pw78 full final

8/2/2019 PW78 Full Final

1/56

PEACEW RKS

BEyond PoWER ShARing

institutionaloptionsforanafghanpeaceprocess

hams nx a Carle hartzell

[ [


2/56

Aboutthe RepoRt

This report, funded by the Royal Norwegian Ministryof Foreign Affairs, is the product of a joint researchproject of the United States Institute of Peace (USIP),the Peace Research Institute Oslo (PRIO), and theChr. Michelsen Institute (CMI) to identify issues andoptions to support durable peace in Afghanistan.Previous publications stemming from the projectinclude several USIP Peace Briefs and two researchpapers: Hamish Nixon, Achieving Durable Peace:Afghan Perspectives on a Peace Process (Oslo: PRIO,May 2011) and Deedee Derksen, Peace from theBottom-Up? The Afghanistan Peace and ReintegrationProgram (Oslo: PRIO, September 2011).

Aboutthe AuthoRsHamish Nixon is project coordinator of the jointresearch project on durable peace in Afghanistan. Hespent five years in Afghanistan working on governancewith the World Bank and the Afghanistan Research andEvaluation Unit, and has researched peace processesand post-conflict governance in Cambodia, El Salvador,and Afghanistan. Caroline Hartzell was a 201011Jennings Randolph Senior Fellow at the United StatesInstitute of Peace and is professor of political science atGettysburg College. She is researching the effects thatcivil war settlements have on post-conflict economicdevelopment, and is coauthor ofCrafting Peace: Power-Sharing Institutions and the Negotiated Settlementof Civil Wars (University Park, PA: Pennsylvania State

University Press, 2007). The authors would like to thankall those who spoke with them in Afghanistan andcolleagues who provided comments on the paper.

Cover photo by Getty Images

The views expressed in this report arethose of the authors alone. They donot necessarily reflect the views of theUnited States Institute of Peace, and donot represent official positions of theUnited States Government.

Unied Saes Insiue Peace2301 Constitution Ave., NWWashington, D.C. 20037

Phone: 202.457.1700Fax: 202.429.6063E-mail: [email protected]: www.usip.rg

Peaceworks No. 78

First published 2011

2011 by the United States Institute of Peace


3/56

Peaceworks December 2011 no. 78

CONTENTS

[ ay gtitd ttlt t th agh lithuld ivlv t titil gt tgv th pid t th igig p

ttlt, -i, d th ty it pt itituti lit gt. ]

Introduction: Peace Is Possible ... 5

Challenges or a Peace Process: Analyzing the Aghan Conict ... 8

Initiating and Structuring a Negotiation ... 12

ransitional Arrangements and Implementation Challenges ... 24

Beyond Power Sharing: Institutional Arrangements or the Long erm ... 32

Conclusions ... 41


4/56


5/56

3

BEYOND POWER SHARING: INSTITUTIONAL OPTIONS FOR AN AFGHAN PEACE PROCESS

Summary

While the need or a peace process to end the conlict in Aghanistan becomes clearerwith each passing month, there are deep doubts about the chances o achieving a settle-ment. While the challenges are ormidable, they are also potentially surmountable withthe right kind o process and settlement.

All the phases o a peace process are linked. Beginning a process requires more than aconducive military situation; it also calls or potential negotiation scenarios that cangenerate the conidence to take the initial steps. he kind o negotiation that takes place

will inluence the comprehensiveness, quality, and thereore sustainability o theoutcome.

The Problem

he Aghanistan conlict has many overlapping structures and root causes, a largenumber o incoherent actors, and rapidly changing causal and escalating dynamics, eacho which presents challenges or a peace process. Within Aghanistan, the struggle oInternational Security Assistance Force (ISAF) troops and increasingly Aghan securityorces against the insurgency is the dominant lens through which Western actors viewthe conlict.

his dimension is actually one layer over a more complex struggle or inluence amongpolitical-military actions with roots in previous decades o conlict, shaped by ethnic,tribal, regional, and economic actors, and increasingly coupled to a broad-based legiti-macy crisis. he character o the conlict will continue to change as military, political,and economic transition proceeds, and it will increasingly revert to a multipolar struggleamong a dynamic combination o political actions with economic dimensions.

Initiating and Structuring Negotiations

he core players in the conlict are complex and potentially incoherent actors who mayhave diiculty cohering around stable negotiating positions and delivering on agree-ments. Any viable peace process will need to encourage a pro-dialogue orientation withinthe actors, generate increased coherence in political positions, and protect the peaceprocess against splits and violent setbacks.

A key task o the prenegotiation stage will be to overcome lack o conidence among thearmed conlict parties through eective third-party engagement, provide support orparties to be eective negotiators, and develop a detailed procedural and substantivenegotiating ramework.

o manage such a complex, incoherent, and dynamic array o conlict and nonconlictparties, a negotiation will likely require time and division o the agenda into incrementalsteps, as well as multiple tracks or accords to acilitate inclusion o diverse interests on keyissues. Mediators and support rameworks may be strong or weak, and play active or more

passive roles as negotiations progress.

Agreeing Transitional Arrangements

Any negotiated settlement to the Aghan conlict should involve a set o transitionalarrangements to govern the period between the signing o a peace settlement, a cease-ire, and the entry into orce o more permanent institutions or conlict management.

Among other tasks, these arrangements need to overcome mistrust and insecurity amongthe parties, address governance challenges, and deine an interim political order that


6/56

4

PEACEWORKS 78

combines stability and continuity with inclusion and reorm, while providing means omonitoring and veriication.

Beyond Power Sharing in Long-Term Institutions

Long-term institutions or distributing decision-making powers must also orm part o a

political settlement. Yet while enabling governance, institutions must not threaten thesurvival o any o the key parties to the conlict or the peace process may not besupported.

Much o the debate about a peace settlement with insurgents in Aghanistan ocuses onlyon political or territorial power sharing through central or provincial executive positions.However, a wide range o potential measures could create opportunities among theconlicting parties to share inluence, as well as balance that inluence with more rolesor noncombatants, civilian political actors, and vulnerable groups.

Power sharing and reorm are not mutually exclusive approaches to addressing the politi-cal dimensions o the conlict. A combination o power-sharing, power-dividing, power-creating, and power-diusing mechanisms can provide groups within divided societies

with assurances that they will not be permanently excluded rom state power and

resources, while generating more eective and accountable governance and establishingthe oundations or a more capable, accountable, and resilient state.

In Aghanistan, this might include clariying or even redeining the powers o the presi-dent, National Assembly, and the courts, modiying the relationship between the centralgovernment and provincial and district administrations, or creating and diusingdecision-making authority among new or existing institutions over issues such asappointments.


7/56

5


Introduction: Peace Is Possible

Te need or a peace process to end the conict in Aghanistan becomes clearer with each

passing month, just as quickly as hopes or one oten seem to recede. Despite many positive

changes, ten years o deepening international involvement, both military and civilian, have

been accompanied by a pattern o mounting violence. Since 2009 a dramatic escalation by

NAO o both conventional counterinsurgency and special operations has certainly cost theinsurgency lives and territory, and the prospects o an outright aliban victory seem negligible.

At the same time, the aliban movement and its allies have shown resilience and exibility,

presenting a consistent tactical challenge in narrowing areas o the south and broad areas in

the southeast and east, while extending their reach in the north.1 A string o high-prole at-

tacks and assassinations has undermined government and NAO claims o increasing security

and eliminated key government allies, while deepening tensions among political actions with

distinct ethnic overtones, raising the specter o a widening civil conict.2

Te United States and its partners seek to transer the bulk o security responsibilities to

the Aghan government by 2014, and there are concerns about the quality, unity, and sustain-

ment costs o regular Aghan security orces and the eect o prolierating irregular units on

government legitimacy.3 A succession o governance crises has undermined the last two na-tional elections, threatens the continued delivery o international development assistance, and

has paralyzed executive-legislative relations, bringing most avenues o institutionalized politics

to a standstill. Te security transition to 2014 will thus be accompanied by a challenging politi-

cal succession and destabilizing economic conditions as output shrinks by upward o one third

on the back o declining security and aid inows.4

Such circumstances heighten the conclusion, already acknowledged by the United States,

its international partners, and the Aghan government, that a diplomatic solution oers the

best chance to avoid a deepening conict. Te Aghan government has pursued both private

and public outreach to neighboring countries and insurgent groups centered on a seventy-

member High Peace Council appointed in late 2010an approach that encountered a serious

setback with the September 20, 2011, assassination o council Chairman Burhanuddin Rabba-ni. Nevertheless, the United States, the aliban, and more recently the Haqqani network have

all claimed participation in preliminary talks, while contact between Gulbuddin Hekmatyars

