Winning Hearts and Minds? - Tufts - Tufts University

©2011 Feinstein International Center. All Rights Reserved.Fair use of this copyrighted material includes its use for non-commercial educationalpurposes, such as teaching, scholarship, research, criticism, commentary, and newsreporting. Unless otherwise noted, those who wish to reproduce text and image filesfrom this publication for such uses may do so without the Feinstein InternationalCenter’s express permission. However, all commercial use of this material and/orreproduction that alters its meaning or intent, without the express permission of theFeinstein International Center, is prohibited.Feinstein International CenterTufts University200 Boston Ave., Suite 4800Medford, MA 02155USAtel: +1 617.627.3423fax: +1 617.627.3428fic.tufts.edu2Feinstein International Center

AuthorsPaul Fishstein is a visiting Fellow at the Feinstein International Center at TuftsUniversity, Medford, Massachusetts. Andrew Wilder is Director, Afghanistan andPakistan Programs, United States Institute for Peace, Washington, District of Columbiaand former Research Director at the Feinstein International Center.AcknowledgementsThe authors wish to thank the following people and institutions and acknowledge theircontribution to the research:• Research colleagues Ahmad Hakeem (“Shajay”), Sayed Yaseen Naqshpa, AhmadGul, Sayed Yaqeen, Faraidoon Shariq, and Geert Gompelman for their assistance andinsights as well as companionship in the field.• Frances Brown, David Katz, David Mansfield, and Astri Suhrke for their substantivecomments and suggestions on a draft version.• William Thompson and DAI for facilitating field research in Paktia Province,AusAID staff and the UNAMA office in Tarin Kot for facilitating field research inUruzgan Province, and the ACTED office in Maimana for facilitating field researchin Faryab Province.• Robert Grant and the Wilton Park conference center for their joint sponsorship ofthe March 2010 conference on “‘Winning Hearts and Minds’ in Afghanistan:Assessing the Effectiveness of Development Aid in COIN Operations.”• Staff of the Afghanistan Research and Evaluation Unit (AREU) for support duringvisits to Afghanistan.Paul Fishstein also wishes to thank the Carr Center for Human Rights Policy at theHarvard Kennedy School for the fellowship during which much of this work was done.The Balkh, Faryab, and Uruzgan analysis benefitted from historical and politicalbackground overviews produced by Mervyn Patterson and Martine van Bijlert, leadinganalysts of these provinces. The section on the evolution of security-driven aidbenefitted from the very substantive contribution of Dr. Stuart Gordon.Thanks go to Joyce Maxwell for her editorial guidance and for helping to clarifyunclear passages, and to Bridget Snow for her efficient and patient work on theproduction of the final document.Thank youGenerous funding for the research was provided by the Afghanistan Research andEvaluation Unit (AREU),the Australian Agency for International Development(AusAID), the Royal Norwegian Ministry of Foreign Affairs, and the SwedishInternational Development Cooperation Agency (SIDA).Cover photoU.S. military and children at health center, HelmandPhoto by © Kate Holt/Integrated Regional Information Networks (IRIN)

ContentsAcknowledgementsGlossary of non-English words and expressionsAcronyms and abbreviationsviiviiiExecutive Summary 21. Introduction 81.1 Purpose, rationale, and description of study 81.2 Methodology 101.3 Similar and related research in Afghanistan 112. Evolution of security-driven aid 132.1 Current approaches 153. Aid and security in Afghanistan 203.1 Provincial background 203.2 Models used for employing aid in the five provinces 214. Drivers of conflict and insecurity 294.1 Corruption, poor governance, and predatory government 294.2 Ethnic, tribal, and factional issues 314.3 Poverty and unemployment 344.4 International military forces 354.5 Religious extremism 364.6 Conflict over scarce resources 384.7 Pakistan and the other “neighbors” 384.8 Opportunities for insurgents to exploit grievances 405. Perceptions of aid projects 415.1 “Nothing, or not enough, was done” 425.2 Inequitable distribution: “They got more than we did” 425.3 Corruption 445.4 “Wrong kind” of projects 465.5 Poorly implemented (low quality) 475.6 Non-sustainable 505.7 Positive views: Some good news 51

6. The stabilizing and destabilizing effects of aid 546.1 Addressing the wrong drivers of insecurity 576.2 Subverted by insurgents 606.3 Poor quality of implementation 616.4 Destabilizing influences of aid 616.4.1 Corruption 616.4.2 Competition over resources: The war-aid economy and perverse incentives 626.4.3 Reinforcing inequalities and creating perceived winners and losers 646.4.4 Regional disparities 647. Summary of conclusions and recommendations 67Annex A: Research Methodology 72Bibliography 78v

MAPMap credit: Hans C. Ege Wenger, Tufts Universityvi

Glossary of non-English words and expressionsarbab/khanarbakaibakhsheeshjihadjihadikarezkomakkuchiskunjalamadrassamahroummalekmujahidinmullahshariashurasudhtak o tooktalibtanzimtashkiltekadarulamawasitawoleswalwoleswalizulmHead of community or tribeTribal security forces indigenous to the Loya Paktia region. The term has beeninformally adopted to refer to irregular local security forces.Financial gift, usually small, offered as a favor to accomplish a taskHoly war, usually referring to the 1979–92 war against the Soviet occupationCommander or political leader who gained his strength during the jihad years(1979–92)Traditional irrigation system that taps aquifers and brings water to the surfacethrough often-lengthy underground canalsHelp, aid, or assistanceNomadsType of animal feedReligious school or training academyDeprived, left out, often with the implication of being discriminated againstLocal leaderGuerillas who fought in the 1979–92 war against the Soviet occupation (literally,those who fight jihad, or holy war)Religious leaderIslamic lawCouncilUsury, excessive interest ratesA bit of noiseIslamic student (singular of taliban)Organization or political partyApproved staffing pattern or list of sanctioned posts in a government officeContractor, one who does a piece of work for paymentSunni religious scholarsPersonal relationship or connection often used to obtain a favor such as employmentor processing of paperworkDistrict administrator or governor; i.e., one who administers a woleswaliAdministrative division within a provinceCrueltyvii

Acronyms and abbreviationsADFANAANPANSFAREUAusAIDCDCCERPCFWCIMICCOINCSODDRDFIDDODDODDEQUIPEUFATAFCOFICGAOGIRoAIDPIEDIMFISAFMODMRRDNATONGONSPODAOECDPDCPRTAustralian Defense ForceAfghan National ArmyAfghan National PoliceAfghan National Security ForcesAfghanistan Research and Evaluation UnitAustralian Agency for International DevelopmentCommunity Development CouncilCommanders Emergency Response ProgramCash-For-WorkCivil-Military CooperationCounterinsurgencyCentral Statistics OrganizationDisarmament, Demobilization, and ReintegrationDepartment for International Development (UK)Department of Defense (U.S.)Department of Defense DirectiveEducation Quality Improvement ProgramEuropean UnionFederally Administered Tribal AreasForeign and Commonwealth Office (UK)Feinstein International CenterGovernment Accountability Office (U.S.)Government of the Islamic Republic of AfghanistanInternally displaced personImprovised explosive deviceInternational military forcesInternational Security Assistance ForceMinistry of Defense (UK)Ministry of Rural Rehabilitation and DevelopmentNorth Atlantic Treaty OrganizationNon-governmental organizationNational Solidarity ProgramOfficial Development AssistanceOrganization for Economic Co-operation and DevelopmentProvincial Development CommitteeProvincial Reconstruction Teamviii

PsyOpsQDRQIPS/CRSSIDASIGARSTARTUN OCHAUNAMAUNFPAUNTAGUSDAUSGPsychological OperationsQuadrennial Defense ReviewQuick Impact ProjectOffice of the Coordinator for Reconstruction and Stabilization (U.S.)Swedish International Development Cooperation AgencySpecial Inspector General for Afghanistan Reconstruction (U.S.)Stabilisation and Reconstruction Task Force (Canada)United Nations Office of Coordinator for Humanitarian AffairsUnited Nations Assistance Mission for AfghanistanUnited Nations Population FundUnited Nations Transition Assistance GroupU.S. Department of AgricultureUnited States Governmentix

EXECUTIVE SUMMARYPolitical and security objectives have alwaysinfluenced U.S. foreign assistance policies andpriorities. Since 9/11, however, development aidfor countries like Afghanistan, Iraq, and Pakistanhas increasingly and explicitly been militarizedand subsumed into the national security agenda.In the U.S. as well as in other western nations,the re-structuring of aid programs to reflect theprevailing foreign policy agenda of confrontingglobal terrorism has had a major impact ondevelopment strategies, priorities, and structures.The widely held assumption in military andforeign policy circles that development assistanceis an important “soft power” tool to win consentfor the presence of foreign troops in potentiallyhostile areas, and to promote stabilization andsecurity objectives, assumes a relationshipbetween poverty and insecurity that is shared bymany in the development and humanitariancommunity.The assumption that aid projects improvesecurity has had a number of implications for theU.S. and other western donors, including: 1) asharp increase in development assistance; 2) anincreasing percentage of assistance programmedbased on strategic security considerations ratherthan on the basis of poverty and need; and, 3) amuch greater role for the military or combinedcivil-military teams in activities that weretraditionally the preserve of development andhumanitarian organizations. At the same time,civilian agencies, including non-governmentalorganizations, have also been increasinglyenlisted in aid and development projects thathave explicit stabilization objectives.Given how widespread the assumption is, andgiven its major impact on aid andcounterinsurgency policies, there is littleempirical evidence that supports the assumptionthat reconstruction assistance is an effective toolto “win hearts and minds,” and improve securityor stabilization in counterinsurgency contexts.To help address this lack of evidence, theFeinstein International Center (FIC) at TuftsUniversity conducted a comparative study inAfghanistan and the Horn of Africa to examinethe effectiveness of aid projects in promotingsecurity objectives in stabilization andcounterinsurgency contexts.This paper presents a summary of the findingsfrom the Afghanistan study. Research wasconducted in five provinces, three in the southand east (Helmand, Paktia, and Uruzgan) whichwere considered insecure and two in the north(Balkh and Faryab) which were consideredrelatively secure, as well as in Kabul city.Through interviews and focus group discussionswith a range of respondents in key institutionsand in communities, views were elicited on thedrivers of insecurity, characteristics of aidprojects and aid implementers (including themilitary), and effects of aid projects on thepopularity of aid actors and on security.Drivers of insecurityThe study first tried to understand the drivers ofinsecurity in the five provinces in order to beable to assess whether aid projects wereaddressing them. The main reported drivers ofconflict or insecurity were poor governance,corruption, and predatory officials; ethnic, tribal,or factional conflict; poverty and unemployment;behavior of foreign forces (including civiliancasualties, night raids, and disrespect for Afghanculture); competition for scarce resources (e.g.,water, land); criminality and narcotics (andcounter-narcotics); ideology or religiousextremism; and, the geopolitical policies ofPakistan and other regional neighbors. Many ofthese factors are complex, intertwined, andoverlapping, so it was difficult to isolate thestrength and influences of each. Respondentsgave notably different weight to the variousfactors in the different provinces. In the southernand eastern provinces, poor governance andtribal and factional conflicts were given moreweight, while in the northern provinces povertyand unemployment were given more weight. Inthe south and east, the actions of theinternational military were reported to be animportant source of insecurity, whereas in thenorth international military forces weregenerally seen as more of a source of security. Acommon theme that cut across many thematic2Feinstein International Center

and geographical areas was that of injustice,including the perceived injustice that a fewcorrupt officials and powerbrokers werebenefiting disproportionally from internationalassistance at the expense of the majority ofAfghans. Insurgents were described as adept attaking advantage of the opportunities offered bycommunities’ grievances and perceptions ofinjustice.Perceptions of aid projectsThe study looked at whether and how aidprojects addressed the drivers of insecurityidentified by respondents and/or were effective atwinning hearts and minds. The research foundthat development projects, rather than generatinggood will and positive perceptions, wereconsistently described negatively by Afghans.Responses suggested that not only were projectsnot winning people over to the government side,but perceptions of the misuse and abuse of aidresources were in many cases fueling thegrowing distrust of the government, creatingenemies, or at least generating skepticismregarding the role of the government and aidagencies. The chief complaints were that projectswere insufficient, both in terms of quantity andof quality; unevenly distributed geographically,politically, and socially; and, above all, associatedwith extensive corruption, especially those thatinvolved multiple levels of subcontracting.Communities did provide positive views on theNational Solidarity Program (NSP), somesignificant and highly visible infrastructureprojects, and long-serving aid agencies that hadestablished relationships with communities.Stabilizing and destabilizing effects of aidWhile the environments in the five provincesdiffered, a number of consistent observationsemerged concerning the effectiveness of aidprojects in promoting short- and long-termstabilization objectives. First, in some areas aidprojects seemed to have had some short-termpositive security effects at a tactical level,including reported intelligence gathering gainsand some limited force protection benefits forinternational forces. In some cases aid projectsalso helped to facilitate creating relationships, inpart by providing a “platform” or context tolegitimize interaction between international andlocal actors who would otherwise find it difficultto meet. However, despite these limited tacticalbenefits, there was little concrete evidence in anyof the five provinces that aid projects werehaving more strategic level stabilization orsecurity benefits such as winning populationsaway from insurgents, legitimizing thegovernment, or reducing levels of violentconflict.The research actually found more evidence ofthe destabilizing rather than the stabilizingeffects of aid, especially in insecure areas wherethe pressures to spend large amounts of moneyquickly were greatest. The most destabilizingaspect of the war-aid economy was in fuelingmassive corruption that served to delegitimizethe government. Other destabilizing effectsincluded: generating competition and conflictover aid resources, often along factional, tribal orethnic lines; creating perverse incentives tomaintain an insecure environment, as was thecase with security contractors who were reportedto be “creating a problem to solve a problem”;fueling conflicts between communities overlocations of roads and the hiring of laborers; and,causing resentment by reinforcing existinginequalities and further strengthening dominantgroups, often allied with political leaders andregional strongmen, at the expense of others.The research found that while the drivers ofinsecurity and conflict in Afghanistan are variedand complex, the root causes are often politicalin nature, especially in terms of competition forpower and resources between and among ethnic,tribal, and factional groups. Internationalstabilization projects, however, tended to laymore emphasis on socio-economic rather thanpolitical drivers of conflict, and thereforeprimarily focused on addressing issues such asunemployment, illiteracy, lack of social services,and inadequate infrastructure such as roads. As aresult, aid projects were often not addressing themain sources of conflict, and in some cases fueledconflict by distributing resources that rivalgroups then fought over.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 3

