away from those that are irreconcilable are usually worth pursuing”. Opportunities have been missed to engage leaders in such groups, including in Afghanistan, Libya, Nigeria and Soma­lia, in ways that might have offered hope of reducing violence.[fn]Crisis Group Report, Exploiting Disorder, op. cit., pp. 45-48.Hide Footnote

A major challenge is often to move as quickly as possible from initial contacts to creation of a framework and recognised process for discussing and defusing tensions. Inclusivity is a recurrent stumbling block: in the rush to set up a mediation process to avert escalation, it is easy to exclude essential participants. Inclusivity is open to interpretation and is often used as shorthand for involving civil society and women in peace processes. As a recent UN report highlighted, this is frequently more a matter of rhetoric than reality: “A study of 31 major peace processes between 1992 and 2011 revealed that only nine per cent of negotiators were women”. There is evidence that such exclusion reduces chances of sustainable settlement. Minority groups in peripheral regions are also frequently excluded, such as the “progressively marginalised” Rohingya in Rakhine State, who have become targets of serial violence during Myanmar’s transition to democracy.[fn]Radhika Coomaraswamy et al., “Preventing Conflict, Transforming Justice, Securing the Peace: A Global Study on the Implementation of United Nations Security Council Resolution 1325”, UN Women, 2015, p. 14. On the importance of civil society inclusion to sustainable settlements, see Desirée Nilson, “Anchoring the Peace: Civil Society Actors in Peace Accords and Durable Peace”, International Interactions vol. 38 (2), 2012, pp. 243-266. Crisis Group Asia Reports N°s 251, The Dark Side of Transition: Violence Against Muslims in Myanmar, 1 October 2013; and 261, Myanmar: The Politics of Rakhine State, 22 October 2014.Hide Footnote

Yemen shows the danger of getting inclusion wrong. In 2011, the Gulf Cooperation Council (GCC), Western powers and the UN stopped immediate violence by setting up a transitional political process, “protecting traditional power centres to prevent war”. This alienated factions in both north and south, including the Huthi movement, which later “thrived by presenting itself as an uncorrupted outsider”. By contrast, pro-government and opposition groups in Mali have made positive efforts to reinforce their political agreements with lower-level pacts “involving local actors and strengthening their trust in a peace otherwise externally imposed”.[fn]Crisis Group Middle East Report N°167, Yemen: Is Peace Possible?, 9 February 2016, pp. 2; Africa Briefing N°115, Mali: Peace From Below? 14 December 2015, p. i.Hide Footnote

There are also risks of processes becoming ends in themselves and of conflict parties deliberately stringing out talks as cover for political games or violence. In eastern Ukraine, “the Minsk process [risks] becoming a substitute for a settlement”, as Russia follows a deliberate strategy of making “parties concentrate more on the process than the settlement”. The Israeli-Palestinian peace process has been criticised as “low-intensity management of the conflict masquerading as the only path to a solution”.[fn]Crisis Group Europe and Central Asia Briefing N°79, Russia and the Separatists in Eastern Ukraine, 5 February 2016, p. 1; Middle East Report N°122, The Emperor Has No Clothes: Palestinians and the End of the Peace Process, 7 May 2012, p. 29.Hide Footnote  In situations where great power interests are at stake, it is difficult to push back. In other circumstances, it may be possible to hustle talks forward by setting timelines and parameters for bargains: though the South Sudan peace process dragged on for nearly two years, regional leaders, the U.S. and China finally strong-armed President Kiir into a peace deal in August 2015.

Alternatively, mediators may try to disaggregate difficult political questions and persuade actors to address specific problems in isolation. This can include focusing on particular economic issues: Crisis Group has argued, for example, that to stop the energy sector falling apart, Libyan factions should hold focused talks on economic governance in parallel with broader political reconciliation efforts. Where no political progress on any level seems likely, it may still be possible to keep open humanitarian talks to minimise suffering or technical military talks to avoid accidents. Crisis Group has highlighted a need for China and neighbours to have functioning capital-to-capital hotlines to handle South China Sea incidents.[fn]Crisis Group Middle East and North Africa Report N°165, The Prize: Fighting for Libya’s Energy Wealth, 3 December 2015; Report, Stirring Up The South China Sea (III), op. cit., noting that hotlines are “important but insufficient” tools for handling potential crises (p. 29).Hide Footnote

Confidence in political processes can also be built by measures such as international observer missions to monitor factions’ behaviour in parallel with talks. In Ukraine, the OSCE’s Special Monitoring Mission has gradually expanded overview of frontline areas; while its position is fragile, it has helped consolidate a reduction in hostilities. Crisis Group has often argued for similar arrangements, such as increased monitoring of the Sudan-South Sudan border in 2014 to address passage of armed groups.[fn]Crisis Group Report, Sudan and South Sudan’s Merging Conflicts, op. cit., pp. 23-24.Hide Footnote  While such presences may help sustain trust, they can easily be marginalised by radical actors: the light UN supervision mission deployed to Syria in support of the Annan peace plan was a courageous effort with little impact.

B. Coercion

The slow progress of many mediations has also led some governments to favour more coercive responses to crises. These may first include relatively limited diplomatic penalties, such as the AU’s threats to suspend members which have had coups, but they have a mixed track record at best. There is minimal evidence repeated resolutions and condemnations from the UN Human Rights Council and General Assembly have influenced Syria’s government since 2011. In some cases, key local players have little knowledge of or respect for the far-off gestures. In others, they do not believe the outrage matters. Crisis Group noted on the CAR crisis that “international organisations always condemn unconstitutional changes of government in Africa, but very rarely try to restore constitutional order by force”.[fn]Crisis Group Report, Central African Republic: The Priorities, op. cit., p. 16, fn. 109.Hide Footnote

This does not mean human rights diplomacy is valueless. In Syria, for example, a UN Commission of Inquiry has gathered much evidence on the use of violence and torture that has challenged false narratives. The Human Rights Council has helped maintain pressure on Sri Lanka to address abuses committed during the war against the Tamil Tigers. Quantitative studies suggest “naming and shaming” helps limit atrocities; nonetheless, statements of concern still often go unheard.[fn]See the Commission’s report “Out of Sight, Out of Mind: Deaths in Detention in the Syrian Arab Republic”, UN document A/HRC/31/CRP.1, 3 February 2016. Jacqueline H.R. De Merritt, “International Organizations and Government Killings: Does Naming and Shaming Save Lives?” International Interactions, vol. 38 (1), 2012, pp. 1-25.Hide Footnote

More stringent forms of coercion include threats of international prosecution of leaders involved in crises and sanctions. The former’s impact is controversial. The International Criminal Court (ICC) appears to have deterred some elites from human rights abuses; other threats of prosecution have been counterproductive. Prospect of prosecution may cause a leader to become more radical. Security Council referral of Libya to the ICC in 2011 left Qadhafi “boxed in”, more willing to fight. In other cases, the evidence is mixed: Crisis Group found that senior Kenyans the ICC accused of crimes relating to the 2007 elections used the cases to “shore up their ethnic bases” before 2013 polls (ironically uniting 2007 foes against the ICC), but also that ICC attention may have helped avert new violence. A wide-ranging study concluded that “the effects of justice mechanisms on the outlook of armed groups or criminal regimes is not likely to hasten an end to atrocities except on rare and unpredictable occasions”, and “states should avoid the use of international justice as an instrumental tool to affect the dynamics of conflict”.[fn]For links to recent and ongoing quantitative studies, see Kevin Burke, “The Deterrent Effect of the International Criminal Court”, Citizens for Global Solutions Blog, 2 March 2015. Jo Becker and Scott Shane, “Hillary Clinton, ‘Smart Power’ and a Dictator’s Fall”, The New York Times, 27 February 2016. Crisis Group Africa Report N°197, Kenya’s 2013 Elections, 17 January 2013, p. 2; Briefing N°94, Kenya After the Elections, 15 May 2013. Anthony Dworkin, “International Justice and the Prevention of Atrocity”, European Council on Foreign Relations, October 2014, p. 44.Hide Footnote

Polarisation over the ICC, especially in Africa, complicates framework diplomacy. Sudan’s President Omar al-Bashir (indicted in 2009 for Darfur) is particularly effective at “mobilising Arab, Islamic and African countries against the court by pitching it as a Western instrument of regime change”. The U.S. and its European allies had prolonged differences over whether to invoke the ICC over Syria, as Washington feared it would hurt diplomacy with Moscow; Russia and China vetoed a French Security Council resolution on this in 2014.[fn]Fouad Hikmat, “Hard Road to Peace After ICC Indicts Bashir”, The East African, 4 March 2009. Ian Black, “Russia and China Veto UN Move to Refer Syria to the International Criminal Court”, The Guardian, 22 May 2014.Hide Footnote

At the least, policy makers should carefully calculate the likely impact of any call for international justice in a crisis. The Security Council referred Libya to the ICC a day after the Human Rights Council instituted a Commission of Inquiry.[fn]Human Rights Council Resolution S-15/1, 25 February 2011. On Security Council-ICC relations, see Eran Sthoeger, “International Courts and Tribunals”, in Sebastian von Einsiedel, David M. Ma­lone and Bruno Stagno Ugarte (eds.), The UN Security Council in the 21st Century (Boulder, 2015), pp. 517-521.Hide Footnote  The latter was arguably a wiser decision and would have been more effective without the former: it sent a message to Qadhafi that he could face legal action one day, but lacked the chilling effect of invoking the ICC. It is impossible to say how Qadhafi would have acted if the Security Council had held back, but the Human Rights Council’s more subtle message was lost.[fn]See also the “Report of the International Commission of Inquiry into Libya”, 8 March 2012 (UN document A/HRC/19/68).Hide Footnote  In some cases, the best way to promote accountability is through special national or hybrid national/international justice mechanisms. Crisis Group has praised the UN-backed International Commission Against Impunity in Guatemala (CIGIC) for empowering citizens and tackling corruption.[fn]Crisis Group Latin America Report N°56, Crutch to Catalyst? The International Commission Against Impunity in Guatemala, 29 January 2016. Though supported by the UN, CIGIC follows Guatemalan law and works closely with domestic prosecutors.Hide Footnote Yet, developing such mechanisms takes time and requires consent from local power brokers that may be unavailable in a fast-moving crisis.

The value of sanctions, both targeted against individuals and entire sanctions regimes, is also often contentious, as are cuts in aid. Sanctions and economic penalties have helped bring some conflicts to a decisive conclusion – as in the EU’s economic isolation of Côte d’Ivoire, 2010-2011 – and appear to have shaped, in combination with many other factors, recent calculations in Iran and Myanmar. Crisis Group has underlined the need to maintain sanctions pressure on Russia over Ukraine. Elsewhere, economic tools have been less effective, especially in changing the short-term thinking of leaders. Denied some EU economic aid in 2015, Burundi simply took cash from elsewhere in its budget.[fn]Crisis Group Briefings N°s 79, Russia and the Separatists, op. cit., p. 2; N°111, Burundi: Peace Sacrificed?, 29 May 2015, p. 3.Hide Footnote

A study found that 22 UN targeted sanctions regimes led to an increase in corruption and criminality in 69 per cent of cases and strengthening of authoritarian rule in 54 per cent, while only 22 per cent could be broadly classed as successful.[fn]Thomas Biersteker et al., “The Effectiveness of United Nations Targeted Sanctions”, The Targeted Sanctions Consortium, November 2013, p. 17. See also Sue Eckert, “The Role of Sanctions”, in The UN Security Council, op. cit., pp. 413-439.Hide Footnote  Crisis Group typically emphasises that sanctions are only worthwhile policy tools if embedded in a wider political strategy, have a clear purpose, enjoy sufficient multilateral support to be effective and are tied to clear conditions for their lifting.