Hezb-e Islami and Kabul has long been quite open.5

Yet, while the desirability o a negotiated outcome is increasingly clear, there are deep

doubts about the chances o achieving a settlement. Tere are two important strands to such

objections: (1) that the aliban themselves are uninterested in negotiating and (2) that Paki-

stan will prevent an eective negotiation in order to preserve its inuence over the outcome in

Aghanistan. Te claim that the aliban leadership will never negotiate, either due to absolutist

ideology or the perception that the war can be won through waiting out the U.S. drawdown,

bears little scrutiny. Te existence o a current o pragmatic, politically thinking, pro-talks ali-

ban has been demonstrated through contacts dating rom 200708, their more recent partici-

pation in preliminary talks, and statements including but not limited to Mullah Omars 2011

Eid message that claim the aliban are ocused on Aghan-specic goals and not international

jihad. Tis evidence suggests recognition o the need to share political power with other groups,

and potential moderation o some views on social issues. Evidence rom eld commanders also

suggests some see talks with the leadership as acceptable.6

Admittedly, the relative inuence o this current within the movement at present is un-

clear. However, this question and the spectacular assassination o Chairman Rabbani do


8/56

6

PEACEWORKS 78

6

not indicate the impossibility o negotiation as has been asserted.7 Continued or intensied

hostilities by branches o the insurgency do not in themselves demonstrate an unshakeable

rejection o talks, just as intensied action by NAO or the Aghan government also does

not. In act, the current escalation reects similar patterns in other conicts as parties seek to

leverage their position in advance o a potential negotiation. A study o relationships between

escalation and negotiation in ten conicts published in 2005 nds that decisions to negotiate

ollow a party s escalation o the conict and that these correlations do, indeed, have a causal

eect.8 Consistent with this reading, or example, the head o the madrassa that trained

Mullah Omar has called or continued armed struggle against U.S. orces in Aghanistan

while oering to mediate between them and the aliban.9 A key challenge o any peace pro-

cess will be to enhance the appeal o negotiation among reluctant insurgents.

Te second viewthat without the acquiescence o the Pakistani security establishment

no settlement is possible, and that that condition is unlikely to be ullledis widely held in

Aghanistan and among parts o the U.S. policy community. Much has been made o Paki-

stani eorts to maintain control over nascent negotiations through the arrest o pro-talks

individuals during 2010, and the public reaction by the Karzai government to the Rabbani

assassination reects this line o thinking.10

A 2011 study suggests that Pakistans policy elitesperceive that their interests increasingly lie with a stable and inclusive Aghan government

potentially including aliban representationthat can protect Pakistans interests, chiey by

preventing excessive Indian inuence. In this light, Pakistans intererence in the peace process

can be seen as hedging that results rom skepticism over the chances o a successul negotia-

tion, which in turn stems rom the ambiguity o U.S. policy toward such talks.11

Pakistani hostility, rather than being absolute, is in this sense part o a vicious circle driven

by uncertain progress or prospects or an intra-Aghan process that would accommodate Paki-

stans concerns. It is possible that such progress might motivate gradual adjustments within

Pakistan, converting a vicious circle o hedging to a more virtuous one o increasing prospects

or peace.12 Certainly, establishing the general parameters o a peace process and uture insti-

tutional options as described in this paper, incorporating Pakistani input and broader regionalagreement, could make a contribution to Pakistans condence that a peace process could meet

its concerns. Furthermore, even i one argues such a change in Pakistani security policy is neces-

sary to achieve a settlement in Aghanistan, it is certainly not sucientto make such a settle-

ment stick. Tereore, an intra-Aghan settlement needs to be hammered out, as prominent

advocates, including U.S. secretary o state Hillary Clinton, have noted.13 While this report

ocuses on the intra-Aghan terms o such a settlement, by no means do we consider that

Pakistans engagement with those terms is not an important key to a lasting settlement.

Te idea that the core o this intra-Aghan settlement will involve the sharing o political

power and positions with aliban representatives either at the central or provincial level is both

widespread in public discourse, and the cause o considerable opposition.14 Tis opposition has

two main strands. On the one hand, leaders o Aghan political or military actions that havea history o opposing the aliban, or who represent minority communities that particularly

suered under their rule, warn that inclusion o the aliban in a power-sharing agreement

would ignite renewed actional struggle.15 On the other, a range o orces that can be broadly

dened as civil society, including traditionally oriented tribal leaders, modernizing political

parties, human rights and womens groups, ear the consequences o accommodation with the

aliban or newound democratic institutions and individual rights.16Te worst outcome rom

the point o view o any o these groups is a closed deal dividing inuence between the current


9/56

7


regime and the alibanin the eyes o many precisely the route pursued by Kabul prior to the

Rabbani killing.

While these issues are ormidable, they are also potentially surmountable by the right kind

o process and settlement. Tis report attempts to contribute to the evolution o a peace process

that may be able to conront such objections about an agreement with the aliban. A peace

process requires bothideas

andwill

. O these, will is clearly most important: without the desire

and eorts o leaders and their ollowers to exit conict and seek peace, no quantity o well-

meaning mediators, policy reports, or peacebuilding programs can lead to a durable settlement.

However, ideas are needed to convert that will into a path toward a settlement, and to increase

the chance o that settlement enduring:

Factors important in creating ripe opportunities or con lict resolution are . . . the avail-ability and increasing acceptability o new sets o basic ideas, principles and concepts oraddressing the conlict and, eventually, crat[ing] viable ormulas and resources [or]peace agreements.17

Indeed, rich experience exists internationally to provide suggestions, lessons, and models, as

a recent call rom the Brookings Institution or a more robust political strategy in Aghanistan

observes:

he international community can oer ideas here, based on previous experiences rang-ing rom the termination o civil wars in Central America and Angola and Rwanda tothe war crimes processes in the Balkans and Liberia to the ruth and Reconciliationprocess in South Arica to the mixed indigenous-international arrangements employedin Cambodia.18

However, to have any chance o success, such ideas must respond to Aghanistans unique

conict dynamics. oo oten international actorswhether analysts, policymakers, or imple-

menterstry to apply lessons that derive rom what they know, rather than what is suited

to the context. Military planners may tend to draw on recent experiences in Iraq, UN o-

cials rom major peacebuilding eorts such as those in Bosnia-Herzegovina or East imor,

and diplomats rom their own countrys experience, such as that o the United Kingdom in

Northern Ireland.Tis report takes a dierent approach. Rather than liting wholesale the experience o other

peace processes that writ large may have limited applicability to the complexities o Aghani-

stan, it examines some specic challenges a peace process in Aghanistan will ace, and then

presents theoretical observations and some real world comparative examples that may be appli-

cable to these challenges. Te report does not aim to recommend a complete or a single set o

institutions or an Aghan peace process, but rather to stimulate discussion on how to connect

the particular challenges an Aghan peace process aces with the possibilities or peace that

careul and innovative arrangements can and have produced even in complex and seemingly

intractable conicts.

o do this, the report develops in its next section a conict analysis to clariy some o the

particular challenges that the Aghan conict presents. Following that, the paper discusses thechallenges o initiating and structuring intra-Aghan negotiations to reach a settlement, ocus-

ing on the need or eective engagement by third parties, a support structure, and a negotiating

ramework with provisions or careul inclusion. Next, it discusses the key challenges and unc-

tions o transitional arrangements that will be needed in a settlement during the implementa-

tion o key provisions. Finally, it details longer-term institutions that can dampen ears about

sharing power with the aliban while addressing root causes o the conict. In particular, the

report outlines a range o examples o power-sharing, power-dividing, power-creating, and


10/56

8

PEACEWORKS 78

power-diusing models that go beyond the current ocus on sharing power with insurgents to

integrate elements o much-needed reorm.

Challenges for a Peace Process: Analyzing the Afghan Conflict

It is beyond the scope o this report, and indeed the abilities o its authors, to comprehensively

analyze or summarize the conict in Aghanistan. Nevertheless, there are patterns that present

challenges to any potential peace process discernible in the available literature on the conict,

and in the primary research conducted under the study o which this report orms a part. I

used careully and exibly, ormal conict analysis rameworks can help ocus attention on

some o the challenges that a peace process must ace.

