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 Youtube
A member of the Revolutionary Armed Forces of Colombia (FARC) sits in his tent in the “Alfonso Artiaga” Front 29 FARC encampment in a rural area of Policarpa, Narino, in southwestern Colombia on 16 January 2017. AFP/Luis Robayo
Report 60 / Latin America & Caribbean

In the Shadow of “No”: Peace after Colombia’s Plebiscite

Revised and ratified after its shock rejection in October 2016’s referendum, Colombia’s peace agreement still lacks sustainable political support. Reversing public distrust will need swift and effective implementation of the accord – including full apologies for past crimes and the visible handover of weapons by insurgents.

  • Share
  • Save
  • Print
  • Download PDF Full Report

Executive Summary

Defeat by a wafer-thin margin in the October 2016 plebiscite on the peace agreement between the government and the Revolutionary Armed Forces of Colombia (FARC) insurgency shocked Colombia’s society and political establishment, as well as the accord’s international backers. With the signed document suspended while rebel combatants tentatively gathered in sites across the country, prospects for an end to 52 years of armed conflict initially darkened. A revised accord, with numerous changes demanded by opposition leaders, was unveiled less than two months later, but the illusion of consensus was short-lived. Indignant that it was not able to review the new text and incensed that though many of its proposals were included, a few key ones were not, the opposition decried the agreement and its ratification in Congress. Peace with the guerrillas is again set to polarise parties and candidates in elections in 2018. A swift, effective start to implementation of the accord is needed to reverse public wariness and political resistance.

Victory in those elections for opponents of the peace agreement would be the harbinger of major challenges to the deal’s sustainability. Concentration of FARC combatants is underway, albeit problematically and with delays, and the six-month timetable for the handover of weapons has been set in motion. However, funding gaps, administrative delays and the political balance of power ahead of 2018 threaten to curtail transitional arrangements and structural reforms aimed at remedying the root grievances of the conflict. The opposition could financially starve institutions, programs or policies in the peace agreement if it comes to power. The terms of transitional justice, measures on rural reform and land access, and community-based approaches to removing coca crops and establishing alternative income-generating activities could all be in danger.

Defending the agreement will be an intrinsic part of the political battle ahead. Persuading a distrustful, urbanised public to give its backing depends in the immediate term on what happens in and around FARC cantonments. Over the next year, successful implementation will be the best way to bolster popular and political support and make it politically costly for opponents to reverse the peace process. Transparency in handover of weapons, full apologies for past crimes, continued progress on humanitarian actions such as de-mining, increased results in the search for victims of forced disappearance and eventual cooperation with the Special Jurisdiction for Peace would underline the insurgents’ commitment to peace and the dangers of reneging on the agreement.

Violence on the ground will also affect support for the agreement in the short term. FARC leaders and troops fear betrayal by the state, and some may seek to hedge their bets in face of the visible opposition from significant political forces. Promised peace talks with the country’s second insurgency, the National Liberation Army (ELN), have not begun, and various armed groups appear to be behind dozens of killings of social leaders that constitute a new wave of terror in remote rural communities. Only resolute commitment by the state to prevent battles for control of illicit economies and protect civilians and ex-combatants will give peace real local-level meaning.

The international community should continue its political support, using its delegates and special envoys to maintain dialogue with all sides and exerting discrete pressure when necessary on opposition leaders to preserve crucial parts of the agreement that could be in jeopardy. It should refrain from making calls for renewed aerial fumigation of coca crops and instead give the agreement on illicit drug substitution a real chance to have effect. It must also use its financial assistance to establish mechanisms for moving resources quickly on behalf of effective implementation on the ground, helping resolve and learn from problems as they arise.

Rebuilding Colombia's Trust in the Peace Process

In this video, our Senior Analyst for Colombia, Kyle Johnson, highlights the main findings of Crisis Group’s report “In the Shadow of “No”: Peace after Colombia’s Plebiscite”. Crisis Group


To build political support for sustainable implementation of the new peace agreement

To the government of Colombia:

  1. Strengthen dissemination of the peace agreement in both rural and urban areas, while increasing protection rapidly for social leaders under threat until the agreement on security guarantees can be implemented.
  2. Establish and fund new institutions and commissions tasked with key roles to implement the peace agreement quickly, while strengthening nascent and galvanising existing bodies to generate early peace dividends for victims and conflict-zone communities.

To the government of Colombia and the FARC:

  1. Continue with the established weapons handover schedule despite delays in FARC arrival at cantonment sites, while adopting a proactive communication strategy, including documenting evidence of the FARC laying down weapons and engaging in reinsertion and of progress on other aspects of implementation.
  2. Keep victims at the centre of the process as implementation begins, increasing cooperation in the search for victims of forced disappearance, releasing all children age fifteen and under in FARC ranks and continuing public apologies for notorious crimes committed in the war.
  3. Explore space for dialogue with the opposition on implementation.
  4. Prioritise improving security for local leaders in the short term with preventive measures, training and strengthened security schemes, while also setting up institutions for protection of FARC members.
  5. Include local and regional authorities more directly during the arms abandonment process and planning of other implementation aspects.

To the opposition:

  1. Reinitiate dialogue with the government on implementation and increase the frequency, strength and level of its denunciations of violence against social leaders.

To the international community:

  1. Continue peace process support by maintaining delegates and special envoys during implementation, supporting citizen security and sustaining funding for international actors with important post-conflict roles, such as the UN High Commissioners for Human Rights and Refugees among other key players.
  2. Press the government and FARC to keep their commitments on time and to involve local actors more extensively.
  3. Continue dialogue with the opposition so as to press for support especially of at-risk parts of the peace agreement, such as rural development, political participation, transitional justice and humanitarian measures.
  4. Make more frequent public statements showing concern for the killing of social leaders and demanding progress in protection and justice.
  5. Support new agreements for major alternative development investments to tackle illicit drug production before pressing for more direct eradication.

To the UN mission:

  1. Finish deployment as quickly as possible, including of the civilian component, to prepare for and receive FARC fighters as they gather in cantonments for weapons handover.
  2. Adopt a proactive communications strategy, publishing frequent updates on FARC concentration and weapons handover, using media beyond regular official reports.
  3. Press the government and FARC to follow the schedule for weapons handover in the peace agreement, despite early and likely future delays.

Bogotá/Brussels, 31 January 2017

I. Introduction

When Colombians voted in October 2016 on the peace agreement between the government and the Revolutionary Armed Forces of Colombia (FARC), the “no” vote edged the “yes” by less than half of one per cent, with a 37 per cent turnout.[fn]For previous Crisis Group work on the peace process, see Latin America Reports N°s 58, Colombia’s Final Steps to the End of War, 7 September 2016; 53, The Day after Tomorrow: Colombia’s FARC and the End of the Conflict, 11 December 2014; 51, Left in the Cold? The ELN and Colombia’s Peace Talks, 26 February 2014; 49, Transitional Justice and Colombia’s Peace Talks, 29 August 2013; 45, Colombia: Peace at Last?, 25 September 2012; and Briefing N°32, On Thinner Ice: The Final Phase of Colombia’s Peace Talks, 3 July 2015.Hide Footnote  An intense process of high-level political dialogue ensued, leading to a new agreement that the government, FARC and many in civil society defend. Voicing dismay at the government, which it accuses of undermining democracy, the opposition has also united, but with the aim of rejecting the new agreement.

Congress has ratified the accord, and the start of the calendar for the insurgency’s weapons handover was set for 1 December, initiating the countdown for the 15,000 FARC combatants and militia members to gather in 26 cantonments across the country. The opposition, despite the Constitutional Court having allowed the congressional ratification procedure, has argued that by relying on the previously established pro-government majorities in both houses of the legislature, President Juan Manuel Santos cheated the people. Attempts to persuade it to support the new agreement have failed.

The context in which peace is to be implemented is far from hospitable. The government will struggle, even with international aid, to fund all the activities envisaged. New institutions the accord requires – some already created – are skeletal, sorely understaffed and unable to undertake the programs they are designed for, such as the Agency for Territorial Renovation; other official bodies, including the attorney general’s office, have proposed policies contrary to those in the agreement.[fn]Crisis Group telephone interview, high-level diplomat, Bogotá, 28 November 2016; interview, government official, 9 December 2016.Hide Footnote Violence against local social leaders has increased, raising doubts about peace benefits and leading to further polarisation between supporters and opponents. The peace process with the National Liberation Army (ELN), Colombia’s second largest insurgency, is yet to begin.

Political support for the peace agreement is weak and will most likely flag as the presidential election campaign begins later this year that will bring a new leader to office in 2018. Full implementation is thus far from guaranteed. Even so, the government and FARC have a window of opportunity to build support via implementation over the next eighteen months that would raise the political cost of not continuing the process from mid-2018 onwards.

The research for this approach included extensive interviews with members of the opposition, FARC and government negotiating teams, members of the Tripartite Mechanism to monitor and verify the ceasefire, pro-peace agreement leaders and politicians, political and legal experts and members of the international community close to the peace talks.

II. Getting to a New Agreement

The journey from plebiscite to new peace agreement hinged on various decisive moments. The starting point was the document’s narrow defeat on 2 October, leading to a complex shift in the balance of political power. Neither government nor opposition could claim a clear mandate. Tensions worsened as renegotiation began on a new text, ending with the crafting of an accord that lacked the stable, sustainable political base that opposition support would have added to that of pro-peace political parties, many victims’ organisations and civil society.

A. What Explains the Plebiscite Result?

The surprise result stemmed from the diversity and levels of commitment of voter bases in the opposition and pro-accord movements. A combination of ex-President Álvaro Uribe’s devoted support, anti-“gender ideology” churchgoers and the most right-wing elements of the divided Conservative party made up the majority of those who rejected the agreement.[fn]“Gender ideology” is the phrase used by groups who claim the accord looks to convert children into homosexuals, attacks the traditional family and seeks to take away parents’ right to educate their children in traditional ways. They say the ideology is in those parts of the agreement that promote special treatment for lesbian, gay, bisexual, transgender and intersex (LGBTI) victims of the conflict. Crisis Group interview, pastor opposed to peace accord, Bogotá, 3 November 2016.Hide Footnote Though he remains extremely divisive, Uribe’s support is stable and high. His followers tend to be more active in promoting their views and encouraging others to vote than those who favour the accords. Uribe backers are in general also more inclined to vote than other groups. The plebiscite was ideal for mobilising his committed base, as it loathes the FARC, rejects its participation in politics and fears Colombia turning into chavista Venezuela. The “religious vote”, primarily concerned about an alleged “gender ideology” smuggled into the agreement, is assumed to have been higher than in previous elections and fundamental to the result. Finally, an unknown part of the “no” vote derived from disinformation targeted at voters according to their region and income level.[fn]‘La estrategia del Sí tuvo muchos desaciertos’: Francisco Gutiérrez”, Semana, 8 October 2016. Jennifer Cyr and Carlos Meléndez, “Colombia’s right-wing populist movement defeated the peace deal. Here’s how we know”, The Washington Post, 4 October 2016. The director of the “no” campaign said different messages were used for distinct population sectors to encourage anger-based voting. These included claims the FARC would receive impunity; images of Santos and FARC leader Timochenko together; that subsidies and pensions for the poor and elderly would be cut; and Colombia would turn into Venezuela. Juliana Ramírez, “El No ha sido la campaña más barata y más efectiva de la historia”, La República, 5 October 2016.Hide Footnote

The high-level battle also favoured the opposition. Ex-President César Gaviria, leading the “yes” campaign, was unable to counter opposition arguments effectively, and no strong, unifying figure similar to Uribe emerged. Vice President Germán Vargas Lleras, a candidate for the role given his power and direct experience of guerrilla violence, was almost completely silent and, if anything, gave indirect support to the opposition.[fn]Vargas Lleras said he supported the agreement, with misgivings on certain issues. The Uribe argument of supporting peace but with changes was conceptually similar, and Vargas Lleras’s lack of campaigning, plus some ideological similarities with Uribe, may have led part of his base to vote “no”. He is due to step down in March 2017 to begin campaigning for the presidency.Hide Footnote “Yes” campaign strategies were also questionable. President Santos’s and other establishment-based parties focused on regional and local politicians, depending mainly on political machines, powerful families and coalition-building, which proved less effective in a single-issue plebiscite than in regular elections and were not even fully activated. Civil society, while vocal, again showed its historical weakness at mobilising votes. Finally, some pro-accord voters may have been complacent due to polls pointing to a big victory.

Areas of higher poverty tended to vote for the accord, except in Bogotá where lower-income groups were strongly “no”. The periphery – defined by measures of typical rural attributes or state capacity – also tended to back the agreement. It has been argued that areas with higher victimisation levels in the armed conflict tended to vote “yes”, but that argument hinges on how victimisation is measured.[fn]Leopoldo Fergusson and Carlos Molina, “Un vistazo a los resultados del plebiscito”, La Silla Vacía, 4 October 2016. If victimisation is measured solely by displacement, there is a clear correlation with “yes” votes, but not when it is measured more generally.Hide Footnote

The opposition’s victory meant its proposals for a new agreement had to be taken into account if the process was to be saved. Early on, some opposition leaders argued that any new accord required their approval. But the close result also allowed the politically-weakened government to divide the opposition (or isolate Uribe) by absorbing some concerns, while maintaining the risky option of a new plebiscite, a tactic that deprived foes of an absolute veto. Moreover, at various stages, the opposition showed itself to be divided. It was not until a new agreement was reached and ratified in Congress, despite the omission of certain of its key concerns, that the opposition could unite in indignation.

B. Positions for a New Agreement

Shortly after the plebiscite, the political and social actors who actively opposed the original agreement handed in their proposed changes. The first document with the entirety of their proposals presented in Havana to the FARC contained more than 260 items from at least ten sources.[fn]Santos no recibe más propuestas sobre el acuerdo de paz”, El Espectador, 20 October 2016. “Propuestas de Gobierno”, Government of Colombia (GOC), 23 October 2016.Hide Footnote Part, especially those touching on rural reform, reflected political interests rather than the concerns of many “no” voters. Despite the diverse, in some cases contradictory universe of proposals, there was a handful the opposition considered indispensable.

The most politically important were unmistakeable. There was a consensus that punishments in the Special Jurisdiction for Peace (SPJ) – the system devised to mete out transitional justice for serious crimes committed during the conflict – must be harsher, especially (in some cases exclusively) for the FARC. The SPJ, the opposition argued, should become part of the normal judicial system. There was near agreement that the Armed Forces must receive preferential treatment, though what that entailed differed between factions. The opposition shifted from demanding permanent prohibition from political office for those convicted of crimes against humanity and war crimes to a ban until sentences were completed.

All opposition groups agreed that there should be no amnesty for drug trafficking, the FARC should hand over its assets to be used as reparations to victims, and FARC use of such resources for political activity should be explicitly prohibited. The opposition was also united in demanding the agreement not have constitutional rank.[fn]The constitutional rank, or bloc, is the series of norms not in the constitution but used as parameters for constitutional control of law, such as treaties. “Sentencia C-067/03”, Constitutional Court, 2003.Hide Footnote Private property, it insisted, must be explicitly respected. Finally, concerns were expressed over “gender ideology” and its alleged effects on what was argued to be traditionally defined family and society.[fn]Crisis Group interviews, opposition representative and negotiators, Bogotá, 19, 20 October, 8, 11 November; Christian pastor, Bogotá, 3 November; senior diplomat, 11 November, all 2016. “Propuestas de Gobierno”, op. cit.Hide Footnote

The government responded by identifying those it considered easy to resolve, those that were difficult but not impossible and those that were held to be completely unviable. The opposition then argued that ruling out anything meant Santos was opening the door to “cheating” “no” voters by discarding key proposals that were inconvenient. Nonetheless, Santos sent his negotiators to Cuba with orders to take tougher stances on certain points, leading to friction with the FARC on the political participation issue, for example.[fn]Santos dice que algunas propuestas para acuerdo de paz son inviables”, El Tiempo, 20 October 2016. “Santos quiere hacer conejo con el acuerdo: Alejandro Ordóñez”, El Espectador, 3 November 2016. Crisis Group interview, senior diplomat, Bogotá, 4 November 2016; opposition negotiator, Bogotá, 8 November 2016; FARC negotiator, Havana, 28 October 2016.Hide Footnote

Once renegotiations began in Cuba, attention turned to the FARC’s reaction. Soon after the plebiscite, the group announced it was committed to peace and, in a 7 October communiqué, to adjusting the agreement so that it could earn broad-based political backing.[fn]Comunicado conjunto: Acuerdo Final, plebiscito y cese al fuego”, GOC and FARC-EP, 7 October 2016.Hide Footnote But it also said it could not yield on eligibility for office, which it considered the essence of the negotiations: converting an armed insurgency into a peaceful political force. It also continued to reject prison sentences, insisting that any harsher punishment must be applied to all actors in the conflict, and opposed including the SPJ within the ordinary justice system. The FARC (and government) disagreed with the proposal to give landowners a permanent assumption of good faith in all land purchases, thus allowing them to avoid prosecution if the land had been stolen without their direct participation. Lastly, it pushed hard for the accord to have constitutional force, but eventually gave way.[fn]Crisis Group interview, FARC negotiator, Havana, 28 October 2016.Hide Footnote

C. Three Renegotiations

A new agreement depended on three negotiation processes. The first was between the opposition and government. In public, both spoke of productive talks; in reality, there was a mutual lack of confidence. The government believed the opposition wanted to drag talks on into the 2018 presidential election, while the opposition was unsure the government would genuinely represent its positions in Havana. After an early back-and-forth, the opposition handed in a document with all its original proposals, some of which were watered down to show flexibility.[fn]Crisis Group interview, opposition negotiator, Bogotá, 8 November 2016.Hide Footnote

The second track involved civil society supporters of the peace agreement and the government, as well as on occasion the FARC. Not so much a negotiation as a defensive move by the pro-agreement camp, organisations, movements and leaders met with Santos to urge him to find a new accord quickly and to retain the original principles. A series of marches across the country and creation of a Peace Camp in Bogotá’s central Bolívar Plaza kept pressure on all sides.[fn]Crisis Group interview, organiser of the Peace Camp in Bolívar Plaza, Bogotá, 12 October 2016.Hide Footnote Some movements travelled to Havana to urge the FARC to persevere in its search for peace.

The third and final negotiation was between the government and the FARC. The government negotiators returned to Havana on 21 October and began a first round of talks, each one lasting roughly twelve hours. After these, the team returned to Bogotá to update the opposition. On 29 October, a new round began with the FARC, eventually leading to the announcement of a revised accord on 12 November. The latter talks had their difficult moments, particularly over FARC’s future political participation.[fn]Crisis Group interviews, high-level diplomat, Bogotá, 4 November 2016; FARC negotiator, Havana, 28 October 2016.Hide Footnote

During the negotiations with the FARC, a stable line of communication was established to keep opposition leaders up to date. However, concern that the government was not properly representing opposition positions was never fully dissipated. The “no” leaders expected further discussion on the new agreement before it was signed, but this never happened. With a few key concerns not addressed and amid politically motivated allegations that the revisions were little more than cosmetic, the opposition finally united against the document.[fn]Crisis Group interviews, opposition negotiators, Bogotá, 8, 11 and 30 November 2016.Hide Footnote

Throughout the renegotiation, the government and FARC felt under great time pressure due to concern the bilateral ceasefire could fall apart, despite an early consensus between them and the opposition that it was necessary. A new, detailed protocol included the pre-grouping of FARC fighters. With the government paying for FARC sustenance after 30 days, maintenance of the cessation of hostilities was possible, but the ceasefire was designed to last only three months.[fn]Crisis Group interview, member, UN mission, Bogotá, 9 November 2016. “protocolo para el cese al fuego y de hostilidades bilateral y definitivo (CFHBD) entre el gobierno nacional y las FARC-EP”, GOC, FARC-EP and UN mission, 13 October 2016.Hide Footnote On 13 November, the army killed two FARC fighters carrying out extortion activities on pretext of being ELN in Santa Rosa del Sur, a southern Bolívar province municipality.[fn]“GOC and FARC-EP violated ceasefire in south Bolivar incident”, Tripartite Mechanism communiqué, 30 November 2016.Hide Footnote While this event highlighted the ceasefire’s fragility, it also showed the robustness of the Tripartite Mechanism.[fn]The Tripartite Mechanism, which includes the government, FARC and UN mission, is to monitor and verify the ceasefire and weapons handover process. The UN mission is charged with investigating possible violations, helping agree protocols when necessary, assuring the parties fulfil their roles and providing recommendations after violations, among other tasks.Hide Footnote Both parties immediately turned to it to investigate, and it found violations by both sides. Government and FARC language then softened, and no further violent actions took place between them.

D. The New Agreement

The new agreement announced on 12 November included numerous changes based on opposition proposals. Some 58 per cent of the opposition’s original public proposals were included completely or partially. Some 58 proposals posited complete changes in form and underlying justification, of which 21 were included completely and six partially.[fn]See “Radiografía del plebiscito y el posplebiscito”, and “Radiografía del nuevo acuerdo: ¿Qué tanto se renegoció?”, both Fundación Ideas para la Paz, n.d.Hide Footnote Contrary to opposition charges, the revisions were beyond cosmetic.

Most of the opposition’s proposals and wording on Comprehensive Rural Reform were included and/or addressed, such as the rural tax system and legal protection for those who bought land in good faith. The same can be said for proposals and wording on political participation, for example on the role of political parties in designing a new statute for the political opposition.[fn]Many of victims’ leader Herbin Hoyos’s original proposals on victims’ participation in politics did not make it into the agreement, as they were either already implicitly there or proposed automatic seats in Congress for victims and a party with the same rights as the FARC, among others. The statute for political opposition is a legal measure to outline the special “guarantees for political parties and movements that declare themselves to be in opposition”. “Acuerdo final para la terminación del conflicto y la construcción de una paz estable y duradera”, GOC and FARC-EP, 24 November 2016, p. 37.Hide Footnote Some proposals on procedures for ending the conflict and on guarantees for ex-combatants’ security were accepted, though the conditions under which ex-combatants could hold office, including automatic allocation of congressional seats, were not altered. Several proposals on illegal drug cultivation were also incorporated, including a FARC legal commitment to hand over all relevant information about the drug trade, and the state’s right to aerially fumigate coca crops, despite suspension of this method in 2015.