The risk of applying sanctions without strong political messaging on goals beca­me clear in tense periods of the Iran nuclear negotiations. Crisis Group warned:

Critical differences exist between how policymakers in Washington and Brussels on the one hand and Tehran on the other view and interpret the sanctions regime. … the West views it as an instrument of coercive diplomacy, primarily designed to pressure Tehran into curtailing its nuclear activities … Iran sees it, and indeed the nuclear issue as a whole, as a thinly disguised pretext to undermine the regime.[fn]Crisis Group Middle East Report N°138, Spider Web: The Making and Unmaking of Iran Sanctions, 25 February 2013, p. 39.Hide Footnote

While the Iranian leadership has not completely shaken off its suspicions of Western intentions, improved communication over the aims of sanctions after 2013 helped make a deal feasible.

In sum, threats of international justice and sanctions may nudge political actors away from escalation but must be tied to a path back from violence.[fn]A partial exception in the sanctions field are arms embargoes, which may limit the deadliness of a war but are porous.Hide Footnote  They should be used sparingly and pointedly.

It is less clear whether military force can be applied in a controlled fashion. Crisis Group does not object in principle to use of force. It has advocated that the U.S. retain its forces and even conduct further military activities in Afghanistan as part of a broader strategy of balancing and engaging with the Taliban in recent years. However, it has also raised concerns about many recent efforts to resolve escalating conflicts by military means in the absence of an overarching political strategy. It warned in 2011 in Libya that “Western calls for military intervention of one kind or another are perilous and potentially counterproductive”. After NATO’s air campaign began, it argued for a ceasefire and search for political settlement, because Qadhafi’s fall could be followed by “a potentially prolonged vacuum that could have grave political and security implications for Libya’s neighbours as well as aggravate an already serious humanitarian crisis”. Recently, it said the Saudi-led operation in Yemen “did more to terrorise civilians than to harm the Huthis”.[fn]Crisis Group Impact Note, “Pushing for a U-Turn in Afghanistan”, 2015; Media Releases, “A Ceasefire and Negotiations the Right Way to Resolve the Libya Crisis”, 10 March 2011 and “Libya: Achieving a Ceasefire, Moving Toward Legitimate Government”, 13 May 2011; and Middle East Report N°167, Yemen: Is Peace Possible?, 9 February 2016, p. 4.Hide Footnote

The arguments against military interventions are as old as thought about warfare: once underway, initially limited campaigns tend to take on their own logic, strategic goals change, and violence can breed resistance that can itself escalate dangerously. Crisis Group thus often urges actors that insist on military action to limit its duration (as in calls for a Libyan ceasefire) and link it to political goals that targets can understand. In Yemen, for example, it urged the Saudis to “communicate specific security requirements” to help end their campaign.[fn]Crisis Group Report, Yemen: Is Peace Possible?, op. cit., p. iv.Hide Footnote

It is also necessary to consider the political implications of covert operations and support to proxy groups, common tools for big powers in recent crises. The U.S. and its allies have given covert support to Syrian rebels and Kurdish groups in Iraq. This is sometimes seen as a limited option, involving few or no (declared) boots on the ground and relatively inexpensive. Yet, it has unintended effects, such as empowering Kurds against the state, exacerbating Baghdad’s challenge to restore a national order and “giving the Kurds not only greater military capability, but also diplomatic cover”. Supposedly trustworthy militias can collaborate with more radical forces. Outside actors may not view such actions as limited: Russia reacted to U.S. and Arab covert support in Syria with its own direct intervention in 2015.[fn]Crisis Group Middle East Report N°158, Arming Iraq’s Kurds: Fighting IS, Inviting Conflict, 12 May 2015, p. 25. “Coalition members see arming the Kurds as a quick security fix, divorced from broader political considerations”, ibid. Jeremy Shapiro (ex-U.S. official), “Obama’s Syria Failure is a Perfect Case Study in How Bad Foreign Policy is Made”, Vox World, 16 March 2016.Hide Footnote  Even when using covert means, governments must recall the political goals they are meant to serve and the diplomatic messages they send.

Military deployments can make immediate and medium-term contributions to peace in many situations, but it is necessary to be realistic about their limits.

Similar considerations apply to another form of military action not always classed as “coercive”: deployment of multilateral peace operations and stabilisation missions. Mandating these is common in post-Cold War conflict resolution. Over 100,000 troops and police are under UN command worldwide, with entities such as the AU, EU and NATO heavily involved. UN officials warn against deploying troops where there is “no peace to keep”, but there is a trend in the Security Council and elsewhere to do just that.[fn]See the data gathered at www.peaceoperationsreview.org. John Karlsrud, “The UN at War: The Consequences of Peace-Enforcement Mandates for the UN Peacekeeping Operations in the CAR, the DRC and Mali”, Third World Quarterly, vol.36 (1), pp. 40-54.Hide Footnote  The AU and African sub-regional bodies have pushed for early deployments to stabilise countries in crisis, including Mali and CAR. There is evidence that deploying peacekeepers leads to significant reduction in violence, but it has risks: African troops have come dangerously close, particularly in CAR, to being a conflict party. The risks are also high for peacekeepers in such cases: the UN has lost over 60 to insurgent attacks in Mali since 2013.[fn]See on statistical evidence in favour of peace operations, Lisa Hultman, Jacob Kathman and Megan Shannon, “United Nations Peacekeeping and Civilian Protection in War”, American Journal of Political Science, vol. 57 (4), pp. 875-891; on risks, Somini Sengupta and Alan Cowell, “Chad, Amid Criticism, Will Pull Troops from Force in Central Africa”, The New York Times, 3 April 2014; on UN Mali casualties, Olga Abilova and Arthur Boutellis, “UN Peace Operations in Violent and Asymmetric Threat Environments”, International Peace Institute, March 2016.Hide Footnote

The long-term presence of peacekeepers can freeze political divisions rather than provide a framework for their resolution. The UN’s DRC mission has faced significant implementation challenges: “political agreements, military operations and attempts at reconstruction have all come up against the same problem … there is a lack of political will to implement agreements and organise indispensable reforms”. Crisis Group has emphasised that “peacekeeping is a tool, not a strategy”, and CAR requires far broader support. Operations often settle into a cycle of “haphazard crisis responses, contingent decisions, and unintended consequences” (otherwise known as dealing with “one damn thing after another”) and can lose strategic direction. In some cases, international forces end up cooperating with governments that lack popular support or target their own citizens.[fn]Crisis Group Africa Briefing N°91, Eastern Congo: Why Stabilisation Failed, 4 October 2012, p. 13; Africa Report N°219, The Central African Crisis: From Predation to Stabilisation, 17 June 2014. Richard Gowan, “The Security Council and Peacekeeping”, in The Security Council, op. cit., p. 752. For Darfur, Chad and Côte d’Ivoire cases, see Colum Lynch “See No Evil, Speak No Evil: U.N. Covers Up for Sudan’s Bad Behavior in Darfur”, Foreign Policy (online), 21 November 2014; and Giulia Piccolino and John Karlsrud, “Withering Consent, but Mutual Dependency: UN Peace Operations and African Assertiveness”, Conflict Security And Development vol. 11(4), pp. 467-472. Recurrent revelations of sexual abuse by UN peacekeepers exacerbate the problem. Marco Odello and Rósín Burke, “Between Immunity and Impunity: Peacekeeping and Sexual Abuses and Violence”, The International Journal of Human Rights, vol. 20(6), 2016, pp. 839-853.Hide Footnote  Military deployments can make immediate and medium-term contributions to peace in many situations, but it is necessary to be realistic about their limits.

C. Incentives

If facilitation is complicated and coercive strategies can backfire, can outsiders respond to looming crises by offering political actors incentives to avoid or limit violence? These can take many forms, including proposals to win over leaders under pressure. The initially successful 2011 political transition plan for Yemen specifically protected President Ali Abdullah Saleh from domestic prosecution. In Darfur, Crisis Group suggested the Security Council could suspend ICC pursuit of President Bashir if Sudan adhered to its promises to end violence and promote reform in the region.[fn]Crisis Group Middle East Report N°125, Yemen: Enduring Conflicts, Threatened Transition, 3 July 2012, p. 1; Africa Report N°152, Sudan: Justice, Peace and the ICC, 17 July 2009. Saleh faced a simultaneous sanctions threat; the UN referred transitional justice questions to the Yemeni national dialogue conference, rather than endorsing a blanket amnesty (Crisis Group communication with UN official, 4 May 2016).Hide Footnote  More broadly, incentives fall into three categories: (i) offers of specific assistance to implement political agreements to avoid violence; (ii) greater economic and political assistance to reduce short- and medium-term risks and tensions; and (iii) efforts to help factions come to long-term agreements on the division of resources or political responsibilities to lower their interests in violence.

Help for implementation of political agreements to avoid or end violence can range from specific proposals to back new political mechanisms to broader efforts to support processes such as disarmament, security sector reform or local forms of transitional justice. In many mediations, such structures are agreed on for political reasons without reference to costs. The South Sudan peace deal, which, as noted, involved a convoluted high-level process, demanded “more than twenty new institutions, [including] the cantonment of tens of thousands of fighters and … other costly provisions”. China helped resolve one obstacle by offering electrical generators for cantonments, but donors are wary of such projects, as hard to evaluate and frequently open to corruption and waste. The head of the UN Peacebuilding Support Office has complained, “in 2013, … support to legitimate politics, security, and justice systems represented only 16 per cent (or $6.8 billion) of the $42 billion in gross development assistance for 31 conflict-affected countries”.[fn]Casie Copeland, “South Sudan’s Peace Needs More than Tents and Generators”, Crisis Group In Pursuit of Peace Blog, 23 February 2016. Oscar Fernandez Taranco, “A New Deal or a New Global Partnership for Conflict-Affected States?”, Africa in Focus, 30 March 2016.Hide Footnote

In South Sudan and other cases, donors need to overcome qualms and support post-conflict mechanisms to help solidify peace agreements (while assessing which mechanisms are truly useful and which may be marginal). Nonetheless, as in other cases of capacity building, it is necessary to recognise that these are inherently political, not solely technical processes. Efforts to disarm or reform corrupt militaries should take into account factions and patronage networks that have grown up inside them and who loses from reform. In cases like Ukraine, though “the inter-penetration of the corrupt political class and super-rich oligarchy is the main obstacle to reform”, it is necessary to offer financial aid to keep the state afloat.[fn]Andrew Wilson, “Survival of the Richest: How Oligarchs Block Reform in Ukraine”, European Council on Foreign Relations, April 2016, p. 3.Hide Footnote

Such economic assistance may at times seem to do little more than buy time and make limited contributions to lasting resolutions. However, outsiders may be able to play a positive, if less direct, role by helping antagonistic leaders and factions identify long-term mechanisms for distributing resources. This is an important element of creating a path away from conflict: if actors agree on possible economic bargains, they may chart political ways forward together.

Resolving Crisis In the Middle East: An Iranian Perspective


Crisis Group, as noted, has for example advocated a push for national consensus on economic governance in Libya, which relies entirely on energy sales, and “the institutions that manage production, sale and export of oil and gas and the wealth they generate … remain the bedrock of what is left of the state and a key to its control”.[fn]Crisis Group Report, The Prize, op. cit., p. 1.Hide Footnote  As a result, much of that war is driven by contests for physical and political control over these assets and institutions. At the same time, a sustainable unity government “is a prerequisite to tackling the complex issues around security and management of the hydrocarbon economy”. It is potentially a point of leverage for outsiders that framing political discussions in terms of protecting and sharing economic gains may persuade divided parties to compromise.