Conict resolution specialists have developed detailed conict-mapping rameworks that

emphasize history, context, and variations in the kinds o actors, the nature o the issues, and

the contrasting belies and orientations that drive a given conict.19 Most rameworks devel-

oped or practitionersinevitably a little reductivetend to stress three main elements:

structures (sometimes reerred to as root causes or conlict issues: what is the conlictabout?including international, regional, historical, and contextual actors);

actors (who are the parties to the conlict?including whether these are primary,secondary, or third parties, their interests and motivations, their relationships,resources, and capacities, and thereore their likely responses to dierent incentives); and

dynamics (triggering or proximate causes o conlict, escalating actors, and changes instructures or issues as conlicts become longer and deeper rooted, or local capacities ormanagement or resolution).20

Te Aghanistan conict exhibits complications in each o these areas: it has many overlap-

ping structures and causes, a large number o incoherent actors, and rapidly changing causal

and escalating dynamics.

Structure: Dimensions o the ConlictIt goes without saying that the Aghan conict actually comprises a number o interacting dis-

putes at international, regional, national, and local levels. One attempt to outline this conict

net describes the international war on terror intertwined with regional disputes between In-

dia and Pakistan, between Pakistan and Aghanistan over the border and with aliban groups

active in both countries, between Saudi Arabia and Iran, and between the United States and

Iran.21 Within Aghanistan, the struggle o International Security Assistance Force (ISAF)

troops and increasingly Aghan security orces against a complex and cross-border insurgency

is the dominant lens through which Western actors view the conict. A well-publicized Sep-

tember 2011 discrepancy in the interpretation o security trends is a telling example o the

dierence o perspective between those who measure progress as a decline in insurgent attacks

and those observing overall violence.22

However, this counterinsurgency conict is one layer over a more complex struggle or

inuence among political-military actions with roots in previous decades o conict, that is in

turn being shaped by ethnic, tribal, regional, and economic actors operating both nationally

and locally. Tere is consequently a ragmentation o understandings o what the Aghanistan

conict is about among dierent actors, or dierent audiences, and in dierent locales. Many

national-level Aghan leaders emphasize Aghanistans victimization in these wider regional

struggles, while internally some ocus more on U.S. and government actions driving the insur-

There is . . . a

ragmentation o

understandings o whatthe Aghanistan conict

is about among

dierent actors, or

dierent audiences, and

in dierent locales.


11/56

9


gency, poor governance and the benets corrupt elites derive rom conict, or historical ethnic

rivalries and patterns o rule.23

Te insurgency does not tend to see or present itsel as an ethnic Pashtun movement and

has achieved modest expansion among other ethnicities primarily through clerical networks

and in remote areas, yet some outside the insurgency view it in ethnic terms.24 Within the

Pashtun sphere, the tribe has a uid, relative, and localized role in the conict, just as it does in

Aghan society more generally.25 While the conict cannot be understood without reerence

to tribal dynamics, these are but one actor or arena among others such as religious and com-

mander networks dating to earlier conicts, competition over opium production and trafck-

ing, and a lack o articulated Pashtun political alternatives.26

In addition to insurgency, regional, and actional dimensions, the Aghan conict increas-

ingly turns on a broad-based legitimacy crisis. Te aliban grew out o and beneted rom

widespread disillusionment with the outcome o short-lived power-sharing agreements among

the mujahedin actions ater 1992 and the resulting civil war and disorder.27 Similarly now,

the conict is or a broad cross-section o stakeholders, a legitimacy crisis stemming rom a

system o power and patronage distribution that has resulted in the capture o the government

by narrow elite interests who enjoy declining legitimacy and seek to mobilize societal tensionssuch as ethnic divisions to maintain their inuence.28 An important structural cause o the

conict thereore remains weak governance, and any peace process will need to address this

issue through reorm as well as power sharing i it is to lead to a durable peace.

Actors: A Cast o Complex and Incoherent Players

Tis multiplicity o conict structures or lenses in turn complicates analysis o the conict

actors. Underlying much conict analysis is an explicit or implicit assumption that actors

in the sense o discrete decision-making unitscan be identied, and their motivations

and resources catalogued.29 For Aghanistan, the number, identity, and nature o the most

relevant actors are matters o interpretation and are contested. As noted earlier, a counter-

insurgency reading o the conict suggests a government and its NAO allies with generallycongruent interests conronted by an insurgency with several components, surrounded by

concentric rings o regional and global players with their own interests. 30 An interpretation

that ocuses on the struggle among political-military actions emphasizes a larger range o

actors engaged in armed and unarmed competition with each other in a violent political

marketplace characterized by neo-patrimonial deal making.31 In turn, analyses o the war

economy emphasize the importance o numerous war proteersso-called malign actor

networksin driving the conict, whether their resources derive rom drugs, or interna-

tional transport and security contracts.32

Te result is a shiting cast o conict actors depending on what aspects o the conict are

seen as most salient. Furthermore, those actors that can be identied exhibit great internal

complexity and incoherence. Much attention in this regard has been paid to the insurgency. Atthe highest level o aggregation, division o the insurgency among three broad networksthe

Quetta Shura aliban, the Haqqanis, and Gulbuddin Hekmatyars Hezb-e Islamiis large-

ly uncontroversial, though some preer to distinguish between a Kandahari and Paktiawal/

Waziristan network o aliban.33 It is airly widely understood that these organizations

have some shared and some distinct objectives and positions, and that a peace process may

need to deal with them separately, exclude one or more o them, and may nd them acting

as spoilers.34


12/56

10

PEACEWORKS 78

Within this macrostructure, there is vibrant debate about the continuity o the aliban

organization, the coherence o its command and control and the impact o accelerated capture

and kill operations, and the likelihood it can be split constructively, or will ragment destruc-

tively, during a peace process.35 Te aliban orm a network o networks that each draw on

dierent group-based motivations. Depending on local dynamics, these groups can be tribal

or subtribal, have roots in jihadi party divisions, or in closer ties o andiwal (allegiance based

on past history o common struggle). Tey may be motivated by grievances against govern-

ment policies or ofcials, revenge or abuses, ideological afnity, the need or support in local

disputes, or desire to commit crimes or extract rents.

However, these disparate building blocks are also mobilized by a central aliban leader-

ship, and bound together through radicalization into an Islamist and increasingly nationalist

ideological rame o reerence.36 Tus,

the system o reerence individual aleban or their leaders allude totribal, nationalist,or Islamistdepends on the circumstances under which a particular decision is takenand on the particular tactical or strategic aim at stake. 37

Te aliban as an organization also eatures a considerable degree o decentralization, de-

spite the totalizing ideological views o the aliban emirate. For example, in general its much-

discussed justice system has not been imposed rom the top down, but rather by establish-

ing aliban courts as available alternatives that adapt to the varying local roles and status o

existing clerical networks.38 While the aliban code o conduct, the layha, establishes a loose

set o centralized policies in military organization and behavior, signicant decision-making

powersor example, over appointment o deputies, dress, non-governmental organization

(NGO) activities, and celebrationshave been decentralized to the provincial level or below.39

In the words o one aliban cleric, Commanders are responsible to Mullah Omar, but have

to respond to the public.40 Tis structure allows a considerable degree o coordination and an

ideological and material inrastructure, while enabling autonomy and resilience.41 However, it

also makes it very difcult to judge the coherence o policy toward a negotiated political settle-

ment, and creates a high probability o splits and independent action by segments whose localconcerns are not met by a peace process: in short, spoilers.

It is not only the insurgency that eatures a complex and potentially incoherent structure.

In act, the Aghan government, the other political actions within the national scene, and even

the international coalition exhibit divisions that prevent eective concerted action in relation

to a peace process. Despite a presidential policy emphasizing reconciliation with the aliban

and a High Peace Council charged with the public pursuit o this policy, interviews with senior

members o both the government and the council suggest both include individuals actively

opposed to negotiations.42 What is sometimes known as the political or legal opposition has

been struggling to cohere around a unied position or to agree to a leadership congura-

tion; describing these actors based on pre-2001 labels such as the Northern Alliance or the

Jamiat-e Islami is increasingly difcult as these are split into camps opposed to or allied toPresident Karzai.43

Despite outwardly parallel mandates to support Aghanistan, the international community

represented by the United States, NAO and its other members, and the United Nations

may be among the least coherent o the groups o actors involved in this conict. One need

look no urther than the diversity o views during development and implementation o the

Aghanistan Peace and Reintegration Program to uncover signicant dierences o analysis

and approach between ISAF, the Aghan government, and the United Nations.44 Even within

The Aghan government,

the other political

actions within the

national scene, and even

the international

coalition exhibit divisions

that prevent eectiveconcerted action

in relation to a

peace process.