Regarding victims and justice, the FARC is to hand over its whole war economy to provide victims reparations. The SPJ system is to be connected to the penal code and judicial system and not have foreign judges nor give NGO reports the same weight in evidence as information from the state authorities. SPJ courts may rule that FARC drug-trafficking can be interpreted as having been for personal gain, not merely to fund armed political activity. Perhaps most importantly in light of “no” campaign rhetoric, the new accord defines the restriction of liberty of convicted guerrilla combatants as obliging them to reside throughout their sentences within a village, under UN surveillance, while doing reparations-oriented work.[fn]Ibid.Hide Footnote

Numerous implementation proposals were added, including clauses on the accord’s financial stability; definition of “gender focus” as the “recognition of the equal rights between men and women, the special circumstances of each …”; and explicit respect for religious liberty.[fn]Ibid, p. 193.Hide Footnote Nor will the agreement have constitutional rank.[fn]Sources for this information include a comparison of the new accord, GOC and FARC-EP, op. cit., and the first document used in Havana to discuss opposition proposals, “Propuestas de Gobierno”, op. cit.; and “Sistematización opciones y propuestas voceros del no y lo acordado en el nuevo acuerdo”, Oficina del Alto Comisionado para la Paz, 22 November 2016. For just changes, see “Documento de trabajo: cambios, precisiones y ajustes”, Office of the High Commissioner for Peace, 12 November 2016. Pablo Abitbol, “Comparación anterior y nuevo Acuerdo Final Gobierno de Colombia – FARC-EP”, n.d.Hide Footnote The government and FARC argued they made changes to 56 of 57 topics discussed with the opposition; FARC political participation was the exception.[fn]“‘Logramos precisiones y cambios en 56 de los 57 temas abordados en nuevo acuerdo’: Santos”, El Espectador, 12 November 2016.Hide Footnote

III. Peace Toward 2018

All opposition leaders and the government negotiating team met the evening of 21 November in Bogotá.[fn]On the morning of 21 November, the presidential candidates of the Democratic Centre Party met with government negotiators to cordially discuss agreement on implementation. But poor communication from the opposition and differences of opinion on the government side, as well as procedural disagreements, set the tone for the evening meeting. Crisis Group interviews, opposition negotiator, Bogotá, 29 November 2016; political expert, Bogotá, 2 December 2016.Hide Footnote It started poorly, and ended worse. Perceptions differed on whether changes had been made to key parts of the accord; whether the new agreement was to be discussed with “no” leaders before signing; and over the way forward. They settled for complete disagreement: the opposition publicly rejected the accord, backtracking on some of its offers of greater flexibility. The polarisation created by the plebiscite, after being briefly camouflaged during the renegotiation process, resurfaced intact during the new ratification process.[fn]No es No”, La Silla Vacía, 22 November 2016; “Comunicado de representantes del No y de las víctimas”, 21 November 2016; Crisis Group interview, opposition negotiator, Bogotá, 30 November 2016.Hide Footnote

A. The Politics of Congressional Ratification

On 29 November, the Senate approved the peace agreement, 75-0; 25 from the opposition took part in debate but abstained, arguing Congress had no legal mandate to approve the accord. The same occurred in the House of Representatives the next day, where the vote was 130-0 (out of 166 taking part). The votes were controversial for reasons that will continue to impair support for the agreement.

Using Congress gave the government and FARC a clear route to ratification, while putting the opposition at a patent disadvantage. The pro-government coalition has a clear majority in both houses, especially on issues relating to the peace process. With Congressional elections not due until 2018, there is no immediate way for the “no” movement to translate its support base into legislative power.

This has led the opposition to argue that the government is undemocratically “imposing” the same peace deal, but the assertion that the congressional ratification is “undemocratic” depends on two claims. The first is that the new accord has only cosmetic changes, which fails to recognise the opposition’s success in getting key proposals into the text. The second contests the government view as to what can be defined legally as a “popular referendum”. The opposition argues that a special congressional vote is not a valid “popular referendum”; the pro-agreement side, including Santos, insists it is. Forced to adjudicate, the Constitutional Court ruled in December that Congress could itself decide on the ratification process.[fn]The Legislative Act for Peace’s fifth article required any peace agreement to go through a “popular referendum”, which at the time included the options of local committees, Congress or a new plebiscite. When the act was passed, the “popular referendum” language referred to the plebiscite. The Constitutional Court’s ruling on the Legislative Act avoided answering if Congress was a valid option, letting that body decide. The response in effect was “yes” as Congress activated the fast-track in December. “Comunicado No. 52”, Constitutional Court, 13 December 2016, p. 2.Hide Footnote The result of these differences is that the opposition has begun to use more extreme language, ratcheting up political polarisation by questioning not just the terms of peace but also the government’s respect for basic democratic tenets.[fn]See the speeches made by the Democratic Centre Party (DCP) Senators Iván Duque, Carlos Holmes Trujillo and Óscar Iván Zuluaga during the referendum debate in the Senate on 29 November 2016.Hide Footnote

The Constitutional Court also allowed Congress to activate the fast-track system laid out in the Legislative Act for Peace for approving the more than 50 laws needed to implement the peace agreement while avoiding the standard four or eight congressional readings of each bill. Considered essential – the FARC even said it would otherwise return to war – fast-track enabled Congress to approve key legislation, including the amnesty law passed at the end of December.[fn]“‘Sin ‘fast track’ volveríamos al monte’”, Semana Video, n.d.Hide Footnote

Avoiding a second plebiscite and securing fast-track procedures for peace agreement legislation have been essential to rapid recovery of the peace process. However, the way in which the government has acted makes rejection of the peace accord and its implementation – partially or wholly – a profitable political platform for 2018, as the opposition will continue to argue there has been no new peace agreement, and that the 2 October plebiscite was thwarted. Implementation in the medium- and long-term thus is at serious risk.

B. To 2018 and Beyond

As in the 2014 elections, peace will be at the heart of the national vote in 2018. The peace agreement will be central in the opposition’s congressional and presidential platforms.[fn]Crisis Group interviews, opposition negotiators, Bogotá, 11 and 30 November 2016.Hide Footnote With Uribe and other leading opposition figures set against the peace agreement and portraying themselves as the saviours of democracy, and with Vice President Vargas Lleras another contender, the likelihood of an anti-agreement candidate winning the presidency is high.[fn]Colombian presidential politics is a mix of traditional patronage networks and political identities, powerful families and opinion-based voting. Uribe will be able to mobilise perhaps around four million supporters for his preferred candidate. Vargas Lleras has the highest favourability of any politician (61 per cent), followed by Uribe (57 per cent), though recent scandals in his Radical Change party, may weaken him. Nonetheless, his patronage networks are unmatched. Marta Lucía Ramírez, Conservative party, had a good first round in 2014, and has a favourable rating of 41 per cent, but her party is regionally weak. On the pro-agreement side, the likely Liberal party candidate, Humberto de la Calle, has a 54 per cent rating. His party, though, can no longer count on votes from its 2014 alliance with Cambio Radical and is still weakened by old divisions. The Greens and Democratic Pole are relatively weak. “Gallup Colombia Poll #116”, December 2016. “Elecciones Presidenciales: Resultados”, Registraduría Nacional del Estado Civil, s.f.Hide Footnote Full implementation of the accord would then be in jeopardy.

Coalition-building will be crucial for the next president and Congress.[fn]Alliances are a constant in presidential elections, as parties make agreements after the first round to support one of two remaining candidates. In 2018, many coalitions will likely already exist due to shared positions on the peace agreement. After the plebiscite, the power of various actors within those coalitions is hard to judge and not static.Hide Footnote The numerous possible presidential candidates within the “no” movement, including Marta Lucía Ramírez, Óscar Iván Zuluaga, Iván Duque, Carlos Holmes Trujillo, and Alejandro Ordóñez, as well as Vargas Lleras, will make competition within and between parties fiercer than normal. In general, the Democratic Centre Party (DCP) starts with an edge, as its vote threshold is high, and Uribe, though barred from a new term, enjoys a certain cult of personality.[fn]“‘La estrategia del Sí tuvo muchos desaciertos’: Francisco Gutiérrez”, Semana, 8 October 2016; Crisis Group interview, political expert, 10 November 2016.Hide Footnote It is difficult to imagine a realistic scenario in which the DCP candidate does not make it to the second round of voting. As other opposition contenders look to increase their vote share, they have tended to portray themselves more radically as saviours of democracy and security.[fn]Alejandro Ordóñez has charged that because of how the peace accord was handled, Santos is consolidating a “dictatorship”. “En Colombia estamos ‘desde hace rato en una dictadura’: Alejandro Ordóñez”, Oiga Noticias, 26 October 2016. Marta Lucía Ramírez has called for a Constitutional Assembly to “redefine the functioning of established state organs, such as the presidency, Congress and the Courts”. “Colombia se está adentrando en una crisis de legitimidad institucional”, Ramírez, 22 December 2016.Hide Footnote

Vargas Lleras will be something of a wildcard. Though he keeps a low profile on the peace issue, the influence of his Radical Change party and the extent of his political patronage networks mean he commands many votes. He regards as his main opponent Humberto de la Calle, the most likely Liberal Party candidate, who, as the government’s chief negotiator with the FARC, is a staunch defender of the peace agreement. However, Vargas Lleras and Uribe have a poor relationship, despite certain gestures from the latter that could be interpreted as an invitation to an alliance.[fn]Tatiana Duque, “La estrategia disidente de Vargas”, La Silla Vacía, 28 November 2016; Crisis Group interview, high-level diplomat, Bogotá, 4 November 2016.Hide Footnote A possible outcome is a second round in the presidential election pitting the DCP against Vargas Lleras, with neither candidate strongly for implementing the whole peace agreement.[fn]Crisis Group interview, pro-accord senator, Bogotá, 30 November 2016. Vargas Lleras never had a strong position on the accord; recent information suggests he may oppose. Tatiana Duque, “Así se prepara Vargas Lleras para cuando le llegue su hora”, La Silla Vacía, 15 January 2017.Hide Footnote If he does not reach the second round, Vargas Lleras’s support would likely be decisive for the winner. He currently appears inclined more toward the DCP than pro-peace agreement parties.

Evangelical Christian churches will also be another major player in the run-up to the election. They are believed to have provided between one and two million votes to the “no” camp, and various religious leaders expect to play a central part in 2018.[fn]Natalio Cosoy, “El rol de las iglesias cristianas evangélicas en la victoria del “No” en el plebiscito de Colombia”, BBC Mundo, 5 October 2016; “El voto evangélico, clave en la victoria del ‘no’ en el plebiscito de Colombia”, El País de España, 13 October 2016. There is no way of knowing how many votes the churches provided, though they are widely credited with contributing two million to the “no” campaign. Some church actors did favour “yes”, but they are perceived as a minority.Hide Footnote The Christian “no” vote, however, is not homogeneous. While references to lesbian, gay, bisexual, transgender and intersex (LGBTI) issues in the peace agreement were ripped out of context and used to anger most Evangelical voters, there is less consensus among these Christian communities on other issues, such as justice, the truth commission and land ownership. Some pastors and followers but not all seem interested in political influence.[fn]Cristianos: ¿el poder decisorio en la política?”, Semana, 29 October 2016; Crisis Group interview, Christian pastor involved in post-plebiscite negotiations, 3 November 2016.Hide Footnote

Crafting coalitions will also be a priority for parties looking to defend the peace agreement. Numerous parties favour peace but by themselves do not provide enough votes to secure a second-round candidacy, as their support is scattered. Creating a pro-agreement alliance would be a step toward assuring that a candidate willing to implement the accord reaches the second round. But such a coalition would feature a wide array of groups that disagree strongly on other issues.[fn]Crisis Group interview, pro-agreement senator, Bogotá, 30 November 2016. For example, Jorge Robledo, Democratic Pole party senator and its possible presidential candidate, favours the accord but strongly opposes the government’s new tax bill, which parties that also support the deal back. “La peor reforma tributaria imaginable: Robledo”, Jorge Robledo, official website, 20 October 2016.Hide Footnote

The pro-agreement coalition would feature the Green Party, a force whose ability to mobilise votes is likely larger than its current representation in Congress; the Liberal and U parties, in which Santos has his roots and currently belongs respectively, and which are unlikely to make it to the second round, having done so in 2014 only because of their alliance with the Radical Change party in the first round; and what remains of the divided, left-leaning Democratic Pole party. Pro-agreement candidates will also have to find the right balance between supporting the accord and distancing themselves from Santos. The tax reform passed toward the end of 2016, which hiked value-added tax by three percentage points, the troubled economic conditions due to declining oil revenues and the president’s unpopularity make it essential that candidates who favour the peace process differentiate themselves from him and his government on other issues.[fn]President Santos’s 60 per cent disapproval rating, “Gallup Colombia Poll #116”, December 2016, is likely to worsen after a very unpopular tax reform in December. In 30 November Bolívar Plaza protests, when the House of Representatives voted on the new peace agreement, some accord protesters also held signs and chanted against the tax reform, believing it was connected to paying for peace and FARC reincorporation.Hide Footnote

Coalitions will also be decisive for forming a majority in Congress.[fn]Eighteen parties are in Congress; five parties competed in the first round of the last presidential election. Regionally-based small parties can amass just enough votes to enter Congress but not enough to be on the presidential ballot. See “Partidos y Bancadas”, Congreso Visible, s.f. “Elección de presidente y vicepresidente – primera vuelta”, Registraduría Nacional de Colombia, s.f.Hide Footnote The plebiscite result suggests the DCP could well increase its Senate representation. In the House of Representatives, however, it may find the going harder. Others tend to have greater regional success, including the Conservative, Liberal and U parties.[fn]Crisis Group interview, political expert, Bogotá, 10 November 2016. In the House of Representatives, the DCP has nineteen seats, six from Antioquia, five from Bogotá and eight from different departments. It has only one governor (Casanare). All but one of 27 Conservative party deputies are from outside Bogotá, as are 36 of 39 Liberals and 35 of 37 U party deputies. See “Elegidos Congreso de la República 2014-2018”, Registraduría Nacional, s.f.Hide Footnote To form congressional majorities, the DCP must count on other parties, making preservation of opposition unity vital for it. Pro-agreement parties will also seek to stay united on the issue to keep their congressional numbers.[fn]Crisis Group interview, opposition negotiator and senator, Bogotá, 30 November 2016.Hide Footnote

If the opposition does take power with a mandate against at least part of the peace agreement, implementation of the most contested areas could end. One option would be to modify laws that were impossible to change when they were originally passed due to the fast-track. Another would be to starve politically and financially key institutions, programs or policies. By underfunding them or undercutting their political importance, it could quickly make the accord an irrelevance.

The government has tried to prevent this by tabling a bill that would oblige future governments to implement the peace agreement, but this could be repealed or ignored after a shift in the balance of power.[fn]Proyecto de acto legislativo 01 de 2016 senado“, law proposed by Interior Minister Juan Fernando Cristo, 19 December 2016.Hide Footnote Pressure from abroad and vocal parts of Colombian society, however, might make it prohibitively costly to jettison the agreement. In that case, a commitment to continue implementing key parts of the agreement, such as transitional justice and humanitarian mechanisms, and not undo progress on other points, could be a viable goal for renewed dialogue between the government and opposition throughout 2017.

IV. Implementation and its Effect on Political Support

During 2017, congressional and presidential support for the agreement will be strong and stable. But popular support for its implementation is fragile and uncertain beyond the short term, putting full application of the accord at risk. A small window exists during which implementation could decisively shift backing in either direction, depending on success in carrying out fundamental parts of the pact and altering conditions in conflict-affected territories.

A. FARC Concerns

The effect on the FARC of political opposition to the peace agreement is likely to become more pronounced in coming months. Throughout the ceasefire, and particularly since the plebiscite, it has faced the risk of increased internal strains. The ceasefire violation in southern Bolívar is telling: the front operating there was unable to control its fighters or had simply continued extortion. A later expulsion of five mid-level commanders in the eastern plains, including Gentil Duarte, who had been put in charge of the faction of the First Front that supported the peace agreement, shows the strains at that level, whether due to political uncertainty, connections to lucrative illegal economies or both.[fn]Crisis Group interview, Tripartite Mechanism member, Bogotá, 9 November 2016. “FARC-EP separa a 5 mandos de sus filas”, FARC-EP, 16 December 2016. In June 2016, a First Front faction in Guaviare announced it would not be part of the agreement. FARC leadership then selected Duarte, a Central High Command member, to lead the part that still supported the peace accord.Hide Footnote

The political climate fosters one concern above all within the FARC: that the government will not fulfil its part of the accord. While the group, in its tenth conference in September 2016, ratified the whole peace agreement by consensus, the possibility of returning to the battlefield resurfaced in December, when Timochenko reminded fighters they might have to prepare for “plan B”: resumption of war. Such threats could increase if implementation is jeopardised.[fn]Crisis Group interviews, FARC members, Yarí plains, 15-25 September 2016. “Timochenko alerta a la tropa de las Farc: preparemos el plan B”, Las 2 Orillas, 8 December 2016.Hide Footnote

FARC dissidents could cause violence, which would produce a chain reaction on implementation and national political support for the agreement. There have already been reports of First Front violence in Guaviare and Vaupés.[fn]Defensoría alerta sobre reclutamiento forzado y extorsiones de bloque disidente de las Farc en Vaupés”, El Espectador, 11 November 2016.Hide Footnote Dissidence in the ranks of the Daniel Aldana Front has been confirmed in Tumaco, though it is not clear whether this is related to the peace process or a schism caused by the character of the front (formerly a mobile column). The death of Don Y, a leader of the dissident faction, at the hands of the FARC in November showed how infighting can lead to violence.[fn]‘Don Y’, el disidente de las Farc que azota a Tumaco”, La Silla Pacífica, 31 October 2016; “Las Farc mataron a ‘Don Y’”, La Silla Pacífica, 16 November 2016. Eduardo Álvarez, “Disidencias de las FARC: ¿Por qué lo hacen? ¿Qué tan peligrosas son?”, Razón Pública, 14 November 2016.Hide Footnote With other armed groups looking to take over Tumaco, home to a dense concentration of coca crops, increased violence there is probable. The removal of five commanders, plus a skirmish between eight dissident members of the 14th front and pro-agreement members of the Teófilo Forero mobile column, provide further examples.

After the weapons handover finishes, the possibility of FARC members returning to violence could grow, through dissidence or individual desertion from the reincorporation program. That program for ex-fighters is notably weak when it comes to a specific approach for mid-level commanders, who are used to handling large sums of money and enjoying political and military power. Many key details of the reincorporation process will only be decided after the census of FARC combatants is finished and Ecomun – the cooperative business the FARC is to run for their reintegration – is created. The political and humanitarian costs of fighters leaving the process to return to violence would be extremely high.[fn]“Lo que tiene que pasar este año para aterrizar los acuerdos”, La Silla Vacía, 10 January 2017. For more on FARC reincorporation, see Crisis Group Report, Colombia’s Final Steps, op. cit.Hide Footnote The already-established National Reincorporation Council and Ecomun will have key roles in keeping fighters involved.

In the immediate future, the first, most critical part of the timetable for both guerrillas and government is the weapons handover to the UN mission.[fn]On D-Day +90, FARC combatants will have to hand over 30 per cent of their weapons; on D-Day +120, another 30 per cent; and on D-Day +150, the remaining 40 per cent. By D-Day +60, the FARC will have had to hand over all light weapons, grenades, munitions and militia arms.Hide Footnote The end of FARC existence as an armed organisation is the crux of the peace agreement and was the government’s principal argument to muster support for quick renegotiation after the plebiscite. Typically, many challenges arise in such processes, including delays, logistical issues and incomplete handover of weapons, and these have already affected FARC concentration. Arms abandonment by the FARC will likely face other problems that the opposition could easily highlight to argue the process is faltering. Discovery of hidden arms, for example, would fuel a wary public’s mistrust. At the same time, lack of medium- and long-term political support for the agreement increases the possibility parts of the FARC will hedge their bets on peace and so risk further undermining public backing.[fn]Crisis Group interview, opposition negotiator, Bogotá, 30 November 2016; Pro-agreement senator, Bogotá, 30 November 2016. For more on the arms handover process, see Crisis Group Report, Colombia’s Final Steps, op. cit.Hide Footnote

A pressing reason for the FARC to prevaricate in this way is the killing of and threats against local social leaders. At least 90 killings and more than 230 threats were recorded in 2016.[fn]Armando Neira, “Asesinatos de líderes sociales, el lunar que deja el 2016”, El Tiempo, 28 December 2016. Eduardo Álvarez, “Quién sigue matando a los líderes sociales en Colombia?”, Razón Pública, 28 November 2016.Hide Footnote It is imperative that the government protect civilians in targeted communities, irrespective of who is doing the killings or whether paramilitaries are behind the wave of violence. If the hypothesis that armed groups moving into new territory and seizing control over illicit economies explains part of the killings, the government must make substantial progress on filling the power vacuum left by the FARC.[fn]Crisis Group interview, high-level diplomat, Bogotá, 28 November 2016.Hide Footnote Until then, the continued killing heightens the risk of fragmentation within the FARC and undermines the perceived benefits of peace locally.

The agreement on security guarantees, for the FARC and other activists and political actors locally, needs to be enforced quickly and effectively. The government should start by strengthening existing individual and collective protection schemes and work with local leaders on steps they can take to mitigate risks. This can be done while the FARC security system is set up during the weapons handover process. Opposition leaders should also increase the frequency and volume of their condemnations of such violence and clearly distinguish their arguments against the peace deal from the actions of violent saboteurs at the local level.[fn]This is not to say there is a connection between the opposition and this violence, but rather that perpetrators might be using opposition arguments to justify their actions.Hide Footnote The international community, already highly concerned by the violence, could raise the international visibility of these attacks by more frequent public condemnations and calls for justice.