Crisis Group has applied similar logic, in different circumstances, to the South China Sea, where competition for underwater energy reserves may lead to clashes. A mechanism is needed to reduce immediate risk of unintended confrontation, while collaborative efforts are explored to exploit the resources. Economic planning may help China and its neighbours envisage a compromise to share control.[fn]See Crisis Group Asia Report N°275, Stirring up the South China Sea (IV): Oil in Troubled Waters, 26 January 2016.Hide Footnote

While early action inevitably centres on short-term issues, policymakers trying to avert conflict should start thinking about long-term end-states they can help local actors imagine and achieve. If this sounds hubristic, one must keep in mind that leaders and their followers may refuse assistance if they believe they will end up in a “wrong peace”, ie, a situation in which their fundamental security and interests will be compromised. Sketching realistic terms for a “good peace” (a situation in which all sides feel secure with the outcome) may help reframe leaders’ risk analyses and calculations, though this may take a long time.

In 2006, Crisis Group proposed “delayed limited enrichment” to resolve tensions over Iran’s nuclear program that could include “a several-year delay in … [Iran’s] enrichment program, major limitations on its initial size and scope, and a highly intrusive inspections regime”. That early Crisis Group concept, unpopular with both sides at the time, proved, Foreign Minister Mohammad Javad Zarif recently acknowledged, to be a major contribution to the deal concluded in 2015.[fn]Crisis Group Middle East Report N°51, Iran: Is There a Way out of the Nuclear Impasse?, 23 February 2006, p. i. “Resolving Crisis in the Middle East: an Iranian Perspective”, Australia National University, 15 March 2016.Hide Footnote

D. Bringing Policy Tools Together

None of the tools available for early action are perfect or even consistently useful. There is a tendency for policymakers to adopt what organisational theorists call the “garbage can” theory of response to crisis: throwing whatever policy tools they have to hand at it and hoping that something will work. This is particularly pernicious where, in the absence of effective framework diplomacy, multiple states and organisations are simultaneously bidding to manage a looming crisis. Different agencies and bureaucracies in powerful states such as the U.S. or diffuse bodies such as the UN often follow confused or internally contradictory policies, despite repeated calls for more integrated approaches. Looking at American policy towards Kurdish groups in Syria, Iraq and Turkey for example, Crisis Group has argued that the U.S. should “unify and clarify its messaging, which has not been understood by many actors in the field, in part because it has not been consistent”.[fn]Michael Lipson, “A ‘Garbage Can’ Model of UN Peacekeeping”, Global Governance, vol. 13 (1), 2007, pp. 79-97. Crisis Group Briefing, Steps Toward Stabilising, 0p. cit., p. 9.Hide Footnote

Even if many calls for greater coherence in crisis response have gone unheeded, it is worth repeating their basic message. Governments and international organisations are likely to perform better in a crisis if they establish clear central planning and oversight functions early to guide overall strategy and can also communicate their strategic goals clearly and explicitly to allies and opponents. In every organisation there are reasons why these tasks are difficult, but in an era when crisis management requires so much framework diplomacy to establish even minimal levels of cooperation, individual players must be clear over their own goals.

V. Conclusion

Though this report has laid out a framework for approaching early warning and early action in an increasingly complex international environment, it must be recognised that all forms of diplomatic engagement are risky and have unpredictable results. It is easy to be pessimistic, but letting conflicts escalate is often risker than addressing them early. Strategic, well-designed engagement predicated on the discipline of close analysis, development of anticipatory relationships and construction of framework diplomacy may all help prevent or limit conflict. As Crisis Group President Guéhenno has noted, “we should approach our responsibilities with humility and set clear limits to our agenda. We must be modest, but we must not be defeatist”.[fn]Jean-Marie Guéhenno, The Fog of Peace: a Memoir of International Peacekeeping in the 21st Century (Washington, 2015), p. 317.Hide Footnote  To the extent resources permit, governments and regional and wider international organisations should thus invest in four key areas:

  • Knowledge and relationships. In addition to economic, demographic and other indicators, policymakers, working directly or through others, should develop the closest possible knowledge of troubled countries’ political systems and those actors who could play essential roles in shaping the outcomes of future crises. Approaches include (i) cultivating channels for frank discussions with leaders and elites; (ii) close monitoring of shifts in political alignments; (iii) deepening contacts with militaries and security services to understand their political positions; and (iv) tracking tensions in volatile peripheral areas.
  • Framework diplomacy. Both in anticipation of and in the immediate run-up to potential crises, policymakers should place early emphasis on constructing diplomatic frameworks and mechanisms among regional and wider powers to discuss policy options and reduce tensions. The faster frameworks can be constructed, whether under the aegis of international organisations or ad hoc, and the more robust the discussions they permit, the more likely they can provide a platform for resolving or at least managing an otherwise divisive crisis.
  • Strategic planning and communication. Where powers or organisations choose to engage directly in a crisis, they must define their goals early and clarify them to both domestic and external players rapidly and explicitly. This is especially true when overt or covert military options are in play.
  • Creating pathways to peace. While decision-makers can use a wide range of tools to try to resolve a crisis, these should be paired with ideas and proposals for a mutually-beneficial peace agreement that can offer a long-term horizon for antagonists to aim at. Outsiders can use back channels, second-track diplomacy and other means to define peace parameters. To do this, however, they need the mix of analysis and relationships required in the first place for effective early warning; indeed, it is hard to know how a conflict can be avoided or halted without a thorough sense of how it begins. Effective preventive diplomacy begins with getting deep inside the dynamic of a conflict – a process that involves grinding analytical work, political risk-taking and uncertain success, and yet can, if done right, create a basis for avoiding unnecessary crises.

Brussels, 22 June 2016

Farmers from the Armenian village of Chinari collect grapes. The land is situated metres away from hills controlled by Azerbaijan. The farmers said shootings were usual, but in 2019, the calm situation allowed them to collect their harvest. September 2019 Jorge Gutierrez Lucena/CRISISGROUP
Report 259 / Europe & Central Asia

Preventing a Bloody Harvest on the Armenia-Azerbaijan State Border

Fighting in July interrupted what had been a stretch of relative quiet on the Armenia-Azerbaijan border. The incidents underscored how quickly and unexpectedly this front can erupt. The two countries should take better advantage of a hotline created in 2018 to avoid dangerous misunderstandings.  

  • Share
  • Save
  • Print
  • Download PDF Full Report

What’s new? Deadly July 2020 clashes between Armenian and Azerbaijani forces left dozens dead, civilians among them, and forced villagers to flee their homes on the Armenia-Azerbaijan state border. Shooting across the trenches along the border is more frequent today than anywhere else on the Nagorno-Karabakh conflict’s front lines.

Why does it matter? Efforts by Baku and Yerevan, including through limited diplomacy, a communication channel set up in 2018 and an agreement between the two sides to safeguard farmers, have largely failed to create conditions that would deter people from leaving border areas. Violence there also risks permanently damaging wider peace efforts.

What should be done? The two sides should use the communication channel to warn each other about planned engineering works or other activities that might be misconstrued and lead to escalation. They should begin talks on limited cooperation to allow farmers to harvest crops, repair water networks and clear mines.

Executive Summary

The mid-July 2020 escalation on Armenia and Azerbaijan’s densely populated state border, which killed over a dozen people and sent women and children fleeing, should sound as a warning. Villagers on both sides of the 230km, trench-lined border have long lived in fear of clashes and landmines. Three decades after the 1992-1994 war over the breakaway region of Nagorno-Karabakh, children go to schools behind ever-thicker cement walls, farmers shun fertile but mine-riddled fields and young people seek their future elsewhere. International mediation efforts have largely ignored the border regions, focused on untangling disagreements over Nagorno-Karabakh itself. But July’s violence shows how critical it is to pay attention to the safety of more than 150,000 civilians living there. Yerevan and Baku should keep channels open to find mutually beneficial ways to cooperate along the border. The shooting should not stop them from exploring collaboration on narrow initiatives to allow children to attend school, farmers to harvest crops, herders to put livestock to pasture, and water to flow to taps and fields.

For both sides, the potential cost in lives and property of violence is higher along the border than in other areas. Neither side has a clear military advantage in the border zone. Military positions and front-line trenches are so close in places that residents can shout to soldiers on the other side. The border is at the crossroads of the three post-Soviet states of the South Caucasus – Armenia, Azerbaijan and Georgia. As such, it is criss-crossed by key roads, railways and pipelines pumping natural resources from Russia to the region or from Azerbaijan and the Caspian Sea to Europe. Both sides not only have an interest in avoiding disruptions to this infrastructure but also stand to benefit from repurposing and repairing sundered cross-border Soviet-era water supply networks.

A looming economic crisis wrought by the COVID-19 pandemic injects even greater volatility into the standoff

Simmering tensions have long choked efforts by both Azerbaijan and Armenia to stem emigration from their border regions. Both sides have built protective walls around public buildings and key roads, dug bomb shelters and instituted crisis training in schools. They have tried a mix of tax incentives, subsidies and other initiatives to spur development and built irrigation works to help farmers. These measures have done little to stop the outflow of people with the means to leave, while some ageing residents have not bothered to repair damaged homes, whose crumbling stands as testament to their fatalism. A looming economic crisis wrought by the COVID-19 pandemic injects even greater volatility into the standoff, making it harder for people in the region to make ends meet. Longer-term climate trends will likely also exacerbate shortages of water and arable land. Only through cooperation can Baku and Yerevan make a lasting difference in the lives of people in the border regions.

The two sides should not now walk away from their breakthrough accords of 2018 and 2019. Those understandings for the first time in over fifteen years reopened direct communication lines between security personnel and political representatives in both capitals. The channels, which have so far only been used in the wake of flare-ups, served in recent days to allow for retrieval of the dead. The 2019 detente also led the parties to agree for the first time to show particular restraint along the border during the harvest, allowing for a modest expansion of farming. The fresh violence now threatens the livelihoods of many facing the impossible choice of leaving their crops to rot or risking their lives gathering their produce for market. Were the two sides to expand communication to include preventive messaging, such as notification of planned engineering works or dates of harvests, they might avoid future misunderstandings or escalations. With time, doing so may also build good-will to allow for further, limited cooperation on areas of mutual interest, such as fixing critical water infrastructure and demining farmlands.

The July clashes, the most serious between the two parties since 2016, also risk hardening attitudes concerning Nagorno-Karabakh. The death of a popular Azerbaijani general in the border skirmishes pushed tens of thousands of angry citizens into Baku’s streets to call for revenge and the return of Nagorno-Karabakh. More violence on the border would only fan those flames. Diplomatic progress on the border not only would be easier to accomplish than progress over Nagorno-Karabakh, but also, at this point, is an essential prerequisite for it.