Conclusions and recommendationsThe following section highlights the mainconclusions and recommendations of this study,which are largely consistent with findings fromseveral other evaluations and studies looking atthe relationship between aid and security inAfghanistan. There is growing awareness bycivilian and military actors of some of the issuesraised here, and steps have been taken to addresssome of them. However, progress has often beenslow because many of the institutional incentivesfor why aid funds are spent in ways that can beineffective or destabilizing remain unchanged.1. Primacy of political over economicdrivers of conflictIn the more insecure areas the reasons identifiedby interviewees for insecurity and opposition tothe government were related most frequently topolitical issues such as the corrupt and predatorybehavior of government actors. Moststabilization initiatives, however, haveemphasized economic drivers of conflict –focusing on poverty, unemployment, illiteracy,delivery of social services, and building ofinfrastructure. In the less insurgency-affectedareas, where poverty and unemployment weregiven as more important drivers of conflict,well-delivered, conflict-sensitive aidinterventions may have been more effective athelping to consolidate stability than aid ininsecure areas was in reversing instability.The population-centered counterinsurgency(COIN) approach of winning the populationaway from insurgents and over to thegovernment struggled to gain traction in partbecause the government’s leadership neverseemed to share the objective of winning overthe population, and instead often pursued apatronage-based approach to buy the support oflocal strongmen. Furthermore, the U.S. andmany of its NATO/ISAF (International SecurityAssistance Force) allies had contradictorystrategies of simultaneously wanting to provideservices and good government to win over thepopulation, but also supporting local strongmenwhose predatory behavior alienated the localpopulation. Aid delivered by or associated withcorrupt officials or strongmen who were in manycases responsible for alienating people in the firstplace has, not surprisingly, proven to be anineffective way of winning people over to thegovernment. Lack of progress on governance hasnot primarily been due to lack of money, but toa lack of political will or a shared strategy on thepart of the government and the internationalcommunity to push a consistent reform agenda.Recommendations:• Focus more on identifying the drivers ofconflict and alienation, and if these areprimarily political, governance, and rule-oflawrelated, do not assume they can effectivelybe addressed through primarily socioeconomicinterventions.• The international community should take abetter-coordinated and more forceful stand oncertain key issues that would help promotebetter governance (e.g., merit-basedappointments into key national and subnationalpositions, more rigorous anticorruptionmeasures including bettermonitoring of donor expenditures, avoidingalliances with notorious strongmen known forcorrupt and predatory behavior).2. Spending too much too quickly can becounterproductive – less can be morePressure to spend too much money too quickly isnot only wasteful, but undermines both securityand development objectives, especially ininsecure environments with weak institutions.However, powerful career and institutionalincentives often contribute to quantity beingprioritized and rewarded over quality. Theseincentives include the strong bureaucraticimperative to grow budgets as much as possible,and to then spend as much money as quickly aspossible in order to justify further budgetgrowth; for Provincial Reconstruction Teams(PRTs), to demonstrate performance duringshort-term rotations based on the quantity offunds expended rather than on the impact thatthe funded activities have had; for manycontractors and NGOs, to generate overheadfunding for headquarters based on program4Feinstein International Center

udgets spent. The experience of the NSP andsome other development projects suggests that,in terms of development, quality should not besacrificed for the sake of quantity. The researchsuggests that in terms of potential stabilizingbenefits as well as positive developmentoutcomes, the process of development, especiallyin building and sustaining relationships, despitebeing time-consuming, is as if not moreimportant than the product of development.Unfortunately, there are few incentives forspending less money more effectively over time.Discussions with individual field-level actors aswell as senior officials confirm that the problemis often not that we do not know what needs tobe done, but rather that institutional incentivesreward getting and spending money. “Less ismore” can never be a reality when “more ismore” is rewarded.Recommendations:• Provide incentives for quality and impact ofaid spending over quantity. Aid money shouldonly be committed when it can be spent in aneffective and accountable manner.• Address the “use it or lose it” problem,whereby budgets are forfeited if not spent, byallowing unused budget amounts to be rolledover into following years, establishing multiyearpredictable funding, and making moreuse of longer-term trust fund-typemechanisms that could be drawn down basedon need rather than annual budget cycles.These approaches would reduce the currentinstitutional incentives and negative effects ofspending too much too fast, while alsoconveying a sense of long-term commitmentto Afghanistan.3. Insufficient attention has been paid tothe political economy of aid inAfghanistanAn important consequence of the pressure tospend too quickly has been inadequateconsideration of incentive structures facingpolicy makers, donors, implementers, andcommunities. Evidence from this as well asother studies indicates that the way in which aidhas been delivered has contributed to instabilitythrough reinforcing uneven and oppressivepower relationships, favoring or being perceivedto favor one community or individual overothers, and providing a valuable resource foractors to fight over. The most destabilizingaspect of the war-aid economy in Afghanistan,however, has been its role in fueling corruption,which delegitimizes both the government andthe international community. Under the currentstatus quo of weak institutions and insecurity,some powerful actors are doing very well, and sohave little incentive to push for change.Recommendations:• Invest more in understanding the politicaleconomy of aid, including local conflictdynamics, the impact of the war-aid economyon these dynamics, the perceived winners andlosers of aid programs, and the role of theseprograms in legitimizing (or delegitimizing)the government.• Give more attention to understanding theincentive structures of national andinternational civilian and military institutionsin terms of aid delivery, and the impact ofthese incentive structures on the effectivedelivery of development assistance.4. Insecurity rather than security isrewardedBecause the primary objective of post-2001 U.S.aid to Afghanistan has not been development forits own sake but rather the promotion of securityobjectives, funding for insecure areas has takenpriority over secure areas. Therefore, the bulk ofU.S. civilian and military development assistancefunds in Afghanistan have been spent ininsurgency-affected provinces in the south andeast. The last several years have seen an evengreater prioritization of the insecure areas despitethe lack of evidence that the aid funds beingspent are promoting stability or improvingattitudes towards the Afghan government andthe international community. The findings fromthis study and other research suggest that aid ismore effectively spent in secure regions wheregood development practice and strongerWinning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 5

oversight is more feasible, and less money has tobe spent on security. The research also suggeststhat in areas where insecurity remained chronicand governance structures broken, spendingresources (e.g., for road building) risks fuelingcorruption (both perceived and real), intercommunalstrife, and competition among localpower-brokers. There is evidence that ininsecure areas local strongmen with militias thatwere being paid to provide security recognizedthe need to perpetuate insecurity. Theprioritization of insecure over secure areas is notsurprisingly being bitterly criticized by Afghansliving in more stable areas, who feel they arebeing penalized for being peaceful.Recommendation:• Reverse the current policy of rewarding insecureareas with extensive aid while effectivelypenalizing secure areas where aid money couldbe spent more effectively and accountably. Investin secure areas and, except for humanitarianassistance, make aid in insecure areas morecontingent on security. While this study did notspecifically examine the demonstration effect thiscould have, it is quite possible that providingincentives for communities to be peaceful wouldbe more effective than the current approach thatis perceived by many Afghans to be rewardinginsecurity.5. Accountability and the measurement ofimpact have been undervaluedThe political need for “quick impact” along withinstitutional imperatives to spend money have inmany cases reduced the incentives for carefulevaluation of project impact. Currently it is noteven possible to get a complete list of the projectsPRTs have implemented with the approximately$2.64 billion in CERP funds appropriatedbetween 2004 and 2010, let alone an indicationof what the impact has been. The study’sfindings have been reinforced by increasingmedia and U.S. agency reports (e.g., SpecialInspector General for AfghanistanReconstruction [SIGAR], USAID Office of theInspector General [OIG]) on funds that havebeen wastefully spent with no (or negative)impact.In an environment with little reliablequantitative data, with numerous independentvariables that make determining correlation (notto mention causality) virtually impossible, andwhere western-style public opinion pollingmethodologies may not be reliable, thedetermination of impact may often have to bemore art than science. Nevertheless, much morefocus should be given to trying to measure theimpact and consequences of aid projects than hasbeen done to date. Recent initiatives by SIGAR,OIG, and staff at the Senate Foreign RelationsCommittee are positive, but they come late inthe game. In addition to the waste of taxpayerresources and negative consequences on theground, the discrediting of all programs forAfghanistan may be collateral damage if aidresources are not spent in a more accountableand effective manner.Recommendation:• Reinforce at all levels the message and cultureof accountability. This is not arecommendation to add several morebureaucratic levels of cumbersome nationaland international oversight mechanisms tooversee inputs, but rather to invest more inmeasuring outcomes. Establish incentivestructures for quality work and carefulassessments of effectiveness and not just forspending money.6. Development is a good in and of itselfThere is considerable evidence that developmentassistance in Afghanistan during the past decadehas directly contributed to some very positivedevelopment benefits, including decreases ininfant and maternal mortality, dramatic increasesin school enrollment rates for boys and girls, amedia revolution, major improvements in roadsand infrastructure, and greater connectivitythrough telecommunication networks. Oneconsequence of viewing aid resources first andforemost as a stabilization tool or “a weaponssystem” is that these major development gainshave often been under-appreciated because theydid not translate into tangible security gains.U.S. development assistance in Afghanistan hasbeen justified on the grounds that it is promoting6Feinstein International Center

COIN/stabilization objectives rather thandevelopment objectives. While in the short termthis has led to much higher levels of developmentassistance in Afghanistan, the failure of theseresources to improve the security situation isnow leading many policymakers to question thevalue of development assistance despite somevery real development gains.Recommendation:• Value development as a good in and of itself.Program development aid first and foremost topromote development objectives, where thereis evidence of impact and effectiveness, ratherthan to promote stabilization and securityobjectives, where this research suggests there islittle evidence of effectiveness.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 7

1. IntroductionU.S. military on patrol, HelmandU.S. Marine Corps photo by Lance Cpl. John M. McCall1.1 Purpose, rationale, and description ofstudyPolitical and national-security considerationshave always influenced U.S. foreign assistancepolicies and priorities. Since 9/11, however, thisinfluence has grown greatly, as development aidfor countries like Afghanistan, Iraq, and Pakistanhas been increasingly and explicitly subsumedunder the national security agenda. The U.S. isnot alone in viewing development through asecurity lens; other Western nations, includingmany of the U.S.’s NATO allies, have, tovarying extents, restructured their aid programsto reflect the prevailing foreign policy agenda ofconfronting global terrorism. The major impactthis has had on development assistance has beenat the policy and practice levels. It is reflected inchanging aid strategies, priorities, and structures.At the same time, a widely held assumption inmilitary and foreign policy circles is thatdevelopment assistance is an important “softpower” 1 tool to win consent and to promotestabilization and security objectives.Counterinsurgency doctrine in particularemphasizes the importance of humanitarian andreconstruction assistance, often in the form of“Quick Impact Projects” that are intended to“win hearts and minds.”The assumption that aid projects improvesecurity has had a number of implications,including the sharp increase since 9/11 in theabsolute amount of funding available from bothU.S. and other Western donors for humanitarian1The term “soft power“ was coined by Joseph Nye in the 1980s to refer to a nation’s ability to influence the preferences and behavior ofother nations not through coercion (hard power), but through projection of attractive national values, levels of prosperity, and openness.See Joseph Nye, “The Benefits of Soft Power,” Working Knowledge (Harvard Business School, August 2, 2004), http://hbswk.hbs.edu/archive/4290.html.8Feinstein International Center