B. Planning Successful Implementation

Weapons handover began poorly. Shockingly little had been done to install adequate infrastructure in the cantonment sites, causing delays in the first steps of the process.[fn]“Comunicado Conjunto Nº 10”, GOC and FARC-EP, 28 December 2016.Hide Footnote FARC fighters will gradually move from the pre-grouping sites as the 26 cantonments are finalised. By mid-January, land to house combatants had been rented in only seventeen. Only in two, Putumayo and Policarpa, Nariño, were FARC fighters able to stay and build the facilities they need to live. According to the government, preparation of cantonment infrastructure is moving fast.[fn]Gobierno acelera el paso para dejar listas las zonas veredales de Farc”, El Tiempo, 10 January 2017; “El 90% de los miembros de las Farc está a 10 km de zonas de desarme”, El Tiempo, 10 January 2017.Hide Footnote A renegotiated protocol has established that delivery of the materials needed to finish construction and the complete concentration of the FARC are to be accomplished by 31 January.[fn]“Acta de acuerdos de trabajo entre el gobierno nacional y las FARC-EP”, GOC and FARC-EP, 17 January 2017.Hide Footnote

While the first three deadlines – FARC concentration, destruction of unstable weapons and transfer of personal and militia-members’ small weapons to the cantonments – were not met, the government has insisted the rest of the handover process will go as scheduled. These targets could easily encounter problems, but it is essential they are met according to the accord’s terms to prevent feeding public mistrust of the group. To avoid this, fighters who arrive first in cantonments could be part of the 30 per cent to hand over their weapons at D-Day +90.

The UN mission and the Tripartite Mechanism to verify and monitor the ceasefire and weapons handover have key roles.[fn]UN: First 2 Deadlines in Colombia Cease-Fire Can’t Be Met”, The New York Times, 11 January 2017. “Ya tenemos 17 zonas arrendadas y 8 en trámite’: Carlos Córdoba”, Semana, 10 January 2017. “Para evitar más muertes, Sergio Jaramillo propone acelerar implementación de acuerdos”, El Espectador, 16 November 2016; “Intervención del Presidente Juan Manuel Santos en el acto de la Firma del Nuevo Acuerdo de Paz con las Farc”, Presidencia, 24 November 2016.Hide Footnote The former has already been engaged in verifying ceasefire violations but has also become entangled to some degree in a few incidents of improper behaviour. Though they were not directly involved, the governor of Antioquia’s accusations that under-age prostitution and heavy drinking by many FARC members in town centres suggested lack of clear information on the ceasefire process. The Tripartite Mechanism later confirmed there was no prostitution but verified a case in which a FARC member violated protocol, leaving the cantonment without permission and drinking and arguing with a civilian in a small hamlet. The governor toured the cantonments after the DCP openly supported him and subsequently backtracked on his original accusations. Still, the Tripartite Mechanism, including the UN mission, in effect became the arbiter between political rivals.[fn]La pelea entre las FARC y el gobernador de Antioquia”, Semana, 28 December 2016. “Mecanismo de monitoreo y verificación communicado de prensa”, Misión de la ONU en Colombia, 30 December 2016. “Uribe dice estar dispuesto a acompañar al gobernador de Antioquia a sitios de preconcentración”, RCN, 29 December 2016. “Así avanza la polémica revisión de las zonas de concentración en Antioquia“, Semana, 7 January 2016.Hide Footnote

More publicly, images of UN mission staff dancing with FARC members on New Years’ Eve led to an opposition outcry and claims that the mission’s credibility and impartiality had been impaired. The members involved were removed, though some questioned whether the incident had not been overblown.[fn]Misión de la onu en colombia separa a observadores de su servicio”, Misión de la ONU en Colombia, 5 January 2016. Marta Ruiz, “El episodio de los verificadores: un escándalo desproporcionado”, Semana, 6 January 2017.Hide Footnote

In its first report, which gained limited media and public attention due to the focus on the alleged scandals, the UN mission stated that 280 observers were in the country, with the number to increase to 450 in January, though the civilian component has lagged behind deployment of the military. A balance between the civilian and military parts of the mission is essential, especially in maintaining strong relationships between the mission, local authorities and communities. By 7 December, the mission had also successfully monitored 183 movements of FARC fighters and dealt with 27 requests for verification, only nine of which could be investigated as eighteen were not within its mandate.[fn]Primer informe de actividades del MM&V”, Mecanismo y Monitoreo y Verificación, 30 December 2016.Hide Footnote

The troubled start to concentration of FARC forces should not obscure the importance of the mission’s role in the future handover of weapons. Complete deployment of personnel across all cantonment sites and regular release of updates to the media beyond its regular reports, with consent of the government and FARC, would help enhance the mission’s effectiveness and public standing, as well as the public’s perception of progress. The mission should also work closely with the government and FARC to push them to follow the weapons handover schedule laid out in the peace agreement, despite early delays.

A robust communications strategy for when the FARC hands over weapons is crucial to gain support for the process. The opposition has not made the early delays the focal point of its complaints, concentrating instead on the relationship between the UN, the government and FARC. Its grievances have been based on mistrust of the guerrillas, which would only be fuelled by an insufficiently transparent weapons handover. To overcome scepticism, the FARC would be well advised to drop its long-held misgivings and allow publication of photos of fighters handing over weapons to the UN mission. Its new media savviness, which has bolstered its poor public image, and its interest in generating political capital suggest it may do so.[fn]Crisis Group interviews, FARC negotiator, Havana, 9 June 2016; high-level diplomat, Bogotá, 28 November 2016. “Nueva estética de las Farc, ¿Estrategia mediática o cambio verdadero?”, Semana, 12 November 2016.Hide Footnote The cost of not doing so could be proliferation of claims that the FARC have held back some weapons.

The handover is due to end six months after D-Day, but the Special Jurisdiction for Peace (SPJ) will not be operating by then. Given that justice was one of most controversial items in the peace accord, opposition scrutiny of the transitional justice system and misgivings about the SPJ will be prominent in the year’s debate. That benefits for FARC fighters, such as amnesties, security measures and reincorporation money will be provided early on, while SPJ sentences will be handed down much later, could create an impression that the FARC is being rewarded without having fulfilled any judicial or truth obligations.[fn]Crisis Group interview, member, international community, Bogotá, 14 December 2016.Hide Footnote As Congressional and presidential campaigns enter their final stretch, the SPJ will likely be hearing its first cases. How it handles them and how the accused behave toward the courts, will surely be closely scrutinised by the opposition and voters.

The FARC should cooperate as much as possible with transitional justice mechanisms, including the SPJ, the truth commission and efforts to find remains of victims of forced disappearance. By proving commitment to fulfilling their obligations even at personal cost, FARC leaders could sway opinions on peace and undercut opposition arguments. Not doing so or using the new judicial mechanisms to defend its war effort, point fingers and/or deny responsibility in high-profile cases would strengthen opinion against the guerrillas and the peace agreement. Members of the Armed Forces must also appear before the SPJ so as not to fuel a perception it is a mechanism aimed only against the FARC and protects state officials.

Public apologies for major war crimes should also remain part of FARC and government approaches to building support for the deal. FARC has apologised for killing eleven deputies in 2007; it should do the same for bombing the Nogal club in Bogotá in 2003. But these should not be public shows of remorse for political purposes.[fn]The public apology in La Chinita for a 1994 massacre had much show and little substance according to an attendee. Crisis Group interview, diplomat, Bogotá, 21 October 2016.Hide Footnote More generally, the focus on and participation of victims in early implementation and other peace activities should remain central, including efforts to find victims of forced disappearance and cooperation with the truth commission. Not doing so would undermine the argument that victims were central to the negotiation and the agreement’s legitimacy. Matters have been made more complicated since the plebiscite by disputes between different opposition and pro-agreement actors over the right to represent victims, a contest that will inevitably continue into the election season.[fn]Opposition actors claimed to speak for FARC victims, including Sofía Gaviria and Herbin Hoyos, as did government and pro-peace agreement leaders.Hide Footnote

The government also needs to continue educational work to create ownership of the accord by local and regional communities, as well as urbanites. Generating that sense would increase the political cost of non-implementation or diluting content. The opposition showed in the plebiscite that connecting the accord to people’s everyday lives, accurately or not, was an effective strategy.[fn]Crisis Group interviews, political activist and university professor, Bogotá, 16 November 2016; opposition negotiator, Bogotá, 30 November 2016. “El No ha sido la campaña más barata y más efectiva de la historia”, La República, 5 October 2016.Hide Footnote Proactive communication to show progress in implementation would be equally important.

A number of other important efforts related to the agreement, especially de-mining, will also be measures of tangible progress. It was recently announced that after 46 devices were destroyed over more than a year, Orejón, in Briceño, Antioquia, no longer has any landmines.[fn]Deicy Johana Pareja M., “El Orejón, la vereda que es ejemplo del desminado humanitario”, El Tiempo, 22 December 2016; “Vereda Orejón, municipio de Briceño (Antioquia)”, Dirección Contra Minas, n.d.Hide Footnote Such efforts, as well as coca crop substitution pilots and other quick-impact projects by different institutions, can help communities feel that peace has brought a major material change in their daily lives and create an expectation of economic and institutional development to come. The 700 projects that the post-conflict ministry recently announced it will begin or carry out in the first 100 days of peace, in addition to its Rapid Response Plan, will be crucial, but there is still a lack of financial and political backing within the government. Adequately managing local communities’ expectations will thus be essential.

Lastly, creation of a space to reinitiate and maintain political dialogue with the opposition on implementation should also be explored. This might begin by inviting “no” leaders to meetings on verification and/or implementation issues and help channel political debate toward serving the stability of the peace process rather than undermining it.[fn]Crisis Group interview, opposition negotiators, Bogotá, 8 and 30 November 2016.Hide Footnote

C. Peace and Other Armed Groups

Implementation of the renegotiated peace agreement faces major challenges at the regional and local levels due to the presence of other armed groups, which will in turn influence national support for the process. Foremost among these groups is the ELN, whose own putative peace process was not aligned with the FARC’s. Negotiations have not begun, though they are scheduled to start on 7 February.[fn]Fase pública de diálogos con el ELN se inicia el 8 de febrero”, El Espectador, 18 January 2017.Hide Footnote Trying to implement the FARC accord where the ELN is active poses acute dilemmas.

ELN violence will affect the government’s ability to implement the peace, especially as the group is expanding its presence and still operates in many priority conflict-affected areas, such as Catatumbo and Arauca (both on the Venezuelan border), as well as Cauca, Nariño and Chocó. Even aspects of the deal that have broad support are affected by the ELN presence, as shown in the Santa Rosa del Sur incident.[fn]“Tripartite Mechanism Communiqué: Government of Colombia and FARC-EP violated ceasefire in south Bolívar incident”, Tripartite Mechanism, 30 November 2016. The FARC fighters involved in the incident had presented themselves as ELN combatants, leading the army to attack them in the belief it would not be violating the ceasefire.Hide Footnote Establishing an effective state presence where control is still contested by armed groups will be costly in lives and resources. The integrity of peace with the FARC will also be at stake in areas where little is likely to change in terms of violence, such as Arauca, Cauca and Catatumbo.[fn]Some recent murders in Cauca, where killings have increased, have been attributed to the ELN. “Tres hombres asesinados en zona rural de Silvia, en el norte del Cauca”, El Tiempo, 20 September 2016. The ELN has also been accused of “killing communists” in Arauca. Carlos A. Lozano Guillén, “Mirador: Carta a Gabino (I)”, Periódico Voz, 2 September 2016.Hide Footnote The situation is aggravated where the FARC and ELN have made agreed or coordinated a transfer of territorial control, such as Cauca and Nariño.[fn]Crisis Group interviews, high-level diplomat, Bogotá, 2 August 2016; government official, Bogotà, 9 December 2016.Hide Footnote

Effective implementation of the FARC accord is also important to convince the ELN it can trust the government to fulfil agreements it signs with insurgents. After the plebiscite, the group internally questioned the merits of trusting the government to deliver on its promises, exacerbating mutual wariness that in any case has tended to be worse than what existed between government and FARC at the start of their negotiation.[fn]During secret government-FARC talks, the latter concluded that Santos was serious about peace. The ELN, years later and during talks to define a negotiation agenda, argued that Santos represented the same old political elite. This, for some in the ELN, has led to a view that negotiations with the government are the correct path, but not now. See Víctor de Currea-Lugo, “Eln dice estar listo para la paz, entrevista con Antonio García”, El Espectador, 9 December 2016. Crisis Group interview, ELN experts, Bogotá, 3 October 2016; Yarí plains, 23 September 2016.Hide Footnote Any further deterioration in ELN confidence in the state would imperil the possibility of a peace accord with it in the near future. It would also be calamitous with regard to public support if ELN presence became a reason for failure to implement the FARC agreement, which in turn would lead the ELN to continue to mistrust the government’s ability to deliver on peace.

Other armed groups, such as neo-paramilitaries and the remnants of the Popular Liberation Army (EPL) will also create difficulties for implementation on a local level.[fn]This report uses the term “neo-paramilitary” instead of criminal bands or Bacrim, per Soledad Granada, Jorge A. Restrepo and Alonso Tobón García, “Neoparamilitarismo en Colombia: una herramienta conceptual para la interpretación de dinámicas recientes del conflicto armado colombiano”, in Restrepo and David Aponte (eds.), Guerra y violencias en Colombia Herramientas e interpretaciones (Bogotá, 2009), pp. 467-499. The Libardo Mora Toro front, the remains of the EPL, operates in the Catatumbo region. It is the only dissident front from the 1991 EPL peace process that still exists.Hide Footnote Both have been moving into areas of former FARC control for some time and will continue to do so unless stronger judicial, political and law-enforcement action is taken against them. Colombia is witnessing a resurgence in coca cultivation, and crop substitution programs designed by the peace accord will take time to have an effect.[fn]According to the UN Office on Drugs and Crime (UNODC), coca cultivation in 2015 increased by almost 40 per cent, to 96,000 hectares. “Colombia: Monitoreo de territorios afectados por cultivos ilícitos 2015”, UNODC, July 2016.Hide Footnote If other armed groups take control of areas with illicit crops before the state can, they could directly stimulate new dynamics of violence and influence the success or failure of crop substitution.

Many social movements believe the neo-paramilitaries – mainly the Gaitán Self-defence Forces (AGC) – are behind the recent increase in killings of local activists, whether for political or economic reasons.[fn]The AGC, also known as the Gulf Cartel, Úsuga Clan or Urabeños, was created and became heavily involved in drug trafficking after the paramilitary demobilisations that ended in 2006.Hide Footnote Evidence for this is patchy, however, and there does not appear to be one specific phenomenon or organisation responsible.[fn]Some interpret the killings as done by paramilitaries under the direct influence of high-level political opposition, though there has been no evidence connecting the two. See Oto Higuita, “¿Por qué están asesinando a los voceros e integrantes del Marcha Patriótica?”, Prensa Rural, 2 December 2016. “¿Quién está ordenando matar a los líderes sociales en Colombia?”, El Colombiano, 27 November 2016. Eduardo González, “¿Quién sigue matando a los líderes sociales en Colombia?”, Razón Pública, 28 November 2016. Crisis Group telephone interview, high-level diplomat, 28 November 2016.Hide Footnote The ELN is behind some killings, as in Arauca; local armed groups connected to local political elites play roles in others, such as in Urabá; and in yet other areas, such as Caguán, the possible arrival of new armed groups might be the main factor.[fn]Crisis Group interviews, government official, Bogotá, 2 December 2016; political analyst, Bogotá, 9 Jul 2016; high-level diplomat, Bogotá, 28 November 2016. Eduardo González, “¿Quién sigue matando a los líderes sociales en Colombia?”, Razón Pública, 28 November 2016.Hide Footnote Even so, the sense that political violence is on the rise makes implementation more difficult, not only because of its effect on FARC’s transition to civilian life, but also because community leaders may come to see participation in peace mechanisms as personally risky.[fn]This is already a risk, as the renegotiated accord weakens community participation. Juanita León, “La gran diferencia entre el Acuerdo I y el Acuerdo II”, La Silla Vacía, 15 November 2016.Hide Footnote

While the agreement includes various initiatives aimed against these groups, including the new investigative unit in the attorney general’s office, and calls for international support to the initiatives, these will be necessary but likely insufficient to constrain new patterns of coercion on the ground. A clear risk exists that implementation of the peace agreement could lead to greater violence in certain areas if the state does not move quickly enough to protect local populations and also to combat neo-paramilitary groups.

There are various explanations for the prospect of a spike in criminal and political violence once the peace accord gets underway. First, armed groups could clash over control of areas the FARC leave, as has occurred between the ELN and AGC. These will be areas with strong illegal economies, trafficking routes and militarily strategic points, such as Tumaco, Chocó and the Nudo de Paramillo, for example. Other reasons are more political. While some concerns of local land-holding elites regarding rural reform aspects were addressed in the new accord, others remain and have led some of those landholders to reject the renegotiated deal.[fn]Crisis Group interview, government official, Bogotá, 9 December 2016.Hide Footnote If the relationship between some of these elites and illegal armed groups and/or actors stays in place but is targeted toward the accord’s land distribution terms, violence could well increase, including in areas traditionally vulnerable to land conflict such as Urabá.[fn]James Bargent, “BACRIM Vuelve a sus Raíces Paramilitares en la Lucha por la Tierra en Colombia”, Insight Crime, 19 July 2013.Hide Footnote

Another possible source of violence prompted by the peace accord’s implementation is continuation of attacks against social leaders due to the perceived imminent opening of the political system regionally and locally, including the sixteen special circumscriptions in Congress for conflict-affected areas.[fn]The peace agreement creates sixteen special constituencies in Congress so that conflict-affected regions can have a stronger voice in legislation and policymaking. The idea is that those who run for these seats not be part of established political parties (including that to be established by the FARC), represent isolated regions and give a voice to victims. See “Acuerdo final para la terminación del conflicto y la construcción de una paz estable y duradera”, GOC and FARC-EP, 24 November 2016, p. 54.Hide Footnote Such local, conservatively-minded political violence has history in Colombia and explains how the paramilitaries gained so much power in the 1980s and 1990s.[fn]Mauricio Romero, Paramilitares y Autodefensas (1982-2003), IEPRI (Bogotá, 2003).Hide Footnote A spike of violence in this spirit would be qualitatively similar to that against the Patriotic Union (UP) in those decades, though it is very unlikely to reach the same level. Too many international actors are already concerned and attentive, thus raising its cost.

D. Institution Building

Violence and security on the ground are not the only issues that could dent political support for the agreement in 2017. Much of the problem for smooth implementation stems from the institutions meant to manage the war-peace transition. Some national-level ones barely exist beyond paper, with little staff or capacity to execute budgets or projects: these include the National Land Agency, the Territorial Renovation Agency and the Agency for Rural Development.[fn]Crisis Group telephone interview, high-level diplomat, Bogotá, 28 November 2016.Hide Footnote If unable to execute on the ground, the state will risk losing the chance to gain local legitimacy. Since they are coming into existence at a time of tight caps on public spending, their financial and political support is also very fragile, especially given the current and historical resistance to rural reform. Facing the threat of being starved of resources or steered toward invisibility over time, as has happened with previous rural-focused institutions, they badly need an injection of high-level support, money and staff.[fn]For a quick review of land reform and institutional issue literature, see La política de reforma agraria y tierras en Colombia Esbozo de una memoria institucional, Centro Nacional de Memoria Histórica (Bogotá, 2013).Hide Footnote

Institutional fragmentation at different levels poses additional dilemmas. Political pressure to combat increasing coca cultivation has led various state actors to adopt rival strategies, some of which could be contrary to the accord’s spirit. This risk will grow if the opposition comes to power in 2018. Local political actors also have insufficient institutional and technical capacity to implement many parts of the agreement, and in some cases, their willingness will be fragile, unless they feel they can obtain financial resources for their regions.[fn]Crisis Group interview, political analyst, Bogotá, 3 December 2016.Hide Footnote

The Rapid Response Plan (RRP), designed by the post-conflict ministry (MPC) but to be implemented with and by various other institutions, needs both political and financial support. The friction between implementing institutions, the MPC and the High Commissioner for Peace’s Office remains a problem, as some institutions feel the latter two overstep their bounds and are too influential. Lack of a clear transition from the RRP to implementation of longer-term aspects of the peace agreement also raises broader concerns.[fn]Crisis Group interviews, government official, Bogotá, 9 December 2016; political analyst, Bogotá, 2 December 2016.Hide Footnote Stronger leadership from above, complemented by international community pressure, is needed.

V. A Role for the International Community

The international community has focused on supporting the peace process and ensuring it concludes with a feasible, robust agreement. It has done so by providing economic resources for state institutions and civil society actors working on peace issues; political backing for the process; delegates from the guarantor and accompanying nations and special envoys from the U.S., European Union and Germany; and technical support on implementation issues, such as support for local justice mechanisms and formalisation of land titles. Such backing remains essential.

A. Implementation and Political Support

In the current political context, a quick start to implementation, with early victories, is ever more required. Though the plebiscite suggested that high-level international support was less effective than anticipated, the delegates and envoys who aided the negotiations should continue to press for the agreement to be carried out. Not only could they be helpful in resolving disputes and influencing the government and FARC, but they will also be able to highlight the broad foreign commitment to peace in Colombia.[fn]The U.S. special envoy to the peace talks, Bernie Aronson, no longer plays this role, and no replacement is in sight. The U.S. Secretary of State nominee, Rex Tillerson, stated in a written response to questions as part of his confirmation process, that the U.S. would have to “review” the new peace agreement to determine which parts it would support. “Trump’s state nominee raises doubts on Colombia peace pact”, The Washington Post, 22 January 2017.Hide Footnote The rural reform pact in particular will face political resistance on different levels, and the international community can play a vital role in raising the cost of obstructing or ignoring it. Financial aid and pushing the government to get key institutions functioning would be major contributions.