13/56

11


the United States government, there is a pattern o divergence between departments such as

State and Deense on Aghan policy as a whole, and a negotiated settlement in particular.45

Tis picture o the core players in the conict is one o complex and potentially incoherent

actors who may have difculty settling on a single set o negotiating positions and delivering

on agreements as a peace process advances. As one comprehensive review o the actors notes,

Low coherence is not good news or the prospects o an accord. Incoherent actors aredi icult and unreliable counterparties in any negotiation. he actors may change coursein midstream, their terms are likely to shit and be retraded, and their commitment toimplementation is always suspect. For these reasons, an Aghan peace process will prob-ably bear little resemblance to the Congress o Vienna, the reaty o Versailles, or the SixParty alks on Korea, all cases in which the participants had pretty clear ideas abouttheir interests, objectives, and limitations.46

A key quality o any viable peace process or Aghanistan will thereore be the ability to

encourage pro-dialogue opinion within actors on all sides, generate increased coherence within

conict actor networks, and protect the peace process against the splits and violent setbacks

that will inevitably arise.

Dynamics: Impact o a Changing Conlict

Finally, an examination o the conict dynamics in Aghanistan illustrates additional chal-

lenges. As described earlier, the conict in Aghanistan is a conuence o intersecting regional

and internal conict structures linking the post-2001 U.S. war on terror with regional ten-

sions, internal historical rivalries, and legacies o past conicts, governance weakness, and war,

aid, and drugs economies. However, the balance o these actors is changing over time, altering

the complexion o the conict as quickly as eorts to resolve it are introduced.

A brie review o the development o the conict ater 2001 illustrates this dynamism. Te

trajectory o aliban resurgence across the country has been examined in some detail.47 While the

top leadership ed to Pakistan, in the south and southeast many mid-level aliban commanders

attempted to melt back to their communities or negotiate a secure exit. In case ater case, the

installation o new local ofcials and the elevation o certain patronage networks ed rivalries thatdrove such men back into the aliban organization.48 In the south, this exploitation o local rival-

ries and events led to widespread disillusionment with the government and oreign orces, giv-

ing the aliban a rank-and-le orce.49 While combinations o actors were at play everywhere,

in some places this process ran more along tribal lines (Kandahar) or between conederacies

o tribes (Uruzgan); in other places ethnicity (Ghazni), political actionalism (Loya Paktia), or

competition over narcotics networks (Helmand) inused the process. Tis reliance on interested

power brokers was combined with a light-ootprint policy that saw oreign troops restricted to

Kabul or engaged in special operations, oten against targets suggested by the same brokers. Te

devastating impact o this lack o a political solution or the aliban and the inadequate provision

o security on the conict has been outlined in considerable detail. 50

When NAO troops were deployed to the south in numbers during 2006, they wereunable to prevent this political process, and some actions o the intensied international en-

gagement, such as the removal o certain regional power brokers or counternarcotics activity,

also exacerbated tensions.51 Te resultant pattern o direct conrontation between a growing

international military orce and the insurgency generated new insurgent strategy and tactics

inuenced by Iraq, as larger scale engagements were supplanted by increased use o suicide

attacks and improvised explosive devices.52 Tus, the post-2006 phase o the conict has been

characterized by escalation on both sides, with ISAF orces approximately quadrupling in


14/56


15/56

13


cations that elements o the insurgency would explore this avenue under certain circumstances.

However, condence on all sides in the possibilities o a negotiated outcome to deliver core

aims or result in sustainable peace is very low, and this is itsel a major barrier to initiation o a

process. As ar as can be discerned publicly, current diplomatic eorts to initiate negotiations

are largely passive, meaning they rely on a presumed continued military stalemate and eorts

to prompt an improved regional situation to alter insurgent willingness to negotiate. Without

a doubt, eorts to eel out the conditions under which Pakistan can be more supportive are im-

portant, and should continue. However, as noted in the introduction, these may also be aided i

they are coupled with concrete possibilities or a negotiation and settlement that reassures the

Pakistani security establishment.

It is oten observed that the willingness to enter into a peace process is related to the

emergence o a mutually hurting stalemate.58 However, or this negative structural condition

to lead to a viable peace process, it should contribute to a broader reevaluation o the armed

groups (and its supporters) chances o achieving key goals through negotiation:

What is more important than external pressures is the armed groups subjective appre-ciation o a negotiated settlement as the irst prizeas something that can actuallydeliver on their bottom line demands.59

Te insurgent organization shits rom a militant position, in which armed conict is the

only option, to a dual strategy open to the opportunities o both military action and negotia-

tions, and eventually to an outright preerence or a negotiated solution. Such a shit typically

has less to do with individual attitudes or splits between hawks and doves than with the

balance o arguments or each approach under the circumstances o the moment.60 In act, the

most eective insurgent interlocutors in such a negotiation are not sot-liners, but rather mili-

tants who have come to view negotiations as a possible route to achieving group aims.61

Beyond the military situation, two important inuences can alter that balance or a non-

state armed group. Te rst is increased condence that entering into a negotiation process

per se will not atally undermine its interests. Te second is the availability o possible out-

comes that can meet enough core goals to be worth exploring. In short, to make the decisionto negotiate seriously, the conict actors may need reassurance both in terms o process and

potential outcomes. A key task o the peace process at the prenegotiation stage will thus be to

overcome this lack o condence among the armed conict parties about a negotiation process.

A well-considered peace process can help generate these reassurances through eective third-

party engagement, support or parties to be more eective negotiators, and development o a

negotiating ramework or road map with both procedural and substantive elements. How-

ever, equally important will be using such a ramework or reassuring other stakeholders that

the resulting process will not be so dominated by armed actors that its outcome will generate

renewed conict or enjoy no legitimacy.

Eective Engagement and Support or an International FacilitatorWhy are the exploratory contacts and increased clarity in the U.S. position in avor o a settle-

ment not producing the required change o orientation on the part o the armed nonstate ac-

tors? As two highly experienced diplomats observe, this lack o progress is caused by

mistrust among all those so engaged, the low level o coherence in the objectives o mosto the players, and the l imited capacity to put together such a complex, multitiered dip-lomatic process. he United States has that capacity, but . . . as one o the major combat-ants, the United States is not well placed to mediate even a procedural accord.62


16/56

14

PEACEWORKS 78

Eective prenegotiation must help address this lack o condence, coherence, and capacity.

What kind o engagement is likely to be constructive in the Aghan conict? First o all it is

important that it be proactive, not based on waiting or insurgents to come to the table:

Successul engagement tends to strengthen the pro-dialogue elements within armedgroups, while political isolation tends to strengthen hardl iners. his suggests that mini-mal levels o engagement need to be the norm, not a concession.63

Knowledgeable analysts o the insurgency describe just such currents o pragmatic pro-

talks and militant aliban, rather than ideological divisions:

here is no organised or recognisable moderate (or any other political) action inthe alebanto counterbalance the religious hardliners. . . . It appears to be moreuseul to dierentiate between currents. . . . On one side, there are pragmatic, politicallythinking, pro-talks aleban who understand that a political solution is desirable but whostill are conservative Islamists. On the other side are those who avour a purely militaryapproach, oten combined with a hypertrophic recourse to terror ist means.64

One actor that works against pro-talks opinion within nonstate armed groups is the like-

lihood that negotiating with state (especially great power) counterparts will leave them at a

disadvantage. Tere are also well-ounded concerns about recognizing violent nonstate groups

as equal negotiating partners, and negotiation structures can vary to reect these concerns.

Inevitably the distribution o power during negotiations is important in generating a decision

to negotiate in the rst place:

parties to a conlictespecially armed groupsare less likely to choose to negotiate ithey consider the process strongly biased against their interests. hey are l ikely to rejectpre-conditions that require them to give up core goals [in] advance or their existingstrategic advantages gained during conlict. . . . Instead they are likely to demand to entertalks on the basis o parity o esteem within the process and demand equal power indecision-making.65

Tis is why considered engagement by a third party can be eective i it provides some

legitimacy and recognition to the nonstate actors, while politically buering state actors rom

the implications o ull recognition o their enemies and their means:

Respect is the basic condition o any negotiation. he opponent must be recognized as aparty with standinga negotiating partner because o its ability to veto any agreementand an actor with identi iable reasons behind its actions. However, respect does not meansympathizing with the terrorists aims and goals or even recognizing their legitimacy.66

Increasingly, there have been high-prole calls or an international acilitatorvia a range

o possible institutional avenues including the United Nations, the Organization o the Islamic

Conerence (OIC), states, private organizations, or high-prole individualsto lead explor-

atory discussions.67 Such third-party engagement is needed not only to coordinate disparate

initiatives toward a coherent track, but also to rebalance the internal relationships o the actors

toward dialogue by reinorcing the sense that they can achieve their aims through negotiation.