and development purposes 2 ; the increasedpercentage of assistance that is programmedbased on strategic security considerations ratherthan on the basis of poverty and need; and themuch greater role for the military or combinedcivil-military teams in activities that weretraditionally the preserve of humanitarian anddevelopment organizations. This assumption ishaving a major policy impact on howdevelopment assistance is allocated and spent,and provides an important rationale for thegrowing “securitization” of developmentassistance. 3 On the one hand, military forceshave become increasingly involved in whatwould previously have been seen as the work ofcivilian humanitarian and development agencies.On the other hand, civilian agencies, includingnon-governmental organizations, have beenincreasingly enlisted in aid and developmentprojects that are seen as having stabilizationobjectives. The assumption has been formalizedin the “comprehensive,” “whole of government,”and “3D” (diplomacy, defense, development)approaches. (See Section 2 and Annex A.)Despite how widespread the assumption is,and despite its major impact on aid andcounterinsurgency policies, there is littleempirical evidence that supports the assumptionthat reconstruction assistance is an effective toolto “win hearts and minds” and improve securityor increase stability in counterinsurgencycontexts. To help address this lack of evidence,the Feinstein International Center (FIC) at TuftsUniversity conducted a comparative study inAfghanistan and the Horn of Africa to examinethe effectiveness of aid projects in promotingsecurity objectives in stabilization andcounterinsurgency contexts. 4Afghanistan provided an opportunity to examineone of the most concerted recent efforts to use“hearts and minds” projects to achieve securityobjectives, especially as it has been the testingground for new approaches to usingreconstruction assistance to promote stability,which in some cases (e.g., ProvincialReconstruction Teams) were then exported toIraq. While other studies have looked at theeffectiveness of aid in promoting humanitarianand development objectives as well as the ethicaland philosophical issues related to merginghumanitarian and security objectives,surprisingly little effort has been given toanalyzing the effectiveness of aid in promotingpolitical and security objectives. Given that asignificant percentage of U.S. foreign aid is nowprogrammed (both explicitly and implicitly) toachieve security objectives, the need todetermine the effectiveness of this use ofdevelopment assistance is real. 5While aid projects are not all designed withstabilization objectives in mind, the study did notdistinguish between military and non-military aid,although in some cases it focused more on militarylinkedaid. While projects that had an explicitstabilization focus might have been of specialinterest, the broad point is that aid in general isassumed to promote stability. Also, while the U.S.is the largest donor and has increased its aidspending by the largest percentage post-2001, thestudy did not intend to be primarily U.S.-focused;therefore, it looked at all aid.2According to Center for Global Development statistics, between 2001 and 2009 U.S. official development assistance more than doubled inreal terms, while Donor Assistance Committee countries’ assistance increased by more than half. See Net Aid Transfers data set (1960–2009), http://www.cgdev.org/content/publications/detail/5492.3The terminology of the “securitization of aid” and much of its intellectual underpinning has been provided by Professor Mark Duffield,who uses it to describe the important role of development aid to support a new system of global governance that helps protect westernsecurity interests. Rather than being primarily about helping the poor through alleviating poverty and promoting development, Duffieldargues that aid is increasingly being used as a governance and security tool to help stabilize and govern unstable and borderland regions sothat they do not threaten the West’s way of life. See, for example, Mark Duffield, “Governing the Borderlands: Decoding the Power ofAid,” Disasters, Volume 25, Issue 4, (December 2001), pp. 308-320; Mark Duffield, Global Governance and the New Wars: The Merging ofDevelopment and Security, (London: Zed Books, 2001); and Mark Duffield, Development, Security and Unending War: Governing the World ofPeoples, (Cambridge: Polity Press, 2007).4This paper focuses on the findings from Afghanistan. For information on findings from the Horn of Africa, see https://wikis.uit.tufts.edu/confluence/pages/viewpage.action?pageId=34807224. For information on the overall aid and security research program, see https://wikis.uit.tufts.edu/confluence/pages/viewpage.action?pageId=19270958.5A March 2010 conference co-sponsored by Feinstein International Center on the use of development aid in COIN operations inAfghanistan noted that “a key theme is the critical lack of monitoring, evaluation, and empirical data available to assess the impact of aid onstability in Afghanistan,” especially given the “otherwise strong traditions of robust after-action reviews” by the military. See Wilton Park,“Winning ‘Hearts and Minds’ In Afghanistan: Assessing the Effectiveness of Development Aid in COIN Operations,” Report on WiltonPark Conference 1022, held March 11–14, 2010 (July 22, 2010). http://www.wiltonpark.org.uk/resources/en/pdf/22290903/22291297/wp1022-report.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 9

After presenting the study methodology below,the paper continues in Section 2 with adiscussion of the evolution of security-driven aidand the effects on current practice. Section 3describes the five provinces included in the studyand how different models of securitizeddevelopment have been used in each of them.Sections 4 and 5 present the views obtained inthe field on the causes of insecurity and thecharacteristics of aid projects. Section 6 discussesthe effectiveness of aid in stabilization inAfghanistan, and Section 7 summarizes thestudy’s conclusions and policy implications. Moredetailed information on the researchmethodology and related issues is contained inAnnex A.1.2 MethodologyResearch was conducted in the five provinces ofBalkh, Faryab, Helmand, Paktia, and Uruzgan,as well as in Kabul city. In these provinces, as innearly all of Afghanistan’s thirty-four provinces,international civilian and military actors areusing humanitarian, reconstruction, anddevelopment aid to promote greater stability andsecurity. The notable differences among the fiveprovinces provided the opportunity to examinethe development-security nexus in very differentcontexts. Balkh and Faryab Provinces in thenorth were much more secure than Helmand,Uruzgan, and Paktia Provinces in the south andsoutheast where the Taliban-led insurgency wasmore active. In the two northern provinces,Pashtuns are a minority ethnic group, whereas inthe south and southeast they make up theoverwhelming majority. Another significantdifference was the variations in approach,budgetary resources, and character of thedifferent NATO/ISAF (International SecurityAssistance Force) nations heading the ProvincialReconstruction Teams (PRTs) in eachprovince—with perhaps the major difference forthe purposes of this study being the much greaterfinancial resources available to U.S.-led PRTs.The study teams used a relatively consistentmethodology in four of the five provincial studyareas (Helmand being the exception), bearing inmind that the varied security and otherconditions allowed or required approachestailored to different areas. Qualitative interviewswith Afghan and international respondents inthe field provided the primary data source.Interviews were conducted during multiplevisits to the provinces between June 2008 andFebruary 2010. In the four provinces as well asin Kabul, a total of 574 respondents (340Afghan, 234 international) were interviewedeither individually or in focus groups at theprovincial, district, and community levels.Separate semi-structured interview guides wereused for key informant and community-levelinterviews. Respondents included current andformer government officials, donors, diplomats,international military officials, PRT militaryand civilian staff, UN and aid agency staff,tribal and religious leaders, journalists, tradersand businessmen, and community members. InHelmand, the methodology consisted ofanalyzing qualitative data from focus groupsconducted in February-March 2008,quantitative data taken from polling data drawnfrom communities in November 2007 andprovided by the PRT, and interviews with keyinformants (e.g., PRT staff, Afghan governmentofficials). Most of the interviews with Afghanswere conducted in Dari or Pashtu, althoughsome with senior government and NGOofficials were conducted in English. The twointernational researchers leading the fieldresearch in Balkh and Faryab Provinces wereexcellent Dari speakers and could directlyinterview Afghan respondents. Afghan researchassistants helped in setting up and conductinginterviews, as well as in note taking andanalysis. Elsewhere, research assistants ortranslators assisted researchers in translatingDari and Pashtu. In all provinces, secondarysources were drawn upon for historicalinformation and background to aid projects.The Balkh, Faryab, and Uruzgan case studiesbenefited from background historical andpolitical overviews written by leading analystsof these provinces.Any research in Afghanistan or other conflictareas requires caution because of the potentialfor respondent bias. This is particularly the casefor research that looks at the types of sensitiveissues raised in this study or includes questionsthat relate to deeply held social norms. Tomitigate these potential biases, the methodologyincluded repeat visits to allow follow-up to and10Feinstein International Center

triangulation of responses, flexible interviewguides that encouraged spontaneous responseswithin specific themes, and the fielding ofteams with extensive local experience.The study relied primarily on the statedperceptions of the wide range of respondentsmentioned above. Where relevant, thediscussion differentiates the perspectives ofdifferent types of actors. The researchersacknowledge the need for caution when basingfindings on the stated perceptions ofrespondents, as respondents’ statements may notalways accurately reflect their perceptions and,in addition, may not match behavior. The studyalso did not aim to measure causality, as thiswas simply too ambitious in an environmentwith so many confounding variables. Still,because aid projects explicitly aim to changeattitudes, perceptions (if captured accurately)are relevant. Moreover, however imperfect, theresearch team believed that in the Afghancontext the qualitative data gathered in indepthinterviews provided a better data sourceand gauge of perceptions than most datacollected using quantitative methodologies,such as public opinion polling.Since the field research was completed in early2010, a number of the issues raised by thefindings have been acknowledged by the U.S.and by ISAF, and some measures have beentaken to mitigate them, as discussed in the text.Security conditions have also changedsomewhat. While Balkh and Faryab are stillrelatively secure, insecurity has widened in thenorth in general, and in the troubled districts inthe provinces of Balkh and Faryab in particular.On the other hand, security in areas ofHelmand has improved since the time of theresearch. Nevertheless, based on more recentvisits and discussions as well as the analysis ofothers’, the researchers feel that the broadconclusions and concerns remain valid and verypolicy-relevant.Additional information on the researchmethodology and related issues is contained inAnnex A.1.3 Similar and related research inAfghanistanA number of other studies and evaluationsexamining the effect of development activitieson security have been conducted in Afghanistanat the same time or later than this one. One ofthe most comprehensive, conducted by ChristofZürcher, Jan Koehler, and Jan Böhnke innortheastern Takhar and Kunduz Provincesbetween 2005 and 2009, concluded thatcommunities that already felt more secure weremore likely to feel positively about aid, and thatany positive effects of aid on the population’sattitude towards the state are “short-term” and“non-cumulative.” 6 Similarly, researchconducted in 2009 by Sarah Ladbury inKandahar, Wardak, and Kabul Provinces foundthat young men joined the insurgency for acomplex combination of reasons, some personaland some related to broader grievances againstthe government and foreign forces.Development projects were seen as being toosmall to have any impact, and as unemployment(or underemployment) was one factor leading tomobilization, respondents expressed the desirefor projects that created employment. The poorquality of aid delivery suggested that moreattention be paid to how services are delivered. 7A quantitative study done with the support ofthe U.S. Army using district-level data for the2002–10 period found that while projects canaffect the number of security incidents in adistrict, in most cases their influence is so smallas to not justify using them as a conflictmitigationtool. 8 Finally, a report prepared inJune 2011 for the use of the U.S. Senate ForeignRelations Committee concluded, afterexamining the evidence from several studies(including the present one and those mentionedabove), that “the evidence that stabilization6J. Böhnke, J. Koehler and C. Zürcher, “Assessing the Impact of Development Cooperation in North East Afghanistan 2005–2009: FinalReport,” Evaluation Reports 049 (Bonn: Bundesministerium für wirtschaftliche Zusammen arbeit und Entwicklung, 2010).7Sarah Ladbury, in collaboration with Cooperation for Peace and Unity (CPAU), “Testing Hypotheses on Radicalisation in Afghanistan:Why Do Men Join the Taliban and Hizb-i Islami? How Much Do Local Communities Support Them?” Independent Report for theDepartment of International Development (August 2009).8Schaun Wheeler and Daniel Stolkowski, “Development as Counterinsurgency in Afghanistan,” Unpublished manuscript (Draft dated June17, 2011).Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 11

Photo by Shirine Bakhat-PontISAF patrol passing through Mazar-e Sharifprograms promote stability in Afghanistan islimited . . . [The] unintended consequences ofpumping large amounts of money into a warzone cannot be underestimated.” 9 One study thatdid show some impact of development assistanceon security perceptions was a World Bankfundedstudy of the National Solidarity Program(NSP) conducted in villages outside of the mainconflict areas between 2007 and 2011 by AndrewBeath and his colleagues. The study found thatthe NSP had overall positive effects oncommunities’ perceptions of economic wellbeing,all levels of government (except thepolice), and non-governmental organizations(NGOs), the security situation, and (weakly)international forces. This study did not, however,find measurable improvements in actual security,although it looked only at short-term outcomes. 10These studies are discussed further in Section 6below.9Majority Staff, Foreign Relations Committee, United States Senate, “Evaluating U.S. Foreign Assistance to Afghanistan,” (June 8, 2011), p. 2,http://www.gpoaccess.gov/congress/index.html.10Andrew Beath, Fotini Christia, and Ruben Enikolopov, “Winning Hearts and Minds? Evidence from a Field Experiment in Afghanistan,”Working Paper No. 2011-14 (Massachusetts Institute of Technology, Political Science Department, undated).12Feinstein International Center

2. Evolution of security-driven aid 11Photo by Cpl. Jeff DrewU.S. military inspecting school construction, HelmandThe role of militaries in delivering aid andreconstruction is not a new phenomenon. Whathas changed, especially since 2001, is the scale ofthe involvement and the purposes underpinningit. According to the Organization for EconomicCo-operation and Development (OECD),between 2002 and 2005, USAID’s share of U.S.official development assistance (ODA) decreasedfrom 50 to 39 percent, while the share of theDepartment of Defense (DOD) increased from 6to 22 percent. 12 While more recent OECD datanow show DOD’s global share of U.S. ODA tobe shrinking, it still plays a dominant role inAfghanistan. Between 2002 and 2010, nearly $38billion was appropriated to “stabilize andstrengthen the Afghan economic, social,political, and security environment so as to bluntpopular support for extremist forces in theregion.” 13 Roughly half of this has gone totraining and equipping the Afghan NationalSecurity Forces (ANSF), while a little overone-third has gone to economic, social, political,and humanitarian efforts. Of the total amount,nearly two-thirds has been allocated to theDOD, with USAID and the Department of Statereceiving lesser amounts. One specific indicatorof the importance of security-driven aid is thedramatic growth of the Commanders’Emergency Response Program (CERP) 14funding in Afghanistan, from zero in 2003 to$1.2 billion in 2010 (Figure 1).11The authors would like to acknowledge the very substantive contribution of Dr. Stuart Gordon to this section.12OECD, “DAC Peer Review: Main Findings and Recommendations, Review of the Development Co-operation Policies andProgrammes of United States” (2006), https://www.oecd.org/document/27/0,2340,en_2649_34603_37829787_1_1_1_1,00.html.13Curt Tarnoff, “Afghanistan: U.S. Foreign Assistance” (Congressional Research Service, August 12, 2010).14CERP is a U.S. military program which provides discretionary funds for PRT commanders to execute local small-scale relief andreconstruction projects. Projects are intended to build good will, trust, and confidence between the local population and the internationalmilitary, thereby increasing the flow of intelligence and turning the population against the insurgents and other anti-government groups.CERP is discussed further below and in Box 2.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 13