International support can also influence FARC decisions, especially if an opposition government proposes changes to, rejects or is unwilling to implement parts of the accord. Given the group’s concerns with full implementation, a change in government could undermine its commitment to peace, leading to fragmentation as some of the organisation return to organised violence. Pressing the FARC to maintain its commitment to peace will be vital. Here the second UN mission, requested in the peace accord to monitor FARC political participation, reincorporation and security guarantees, will have a vital role in maintaining trust between the guerrillas and government. It will also be politically contentious, as its mandate covers the most controversial issue in the new agreement, namely the FARC’s participation in politics. It will need to be functioning quite soon in order to respond to its mandate; early preparation to take advantage of the period before its mandate begins is essential.

Financial aid will also be vital, especially beyond 2017. Colombia is currently unable to fully afford its post-conflict pledges, something that the international community has committed to make good. In the longer term, financial support might help persuade a new government to honour disputed aspects of the accord. Partners could also work directly with local governments, providing financial and technical assistance and ensuring that political differences between local, regional and national levels do not impede implementation.[fn]Crisis Group interview, political analyst, Bogotá, 3 December 2016.Hide Footnote Finally, international non-state actors will also need funding, such as the UN High Commissioners for Refugees and Human Rights, among others, who have important post-conflict roles on displacement, border issues and violence, all risks to a successful transition from war to peace.

B. The Special Issue of Drugs

It is unlikely that a decrease in coca cultivation resulting from application of the peace agreement will occur before 2018. New programs need time and will be largely emasculated if support wavers. Recent coca production increases have made the drug issue important again to preventing violence, but also politically critical. The opposition points to rising hectarage to argue that drug policy is not working due to concessions made to the FARC, including prohibition of aerial fumigation. Drug policy also is a source of tension within the government and between Bogotá and local communities.[fn]“Procurador colombiano acusa a Santos de proteger cultivos de las FARC”, El Nuevo Heraldo, 18 April 2016; “Uribe considera que fin de aspersiones con glifosato es exigencia de las Farc”, El Nuevo Heraldo, 10 May 2015. Between July and September 2016, protests by coca growers, mainly in Putumayo, lasted 39 days, as peasants rejected the use of fumigation chemicals applied on the ground during manual eradication. Peasants in Putumayo also have voiced concerns over drug policy and implementation of the peace agreements on crop substitution. See “Razones del paro cocalero en Putumayo”, El Espectador, 19 August 2016; “Levantan protesta cocalera en Putumayo”, El País de Cali, 7 September 2016; Crisis Group interviews, local leaders and coca growers, 20-24 March 2016. Within the government, new Attorney General Néstor Martínez has called for fumigation to be reinstated but with new chemicals. “Fiscal pide volver a la fumigación aérea contra los cultivos ilícitos”, El Tiempo, 4 September 2016.Hide Footnote When published this year, cultivation data will show another increase in 2016, before crop substitution programs derived from the peace deal begin. This may increase calls for traditional policy, including forced eradication. The government plan to substitute and forcefully eradicate 50,000 hectares each in 2017 is probably unreachable and will also create serious tensions on the ground. The forced eradication will also produce unnecessary tension with the FARC and close the state’s window to gain legitimacy in areas highly affected by coca cultivation.

The international community, especially the U.S., should give the agreement on illicit drugs a chance to prove itself and not expect immediate decreases in illicit crop cultivation. The focus should instead be on strengthening interdiction within and outside Colombia and supporting the rapid and effective implementation of the relevant points of the peace agreement. Prioritising such implementation over national and international political interests related to traditional counter-narcotics policy will be critical, not least because returning to costly forced manual eradication, the results of which are easily reversible, is no guarantee of success.

VI. Conclusion

Colombia has signed and ratified the peace agreement, and the whole of the FARC will soon be in cantonment sites, where they will hand over their weapons and begin transition to civilian life. While this is cause for celebration, how peace was signed and approved was highly controversial, and the agreement appears to lack broad, stable, sustainable political support. Implementation is threatened on several fronts, and with a united and strengthened opposition, the future looks somewhat bleak for pro-agreement leaders. The peace deal is likely to be a target for multiple grievances in the 2018 presidential election, which may produce a result not unlike that of the 2 October 2016 plebiscite.

Substantial and rapid progress on implementing crucial aspects of the agreement is needed in 2017 to shift the balance in favour of the accord. The difficult national and local contexts – a financial shortfall, a stuttering peace process with the ELN, weak institutions and internal government rivalries and high levels of targeted killings in rural areas – mean implementation faces concrete threats that have been partly aggravated by political disputes over the peace agreement. If efforts to apply the accord do not overcome these initial hurdles, parts of it may be condemned to failure before they have a chance to succeed. If that happens, FARC commitment to peace, the possibility of a similar negotiation with the ELN and prospects for addressing the root issues in the long armed conflict will all be in doubt. The immediate political battle to finalise the agreement has been won, but it is premature to declare victory for peace.

Bogotá/Brussels, 31 January 2017

Appendix A: Map of Colombia

Map of Colombia AB Carto/International Crisis Group
Ruined houses at the entrance to Kubatly town. All rights reserved
Report 255 / Europe & Central Asia

Digging out of Deadlock in Nagorno-Karabakh

If they move quickly, Armenia and Azerbaijan could break out of their long impasse over the disputed territory of Nagorno-Karabakh. They could pursue quiet talks on thorny issues – settlements, peacekeepers and final status – but along separate tracks rather than in a single package.

What’s new? An opportunity has opened to reset deadlocked talks between Baku and Yerevan over the breakaway region of Nagorno-Karabakh. The parties are a long way apart, but negotiations could help prevent a new escalation after years of growing militarisation and lay the groundwork for the conflict’s eventual resolution.

Why does it matter? The window may close if Baku and Yerevan do not act. Already the thaw in Armenia-Azerbaijan relations shows signs of frost. Without talks on key issues – the future of areas adjacent to Nagorno-Karabakh and people currently residing there, prospects for international peacekeeping, and Nagorno-Karabakh’s status – positions risk hardening further.

What should be done? On the adjacent territories, temporarily freezing new settlement construction in return for Azerbaijan refraining from legal action or new sanctions could improve prospects for talks. For peacekeepers, the OSCE High-Level Planning Group could reassess options. On Nagorno-Karabakh’s status, the parties remain far apart but informal talks could still be worthwhile.

Executive Summary

A narrow opening to breathe life into the moribund peace process between Azerbaijan and Armenia over the breakaway territory of Nagorno-Karabakh risks closing. If it does, Baku and Yerevan may not only lose the gains they have recently made but also bury the peace process for some time. Yerevan and Baku would be wise to act fast. They could start talks on issues underpinning the standoff: the future of territories adjacent to Nagorno-Karabakh in which Armenian settlers have made their homes; a potential role for international peacekeepers; and, the core issue, Nagorno-Karabakh’s status. On the adjacent territories, a time-bound freeze on new settlements in return for Azerbaijan’s pledge to pause any international legal action or new sanctions could check a gnawing problem and help unlock talks on other core disagreements. On prospects for peacekeeping, the Organization for Security and Co-operation in Europe’s (OSCE) High-Level Planning Group (HLPG), set up in the 1990s to plan for such missions, could assess options anew. The parties are bitterly divided on Nagorno-Karabakh’s status but starting discreet, informal talks could still be worthwhile.

In early 2019, progress seemed palpable. A new government in Yerevan said it was ready to seek a compromise solution. Baku appeared to be more open to exploring ways to resolve the dispute. The two countries’ relations, acrimonious since a 1992-1994 war and further damaged by clashes in 2016 that killed hundreds of people, slightly thawed. Renewed diplomatic engagement between the two reduced flare-ups and created a more favourable environment for negotiations. The Armenian and Azerbaijani governments agreed to launch humanitarian projects near the front lines and let journalists and relatives visit detainees in their respective capitals.

But the rapprochement has not led to renewed peace talks. Discussion between the two sides on their main points of disagreement over Nagorno-Karabakh have been suspended for more than a decade. Years of estrangement have hardened positions: Yerevan, Baku and the de facto authorities in Nagorno-Karabakh’s capital Stepanakert continue to make uncompromising demands regarding Nagorno-Karabakh’s ultimate fate. Moreover, over recent months Armenia-Azerbaijan relations have cooled again as each leader issued tit-for-tat claims over Nagorno-Karabakh that the other considered provocative. If the two sides fail to build on the cornerstones laid in 2019, the relative calm may not hold.

Finding a way forward ... will be no small challenge.

A renewed effort to seek compromise could help prevent tensions from once again spiralling. Specifically, the parties could revisit three issues over which they have been at loggerheads since the 1992-1994 war. The first involves the fate of territories adjacent to Nagorno-Karabakh. Azerbaijanis were forced to flee these areas during the war. Settlers – mostly ethnic Armenians displaced from Azerbaijan itself – moved in. Stepanakert now exerts authority over and funds settlements that have expanded to most of the area between Armenia and Nagorno-Karabakh. Settlers contribute significantly to the breakaway region’s economy, mostly through booming agriculture, and have strong ties to homes and communities they have built from the ground up. Finding a way forward that meets the interests of both settlers and people displaced from the adjacent areas, and also involves the return of those areas to Azerbaijan, will be no small challenge.

One option to nurture conditions for talks might be for Armenia to persuade Nagorno-Karabakh’s de facto authorities to suspend plans for new settlements and, in return, for Azerbaijan to pledge not to act on plans to pursue settlement-related complaints in international courts or impose further sanctions for a set period. Yerevan argues that decisions regarding settlement expansion are in Stepanakert’s hands. In reality, however, Armenia has considerable influence as Nagorno-Karabakh’s main security guarantor, provider of around half of its budget and main market for its products. For its part, Baku is likely to oppose such reciprocal steps, fearing that pausing legal action in return for a settlement freeze would risk appearing to accept existing settlements at a time when it feels there is greater international support for its stance. But it could reiterate publicly its position that the settlements violate international law even while pledging to halt new sanctions or legal action, and thus signal it rejects the continued existence of those settlements that are in place.

The second issue revolves around the composition and mandate of a potential international peacekeeping or monitoring mission. Such a mission could help minimise violence, create conditions for a peace deal and monitor or enforce such a deal if and when one is reached. While proposals have been circulated intermittently since 1994, particularly by Russia, no such force has ever deployed. The parties have both tended to oppose a military force or one with an outsized Russian role. An OSCE HLPG was set up in the 1990s to plan for such missions but – in the absence of progress in talks – has foundered. With the support of the parties, the OSCE could reinvigorate it and task it with a specific, time-delimited (perhaps one year) mandate to define a set of options. These could then form the basis for the parties’ discussions on such a mission.

The last issue is Nagorno-Karabakh’s independence claim, at the conflict’s core and the hardest to resolve. Armenia and Stepanakert insist on statehood. Baku is at most prepared to offer Nagorno-Karabakh self-rule within Azerbaijan. Though the parties share little common ground, there are tentative signs of movement. In Azerbaijan, senior officials have begun exploring precisely what granting the region autonomy would entail and how a referendum on its status could be organised. Their ideas remain far from anything Yerevan or Stepanakert would accept; nor do they reflect an accurate grasp of life and governance in Nagorno-Karabakh today. They could, however, offer an opening for discussion. Given the sensitivity of the issue and the distance between the parties, any talks on status would likely have to start discreetly and semi-formally.

While past dialogues have failed mostly due to disagreement and distrust between the parties, the fact that the three issues have always been discussed together, as a single package, arguably has not helped. The three are interconnected, and progress on any requires (and could enable) progress on the others. But parties have been slow to act on the first two – the settlements and the potential role of international peacekeepers or monitors – for fear that doing so could influence future discussions of the third, Nagorno-Karabakh’s status. To mitigate this constraint, the parties could pledge that any agreement reached would be without prejudice to talks on other issues.

Direct talks between the parties inevitably entail risks. They could highlight the distance between the two sides’ positions, thereby fuelling mutual anger and potentially reversing the past months’ gains. But years of continued stalemate have put a potential solution further out of reach and isolated Armenians and Azerbaijanis from one another. The more time goes by, the more facts on the ground will be entrenched, the harder they will be to reverse and the graver the risk of war. If talks might make matters worse, their continued absence almost certainly will. Getting back to the table will be difficult but is the only way Armenia and Azerbaijan can start digging out of their deadlock.

Baku/Yerevan/Stepanakert/Tbilisi/Brussels, 20 December 2019

Flowers have been placed here, 100m from the Line of Contact, in memory of a girl shot dead during the 2016 escalation in Hasanqaya, a village located in Azerbaijan's Tartar region.  CRISISGROUP/Jorge Gutierrez Lucena

I. Introduction

The coming to power in 2018 of a new government in Yerevan raised hopes of a reset in relations between Armenia and Azerbaijan. The two countries have been deadlocked for over two decades over Nagorno-Karabakh, which declared independence from Azerbaijan in 1991. The 1992-1994 war that followed pitted Azerbaijan’s armed forces against Nagorno-Karabakh rebels backed by the Armenian army. It ended with Nagorno-Karabakh’s de facto independence and a self-proclaimed government based in Stepanakert. Armenian forces also took effective control over seven regions adjacent to Nagorno-Karabakh.[fn]UN Security Council Resolutions 822, 853, 874 and 884, all adopted during the 1992-1994 war, as well as UN General Assembly Resolution 62/243 adopted in 2008, refer to these territories as occupied.Hide Footnote Tens of thousands died in the fighting. Although exact numbers are contested, well over 400,000 ethnic Azerbaijanis were displaced from the territories adjacent to Nagorno-Karabakh and some 40,000 from Nagorno-Karabakh itself.[fn]See “Всесоюзная перепись населения 1989” [Soviet Census 1989], which lists 440,000 inhabitants in total on those territories. Official Azerbaijani statistics count over 700,000 Azerbaijanis displaced from the adjacent territories, but also include descendants of those who initially fled. “On the districts bordering Armenia or Nagorno-Karabakh, territories of which are either occupied or affected by the Armenian armed forces”, official website of the State Committee for the Affairs of Refugees and Internally Displaced Persons of the Republic of Azerbaijan.Hide Footnote In addition, hundreds of thousands of Armenians from throughout Azerbaijan and Azerbaijanis from throughout Armenia fled their homes during the war.[fn]See “Soviet Census 1989”, op. cit., which lists a total pre-conflict Azerbaijani population in Armenia of 84,860 and a pre-conflict Armenian population in Azerbaijan of around 245,000, excluding around 145,500 Armenian living in the Nagorno-Karabakh Autonomous Oblast (NKAO). The Armenian government estimates that there are now more than 360,000 refugees from Azerbaijan. Also see “Azerbaijan: Analysis of Gaps in the Protection of Internally Displaced Persons”, UN High Commissioner for Refugees, October 2009, which estimates that 200,000 Azerbaijanis fled Armenia.Hide Footnote

Although a May 1994 ceasefire ended open conflict, peace has been elusive. Since the ceasefire, the conflict parties have reviewed and rejected several plans proposed by international mediators. Armenia has continued providing political, military and financial support to the breakaway region, which Baku views as Armenia-occupied Azerbaijani territory. Tension occasionally has led to clashes, the worst of them in the spring of 2016. Then, four days of fighting killed hundreds, although again exact numbers are disputed. It left Azerbaijan in control of slightly more territory in Nagorno-Karabakh and the adjacent territories than before. It also left the combatants thinking about a rematch.

Armenian and Azerbaijani leaders have taken steps to reverse what had seemed a slide toward a new war.

In the last eighteen months, however, Armenian and Azerbaijani leaders have taken steps to reverse what had seemed a slide toward a new war. Direct leadership contacts and communication channels between security personnel and political representatives in capitals have minimised flare-ups and casualties. Both countries’ leaders also agreed to launch humanitarian projects and support visits of relatives of detainees held in each other’s capitals as well as of journalists, the first of which occurred in November.[fn]“Leyla Abdullayeva answers the question of media regarding the mutual visits of journalists from Azerbaijan and Armenia”, official website of the Republic of Azerbaijan, 25 November 2019; “Answer by spokesperson of the MFA of Armenia on the question about the journalist exchange programme”, official website of the Ministry of Foreign Affairs of the Republic of Armenia, 26 November 2019.Hide Footnote This slight thaw marks a substantial shift. It is the first reversal in what had been a steady decline in relations since the April 2016 clashes.

When they are ready to come to the table, the parties will have a quasi-roadmap at hand. If, despite decades of negotiations, they have never settled on a peace plan, Baku and Yerevan have agreed to a framework for talks. It begins with the core principles of the Helsinki Final Act, which mediators and the parties endorsed during the Organization for Security and Co-operation in Europe (OSCE) Ministerial Summit in Madrid in 2007. These include refraining from the threat or use of force, preserving states’ territorial integrity, and protecting the equal rights and self-determination of peoples.[fn]“Fifteenth Meeting of the Ministerial Council 29 and 30 November 2007: Statements and Declarations by the Ministerial Council Decisions of the Ministerial Council”, OSCE, 30 November 2007; “Seventeenth Meeting of the Ministerial Council 1 and 2 December 2009: Statements and Declarations by the Ministerial Council Decisions of the Ministerial Council”, OSCE, 2 December 2009.Hide Footnote

Based on these principles, in 2009-2012 the OSCE Minsk Group, co-chaired by the U.S., Russia and France, proposed six additional elements as a guide for talks, which neither Baku nor Yerevan has ever publicly rejected:[fn]Since 2009, the de facto authorities of the Nagorno-Karabakh entity voiced repeated concerns over the elements. For example, see “Statement of the MFA of NKR”, official website of the Ministry of Foreign Affairs of the Republic of Artsakh, 15 July 2009.Hide Footnote

  • Creating an interim status for Nagorno-Karabakh that provides guarantees for security and self-governance;
  • Returning the territories surrounding Nagorno-Karabakh to Azerbaijani control;
  • Building a corridor linking Armenia to Nagorno-Karabakh;
  • Determining the final legal status of Nagorno-Karabakh through a legally binding expression of will;
  • Upholding the right of all internally displaced persons (IDPs) and refugees to return to their former places of residence; and
  • Granting the parties international security guarantees that would include a peacekeeping operation.[fn]“Statement by the OSCE Minsk Group Co-Chair countries”, OSCE, 10 July 2009.Hide Footnote

If the parties appeared to accept these elements, they understood them differently. Armenia viewed “interim status” warily, but agreed because it expected that the proposed referendum, held in Armenian-majority areas, would culminate in Nagorno-Karabakh’s independence. Even if some ethnic Azerbaijani IDPs were to return, their numbers would be insufficient to sway the result.[fn]Crisis Group interviews, current and former Armenian officials, analysts, December 2017-March 2019.Hide Footnote For its part, Azerbaijan assumed that interim status, which would involve Azerbaijani rule in some form for as long as it lasted, could be indefinite – as no deadline was set for the referendum – or at least offer an opportunity for Azerbaijanis to reintegrate and, officials say, win over the Armenian population.[fn]Crisis Group interview, former senior Azerbaijani official, Baku, March 2019.Hide Footnote Baku also saw an opportunity to restore control over the adjacent territories “without a shot being fired”.[fn]President of Azerbaijan: ‘Our patience also has limits’”, Euronews, February 2010; Crisis Group interviews, current and former Azerbaijani officials, analysts, May 2018-March 2019.Hide Footnote

Given the distance between the two sides’ understanding of where the six elements would lead, it is perhaps not surprising that progress since has stalled. There is no agreement on interim, let alone final, status, the adjacent territories remain under Armenian control, IDPs are still displaced and no international peacekeepers or monitors have deployed. To break out of this deadlock, the parties must find ways to resolve three main areas of disagreement:

  • the fate of seven adjacent regions in which thousands of ethnic Armenians have settled and which are under the effective control of the de facto authorities in Nagorno-Karabakh;
  • the mandate and composition of an international peacekeeping or observer mission that could buttress any political agreement; and
  • Nagorno-Karabakh’s ultimate status.

Thus far, all efforts to tackle the three issues have sought to do so in toto. The three are interdependent: resolution of the conflict will require a single comprehensive agreement, not piecemeal understandings. But failure to look at each issue independently has hampered discussion of any of them.

This report examines these three issues with an eye to finding ways to break the impasse. It is based on interviews with local and international officials, experts, and members of the general population residing in Armenia, Azerbaijan and Nagorno-Karabakh and the adjacent areas in 2017-2019. It factors in the parties’ legal and political positions, but does not advocate any particular stance in the ongoing dispute, simply aiming to help the parties overcome a debilitating stalemate and take advantage of a slight thaw in relations. Geographical names reflect the usage of the pre-war years in the 1990s. The report acknowledges that the current population of the de facto Nagorno-Karabakh entity does not include ethnic Azerbaijani IDPs forced to flee the territory during the 1992-1994 war.