Tis will place complex demands on acilitation, including sustaining contact with nonstate

actors that remain proscribed and potentially incoherent, supporting the credibility o the

process despite limited international acceptance o the nonstate actors goals, supporting the

negotiating capacity o the actors, and determining tricky logistics and security arrangements.

Considerable and careully considered support will be needed to carry out this challenging set

o tasks and thus generate condence.

It is likely no one acilitator or organization alone can bring all that is required to the

table(s). Instead o ocusing on the choice o a single mediation arrangement, some combina-

tion o a mediator or acilitator with a group o riends or contact group that brings diverse

resources to the task may be useul, though there are potential pros and cons:


17/56

15


Potential beneits o grouping the external actors in some way include enhancing theleverage o the mediator, raising the visibility o the peace process, preempting rivalmediation initiatives, and preparing or sustained support in implementation. However,groups also have disadvantages. he question o composition is delicate, as smallgroups, although undoubtedly more eective, risk excludingand thus oendingsigniicant potential partners. . . . In regionally intertwined conlicts, or conlicts thattake place in the shadow o a regional power, what to do about the neighbors will be aconstant concern.68

Such a group might consist o states, as with early examples o such mechanisms in the

Central American peace processes or in the standing international conerence suggested in

the Century Foundation report or Aghanistan (see box 1).69 Groups under United Nations

auspices have generally been a mix o permanent members o the Security Council, interested

regional actors, and midsize donor states or helpul xers with experience o the conict that

can bring a range o resources to bear.70 Te mechanism must be able to build the credibility o

the process and increase condence in its lack o bias among all the parties. Tis requirement

is one motivation or the requent mention o the OIC, Saudi Arabia, or other Gul states as

potential mediators or the Islamic context o Aghanistan. In Aghanistan, as was the case in

Central America, it is likely that neighboring countries will need to act as parties to a regional

peace process, rather than members o a third-party support group. With these considerationsin mind, one can picture a range o possible groups o riends rom among one or more inter-

ested midsized Western states, sizeable distant Muslim states, along with Islamic partners who

can reassure the insurgent parties, such as Saudi Arabia, or interested but removed regional

powers, such as urkey or the OIC.

However, such support mechanisms need not be comprised only o states or multilateral

organizations, though these may certainly be important or the heavy liting o logistics, secu-

rity, and implementation guarantees. Recently restored negotiations between the government

o the Philippines and the Moro Islamic Liberation Front (MILF) eature a novel orm o

international contact group that brings together our third-party governments with our in-

terested international nongovernmental organizations (INGOs) with wide expertise in peace-

making and peacebuilding generally, and in the Philippines specically (see box 2).While this example corresponds to an ongoing negotiation, the introduction o NGOs

suggests possibilities or new kinds o support or mediation even in exploratory stages. In

Aghanistan, there might be potential to increase the credibility o the process among the non-

state actors by including nongovernmental or educational institutions with particular Islamic

credentials, as well as organizations specializing in mediation or peacebuilding. Such engage-

ment will need to come through an active third party, and may require a mechanism or diverse

support that can provide legitimacy and credibility to the process, exible negotiation support,

and some heavy liting as well.

Supporting the Parties Negotiating Capacity

Eective third-party engagement can help reassure parties and thus help induce and support

a negotiation. However, governments, armed groups, political parties, and civil-society actors

who have been immersed in a conict environment or an extended period may be ragmented,

lack negotiating skills or clearly developed positions, and be unaccustomed to the give and

take o negotiations. Some groups may need technical support to articulate their negotiation

strategy and to develop skills and condence in their ability to negotiate an agreement. 71

Accordingly, third parties should attempt to determine whether divided groups need help rec-

onciling their positions and developing a common bargaining position beore attempting to

Eective third-partyengagement can help

reassure parties and thus

help induce and support

a negotiation.


18/56

16

PEACEWORKS 78

negotiate with other groups. Such support may increase their condence to pursue a negoti-

ated outcome:

When leaders are more conident in the prospects o attaining their interests throughpoliticalrather than militarymeans and eel themselves able to skillul ly negotiate toachieve their objectives, then a negotiated process become a more attractive option.72

raining and capacity building that enables actors to analyze the sources o conict and

identiy potential solutions, engage in policy ormulation, and negotiate eectively can in-

crease the probability that the negotiation process will be successul. However, groups may

react i they see rival groups receiving support to enable them to play a more eective role,

and thus support needs to be approached with sensitivity by third parties, perhaps by using

one o the models o support groups described earlier. But properly preparing groups or

the possibility that their rivals may also receive support can help to alert them to the role

reciprocity plays in the negotiation process. Tird parties should consider these types o sup-

Box 1. Te Contadora GroupFraming the Principles o Central American Peace

The Contadora group, comprising interested but not directly involved states o Mexico, Venezu-ela, Panama, and Colombia, was ormed in January 1983 to support a peace process in CentralAmerica, where several insurgencies were tied in with regional conicts and Cold War dynamics. Thegroups oreign ministers met with the Central American governments to try to develop a peace planocusing on regional and internal causes o the conicts. Three core elements o the plan, knownas the 21 Objectives, were the termination o oreign intererence and the nonuse o neighbor-

ing territories to support insurgencies; respect or human, political, civil, economic, and social rightsand measures to democratize; and the establishment o an appropriate verifcation and monitoringsystemlater requested rom the United Nations.1

While the Contadora plan in itsel could not bring about an end to the conicts, as conditions didnot allow internal peace talks with insurgents in Nicaragua and El Salvador until later in the decade,these principles ormed the backbone o all subsequent eorts, including those that ormed theSalvadoran accords. Individual states rom the group also continued to provide important assistance,with Spain providing the idea or a joint multiparty monitoring mechanism in El Salvador (COPAZ,discussed later in this report).

1. Dario Moreno, The Struggle or Peace in Latin America (Gainesville, FL: University Press o Florida, 1994), 5560; 21 Objectives Docu-ment (September 9, 1983), reproduced in Jack Child, The Central American Peace Process, 19831991: Sheathing Swords, Building Conf-dence (Boulder, CO: Lynne Rienner Publishers, 1992), 174177.

Box 2. Te Philippines International Contact GroupMixed Composition Support

The 2008 breakdown o peace negotiations between the government o the Philippines and theMILF over provisions or the Autonomous Region o Muslim Mindanao (ARMM) later producedan agreement to create an international contact group empowered to attend and observe thenegotiations, visit and advise the parties to conict . . . and meet with the parties upon request toresolve outstanding issues, under the coordination o the Malaysian acilitator o the negotiations.The agreement specifed a preerence or the inclusion o countries rom the OIC and the EuropeanUnion and allowed or the inclusion and support o international NGOs.1

In addition to the United Kingdom, Turkey, Japan, and Saudi Arabia, the group includes the AsiaFoundation (a regional INGO with broad activities), Conciliation Resources (a peacebuilding resourceand training organization), the Centre or Humanitarian Dialogue (a mediation and mediationsupport organization), and Muhammadiyah (an international Islamic NGO). Beore and ater theormation o the group, these organizations have provided numerous orms o active support and ex-pertise to the peace process locally and nationally. Their inclusion leverages these experiences andresources but also enables, through their local partners, an organic connection between the peaceprocess and the broad scope o civil society in the Philippines involved in peacebuilding.

1. Claudia Homan, Peace Negotiations in the Philippines: The Government, the MILF and International NGOs, Peace Brie, United StatesInstitute o Peace, 2011.


19/56

17


port an investment in the peace process, in later policy processes, and in good politics ater

the settlement.73

One example o multipronged third-party support to build the capacity and coherenceo actors or a negotiation process took place in Mozambique. Although the years since the

signing o the countrys peace accord have not seen the Mozambican National Resistance

(RENAMO) become a particularly eective opposition party, the support the group received

in the run-up to and during the peace negotiations helped to secure its commitment to the

peace process (see box 3).