Figure 1Source: Special Inspector General for Afghanistan Reconstruction,“Quarterly Report to the U.S. Congress” (April 30, 2011),142–3.The contemporary origins of militaryinvolvement in delivering assistance lie in theAllied preparations for the invasions of NorthAfrica and Western Europe in the course ofWorld War II. However, it is perhaps morestrongly associated with the Cold Warcounterinsurgency campaigns of the 1950sthrough the 1970s, principally the Britishexperiences in Malaya, Oman, and Aden(Yemen), and the U.S. experience in Vietnam.The phrase “hearts and minds” is usuallyassociated with Field Marshal Sir GeraldTempler, 15 and his ultimately successful conductof the British-led counter insurgency campaignin Malaya (1948–60). Since the “MalayanEmergency,” the phrase has often been used as aform of shorthand for the overall Britishapproach to counter-insurgency: emphasizingwinning the “hearts and minds” of thepopulation through securing the support of thepeople. The approach shaped British strategyboth in Malaya and in dealing with the MauMau rebellion in Kenya. In the 1970s, it wasinfluential in Northern Ireland. The kernel ofthe strategy was to establish secure zones, useminimum force, apply development, and addresspolitical grievances that underlay the rebellions— all in order to turn the population against theinsurgents. At the same time, outside of thesecure areas, the strategy was to implementmilitary measures designed to inflict attrition onthe military component of the insurgency. Thisapproach has been contrasted with tactics thatstress more conventional military means, are lessfocused on developing the support of thepopulation, and are less concerned with avoidingcivilian casualties.The U.S. experience began as Civil Affairs inWorld War II but has echoed the British path inits association with counter-revolutionary warfare,particularly in programs such as the CivilOperations and Revolutionary DevelopmentSupport Program (or CORDS) during theVietnam War. These counter-insurgencyapproaches tended to bring together efforts toseparate the population from the insurgents whileproviding a variety of reconstruction programs towin over the sympathy of the population. Thephrase “hearts and minds” was also associatedwith the U.S. military and strategies adopted tocontain the communist insurgency in Vietnam.President Lyndon B. Johnson is quoted in May1965 when he argued that U.S. victory would bebuilt on the “hearts and minds of the people whoactually live out there. By helping to bring themhope and electricity you are also striking a veryimportant blow for the cause of freedomthroughout the world.” 16 This approach shapedboth U.S. strategy and rhetoric on the war inIndo-China and led to efforts to coordinatedevelopment and security approaches that wouldcounter communist propaganda and isolate theinsurgents from the people. Under Johnson theU.S. “committed itself to ‘pacification’ of SouthVietnam by providing both security anddevelopment support. U.S. officials, both civilianand military, would provide ‘advice’ and resourcesfor economic development projects, such asrebuilding roads and bridges, while the militarywould train and equip South Vietnam’s police andparamilitary groups to hunt down insurgents.” 1715John Cloake, Templer: Tiger of Malaya: The Life of Field Marshal Sir Gerald Templer (London: Harrap, 1985). Footnote 1 states that Templerfirst used the term on April 26, 1952.16Lyndon B. Johnson quoted in Francis Njubi Nesbitt, “Hearts and Minds and Empire.” Foreign Policy in Focus (March 20, 2009), http://www.fpif.org/articles/hearts_and_minds_and_empire.17Ibid.14Feinstein International Center

2.1 Current approachesRecognition that the causes of instability arecomplex has driven the formulation of the variousmodels, such as the“ comprehensive approach,”“3D” (defense, diplomacy, development), “wholeof government,” and “integrated approach.”These trends are discernible in a range ofinternational organizations (particularly the EUand NATO) as well as between and withinministries in individual states. For example, in2004, the UK government established a Post-Conflict Reconstruction Unit (renamed theStabilisation Unit in 2007), jointly owned by theMinistry of Defense (MOD), Foreign andCommonwealth Office (FCO) and Departmentfor International Development (DFID); in 2004,the U.S. Congress appropriated funds for the U.S.State Department to establish the Office of theCoordinator for Reconstruction and Stabilization(S/CRS); and in 2005, the Canadian governmentestablished the Stabilisation and ReconstructionTask Force (START) within its Department ofForeign Affairs and International Trade.The various models are all derived from thesense that the origins of conflict are largelysocio-economic in nature, and that state failureis a consequence of the breakdown of publicservice delivery. Militaries have been attractedto comprehensive or integrative approaches fora variety of reasons related both to theories ofconflict causation and resolution and to thenecessity for “soft power” to contribute to“force protection.” 18 In terms of theories aboutwhat causes conflicts and how they are resolved,conflict is frequently portrayed as a product oflow levels of development as well as politicaland social marginalization. In the course of aninternational military intervention, it is oftenassumed that tactical military progress cannotbe consolidated or translated into strategicsuccess, and viable states cannot be built,without the host government constructinglegitimacy through the provision of publicservices. In terms of force protection, thetheory is that when an international militaryforce provides infrastructure and “facilities”(such as public health clinics, wells, and schools)the population will be encouraged intocollaborative relationships with theinternational military—reducing opportunitiesfor insurgents and providing intelligence to thecounter-insurgent forces.The U.S. DOD has made considerable efforts todevelop the capacities for “stability operations” 19and to link these with the work of theDepartment of State and USAID—the result isan approach to security that makes afundamental break with the past. The principalchange has been a reorientation to meet adifferent perceived threat. The 2002 U.S.National Security Strategy stated that the U.S.“is now threatened less by conquering statesthan we are by failing ones.” Then Secretary ofDefense, Robert M. Gates, argued that even theemphasis on regime change that dominatedbetween 2001 and 2003 had changed:Repeating an Afghanistan or an Iraq—forcedregime change followed by nation-building underfire—probably is unlikely in the foreseeablefuture. What is likely though, even a certainty, isthe need to work with and through localgovernments to avoid the next insurgency, torescue the next failing state, or to head off thenext humanitarian disaster.Correspondingly, the overall posture and thinkingof the United States armed forces has shifted—away from solely focusing on direct Americanmilitary action, and towards new capabilities toshape the security environment in ways that obviatethe need for military intervention in the future. 20In 2005 the DOD issued “DOD Directive3000.05” (2005), emphasizing the importance18Force protection consists of preventive measures intended to reduce hostile actions against military personnel, resources, facilities, andinformation.19DOD defines stability operations as “An overarching term encompassing various military missions, tasks, and activities conducted outsidethe United States in coordination with other instruments of national power to maintain or reestablish a safe and secure environment,provide essential governmental services, emergency infrastructure reconstruction, and humanitarian relief.” See Headquarters Departmentof the Army, “Stability Operations,” FM 3-07 (December 2008), Glossary, p. 9.20Robert M. Gates. (U.S. Secretary of Defense) “U.S. Global Leadership Campaign.” Speech given on July 15, 2008 at the U.S. GlobalLeadership Campaign (Washington, DC). http://www.defenselink.mil/speeches/speech.aspx?speechid=1262.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 15

of stability operations asa core U.S. military mission that the Department ofDefense shall be prepared to conduct and support.They shall be given priority comparable to combatoperations and be explicitly addressed and integratedacross all DOD activities including doctrine,organizations, training, education, exercises,materiel, leadership, personnel, facilities, andplanning. 21The directive emphasized that “stabilityoperations were likely to be more important tothe lasting success of military operations thantraditional combat operations” and “elevatedstability operations to a status equal to that of theoffense and defense.” 22 This represented asignificant reorientation of the U.S. military’straditional focus on “warfighting.”More recently, this emphasis on the identificationof instability in other states as a threat to U.S.interests was reflected in both the 2008 NationalDefense Strategy and the 2010 QuadrennialDefense Review (QDR), which noted that“preventing conflict, stabilizing crises, andbuilding security sector capacity are essentialelements of America’s national securityapproach.” 23 Understandably, this thinking hasplayed an important role in driving forcepreparations within the DOD, strengthening thesignificance of stability operations andbroadening the range of competencies thatwould be required in a future U.S. military.The changing emphasis in the U.S. is alsoconsistent with the changing nature of UNpeacekeeping missions, increasinglycharacterized by complex mandates spanning“immediate stabilization and protection ofcivilians to supporting humanitarian assistance,organizing elections, assisting the developmentof new political structures, engaging in securitysector reform, disarming, demobilizing andreintegrating former combatants and laying thefoundations of a lasting peace.” 24 This integrationof diplomatic, human rights, military, anddevelopment responses has been driven primarilyby the requirement to effectively consolidatefragile peace agreements and make the delicatetransition from war to a lasting peace 25 —thefragility of peace often being “ascribed to a lackof strategic, coordinated and sustainedinternational efforts.” 26 A significant amount ofliterature documents the increasing size andcomplexity of, particularly, the civiliancomponents of peace missions 27 (arguablybeginning with the deployment of the UNTAGmission in Namibia in 1989 28 ) and thediversification and growing importance ofnon-military tasks within UN mandates. Evenwhere the UN has deployed solely civilianmissions, their proximity and relationship tomilitary, peace-building, or state-buildingmissions in support of a government authorityhas raised the same issue for some critics: theassociation of humanitarian and developmentresponses with one of the belligerentsundermines the UN’s independence andneutrality.Despite the lack of evidence that “hearts andminds” activities can generate attitude orbehavior change, these broader strategic trends21“DOD, Directive 3000.05: Military Support for Stability, Security, Transition, and Reconstruction (SSTR) Operations” (November 28,2005). p.2. As noted on the next page, the directive also clarifies that DOD sees its role in U.S. government plans for SSTR as part ofinteragency partnerships.22FM 3-07, “Stability Operations,” p. vi.23Lauren Ploch, “Africa Command: U.S. Strategic Interests and the Role of the U.S. Military in Africa” (Congressional Research Service,July 22, 2011), http://opencrs.com/document/RL34003/24Espen Barth Eide, Anja Therese Kaspersen, Randolph Kent, and Karen von Hippel, “Report on Integrated Missions: PracticalPerspectives and Recommendations,” Independent Study for the Expanded UN ECHA Core Group (May 2005), p. 3.25“In Larger Freedom: Towards Development, Security and Human Rights for All,” A/59/2005 (March 21, 2005), paragraph 114.26Eide et al., “Report on Integrated Missions,” p. 3.27For more on the changing nature of peace operations, see Bruce Jones and Feryal Cherif, “Evolving Models of Peacekeeping: PolicyImplications and Responses,” (Center on International Cooperation, NYU, September 2003), http://www.peacekeepingbestpractices.unlb.org/.28The United Nations Transition Assistance Group (UNTAG) was a UN peacekeeping force deployed in April 1989 with a very broadmission in the transition from then South-West Africa to an independent Namibia. See, United Nations, “Namibia – UNTAGBackground” (undated), http://www.un.org/en/peacekeeping/missions/past/untagFT.htm.16Feinstein International Center

Photo by Cpl. Andrew CarlsonU.S. military and construction, Helmandand the hearts and minds approach have had adramatic impact on doctrine and tactics relatedto stabilization and on the organization of theU.S. military. Department of Defense Directive(DODD) 3000.05 defined the broad objectivesof stability operations, stating that theirimmediate goal was to “provide the localpopulace with security, restore essential services,and meet humanitarian needs. Long-term goalsthat reflect transformation and fostersustainability efforts include developing hostnationcapacity for securing essential services, aviable market economy, rule of law, legitimateand effective institutions, and a robust civilsociety.” 29 The U.S. Army’s manual, Tactics inCounterinsurgency, states that “at its heart acounterinsurgency is an armed struggle for thesupport of the population.” 30 This populationcentricview of conflict is rooted in theassumption that conditions such as poverty,unemployment, illiteracy, and unmet aspirationsare the fuels that drive insurgencies and that theremedy is humanitarian, reconstruction, anddevelopment assistance. The idea underpinningthese perspectives is that project-based assistance,including small-scale “Quick Impact Projects,”can capture the “hearts and minds” ofbeneficiary populations and lead to both achange of attitude towards the government andincreasing co-operation with the internationalmilitary. The latter is most likely to be seen interms of intelligence sharing—e.g., identifyingimprovised explosive devices (IEDs) or providinginformation on insurgents—a key component inprotecting one’s own forces. The deterioratingsecurity situation in parts of Afghanistan, whichforced many traditional humanitarian anddevelopment organizations to suspend activitiesand withdraw staff, also contributed to thegrowing role of the military and PRTs indirectly supporting reconstruction anddevelopment activities as they felt compelled tostep into the void. 3129FM3-07, “Stability Operations,” p. 1-1530U.S. Department of the Army, Tactics in Counterinsurgency, FM 3-24.2 (April 2009), p. ix.31Stewart Patrick and Kaysie Brown, “The Pentagon and Global Development: Making Sense of the DOD’s Expanding Role,” WorkingPaper Number 131 (November 2007).Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 17