Baku/Yerevan/Brussels, 24 July 2020

Preventing a Bloody Harvest on the Armenia-Azerbaijan State Border


I. Introduction

The Armenia-Azerbaijan border zone was not always a place of entrenched rival militaries and palpable hostility breaking out in periodic shooting. Older residents remember the close ties that bound people across the boundary the Soviet Union had drawn between its two republics. The Azerbaijani side was home to the biggest regional markets, while some Armenians were fluent in Azeri and studied or completed their military service in the republic of Azerbaijan.[fn]Trade persisted after the 1992-1994 war in neighbouring Georgia, stopping only in 2006, when Georgian authorities closed the Sadakhlo market for reasons unrelated to the Armenia-Azerbaijan conflict. See “From War Economies to Peace Economies in the South Caucasus”, International Alert, 2004, pp. 218-226. The market was officially shut down in 2007. See “Sadakhlo Neutral Zone for Armenian-Azerbaijani Contacts: Use of Trade as a Tool for Conflict Transformation”, Caucasus Edition, 1 April 2011.Hide Footnote

Yet it is in these once-intermingled border villages that ethnic violence between Armenians and Azerbaijanis first erupted in the late 1980s, stirred by irredentism in Nagorno-Karabakh. In 1991, upon the Soviet Union’s collapse and independence for both countries, the former administrative border between two republics became an international border. Three western Azerbaijani districts lie on one side of the trench-lined divide and the Armenian region of Tavush on the other. The ethnic violence and ensuing 1992-1994 war over Nagorno-Karabakh left hundreds of people displaced from centuries-old settlements on both sides of the border.[fn]Most Armenians displaced from the border area come from Artsvashen – an Armenian village inside Azerbaijan that saw clashes in August 1992. Around 710 families had to flee their homes. See “Artsvashen’s fall was due to its location, says village governor”, Tert.am, 3 August 2013; and Tatul Hakobyan, “Armenian Border: Tavush”, ANI Armenian Research Centre, February 2015, p. 10-11 (Armenian). Around 7,803 Azerbaijanis were displaced from seven villages in Gazakh district (Baganis Ayrym, Kheyrymly, Ashaghy Askipara, Barkhudarly, Sofulu, Gyzylhajyly and Yukhary Askipara). See “Consequences of the Aggression of Armenia against Azerbaijan”, State Statistical Committee of Azerbaijan, 2019.Hide Footnote Though recognised by the UN, the border remains undemarcated, and soldiers from both countries are dug into positions on both sides.[fn]Armenia and Azerbaijan never officially delimited their borders after the Soviet Union’s demise. Each army controls patches of territory that lie on the other’s side of the UN-recognised border, which corresponds to the Soviet-era administrative line.Hide Footnote More than 150,000 people live close to the front lines, with 26 Armenian and 84 Azerbaijani villages within 10km of the border.[fn]Today, on the Armenian side, there are 26 villages and one town, which together have a population of more than 38,000, within 10km of the border. Around sixteen of them, with around 14,500 people, are between or right next to the trenches: Chinari, Movses (Mosesgegh), Aygepar, Nerkin Karmiraghbyur, Paravakar, Kirants, Vazashen, Berkaber, Voskepar, Baghanis, Voskevan, Koti, Barekamavan, Dovegh, Berdavan and Kayan. For demographic data, see the de facto population figures in the Tavush region at “Marzes of the Republic of Armenia and Yerevan City in Figures, 2011”, Statistical Committee of the Republic of Armenia, 2011 (Armenian). A marz is an Armenian administrative province. On the Azerbaijani side, there are 84 villages, with a total population of over 110,000, within 10km of the border. The biggest villages, home to some 36,000, are in the Gazakh district. Crisis Group counted the number of villages using the OpenStreetMap database, and completed the population data using the 2009 census. Another census was carried out in 2019, but a thorough breakdown for the border regions is not publicly available. See the “Azerbaijan Population Census 2009”, Statistical Committee of Azerbaijan. Demographic data on Gazakh’s border villages is available at the official website of Gazakh district’s executive authority.Hide Footnote

Tensions in the area have remained high for years, reaching a zenith in 2014, when several weeks of clashes drove many locals to emigrate.[fn]Crisis Group interviews, village residents, Berkaber and Nerkin Karmiraghbyur (Armenia), December 2017 and November 2018. See also “Towards a Secure Future: Community Voices in the Border Areas of Armenia and Azerbaijan”, Saferworld, January 2015.Hide Footnote Another uptick came in 2016, when fighting erupted in the Nagorno-Karabakh conflict zone.[fn]The Nagorno-Karabakh conflict zone includes the Soviet-era Nagorno-Karabakh Autonomous Oblast and the seven Armenian-controlled adjacent territories. See a detailed map at Crisis Group’s The Nagorno-Karabakh Conflict: A Visual Explainer.Hide Footnote The Armenian and Azerbaijani armies faced off along the international border, moving heavy vehicles and artillery closer to the trenches.[fn]Crisis Group interviews, village residents, Paravakar and Berkaber (Armenia), December 2017 and November 2018.Hide Footnote An urgent ceasefire on the fourth day of fighting in Nagorno-Karabakh pulled them back from the brink.

While the border regions have seen relative calm since then, tensions have been rising again in the last two years. Even before the July 2020 skirmishes, the international border area had been the most active stretch of the front line. From 2016 to 2018, clashes along the border accounted for only 7 per cent of all soldiers and civilians killed or wounded in the conflict, but in the last two years the proportion has doubled.[fn]According to official Armenian and Azerbaijani reports, at least nine people were killed and six wounded in violence along the border from September 2018 to the July 2020 clashes. There was one civilian among the dead and two among the wounded. The sides reported at least seven soldiers killed from April 2016 through September 2018. See a detailed database of the relevant official reports at Crisis Group’s The Nagorno-Karabakh Conflict: A Visual Explainer.Hide Footnote Most of these have taken place along the northern part of the border, close to Georgia; the latest fighting flared up on 12 July to the south, between Movses in Armenia and Agdam in Azerbaijan – mountain villages on opposite sides of the border, both surrounded by trenches.

Azerbaijani sniper fire can hit almost everywhere in Movses, a village in Armenia’s north-eastern region of Tavush, pictured here from the Armenian side. The Azerbaijani positions run along the flanks of the high mountain in the distance on the right. CRISISGROUP/Jorge Gutiérrez Lucena

What sparked the July escalation remains unclear; each party accuses the other of conducting the first strike.[fn]“No. 158/20, Statement by the Ministry of Foreign Affairs of the Republic of Azerbaijan”, official website of the Ministry of Foreign Affairs of the Republic of Azerbaijan, 12 July 2020; “Comment by the Spokesperson of the Foreign Ministry of Armenia on the ceasefire violations by the Azerbaijani armed forces in the direction of the Tavush region”, official website of the Ministry of Foreign Affairs of the Republic of Armenia, 13 July 2020.Hide Footnote Yerevan’s defence minister said an Azerbaijani jeep drove close to an Armenian outpost, catching Armenian soldiers off guard: “[F]or as yet unknown reasons, Azerbaijani servicemen got out of the car and moved toward our position”.[fn]“PM: “Armenia’s armed forces keep the situation under full control: no provocative action goes unanswered””, official website of the Prime Minister of the Republic of Armenia, 13 July 2020.Hide Footnote Rather than employing “some form of communication” to retrieve the vehicle, which they abandoned after Armenian troops issued a warning, he said, the Azerbaijanis began shelling.[fn]Ibid.Hide Footnote Baku rejected this version of events, saying its soldiers were on a routine patrol when they were shot at.[fn]Crisis Group interview, senior Azerbaijani official, 21 July 2020.Hide Footnote “The Armenian side ambushed them in a premeditated attack. … The soldiers on patrol had to retreat to their posts, and then the Armenians opened fire on our post with artillery”.[fn]Ibid.Hide Footnote  A few days earlier, Baku’s ambassador to Moscow had stressed that the vehicle was “not a tank. They could have immediately called – got in touch by radio. They have a local way to contact each other, a hotline”.[fn]“Посол Азербайджана в России назвал провокацией со стороны Армении инцидент с применением артиллерии в Товузском районе”, Ekho Moskvy, 14 July 2020 (Russian).Hide Footnote

Violence takes a heavy toll on civilians on both sides of the border.

The mutual accusations of failure to use communications channels before resorting to violence show awareness of these channels’ potential. But in this case the channels availed the sides little: the July fighting wound up being a rare instance in which the sides used heavy weapons before it tapered off. As of 21 July, Azerbaijan was reporting twelve of its military personnel, including a well-regarded general, and one civilian killed, while Armenia was reporting four military casualties and one civilian wounded.[fn]Please see Crisis Group’s The Nagorno-Karabakh Conflict: A Visual Explainer for more details.Hide Footnote

Such violence takes a heavy toll on civilians on both sides of the border. Even in tranquil periods, daily life is constrained by fear of renewed clashes and limited economic prospects. Families keep their children out of school for long periods because several have been hurt on the way there and back. Graves located in no-man’s land or near minefields are left untended – in painful violation of deep-seated traditions. When violence breaks out, as in mid-July, the elderly, women and children are evacuated. Some return within days, but others do not. In the aftermath of the most recent clashes, a villager told Crisis Group that evacuees “are waiting to see what will happen. … It is not safe here”.[fn]Crisis Group telephone interview, village resident, Alibeyli, Tovuz district (Azerbaijan), July 2020.Hide Footnote

Those who can get out do so, usually to look for work in Baku, Yerevan, Russia or Turkey. It is mostly young men who depart, leaving wives, children and parents of pensioner age behind.[fn]Crisis Group interviews, local officials, international organisation representatives, village residents, Ijevan and Yerevan (Armenia), Tovuz and Gazakh districts (Azerbaijan), October-November 2018, June-July 2019 and September 2019.Hide Footnote Reliance on remittances from male family members has led to selective abortion – families terminating pregnancies if the baby is a girl – in the region.[fn]Crisis Group interviews, village resident, international organisation representatives, priest, Armenia, October-November 2018. For more about the problem of selective abortion, see “Listen to Her: Gendered Effects of the Conflict over Nagorno-Karabakh and Women’s Priorities for Peace”, Kvinna till Kvinna Foundation, p. 26; Suzanne Moore, “‘We lose 1,400 girls a year. Who would our boys marry?’: Armenia’s quandary”, The Guardian, 22 February 2018.Hide Footnote “People leave because they lost their last means of living”, said an Armenian priest.[fn]Crisis Group interview, priest, Berd (Armenia), November 2018. Berd was hit by a drone strike during the July 2020 fighting. See “Comment by the Foreign Ministry Spokesperson on the Shelling of the City of Berd by the Azerbaijani Armed Forces”, official website of the Ministry of Foreign Affairs of the Republic of Armenia, 14 July 2020.Hide Footnote “Even if we were allowed access to all our lands … there would be no young people to work there”, said a female teacher in neighbouring Movses.[fn]Crisis Group interview, village resident, Movses (Armenia), November 2018. See also “Announcement by the Ministry of Defence of the Republic of Armenia”, 13 July 2020.Hide Footnote Attempts by Baku and Yerevan to improve security and spur growth have done little to alleviate the area’s plight.

In the wake of the July clashes, this report looks at insecurity and ways to reduce it in the Armenia-Azerbaijan border zone. It is based on dozens of interviews conducted in Baku, Yerevan and localities on both sides of the border from November 2018 through December 2019, as well as telephone interviews since then. It includes the views of former and current officials, diplomats, military and security officers, NGO workers, independent experts, clerics and tens of residents in the border regions. In Armenia, field research took place in the biggest towns and border villages of the Tavush region. In Azerbaijan, it took place in four districts – Gazakh, Tovuz, Gadabay and Aghstafa – and approximately 25 villages during June-July 2019. Crisis Group had official permission to travel to these Azerbaijani districts and meet with Border Service officials in other villages. Officials denied a request to visit sensitive outposts in Gazakh and Aghstafa. The report does not cover the Azerbaijani exclave of Nakhchivan, where greater calm reigns and where the Armenian and Azerbaijani militaries are dug in along a 246km border.

