icon caret Arrow Down Arrow Left Arrow Right Arrow Up Line Camera icon set icon set Ellipsis icon set Facebook Favorite Globe Hamburger List Mail Map Marker Map Microphone Minus PDF Play Print RSS Search Share Trash Crisiswatch Alerts and Trends Box - 1080/761 Copy Twitter Video Camera  copyview Whatsapp Youtube
US Special Representative for Afghanistan Reconciliation Zalmay Khalilzad (L) and Taliban co-founder Mullah Abdul Ghani Baradar (R) shake hands after signing the peace agreement. Fatih Aktas /ANADOLU AGENCY
Briefing 160 / Asia

Twelve Ideas to Make Intra-Afghan Negotiations Work 

Eighteen years after the U.S. war with Afghanistan’s Taliban began, all sides are taking the first formal steps toward a political settlement. From designating a neutral mediator to agreeing on “rules of the road”, Crisis Group lays out twelve prerequisites for keeping the talks going.

What’s new? On 29 February, the U.S. and Taliban signed an agreement on a phased U.S. military drawdown, Taliban guarantees to sever ties with terrorist groups, and swift initiation of peace negotiations among Afghan parties to the war. These intra-Afghan negotiations could commence as soon as 10 March.

Why does it matter? Intra-Afghan negotiations would be the first formal step to politically settle Afghanistan’s conflict since the U.S. toppled the Taliban regime in 2001. The U.S.-Taliban deal sets the stage for those talks, but it does not resolve issues among the Afghan parties that could prevent them from making progress.

What should be done? All parties have crucial preparations to make, both before intra-Afghan negotiations start and during the talks’ early stages. Crisis Group has identified twelve key points that could make the difference between a successful beginning to a peace process and delays or early stagnation.


The U.S. and Taliban broke new ground in Afghanistan’s conflict on 29 February, when U.S. Envoy Zalmay Khalilzad and Taliban chief negotiator Mullah Abdul Ghani Baradar signed an agreement in Doha, Qatar. The terms of the agreement centre on a fourteen-month timeline for a phased drawdown of U.S. military forces in exchange for Taliban pledges to sever ties with terrorist groups and deny them safe haven. Most significantly, this deal is a precursor to the next phase, an Afghan peace process: the Taliban is obliged to take part in intra-Afghan negotiations shortly after the agreement’s signing, possibly as soon as 10 March, a date specified in the text. The U.S.-Taliban accord sets no parameters for these talks, however. With the clock ticking, Crisis Group has pointed to twelve matters on which the parties should urgently come to terms before negotiations begin – lest this historic chance at peace in Afghanistan be lost.

The 29 February agreement is a welcome step toward ending the world’s deadliest conflict. The U.S.-Taliban talks stalled in September 2019, following President Donald Trump’s declaration that the process was “dead”. Contacts quietly restarted the following month, facilitated a high-profile prisoner exchange in November and officially resumed in December. The resumption of talks came with a new demand from Washington: some meaningful form of violence reduction to be carried out before signing an agreement. The Taliban proposed a seven-day “reduction in violence”, the details of which were then fleshed out and eventually implemented, starting on 22 February. This reduction in violence, although falling short of a total ceasefire, entailed a countrywide cessation of offensive operations by the Taliban, the U.S. military and Afghan government forces. All sides largely adhered to it.

The tight timeline for the start of intra-Afghan negotiations is critical to maintain the momentum of the peace process and to avoid the very real possibility of a resurgence in violence that could stem from any delay. This time pressure should push all sides to prepare urgently for the logistics, structure and content of intra-Afghan negotiations, with the support of regional and other governments that have an interest in successful peace talks.

A process as difficult as peace talks aimed at ending decades of war in Afghanistan is unlikely to get off to a productive start without thorough and urgent preparation.

Even with the negotiations possibly only days away, however, there is still much left to be decided and done to prepare for the negotiations: the parties have yet to name a venue for the talks; agree on an agenda (save for the Taliban’s public commitment to discuss a ceasefire “early on”); or designate the members of negotiating teams. Putting together the negotiating team is a problem particularly on the Afghan government’s side, due to a tense standoff among political figures over presidential election results. And after heavy U.S. involvement in the peace process so far, U.S. intentions regarding its role in shaping or participating in the next-stage negotiations are ambiguous – nor is it apparent what sort of U.S. involvement the Afghan negotiating sides would welcome.