The need to develop interagency partnershipshas been a recurring theme across recent U.S.security policy. 32 In the preface to FM 3-07“Stability Operations,” Lieutenant GeneralCaldwell, argues that at the heart of the programto enhance stability operations is “acomprehensive approach . . . that integrates thetools of statecraft with our military forces,international partners, humanitarianorganizations, and the private sector.” 33 The sumof these changes is an organizationalreconfiguration around a military capable ofintervening in fragile states and having thepotential to assume responsibility for a range oftasks that had traditionally been delivered bygovernment civilian agencies and developmentand humanitarian organizations. While thisguidance recognized that “many stabilityoperations tasks are best performed byindigenous, foreign, or U.S. civilianprofessionals,” it situated the militarystabilization response within a broader “whole ofgovernment” 34 approach, but made clear that“U.S. military forces shall be prepared toperform all tasks necessary to establish ormaintain order when civilians cannot do so.”Officially, U.S. development agencies haveembraced this relationship. According to USAIDAdministrator Rajiv Shah, “in the most volatileregions of Afghanistan, USAID works side-bysidewith the military, playing a critical role instabilizing districts, building responsive localgovernance, improving the lives of ordinaryAfghans, and—ultimately—helping to pave theway for American troops to return home.” 35While policy has referred to “partnerships” andintegration of military and civilian agencies, themilitary’s relationship with the civilian branchesof government and the external humanitariancommunity is unlikely to be an equal one. AU.S. military with significantly more financial,personnel, and material resources, and fargreater reach than either USAID or the StateDepartment, inevitably makes the military’s rolestrong and potentially dominant.While the Afghan and Iraqi funding surges areclearly only temporary, DOD’s approach tostabilization, and the significant role of themilitary in delivering ODA, is likely to endure. 36Some argue that the DOD’s expanded assistanceauthorities” threaten to “displace or overshadowbroader U.S. foreign policy and developmentobjectives in target countries and exacerbate thelongstanding imbalance between the militaryand civilian components of the U.S. approach tostate-building.” 37The focus on integrated and comprehensiveapproaches to secure the support of thepopulation has led humanitarian anddevelopment actors to criticize both stabilizationand counterinsurgency doctrines for leaving littleroom for the fundamental humanitarianprinciples of independence, impartiality, andneutrality. Assistance is increasinglyinstrumentalized behind security and politicalobjectives. Patrick and Brown argue that thesetrends are potentially damaging and raiseconcerns that “U.S. foreign and developmentpolicies may become subordinated to a narrow,short-term security agenda at the expense ofbroader, longer-term diplomatic goals andinstitution-building efforts in the developingworld.” 38Many aid practitioners and agencies areconcerned by the increased politicization andsecuritization of aid because this violates32As featured in the National Military Strategy, the National Defense Strategy, and the 2010 QDR: DOD, The Quadrennial Defense ReviewReport (February 2010), p. 73. The FY2011 Defense Budget Request is available at http://www.defenselink.mil/comptroller/budget.html.33Lieutenant General and former Commander of the U.S. Army Combined Arms Center, William H. Caldwell, as quoted in the preface toFM 3-07 “Stability Operations.”34DOD defines “whole of government” as “an approach that integrates the collaborative efforts of the departments and agencies of theUnited States Government to achieve unity of effort toward a shared goal.” See FM3-07, “Stability Operations,” Glossary, p. 10.35Cited in Scott Dempsey, “Is Spending the Strategy?” Small Wars Journal (May 4, 2011), p. 2. http://smallwarsjournal.com/jrnl/art/is-spending-the-strategy.36Some respondents for this study questioned whether this expanded role has been accepted wholeheartedly as being of equal status by themilitary, or whether there will be resistance and reconsideration in light of experience in Afghanistan and elsewhere.37Scott Dempsey.38Stewart Patrick and Kaysie Brown.18Feinstein International Center

Photo by Paul FishsteinSchool, Balkhfundamental humanitarian principles, and severalinterviewed for this study expressed concernabout the negative consequences for civilian aidworkers of the “blurring of lines” that resultedfrom the direct role of the military in thedelivery of humanitarian and developmentassistance. However, few questioned thefundamental assumption of a causal relationshipbetween aid and security. In fact, some aidworkers seemed to believe more strongly in thisrelationship than military respondents, and werevery supportive of using aid to promote conflictresolution, stabilization and peacebuildingobjectives as long as this was done by civilianrather than military agencies and personnel. Theobjective of this study, however, was not toexamine whether humanitarian principles werebeing violated, or the impact of the securitizationof aid on humanitarian actors, but rather toexamine the fundamental question – isdevelopment assistance effective in promotingstabilization and security objectives?Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 19

3. Aid and security in AfghanistanNational ring road in FaryabPhoto by Geert Gompelman3.1 Provincial backgroundThe five study provinces (see map on page vii)were selected in part due to their variedcharacteristics, and in part due to morepragmatic considerations of local contacts, donorinterest, and logistical support. The threesouthern and eastern provinces (Helmand,Uruzgan, Paktia) were selected due to theirinsecurity and because of the different aidapproaches taken in each (see Section 3.2). Thetwo northern provinces (Balkh, Faryab) wereselected due to their relative security; theyprovide a counterpoint to the less-secureprovinces by illustrating the response to aid in amore-secure environment. In 2007, the fiveprovinces collectively accounted for 27 percentof government and donor assistance. UnlikeHelmand, Uruzgan, and Paktia, Balkh andFaryab are non-Pashtun-majority provinces.Beyond that, each of the five provinces has itsown varied history, geography, economy, andpolitics.Helmand Province lies in southwesternAfghanistan, bordering Pakistan’s BalochistanProvince to the south. Nearly 95 percent of thepopulation is Pashtun, living alongside smallnumbers of the Tajiks, Hazaras, Uzbeks, Baluch,and Sikhs who were settled in the 1960’s as partof the major Helmand-Arghandab Valleyirrigation scheme. The dislocations of the lastthirty years are seen as having resulted in manyunsettling social and political changes. Helmand20Feinstein International Center

produces a variety of field and horticulturalcrops, but has become synonymous with opiumproduction and trade. The provincial center isLashkar Gah.Uruzgan Province lies in south-centralAfghanistan, bordering Kandahar and HelmandProvinces to the south and southwest,respectively. Approximately 90 percent of thepopulation is Pashtun, with much of the formerHazara population having been separated intothe new province of Dai Kundi in 2004. 39Uruzgan has historically been remote, poor,conservative, and violent, even by Afghanstandards. Politically and tribally, Uruzgan is partof “greater Kandahar,” and is also the origin ofmany of the Taliban’s original leaders, includingMullah Mohammad Omar, who was born inDeh Rawood District. As in Helmand, tribalstructure is generally considered much lesscohesive and more fragmented than in the eastand southeast of the country. The provincialcenter is Tarin Kot.Paktia Province lies in southeastern Afghanistan,bordering Pakistan’s Federally AdministeredTribal Areas (FATA) of Kurram, NorthWaziristan, and South Waziristan. Ninety-onepercent of the population is Pashtun, with mostof the rest being Tajik. The Paktia tribes havehistorically had a “special relationship” withKabul that provided dispensations such asexemption from military service, and tribalstructures remain strong. Paktia has limitedindustry, and major livelihoods are cuttingwood, smuggling, and labor migration to theGulf states. The provincial center is Gardez.Balkh Province lies in a strategic location innorthern Afghanistan, bordering Uzbekistan andTajikistan. It is a largely Tajik province, althoughwith a mix of other ethnic groups, includingPashtun enclaves. The provincial center Mazar-eSharif is the de facto political, economic, andadministrative hub of northern Afghanistan.Emerging relatively unscathed from the war,Balkh retains significant Soviet-era influence,including remnants of pre-war industrialdevelopment and commercial linkages with thecentral Asian states.Faryab Province also lies in northernAfghanistan, bordering Turkmenistan to thenorth. Faryab is one of two Uzbek-majorityprovinces, with the remainder of the populationmade up of pockets of Tajik, Pashtun, Turkmen,and other smaller ethnic groups. Althoughlocated on the ring road, the province hashistorically been an economic backwater, withlow-productivity agriculture, limitedhorticulture, and virtually no industry. Theinsecure, Pashtun-dominated district ofGhormach officially lies in Badghis Province, butis temporarily being administered from Faryab’sprovincial center, Maimana, and was thereforeincluded within the scope of this study.Additional background information on theprovinces is available in the individual provincialcase studies. 403.2 Models used for employing aid in thefive provincesStabilization through aid projects was an overallstrategy of some donors, notably the U.S., andnot just through the PRTs. The study thereforelooked broadly at the effect of all types of aid onstabilization, with special focus on aidimplemented by or through the PRTs. Themodels employed by PRTs in each of the fivestudy provinces differed according to the leadnation with respect to strategies, objectives (e.g.,consent-winning, force protection,development), aid presence, levels of resources,civilian-military relationships, relationship withlocal and Kabul administration, role of QuickImpact Projects (QIPs), the use of CERP funds,relation with a political strategy (e.g., “inkspot”),41 and the extent to which they followedan integrated, comprehensive, or whole-ofgovernmentapproach.39As in other areas of Afghanistan, the assignment of certain districts to one province or another is not always clear. Gizab District contains asignificant Hazara population, but may or may not be part of Dai Kundi. For some matters it appears to be in Uruzgan, for others in DaiKundi.40The case studies are accessible at https://wikis.uit.tufts.edu/confluence/pages/viewpage.action?pageId=19270958.41The “ink-spot“ strategy is where military forces occupy a number of small, disconnected areas, and gradually extend their influence untilpockets (“ink spots“) became connected, leaving only small and isolated pockets of resistance.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 21

BOX 1: Provincial Reconstruction Teams (PRTs)Embodying the notion of civil-military collaboration, the first Afghanistan PRT wasestablished in Paktia Province by the U.S. in late 2002, and by 2011, 26 more had beenestablished throughout Afghanistan. Initially called “Joint Regional Teams” due to theirjoint civilian-military composition, they were intended to be an interim structure forfacilitating improvements in security and reconstruction and for supporting the extension ofcentral government authority to areas beyond Kabul. They were seen as a compromisebetween the expansion of ISAF proposed by President Karzai and the UN and the reluctanceof the U.S. to go along with a major expansion.The PRT terms of reference adopted in January 2005 identified a very general mission:“assist[ing] The Islamic Republic of Afghanistan to extend its authority, in order to facilitatethe development of a stable and secure environment in the identified areas of operations, andenable SSR [security sector reform] and reconstruction efforts.” 42 Over time, the agenda ofthe PRTs has broadened; in addition to doing short-term Quick Impact Projects, PRTs haveexpanded their roles in governance and developmental initiatives, either because NGOs andother civilian agencies could not operate in the insecure contexts of the south and east or forovertly political reasons. These roles involve promoting the capacity of local administrationthrough engagement with local stakeholders; promoting budget execution, businessdevelopment, agriculture, public health initiatives, and governance; and supporting thedelivery of basic social services. PRTs have also been used to strengthen troop-contributingnations’ interagency or interdepartmental efforts through harnessing all aspects of nationalpower and leverage—diplomatic, economic, reconstruction, and counterinsurgency efforts.Despite a common name and mission, the 27 PRTs have a varied structure and make-upreflecting the characteristics of the local area as well as the philosophies, legal restrictions,administrative arrangements, priorities, and available resources of the lead nation. There arecurrently 14 PRT lead nations, although most PRTs are shared by two or more nations. Anumber of PRTs cover more than one province. PRTs can also provide a base for non-ISAFpersonnel (e.g., Police Mentoring Teams, U.S. Army Corps of Engineers, DFID, U.S.Department of State, USAID, SIDA).PRTs have had strong political support in Washington, but have been controversial withboth the humanitarian community and the Afghan government. Within the humanitariancommunity, PRTs have been viewed warily as representing the militarization of aid (withthe risk that projects will be developed for short-term security rather than long-termdevelopment); potentially duplicating the work of humanitarian and development agencies;and, as putting at risk relationships with communities due to the “blurring of lines” betweencivilian and military actors, and the “shrinking of humanitarian space” for aid agencies tooperate in an independent and impartial manner. Lobbying by humanitarian anddevelopment agencies has resulted in improved dialogue and recognition of the conflictingneeds of these institutions, although fundamental issues remain.The Afghan government has often seen PRTs as a parallel and in some senses competingstructure that takes away legitimacy from the government and potentially prioritizes projectsBox 1 continued on next page42PRT Executive Steering Committee, January 2005. Cited in ISAF PRT Handbook, Edition 4.22Feinstein International Center

Box 1 continued from previous pageand activities that are not in line with Afghan government priorities. As a conference reportnoted, PRTs’ “experience as brokers of development has led to the formation of parallelgovernance and funding structures, which substitute government functions and formunintentional competition for legitimacy.” 43 In February 2011, President Karzai publiclydescribed PRTs as “structures parallel to the Afghan Government” that were “hindering theAfghan government’s development and hindering the governance of Afghanistan.” 44 Othercritics have noted “the flexibility of the concept has developed into an incoherent network oflead-nation-driven units that could run counterproductive to Afghan ownership.” 45 On thepositive side, PRTs provided a useful mechanism for smaller NATO allies to takeresponsibility for relatively small geographic regions. They also resulted in much moreattention and resources to sub-national levels, and, over time they contributed to a betterunderstanding among international personnel of Afghanistan beyond Kabul city. 46In Helmand, where due to insecurity theinternational military played a large role insupporting the Afghan government, UKgovernment institutions were somewhat dividedon the role of projects, especially QIPs, inpromoting stability. While there was consensusthat consolidation of Afghan state institutionswas the key to ultimate “stability,” it wasrecognized that this would take time and wouldneed to be augmented by international supportto provide security and “kick-start” at least someservice delivery. Consequently QIPs weregenerally seen as rapidly implementable projectsthat might “serve as down payments on promisesof political and economic progress.” 47Notwithstanding the areas of consensus on therole of projects in stabilization, each UKgovernment department had a different view ofthe utility, underlying purpose, and benefits ofthe QIPs program. The FCO tended to viewthem more as instruments of politicalengagement or strategic communication whilethe military tended to place great emphasis onthe role of both development and QIPs inparticular in consolidating tactical militarysuccesses. Most military interviewees argued thatreconstruction projects would deliver a morecooperative civilian population that, amongother things, would be more willing to shareintelligence information such as the location ofimprovised explosive devices (IEDs) and themovement of Taliban fighters. Aid projects wereused to try to “win consent” from localpopulations to work in a given area, 48 and theirprincipal benefits were often described in termsof short-term force protection. A particularimplication of this view of QIPS was that itraised the military’s expectations that civiliandevelopment and stabilization officials wouldrapidly follow the front-line troops andimmediately begin highly visible reconstructionor infrastructure projects. The military’s focus43Wilton Park, “Winning ‘Hearts and Minds,’” p. 14.44BBC News, February 8, 2011 at http://www.bbc.co.uk/news/world-south-asia-12400045.45Oskari Eronen, “PRT Models in Afghanistan: Approaches to Civil-Military Integration”. Crisis Management Center Finland, CivilianCrisis Management Studies, Volume 1: Number 5/2008. p. 1 (August 2008).46For more on the positive views of PRTs, especially in insecure areas, see Carter Malkasian and Gerald Meyerle, “ProvincialReconstruction Teams: How Do We Know that They Work?” (March 2009). http://www.strategicstudiesinstitute.army.mil/pubs/display.cfm?pubID=911.47Bennett et al., Country Programme Evaluation, Afghanistan. Evaluation Report EV696. p. 15 (DFID, London, May 2009).48According to the UK Ministry of Defense, consent winning activities are “simple projects that gain consent of the local populace . . . tocreate a permissive environment. . . . These activities are rarely a long term solution but must be part of the overall strategic development.They are intended to gain the goodwill of the community in order to initiate the engagement required to identify, plan and implementlonger term programmes. Examples of consent winning activities include the provision of electricity and water, the removal of litter, theopening of markets and the repairing of roads.” United Kingdom, Ministry of Defense, British Army Field Manual: Countering Insurgency,Vol. 1, Part 10 (January 2010).Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 23