II. Efforts to Improve Security, Stem Emigration

Neither side has a clear military advantage in the border zone, leading to an unstable standoff characterised by frequent ceasefire violations that nevertheless stop short of the fighting seen in the Nagorno-Karabakh conflict zone.[fn]In the Nagorno-Karabakh conflict zone, Armenian forces have a clear advantage. They control most strategic hilltops, especially along the northern and north-eastern sections of the front line. Azerbaijani trenches are located close to civilian areas, especially near the middle of the front line. Almost 300,000 Azerbaijani civilians live within 15km of the trenches. On the Armenian side, no more than 7,000 people live so close to the trenches. See Crisis Group Europe Report N°244, Nagorno-Karabakh’s Gathering War Clouds, 1 June 2017, pp. 4-7.Hide Footnote Settlements in proximity to the trenches render civilians on both sides equally vulnerable. While the prospect of civilian casualties may discourage the sides from using heavy weaponry, it also means that any escalation risks taking a fearsome human toll.[fn]Crisis Group interviews, military officers, Tavush region (Armenia), November 2018.Hide Footnote The resulting anxiety is an omnipresent feature of life in the area, stunting development and overshadowing steps by both governments to reassure their populations.

Efforts by both sides to bolster their position through security relations with neighbouring states have acted as a mutual deterrent: Armenia is a member of the Russia-led Collective Security Treaty Organisation (CSTO), which calls upon members to come to one another’s aid in case of outside attack, while Azerbaijan has strengthened its ties with Turkey.[fn]The 2012 CSTO declaration does not make clear what kind of support a member state can expect in case of foreign aggression. One line reads, “In case of need the mechanism for joint consultations will be brought into action”. The other line contains a commitment from all the member states not to take part in actions that might harm any one of them. See more at “Declaration of the Member States of the Collective Security Treaty”, official CSTO website, 26 April 2012 (Russian).Hide Footnote Since 2015, the CSTO has issued regular statements of concern about clashes along the border, as it did on 14 July.[fn]Crisis Group interview, expert in Russia-Armenia military relations, Yerevan, November 2019. See also “The CSTO Secretariat commentary on the situation on the Armenian-Azerbaijani border that arose on July 12, 2020”, official CSTO website, 14 July 2020.Hide Footnote In the words of a former Armenian official, CSTO membership is “a security umbrella for us, although one with holes in it”.[fn]Crisis Group interview, former official, Yerevan, November 2018.Hide Footnote Russia voiced “extreme concern” over the flare-up in fighting and said it was ready to mediate.[fn]“Russia ready to mediate talks between Armenia, Azerbaijan", Moscow Times, 17 July 2020.Hide Footnote Baku, which is not a member but nonetheless has its own close ties to Moscow and other CSTO members, says only a full-fledged incursion into Armenia by its military – an eventuality it views as unlikely – would trigger CSTO intervention.[fn]“Azerbaijan building up forces in Nakhchivan”, Eurasianet, 10 August 2017.Hide Footnote  After the July clashes, Azerbaijani President Ilham Aliyev complained about Armenia’s appeal to the CSTO: “What does the CSTO have to do with it?”[fn]“President Ilham Aliyev chairs meeting of Cabinet of Ministers”, Azvision, 15 July 2020.Hide Footnote

Both sides also worry that ceasefire violations along the border might damage strategic roads, railways and energy infrastructure.

While the United States, Russia and the European Union (EU) all called on both sides to show restraint amid the July clashes, Turkey threw its full weight behind Azerbaijan, its statements replete with bellicose rhetoric. Armenia will be “buried under their own plot, drown in it, and will absolutely pay for what they did”, Turkey’s defence minister said.[fn]“Turkish Defence Minister Akar: Armenia will definitely pay for what they did”, Anadolu Agency, 16 July 2020 (Turkish).Hide Footnote  Azerbaijan’s ties with Turkey were strengthened by an agreement “on Strategic Partnership and Mutual Assistance”, signed in 2010 just days before Yerevan extended Russia’s lease for military bases in Armenia.[fn]“Azerbaijan and Turkey sign agreement on strategic partnership and mutual assistance”, Trend, 16 August 2010; “Russia secures military presence in Armenia until 2044”, Euractiv, 23 August 2010.Hide Footnote The Baku-Ankara agreement prioritises military cooperation, including mutual assistance in the event of an attack or act of aggression upon either country.[fn]Article 2 of the agreement stipulates that the form and volume of such assistance shall be agreed upon without delay. Article 5 demands that neither party take part in alliances and actions that may be directed against the independence, sovereignty or territorial integrity of the other; or allow its territory to be used for acts of aggression or violence against the other. The full version of the agreement is available in Azerbaijani at the E-Qanun database.Hide Footnote In response to Ankara’s backing of Baku in July, Armenia accused Turkey of “neo-Ottoman policies” and meddling in the South Caucasus.[fn]“Interview of the Foreign Minister of Armenia Zohrab Mnatsakanyan with Sky News Arabia”, official website of the Ministry of Foreign Affairs of the Republic of Armenia, 19 July 2020.Hide Footnote

Both sides also worry that ceasefire violations along the border might damage strategic roads, railways and energy infrastructure. A major road and a railway linking Azerbaijan, Georgia and Turkey run near the Azerbaijani side of the border. The 692km South Caucasus pipeline, which transports natural gas from the Shah Deniz field to the Georgian-Turkish border, lies some 15km from Azerbaijan’s border districts of Aghstafa and Tovuz. The 1,768km Baku-Tbilisi-Ceyhan pipeline also passes through Aghstafa and Tovuz, carrying oil from Azerbaijan to Georgia and Turkey.[fn]See the fact sheet about the Baku-Tbilisi-Ceyhan pipeline at the BP Azerbaijan website.Hide Footnote On the Armenian side, the main highway and a railway connecting the country to Georgia, Russia and other parts of Europe pass through the northern Tavush region, as does a Russian pipeline bearing gas to Armenia.

A boast by Azerbaijan’s military amid heightened tensions on 16 July that its new missile system had ample range to reach a Soviet-era nuclear power plant near Yerevan highlighted both the strategic importance of critical infrastructure and constraints against further escalation.[fn]“Azerbaijani MoD: our army’s missile systems allow us to hit Metsamor nuclear power plant with high accuracy”, APA, 16 July 2020.Hide Footnote After the threat made international headlines and prompted outrage in Yerevan, a high-ranking Azerbaijani official walked it back, saying, “Azerbaijan has no policy of targeting critical strategic facilities”.[fn]“Hikmat Hajiyev: Armenia has deliberately turned Metsamor issue, which poses serious threat for region, into show”, APA, 21 July 2020.Hide Footnote

A wall protects a school from potential Armenian sniper fire in Bala Jafarli village, in the Azerbaijani region of Gazakh. CRISISGROUP/Zaur Shiriyev

Despite the structural checks on escalation by both sides, residents feel the danger of stray bullets acutely and daily. Efforts by both sides to boost villagers’ sense of security with school safety drills and protective barriers around public buildings are no more than a band-aid. “When my son is a minute late [coming home] from school, I fear”, a mother of two said.[fn]Crisis Group interviews, village residents, Haciali and Alibeyli, Tovuz district (Azerbaijan), June 2019.Hide Footnote The barriers have eased concerns among some but serve as permanent reminders of insecurity for others. A 70-year-old who bricked up a window 100m from military positions said, “I feel like I am living in a prison”.[fn]Crisis Group interview, village resident, Mezem, Gazakh district (Azerbaijan), June 2019.Hide Footnote Residents erect most of the walls themselves, using building materials supplied by the respective governments, sometimes resulting in flimsy constructions that cannot withstand sniper fire.[fn]Crisis Group interviews, humanitarian workers, Gazakh and Tovuz districts (Azerbaijan), June-July 2019.Hide Footnote The barriers become useless if troops on the other side move their positions. Residents clamour for more and better protection in the absence of a palpable lowering of tensions. One exasperated Armenian village head said, “One can’t put the whole village behind concrete walls”.[fn]Crisis Group interview, village head, Koti (Armenia), October 2018.Hide Footnote

On both sides of the border, security services hold sway and treat outsiders with suspicion. In recent years, Baku in particular has tightened security measures in the region. Amendments to the laws on state secrets and media enacted in late 2014 limit access by journalists and other non-residents. These measures became more stringent after the April 2016 escalation, according to locals, possibly due to a May 2017 presidential decree on “rules for providing an enhanced security regime in residential areas within the front lines”, though the “rules” are not public.[fn]Presidential decree on “Rules for providing enhanced security regime in residential areas within the front lines”, E-Qanun database, May 2017 (Azerbaijani).Hide Footnote Baku has opened new police stations in most border villages and installed new surveillance cameras.[fn]Seymur Kazimov, “New curbs on conflict zone reporting in Azerbaijan”, Institute for War and Peace Reporting, 7 October 2014; “13 new police stations built in the border villages of Gazakh district”, AIA, 10 October 2018 (Azerbaijani).Hide Footnote Despite the changes, security in the border villages is largely a question of topography. In some villages, Azerbaijani forces are dug into strong defensive positions. In others, the Armenian military holds the higher ground.

On both sides of the border, security services hold sway and treat outsiders with suspicion.

In a bid to present a more human face to residents, Baku replaced the military with the State Border Service in the Gazakh district and parts of Aghstafa in December 2018. The change brought new restrictions on movement, especially for farmers and agricultural labourers, who must get daily permission from the Border Service to work in their fields.[fn]According to Azerbaijan’s 1995 law on the state border, “Persons who do not reside permanently on the border line or in areas between the state border and technical fences of border security agencies are allowed to enter there only when products or services are required. The procedure for the release and temporary stay of such citizens, as well as the rules for water use, agricultural and other activities there, shall be established by the relevant executive authority in consultation with local authorities”. Crisis Group translation from the Azerbaijani.Hide Footnote  “We are now unable to secure our daily income because we can only get into our farmlands for part of the day”, a farmer in Qushchu Ayrim in the Gazakh district said.[fn]Crisis Group interviews, village residents, Quschu Ayrim, Gazakh district (Azerbaijan), July 2019.Hide Footnote  Border guards are also conducting more patrols, including occasional checks of identity documents, in residential areas. The new measures have done little to assuage safety concerns among those living near the front lines and may even have raised tensions, as they provoked alarm on the Armenian side.

Both Yerevan and Baku have also sought to stimulate development in the border villages – as part of efforts to stem emigration. Azerbaijan has repaired roads; increased financial and technical support to small businesses; carried out irrigation work; and granted some villages special status making residents there eligible for gas and electricity subsidies. For its part, the Armenian government has introduced subsidies for electricity, natural gas and drinking water.[fn]See the decree that lists all the subsidies: “Decision by the Government of the Republic of Armenia N 144-N”, 18 December 2014 (Armenian).Hide Footnote  It has spent almost $40 per person in the area annually since 2015.[fn]The estimate is based on population figures and the amounts allocated by the Armenian government in 2015-2018.Hide Footnote Yerevan has also abolished two business taxes to attract investors to the area and is considering getting rid of more.[fn]Crisis Group interview, official, Ijevan (Armenia), September 2019. The cuts came in income and value-added taxes. See “Law on Tax Breaks for Enterprise in the Border Zone”, 20 November 2014 (Armenian). The Armenian parliament is discussing cutting the revenue tax in the border zone from 23 to 10 per cent. See the Tax Code Amendment (Armenian) proposed in November 2019.Hide Footnote  “Our soldiers need to see life to understand who they are protecting”, said a military commander in the Tavush region.[fn]Crisis Group interview, military commander, Tavush region (Armenia), November 2018.Hide Footnote

Crisis Group’s Olesya Vartanyan visits a glove factory set up to promote business and help repopulate the Armenian village of Movses, close to the international border with Azerbaijan. CRISISGROUP/Jorge Gutiérrez Lucena