A process as difficult as peace talks aimed at ending decades of war in Afghanistan is unlikely to get off to a productive start without thorough and urgent preparation. Crisis Group proposes twelve steps that can be taken to bolster the prospects for sustaining intra-Afghan talks beyond an opening round and eventually producing a political settlement to the conflict:

  • Confirm one location as the venue for talks, with a host that can play an effective facilitating role;
  • Designate a neutral mediator;
  • Decide on the structure of the negotiations;
  • Pre-negotiate the initial agenda;
  • Agree to “rules of the road” for the talks;
  • Identify easily agreed-upon principles early, and build on that foundation;
  • Maintain patience and persistence;
  • Preserve reduced violence;
  • Meet continually;
  • Create a “Friends of the Process” forum;
  • Make technical assistance available;
  • Agree on the talks’ overall objective.

I. Confirm a single location as the venue, with a host that can play an effective facilitating role

Unusually for a major diplomatic event that is expected to commence so soon, there is no confirmed venue for the intra-Afghan negotiations.[fn]For the full text of the U.S.-Taliban agreement, as well as a U.S.-Afghan government Joint Declaration also signed on the same day, see “Agreement for Bringing Peace to Afghanistan”, U.S. State Department, 29 February 2010.Hide Footnote It is not even clear whether talks will take place in one venue or in several, with rotating host governments. The latter approach is unwise, because it could pose several unnecessary challenges. Most important, various host governments could bring different agendas of their own into play or, at the very least, confront the negotiating teams with different styles and expectations of their own involvement in the process. Logistical complications would almost certainly arise, consuming time to sort out varying diplomatic protocols and the mechanics and funding of travel and accommodation. The logistical discontinuities and friction thus introduced into the process could produce delays and unproductive gaps of time between meetings. Moreover, this approach would necessarily break up the negotiations into rounds instead of enabling continuous negotiation, thereby risking loss of momentum.

The parties should agree on a single host country, one that has experience organising and supporting negotiations but no political agenda of its own related to the substance of the talks. An experienced host could formally or informally play a facilitating role, helping keep negotiations on track.

II. Designate a neutral mediator

The U.S. – which, for now, is the primary catalyst of the peace process – should take the lead in urging the designation of a neutral mediator. No other actor has the requisite leverage with the negotiating parties and with other governments that can also press the parties to agree to a mediator. The mediator could be associated with the government that hosts the talks but need not be. The most important qualifications are that the mediator be a person experienced in negotiations, trusted by the negotiating parties and of a stature that enables engagement with the parties and other governments at the highest levels. The two negotiating parties could formally issue the invitation to mediate.

Peace processes require process managers. In this case, with neither of the Afghan sides unequivocally having the upper hand on the battlefield, neither is in a position to claim for itself the leading role in managing the process. Moreover, with the parties’ negotiating positions still quite vague, it is unlikely that talks held solely in a plenary format will be productive. A mediator could assist by keeping the content and progression of the talks focused, by helping negotiate the agenda before or as talks begin, and by helping establish benchmarks for progress. A mediator also could float compromises and alternatives when discussions reach deadlock, as well as conduct go-between “proximity talks” to keep the process moving forward when plenary meetings are unproductive. In addition, a mediator supported by a secretariat composed of neutral technocrats could manage the development and iterations of draft text for any agreements that talks produce.

The UN could be involved at an institutional level in providing the secretariat or other technical functions and/or in backing the appointment of a mediator politically through the Security Council.[fn]See Crisis Group Special Briefing N°2, Seven Opportunities for the UN in 2019-2020, 12 September 2019.Hide Footnote A UN role is not essential but may be helpful in confirming and signalling broad international support for the peace process.

The U.S., as the external actor with the greatest leverage, will need to play a continued role in propelling the peace process but – because it is not neutral in the Afghan conflict – should not serve as the formal mediator. Furthermore, if the U.S. were to serve as either a formal or de facto mediator, its own domestic political considerations might drive the pace and possibly the substance of talks, risking the adoption of expedient solutions rather than the talks being conducted more strictly on the basis of how best to ensure a durable outcome.