Photo courtesy Australian Government Department of DefenceAustralian military on patrol, Uruzganwas clearly on the quick delivery of visibleprojects in quantities sufficient to signal acommitment to a community by ISAF and thegovernment of Afghanistan. The UK militarypressed hard for existing restrictions on the useof money to be relaxed and for commanders tohave the same levels of autonomy and delegatedauthority as their U.S. military colleagues.For DFID, QIPs were frequently described asmechanisms for ensuring rapid delivery ofcommunity-based programs that could serve as abridge to future and more sustainabledevelopment initiatives. As an aid instrument,QIP-type “projects” were generally recognizedas having their place, potentially offering a rangeof benefits: enabling donors to target their effortsmore specifically than they could using onlybudgetary support measures;enabling work through a widervariety of implementingpartners; enabling work insituations of poor governmentcapacity; providing anopportunity to work outside ofgovernment institutions; andlimiting elements of fiduciaryrisk. 49In Uruzgan, until the fall of2010, the PRT was led by theNetherlands but supported byAustralian and U.S. militaryforces, diplomats, and aidofficials. The Dutch supportedsome longer-term developmentefforts, but also invested muchof their aid resources toinfluence conflict dynamics andpromote stabilizationobjectives. By viewing theirengagement in Uruzgan as areconstruction rather than afighting mission, in partbecause that was more palatableto the population at home, theDutch approach to stabilizationtended to focus on addressinglocal grievances that werefueling conflict rather than on defeating theenemy or convincing insurgents to join thegovernment. This approach was characterized bythe employment of highly experienced politicaland cultural advisers who conducted detailedanalyses of local conflict dynamics and sought toshape and support positive forces with the use ofaid. The approach also relied somewhat on thenotion of “plausible deniability”—that some aidfunds for local influential persons or tribal groupswho were disgruntled with the government orsympathetic with the Taliban were useful, butonly if they were not visible. This meant thatsome aid was given discreetly, through smallscale“under the radar” projects with nobranding or taking credit and with littleoversight. The Dutch military also made smallsums available for civil-military cooperation49N. Leader and P. Colenso, “Working Effectively in Conflict-Affected and Fragile Situations,” Briefing Paper E: Aligning with Local Priorities,DFID Practice Paper (DFID, London, March 2010), p. 5.24Feinstein International Center

(CIMIC) work, which focused on forceprotection by seeking to win local support foroperations. As noted above, the Uruzgan PRTwas shared with the U.S. and Australians, whoseapproaches were sometimes at odds with theDutch. According to one interviewee, while theU.S. tended to think in terms of “high-valuetargets,” the Dutch thought in terms of “highvaluefacilitators.” 50 As a consequence, someAustralian and U.S. military personnel expressedconcern that the Dutch were too soft on theTaliban and were being politically naïve.Similarly, the Dutch had actively campaigned toget UNAMA to establish an office in Tarin Kot,in contrast to Gardez, where the (US) PRTinteracted much less with UNAMA. TheAustralians tended to focus more directly onreconstructing or building new infrastructure.The Australian Defense Force (ADF) and theAustralian Agency for International Development(AusAID) were also active in Uruzganimplementing and supporting stabilizationactivities and capacity-building projects in theareas of health, education, agriculture, water, androads. The geographic focus of their activities wason the areas north of Tarin Kot in the NorthDorafshan-Baluchi Valley and in Chora District,as well as in the Mehrabad area east of Tarin Kot.The Australian stabilization projects, jointlyplanned by ADF and AusAID officials, wererelatively small (less than $10,000), short-termprojects often directly implemented by the ADFand designed to legitimize initial interactions ofthe ADF with local community leaders ininsecure areas. If the security situation permitted,AusAID would start playing more of a lead role inintroducing more medium-term projects intendedto increase the legitimacy of the Afghangovernment by strengthening the capacity of linedepartments such as the Ministries of PublicHealth, Education, and Rural Rehabilitation andDevelopment to provide services. In more stableareas, ADF engineers continued to play a role inmanaging and overseeing infrastructure projectssuch as schools, clinics, and irrigationinfrastructure. Some interviewees noted that thisapproach was a very expensive way to buildinfrastructure, but that the strong ADF oversightrole resulted in better quality structures and fewerallegations of corruption. A specific area of focuswas vocational training and skills development forthe construction sector provided through theTrade Training School operated by the ADF atthe PRT.In Paktia, the PRT, like other U.S.-led PRTs,relied on a mix of funding, including CERP (seeBox 2), USAID, U.S. Department of Agriculture(USDA), and other military sources. Due to theinsecure environment, and far-greater humanresources and logistical capacity, the militaryplayed a much more central role than their U.S.government civilian counterparts in selecting,implementing, and managing the large andcomplex portfolio of aid projects that wereplanned or underway (approximately US$90million in 2009–10 alone). PRT-funded activitiesprimarily focused on infrastructure projects suchas roads, schools, clinics, district offices, andother government buildings. The centralobjective of these projects was to help strengthenconnections between the government and thepeople in order to help build public confidencein the government at the expense of insurgents.PRT teams visiting the fourteen districts ofPaktia would work with provincial and districtofficials to identify potential projects; theengineering team would then conduct afeasibility study, develop a scope of work, solicitproposals, review bids, award a contract, andmonitor implementation. This posed manydaunting challenges given the large amount offunds to be spent, insecure environment, limitedcapacity of contractors, language and culturalbarriers, and the capacity constraints andlogistical challenges of trying to monitor somany projects. With the objective of “extendingthe reach of the government,” PRT personnelwent to considerable effort to involve provincialand district officials in identifying projects. Theyalso involved the Provincial DevelopmentCommittee (PDC), although this tended toresult in long wish lists of activities largelyfocused in and around the provincial center,Gardez, where most PDC members lived. ThePRT also prepared short lists of construction bidsfor projects and then involved the governor inthe process of selecting and awarding thecontracts. While intended to demonstrate that50Interview with international analyst, Tarin Kot, Uruzgan Province, January 20, 2009.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 25

the government was now in the driver’s seat indecision-making, interviews indicated that thiswell-intentioned initiative may have backfiredand served to delegitimize the government, asthe governor was widely viewed to be receivingkickbacks on the contracts being awarded.BOX 2: CERP fundingFirst developed in Iraq, where the Coalition Provisional Authority used Iraqi funds and assetsseized prior to the invasion to establish a development fund, the Commanders EmergencyResponse Program (CERP) was intended to enable military commanders in both Iraq andAfghanistan to respond to urgent, small-scale, humanitarian relief and reconstructionprojects and services that immediately assist the indigenous population and that can besustained by the local population or government. In Afghanistan, PRTs first received CERPfunds in 2004, and during the period FY2004 to FY2010, the U.S. Congress allocated $2.6billion to CERP, with $1 billion in 2010 alone.CERP has enjoyed enormous support within the U.S. military. Robert Gates, formersecretary of defense, reinforced this when testifying before the Senate AppropriationsCommittee by arguing that CERP funds “can be dispensed quickly and applied directly tolocal needs, [and] they have had a tremendous impact—far beyond the dollar value—on theability of our troops to succeed in Iraq and Afghanistan. By building trust and confidence incoalition forces, these CERP projects increase the flow of intelligence to commanders in thefield and help turn local Iraqis and Afghans against insurgents and terrorists.” 51 According tothe U.S. Army, “CERP provides a quick and effective method to institute an immediatepositive impact on the Afghan people. The keys to project selection are: (1) quickexecutability; (2) local national employment; (3) benefit to the Afghan population; (4) highvisibility to the local populace; and (5) sustainability by GIRoA [Government of the IslamicRepublic of Afghanistan].” 52Regulations governing the type of expenditure have been extremely flexible. The U.S. GAOrecords that the undersecretary of defense’s (comptroller) guidance authorizes 19 differentusages “including transportation, electricity, and condolence payments.” The funds couldalso be linked with other U.S. programs such as those funded by USAID. This flexibility isreflected in the delegated funding authority of PRT commanders, with authorization tospend up to $25,000 per project before seeking approval from higher authority. The centralcommander in Afghanistan may authorize projects up to $2 million. As overall funding hasrisen, so has the size of individual projects.Most ISAF troop-contributing nations followed the U.S. lead and sought to develop quickimpact funding of one sort or another, although with considerably less money. However, theapproaches have tended to differ and often reflect national policy frameworks for using themilitary in development and disaster relief work. The Nordic states, for example, tend tohave national policies that restrict the role of the military in these areas.Development actors and policy makers have criticized CERP for the lack of controls onfunding, for being donor-driven, and for being unsustainable. Sen. Claire McCaskill, chairBox 2 continued on next page51Defense Secretary Robert M. Gates, Testimony to U.S. Senate Committee on Appropriations, February 2007. See, http://usacac.army.mil/cac2/call/docs/10-10/ch-8.asp.52U.S. Forces-Afghanistan, Money as a Weapon System-Afghanistan (MAAWS-A), USFOR-A J8 Publication 1-06 (15 May 2009), Appendix B-1.26Feinstein International Center

Box 2 continued from previous pageof the Senate Homeland Security and Governmental Affairs Committee’s contractingoversight panel, has been vocal in pressing for greater accountability and arguing that theprogram has evolved from the original concept of small-scale, rapidly implemented quickimpact projects into a major infrastructure program. She argues in the Military Times thatthere is “a disconnect between what the commanders in the field want to have happen andwhat actually happens,” and asserts that CERP is in danger of being substituted for longtermreconstruction programs, but without oversight and effective management. 53 TheJanuary 2011 report of the Special Inspector General for Afghanistan Reconstruction notesthat “27 of the 69 CERP projects it examined in Laghman province were at risk or havequestionable outcomes.” 54 The report warned that $49.2 million was at risk of being wastedin that province alone.In Balkh, the PRT, run jointly by Sweden andFinland, was responsible for the four northernprovinces of Balkh, Jawzjan, Samangan, andSar-e Pol. The PRT also functioned as a base forcivilian representatives of the U.S. Departmentof State and USAID, and for SIDA DevelopmentAdvisors. The Swedish approach differs from thecounter-insurgency approaches pursued byseveral other troop-contributing states in thesouth. Virtually all of the $60 million thatSweden gives Afghanistan is channeled throughthe central government, although 20 percent ofthat is earmarked for Balkh, Jawzjan, Samangan,and Sar-e Pol through a specially designated“Northern Fund.” As of 2009, Sweden also had aseparate $1.4 million annual allocation availablefor the north to support development projects,plus an additional $1.4 million for private sectordevelopment. All this aid is aimed at twoobjectives: building the long-term capacity ofPhoto by Paul FishsteinDestroyed bazaar, Balkh53Michelle M. Stein, “Lawmakers Question CERP Funds in Afghanistan,” Military Times (August 8, 2011), http://medilldc.net/2011/08/lawmakers-question-cerp-funds-in-afghanistan/.54Special Inspector General for Afghanistan Reconstruction, “Quarterly Report to the United States Congress” (January 2011).Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 27

the government while in the meantime helpingto ensure services. Unlike most other PRTs, theSwedes separate development and militaryfunctions, and their funds are administeredthrough civilian channels in Stockholm, Kabul,and Mazar. While the development and militaryfunctions are separate, collaboration acrosscivilian and military institutional channels isencouraged. The overall budget for Finnishassistance to Afghanistan in 2009 wasapproximately $15.2 million, of which $695,000(5 percent) was allocated to PRT projects. LikeSweden, Finland puts its funding largely throughthe central government, using various trust fundmechanisms. Separate from the PRT butoperating in the same geographic area, U.S.forces located at the Deh Dadi airbase just westof Mazar and at other area bases were providingtraining for the ANSF, and were supportingactivities with “humanitarian” funds and goodsin-kind.The Balkh PRT oversees three small-scaleproject funds, one funded by the Finns andlargely focused on QIPs such as support toANSF (e.g., for concertina wire), gravel roads,the Shrine of Hazrat Ali in Mazar, a microhydroelectricturbine in Sholgara, medicalsupplies, and school books. Another is fundedby the Swedish military intended entirely forwinning hearts and minds and creatinggoodwill for soldiers. A third, the “chaplain’sfund,” is supplied in part by money paid forrefreshments in the PRT canteen and personaldonations by soldiers. The chaplain’s fund is notexplicitly for winning hearts and minds, but isused for vocational training and other activitiesin Dawlatabad and Khulm with a focus onwomen and children. However, because thefund supports activities in communities, whichare similar to hearts and minds activities, it mayaffect communities’ perceptions of the military.Unlike most PRTs, the Norwegians in Faryabdo not have CIMIC projects or a CIMICadvisor. Norwegian development funding islargely channeled through national programssuch as the NSP and the Education QualityImprovement Program (EQUIP), 55 whileassistance in Faryab itself is channeled mostlythrough international NGOs. The PRTtherefore directly funds a very limited numberof aid activities. The reluctance on the part ofthe Norwegian military to conduct quickimpact projects partly reflects national policieson limiting the role of the military in aid workbut also the negative experiences prior to June2007. The event that led to this policy shift wasthe June 2007 visit by the deputy ministers offoreign affairs and defense to the hospital inMaimana, which had been refurbished by thePRT. This project had a number of problemsstemming from the PRT’s lack ofunderstanding of the local context and lack ofexperience in implementing developmentprojects. The principal consequence has beenthat Norway contracts through NGOs. Forexample, an international NGO was contractedto rehabilitate a prison in the province throughsub-contracts to private contractors.In all of the provinces, in addition to theprojects described above, other military unitsand civilian institutions undertook “hearts andminds” type projects. Military PsychologicalOperations, or “PsyOps,” had a small fund touse to change attitudes. Projects includedpainting mosques to encourage religious leadersto preach positive things about ISAF, givingpencils and other supplies to teachers to makethem look good, or providing radios to villageelders to enhance their status. PsyOps had alsoprovided funds for loudspeakers at mosques,lights at Afghan National Police (ANP) checkposts, and jackets for the Afghan NationalArmy (ANA) and ANP. Respondents describedthe PsyOps perspective very much astransactional, i.e., “relationship building”activities for which “they owe us one.” As onemilitary official put it, “all of this is bribes, sothat they talk.” 5655Begun in 2004 and expanded in 2009, the World Bank-funded EQUIP program aims to improve both access to and quality of basiceducation by providing grants directly to communities for the rehabilitation of school buildings and for obtaining teaching andlearning materials.56Interview with international military official, Mazar-e Sharif, Balkh Province, January 15, 2010.28Feinstein International Center