But these policies have not succeeded in attracting more investment to the region or slowing emigration.[fn]According to 2001 and 2011 census data, the population in four Armenian villages (Movses, Aygedzor, Chinari and Barekamovan) fell by 25-40 per cent; in nine villages (Nerkin Karmiraghbyur, Vazashen, Aygepar, Kayan, Voskepar, Koti, Artsvaberd, Verin Karmiraghbyur and Choratan) by 15-24 per cent; and in five villages (Aygehovit, Paravakar, Berdavan, Tsaghkavan and Berkaber) by 10-14 per cent. See “The Results of 2001 Population Census of RA” and “Marzes of the Republic of Armenia and Yerevan City in Figures, 2011”, Statistical Committee of the Republic of Armenia.Hide Footnote What little money flows into the region comes from the Armenian diaspora, who donate to medical facilities and schools. In a few villages, Armenian expatriates have opened small businesses employing local women.[fn]There is a glove factory, for example, and a cannery, both producing for export to Russia. Crisis Group interviews, local businessmen, Movses, Berd and Aygedzor (Armenia), September 2019.Hide Footnote “It’s more about charity than a real business”, said the manager of one such enterprise.[fn]Crisis Group interviews, local businessmen, Berd (Armenia), September 2019.Hide Footnote

The economic fallout of the COVID-19 pandemic will only make matters worse. In Azerbaijan’s border districts, which account for some 7 per cent of the country’s agricultural production including 35 per cent of its potato crop, locals have faced bureaucratic delays obtaining permission to continue work during lockdowns while their produce rotted.[fn]The Azerbaijani districts of Gazakh, Tovuz, Gadabay and Aghstafa also accounted for 13 per cent of the country’s grapes, 12 per cent of its sugar beets, and 5 per cent of its cereals and beans. “The Agriculture of Azerbaijan”, Statistical Yearbook, 2019.Hide Footnote Others lacked the field hands to sow their crops.[fn]Crisis Group telephone interviews, farmers, Gazakh district (Azerbaijan), 1 May 2020.Hide Footnote The collapse in oil prices has also hit Azerbaijan hard: it will likely drive up unemployment and reduce the budget allocations for border areas. In Armenia, the post-Soviet country seemingly worst affected by the coronavirus per capita with more than 35,000 confirmed infections, an economic contraction is also expected.[fn]See the latest updates on numbers of COVID-19 patients on the official website of the Ministry of Health of the Republic of Armenia.Hide Footnote Remittances, which account for some 13 per cent of GDP in Armenia and almost 3 per cent in Azerbaijan, will likely shrink.[fn]See Global Remittances Guide by the Migration Policy Institute.Hide Footnote The economic pressure may push farmers back to precarious farmlands despite the insecurity in many areas.


III. Preventive Messaging

In September 2018, the then newly elected Armenian prime minister, Nikol Pashinyan, and Azerbaijani President Ilham Aliyev agreed to set up a new communication channel amid an international push to lower tensions following clashes over Nagorno-Karabakh.[fn]“Release of the Press Service of the President”, official website of the President of Azerbaijan, 28 September 2018; Emil Sanamyan, “‘Agreed to reduce tensions’: Aliyev, Pashinyan talk in Dushanbe”, CivilNet, 30 September 2018.Hide Footnote The leaders reached agreement outside the usual OSCE Minsk Group format, co-chaired by representatives of Russia, U.S. and France, that governs the Nagorno-Karabakh peace process. Pashinyan and Aliyev clinched what one diplomat described as an unscripted accord on the sidelines of a Russia-led summit of post-Soviet states in Dushanbe. It listed technical steps to reduce tensions; Pashinyan kept the original piece of paper and Aliyev photographed it for the record on his mobile phone.[fn]Crisis Group interview, diplomat, Tbilisi, October 2018.Hide Footnote The communication channel they put in place is used to discuss incidents along the front lines. Although the respective defence ministries operate the line, communications through this channel go directly to the top security services personnel who brief the two leaders.

The Armenian village of Berkaber, seen from a hill nearby. The village, in the region of Tavush, is separated from Azerbaijan’s district of Gazakh by the Joghaz Water Reservoir. CRISISGROUP/Jorge Gutiérrez Lucena

Since then, prospects for resetting the peace process have dimmed, and the July skirmishes further narrowed space for discussion. After the worst of the clashes, Pashinyan urged a resumption of talks mediated by the OSCE Minsk Group co-chairs, repeating a call for more rigorous international monitoring of ceasefire violations.[fn]“PM Pashinyan: ‘An international system of reliable ceasefire monitoring needs to be established’”, official website of the Prime Minister of the Republic of Armenia, 19 July 2020.Hide Footnote In the past, Russia has offered to deploy its troops to the conflict zone to deter any further escalation, but both Yerevan and Baku turned the Kremlin down, wishing for a more international peace monitoring presence in the region.[fn]See Crisis Group Europe Report N°255, Digging out of Deadlock in Nagorno-Karabakh, 20 December 2019.Hide Footnote

What is clear is that neither side appears to have availed itself of the existing communication channel.

In any case, officials in Baku said now was not the time for talks.[fn]“There were many opportunities before the incident to work on security-related cooperation in the border areas. … Not now”. Crisis Group interview, senior Azerbaijani official, Baku, 21 July 2020.Hide Footnote The death of a well-respected Azerbaijani general in the July fighting stirred one of the largest street protests in recent years. Major General Polad Hashimov was the highest-ranking official killed on the battlefield since the 1994 ceasefire. In the wake of his death, tens of thousands of protesters took to the streets, demanding that Baku go to war to return Nagorno-Karabakh to Azerbaijani control, with a small number storming Azerbaijan’s parliament in anger.[fn]“Unauthorised procession of thousands of people through streets of Baku lasted all night”, Turan Agency, 15 July 2020.Hide Footnote Aliyev praised the demonstrations as “another picture showing the unity of the people and power”, though he denounced the incursion into parliament.[fn]“Azerbaijani opposition hit with new arrests following Aliyev speech”, OC Media, 17 July 2020; “Azerbaijan fires foreign minister”, Eurasianet, 16 July 2020.Hide Footnote A day later, he sacked his foreign minister, Elmar Mammadyarov, who had been the country’s key emissary in peace talks with Armenia for sixteen years, accusing him of leading “useless negotiations” to allow the World Health Organization access to Nagorno-Karabakh to help its residents deal with the pandemic.[fn]Crisis Group Europe Briefing N°89, The COVID-19 Challenge in Post-Soviet Breakaway Statelets, 6 May 2020.Hide Footnote Aliyev replaced Mammadyarov with the relatively low-profile education minister.

As noted above, the origins of the escalation remain murky. What is clear is that neither side appears to have availed itself of the existing communication channel.[fn]“PM: ‘Armenia’s armed forces keep the situation under full control: no provocative action goes unanswered’”, official website of the Prime Minister of the Republic of Armenia, 13 July 2020.Hide Footnote It was only used after calls by the OSCE and EU to do so, and then only in order to allow Azerbaijan to safely retrieve its dead.[fn]“An Armenian Defence Ministry’s video showing Azerbaijanis collecting corpses of their soldiers from the neutral zone”, Sputnik Armenia, 16 July 2020 (Russian).
Hide Footnote
This is nothing new. The two sides rarely use the channel to relay information on planned activities on the front – such as troop movements or reinforcement of trenches – and thereby to minimise risks of misunderstandings.

Nor was it the first time that the channel could have come in handy. For instance, it could have lowered the tensions that have regularly emerged since December 2018, after Baku replaced soldiers with border guards in Gazakh and a part of Aghstafa bordering Armenia’s Tavush region. Baku maintains that it intended the move as an olive branch – but it appears to have backfired.[fn]“Border guards replace army forces in Gazakh and Aghstafa regions”, Turan Agency, 14 December 2018. According to the OSCE Polis’s Country Profile: Azerbaijan, “the State Border Service is the body of the central executive power and reports directly to the President of the Republic of Azerbaijan”.Hide Footnote “We wanted the demilitarisation of the border areas. Therefore, we deployed the border guards”, a senior government official said.[fn]Crisis Group interview, senior Azerbaijani official, Baku, 21 July 2020.Hide Footnote The border guards began building new bases as they moved in. “The establishment of a border security system requires land clearing and considerable digging”, one border guard official explained.[fn]Crisis Group interviews, security officials, Gazakh district (Azerbaijan), June 2019.Hide Footnote Across the border, however, “no one could understand why they were digging”, said an Armenian commander.[fn]Crisis Group interview, military commander, Tavush region (Armenia), November 2018.Hide Footnote The unusual activity may partly account for the rise in tensions. Today, 60 per cent of all clashes between Azerbaijan and Armenia occur in the border regions.[fn]The estimate is based on official reports from Armenian, Azerbaijani and de facto Nagorno-Karabakh sources. See a detailed database at Crisis Group’s The Nagorno-Karabakh Conflict: A Visual Explainer.Hide Footnote Thirteen of the sixteen incidents since September 2018 have taken place in the area that the border guards control.[fn]Ibid.Hide Footnote

Top brass from the two countries fear that allowing communication between lower-ranking officers on the front lines could lead to explosive misunderstandings.

In the last decade, the Minsk Group co-chairs have on occasion floated ideas for strengthening the ceasefire regime along the international border, including by opening a direct line of communication between military commanders.[fn]For some ideas, see “Snipers must be withdrawn to avoid further incidents on line of contact, says OSCE Chairperson during visit to Armenia”, press release, OSCE, 18 March 2011; “Joint Statement by the Heads of Delegation of the Minsk Group Co-Chair Countries and the Foreign Ministers of Armenia and Azerbaijan”, OSCE, 6 December 2011.  Hide Footnote Commanders in trenches on both sides of the border in Nakhchivan have been communicating via hotline since the mid-1990s. But attempts to make similar arrangements along other sections of the border have gained little traction.

Today, neither side believes it is possible to revive contact when there has been so little for so long. Top brass from the two countries fear that allowing communication between lower-ranking officers on the front lines could lead to explosive misunderstandings. “Azerbaijanis and Armenians are emotional people, and given that both sides have mostly young soldiers and officers, today such a hotline is risky”, a high-level commander in Baku said.[fn]Crisis Group interviews, senior Azerbaijani defence ministry officials, Baku, November 2019.Hide Footnote Echoing these concerns, an Armenian field officer in the Tavush region said communication should take place at a senior level to protect the chain of command: “If we were to speak with those in the trenches, and their senior commanders were to give conflicting orders, who would they listen to?”[fn]Crisis Group interviews, military commanders, Tavush region, Yerevan, November 2018 and February 2019. Hide Footnote

The best way forward could be an agreement between Baku and Yerevan to make greater use of existing communication channels, including to offer updates on planned construction works or other activities along the front lines. Even before the 2018 Dushanbe agreement, some international humanitarian organisations had sought to introduce such preventive messaging. Every summer, the International Committee of the Red Cross seeks information from farmers on the planned harvests near the trenches and shares it in an official letter to both sides in hopes of encouraging restraint.[fn]Crisis Group interviews, diplomats, Yerevan, Baku, October 2018 and July 2020.Hide Footnote  Military commanders along the border, who are closest to the situation on the ground, should elaborate on these messages and give them to superiors, who would then relay them via security agencies to the other side. An expanded communication channel might include updates on mutually acceptable activities such as farming, waterway maintenance and demining (all discussed in detail in sections below). For Baku, expanded communication could allow the complete deployment of its border guard, which it halted amid heightened tensions.