III. Decide on the structure of the negotiations

Several questions regarding the structure of the negotiations will need to be addressed early on – ones that again, a neutral mediator or host could help the negotiating parties identify and resolve. Before talks start, a decision will be required as to whether, when the parties meet in plenary format, there will be two negotiating teams – one on the Taliban side and one on the government side – or whether there will be several, with multiple non-Taliban political factions represented separately. The dispute over the outcome of the September 2019 presidential election exacerbates the difficulty of resolving this question. It is unlikely that this particular issue will be resolved without the U.S. – which, as a result of its financial and military support, has the greatest leverage over the Afghan government – arbitrating the result among the current government and its political opposition. If there is a unified delegation sitting across from the Taliban, agreement will be needed not only on its composition but also on how it will make decisions internally.

Another structural issue is whether there should be tiers of negotiators, with a senior level of negotiators and a more technical level (the latter possibly functioning in multiple working groups). These two levels could meet in parallel or sequentially on an issue-by-issue basis. The technical level could be organised in a single working group or in multiple thematic working groups, depending on the number of trusted personnel both sides have available. Including a technical level of negotiators is likely to be useful in working out text for consideration by the senior level and in exploring potential compromises in a somewhat less heated environment.

An additional option to consider is whether there should be an expanded or separate format for negotiations or dialogue that includes civil society representatives. This idea could prove difficult because it is not apparent there would be such representatives associated with the Taliban side.

IV. Pre-negotiate the initial agenda

Beyond the initial agenda, once the talks start, the parties will need to agree on a medium-term agenda and sequencing for issues before they broach the most contentious topics.

The lack as yet of a designated host or mediator, as well as the focus until 29 February on concluding the U.S.-Taliban agreement, has impeded development of an agenda for the talks. Agreement on the agenda before talks start could help prevent (or at least minimise) confusion or contention during the opening phase. It could also lessen the chances of provocation or obstruction, such as highly sensitive topics being raised too quickly. The parties should use the days remaining before the launch of intra-Afghan negotiations to develop the agenda for the opening period of talks. If a host or mediator is quickly identified, either could assist by helping negotiate the agenda through shuttle diplomacy between the two sides. Beyond the initial agenda, once the talks start, the parties will need to agree on a medium-term agenda and sequencing for issues before they broach the most contentious topics.

Afghan political leaders and the Taliban may engage in preliminary discussions on an opening session agenda without the aid of a mediator. Some reports suggested that an Afghan government delegation that visited Doha to discuss prisoner exchanges might also have tried to approach the Taliban on preparatory aspects of intra-Afghan negotiations.[fn]“Six-member govt team en route to Doha to meet Taliban”, TOLO News, 27 February 2020.Hide Footnote Yet the Taliban’s swift denial that it would engage with the Afghan government at this stage is evidence of a mediator’s usefulness in facilitating this step.[fn]J. P. Lawrence, “US set to sign peace deal with Taliban in Doha”, Stars and Stripes, 28 February 2020.

V. Agree to “rules of the road” for the negotiating process

Peace negotiations often commence with agreement, in writing, on “rules of the road”, setting ground rules for how the talks will operate and providing a broader sense of direction. Such rules can include pre-agreed resolution measures for occasions on which certain points on the agenda prove intractable (such as deferring issues at the request of either side without requiring mutual agreement or asking for a mediator’s compromise proposals).   

The rules can help set expectations – for the negotiators as well as for the Afghan public – regarding the pace of talks, including whether the negotiations will be a single, continuous process or segmented into discrete phases. They also can cover communications with the media, understandings regarding confidentiality and, if there is a mediator or secretariat, what their functions will be.

VI. Identify easily agreed-upon principles early, and build on that foundation

A common device in negotiations is to begin with relatively more easily agreed-upon issues and gradually move toward more controversial ones. This approach can help build and sustain momentum. More specifically, peace processes often produce intermediate agreements – or understandings short of firm agreement – which build upon each other as the process progresses. Producing agreement on a set of substantive principles would be a way to begin generating traction, fostering the negotiators’ mutual confidence and reassuring the public that the peace process has potential to produce results. Depending on the talks’ rules of the road, early rounds of negotiation could conclude by issuing statements of agreed principles, incrementally building toward a more comprehensive settlement.