4. Drivers of conflict and insecurityISAF Photo by U.S. Navy Mass Communications Specialist 1st Class Eric S. DehmU.S. military and Afghan National Police, FaryabGiven the assumed link between lack ofdevelopment and insecurity, the study first triedto understand the drivers of insecurity in the fiveprovinces in order to assess whether aid projectswere addressing them. When asked about thedrivers of conflict or insecurity in their areas,while respondents identified mostly similarfactors, a number of notable differences existedamong provinces in the weight respondents gaveto the various factors. 57 The main reporteddrivers of conflict or insecurity were poorgovernance, corruption, and predatory officials;ethnic, tribal, or factional conflict; poverty andunemployment; behavior of foreign forces(including civilian casualties, night raids, anddisrespect for Afghan culture); competition forscarce resources (e.g., water, land); criminalityand narcotics (and counter-narcotics); ideologyor religious extremism; and, the geopoliticalpolicies of Pakistan and other regional neighbors.As discussed below, many of these factors arecomplex, intertwined, and overlapping, soisolating the strength and influences of each asseparate analytical factors is difficult.4.1 Corruption, poor governance, andpredatory governmentA widely reported cause of insecurity in all fiveprovinces was the poor quality of governance, inparticular the corrupt and predatory activities oflocal officials and police, as this was described asalienating the population and providing anopportunity for the Taliban. In the most insecureprovinces of Helmand, Paktia, and Uruzgan itwas reported as among the most importantfactors. The belief that the government (and itsinternational supporters) had failed to deliver ongovernance took a number of different forms.Consistent accusations included obtainingpositions and contracts through nepotism,favoritism, and bribery rather than through57While the study focused on insurgency-related insecurity, during interviews a broad, flexible definition of “insecurity” was used.This was justified due to the unclear lines between political and criminal insecurity, and between insecurity that originated eitherwith non-state actors or with the state itself.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 29

merit; flagrant and extensive abuse of authorityin extortion and in the illegal occupation andre-selling of land; arbitrary detention both as asource of revenue and as a way to neutralizeone’s rivals; distribution of patronage toreinforce one’s own position and marginalizeothers; control and protection of narcoticsproduction and trade as well as other illegalenterprises; and, maintenance of illegalcheckposts on the road to collect revenue. Asone UN official in Uruzgan put it, “it’s abusiness not a government.” 58Described as especially galling was that corruptgovernment officials were often the sameunsavory individuals who played a deeplyalienating role in previous eras (e.g., 1992–6)—and who, to widespread applause, were swept outby the Taliban regime. In some areas the currentgovernment is seen as largely composed of thesame local strongmen (e.g., Sher MohammadAkhundzada in Helmand, Jan Mohammad Khanin Uruzgan, Mohammad Atta in Balkh) whoplayed a role in past corrupt regimes, and whowere allowed to reinstate themselves post-2001.In Helmand and Uruzgan, for example, a smallset of jihadi leaders who had been driven out bythe Taliban regime had re-emerged as Karzai’sallies in 2001 and promptly took advantage oftheir new political power and institutionalpositions to capture patronage positions anddevelopment funding and to systematicallyexclude and marginalize their rivals. This wasdone through their personal militias as well asgovernment security institutions.The free rein of corrupt local officials, especiallywithin the ANP and the justice system, wheremost believe that favorable outcomes are onlyavailable if one is powerful or willing (and able)to pay large bribes, has reinforced the sense ofimpunity and lack of redress for ordinarycitizens. As a group of tribal elders in Paktianoted, “today’s laws are just for poor people.” 59The release of powerful commanders accused ofviolent and other criminal acts was cited inFaryab as a source of frustration. In Paktia, asone elder put it,Paktia has lots of problems, but the issue of lack ofclinics, schools and roads are not the problem. Themain problem is we don’t have a good government. .. . Without a clean government, millions of dollarsare stolen. If you increase the amount of money itwill also be useless because the government willsimply steal more. There’s a growing distancebetween the people and the government, and this isthe main cause of the deteriorating securitysituation. 60In Uruzgan, according to an internationalofficial, “the lack of accepted and effectivegovernment is the most important cause ofinsecurity. . . . If the local community has notrust in government then they will turn to theTaliban.” 61In Faryab, the appointment process for officialpositions was described as having played aparticularly destabilizing role. In large part thiswas because the dominant (Uzbek) politicalparty, Jumbish, has often been at odds with acentral government seen locally as Pashtundominated.As a result, appointments were oftenseen as intentionally undermining Jumbish’sposition and destabilizing the province. Inaddition, while the current governor, Abdul HaqShafaq, an ethnic Hazara from Sar-e PolProvince, was described in more positive terms,the provincial administration was said to rely on“factional networks, wasita (connections),nepotism, cronyism, and bribery in makingprovincial and district appointments,” 62 whichmeant that power and influence is mostly in thehands of Jumbish supporters.As would be expected, in all five provincescorruption was said to take place according tothe tribal and ethnic lines described below,underlining the overlapping nature of the drivers58Interview with UN official, Tarin Kot, Uruzgan Province, February 2, 2010.59Interview with tribal elders, Paktia Province, January 11, 2009.60Interview with local community leader and former mujihadin commander, Paktia Province, January 10, 2008.61Interview with PRT civilian official, Tarin Kot, Uruzgan Province, January 31, 2010.62Geert Gompelman. Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan’s Faryab Province (FeinsteinInternational Center, Tufts University, January 2011), p. 3.30Feinstein International Center

of insecurity. A consistent theme in Helmandand Uruzgan was that a select group of triballyaffiliated strongmen took advantage to securegovernment positions and gain access togovernment and development funds toconsolidate political and economic power amongtheir own people. In Faryab, given thecorrelation between ethnicity and politicalparties, the relations between ethnic groups isalso intertwined with party politics, especiallygiven the dominant role of the Jumbish party. InBalkh, perceived ethnic bias and bad governancesometimes came together as a source ofgrievance; cases were reported where Pashtunrefugees or IDPs had returned to their homeareas but due to bias were unable to get supportfor reclaiming their land, which has alienatedthem from the government.While most emphasis was on “administrativecorruption,” many respondents also drewattention to “moral corruption,” often inreference to how Western liberal values hadseeped into a government and society that wasnot ready for them. Even “democracy” was citedas a source of conflict; in the words of Paktiaelders: “now we have ‘democracy.’ People dowhatever they want—loot, steal, and only thinkabout themselves,” 63 or “in the name of‘democracy’ people wear jeans, eat kabobs, anddrink, but the culture and traditions of thiscountry don’t permit this kind of thing. Peoplein government think that wearing jeans anddrinking wine equals democracy.” 644.2 Ethnic, tribal, and factional issuesEthnic and tribal issues were mentioned in allfive provinces as a significant driver of insecurity,although this varied in form and intensity fromplace to place. In the insecure areas, much of theconflict is due to local tribal dynamics, which isthen labeled “Taliban” versus “government.” InHelmand, tribal issues were given first place, dueto the post-2001 “carve-up” of institutions andthe consequent on-going concentration ofpolitical and economic power, includingpatronage and development resources, in thehands of a limited number of tribal groups. Inparticular, the Zirak tribes (including PresidentKarzai’s Popalzai, as well as the Barakzai,Achakzai, and Alikozai) were the clear winners,Photo by ©UK MOD Crown Copyright 2011Joint PRT/Afghan National Police patrol, Lashkar Gah63Interview with tribal elders, Paktia Province, January 11, 2009.64Interviews with Paktia elders, Paktia Province, January 12, 2009. See Anna Larson, “Toward an Afghan Democracy? ExploringPerceptions of Democratisation in Afghanistan,” Discussion Paper (Afghanistan Research and Evaluation Unit, September 2009). Also,“Afghans Need to Find a New Model of Democracy” (October 28, 2009), Financial Times, www.ft.com/cms/s/0/32d3b638-c362-11de-8eca-00144feab49a.html.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 31

and benefited at the expense of the smallerPanjpai tribes, which were the clear losers. 65 Thisimbalance created resentment and conflict andundermined prospects of an inter-tribal politicalsettlement within the province. At the sametime, the political cover offered by strongmanAkhundzada’s links with President Karzai andthe U.S. military, which saw him as useful in thefight against the Taliban, made him untouchable,leaving those who suffered at his hands with nooptions for redress. All of this made thepowerbrokers’ rivals open to Taliban infiltrationand offers of protection. 66In Uruzgan, the key conflicts also revolvedaround inter- and intra-tribal power struggles.As in Helmand, many of those who had beendeposed by the Taliban and then reinstatedpost-2001 resumed their predatory behavior andpursued their rivals in other tribes, in this caseabetted at times by U.S. special forces in theguise of pursuing former Taliban officials as partof the “war on terror.” Former governor JanMohammad Khan (killed in an attack on hisresidence in Kabul in July 2011), the mainPopalzai representative in the province, tookadvantage of his jihadi commander networks,position as provincial governor, close personalrelationship with President Karzai, and status astrusted partner of the U.S. in the “war on terror”to exert tight control over the province. He wasadept at appointing his loyalists to serve in keypositions such as district governors, police chiefs,other government officials, and as members ofirregular armed groups. Much of this involvedstrengthening the Popalzai’s position in theprovince. Many of his rivals were forced to leavethe area after being targeted by local authoritiesor international forces (acting on the basis of“tips” received from local informants that theywere either Taliban or involved in the narcoticstrade). Or they simply became disgruntled thatthey were marginalized and not brought intothe government network or sought for advice,as people of their social standing would expect.Some groups (e.g., Ghilzai 67 elders) were in factpro-government, until they were labeled as“Taliban.” Often, targeted individuals werepowerless to deal with their rivals, who were inpositions of power. With no other avenue forredress, and in a culture that requires revengeto maintain personal honor, they consideredjoining the insurgency as their only option,both out of resentment and out of the need todefend themselves and their honor. Accordingto one international official, “When JMK [JanMohammad Khan] was governor, hedestabilized the Ghilzai to such an extent theyhad to go to the Taliban for security against thegovernor. There didn’t use to be lots ofTaliban—JMK is responsible for creating somany.” 68This illustrates the overlapping nature of thedrivers of insecurity; while these abuses occurredalong tribal lines, they were perpetrated by actorsserving in government positions and thus are alsoexamples of poor governance. As one respondentnoted, “in the southern provinces the problem isbetween the tribes, not between the governmentand the Taliban. . . . The tribal problem wasalready there but the government turned it into abig problem.” 69In Paktia, both ethnic and tribal issues were citedas sources of insecurity. The perceiveddisempowerment of the Pashtuns in general andPaktia Province in particular in favor of otherethnic groups and areas, especially in the earlydays of the Karzai government, was given as areason why people were not supporting thegovernment. A similar grievance frequentlyexpressed in interviews was the perceived lack ofsupport for Paktia tribes such as the Zadran,65Tribal Analysis Center, “The Quetta Shura: A Tribal Analysis” (Williamsburg, VA: Tribal Analysis Center, October 2009), p. 14. athttp://www.tribalanalysiscenter.com/PDF-TAC/Quetta%20Shura.pdf. Cited in Winning Hearts and Minds? Examining the Relationshipbetween Aid and Security in Afghanistan’s Helmand Province, Stuart Gordon (Feinstein International Center, Tufts University, April 2011).66The Helmand case study focused on the UK’s stabilization program between 2006 and 2008, and therefore reflects the situation up to thattime. Akhundzada served as governor from 2001 to 2005.67Largest of the four Pashtun tribal confederations, the Ghilzai are primarily located in the east, and so are a minority in the south.Historically they been rivals of the Durrani confederation, of which Jan Mohammad Khan was a member.68Interview with international civilian official, Tarin Kot, Uruzgan Province, January 31, 2010.69Interview with Afghan employee of USAID contractor, Tarin Kot, Uruzgan Province, July 6, 2009.32Feinstein International Center