IV. Making Farming Safe

With farmers now too fearful to venture out, the forthcoming harvest in the border regions may be left rotting for the birds and the worms. Even before the July 2020 clashes, the abundance of unmarked minefields and restrictions on access to land prevented border residents from fully exploiting opportunities for farming, which is often their only source of income. Locals cultivate fruits, vegetables and grains where they can, but much fertile land remains beyond reach. Others raise livestock in the foothills, running the risk of stumbling upon mines or being targeted by snipers.[fn]Crisis Group interviews, village residents, Tovuz and Gazakh districts (Azerbaijan), June 2019.Hide Footnote  Additional job opportunities are scarce. Some youths serve on the front line, but few remain with the army after completing their compulsory military service.[fn]Crisis Group interviews, military commanders, Tavush region (Armenia), November 2018.Hide Footnote  Nor does the military presence in the region bring much commerce to villages.[fn]Crisis Group interview, official, Ijevan (Armenia), September 2019.Hide Footnote

The Armenian and Azerbaijani leaders underscored the importance of making farming safe in the spring of 2019, when they tentatively pledged to avoid ceasefire violations during that fall’s harvest.[fn]“Joint Statement by the Foreign Ministers of Armenia and Azerbaijan and the Co-Chairs of the OSCE Minsk Group”, OSCE, Vienna, 29 March 2019. “Foreign Minister Sergey Lavrov’s remarks and answers to media questions at a joint news conference following talks with OSCE Secretary General Thomas Greminger”, Ministry of Foreign Affairs of the Russian Federation, 31 October 2019.Hide Footnote Two weeks after their first official meeting in March, the two sides formalised their agreement, but no details were made public.[fn]“Joint Statement by the Foreign Ministers of Azerbaijan, Armenia and Russia, and the Co-Chairs of the OSCE Minsk Group”, OSCE, 15 April 2019.Hide Footnote Although the agreement appears to have led to a decrease in incidents during the 2019 harvest, farmers were left ignorant of the high-level diplomacy. Officials on both sides say they refrained from commenting on the agreement for fear of being blamed for encouraging residents to cultivate lands close to the border in the event that tensions escalated.[fn]Crisis Group interviews, senior Armenian and Azerbaijani officials, Yerevan and Baku, November 2018 and September 2019.Hide Footnote

Preparation for the annual harvest in fields near Farahli village, in Gazakh region, Azerbaijan. CRISISGROUP/Zaur Shiriyev

Over the past decade, residents on both sides have lost access to acres of land they once farmed.[fn]Crisis Group interviews, residents, local officials, Nerkin Karmiraghbyur, Chinari, Movses, Berkaber, Paravakar, Vazashen, Koti and Ijevan (Armenia), December 2017, November 2018 and September 2019. Crisis Group interviews, local officials, Gazakh and Tovuz districts (Azerbaijan), June-July 2019.
Hide Footnote
“We were able to use the fields until 2002, but then the military put up barbed wire in the hills and mined more areas near the border”, an Azerbaijani villager in the Gadabay region said.[fn]Crisis Group interview, village residents, Goyalli, Gadabay district (Azerbaijan), July 2019.Hide Footnote Farmers are reluctant to expose tractors or combine harvesters to gunfire or shelling. Azerbaijani villagers in Asrik-Jirdakhan work their farms at night because almost all their land is within Armenian troops’ shooting range.[fn]Crisis Group interview, village residents, Asrik-Jirdakhan, Tovuz district (Azerbaijan), June 2019.Hide Footnote The fear is shared across the border. In the Armenian village of Nerkin Karmiraghbyur, residents have built greenhouses to compensate for the loss of farmland. “Many are still afraid of attacks on greenhouses because they are … very visible, but greenhouses help us make some money”, said one woman, who repaired her house for the first time in 30 years in 2019 when she first felt secure enough to undertake building work close to the trenches.[fn]Crisis Group interview, village resident, Nerkin Karmiraghbyur (Armenia), September 2019.Hide Footnote But in July, when at least six houses were damaged and gas supplies cut in her village, she left with her grandchildren.

In Azerbaijan, disputes over land reform as well as the special fiscal status granted to a few border villages complicate matters.

Before the July clashes, some farmers, out of desperation, had encroached on land closer to the trenches, though most remained wary.[fn]Crisis Group interviews, village residents, Tavush region (Armenia), September 2019. The residents of Chinari, Paravakar and Nerkin Karmiraghbyur said they would try to gain access to around 1,000 more hectares in 2020. Crisis Group interviews, village residents, local officials, Chinari, Paravakar and Nerkin Karmiraghbyur (Armenia), September 2019.Hide Footnote On the Armenian side, a modest uptick in enterprise was noticeable amid the relative calm since the 2018 Dushanbe agreement. Some farmers there told Crisis Group they planned to invest in new seeds and technology, which they had previously been reluctant to do due to fears of renewed tensions.[fn]Crisis Group interviews, village residents, Chinari, Berkaber and Nerkin Karmiraghbyur (Armenia), September 2019.Hide Footnote No such boldness was apparent on the Azerbaijani side, however, and on both sides, habits formed over years of worry and want are hard to break. Across the border, one Armenian harvester said she could not shake her anxiety from years “when we had to run, hide, lay on the ground … thinking that, at any moment, any of us can be shot” despite the relative calm of past years.[fn]Crisis Group interview, village resident, Chinari (Armenia), September 2019.Hide Footnote Her home village of Aygepar was rocked by fresh shooting in the recent violence.

In Azerbaijan, disputes over land reform as well as the special fiscal status granted to a few border villages complicate matters. The post-Soviet land reform enables locals to rent land from municipalities for agriculture. Farmers in villages where insecurity halted the reform’s rollout cannot own or plough land.[fn]Crisis Group interviews, village residents, Haciali, Vahidli, Agdam and Alibeyli, Tovuz district (Azerbaijan), June 2019.
Hide Footnote
As such, their livelihoods are more precarious, and they take greater financial risks, when renting land, for instance, and stand to lose more when fighting disrupts commerce. The “border village status”, meanwhile, confers tax breaks as well as gas and electricity subsidies.[fn]Crisis Group interviews, border village authorities, Azerbaijan, June-July 2019.Hide Footnote Where villages do not have it, especially in the Gazakh and Gadabay districts, residents are resentful. They are likelier to push for access to farmland in riskier border regions. “If there isn’t enough agricultural production in the village, then we cannot afford the electricity and gas prices”, said a 70-year-old villager in Qaralar. New Azerbaijani legislation could give some relief by allowing farmers access to subsidies via an electronic system.[fn]“Azerbaijan's e-agriculture system to provide farmers with necessary services”, Azernews, 6 February 2020.Hide Footnote  But farmers need to apply well in advance of sowing season, when they may not yet know which land is safe.

As one border resident said in September 2019, efforts to improve farming in the region are like “sand castles”, carefully built only to disappear the moment shots are fired.[fn]Crisis Group interview, village resident, Berkaber, September 2019.Hide Footnote  His words seemed prescient. Activity ground to a halt amid the July clashes, causing people on both sides to suffer. One farmer in the Azerbaijani village of Agdam, which was shelled in mid-July, said he was already having trouble finding people to work the fields: “After these clashes, the fear of those people increased”.[fn]Crisis Group telephone interview, village resident, Agdam, Tovuz district (Azerbaijan), 17 July 2020. “As a result of artillery fire by Armenia, house and yards damaged in Azerbaijan’s Tovuz”, APA, 14 July 2020.Hide Footnote Likewise, in the Armenian village of Movses, when shooting started near his house, Sargis Arakelyan took his wife and 92-year-old mother to safety at the other edge of the village.[fn]Crisis Group telephone interview, village resident, Movses, July 2020.
Hide Footnote
Speaking to Crisis Group by telephone, he said his more pressing concern was whether locals would be able to collect the forthcoming harvest: “Some people already try to go to the farmlands, but only a few. The tragedy is that this year the harvest should have been particularly rich”.[fn]Ibid.Hide Footnote

Yerevan and Baku should redouble efforts to allow farmers to safely harvest crops and herders to tend livestock.

Even in areas unaffected by the flare-up, people dare not venture out. “If there is shooting in Tovuz’s Agdam village, I naturally think that if I go to my fields around Gazakh’s Dash Salahli village, they will shoot me, too”.[fn]Crisis Group telephone interview, resident, Gazakh district (Azerbaijan), 17 July 2020.Hide Footnote

Yerevan and Baku should redouble efforts to allow farmers to safely harvest crops and herders to tend livestock – all the more so as people in the region feel the pandemic’s financial bite. The two sides have found ways in the past to put aside deeply divisive territorial disputes to cooperate on limited humanitarian measures. At a minimum, they should inform each other of any plans for works on land near the trenches – offering the opposing side a chance to raise security concerns. In the future, they might organise talks among local and military officials on both sides about the tracts of land that residents are eyeing. The evolving security environment, with the deployment of border guards on the Azerbaijani side, will require an especially nuanced approach to allow for effective communication not only with the Azerbaijani military but also with its border guard service. The imminent harvest makes the development of preventive messaging between the sides all the more important.

V. Repairing Water Supply Networks

Water was once abundant in the border zone, thanks to a network of reservoirs and irrigation pipes, but today shortages are chronic.[fn]Crisis Group interviews, residents and officials, diplomat, Koti, Berkaber, Ijevan and Yerevan (Armenia), November 2018 and September 2019.Hide Footnote  After the 1992-1994 war, it became too dangerous to maintain the water supply system, which criss-crosses the front lines, and it fell into disrepair.[fn]Tatul Hakobyan, “Armenian Border: Tavush”, op. cit. Crisis Group interviews, residents, regional officials, Koti, Chinari and Ijevan (Armenia), November 2018 and September 2019.Hide Footnote Villagers began blocking supply channels to satisfy their own needs.[fn]Crisis Group interviews, residents and regional officials, diplomat, Tavush region, Ijevan and Yerevan (Armenia), November 2018 and September 2019.Hide Footnote Today, a mere handful of households draw their water from reservoirs fed by mountain rivers. In some places the dams are so outdated they may themselves pose a threat to nearby residents.[fn]Crisis Group telephone interviews, diplomats, Baku and Yerevan, July 2020.Hide Footnote Many households have access only to unfiltered well water, which they describe as “salty”.[fn]Crisis Group interviews, village residents, Gazakh district (Azerbaijan), June-July 2019.Hide Footnote In some villages, running water is limited to two to three hours per day.[fn]Crisis Group interviews, village residents, Berkaber, Koti, Paravakar and Chinari (Armenia), November 2018 and September 2019.Hide Footnote “It’s only enough to prepare dinner”, said a woman from Berkaber, a village sitting next to a reservoir in disuse.[fn]Crisis Group interview, village resident, Berkaber (Armenia), November 2018.Hide Footnote

Officials on both sides are under pressure from villagers to fix water shortages, and demand will only grow if farming expands amid long-term climate trends.[fn]Crisis Group interviews, village heads, regional officials, Koti, Berkaber, Chinari, Paravakar and Ijevan (Armenia), November 2018 and September 2019.Hide Footnote Already fields are parched, with farmers praying for rainfall in ever-warming summers. “There wasn’t enough rain this year, and I lost all my investments”, one farmer said.[fn]Crisis Group interviews, village residents, Agbulag, Tovuz district (Azerbaijan), June 2019.Hide Footnote While locals are largely unaware of climate dynamics, experts say less rain and higher temperatures due to climate change are exacerbating water scarcity in the region.[fn]“Climate Risk Profile Armenia” and “Climate Risk Profile Azerbaijan”, USAID, June 2017.Hide Footnote Aggregate river flows in Armenia and Azerbaijan are projected to decrease by 11.9 per cent by 2030 and 37.8 per cent by 2100.[fn]Ibid.Hide Footnote