Crisis Group's expert Zaur Shiriyev (seated) meets with Azerbaijani IDPs living close to the Line of Contact. CRISISGROUP/Jorge Gutierrez Lucena

II. Territories Adjacent to Nagorno-Karabakh

In 1994, Armenian forces took control of the seven Azerbaijani districts adjacent to Nagorno-Karabakh. In five (Jebrail, Zangelan, Kubatly, Lachin and Kelbajar) they took full control, while seizing only parts of Agdam and Fizuli. Settlers soon followed. Today, they comprise around 11 per cent of the combined population of the adjacent areas and Nagorno-Karabakh itself – the territories that Azerbaijan views as occupied by Armenia and their numbers continue to grow.[fn]Estimate based on information from the current and former de facto officials, who consider that 147,000 people live in the areas under their control; about 15,000 live west and south west of Nagorno-Karabakh, while around 2,000 live in smaller settlements south and east of the region. See “Demographic handbook of Artsakh 2019”, National Statistical Service of the Republic of Artsakh, 2019. This information was cross-checked with other public sources; for more details, see Appendix C.Hide Footnote They represent a major challenge for both Armenia and Azerbaijan. Neither country has publicly expressed willingness to discuss the settlements as part of peace talks, even as Azerbaijan continues to demand the return of those territories.[fn]Crisis Group interviews, officials, Yerevan, Stepanakert, Baku, 2017-1018.Hide Footnote

A. History

Before the war, the seven districts were populated predominantly by Azerbaijanis. This population fled during the fighting. Afterward, Armenian and de facto authorities in Stepanakert saw limited settlements as a way to establish control over strategically important territory, notably the one road connecting Armenia with Nagorno-Karabakh, which runs through the town of Lachin.[fn]Crisis Group interview, former de facto officials, Stepanakert and Yerevan, March-April 2018.Hide Footnote According to a former de facto official, a secret order issued by the de facto authorities, under Yerevan’s supervision, called on ethnic Armenians to settle in the town and a handful of nearby villages in order to control that road.[fn]Crisis Group interview, former de facto official, Yerevan, April 2018.Hide Footnote The de facto authorities felt that four settlements in Lachin district would suffice.[fn]Ibid.Hide Footnote

Some Armenian activists and war veterans had bigger plans, however. Instead of limiting settlements to Lachin, they argued that it was ethnic Armenians’ “moral right” to settle land that centuries ago was part of the Kingdom of Greater Armenia.[fn]

Many Armenian nationalists argue that these territories are part of Artsakh, a region in the Kingdom of Greater Armenia, which existed for almost 600 years starting in the 2nd century BC. They refer to history in general and to specific artefacts, sites and monuments of cultural and religious significance. Since 1995, de facto authorities have asserted different names for towns, villages and districts in these territories. Some were picked from Armenian history books, while others correspond to the names of Armenian towns and villages in eastern Turkey under the Ottoman Empire. Crisis Group interview, former de facto senior official, Yerevan, April 2018; Crisis Group interviews, leaders of resettlement process, Lachin and Kelbajar districts, Stepanakert, December 2017, March 2018. Also see fn 28.

Hide Footnote Through media campaigns and Armenian charities, they encouraged ethnic Armenians to move to not only the town of Lachin and nearby villages but all the adjacent territories. Areas between Armenia and Nagorno-Karabakh – namely Lachin, Kelbajar, Kubatly and Zangelan – and along the roads connecting Stepanakert to Agdam district saw the most substantial growth.[fn]Crisis Group interviews, former de facto senior official and leaders of resettlement process, Yerevan, Stepanakert, Lachin and Kelbajar districts, December 2017, March-April 2018.Hide Footnote By 1995, Lachin district housed twelve settlements instead of the planned four.[fn]There are no signs that military authorities took part in fostering the settlement process. The first military units were deployed in the main towns of the adjacent districts. Their bases remain fenced in and personnel rotate on a regular basis. No military personnel have permanent homes in any nearby settlements. Crisis Group interviews, December 2017, March 2018. Whose army is present in the conflict region is disputed. Stepanakert insists that these are its troops and that any linkages with Armenian military personnel or institutions take place only through special bilateral agreements. Because Baku rejects the possibility of an independent Karabakh force, it views military personnel in the area as occupying Armenian forces.Hide Footnote By 2004, there were about 13,500 permanent residents in dozens of new villages across the four districts between Nagorno-Karabakh and Armenia.[fn]Crisis Group interviews, former de facto officials, leaders of resettlement process, settlers, December 2017, April-March 2018.Hide Footnote

The settlers arrived at the ruins of Azerbaijani villages destroyed during the war. In some areas, the Armenian forces had burned homes and other infrastructure and mined the land to prevent Azerbaijanis from returning.[fn]Crisis Group interviews, Armenian veterans, December 2017.Hide Footnote Most settlers were already socially or economically vulnerable. The majority were ethnic Armenians displaced from neighbouring regions of Azerbaijan during the conflict or migrants from nearby mountainous areas of Armenia in search of free housing and land.[fn]De facto authorities have registered up to 30,000 people as “Armenian refugees from Azerbaijan”. Another 60,000 ethnic Armenians are considered IDPs by Stepanakert as they come from Shahumyan district. This is a district in Azerbaijan where Armenians have a long history. De facto authorities consider it a part of Nagorno-Karabakh “occupied by Azerbaijan”. Crisis Group interviews, de facto officials and civil society representatives, March 2018. The 2005 Report of the OSCE Fact-Finding Mission also mentions victims of 1988 Armenia’s earthquake among the new settlers; see “Report of the OSCE Fact-Finding Mission (FFM) to the occupied territories of Azerbaijan surrounding Nagorno-Karabakh (NK)”, OSCE, 2005.Hide Footnote

Yerevan strongly opposed attempts to settle the surrounding territories except for the Lachin road, to which “no Azerbaijani was going to return anyway”.

The settlements, especially those outside Lachin, made the Armenian government nervous. While some Armenian charities offered limited help with relocation costs, Yerevan refused to provide direct assistance for fear of international criticism and Azerbaijani legal action.[fn]Crisis Group interviews, settlers and former de facto officials, Stepanakert and Yerevan, December 2017, March-April 2018. Armenian leaders have never provided direct support to settlements. But Yerevan has provided budget support to Stepanakert since the 1990s, part of which the latter has used to finance the settlements since 2006. For details see Appendix C.Hide Footnote Several of those who founded settlements say that Yerevan even tried to prevent them doing so.[fn]Crisis Group interviews, founders of settlements and former de facto officials, Stepanakert and Yerevan, March-April 2018.Hide Footnote One former de facto official who sought financial support from private sources in Yerevan to improve living conditions in Zangelan and Kubatly reported consistent obstacles throughout his tenure, which lasted into 2004. Change, he said, came only after he left office and Stepanakert took full control of the territories in 2006.[fn]Crisis Group interview, former de facto official, Yerevan, April 2018.Hide Footnote A politician with close links to the Armenian leadership of the 1990s confirmed that Yerevan strongly opposed attempts to settle the surrounding territories except for the Lachin road, to which “no Azerbaijani was going to return anyway”.[fn]Crisis Group interview, Armenian politician, Yerevan, July 2018.Hide Footnote

With limited resources, settlers throughout the 1990s lacked proper construction materials or equipment. To repair schools and other public buildings, they tried to raise private funds locally and in Armenia.[fn]Crisis Group interviews, settlers, Lachin and Kubatly districts, December 2017.Hide Footnote The settlements were isolated, with minimal access to public goods or services such as electricity or telephone connections.[fn]In some settlements, residents took part in local parliamentary elections organised by de facto authorities since 1997. In 1998, de facto authorities mentioned the adjacent territories in a law “on administrative division”; see “ԼՂՀ ՎԱՐՉԱՏԱՐԱԾՔԱՅԻՆ ԲԱԺԱՆՄԱՆ ՄԱՍԻՆ” [About NKR’s Administrative Division], official website of the National Assembly of the Republic of Nagorno-Karabakh, adopted on 16 June 1998. In 2000, the de facto parliament invited representatives of the settlements to take part in its sessions as “observers”. Crisis Group interviews, current and former de facto officials, member of de facto parliament, former and current heads of settlements, Stepanakert, Lachin, Kubatly, Zangelan and Kelbajar districts, December 2017, March 2018.Hide Footnote

In 1998, de facto authorities in Stepanakert began to exert control over the settlements, starting in Lachin and continuing with Kelbajar, though their investment in those areas remained minimal.[fn]Crisis Group interviews, former de facto officials, Yerevan and Stepanakert, March-April 2018. In 1998, de facto authorities assigned new names to the main towns and administrative units. Lachin town was renamed Berdzor. Lachin, Kubatly and Zangelan districts were merged into one administrative unit called Kashatagh. Zangelan was renamed Kovsakan, Kubatly to Sanasar. Kelbajar town was renamed Karvachar, and the district was renamed Shahumyan to recall the territory to the north of Nagorno-Karabakh, which the de facto leadership considers “occupied by Azerbaijan”. Agdam merged with a new Askeran district and was renamed Akna. Fizuli became part of Martuni district and was renamed Varanda. Jebrail district was merged with Hadrut and its main town renamed Jrakan. For some of the names, see “About NKR’s Administrative Division”, op. cit.Hide Footnote De facto officials deployed, and the authorities took on partial salary payment for local teachers, workers responsible for public cultural events, and nurses (of whom there are few, and only in some settlements).[fn]The 1999 budget law already included an explicit reference to expenditure in Kelbajar, see “ԼՂՀ 1999Թ. ՊԵՏԱԿԱՆ ԲՅՈՒՋԵԻ ՄԱՍԻՆ” [About 1999 State Budget of NKR], official website of the National Assembly of the Republic of Nagorno-Karabakh; “ԼՂՀ ՕՐԵՆՔԸ ԼՂՀ 1999 ԹՎԱԿԱՆԻ ՊԵՏԱԿԱՆ ԲՅՈՒՋԵԻ ՄԱՍԻՆ” [Decree to the Law on 1999 State Budget of NKR], Database of the Ministry of Justice of the Republic of Nagorno-Karabakh, 30 December 1998.Hide Footnote Local residents did not always welcome the new authorities.[fn]Crisis Group interviews, settlers and former and current heads of settlements, Kelbajar district, December 2017.Hide Footnote One of the first de facto police officers deployed to Kelbajar reports that due to hostility from inhabitants he spent several nights in his car instead of asking for shelter at a local house.[fn]Crisis Group interview, de facto official, Kelbajar, December 2017.Hide Footnote Even if other districts were more welcoming, Stepanakert’s involvement failed to bring what settlers wanted most: real financial support.[fn]Crisis Group interviews, settlers, Kelbajar, Lachin, Kubatly and Zangelan districts, December 2017.Hide Footnote

International attention to the settlements continued to make Yerevan uneasy. After a 2005 fact-finding mission, the OSCE Minsk Group co-chairs explicitly called for an end to new settlements.[fn]“Report of the OSCE Minsk Group Co-Chairs’ Field Assessment Mission to the occupied territories of Azerbaijan surrounding Nagorno-Karabakh”, OSCE, 24 March 2011; “Executive summary of the Report of the OSCE Minsk Group Co-Chairs’ Field Assessment Mission to the occupied territories of Azerbaijan surrounding Nagorno-Karabakh”, OSCE, October 2010.Hide Footnote They also urged the parties to reach agreement on the territories’ fate, which they saw as the only way to avoid settlers laying down deeper “roots and attachments to their present places of residence”.[fn]“Letter of the OSCE Minsk Group co-chairs to the OSCE Permanent Council”, p. 3.Hide Footnote An agreement, they thought, could also make it possible to end the settlers’ “miserable and isolated” living conditions.[fn]Ibid.Hide Footnote One former de facto official told Crisis Group that these statements led Yerevan to instruct Stepanakert to halt even the basic financial support it was providing. “Because of [the co-chairs], people were spending winters with holes in their roofs, and I could not help them”, he said.[fn]Crisis Group interview, former de facto official, Yerevan, April 2018.Hide Footnote Indeed, many former and current de facto officials and politicians continue to blame the OSCE Minsk Group for, in their view, forcing settlers to live for years in poverty.[fn]Crisis Group interviews, former and current de facto officials, politicians, Kelbajar, Lachin, Kubatly and Zangelan districts, Stepanakert, Yerevan, December 2017, April 2018, October 2019.Hide Footnote

This changed in 2006, when the de facto government in Stepanakert adopted a constitution claiming full but temporary jurisdiction over the adjacent territories and thus the settlements.[fn]Article 142 of the Nagorno-Karabakh constitution says: “Until the restoration of the state territorial integrity of the Nagorno-Karabakh Republic and the adjustment of its borders public authority is exercised on the territory under factual jurisdiction of the Republic of Nagorno-Karabakh”.Hide Footnote The constitution was recognised only by the de facto Nagorno-Karabakh authorities. Nonetheless, it provided a framework through which Stepanakert began to increase services throughout the surrounding settlements.[fn]One popular program among this predominantly poor population aims to increase birth rates through financial allowances and special banking preferences for each child born to a family in the de facto region: after the birth of a fifth or sixth child, families often receive free housing. Some families in the settlements with four or more children have received help from the de facto government with housing repairs or been given a new home. See the de facto government’s decrees to this effect on its official website.Hide Footnote

Perhaps most significantly, Stepanakert’s greater involvement after 2006 has jumpstarted agriculture in the area.[fn]The program officially started with the 2007 establishment of the agriculture fund, which offered farmers preferential credits for grain and fertilisers. New technology and equipment helped increase yields. Produce was sold inside the de facto region, as well as in Armenia. The program led to a sharp increase in demand for agricultural land, with the size of cultivated land tripling in ten years. About 80 per cent of the economically active population is now engaged in farming, which is the leading source of employment after the local army. Crisis Group interviews, senior de facto official, de facto officials and parliamentarians, Stepanakert, December 2017, March 2018. Also see “Made in Artsakh: Как бизнесмены подняли с колен непризнанную республику” [Made in Artsakh: How businesspeople raised the unrecognised republic from its knees], Sekret Firmy, 14 October 2015.Hide Footnote Although agricultural programs are meant to span the whole region, they have little success in Nagorno-Karabakh’s mountainous terrain. In the settlements, however, agriculture boomed. This helped improve living standards and attract new settlers. Entrepreneurs began to lease large plots of land, employing other settlers to work them. Over time, settlers began to organise their own plots. “Nine years ago, when we first arrived, we worked only on 150 hectares of land”, said a settler in Kubatly. “Now there is not a piece of land to spare”.[fn]Crisis Group interviews, settlers in Kubatly district, December 2017.Hide Footnote

Baku has closely followed these developments. Since the 2005 OSCE fact-finding mission, Azerbaijan has increasingly emphasised the growing settlements and illegal economic activity at international organisations and in bilateral discussions with foreign partners.[fn]Illegal settlement of Armenians in the occupied territories of Azerbaijan by Armenia as a gross violation of the principles of international law”, Council of Europe (CoE), 28 June 2006; “The situation in the occupied territories of Azerbaijan”, UN General Assembly (UNGA) 62/243, 25 April 2008; “Annex to the letter from the Permanent Representative of Azerbaijan to the United Nations addressed to the Secretary-General”, 27 April 2010; “Speech by Ilham Aliyev at the 65th session of the United Nations General Assembly”, official website of the President of the Republic of Azerbaijan, 23 September 2010.Hide Footnote Since 2016, the Azerbaijani foreign ministry has disseminated regular reports and satellite imagery of settlement expansion.[fn]For more details, see “Illegal economic and other activities in the territories of Azerbaijan occupied by Armenia”, Ministry of Foreign Affairs of the Republic of Azerbaijan, 2016; and “Illegal activities in the territories of Azerbaijan under Armenia’s occupation: Evidence from satellite imagery”, Report by Azercosmos OJSCo and the Ministry of Foreign Affairs of the Republic of Azerbaijan, 2019.Hide Footnote Some Azerbaijani officials suggest that the settlements could be cause for future sanctions and legal action against Armenia.[fn]Crisis Group interviews, officials, Baku, May 2018-March 2019. Also see “Azerbaijan calls on PACE to impose sanctions on Armenia”, Azernews, 23 June 2015; “Azerbaijani Embassy in Washington calls on the international community to consider sanctions against Armenia”, New Azerbaijan Party website, 6 August 2014.Hide Footnote

B. The Settlements Today

Yerevan’s apprehension, Baku’s protests and the OSCE Minsk Group’s appeals have not constrained the settlements’ growth. Today, about 17,000 ethnic Armenians live in the territories between Armenia and Nagorno-Karabakh.[fn]The population of the Nagorno-Karabakh conflict region is about 140,000. See “The Results of 2015 population census of the Republic of NKR”, National Statistical Service of the Republic of Artsakh. Before the war in 1990s, the population numbered around 190,000. See “Soviet Census 1989”, op. cit. For more details on demography in the adjacent territories, see Appendix C.Hide Footnote Among them are some 4,000 children and young people born since 1994, a faster rate of growth than in Nagorno-Karabakh itself.[fn]De facto officials and politicians argue that about 10,000 children were born in the adjacent territories. Crisis Group could not find evidence for this claim. According to the statistics of the de facto authorities in 2004-2018, 3,889 children were born in the most populated districts of Kelbajar, Kubatly, Lachin and Zangelan. Crisis Group estimates based on figures published by the local statistical office. See “The Demographic Handbook of Artsakh 2019”, op. cit.; “The Regions of NKR in Figures 2010-2016”, The National Statistical Service of the Republic of Artsakh, 2016; “The regions of NKR in figures 2008-2014”, The National Statistical Service of the Republic of Artsakh, 2015; “The regions of NKR in figures 2003-2009”, The National Statistical Service of the Republic of Artsakh, 2010. For more demographic details, see Appendix C.Hide Footnote

Life remains tough in the settlements. Settlers have restored water and power supplies, but public transport between settlements and other destinations remains non-existent. The trip from either Stepanakert or any of the closest towns in Armenia along the remains of winding, damaged roads to settled areas can take up to a day.[fn]Crisis Group interviews, settlers in Zangelan, Kubatly and Kelbajar districts, December 2017.Hide Footnote With public hospitals far away and bad roads limiting access to emergency health care providers, people have adapted. Home births, for example, are typical.[fn]Ibid.Hide Footnote While many settlers are displaced victims of war, they have been unable to receive foreign aid because the settlements are illegal according to international law.[fn]Crisis Group interviews, representatives of international humanitarian organisations, March 2018.Hide Footnote

But if public transport, connectivity and health-care access have not much improved, the agricultural sector continues to expand rapidly.[fn]Crisis Group interviews, settlers in Zangelan, Kubatly and Kelbajar districts, December 2017.Hide Footnote Lachin, Kubatly and Zangelan districts dominate local agricultural production, accounting for more than one quarter of Nagorno-Karabakh’s and the adjacent territories’ output (for both export and local consumption) in 2016.[fn]See “The regions of NKR in figures 2000-2016”, op. cit.Hide Footnote Indeed, farming and the construction of small- and medium-sized hydroelectric power stations in the adjacent territories have significantly contributed to Nagorno-Karabakh’s economy and the de facto government’s revenues.[fn]For more details on agriculture in the adjacent territories, see Appendix C.Hide Footnote

To maintain growth driven by agricultural expansion, Stepanakert now plans to start developing previously unsettled areas.[fn]Crisis Group interviews, de facto leadership, Stepanakert, December 2017, March 2018.Hide Footnote Parts of Jebrail and Fizuli districts, to the south and east of Nagorno-Karabakh, had been largely left settlement-free, possibly due in part to pressure from Yerevan, which sought to leave itself the option of a peace deal that would return those areas to Baku’s control.[fn]Ibid.Hide Footnote Increasing demand for land, however, has made de facto officials and the Nagorno-Karabakh population more determined to maintain control of those areas. Even those who once saw the territory as subject to a bargain now want to hold on to it. Settlers have cultivated unsettled land along all major roads in the territories, up to the rear positions of Armenian troops along the line of contact with Azerbaijani forces. Areas near the Araks River on the Iranian border have proven particularly promising for farming.[fn]Crisis Group interviews, de facto leadership and parliamentarians, Stepanakert, December 2017 and March 2018.Hide Footnote

In October 2017, Nagorno-Karabakh’s de facto president, Bako Sahakyan, identified expanding the settlement of the adjacent territories as a priority for 2017-2020.[fn]“Speech of President Sahakyan at the enlarged consultation dedicated to the key points of the 2017-2020 Artsakh Republic President Program”, official website of the President of the Republic of Artsakh, 16 October 2017.Hide Footnote In 2018, his government allocated $800,000 in the de facto entity’s annual budget to populate and develop new settlements, the first time that funds were earmarked for this purpose.[fn]Crisis Group interviews, de facto leadership and parliamentarians, Stepanakert, December 2017, March 2018. See “Արցախի Հանրապետության 2018թ. պետական բյուջեի մասին” [About State Budget of Artsakh government for 2018], Ministry of Finance of the Republic of Artsakh, 2018. The 2019 budget does not include a similar line. See “Արցախի Հանրապետության 2019թ. պետական բյուջեի մասին” [About State Budget of Artsakh government for 2019], Ministry of Finance of the Republic of Artsakh, 2019.Hide Footnote In early 2019, it unveiled plans for a new settlement in Fizuli with several rows of houses close to fertile farmlands.[fn]Արաքսի հովտում նոր գյուղ է ձևավորվում” [A new village getting founded in the Araks valley], Armenia Public TV on YouTube, 28 February 2019.Hide Footnote At the time of writing, however, no new settlement was reported either in Fizuli or Jebrail.

If the Armenian government decides [to transfer the territory to Azerbaijani control], I am ready to take up arms against them

As settlements grow, so does opposition to returning the lands to Azerbaijan.[fn]For instance, de facto officials, analysts and civil activists voiced similar sentiments during Crisis Group interviews in Stepanakert in July 2015.Hide Footnote Stepanakert and the settlers increasingly question any sort of peaceful coexistence with ethnic Azerbaijanis.[fn]Crisis Group interviews, de facto officials and politicians, Stepanakert, December 2017, March 2018.Hide Footnote Many settlers say that they have lived for twenty years in extremely tough conditions, without financial support, and will fight to stay. “If the Armenian government decides [to transfer the territory to Azerbaijani control], I am ready to take up arms against them”, said an Armenian villager in Lachin district.[fn]Crisis Group interview, resident of settlement, Lachin district, December 2017.Hide Footnote The Armenian diaspora’s backing, which has kept these communities afloat in the absence of other assistance, strengthens settlers’ belief that these territories are “primordially Armenian lands”.[fn]Crisis Group interviews, de facto officials, politicians, experts, Stepanakert and Yerevan, December 2017, March-April 2018. See comment by the representative of the Tufenkian Foundation, an Armenian diaspora organisation that has been the main provider of financial support to projects in the adjacent areas, in “For Armenians, they’re not occupied territories – they’re the homeland”, Eurasianet, 6 August 2018. For more detail on the support of the Armenian diaspora organisations, see Appendix C.Hide Footnote April 2016’s outbreak of fighting hardened these positions.