Initial principles in Afghanistan’s case might touch on matters such as anti-corruption, Islamic values in governance and justice, human dignity, and Afghanistan’s sovereignty, independence and territorial integrity.

VII. Maintain patience and persistence

Serious divisions exist among the factions within Afghanistan’s anti-Taliban political sphere. The 2019 presidential election results remain contested by President Ashraf Ghani’s chief opponent, Abdullah Abdullah, and many of his supporters.[fn]See Mujib Mashal, Najim Rahim and Fatima Faizi, “Ghani named Afghan election winner. His opponent claims victory, too”, The New York Times, 18 February 2020.Hide Footnote The government side’s negotiating team potentially could consist of both government representatives and opposition politicians (though, as noted above, the team’s structure is not yet determined), a framework that might make it difficult for the team to present clear, unified negotiating positions.

On the Taliban’s side, there is little evidence that the movement’s leaders (much less its wider membership) have held comprehensive dialogue on critical substantive questions likely to arise during the intra-Afghan talks.

In other words, both parties have a great deal of internal deliberation ahead of them before they can present coherent positions and be in a position to debate with the opposite side. What is more, peace negotiations have sometimes involved actors engaging with their constituents publicly, consulting with civil society and community leaders who may not be directly involved in the talks – a process that can be time-consuming. Leaders on both sides also will need to neutralise disruptions on the part of spoilers, including those pushing for maximalist positions. For these reasons, as well as the inherent difficulty of reaching compromise among antagonists who have been in conflict for decades, intra-Afghan negotiations are unlikely to progress smoothly and speedily. Both the Afghan sides and their external supporters will need to exercise patience and persistence.

VIII. Preserve reduced violence

No matter when intra-Afghan negotiations formally commence, the Afghan government and Taliban should continue to keep violence as low as possible. Ideally, this would be an extension of the same conditions as the seven-day “reduction in violence” that preceded the signing ceremony for the U.S.-Taliban deal, though there is no explicit Taliban agreement to this.[fn]Rebecca Kheel, “Pompeo: Afghanistan ‘reduction of violence is working’”, The Hill, 25 February 2020.Hide Footnote

Requiring a comprehensive ceasefire as a precondition for intra-Afghan talks, however, could cause the process to break apart on the launching pad. Even though there already is an understanding that the subject of a ceasefire will be discussed early on, it may be some time before negotiators reach agreement on a complete nationwide ceasefire.[fn]On the Taliban’s commitment to discuss a ceasefire, see Jonathan Marcus, “On US-Taliban talks: A prelude to all-encompassing Afghan deal?”, BBC, 15 February 2020. The publicly released agreement confirmed this point.Hide Footnote That should not stop both sides from continuing to contain violence and limit offensive operations to a significant extent. U.S. officials have suggested that Washington will retain the right to conduct airstrikes in defence of Afghan government forces, an escalation that could spur Taliban members to reprisal and lead to further escalation.[fn]David Brennan, “U.S.-led forces urge Taliban to honor Afghanistan truce: ‘This opportunity for peace shouldn’t be missed’”, Newsweek, 22 February 2020.Hide Footnote The best way to avoid such a spiral, and the negative impact it could have on intra-Afghan talks, is to prolong and normalise an atmosphere of reduced violence.

IX. Meet continually

Even in the event of major disagreement, provocations or unrelated distractions, the negotiators should strive to meet continually in order to maintain momentum. Specific terms may be outlined in the talks’ rules of the road, to include an option for technical teams to continue meeting even if senior negotiators break for consultation with their respective leaderships.

Each side should commit to the principle of meeting continually even in the event of violence carried out by the other side. Intra-Afghan negotiations will not only be vulnerable to a number of external and internal spoilers, but, as noted above, each side will need to engage in difficult internal debates on the substantive issues. Halting talks altogether for extended periods could sap momentum, feed scepticism on either side about the other’s good faith and provide greater opportunity for spoilers.