Another example o a more structured technical support program was the Palestine Nego-

tiation Support Unit (NSU), established in 1999 with mixed results (see box 4).

Establishing a Negotiating Framework

o urther enhance the chances that the parties will accept a negotiation as a viable route

to important preerred outcomes, the goal o third-party engagement and early negotiation

support should be to achieve as comprehensive and explicit a ramework or negotiation as

possible. Ideally, such a ramework will have two broad elements, each contributing to reassur-

ing the parties that negotiation is the way orward:

a proceduralagreement that outlines how negotiations will proceed in terms o mediation,representation, sequencing, and decision making, as well as practical but diicult issueso logistics and security;

Box 3. MozambiqueCapacity Support or Negotiations and ransormation

Mozambiques ten-year civil war ended with a settlement signed by the governing party, the Libera-tion Front o Mozambique (FRELIMO), and RENAMO, in 1992. At the start o the negotiation pro-cess, FRELIMO had signifcant advantages, including a well-established party structure, a tradition ounity, and experience as the government in power. RENAMO, on the other hand, lacked ideologicalcoherence, knowledge o constitutional and electoral processes, and experience carrying out basicadministrative and political tasks. It remained wary during early diplomatic activity prior to ormaltalks beginning in Rome in 1990.

South Arica, a RENAMO ally in the conict, began to consider the groups uture in light o unold-ing political changes in southern Arica, and shited its support away rom military means towardencouraging and assisting the group to ormulate and consolidate its political demands. During thenegotiations, logistical and fnancial assistance provided by the international community enabledRENAMO leaders to participate on an equal ooting.

The United Nations ostered a plan to hold democratic elections in 1994, giving RENAMO twoyears to develop into an opposition party. Ater the signing o Protocol 111 o the General PeaceAccords that dealt with the electoral laws and guarantees o logistical support or RENAMO in thecities, the United Nations set up a trust und to support the rebel groups transormation. Resourc-es in the amount o US$17 million gave RENAMOs leadership a realistic chance o competing inthe multiparty elections and provided its ormer military commanders with salaries, houses, ofces,and vehicles.

Funds rom the international community bolstered Aonso Dhlakamas leadership position, allowingRENAMO to pay o military leaders and other ofcials it could no longer use; maintain the loyaltyand services o selected party leaders; and attract new leaders and activists. Initially, fnancial supportwas disbursed directly to the party leadership or discretionary use. Ater the elections, party undingwas to become contingent upon the partys ability to win ofce.1

1. Giovanni M. Carbone, Emerging Pluralist Politics in Mozambique: The FRELIMO-RENAMO Party System, Working Paper no. 23 (LSECrisis States Programme, 2003), http://eprints.lse.ac.uk/28268/1/WP23GC.pd; Carrie Manning, The Politics o Peace in Mozambique: Post-Confict Democratization, 19922000(Westport, CT: Praeger, 2002); Severine Rugumamu and Osman Glba, Studies in Reconstruction andCapacity Building in Post-Conict Countries in Arica: Some Lessons o Experience rom Mozambique (Arican Capacity Building Founda-tion, 2003); Jeremy Weinstein, Mozambique: A Fading U.N. Success Story, Journal o Democracy13, 1 (2002): 141156.


20/56

18

PEACEWORKS 78

Box 4. Te Palestine Negotiations Support Unit

The Palestine Negotiations Support Unit was established in 1999 through a request rom the Pales-tine Liberation Organization (PLO) to the British government or technical and fnancial support in therun-up to the (eventually ailed) fnal-status negotiations in 20002001. The unit was partially undedby the United Kingdom and later several other European nations with a mandate to provide legal,communications, and policy advice to Palestinian negotiators. The unit was integrated into the PLOsNegotiations Aairs Department in Ramallah and drew on the expertise o Palestinian-American and

other Western-trained lawyers and ofcials.The unit has a legal and policy department that aimed to strengthen and refne existing Palestiniannegotiation positions, develop new positions, and contribute to resumption o permanent statusnegotiations. It provides advice on the permanent-status issues (security, settlements, Jerusalem,reugees, borders, and water), as well as other issues (economic relations, compensation, agriculture,tourism, health, transport, energy, telecommunications, and archaeology). The CommunicationsDepartment publicizes and explains Palestinian negotiation positions.1

The unit was embroiled in controversy in early 2011 ater it was determined three employees werebehind a leak to Al Jazeera o 1,600 key internal documents on Palestinian negotiation strategy overa ten-year period, prompting the resignation o the Palestinian Authoritys chie negotiator. 2

1. The authors are grateul or a suggestion by Michael Keating o the United Nations Mission in Aghanistan (UNAMA) regarding the NSU, per-sonal communication, July 15, 2011. PLO Negotiations Aairs Department Web site, About Us, www.nad-plo.org/etemplate.php?id=182.

2. Seamus Milne and Ian Black, The Story Behind the Palestine Papers, Guardian, January 24, 2011.

a statement o undamental principles that outlines core substantive elements and limits,reassuring the parties that primary or existential interests will not be sacri iced and speci-ying agenda items that need to be negotiated.

A procedural agreement should determine how parties are represented and how many are

included, thus suggesting how power will be structured during negotiations. Decision-making

ormulae, the makeup o delegations, logistical questions, the role and powers o the third

party suggested time rames and pacing, and communications, recording, and condentiality

procedures can all contribute to clarity regarding the direction o the peace process. wo o the

most important procedural issues that will need to be considered are the role o the third party

and the structure o the negotiation. A mediator can play a stronger or weaker, a more or less

active, role in a negotiating process. A strong mediator with solid international backing may beable to commit these resources to pushing a high-prole plan with top leaders, and to mobilize

signicant rewards or coercive provisions. A lower-prole mediator rom a nongovernmental

institution may have more exibility in methods and ormality, perhaps more easily gaining

condence and thus inuence rom knowledge or inormational power.74

Te mediator may play a very active role, suggesting solutions and ormulae and proposing

texts based on consultation with the parties, or a more passive one that ocuses on bringing the

parties together and letting them resolve key questions. A airly active mandate to integrate po-

sitions and suggest ormulae can be an important way to support the balance between nonstate

parties and states by elevating the position o the less legitimate actor while providing a buer

or the governments involved. For example, in El Salvador, Salvadoran government negotiators

tried to push a minimal role or the United Nations mediator and continuous negotiations to

capitalize on the weaknesses o a nonstate insurgent actor that had several component orga-

nizations and lacked coherence, while guerrilla groups sought an active mediation that could

suggest settlement ormulae to rebalance the relationship (see box 5). What kind o role is ap-

propriate will depend on the nature o the mediator and the kind o support system available.

Te structure o substantive talks might consist o direct or shuttle diplomacy, be conduct-

ed in parallel according to theme or sequenced over time, and held in continuous or punctu-

ated sessions. In Aghanistan one might imagine a military-security negotiation ocused on


21/56

19


cease-res or cantonment, uture security arrangements, and verication measures such as

military observation. Tis will primarily involve the United States, the Aghan government

and insurgents, and perhaps the United Nations. A parallel political and social negotiation

might involve wider inclusion o Aghan groups, perhaps with additional mediation rom

respected gures rom within or outside Aghanistan, particularly with legitimate Islamic

credentials and reputation.

However, such an approach may ounder on the inability to separate these issuesa lasting

cease-re is unlikely to be agreed in the absence o political and social agreements. An alter-

native approach would be a series o consecutive negotiations, building upon each other and

perhaps with varying participation, aiming to build momentum by generating a track record

o mutual accommodation. For example, an agreement on allowing humanitarian access and

limited cease-res might come rst to initiate the process. alks on the uture electoral system

or power-sharing institutions and governance o the security orces might lead to a broader but

armed cease-re. Tis progress may be ollowed by nal negotiations over demobilization, the

withdrawal o international orces, and the means or verication o counterterrorism provi-

sions including a clear break between insurgents and Al Qaida militants, as these are likely to

be among the last cards played in a peace process.