Nor is it clear that settlers would be willing to move elsewhere if offered compensation, as some Armenian officials and politicians suggest.[fn]Crisis Group interviews, former officials and politicians, Yerevan, December 2017, April 2018.Hide Footnote A minority might: “If people ask us to leave, we will not stay”, said a settler in Jebrail district.[fn]Crisis Group interview, resident of settlement, Jebrail district, December 2017.Hide Footnote But given the settlements’ growing economic importance, the investments settlers have made in creating homes and communities for themselves, and the narrative that the land is Armenian, financial incentives may not suffice. One diplomat suggested that attempts at resettlement would prompt a “tsunami of protest” from both settlers and inhabitants of Nagorno-Karabakh itself, to say nothing of Armenian nationalists at home and within the diaspora.[fn]Crisis Group interview, diplomat, Tbilisi, December 2017.Hide Footnote He cautioned that the issue of settlements could become a Pandora’s box: adding one more insurmountable issue to an already long list.[fn]Crisis Group interview, diplomat, Tbilisi, December 2017.Hide Footnote For Armenian officials pessimistic that peace talks will get anywhere, avoiding any discussion of the settlements, and thus allowing their growth and postponing decisions on their fate, is preferable to trying to resolve the question now.[fn]Crisis Group interviews, former and current officials, Yerevan, March 2018, July 2018, October 2019.Hide Footnote

A legal dimension further complicates the situation. In 2015, the European Court on Human Rights ruled that Armenia exercises effective control over the territory of Nagorno-Karabakh and the surrounding areas and therefore bears responsibility for them under the European Human Rights Convention.[fn]“Case of Chiragov and others v. Armenia”, European Court of Human Rights, 16 June 2015. The European Court of Human Rights issued two similar decisions in Chiragov and Others v. Armenia and Sargsyan v. Azerbaijan, which discussed the issue of financial compensation to displaced people as a result to the ongoing conflict. Those decisions held that due to a lack of political solution to the Nagorno-Karabakh conflict, the applicants should be awarded compensation as just satisfaction in respect of pecuniary and non-pecuniary damage.Hide Footnote The case has increased Armenia’s concern that discussing settlements would amount to an admission of occupation, and thus its legal responsibility for Nagorno-Karabakh and the adjacent districts.[fn]Crisis Group interviews, former and current officials, Yerevan, March 2018, July 2018, October 2019.Hide Footnote

Since 2016, Baku has intensified international outreach to warn countries against contacts with and aid or investment to settlements in the adjacent areas. It has gathered satellite imagery documenting the settlements’ expansion and hired lawyers to build its case for how they violate international law.[fn]“Baku calls foreign companies to avoid illegal activity at the occupied territories”, Turan Agency, 23 July 2018.Hide Footnote The Azerbaijani government is now considering filing lawsuits in the European Court of Human Rights against the Armenian government and individuals engaged in the regions. It hopes that victory in these cases will lead to asset freezes and other penalties against those responsible.[fn]Crisis Group interview, senior officials, Baku, November 2019.Hide Footnote While Yerevan does not fear that such cases or sanctions would substantially harm the Armenian economy, it does worry that the cases may bolster Azerbaijan’s stance that the territories are occupied.[fn]Crisis Group interviews, senior officials, Yerevan, October 2019.Hide Footnote

C. Steps to More Constructive Talks on Settlements

Space for starting any conversation on settlements is limited. Azerbaijan’s current position is clear: the settlements are illegal, and their continuation is creeping expropriation of Azerbaijani territory.[fn]“Speech by Ilham Aliyev at the 65th session of the United Nations General Assembly”, op. cit.; “Letter dated 10 April 2017 from the Permanent Representative of Azerbaijan to the United Nations addressed to the Secretary-General”, op. cit.Hide Footnote Azerbaijani officials fear that opening talks could implicitly signal recognition of the settlements. Baku wants to link any discussion of the settlements to the return of the adjacent territories and IDPs.[fn]Crisis Group interview, senior official, Baku, November 2019.Hide Footnote Yerevan and Stepanakert usually dispute that the territory in question is occupied; they also cite security requirements as their rationale for maintaining control of the land.[fn]Crisis Group interviews, officials and politicians, Yerevan and Stepanakert, 2017-2019.Hide Footnote For Yerevan, any talks must be tied to Azerbaijani compromises including on Nagorno-Karabakh’s status. As noted, it also fears that any talks in which it is perceived as representing Stepanakert’s interests regarding the settlements could be read as a tacit admission of occupation, bolstering Azerbaijani claims in international courts regarding the settlements and potentially serving as fodder in a future Azerbaijani campaign to convince other states to impose sanctions against Armenia.[fn]Crisis Group interviews, officials and politicians, Yerevan, December 2017, April 2018.Hide Footnote

But if prospects for substantive discussions between Armenia and Azerbaijan regarding the settlements appear remote, failure to talk incurs mounting risks. The gap is widening between the sides: Azerbaijanis believe that the territories could be easily returned, while Armenians are increasingly convinced that they should not be. Neither side discusses how Armenian settlers could coexist with returning Azerbaijani IDPs; absent such a dialogue, there is no hope of reversing resistance to a possible compromise within those two groups.

One option for paving the way for more substantive talks might be for Yerevan and the de facto authorities to cease settlement construction in return for Baku pledging to abstain from advocating settlement-related sanctions or filing any related claims in international courts against Armenia. This moratorium could be time-delimited (perhaps for one year) to provide a window for talks to show progress. Without it, the settlement expansion that Stepanakert plans would create more obstacles to a resolution, further harden positions and bolster constituencies against peace in both countries.

Such an arrangement would have advantages for Yerevan. As things stand, new settlements strengthen the belief among settlers and Armenian nationalists that no territory can be given up. By turning a blind eye to them, Yerevan sets the stage for the problem to grow and complicate any eventual discussions. Building new settlements also risks making the situation on the ground even more unmanageable as existing ones are plagued by poverty, dilapidated housing and shoddy infrastructure.

Yerevan would likely have to apply substantial pressure to persuade the de facto leadership to accept a one-year moratorium on further settlements, but the de facto entity’s dependence on Yerevan gives it considerable influence. Armenia is Nagorno-Karabakh’s main security provider and represents it in the official talks with Azerbaijan. Yerevan also supplies around half of Stepanakert’s budget and remains the main market for Nagorno-Karabakh’s products.[fn]For more details, see Appendix C.Hide Footnote

Azerbaijani officials understandably fear that accepting an arrangement ... could signal that Baku accepts existing settlements.

For its part, Baku would likely have strong reasons to reject reciprocal steps along these lines, but there might be ways to address its concerns. Azerbaijani officials understandably fear that accepting an arrangement whereby settlement expansion ceases in return for a freeze on sanctions or legal action could signal that Baku accepts existing settlements. They may also sense they have international winds in their sails given their chairmanship of the Non-Aligned Movement, a grouping set up for nations that backed neither the U.S. nor the Soviet Union during the Cold War, and support for Azerbaijan’s position on Nagorno-Karabakh from the Organisation of Islamic Cooperation.[fn]“The Aggression of the Republic of Armenia against the Republic of Azerbaijan”, Organisation of Islamic Cooperation Resolution No. 10/11, in “Resolutions on Political Affairs adopted by the Eleventh Session of the Islamic Summit Conference”, March 2008; “Cairo Final Communique of the Twelfth Session of the Islamic Summit Conference”, 6-7 February 2013; “OIC adopts two resolutions on Azerbaijan”, Azernews, 25 January 2018.Hide Footnote

But accepting a freeze does not mean accepting the settlements. Baku could make public statements reiterating its view that the settlements are illegal under international law even while pledging to put any plans for new sanctions or legal action on hold for a delimited time. This would signal that it does not accept the legality or continued existence of those settlements that are in place. Moreover, a freeze on settlements would serve Azerbaijan’s interests as much as those of Armenia. The settlement expansion threatens to make any prospective return of the adjacent territories harder and costlier for Azerbaijan.

Another option to help break the deadlock on settlements might be a comprehensive, independent assessment of the situation in the adjacent territories. This would in turn help counter misperceptions and potentially lay the groundwork for an informed discussion about those territories’ future. Past OSCE fact-finding missions, in 2005 and 2010, were intended to document the settlements’ existence. Both were brief, enjoyed limited access and lacked relevant expertise. In 2005, the team identified a small number of settlements and reported the destruction of infrastructure and high levels of poverty among the local population.[fn]“Report of the OSCE Fact-Finding Mission (FFM) to the occupied territories of Azerbaijan surrounding Nagorno-Karabakh (NK)”, OSCE, 28 February 2005.Hide Footnote As noted above, OSCE Minsk Group co-chairs recommended that Yerevan and Stepanakert prevent new settlements on the back of that report.[fn]Ibid.Hide Footnote The 2010 mission published only an executive summary, which echoed the 2005 findings. It called on “the leaders of all the parties to avoid any activities in the territories and other disputed areas that would prejudice a final settlement or change the character of these areas”.[fn]“Executive summary of the Report of the OSCE Minsk Group Co-Chairs’ Field Assessment Mission to the occupied territories of Azerbaijan surrounding Nagorno-Karabakh”, op.cit.Hide Footnote

In 2018, the question of an assessment resurfaced. First, at the January meeting of the Armenian and Azerbaijani foreign ministers, the OSCE Minsk Group co-chairs discussed a possible new mission.[fn]“Press Statement by the Co-Chairs of the OSCE Minsk Group”, OSCE, 18 January 2018.Hide Footnote Baku was supportive, hoping that an assessment would draw fresh international attention to the settlements and more support for Azerbaijan’s demand for the territories’ unconditional return.[fn]Crisis Group interview, senior Azerbaijani diplomat, Baku, May 2018.Hide Footnote Some de facto officials in Stepanakert were also positively inclined but, for their part, hoped an assessment would demonstrate that the territories’ immediate return was no longer feasible.[fn]Crisis Groups interviews, de facto officials, Stepanakert, December 2017, March 2018.Hide Footnote Yerevan never made its position public.[fn]Crisis Group interviews, officials, Yerevan, July 2018.Hide Footnote In late 2018, after the change in Armenian leadership, Baku proposed a fact-finding mission. This time, Yerevan conditioned a mission on a similar assessment in the relatively small territories of the Soviet-era Nagorno-Karabakh oblast currently controlled by Azerbaijan.[fn]These are parts of Martakert and Martuni regions of Nagorno-Karabakh that both Armenian and de facto authorities consider “occupied” by Azerbaijan. In 2015, one of these territories started building a settlement for around 1,100 people, mainly IDPs from the Nagorno-Karabakh conflict zone. For more details about the settlement, see “Məcburi köçkünlər üçün yeni salınmış qəsəbədə Novruz şənliyi” [IDPs celebrate Novruz in a newly built settlement], official website of the Tartar Regional Administration, 20 March 2018. Stepanakert also claims authority over the former Shahumyan region, which was mainly populated by the ethnic Armenians, who fled the region during the war in 1990s. Stepanakert declared this region part of its territory when announcing independence in 1991.Hide Footnote Baku rejects this idea because, as one Azerbaijani official told Crisis Group: “Those territories are part of the internationally recognised lands of Azerbaijan and should undoubtedly be under Azerbaijan’s legal authority”.[fn]Crisis Group interview, senior Azerbaijani official, November 2019.Hide Footnote

In principle, a new, comprehensive survey of the facts on the ground could prove valuable to all sides. By providing a neutral outside perspective, the assessment could give Yerevan, Baku and Stepanakert a shared understanding of realities and constraints on the ground and help them look for pragmatic solutions that would best serve those affected by their decisions. The problem is that the disagreement between the two sides on an assessment’s purpose for now appears unbridgeable. For Armenia and the de facto authorities, it would have to focus primarily on settlers’ needs – their livelihoods, access to health care and education and other aspects of their socio-economic well-being – and thus potentially open up the area to international humanitarian organisations that for now do not work in the settlements.[fn]Representatives of two such organisations told Crisis Group they would be ready to do so if Armenia and Azerbaijan reached an agreement, endorsed by international mediators, to give them access. Crisis Group interviews, representatives of international humanitarian organisations, Yerevan, Tbilisi, March and November 2018, January 2019.Hide Footnote Azerbaijan rejects an assessment with that focus, fearing it would solidify the settlements’ existing status. For Baku, the main purpose of an assessment should be to draw international attention back to the settlements and their illegality and pave the way for displaced Azerbaijanis to return.[fn]According to a senior Azerbaijani official, a new survey must assess “the factual situation in order to prepare for the eventual safe and dignified return of Azerbaijani IDPs and in accordance with UNGA Resolution 62/243, for ‘creating appropriate conditions for this return, including the comprehensive rehabilitation of the conflict-affected territories’”. Crisis Group interview, senior Azerbaijani official, Baku, November 2019.Hide Footnote

While for now it appears unlikely that a way through exists, the co-chairs could continue to explore options with the two sides to see if there is some space for compromise.

Armenians in the Ishkhanadzor settlement on the main road between Zangelan and Kubatly districts. It is one of the biggest and fastest growing settlements in the region. All rights reserved

III. Prospects for an International Mission

Proposals for an international peacekeeping or monitoring mission to Nagorno-Karabakh are as old as the peace talks themselves. Here, too, breakdowns in communication have stymied useful discussions. Four questions pertaining to an international mission are relevant. The first concerns any potential mission’s mandate, including whether it would be military or civilian and whether it would deploy before or after a peace agreement. The second relates to the role of the OSCE’s HLPG, established after the 1992-1994 war to explore peacekeeping options. The third involves a potential expansion of the regional Office of the Personal Representative of the OSCE Chairman-in-Office (PRCiO), established in 1995. The last concerns Russia’s role.

A. Mandate and Makeup

The main questions concern what an international mission would do – whether helping reduce risks of violence even absent a peace deal or deploying after a deal to monitor or enforce its provisions – and, following from that, who it would comprise. Officials, military officers and civil society representatives in Yerevan, Baku and Stepanakert mostly oppose any foreign military deployment, which they fear could lead to “an occupying force”.[fn]Crisis Group interviews, political and military officials, Baku, Yerevan and Stepanakert, December 2017, March-May 2018, February-March 2019, October 2019.Hide Footnote Armenian, Azerbaijani and de facto officials tend to prefer that any mission be civilian-led, limited to observation and armed only for self-defence.

Agreement ends there, however. The two sides’ perspectives on a mission’s potential role and conditions under which it would deploy differ, based on their contrasting visions of peace. Yerevan and Stepanakert oppose any mission that would require Armenians to withdraw their military forces. They see themselves as guarantors of ethnic Armenians’ security and are unwilling to surrender that role to outsiders.[fn]Crisis Group interviews, de facto political and military officials, Stepanakert, December 2017, March 2018, October 2019. In its official statements, Armenia advocates for a heavily armed peacekeeping force in case of a peace agreement that would discuss a need for withdrawal of the Armenian troops from the conflict zone. (See Armenia’s “Statement on the 2019 Programme Outline”, PC.DEL/652/18, 22 May 2018; “Statement on the 2020 Programme Outline”, PC.DEL/476/19, 6 May 2019.) Crisis Group interview, official, Yerevan, October 2019.Hide Footnote They do, however, feel that a civilian observer mission deployed as peace talks continue would signal international commitment and build confidence.[fn]Crisis Group interviews, officials, de facto political and military officials, Stepanakert and Yerevan, December 2017, March-April 2018, February and October 2019.Hide Footnote

In contrast, Baku would like a mission to deploy as part of an agreement that includes the withdrawal of Armenian (whether they report to Yerevan or Stepanakert) forces. It supports the demilitarisation of the lines between Nagorno-Karabakh and the adjacent territories. That said, Baku is sceptical of any international mission absent agreements on the return of both the adjacent territories to Azerbaijani control and Azerbaijani IDPs to those territories and Nagorno-Karabakh itself, fearing that it could entrench the status quo. Were such agreements in place, a mission could monitor their implementation and undertake some policing functions in areas with mixed populations. This, Baku believes, could support “integration and peaceful coexistence” between Armenians and returning Azerbaijanis.[fn]Crisis Group interview, senior Azerbaijani diplomat, Baku, May 2018.Hide Footnote

B. The OSCE’s High-Level Planning Group

Although peacekeepers have never been deployed to Nagorno-Karabakh, the OSCE High-Level Planning Group, which reports annually to the OSCE Permanent Council, has had a nearly three decades-long mandate, approved at the 1994 OSCE Budapest Summit, to figure out how they could be.[fn]The office of the Personal Representative of the OSCE Chairperson-in-Office is the only international mission to promote peace in relation to Nagorno-Karabakh, but it consists of only six unarmed international members and has no mandate to monitor the conflict, let alone keep the peace. For more, see “Mandate of the Personal Representative of the OSCE Chairman-in-Office on the conflict dealt with by the OSCE Minsk Conference”, European Parliament, 15 June 2011 and the official website of the OSCE Minsk Group. The HLPG’s mandate, adopted by the Chairperson-in-Office (CiO) on 23 March 1995, tasks it: a) to make recommendations for the Chairperson-in-Office on developing as soon as possible a plan for the establishment, force structure requirements and operation of a multinational OSCE peacekeeping force; and b) to make recommendations on, inter alia, the size and characteristics of the force, command and control, logistics, allocation of units and resources, rules of engagement and arrangements with contributing states. See “Survey of OSCE Field Operations”, OSCE Secretariat, June 2019.Hide Footnote But the HLPG is handicapped by the absence of progress toward a peace agreement, without which peacekeeping scenarios and its operational requirements remain theoretical or outdated. Reportedly, the HLPG presented four options for a multinational peacekeeping mission in 1995, ranging from traditional armed peacekeeping to unarmed observer/monitoring missions. But the OSCE acknowledged soon after that “conditions which would allow the deployment of such an operation are […] still lacking”.[fn]“Fifth meeting of the Ministerial Council: Chairman’s summary decisions of the Budapest Ministerial Council Meeting”, OSCE, 8 December 1995.Hide Footnote

While successive OSCE Chairmen-in-Office have approved the HLPG’s work, some OSCE personnel worry about its capacity. Its annual papers are often repetitive and fail to reflect changing political realities.[fn]Crisis Group interviews, diplomats, Vienna, November 2017.Hide Footnote Former OSCE personnel argue that plans developed in the 1990s are also out of step with modern UN peacekeeping standards.[fn]Ibid.Hide Footnote The HLPG comprises nine people, based in Vienna. The number is a fraction of the 31 originally planned though the organisation is not fulfilling its full mission and could ramp up if needed.[fn]Crisis Group interviews, former OSCE personnel, Vienna, November 2017.Hide Footnote It reported on conditions in the conflict zone in 1995-1997 and, in 2007, prepared an estimate of the costs for a military peacekeeping and an unarmed observer mission.[fn]The group also updated its operating plans in 2007. “Annual Report 2008”, OSCE, 2009.Hide Footnote Today, however, HLPG staff rarely visit the area.[fn]Crisis Group interviews, OSCE current and former staff, Armenian, Azerbaijani and de facto officials, November 2017, May, July and October 2018, October and November 2019.Hide Footnote To do so, they require – and often fail to secure – explicit consent from Baku, Yerevan and Stepanakert.[fn]Yerevan and Stepanakert have consistently raised concerns over a Turkish former military officer on the HLPG staff because Turkey that openly supported Azerbaijan during the April 2016 escalation. Crisis Group interviews, officials and de facto officials, Yerevan and Stepanakert, December 2017, October and November 2019.Hide Footnote

For more than a decade, attempts to improve the HLPG’s capacity have been blocked by the fact that both sides link it to their wider disagreements, when they need not do so. Each has politicised the body’s work by seeking to use it as a tool to advance their agendas: in Azerbaijan’s case, to push its demand that Armenian forces withdraw from Nagorno-Karabakh and the adjacent areas; and in Yerevan and Stepanakert’s case, to force a discussion of Nagorno-Karabakh’s status mission.[fn]Crisis Group interview, official, Yerevan, October 2019.Hide Footnote This has hobbled the body’s ability to provide the technical expertise needed to support discussions of new security arrangements for populations living in the conflict zone to support the peace process. Neither Yerevan nor Baku has called for the body’s dissolution, for fear of appearing the spoiler. Yet both sides voice frustration with its functioning, suggesting an opening for agreeing to reform efforts by the HLPG’s own staff.[fn]Crisis Group interviews, OSCE, Armenian and Azerbaijani former and current officials, Vienna, Tbilisi, Yerevan, Baku, November 2017, May and October 2018, July and October 2019.Hide Footnote

The OSCE, co-chairs and parties to the conflict could take steps to make the HLPG more useful. They could grant it the necessary access and resources and charge it with a time-delimited (perhaps one year) task of defining an updated set of scenarios and options for international peacekeepers or monitors. The resulting report would be provided to the OSCE, co-chairs and parties. The latter should then disseminate it and foster expert debate in Baku, Yerevan and Stepanakert. Updated plans would allow all involved to assess the value and feasibility of an international mission, whether before or after a final peace agreement is defined. For its work in Nagorno-Karabakh and the adjacent areas, the HLPG would need to cooperate with the OSCE’s Office of the PRCiO, which has unrestricted access to the de facto authorities in the region.

C. The OSCE’s Office of the PRCiO

Another modest step forward could be for the parties to reach agreement on an increase of personnel in the Personal Representative’s office, which is responsible for monitoring along the Line of Contact and the Armenia-Azerbaijan border, and other trust-building measures.[fn]For the full list of responsibilities of the PRCiO, see “Mandate of the Personal Representative of the OSCE Chairman-in-Office”, op. cit.Hide Footnote Since the office’s establishment in 1995, its six international staff have been based in Georgia’s capital Tbilisi.[fn]The Personal Representative does not have a permanent base and travels between Baku, Tbilisi, Yerevan and Stepanakert on a regular basis.Hide Footnote Twice a month they visit the conflict zone for a short-term monitoring exercise that usually lasts several hours. The time and location of their visits are agreed with Baku, Yerevan or Stepanakert – depending on which front lines they are visiting – beforehand.[fn]See Crisis Group Europe Report N°244, Nagorno-Karabakh’s Gathering War Clouds, 1 June 2017. Who on the Armenian side is communicating details depends on the location of the visit: if it takes place along the Armenian-Azerbaijani border, then the logistics are settled between Baku and Yerevan; if it takes place in the Nagorno-Karabakh conflict zone, then Baku and Stepanakert make the arrangements.Hide Footnote

The question of reinforcing its staff numbers has been on the table since the aftermath of the April 2016 escalation. Back then, Baku and Yerevan agreed in principle to increase the office’s staff, envisaging adding between four and six international personnel to the current six.[fn]Crisis Group Commentary, “Politics and Security Hold Each Other Hostage in Nagorno-Karabakh”, 17 January 2018.Hide Footnote That agreement was endorsed by the three Minsk co-chairs (the U.S., France and Russia).[fn]“Joint Statement of the Minister of Foreign Affairs of the Russian Federation, Secretary of State of the United States of America and State Secretary for Europe Affairs of France”, OSCE, Vienna, 16 May 2016.Hide Footnote It was then re-confirmed by the Armenian and Azerbaijani leaders in a joint statement with the Russian president.[fn]Совместное заявление президентов Азербайджанской Республики, Республики Армения и Российской Федерации по нагорнокарабахскому урегулированию” [Joint Statement of the Presidents of the Republic of Azerbaijan, Republic of Armenia and the Russian Federation], official website of the Kremlin, 20 June 2016.Hide Footnote Since then, both Armenia and Azerbaijan have, in principle, remained committed to that deal.