X. Create a “Friends of the Process” forum

It is a central tenet of peace negotiation theory, backed by global historical precedent, that external support for negotiations is a critical factor in their successful conclusion.[fn]ACCORD, “Ending war: the need for peace process support strategies”, Conciliation Resources, 2009.Hide Footnote This may be particularly true for Afghanistan, given the involvement of the country’s neighbours and others in over forty years of conflict and the related refugee crises. A forum for regional and other international stakeholders should be established to provide this integral support for intra-Afghan talks.

The talks’ designated mediator, if there is one, or the host could convene this forum. Otherwise, a government the parties designate, the UN or possibly the Organization of Islamic Cooperation could perform the convening function. To fit Afghanistan’s particular context, this forum could consist of two tiers: a smaller group of representatives from regional countries that meets more regularly, and a larger group that includes donor nations providing financial and development support. This format would have the additional benefit of tying successful negotiations to coordinated international commitments for future aid in support of implementation. The mediator and the parties themselves could brief the forum on progress at regular intervals.

For the smaller group, the essential participants could include China, India, Iran, Pakistan, Russia and the U.S. Partly because of their past interference in Afghanistan as well as the tensions between some of them (but tangential to Afghanistan) that play out there, these countries have had and likely will continue to have the greatest external influence over whether a peace process is sustained and produces a durable result. Moreover, the large numbers of Afghan refugees in Iran and Pakistan justify their seats at the table. Afghanistan’s Central Asian neighbours to the north could potentially be included as well. The U.S. will almost certainly resist Iran’s participation, but excluding Tehran would be counterproductive for the reasons mentioned here.

XI. Make technical assistance available

If intra-Afghan negotiations progress, they will eventually come to issues for which technical assistance will be important. Examples of such issues include constitutional reform, distribution of governance authority and security institution reform (including possibilities for disarmament, demobilisation and integration of former fighters into civilian life), as well as particularly sensitive topics such as women’s rights.

Technical experts are familiar with challenges and misconceptions that commonly arise in peace negotiations, and, as with a good mediator, can prove helpful in suggesting compromises and creative alternatives to move talks forward.

The parties themselves will, of course, call upon their own experts. But a mediator or the host government could also organise relevant technical experts and ensure that they are prepared to assist as needed. These may consist of UN specialists, non-governmental organisation personnel and individual subject matter experts, as well as diplomatic staff of the government hosting talks who have relevant experience. Technical experts are familiar with challenges and misconceptions that commonly arise in peace negotiations, and, as with a good mediator, can prove helpful in suggesting compromises and creative alternatives to move talks forward.

XII. Agree on the overall objective of the talks

In addition to the practical measures identified above, it will be important for the negotiators to consider and agree upon the kind of outcome they are aiming to achieve. The parties will need to confront whether the talks are intended to resolve the main substantive grievances underlying the conflict, or whether they are meant to be a way station in a longer-term process of ameliorating those grievances.

The parties will need to grapple with this issue not only conceptually but also in terms of what tangible results they hope to produce. Will the goal of intra-Afghan negotiations be to conclude a comprehensive peace agreement that addresses the major substantive political and security issues? Will it be to conclude a thinner agreement that establishes interim governing arrangements and a process map for resolving the major substantive issues sometime down the road? Or will it be something in between? An overly ambitious goal would risk bogging down the process, but an insufficiently ambitious goal would probably be worse, because it would risk sacrificing an agreement’s durability to expediency in reaching it. An outcome that defers too many contentious issues to a later date, when external pressure and support for the process will probably have waned, could readily collapse.


Intra-Afghan negotiations will face a host of challenges, from the days leading up to their commencement all the way to an eventual political settlement, if the process gets that far. Practical measures such as those suggested here by no means guarantee success, but the failure to organise efficiently the practical aspects of a peace process can squander opportunity or at a minimum invite unnecessary delay and friction between the parties. More broadly, defining the scope of ambition for the intra-Afghan talks will be an early critical challenge. The parties will need to determine what kind of concluding settlement they envision, if the talks are to have a chance of producing it.

Kabul/Washington/Brussels, 2 March 2020