A key advantage o such a process is that it can be paced in such a way that there is an

opportunity or shuttling and consensus building not just between but also within incoherent

parties. In South Arica, where the groups goals and command structures were quite clear, the

Box 5. El SalvadorSequential Negotiations, Resilience, and Consensus Building

The Salvadoran peace agreements were reached sequentially, with consecutive agreements buildingon earlier pacts. Ater transormations within both the regime and the insurgent Farabundo Mart Na-tional Liberation Front (FMLN, itsel comprised o fve separate insurgent parties) that enabled bothto envision negotiating their core objectives, talks began in 1989.1 Violent escalation on both sidesollowed, but served to weaken militant elements and heighten momentum or a settlement. Thegovernment wanted minimal UN involvement and continuous negotiations, while the FMLN needed

active UN mediation and long breaks to consult its feld commanders to balance the playing feld.2

The resulting road map signed in Geneva in 1990 specifed alternating continuous and shuttlediplomacy, and set an agenda o uture agreements on the armed orces, human rights, the judicialsystem, the electoral system, constitutional reorm, economic and social issues, and UN verifcation.How to time the beginning o the cease-fre was a key question: it was agreed that each o thesesubstantive issues would be negotiated prior to an armed cease-fre, which would then last untilprovisions or demobilization and reintegration o combatants were agreed.

However, the uture o the Salvadoran Armed Forces was a major point o disagreement, and couldnot be settled frst. Alvaro de Soto, the secretary-generals appointed mediator, passed over thisagenda item and provided text or an accord on human rights that was agreed easily, rebuildingsome momentum. Subsequent active UN promotion o solutions or the security orces achievedseveral interrelated agreements, including purging the military with an ad hoc commission, dis-solving certain special orces battalions and the police, and creating a new National Civilian Police.In a parallel process, constitutional reorms agreed to by the government, insurgents, and civilianpolitical parties included a powerul human rights ombudsperson, a requirement that Supreme Court

justices be approved by two thirds o the National Assembly, and a new Supreme Electoral Tribunalwith multiparty representation in its leadership.

In order to achieve agreement on sequencing its demobilization with these other steps, the FMLNneeded time and opportunity to intensively lobby its own commanders, even bringing them to Mexi-co to observe negotiations.3 Following the resolution o these key issues, a compressed agendaallowed settlement o outstanding points and details quite quickly, and an ad hoc temporary cease-fre became permanent when a comprehensive accord was signed in January 1992.

1. United Nations Department o Public Inormation, The United Nations and El Salvador, 19901995, 11.

2. Alvaro de Soto, Ending Violent Conict in El Salvador, in Herding Cats: Multiparty Mediation in a Complex World, ed. Chester Crocker,Fen Osler Hampson, and Pamela Aall (Washington, DC: United States Institute or Peace Press, 1999).

3. Sir Marrack Goulding, Peacemonger(London: John Murray, 2002), 229230.


22/56

20

PEACEWORKS 78

Arican National Congress still required several months to bring lower-level cadres around to

some aspects o the agreements. In addition, when difculties or setbacks arise, attention can

be shited to dierent agenda items or onto new tracks, as occurred in El Salvador during the

sequential negotiation o six separate accords (see box 5). Tese eatures enhance the resilience

o the peace process compared to a big bang peace conerence to produce a comprehensive

settlement. As also occurred in El Salvador, as momentum builds the conguration o talks can

be adjusted to cover ground more quickly.

A procedural ramework or Aghan negotiations might take a number o orms, but

the need to generate condence among the parties, especially insurgents, suggests a ew

eatures that might be helpul. A mediation mechanism will be needed that can be exible,

that can draw upon wide-ranging resources, including some that coner Islamic legitimacy,

and that can actively promote solutions around sensitive issues like representation o the

parties. A negotiation structure that can gradually generate momentum and provides space

to build coherence within and between the multiple incoherent actors is also needed, sug-

gesting that at least initially a procedural accord might seek to set a sequential and punctu-

ated agenda. O course, practical logistics and security issues will also be crucial, but they

will depend on the mediation and support arrangements discussed in the previous sectionand lie beyond the scope o this report.

Once a procedure is established, one might assume the natural sequence is that parties decide

to negotiate, and then through negotiations determine the nature o a settlement. In act, it is

more likely that urther elaboration o the terms o a settlement will help initiate the process:

experience suggests that parties can only enter into negotiations when they have someidea o the parameters o a settlement. A ramework document outlining these param-eters has oten been an eective element in bringing about a ceaseire and peaceprocess.75

A negotiating ramework may benet rom an agreement or declaration that establishes

some undamental principles o the peace process, lays out key issues or negotiation, and may

even suggest the overarching structure o a peace agreement. Tis process does not need to go

into great detail, but a broad outline can help clariy areas o common interest and key issues ornegotiation, and help reassure stakeholders on some undamental issues. In South Arica a road

map between the government and the Arican National Congress (ANC) developed in a series

o minutes addressing issues such as dening political oenses, releasing prisoners, suspending

armed action by the ANC, leveraging national, regional, and local structures to address situations

o conict, and reaching agreement or the commencement o constitutional exploratory talks.76

In Aghanistan such principles might try to incorporate core concerns o dierent stake-

holders to the conict. For example, goals o the peace process might be stated along the ol-

lowing lines:

establishing the conditions or Aghanistans independence rom any oreign intererenceand the removal o oreign orces o any origin rom its soil;

preventing the use o Aghanistans territory or attacks outside the country or against theinterests o other states;

ending the suering o the Aghan people by achieving a cease-ire, protecting civilians,and establishing means or resolving conlicts through peaceul means;

promoting the development and well-being o Aghanistans people through thestrengthening o national unity and participation o all members o society in nationallie, governance in accordance with Islamic principles, and respect or the rights o men,


23/56

21


women, and children, including access to education, economic development, justice, andprotection rom crime and corruption;

providing or adequate veriication acceptable to all signatories.

Without doubt, even principles such as these may prove difcult, but any agreement, even

i purposely vague, may be o use in generating momentum. Additionally, it may be possible to

establish key structural elements o the agreement by linking these goals together with specicmeans or elements o a procedural ramework in some orm ox will do this when y has done that.

For example, the removal o all oreign troops may be linked to the establishment o an agreed

mechanism or international verication that terrorist groups cannot operate rom Aghanistan

and that all Aghan groups have broken ties with Al Qaida militants.

Te development o such a negotiating ramework may occur through engagement by a

mediator or members o a support group with leaders o the conict parties and other stake-

holders through inormal channels or shuttle diplomacy. However, the process can also be

prompted by unilateral proposals or declarations o basic principles by third parties or even the

primary conict actors. For example, in the Mozambique conict, a negotiating ramework

grew out o unilateral declarations by the FRELIMO o twelve principles or direct dialogue,

answered by an alternative sixteen-point declaration by the opposition RENAMO. Tese

claried both the considerable gaps that needed to be overcome procedurally, while establish-

ing agreement on seeking a peaceul settlement.77

Alternatively, ideas to ll out a negotiating ramework can emerge through track II or track

1.5 eorts that ocus on dialogue among unofcial representatives or mid-level players beore

an ofcial mediation begins, as occurred through the inter-ajik dialogue (see box 6). A track

II process can also contribute to solving a problem that has been prominent in engagement

eorts in Aghanistan, namely identiying who to talk to.

It is a kind o catch-22 that a conict actor may not have enough condence to enter into

negotiations without some assurances about the process and outcomes, yet these will not be

certain until a negotiation takes place. Developing a negotiating ramework provides a bridge

between phases o the peace process, and developing such rameworks can help transorm a

passive strategy or initiating negotiations into a more active one. However, beyond the con-

icting parties, mechanisms in the negotiating ramework or the inclusion and inuence o

other interests are also very important or the viability and durability o the peace process as

it progresses.