What stops its implementation is disagreement over where the new staff will be located. Yerevan and Stepanakert would want new staff based not in Tbilisi but on either side of the line of contact, with two or three on each side.[fn]Crisis Group interviews, officials, Stepanakert and Yerevan, May 2017, April 2018, October 2019.Hide Footnote Some Armenian and de facto officials also want PRCiO’s staff to enjoy unlimited access to the front line without having to request permission in advance from the parties.[fn]Crisis Group interviews, officials, Stepanakert and Yerevan, October and November 2019.Hide Footnote Armenia thus hopes to deter any potential assault from Azerbaijan. Baku rejects those ideas, which run contrary to its core demand for peacekeepers or monitors to deploy only after an Armenian troop withdrawal. It supports an increase only if staff are based in Tbilisi and their modus operandi remains unchanged. Baku also insists that any increase be temporary, with a duration probably of one year. That way, “if we don’t see any progress on the settlement of the conflict, we reserve the right to return to earlier arrangements”, said one senior Azerbaijani diplomat.[fn]Crisis Group interview, senior official, Baku, November 2019.Hide Footnote

Despite the parties’ deeply rooted concerns, the staff increase might still be possible and would allow the PRCiO to better monitor the front lines and fulfil other trust-building steps, potentially including humanitarian projects. One way out might be assigning new staff to Tbilisi, but with an agreement that they would visit the conflict zone more often. Moving forward even on a modest increase of four to six personnel on the basis of the parties’ agreement would be a confidence-building step for wider discussions on potential security provisions for people living in affected areas. Revising the PRCiO’s modus operandi is a taller order. Azerbaijan is unlikely to give up its requirement that monitors seek its permission before travelling to the front lines. Doing so would make it resemble too closely a monitoring mission, which Baku finds unacceptable absent an agreement on the return of adjacent territories and IDPs. The sides could continue discussing this option, possibly with the involvement of the HLPG’s reformed staff and the support of the OSCE Minsk Group co-chairs.

D. Russia’s Role

Azerbaijani and Armenian fears of Russian domination of any mission pose another obstacle to the agreement on peacekeepers deploying. Russia has played a leading role in mediating the Nagorno-Karabakh conflict for more than a decade. It is the only regional power to have publicly pledged peacekeepers, offering to deploy troops to Nagorno-Karabakh at least twice since the 1992-1994 war (first shortly after the 1994 Bishkek protocol was signed and more recently since 2015, as discussed below).[fn]Vladimir Kazimirov, “Мир Карабаху” [Peace for Karabakh], International Relations, 2nd edition (Moscow, 2015), p. 184. Crisis Group interviews, Russian diplomats, Yerevan, December 2017.Hide Footnote Current and former Russian officials argue that a Russian-led peacekeeping force could help prevent further hostilities, even absent a wider agreement.[fn]Crisis Group interviews, Russian diplomats, Yerevan, December 2017.Hide Footnote Indeed, Russia is the only world power still actively involved with Nagorno-Karabakh. For more than a decade it has been the main player in the OSCE Minsk Group. The two other co-chairs – France and the U.S. – are increasingly disengaged.[fn]Crisis Group interviews, officials, foreign diplomats, Yerevan, Baku, March-May 2018.Hide Footnote The EU has no formal role.[fn]Some experts still suggest a possible role the EU could play in supporting confidence-building measures between the conflict actors. See “Nagorno-Karabakh: Is it time to bring peacekeeping and confidence building back on the agenda?”, European Policy Centre, 29 September 2016.Hide Footnote Yerevan views Turkey, another significant regional power, as Azerbaijan’s backer, and thus not a plausible actor for negotiations or peacekeeping.[fn]Crisis Group interviews, officials, Yerevan, December 2017, April 2018.Hide Footnote

Yet Russia’s offers to deploy troops to Nagorno-Karabakh have consistently been rebuffed by Armenia, Azerbaijan or both.[fn]Crisis Group interviews, current and former officials, military, Baku, Yerevan, Stepanakert, 2017-2018.Hide Footnote The two countries share fears that a mission led by Russia or comprising its forces would fail to solve the conflict while creating new dangers.[fn]Crisis Group interviews, officials, Baku, Yerevan, December 2017, March-May 2018.Hide Footnote Sceptics argue that peacekeepers would serve as cover for an increased Russian military presence in the region. Yerevan and Baku also fear a Russian mission could undermine their sovereignty and increase their dependence on Moscow.[fn]Crisis Group interviews, officials, Baku, Yerevan, December 2017, March-May 2018.Hide Footnote Both maintain good relations with Russia, but with the Kremlin selling weapons to the two of them and working to increase its influence throughout the broader region (including in Turkey, Iran, and Georgia), neither fully trusts Moscow’s motives.[fn]Crisis Group interviews, foreign diplomats, Yerevan, December 2017, April 2018.Hide Footnote

Moscow’s most recent offer to send peacekeepers was part of the so-called Lavrov Plan. First mooted by Russian Foreign Minister Sergei Lavrov in 2015 and reiterated in 2016, following the outbreak of hostilities, it was never formally acknowledged by the Kremlin. The plan proposed pairing the deployment of Russian armed forces to Nagorno-Karabakh with a gradual withdrawal of Armenian forces from the adjacent territories and granting “interim status” to the breakaway entity for an unidentified period of time. It did not offer any clarity on what a referendum or longer-term status would look like.[fn]Crisis Group interviews, foreign diplomats and officials in Baku, Yerevan, May 2018.Hide Footnote

Azerbaijani ruling elite has increasingly accepted that Russia will be part of any resolution to the conflict.

Azerbaijani leaders, backing away from their usual opposition to a Russian presence, first supported the concept, hoping it could lay the groundwork for the return of the adjacent territories and reduce the number of Armenian forces in the conflict region. Baku’s willingness to consider the Russian proposal, despite its reservations about Moscow’s role and armed peacekeepers, also reflected its recognition that Western disinterest meant that no other outside power was likely to get involved. Over time, the Azerbaijani ruling elite has increasingly accepted that Russia will be part of any resolution to the conflict. Indeed, in 2016, the Azerbaijani government was open to a Russian-led peacekeeping mission in Nagorno-Karabakh, which it hoped would replace Armenian forces and guarantee security for settlers until the territories’ return to Azerbaijan’s control (at which point IDPs could return).[fn]Crisis Group interviews, senior officials, Baku, May 2018.Hide Footnote

Yerevan, for its part, rejected the proposal. The central problem was the lack of provisions for resolving Nagorno-Karabakh’s final status and Armenian leaders’ fear that such peacekeepers would effectively eliminate future prospects for the region’s independence.[fn]Crisis Group interviews, officials, foreign diplomats, Baku, Yerevan, Tbilisi, 2017-2018.Hide Footnote Some Armenian officials reportedly said the paper could have been drafted in Baku, given how closely they believed it hewed to Azerbaijani interests.[fn]Crisis Group interview, diplomat, January 2017.Hide Footnote

Today, the Lavrov proposal is off the table. Not only does Armenia continue to reject it, but Azerbaijan, too, has reverted to past calls for “balanced peacekeeping forces” comprising contingents from several countries, not only Russia.[fn]Crisis Group interviews, senior officials, Baku, February-May 2019.Hide Footnote Some in Baku have even returned to old formulae, arguing that if co-chair countries contribute to the peacekeeping force, their individual contributions should make up no more than 10 per cent of the mission’s personnel – thus keeping Russian numbers down – and that peacekeepers should be unarmed.[fn]Crisis Group interviews, senior officials, Baku, May 2018.Hide Footnote As for Armenia, it remains ready to support a multinational presence with Russian participation while negotiations continue, as long as it is unarmed, has an observation mandate only and does not require Armenian troops’ immediate withdrawal.[fn]Crisis Group interviews, officials, Stepanakert, Yerevan, December 2017, March-April 2018.Hide Footnote

The HLPG likely has to factor Armenian and Azerbaijani uneasiness over a large Russian contingent into its planning. Objections in Baku and Yerevan do not necessarily rule out Russia playing an important role. Indeed, Azerbaijan’s previous acceptance of the Lavrov plan shows that it at least could agree to a Russian-led mission under appropriate conditions. But the HLPG might explore a formula that allows for a mix of forces without a majority from any one state, which would temper fears of excessive Russian influence.

Sarsang water reservoir in the north of Nagorno-Karabakh. All rights reserved

IV. Nagorno-Karabakh’s Status

Whether Nagorno-Karabakh will remain part of Azerbaijan or become an independent state is the conflict’s central question. Both Armenian and de facto leaders demand independence, which for many of them would serve as a stepping-stone to Nagorno-Karabakh’s eventual unification with Armenia. Azerbaijan, on the other hand, considers independence anathema. It is willing to allow Nagorno-Karabakh substantial autonomy within Azerbaijan, though has never presented a clear and detailed proposal on what that would look like.[fn]Azerbaijan’s position is founded on the 1996 Lisbon Declaration adopted at the annual OSCE Summit. Baku refers to the 1996 OSCE summit declaration that stated that the status of Nagorno-Karabakh should be “defined in an agreement based on self-determination which confers […] the highest degree of self-rule within Azerbaijan”. Armenia did not endorse this statement. See Annexes 1 and 2 in “Lisbon Document 1996”, OSCE, 2 December 1996.Hide Footnote All previous debates on status were spurred by proposals from foreign mediators.

Both sides claim strong ties to the territory. While Nagorno-Karabakh has long been home to many ethnic groups, Armenians have been the majority for centuries and Armenian culture and society have deep roots there. The region also figures prominently in Azerbaijan’s history, literature and art. The suffering of hundreds of thousands of Azerbaijanis forced to flee Armenia, Nagorno-Karabakh and the adjacent territories during the 1992-1994 war has kept alive Baku’s demands for the territory’s return. In the words of one Azerbaijani politician, “you cannot concede part of your identity, especially when it is tangible and visible like pieces of land”.[fn]Crisis Group interview, official, Baku, September 2018.Hide Footnote

Both countries agreed to these [three] principles believing they could serve their own aspirations.

With both parties making unyielding claims, the only breakthrough in 25 years of negotiations came at the 2007 Madrid OSCE Ministerial meeting – and even then, progress was limited. Armenia and Azerbaijan agreed to the three basic principles, which were later developed into the six elements (all outlined in this report’s introduction). These included granting Nagorno-Karabakh a temporary “interim status” that would end after “a legally binding expression of will” – a referendum, in other words – to determine the region’s final status. Details on how the vote would work, what question or questions it would ask and who could cast ballots were left to further negotiations. As described, both countries agreed to these principles believing they could serve their own aspirations regarding status, not to find solutions that the other would accept.[fn]Since 2009, the de facto authorities of Nagorno-Karabakh entity voiced repeated concerns over the elements. For example, see “Statement of the MFA of NKR”, op. cit.Hide Footnote

Since that time, deadlock, militarisation and the 2016 clashes have not only worsened prospects for Baku and Yerevan to reach agreement on status but also closed space for discussion of the issue between the two societies. Baku’s legal restrictions on civil society, widespread support among both sides’ populations for military action during and after the 2016 clashes and collapsing faith in negotiations have strongly discouraged Azerbaijanis from talking to Armenians.[fn]Crisis Group interviews, pro-government analysts, Baku, April 2018.Hide Footnote In both Azerbaijan and Armenia, only a handful of activists spoke out against the 2016 confrontation, highlighting how small (and narrowing) any potential lobby for dialogue is on both sides.[fn]“Assessment of the April 2016 Conflict Escalation between Armenia and Azerbaijan: A Critical Analysis of the Reaction of the ‘Track Two’ Community in Azerbaijan”, Eurasia Partnership Foundation/Caucasus Research Resource Centre, 2018.Hide Footnote For its part, Yerevan is hostile to any conversation that might throw into question Nagorno-Karabakh’s future independence. Stepanakert is determined not to give up the self-governance it has enjoyed for two and a half decades.[fn]Crisis Group interviews, Armenian officials and the de facto officials, Yerevan and Stepanakert, March-April 2018; Yerevan, February and October 2019.Hide Footnote All this leaves little room for discussion between the conflict-torn populations about Nagorno-Karabakh’s future and further hardens public opinion across the board.[fn]“Azerbaijan detains activists amid Karabakh tensions”, Institute for War and Peace Reporting, 15 August 2014; on the situation of Azerbaijan’s civil society in 2014-2017, see “2018 Country Reports on Human Rights Practices: Azerbaijan”, U.S. Department of State, 2018.Hide Footnote

Moreover, after the thaw of early 2019, Baku-Yerevan relations appear to have become frostier again over recent months, with the leaders exchanging tit-for-tat claims on the disputed territory, sometimes referred to as the battle of punctuation marks. In August 2019, Prime Minister Pashinyan appeared to toughen his rhetoric, declaring: “Artsakh [Nagorno-Karabakh] is Armenia, period”.[fn]“Nikol Pashinyan attends opening of 7th Pan-Armenian Summer Games”, official website of the Prime Minister of the Republic of Armenia, 5 August 2019.Hide Footnote The next day, his foreign ministry downplayed the comments.[fn]“The Statement of the Ministry of Foreign Affairs of the Republic of Armenia”, official website of the Ministry of Foreign Affairs of Armenia, 6 August 2019.Hide Footnote But the comment was welcomed by hardliners in Armenia and denounced in Azerbaijan. Two months later, Azerbaijani President Aliyev retorted: “Karabakh is Azerbaijan, exclamation point”.[fn]Ilham Aliyev, speech to the 16th Annual Meeting of Valdai International Discussion Club, 3 October 2019, cited on the official website of the President of the Republic of Azerbaijan.Hide Footnote

The impasse over Nagorno-Karabakh’s status has precluded discussion of just about anything else. Both sides fear compromising on issues ranging from broader economic and humanitarian cooperation to short-term confidence-building measures for fear of undermining their positions on status.[fn]Crisis Group interviews, Azerbaijani officials, Baku, March 2019; Armenian and de facto officials, December 2017, February and October 2019.Hide Footnote

There are, however, small signs of movement. In contrast to its predecessors, the new Armenian administration expresses more interest in finding a compromise to resolve the conflict. For the first time, Yerevan speaks publicly about its readiness to take into account the interests of people living in not only Armenia and Nagorno-Karabakh, but also Azerbaijan.[fn]According to Pashinyan, “[A]ny solution to the Karabakh issue should be equally acceptable to the people of Armenia, the people of Artsakh [Nagorno-Karabakh] and the people of Azerbaijan”. Quoted in “Nikol Pashinyan, Bako Sahakyan co-chair joint meeting of Security Councils of Armenia and Artsakh”, official website of the Security Council of Armenia, 12 March 2019. All Armenian governments since the 1992-1994 war have declined to recognise Nagorno-Karabakh’s independence, which they have claimed signals their readiness to find a compromise with Azerbaijan on the breakaway region. Crisis Group interviews, former and current officials, Yerevan, March-July 2018, October 2019.Hide Footnote Azerbaijanis increasingly discuss what a peace process and Nagorno-Karabakh self-rule might entail, with some analysts and officials thinking creatively about possible next steps on the main status-related issues in the Madrid principles: interim status, the referendum and final status.[fn]Crisis Group interviews, Azerbaijani officials, experts and civil society members, Baku, May 2018, March 2019, November 2019.Hide Footnote Baku has made no proposal to its Armenian counterparts and the ideas circulating, which do not envisage the region’s independence, remain far from anything to which Yerevan, let alone Stepanakert, will agree. Still, they can be seen as suggesting a desire to talk rather than fight.

If Baku turns such ideas into official proposals and Yerevan is ready to make good on its expressed desire to find a peaceful solution acceptable to Azerbaijan and reciprocate with its own suggestions that factor in Stepanakert’s views, the resulting conversation in theory could help both sides better understand one another. Such discussion will not resolve the question of status any time soon, but it can perhaps help identify potential areas of cooperation.

A. “Interim Status”

President Aliyev repeatedly argues that “the people and state of Azerbaijan will never allow the creation of a second Armenian state on Azerbaijan’s historical lands”.[fn]“Azerbaijan won’t allow creation of second Armenian state on its historical lands – Ilham Aliyev”, Azvision, 31 October 2017. For a similar presidential statement, see Ilham Aliyev, speech at the official reception on Republic Day, cited in the official website of the President of the Republic of Azerbaijan, 27 May 2014.Hide Footnote This attitude is shared among Azerbaijani public figures and opposition groups.[fn]Crisis Group interviews, current and former officials, opposition members, Baku, May 2018- February 2019.Hide Footnote But something officials term “high-level autonomy” is, they say, palatable. This would grant Nagorno-Karabakh substantial self-rule within Azerbaijan.

In discussions of “interim status”, Azerbaijanis have mentioned retaining current de facto governing structures, but as part of Azerbaijan.[fn]According to Taleh Ziyadov, Azerbaijan could temporarily recognise the de facto governing arrangements in Nagorno-Karabakh until its final status is determined. In exchange, Armenia should return Azerbaijani territories adjacent to Nagorno-Karabakh. See Taleh Ziyadov, “Nagorno-Karabakh Negotiations: Though the Prism of a Multi-Issue Bargaining Model”, International Negotiation, vol. 15, no. 1 (March 2010), p. 107.Hide Footnote To sweeten the deal, they suggest Nagorno-Karabakh could have the authority to conduct a limited foreign policy on an agreed set of issues, which it cannot do today as an entity not recognised by any nation-state. The autonomous Nagorno-Karabakh region would extend only to the region’s Soviet borders. Indeed, in return for Nagorno-Karabakh’s interim autonomy, Baku would expect Yerevan and Stepanakert to give up claims or control over the adjacent territories.

Azerbaijani officials see such a formula as a substantial concession, but one they say they would be willing to pursue if Yerevan and Stepanakert agree.[fn]Crisis Group interviews, senior officials, Baku, May-June 2018, March 2019.Hide Footnote “This is the greatest compromise Azerbaijan can offer”, one said.[fn]Crisis Group interview, senior official, Baku, June 2018.Hide Footnote They also feel it could offer many mutual advantages, including contacts between the two societies and the return of IDPs. “Communication on demining will open up, investment to the [Nagorno-Karabakh] region will be launched”, one senior official said.[fn]Crisis Group interview, official dealing with the Nagorno-Karabakh conflict, Baku, May 2018.Hide Footnote Azerbaijani proponents argue that this would help both societies overcome their prejudices and smooth the way for resolving the question of Nagorno-Karabakh’s final status.[fn]Crisis Group interviews, senior officials, Baku, May-June 2018, March 2019.Hide Footnote

A proposal of interim autonomy along these lines would be rejected out of hand by Yerevan and Stepanakert.[fn]Crisis Group interview, de facto official, Stepanakert, November 2019.Hide Footnote Even if the move proposes no real change in the self-governance and life of the de facto entity, and is billed as “interim autonomy”, many Armenians will suspect that it is an Azerbaijani attempt to attain its goals in the “interim” only then to cement them permanently. Even were Yerevan somehow to accept the idea of Azerbaijan granting Nagorno-Karabakh an indefinite “interim status” in exchange for the return of adjacent territories (another step that Armenians presently adamantly refuse), many Armenians would want additional security arrangements to protect settlers given the likely inflow of Azerbaijani IDPs to those areas.[fn]Crisis Group interviews, officials of Armenia and de facto Nagorno-Karabakh, Yerevan, October 2019.Hide Footnote Armenian officials have avoided offering “interim status” proposals of their own, for fear of undermining their position on Nagorno-Karabakh’s independence.