Armenia and Azerbaijan have both made substantial investments to break dependence on the other side but with limited success. They both lack expertise and capacity to implement adequate solutions.[fn]Crisis Group telephone interviews, diplomats, Baku and Yerevan, July 2020.Hide Footnote The Azerbaijani government built a reservoir in Tovuzchay with a capacity of 20 million cubic metres.[fn]“Ilham Aliyev inaugurated Tovuzchay water reservoir”, press release, official website of the President of Azerbaijan, 16 February 2016.Hide Footnote But most residents of the district where it is located still lack water.[fn]Crisis Group interviews, village residents, Haciali and Alibeyli (Azerbaijan), June 2019.
Hide Footnote
With the help of donors, Baku has also drilled hundreds of sub-artesian wells.[fn]Crisis Group interview, local official, Tovuz district (Azerbaijan), June 2019.
Hide Footnote
The wells may also soon run dry. Residents also blame the well water’s poor quality for health problems.[fn]Crisis Group interviews, village residents, doctor, Dash Salahli, Gazakh district (Azerbaijan), June 2019.Hide Footnote “When we go to doctors in Baku, and they find kidney stones, they ask: ‘Are you from the Gazakh district?’”, one woman said.[fn]Crisis Group interview, village resident, Dash Salahli, Gazakh district (Azerbaijan), June 2019.Hide Footnote For its part, the Armenian government is mulling an investment of around $30 million in new pumping stations.[fn]Crisis Group interview, regional governor, Ijevan (Armenia), September 2019.Hide Footnote It has, however, baulked at the cost of a new reservoir.[fn]Ibid. See also “Armenia decided to construct 12 new water reservoirs”, RIA Novosti, 6 February 2018 (Russian).Hide Footnote Instead, villages raise funds from the Armenian diaspora to lay pipes coming down from the mountains, but this costly supply is far from enough.[fn]Crisis Group interviews, village heads, regional officials, Koti, Berkaber, Chinari and Ijevan (Armenia), November 2018, September 2019.Hide Footnote Some village heads have rationed water and publicly shamed or cut off supply to residents who exceed their quotas.[fn]Crisis Group interview, village head, Chinari (Armenia), October 2018.Hide Footnote

While decades of tensions have prevented cross-border cooperation, some tentative steps might serve the two nations’ interests.

A more strategic approach to the water problem in the region would help, but ultimately neither side can resolve the water supply problems without the other. While decades of tensions have prevented cross-border cooperation, some tentative steps might serve the two nations’ interests. One might be the resumed use of the Joghaz reservoir. Built in the early 1970s, the Joghaz reservoir once supplied water to almost 30 Armenian and Azerbaijani villages.[fn]Crisis Group interviews, village residents, regional officials, diplomat, Berkaber, Koti, Ijevan and Yerevan (Armenia), November 2018 and September 2019. Also see “Berkaber: border village, reservoir, people”, CivilNet, 5 May 2020 (Armenian).Hide Footnote Now it services only a few nearby households.[fn]Crisis Group interviews, village head, residents, Berkaber, November 2018 and September 2019. Also see Olesya Vartanyan, “Armenia Elections Boost Hopes for Peace with Azerbaijan”, Crisis Group Commentary, 10 December 2018.Hide Footnote Trenches stretch along the shores, and soldiers face off mere metres from each other on the dam. Three derelict pumping stations need to be fixed in order to restore water supplies to adjacent Armenian and Azerbaijani villages.[fn]Crisis Group interviews, diplomat, regional officials, Ijevan and Yerevan (Armenia), December 2018 and September 2019.Hide Footnote Engineering works are impossible, however, without a clear, detailed accord and a commitment from both sides.[fn]Crisis Group interviews, military commanders, regional officials, Armenia, November 2018 and September 2019.Hide Footnote

Such circumscribed technical talks, avoiding discussion of Nagorno-Karabakh, could be possible with the blessing of the leadership in both countries. An independent international actor such as the Red Cross could help facilitate limited discussions between local representatives and security officials about repair of specific water infrastructure of mutual benefit.[fn]Crisis Group interview, diplomat, Yerevan, December 2018.Hide Footnote Such cooperation could then extend to other areas of the border where water distribution systems are in disrepair but remain inaccessible even for regular maintenance due to minefields or trenches nearby.[fn]Crisis Group interviews, village residents, Aghstafa, Tovuz and Gazakh districts (Azerbaijan), June-July 2019.Hide Footnote It would also be an important trust building measure that could bolster the peace process and reassure residents in the border areas. “If one starts bringing water from the other side, the message will be clear: ‘We have no intention of fighting with you anymore’”, said Ahmed, an 80-year-old resident of Agdam village in Tovuz district.[fn]Crisis Group interview, village resident, Agdam, Tovuz district (Azerbaijan), June 2019.Hide Footnote

80-year-old Ahmed, a local resident of Agdam village of Tovuz region, thinks the time is right to act for peace. CRISISGROUP/Zaur Shiriyev

VI. Clearing Landmines

Landmines and unexploded ordnance are perhaps the most pernicious and complicated of border residents’ concerns. In some villages, mines have turned more than 500-1,000 hectares into no-go zones. There is no reliable data on the number of mines or unexploded ordnance in border villages, but the region is among the most heavily contaminated in Azerbaijan and Armenia.[fn]Mine Action Profiles of Armenia and Azerbaijan, Landmine and Cluster Munition Monitor, 12 November 2018. After recent clashes, Azerbaijan’s state demining agency ANAMA collected artillery and mortar shells from in and around the villages of Agdam and Dondar Gushchu. “ANAMA took action: shells fired by Armenians”, Axar, 15 July 2020.Hide Footnote Demining has long been stalled by mutual concerns that it might shift the balance on the front lines and weaken deterrence. The painstaking and thorough nature of mine clearance work also rubs up against military concerns about secrecy.

A sign posted by Azerbaijan’s de-mining agency warns residents “Don't enter dangerous areas” in Aghstafa, Tatli village, Azerbaijan. CRISISGROUP/Zaur Shiriyev

In Azerbaijan, the state demining agency (ANAMA) clears mines, raises awareness and offers first aid training in partnership with the education ministry and Azerbaijan Red Crescent Society.[fn]See details about Mine Risk Education at the official ANAMA website. “Operative information of ANAMA’”, 15 July 2020.Hide Footnote There have been fewer accidents since 2016, with explosions triggered by cows or disoriented herders straying into no-man’s land.[fn]Beginning in 2016 and continuing to January 2020, ANAMA responded to 132 emergency calls on the hotline from locals in Gazakh, Tovuz, Gadabay and Aghstafa districts. See “Operations” on ANAMA’s official website.Hide Footnote “One of our herders stepped on a mine while his cattle were grazing”, a farmer said of one accident that led to a leg amputation. “By the time he realised [he was in a minefield], it was too late”.[fn]Crisis Group interviews, village residents, Asrik-Jirdakhan, Tovuz district (Azerbaijan), June 2019.Hide Footnote ANAMA says it cannot work within 2km of the border at present.[fn]Crisis Group interview, ANAMA representative, Baku, July 2020. See also the “ANAMA Principles” published on the agency’s official website.Hide Footnote “Our workers’ safety is also a priority”.[fn]Crisis Group interview, ANAMA representative, Baku, March 2019.Hide Footnote In Armenia, demining has never taken place in the border region of Tavush, leaving at least twenty contaminated areas.[fn]Crisis Group interview, demining specialist, Yerevan, October 2018. See also “Dangerous Areas of Tavush Region”, Armenian Center for Humanitarian Demining and Expertise working paper (Armenian), shared with Crisis Group in October 2018; and “FSD Non-technical Mine Action Survey”, Armenian Center for Humanitarian Demining and Expertise, May 2013.
Hide Footnote
The Armenian Center for Humanitarian Demining and Expertise says it cannot guarantee its employees’ safety, as they would need to enter areas close to the trenches. They fear that Azerbaijani soldiers will mistake the hulking demining trucks for military vehicles.[fn]Crisis Group interview, demining specialist, Yerevan, October 2018.Hide Footnote

Modest attempts to open discussion on the issue have been tied to diplomacy over Nagorno-Karabakh. An attempt in the early 2000s to put together a survey of where mines lie hidden saw months of detailed talks among local officials and military representatives, before disagreements between the countries’ leaders over Nagorno-Karabakh halted progress.[fn]“Landmine Monitor Report 2002: Toward a Mine-free World”, Human Rights Watch, 2002, pp. 609-610.Hide Footnote “Politics spoiled what should have been only in the hands of professionals”, said a veteran demining specialist in Armenia.[fn]Crisis Group interview, demining specialist, Yerevan, October 2018.Hide Footnote In 2017, Azerbaijan blocked the work of the OSCE office in Yerevan, accusing it of going beyond its mandate by engaging in demining activity in the Nagorno-Karabakh conflict zone.[fn]The issue was discussed at the OSCE Permanent Council in 2017. See the public statements by Azerbaijan and Armenia (made on 19 January 2017) and the U.S. (made on 4 May 2017).
Hide Footnote
Armenian officials deny it, saying a detailed investigation showed no evidence of work close to the front lines.[fn]Crisis Group interview, official, Yerevan, October 2018. Also see “Letter of Azerbaijan's permanent representatives in OSCE to the Chairperson of the OSCE Permanent Council”, Ministry of Foreign Affairs of the Republic of Azerbaijan, 10 February 2017.Hide Footnote

Even with political will at the top, demining experts say the task at hand is gargantuan,

Any new attempts to discuss demining would likely need to occur in tandem with or following improvements in preventive messaging, respect for the ceasefire during the harvest and repairs to essential water supply networks. Even with political will at the top, demining experts say the task at hand is gargantuan, necessitating special security guarantees and additional funds that would likely need to come from donors.[fn]Crisis Group interview, demining specialist, Yerevan, October 2018.Hide Footnote The focus, for now, should be on the humanitarian imperative of clearing the mines that pose the greatest danger. Discussions might begin by delineating areas in some border villages where, for example, landmines block access to the only source of water or particularly valuable farming areas. At the very least, the sides should aspire to clearly demarcate where landmines may be buried near farmlands and water sources to prevent accidents.[fn]Crisis Group interview, demining specialist, Yerevan, October 2018.Hide Footnote

VII. Conclusion

For three decades, what little diplomacy Armenia and Azerbaijan have engaged in has sought to reconcile their leaders’ intransigent views on Nagorno-Karabakh. Humanitarian projects, especially in the border regions, have fallen by the wayside. It is regrettable that these projects have languished: not only do they arguably offer the clearest opportunities for compromise and mutual benefit, but their focus is also squarely on improving the lives of populations most affected by the daily tensions on the front lines. Residents on both sides of the border share the same fears: “The pandemic has already affected our lives. Water shortages have affected us also, and these clashes have become another burden”.[fn]Crisis Group telephone interview, village resident, Kokhanebi, Tovuz district (Azerbaijan), 17 July 2020.Hide Footnote

Easing tensions could prevent deaths and injuries among farmers and seasonal workers, who are readying to collect the next harvest.

The July clashes should serve as a warning and call to action. As a first step, the two countries should recommit to using their existing communication channel. It should be expanded to include preventive messaging to avoid tensions and apparent misunderstandings like those that developed around Azerbaijan’s deployment of border guards to replace its armed forces. Easing tensions could prevent deaths and injuries among farmers and seasonal workers, who are even now readying to collect the next harvest from August to September. Over time, such messaging may also help dispel suspicion between military commanders along the front lines. Repairing water supply networks and demining borderlands will be lengthy and complex processes; that is even more reason to begin discussions on these issues before new pandemic-related economic pressures and climate change worsen conditions for farmers and herders in the region.

Baku and Yerevan have both invested heavily in attempts to mitigate problems related to insecurity and water shortages, with little success. Even minimal cooperation would cost them less. It may be the only way to stop the emptying of villages on both sides of the border, as people who love their homeland nonetheless find themselves forced to seek better, safer lives elsewhere.

Appendix A: Map of Armenia-Azerbaijan State Border