The Importance o Broadening Inclusion in the Peace Process

Te preceding discussion has ocused on gaining the consent o the conicting parties to

negotiate. Oten this process is delicate and undertaken through condential channels

ocusing on the leaders or representatives o conict parties beore a ormal process can

begin. Experienced diplomats or conict resolution theorists oten stress the need or con-dentiality during this phase, though some emphasize that a separate public arena still has

a contribution to make by providing ideas and honing understandings o the conict during

prenegotiations.78

However, there are a number o arguments or peace processes to include other actors and

interests, with interrelated normative and pragmatic dimensions. Te normative argument

is that civilians and vulnerable groups are oten the biggest victims o armed conict, and

thereore deserve to have a voice in eorts to resolve it. Leaders and groups that have pursued


24/56

22

PEACEWORKS 78

their aims through orce should not be rewarded with the only seats at the table. Instead the

peace process should try to redress the imbalance between the powerul and the powerlessin conict situations to prevent widespread alienation rom the outcome and bolster the

protection o vulnerable groups such as religious minorities and women, a particular concern

in Aghanistan.79

However, this inclusion should not be considered simply desirable yet optional. Case stud-

ies by practitioners and increasingly systematic studies by researchers are showing that peace is

more durable when peacemaking goes beyond the inclusion o conict elites to provide struc-

tures or civil society and mass inclusion, representation, or consultation.80 Te causes o this

positive impact may include wider acceptance o a settlement because groups eel they were

consulted, a settlement that deals better with root causes o the conict because it addressed a

wider cross section o interests, or better preparation or peaceul politics in the uture by not

only privileging armed groups.81In short, the way that the negotiations take place is important or the quality and durability

o the outcome they are likely to produce:

he complexity o a conlict situation may require a comprehensive response. It mayrequire a negotiation structure capable o addressing a number o interconnected con-licts within the state or region. . . . Substantively, a comprehensive negotiation agendadeals with the multiple causes o conlict and addresses the needs and rights o the widersociety as well as those o the belligerents.82

Te impact o lack o inclusion on the resilience o a peace settlement is a particular con-

cern in situations where the government and the armed groups lack a strong social support

base and thus neither are seen as legitimate representatives o public interests.83 As described

earlier in this report, the Aghan conict consists o several overlapping structures, and the

legitimacy o the conict parties is a particular issue. Aghan politics over the last decade has

been largely rebuilt around the leaders and networks o jihadi parties with ethnic bases, privi-

leging them over other orces that may be legitimate. Te Aghan government is widely seen as

corrupt and captured by narrow interests, the insurgency is considered a tool o oreign inu-

ences, and historical jihadi ethnic leaders are losing credibility even as they stoke ethnic ears to

bolster their positions.84 Just as the aliban do not represent majority Pashtun opinion, a wider

range o social orces exists in mainstream Aghan society, and their exclusion has weakened

the center ground o politics.

Box 6. Te Inter-ajik DialogueHelping to Frame the Peace Process

The inter-Tajik dialogue was initiated in 1993 via the Dartmouth Conerence, a long-standing bilateralU.S.-Russian inormal negotiating mechanism, to see i a group can be ormed rom within the civilconict to design a peace process or their own country. The participants were chosen rom amongthe second or third layer o decision makers in the conict actions. A conict-mapping exercisewas carried out, identiying key problems or a peace process, such as allowing displaced people toreturn home or developing a platorm to represent the dispersed oppositionlater resembled by

the United Tajik Opposition. The dialogue group recorded its deliberations and contributed to theperception that negotiations were viable, meeting six times beore track I negotiations began.

As ofcial UN-mediated negotiations began in April 1994, the dialogue prepared the frst o eigh-teen memoranda that conveyed ideas and suggestions about the peace process, and membersserved on the ofcial delegations. The dialogue had an impact on the measures to monitor imple-mentation o the eventual General Agreement through a multiparty commission. While the dialoguewas very ocused on the conict parties, it continued to meet ater the agreement and implementa-tion concluded, turning its attention to civil society and other issues.1

1. Randa M. Slim and Harold H. Saunders, The Inter-Tajik Dialogue: From Civil War toward Civil Society, in Politics o Compromise: TheTajikistan Peace Process, ed. Kamoludin Abdulaev and Catherine Barnes (London: Conciliation Resources, Accord, 2001), 4447.


25/56

23


I the multilayered structure o the Aghan conict is to be adequately addressed, it is im-

perative that a negotiation process broaden beyond combatant groups or leaders to encompass

diverse representatives o other social groups and orces. Contrast this need with the approach

taken in ormulating the Bonn Process in 2001, since acknowledged by Lakhdar Brahimi

himsel as incomplete:

he group assembled in Bonn did not represent the people o Aghanistan, eitherdirectly or indirectly. he UN veteran and ormer Algerian oreign minister LakhdarBrahimi . . . repeatedly stressed that no one would remember how unrepresentative themeeting had been i the participants managed to ashion a process that would lead to alegitimate and representative government.85

In addition, such inclusion can have the advantage o diluting the impact o allowing

the insurgency representation, and increase the palatability o negotiating with the aliban

or both internal opponents and skeptical international actors. For these reasons, any nego-

tiating ramework should include as a undamental principle the incorporation o a range

o unarmed actors and representatives throughout the peace process.

What should be the specic means or ensuring this kind o inclusion takes place in

Aghanistans peace process? Tere is a range o experience or including noncombatants in

peace processes, and the best combination should be determined by Aghan stakeholders.However, three broad approaches are sometimes described:

representative participation through political parties or other organizations;

consultation , in which parallel public processes in luence and contribute to negotiations;

direct participation, where civil society is able to participate in negotiations directly, eitherat local or national levels.86

o this list can be added debate acilitated by the media, and ratication or legitimation

o negotiated agreements through election or reerenduma more ex post orm o consul-

tation. Each o these modes o participation may have a role to play in Aghanistan, and

Lisa Schirch has developed in some detail examples or consideration in Aghanistan, such

as a civil-society assembly as was created in Guatemala, or local-level direct conict resolu-

tion as in Mali and already ound in Aghanistan through diverse initiatives. 87 Whatever

means are chosen, a ew key issues specic to Aghanistan need to be considered.

First, Aghan civil society must be viewed broadly, going beyond Western-style nongov-

ernmental, womens, or proessional organizations to encompass a wide range o customary,

tribal, and religious networks.88 It should not be expected that this sector can or will speak

with one voice, especially on issues as controversial as an acceptable peace settlement and

the political uture o the country. However, consultation with these social orces in recent

years has repeated a quite supercial and narrow pattern. It has typically involved the con-

vening o large assemblies under the Loya Jirga (Grand Assembly) label to ratiy policies

already determined among key domestic and international actors, and/or the opportunity

to present a list o recommendations by a nominal number o civil-society representatives.

Tis pattern has been repeated in both domestic ora such as the National Consultative

Peace Jirga in June 2010, and international conerences like the London conerence o 2010

and the Bonn international conerence in December 2011. Future arrangements need to go

beyond imagining civil society as a coherent sector and structure mechanisms to represent

the diversity o interests beyond the conict parties and major political actions.

A second issue is how actively civil society should relate to the peace process. Te de-

scribed pattern o consultation has meant that civil society in Aghanistan is cast in a passive or


26/56

24

PEACEWORKS 78

reactive role, having to respond to political steps and developments that are determined with-

out its input. Given the tul progress o engagement with the insurgency, and doubts about

the rapid development o a regional atmosphere more conducive to progress, is there scope or

civil-society actors to take a more active role in raming a potential peace process, as womens

networks did in Liberia in 2003 (see box 7)?

Tis approach implies a considerably more active role or civil society than the current one

o hoping to be recipients o commitments that the conict parties will widen a uture negotia-

tion to noncombatant representatives:

I civil society organizations and a broader proportion o the overall public are sui-ciently prepared to engage in peacemaking, it can both create a climate conducive tonegotiations and help to ensure that the social inrast ructure is developed or their voicesto be heard at ormal peace talks.89

What structures should be adopted to advocate or representation? Should civil society

seek to create broad alliances o noncombatants to advocate or civilian representation in a

peace process, or pursue specic issue-based agendas, or example, in ocusing on issues related

to women specically or victims o violence?

Finally, how can these eorts be supported? Just as initiation o a peace process may be en-

couraged by support to conict parties, it should be asked what can be done or noncombatantsand civil society in their eorts to overcome the disadvantages o their nongovernmental and

pacic status to play meaningul roles. Such support may be material, but it can also be moral

through consistent assertion o the need or broad participation beyond the conict parties,

and institutional through promotion and support o new mechanisms and reorms that oster

ongoing participation.

Progress toward mechanisms or deliberation, debate, and the emergence o an active peace

agenda among civil society should also translate to a broader conception o power sharing than

the division o spoils or inuence among the current government and the insurgents. In this

sense, the shape o the negotiation process is in act intimately connected to the orms o power

sharing that emerge rom it:

he negotiations at the center o most war-to-peace transitions have proound implica-tions or the political settlement and even the state itsel. . . . A process that includesormerly excluded groups can lead to a more inclusive and thereore resilient settlementin the uture.90

Transitional Arrangements and Implementation Challenges

A key dimension o any negotiated settlement to the Aghan conict will be a set o transi-

tional arrangements to govern the period between the signing o a peace settlement, a cease-

re, and the entry into orce o more permanent institutions or the management o conict.

Tese interim arrangements are important because t

pw78 full final

Documents