B. Referendum on Final Status

Baku accepts that a referendum on the future status of Nagorno-Karabakh would be legally binding. But it believes that the timing of the referendum and the question of voter eligibility require further negotiation. Otherwise, according to one former senior Azerbaijani official, “Armenians would vote for independence and will get it, and Azerbaijan will have to agree”.[fn]Crisis Group interview, former senior official, Baku, September 2018.Hide Footnote

Among the alternatives Baku has considered is the prospect of two separate votes – one for the Armenian majority, another for Azerbaijani IDPs, such that both groups must agree in order for the vote to be binding (this can be thought of as the Cyprus model).[fn]Crisis Group interview, senior official, Baku, September 2018. In 2004, parallel referendums on a reunification plan were held in the Republic of Cyprus and the breakaway Turkish Republic of Northern Cyprus. Both communities had to approve the plan on offer for it to be implemented. In the event, Turkish Republic of Northern Cyprus voters approved the plan, but Republic of Cyprus voters rejected it. Thus, the plan was rejected. For more see Crisis Group Europe Report N°171, The Cyprus Stalemate: What Next?, 8 March 2006.Hide Footnote Baku has helped foster institutions for Nagorno-Karabakh Azerbaijanis in part so they are organised to participate in such a vote – or, indeed, in negotiations – in the future. One such institution is the Azerbaijani Community of the Nagorno-Karabakh Region, created in 2006 by a group of IDPs from Nagorno-Karabakh in the form of a public union. Most members are from Shusha, where over 20,000 ethnic Azerbaijanis lived before the war.[fn]Before the 1992-1994 war, the entire population of the Shusha city was 23,156: 92 per cent of residents were ethnic Azerbaijanis and around 7 per cent were ethnic Armenians. The total pre-war ethnic Azerbaijani population of NKAO was around 40,000, while Armenians comprised 150,000. See “Soviet Census 1989”, op. cit.Hide Footnote Community leaders say that the population of Azerbaijanis displaced from the Soviet-era Nagorno-Karabakh Autonomous Oblast has since grown to over 60,000, roughly a third of Nagorno-Karabakh’s current population which is estimated at 150,000.[fn]Various sources cite different figures for total numbers of IDPs both at the time of displacement and in the community today. These figures are difficult to verify and remain debated. 2009 Census in the Republic of Azerbaijan, Azerbaijan State Statistics Committee, 1st Volume, Baku, 2010.Crisis Group interviews, Azerbaijani community leaders, Baku, March 2019.Hide Footnote

The Cyprus model – in effect, giving a veto to displaced Azerbaijanis – is almost certain to be rejected out of hand by Yerevan and Stepanakert. Armenia endorsed the six elements of the OSCE Madrid Principles because it saw in them a clear prospect for Nagorno-Karabakh’s independence through a public vote recognised by Azerbaijan.[fn]Crisis Group interview, Armenian historian, Yerevan, December 2017.Hide Footnote In the words of a senior Armenian diplomat, the idea of a referendum “is a good face-saving tool for Azerbaijani leadership that otherwise cannot declare Nagorno-Karabakh ‘independent’”.[fn]Crisis Group interview, Armenian official, Yerevan, April 2018.Hide Footnote Some Armenian officials and politicians are ready to discuss voting rights for Azerbaijani IDPs, though as part of a single vote rather than having their own separate vote (as in the Cyprus model).[fn]Crisis Group interviews, officials, parliamentarians, politicians, Yerevan, December 2017, April 2018, June and November 2019.Hide Footnote As the number of IDPs is insufficient to make Azerbaijanis the majority in the region, letting them vote would not block independence.

C. Final Status

For final as for interim status, Azerbaijan may accept considerable autonomy for the region but not independence. In October 2016, President Aliyev surprised many when he referred to “an autonomous republic” of Nagorno-Karabakh in an interview.[fn]“Алиев: Нагорный Карабах может стать автономной республикой” [Aliyev: Nagorno-Karabakh can become an autonomous republic], Sputnik, October 2016.Hide Footnote No Azerbaijani leader had ever used the word “republic” in this context, always referring to it as a “region”. Almost immediately, then OSCE Minsk Group U.S. Co-Chair James Warlick lauded the statement and welcomed the president’s decision to start “discussions on status”.[fn]“Aliyev’s remarks on Karabakh’s status must be discussed: OSCE envoy”, Panarmenian.net, 26 October 2016.Hide Footnote

Some Azerbaijani politicians advocate skipping “interim status” and a referendum and negotiating Nagorno-Karabakh’s final status from the outset.[fn]Crisis Group’s interviews, Azerbaijani politicians, Baku, May-June 2018.Hide Footnote Their stated logic is that the effort that would go in to negotiating a referendum and interim status would be better spent sorting out a more lasting way forward. Although jettisoning the OSCE’s six principles implies a difficult negotiation in its own right, the same questions and issues arise in discussions of interim status suggesting it is not impossible that such conversations could lead to a final status resolution that avoids the intermediate step.

While no Azerbaijani official has publicly offered details of how Nagorno-Karabakh’s self-rule under Azerbaijan would work, in interviews with Crisis Group several discussed models that involve minimal subordination to Baku. These tend to envision a future in which Nagorno-Karabakh authorities:

  • Can reject decrees or laws from Baku related to self-governance in the region;
  • Enjoy considerable self-governance, including in educational and cultural policy, public health, some branches of the economy, law enforcement and postal services, among others;
  • Are subject to Azerbaijan’s judicial and customs systems;
  • Can establish economic representation in foreign countries;
  • Have a role in formulating foreign and security policies (but no veto over Azerbaijan’s policies in those areas);
  • Maintain a demilitarised zone with no armed forces inside the region.

At the same time, Baku would want to ensure that returning Azerbaijani IDPs are well represented in Nagorno-Karabakh’s governance structures, including as elected and appointed officials.[fn]The list is prepared based on Crisis Group interviews with Azerbaijani senior officials, Baku, May-June 2018.Hide Footnote

These options draw substantially on the experience of European countries that resolved territorial disputes without changing their borders. Many in Baku point to the Åland Islands, Northern Ireland and South Tyrol as examples:

  • The Åland Islands have a Swedish-speaking majority but are part of Finland. They are demilitarised, with self-rule, their own police force, a flag and other attributes of a sovereign entity. International treaties signed by Finland have to be ratified by the Åland parliament to have legal force on the islands.
  • Northern Ireland is an appealing model to many in Baku because the territory, like Nagorno-Karabakh, fought a war for independence. In addition to its own self-rule and distinctive state attributes, Belfast controls an independent judiciary, but the highest court of appeal remains the UK Supreme Court.[fn]Northern Ireland Law: Legal System – An Introduction to Northern Ireland Law”, Oxford LibGuides.Hide Footnote It also has a sovereign parliament, as well as representation in the British parliament and central government in London.
  • South Tyrol in the north of Italy has three official languages, reflecting a diverse local population. Along with executive and legislative sovereignty, its local government enjoys fiscal independence.

None of these models is directly comparable to Nagorno-Karabakh. The Åland Islands, for example, enjoy more authority than Baku wants to grant Nagorno-Karabakh. Moreover, neither Europe’s relatively long democratic history nor the supranational umbrella of the EU is present in the South Caucasus.[fn]“Azerbaijan is not Finland and Armenians are not Swedes”, Aravot, October 2014; Crisis Group interview, foreign diplomat, Baku, March 2019.Hide Footnote A Nagorno-Karabakh solution would also have to align with the region’s specific needs. Nevertheless, “although not all features are relevant to Nagorno-Karabakh, learning about these models should provide a reference point”, suggested an Azerbaijani analyst.[fn]Crisis Group interview, analyst, Baku, April 2019.Hide Footnote

The principal problem, however, is that even if Baku is open to granting Nagorno-Karabakh considerable autonomy, anything short of independence is unacceptable to Yerevan and Stepanakert. “There will be no return to the [early] 1990s”, a senior Armenian official told Crisis Group. The Armenian side rejects any plan in which Nagorno-Karabakh returns to Baku’s direct control, whether in the short or long term.[fn]Crisis Group interview, senior Armenian official, February 2019.Hide Footnote Stepanakert says it has built a functional entity whose economy is growing despite political and legal pressure from Baku.[fn]Crisis Group, de facto senior official, Yerevan, October 2019.Hide Footnote

That said, the fact that Baku is putting ideas forward could at least create space for discussions on governance and security in Nagorno-Karabakh. Baku should formalise proposals that delineate how autonomy can guarantee the rights and meet the needs of both local Armenians and displaced Azerbaijanis who may seek to return. While this will not be enough to convince Yerevan or Stepanakert, it might at a minimum start a conversation.

D. Talks on Status

While the two sides are far from one another on Nagorno-Karabakh’s interim status, a referendum and its final status, discussions of these issues could still be beneficial. The converse is almost certainly true: in the absence of such talks, the gulf separating the two sides and societies on each issue is likely to widen. These conversations could begin to lay the groundwork for eventual compromise, however difficult it is to envisage today.

Putting out concrete options on interim and final status might be useful. Baku should convert some of the ideas circulating among official and experts into formal proposals. For its part, the Armenian leadership and the de facto authorities in Nagorno-Karabakh could demonstrate their interest in compromise by engaging constructively and offering their own ideas. Insofar as Azerbaijani proposals underestimate the true extent of self-rule and institutionalisation of the de facto authorities in Nagorno-Karabakh today, Yerevan and Stepanakert can offer Azerbaijan a more accurate picture, which might lead to adjustments at least in Azerbaijani perspectives. Armenian suggestions are unlikely to be initially received in Baku any more warmly than Azerbaijani ideas are in Yerevan or Stepanakert, but they might similarly help set a starting point for talks.

The enormous sensitivity of Nagorno-Karabakh’s independence demand means that any engagement by the two parties on the issue is best undertaken initially through semi-formal or informal channels, lest the publicity of formal talks lead to greater acrimony. All recent attempts to start formal discussions have deepened confrontation between Yerevan and Baku. None has led to serious debate on any of the main issues.

A final challenge is Stepanakert’s role in talks. Since spring 2018, the new Armenian government has demanded that the de facto leadership join. Azerbaijan has always disputed the de facto leadership’s participation. In any case, while Stepanakert was part of 1994 and 1995 ceasefire agreements, it lost its independent negotiating role when its first president, Robert Kocharyan, became Armenian president in 1998.[fn]See Lévon Ter-Petrossian, Armenia’s Future, Relations with Turkey and the Karabagh Conflict (New York, 2018), p. 118.Hide Footnote Before 2018, there was little question that Yerevan represented Stepanakert’s interests, as successive Armenian leaders had personal ties to the region and the 1992-1994 war. Pashinyan, in contrast, has no such connection. Baku has rejected calls for Stepanakert to return to the table, arguing that Nagorno-Karabakh is under Armenian occupation and that for its leadership to participate would imply the de facto authorities represent an independent entity, potentially strengthening their statehood claim.[fn]Crisis Group interview, senior Azerbaijani official, Baku, March 2019.Hide Footnote Baku countered that if Stepanakert joins talks, so should Azerbaijanis displaced from the region.[fn]“Nikol Pashinyan, Bako Sahakyan co-chair joint meeting of Security Councils of Armenia and Artsakh”, op. cit. Baku references the 1992 Commission on Security and Cooperation in Europe Helsinki meeting and the definition of the parties of the Nagorno-Karabakh conflict: “elected and other representatives of Nagorno-Karabakh will be invited to the [Minsk] Conference as interested parties”, which it sees as an acknowledgement that the Nagorno-Karabakh Azerbaijanis have equal rights to the region’s Armenians in the negotiations process. The Armenian side does not agree and references other documents from the past, including ceasefire document, which contains signatures of representatives from Baku, Yerevan and Stepanakert.Hide Footnote Both Yerevan and Stepanakert reject that idea, arguing that Baku represents the IDPs.[fn]Crisis Group interviews, officials and politicians, Yerevan and Stepanakert, April, June and November 2018, April and October 2019.Hide Footnote

Disagreement over Stepanakert’s role strengthens the case for informal talks, at least as a starting point. One option could be discussions among Baku, Yerevan and Stepanakert, initially including only independent experts, former officials or civil society groups endorsed by the leaderships in the three places, potentially with lower level or no officials participating from the outset. To meet Baku’s demand that the concerns of Azerbaijani IDPs be heard, representatives from this community could be included as part of the delegation endorsed by Azerbaijan. Such a dialogue would not replace direct talks between the sides, but take place in parallel.

Both sides ought to agree on two key principles: first, to keep talks on status separate from those on adjacent territories and international missions; secondly, that any agreement on steps prior to final status determination will be without prejudice to that determination and would be revisited if incompatible with that decision once made. Indeed, in the past, fear of setting precedents that could limit manoeuvre on final status has hindered conversation on other topics.

An Armenian villager shows Crisis Group's expert Olesya Vartanyan the positions of the Armenian and Azerbaijani militaries. CRISISGROUP/Jorge Gutierrez Lucena

V. Conclusion

The nearly three decades since the 1992-1994 war have made peace harder. Antagonism and distrust have grown. The two societies have less interaction, even as the line of conflict is more militarised. But while the chasm between them is huge, the recent relative thaw in relations between Baku and Yerevan offers a modicum of hope and a window of opportunity.

Armenia and Azerbaijan ought to take advantage of this situation by initiating direct talks about the issues underpinning their standoff: the adjacent territories’ fate, a potential role for international peacekeepers or monitors, and Nagorno-Karabakh’s status. On the first issue, a preliminary agreement could trade a temporary freeze on new settlement construction for Azerbaijan’s pledge to pause taking legal action or imposing new sanctions. For the second, the HLPG can be tasked with, and provided the access to carry out, a renewed assessment of peacekeeping or monitoring options under various contingencies. This could be coupled with a compromise agreement to expand the OSCE’s Office of the PRCiO. These steps might build some trust and enable further discussions. The distance between the parties on the third issue – status (final or interim) and how a referendum might be organised – means that even conversation on these matters would be a radical step and would likely need to take place initially through semi-formal or informal channels.

There are inevitable pitfalls to reopening dialogue. Discussions when views are so opposed could fuel anger, between the two governments and among the broader public, especially if information is not carefully managed. Yet without talks, opinion in Armenia and Azerbaijan will likely continue to drift further apart. Failure to reset the peace process could also reverse progress that has been made on the front lines. Direct talks could minimise risks of a new war and rekindle a measure of hope in resolving the Nagorno-Karabakh conflict. The last several months of calm are a significant improvement over years of flare-ups that risked triggering wider violence. If the two sides can exploit this interlude, they might be able to create a new norm of engagement that helps them at least discuss some of their differences. Over time this might create openings for the broader and lasting settlement that has been so elusive for so long.

Baku/Yerevan/Stepanakert/Tbilisi/Brussels, 20 December 2019

Our newest Visual Explainer features a uniquely detailed map of the front lines and interactive data charts on reports of casualties and on the use of heavy weaponry, drones and special operations in the conflict zone.

Appendix A: Detailed Map of the Conflict Zone


Appendix B: Map of the Conflict Zone in a Regional Context


Appendix C: Data Collection Methods

Information provided during field trips to Nagorno-Karabakh and the adjacent areas was cross-checked with reports regularly published by the statistical offices of Armenia and the de facto entity along with press reports, local legislation, institutional websites and materials made public by diaspora organisations active in the region.[fn]The de facto Nagorno-Karabakh authorities routinely publish legislative acts on a dedicated portal. As of January 2018, this included about 12,000 acts dating from 1992. In some cases, detailed annexes to budget laws have not been published, but they still often include significant information that corroborates data from other sources.Hide Footnote All these sources have been systematically analysed and compared with each other. As a whole, data published by local authorities are generally internally coherent and primarily produced for local consumption and for administrative purposes. They have also previously been referenced by international organisations, as well as by the Azerbaijani government in official statements.[fn]“Letter dated 15 August 2016 from the Permanent Representative of Azerbaijan to the United Nations addressed to the Secretary-General”, Permanent Representative of Azerbaijan to the UN, 16 August 2016.Hide Footnote

There is no direct correspondence between the current administrative subdivisions of the de facto entity and the formerly used borders of the adjacent territories. The most populated parts correspond with the districts that the de facto authorities call Kashatagh and Shahumyan, respectively located to the west and south west of Nagorno-Karabakh. They bring together large parts of four administrative units known as Kelbajar, Lachin, Zangelan and Kubatly. For these territories, separate statistics are more readily available. Therefore, unless otherwise noted, data from Kashatagh and Shahumyan form the basis of this report.

Demography of the adjacent territories

Around 11.48 per cent of the population of areas now controlled by the de facto Nagorno-Karabakh authorities lives in the adjacent territories. As a whole, the total estimate of about 17,000 people corresponds to the number of registered residents, and only slightly overestimates the number of people actually living there:

  • Almost 15,000 registered in areas west and south west of Nagorno-Karabakh (Kelbajar, Lachin, Kubatly and Zangelan). According to de facto authorities, there are 14,913 people living in this area as of January 2019. Local administrative documents, including the number of pupils recorded in schools in areas south west of the former (Soviet-era) Nagorno-Karabakh Autonomous Oblast, present figures that are about 17 per cent lower than the number of children in population statistics, which may partly be due to underreported emigration from these areas.[fn]
    See, in particular, “The Demographic Handbook of Artsakh 2019”, op. cit.; decision by the de facto authorities on education and schools for the academic year 2018-2019, Decision 1047-N, 24 December 2018.Hide Footnote
  • In addition, up to 2,000 people mostly live in settlements near the former town of Agdam. The 2005 OSCE mission that visited these villages confirmed that about 800-1,000 people were living in areas outside of Nagorno-Karabakh’s borders in Agdam district.[fn]See “Report of the OSCE Fact-Finding Mission (FFM) to the occupied territories of Azerbaijan Surrounding Nagorno-Karabakh (NK)”, 2005.Hide Footnote The number of residents in these settlements has increased substantially in recent years, in part due to external assistance. Diaspora organisations working in these areas report numbers of residents that are significantly higher than those in the 2005 census of the de facto authorities (eg, Nor Maragha was reported to have 516 residents in 2013 compared to 349 in 2005; and Ukhatsar was reported to have 285 residents in 2017 compared to 144 in 2005).[fn]“De Facto and De Jure Population by Administrative Territorial Distribution and Sex”, National Statistical Service of the Republic of Artsakh; see himnadram.org for reports on projects conducted by the All Armenian Fund.Hide Footnote

Chart 1. Natural population increase in adjacent territories 2003-2018
(including only areas west and south west of former NKAO)

Office for statistics of the de facto Nagorno-Karabakh entity.

Due to both natural growth and migration, settlements have disproportionately contributed to the de facto entity’s population growth, accounting for about one third of the total growth recorded in the 2010-2015 period.[fn]Estimate based on accounting for migration and natural growth as reported by the Nagorno-Karabakh statistical office in adjacent territories west and south west of the former (Soviet-era) Nagorno-Karabakh Autonomous Oblast. See “The Regions of NKR in Figures 2010-2016”, op. cit.Hide Footnote Relatively small number of pensioners and recorded deaths, as well as the relatively high number of births and school-age children recorded in these areas, strongly suggest that the settlements are mostly inhabited by working-age people and children. As of 2017, about 8 per cent of residents in adjacent territories south west of Nagorno-Karabakh are pensioners, while this figure for other areas controlled by the de facto authorities is well over 20 per cent.[fn]See “The Regions of NKR in Figures 2010-2016”, op. cit.Hide Footnote Given this demographic trend, natural growth is due to remain remarkably strong, in line with the data recorded for the last decade, with between four and seven times more births than deaths recorded in any given year. Migration patterns to and from adjacent territories have been more volatile, with periods of outbound migration (2005-2009, 2016-2018), as well as periods of inbound migration (2010-2015).

Agriculture in the adjacent territories

As of 2017, the adjacent territories accounted for almost one third of the total agricultural output recorded in the Nagorno-Karabakh’s official statistics, nearly doubling their relative weight in agricultural production in less than a decade. The continuous growth and integration of the settlements is reflected in local legislation. For example, the de facto law with all Nagorno-Karabakh cadastre codes published in 2000 had relatively few locations in the adjacent territories; the correspondent de facto law issued in 2005 added about 40 new cadastre locations, most of them in the adjacent territories.[fn]List of cadastral codes, N 9-01 / 01 (2000), 9 June 2000; List of cadastral codes, N 15-K (2005), 11 March 2005.Hide Footnote

Chart 2. Agricultural output of adjacent territories 2013-2018
(including only areas west and south west of former NKAO)
Inflation adjusted at constant 2010 prices

Source: Office for statistics of the de facto Nagorno-Karabakh entity; World Bank for inflation (FP.CPI.TOTL) and ex-change rate (PA.NUS.FCRF).

Assistance to Nagorno-Karabakh and adjacent territories

Over the last decade, Armenia has directly financed between 50 and 60 per cent of Nagorno-Karabakh’s budget. The amount of the transfer (often referred to in local documents as an “interstate loan”) is recorded both in statistical yearbooks issued by the de facto authorities, as well as in Armenia’s own budget law, and the figures between these sources correspond.[fn]See, for example, “Law on the State Budget of the Republic of Armenia for 2018”, official website of the Government of Armenia, 8 December 2017; “Artsakh in figures 2018”, National Statistical Service of the Republic of Artsakh, 2019.Hide Footnote Armenia provides other forms of assistance, including training and occasional in kind donations ranging from cars to computer servers.[fn]See, for example, Armenia’s government’s Decision on providing property to NKR, Decision 184-A, 26 February 2009; Armenia’s government’s Decision on donation to NKR, Decision 151-A, 15 February 2018.Hide Footnote Partly thanks to this assistance, the budget per capita of de facto authorities is about 20-30 per cent higher than in Armenia.[fn]Based on figures on consolidated budget expenditure and total population included in the respective statistical yearbooks. See “Artsakh in figures 2018”, op. cit.; Statistical Yearbook of Armenia 2018, Statistical Committee of the Republic of Armenia, 2018.Hide Footnote According to de facto budget laws, communities in adjacent territories receive a higher subsidy per capita than other parts of Nagorno-Karabakh.[fn]Details on budget subsidies to communities in Nagorno-Karabakh are included in local budget laws, see “NK budget law for 2018”, Ministry of Finance of the Republic of Artsakh, 21 December 2017.Hide Footnote

Chart 3. Armenia and de facto Nagorno-Karabakh’s budget expenditure, USD per capita 2013-2018

Armenia’s Office for statistics; Office for statistics of the de facto Nagorno-Karabakh entity; World Bank for exchange rate.

No single public document summarises the total amount of assistance Armenian diaspora organisations provide to initiatives in Nagorno-Karabakh. Notable activities in the adjacent territories include the Vardenis-Martakert highway, as about 50 out of its total 115km are located in these areas (out of a total cost of $35 million, the Armenia Fund has reportedly contributed approximately $15 million).[fn]See Hetq.am, “Second Highway Linking Armenia and Artsakh Officially Opens”, Hetq.am, 1 September 2017; “Vardenis-Martakert Highway”, Armenia Fund, September 2017.Hide Footnote The U.S.-based Tufenkian Foundation has conducted a large part of its activities in Nagorno-Karabakh’s adjacent territories, which reportedly include spending $900,000 on building a new village in Jebrail district.[fn]See “Expansion and Development of the Arajamugh Village”, Tufenkian Foundation, 18 November 2017.Hide Footnote According to estimates based on official documentation that U.S.-based non-profits file with tax authorities, the Tufenkian Foundation has spent about $2 million in the adjacent territories between 2003 and 2015, and has facilitated additional assistance from other donors and foundations.[fn]See “Tufenkian Foundation, Inc.”, ProPublica’s Nonprofit Explorer.Hide Footnote