which historically was the dominant local tribeand enjoyed a special relationship with Kabul.Another complaint in Paktia was about the lackof support for the arbakai, 70 which was consideredemblematic of a broader marginalization oftraditionally influential Pashtun tribes.In Balkh, which is dominated by thepredominantly Tajik Jamiat-i Islami politicalparty, regionally headed by Governor AttaMohammad Noor, ethnicity was cited as anissue, but in a very different way. The threeinsecure districts of the province arepredominantly Pashtun, and while mostrespondents gave economic reasons forinsecurity, the insurgency is still seen asessentially ethnic. Insurgents from southernAfghanistan are believed to be using ethnic,kinship, and smuggling networks to establish apresence in the north and win over sections ofthe population to their cause. Respondents alsonoted the perception that Pashtun communities,especially those that had returned from Pakistan,were generally more conservative and cited caseswhere they had become embroiled in disputeswith their neighbors over social issues such asgirls attending school. At the same time, manynoted a widespread belief among Pashtuns aswell as others that Pashtuns were beingdiscriminated against both by other ethniccommunities and by the authorities. Cases werecited where, due to ethnic bias, Pashtunsreturning from Pakistan were obstructed byauthorities in reclaiming their land. Such caseswere seen to lead to alienation, with thedispossessed groups more open to antigovernmentactivities. According to oneinternational military official, “the Pashtunpockets—they feel somewhat left out of thegovernment. They have ideological views, butmainly they are just frustrated with thegovernment. They are often treated unfairly.” 71Some respondents attributed the relative lack ofconflict in Balkh in part to the dominanceexerted by Governor Atta, 72 whosecomprehensive control over the provincialadministration and the security forces leaves littleroom for others to contest power—and henceless room for violence. 73In Uzbek-majority Faryab, much of the historyof conflict revolves around ethnicity. After 2001,many Pashtuns were forced out of the province,either to internally displaced person (IDP) campsor to Pakistan, in retaliation for harsh treatmenttowards Uzbeks and Tajiks during Taliban rule.Their return to their home areas in the provincehas caused tension, which is given added strengthby asymmetric perceptions: Pashtuns claim thattheir areas are underserved, while non-Pashtunsbelieve that the Pashtuns are there due to acentral government that is dominated byPashtuns and therefore supports them. TheTaliban are said to be exploiting thesegrievances, at least to get unhindered presenceand mobility in the area, although sometimesbowing to the demand from locals that they notcommit acts of violence there.In Faryab, factional politics were also cited as animportant driver of insecurity, although due tothe high correlation between ethnicity andpolitical parties, ethnicity and factional politicsare intertwined and hard to separate. Thedominant political party in Faryab is thepredominantly Uzbek Jumbish-i Milli, led byGeneral Abdul Rashid Dostum. The failure ofthe major post-2001 disarmament programs,Disarmament, Demobilization, andReintegration (DDR) and Disbandment ofIllegal Armed Groups (DIAG), to substantivelyweaken local commanders, as well as theperceived re-emerging threat of the Taliban, hasresulted in militia leaders and commanders70Tribal security forces. Although arbakai are indigenous to the southeastern Loya Paktia region (encompassing the provinces of Paktia,Paktika, and Khost), the term has been informally adopted to refer to irregular local militias in other parts of Afghanistan. The arming,organizing, and empowering of tribal groups as self-defense or counterinsurgency forces has been one of the more contentious policyissues in post-2001 Afghanistan, largely due to the negative experience of the early 1990s. The complaints heard in Paktia about the lackof support for the arbakai were in interviews conducted prior to the more recent policy of promoting the establishment of local militiaforces, for the most part under the Afghan Local Police initiative.71Phone interview with international military official, Mazar-e Sharif, Balkh Province February 5, 2010.72See Dipali Mukhopadhyay, “Warlords as Bureaucrats: The Afghan Experience” (Washington, DC: Carnegie Endowment for InternationalPeace, August 2009).73In recent months, Mazar-e Sharif has experienced a number of violent events, including the April 1, 2011, demonstration and mob attackon the UNAMA compound in which at least twelve persons (seven foreigners and five Afghans) were killed, and the July 20, 2011,bicycle-mounted IED explosion in which four were killed.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 33

having extensive caches of weapons. As a result,a local conflict over resources in Faryab canquickly escalate into a major security situationbecause protagonists are likely to besimultaneously tied in to overlapping ethnic andpolitical networks supported by well-armedmilitias.Perhaps ironically, the degrading of tribalcohesion has been a source of conflict, but also alimiting factor in resolving what might havepreviously been simple disputes. In Faryab, it wasnoted that elders can resolve local conflicts andlimit Taliban incursions, but only if there is arobust, existent tribal structure and identity inthe area, which is not always the case.Individuals competing for influence mayexacerbate problems to serve their personalpolitical agendas. In Helmand, the fragmentationof tribal power (which has historical roots in thesettlement of different tribal and ethnic groupsafter Kajaki Dam was built and large swathes ofnew land irrigated in the Helmand Valley in the1960s, and more recently due to factors such asthe distortions caused by the narcotics industry)means that elders may compete for influence,with no clearly dominant person or group. Insome cases, tribal groups can contain bothpro- and anti-Taliban factions. In GhormachDistrict bordering Faryab and BadghisProvinces, lack of tribal cohesion was describedas a driver of insecurity, some of which resultedfrom Taliban struggles for leadership or divisionof spoils. Both inter- and intra-tribal splits existamong both pro-government and pro-Talibancamps, many of which are based on personal orcriminal competition.4.3 Poverty and unemploymentIn the two northern provinces of Balkh andFaryab, poverty and unemployment wereoverwhelmingly reported as the main causes ofinsecurity, with a dominant narrative being thatthe typical fighter joined the insurgencyprimarily because of economic necessity. InHelmand, Paktia, and Uruzgan, poverty was notgiven as the most important driver of insecurity,although it was said to make people morevulnerable to other factors. Especially in thenorthern provinces, poverty and unemploymentwere typically described as being “an open doorfor the Taliban” in that unemployed men couldbe enlisted in insurgent groups. As onerespondent stated, “Lack of jobs is the mostimportant cause of insecurity. If young peopledon’t get jobs and can’t go to Iran [for work],what will they do?” 74 Another asserted, “thosewho can’t go—how can they survive? Some aretempted by bandits or insurgents and given$100–500 to plant an IED [improvised explosivedevice]. They are not supporters of Taliban, butdo it for the money.” 75 Many used the languageof desperation, family responsibility, or honor,saying, “if a man has no other way to feed hisfamily, then he will do anything.” As a tribalelder in Paktia put it, “if someone has six peopleat home to support and has no money, he has tosupport insurgents to get money. The motivationto join the Taliban is not religious or tribal orethnic based but unemployment.” 76 In Balkh, thelack of compensation (in the form of developmentaid) anticipated as payment for eliminating opiumpoppy cultivation from 2007 was described ascreating even more motivation for people to jointhe insurgency; “people are hungry and can eithergrow opium or join al Qaeda and the Taliban. Weeliminated opium in Balkh, but we haven’tprovided livelihoods support for the people.” 77Even among those who didn’t completely acceptthe notion that poverty was the fundamental anddirect driver of insecurity, many stated that itmade people more susceptible to other factors.For instance, in Balkh, the migration of youngmen to the south to work on the poppy harvestwas associated with turning towards antigovernmentgroups. In Faryab, poor families inrural areas without schools were said to have noother option than to send their boys to74Interview with NGO staff member, Maimana, Faryab Province, January 14, 2009.75Interview with UN employee, Maimana, Faryab Province, January 14, 2009.76Interview with tribal elder and district governor, Paktia Province, January 20, 2009.77Interview with senior provincial official, Mazar-e Sharif, Balkh Province, January 2010. Balkh Province was declared “poppy-free“ in2007, a state that was attributed largely to the efforts of Governor Atta Mohammad Noor.34Feinstein International Center

unregistered madrassas either within the provinceor in Pakistan, where they were at risk ofbecoming radicalized.While poverty was frequently cited as anexplanation for insecurity, some skepticism iswarranted on the linear relationship givenbetween poverty and joining the insurgency.While poverty is a widespread condition inAfghanistan, insurgency is less so. Also, that therole of unemployment was seen as much strongerin the north suggests that it may be more relatedto criminality in some areas. This issue isdiscussed further in Section 6 below.4.4 International military forcesMore than on any other factor, the insecure andsecure provinces diverged in reported views onthe role of the international military in creatinginsecurity. In the three insecure provinces, thebehavior of the international military forces wasreported to be an important source and driver ofinsecurity. The lists of objectionable behaviorswere consistent: civilian casualties, night raids,house searches, population displacement,destruction of infrastructure, and aggressivebehavior. Many of these were summed up,especially in Paktia, as a disrespect for Afghanculture, religion, and traditions. As a formerTaliban commander and deputy of an Islamicparty said, “the mistakes of the Coalition arewhy aid projects have no security benefit. Whenthey bomb, kill innocent civilians, use dogs tosearch our houses and women, and even harmwomen and children, how can they expectpeople to support them?” 78 As a Westerneducatedbusinessman and former provincialgovernor noted, “people are slowly but surelycoming to the conclusion that they are anoccupied country. As a result of the bombings,house searches, being bitten by dogs, people arethinking that the U.S. is worse than theSoviets.” 79 In Uruzgan, as noted above, thetargeting of “Taliban” by international forceswho were unknowingly manipulated by localelders into settling scores with tribal rivals wascited as contributing to the dissatisfaction withthe international military; i.e., “people useinternational and national forces to pursuepersonal and tribal grievances.” 80 Arrests andraids have also resulted in the flight of influentialelders, which has had the further negative effectof weakening the social fabric—and leading tofurther insecurity.While in the two urban areas of Helmand,Lashkar Gah, and Gerishk, respondents were notquite as negative about the international forces asthey were in areas of the province seeing muchmore military activity, the overwhelmingmajority of respondents reported that ISAF’spresence had made the security situation worse.In Uruzgan, in some cases even having contactswith Coalition forces reportedly causedinsecurity. While the Dutch were givenrelatively high marks for their interactions,several Afghan interviewees noted that themore-aggressive operations of U.S. andAustralian special forces, especially night raids,were considered problematic. In and aroundTarin Kot, Afghans distinguished between theDutch “soft-knock” approach and the Australianmore aggressive one.In the two more-secure provinces, Balkh andFaryab, the international military forces were notreported to be the de-stabilizing factor they werein the insecure ones. In Faryab, respondents weregenerally positive about the PRT, due to itsculturally sensitive interactions withcommunities, the absence of civilian casualties,and resources that had been brought to theprovince. 81 In Balkh, there was less of an“allergic reaction” towards military actors andmore of a willingness to engage with them. No78Interview with former Taliban commander and deputy of an Islamic political party, Paktia Province, January 5, 2009.79Interview with former provincial governor, Kabul, January 5, 2009.80Interview with Afghan UN official, Tarin Kot, Uruzgan Province, July 6, 2009.81According to research by Christian Michelsen Institute, trust in ISAF decreased significantly between 2006 and 2009, although thisreferred to ISAF in general and not just the local (Norwegian) forces. See Arne Strand and Organisation for Sustainable Development andResearch (OSDR), “Faryab Survey: Comparison of Findings from Maymane 2006 and 2009” (Bergen, Norway: Christian MichelsenInstitute, 2009). p. 4. Public opinion polls have consistently shown that the Afghan population supports the presence of internationalforces. This should, however, be put in the context of “compared with what,” as it may reflect not so much a fondness for internationalforces as the fear that their withdrawal will lead to a situation similar to the civil war of 1992–6. In general, polling data on thepopulation’s view of the international military should be looked at carefully.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 35

doubt this pragmatism about engaging with themilitary was conditioned by a context in whichfew insurgents and little armed conflict existed.As noted by a former Afghan military officer, “Ifthe area is peaceful, then the people willwelcome the PRTs. If the area is insecure, thenpeople will be afraid.” 82 Similarly, an AfghanNGO official noted that “north and south aredifferent. In the north, people are not seen bythe military as enemies. There have been nocivilian casualties, so relations are better.” 83Even in Balkh, however, while respondentsgenerally did not ascribe insecurity to theinternational military forces, there were severalexceptions. First, aggressive and uncoordinatedactivities of U.S. military and Special Forceswere described as creating problems, an issuethat was mentioned by (non-U.S.) militarypersonnel as well as Afghan respondents. 84CIMIC personnel described working withcommunities over a period of months to buildconfidence, only to have it underminedovernight by a military operation they had noprior information about. Second, payments madeby the military to maleks (local leaders) were seenas de-stabilizing in the long term: while theymight achieve a short-term aim, they coulddecrease stability by creating jealousy andproviding perverse incentives for others todemand the same sort of payments. Third, roadblockages and closures due to patrols and othermovements of forces were reported as creating illwill. The influence of the effect of internationalmilitary forces on community perceptions isdiscussed in greater detail below.4.5 Religious extremismThe way in which respondents directly orindirectly connected religious and ideologicalissues with insecurity varied, and in many casestheir responses seemed to conflate religious andmoral principles as well as the groupsrepresenting them.First, religion and ideology were said to play arole in mobilizing young men to act against thegovernment and international forces andtherefore in legitimizing violence. In Helmandthis was largely seen to be done indirectlybecause it legitimized other, existing grievancessuch as alienation from the corrupt governmentand anger at the actions of international forces.This is not surprising, as complaints about theadministrative and moral corruption of thegovernment and its supporters were often put interms of religious principles. It was not alwaysclear where the sense of injustice ended and thereference to religious issues began, as inAfghanistan the sense of justice is informed byand closely linked with religious principles.Many of these principles related to the notion ofinjustice. (See Box 3.)Second, religion was connected with insecurityin that the Taliban were seen to be acting in thename of religion. Respondents varied in thedegree to which they gave legitimacy to theTaliban’s connection with religion or instead sawit as simply being instrumentalized by them;many respondents accused the Talibanthemselves of being un-Islamic, primarilybecause they embrace violence against civilians,and not necessarily due to their attacks on thestate or the international forces.Third, more extremist religious views wereassociated with insecurity in terms of conflictwith neighbors. As noted above, a number ofPashtun communities had returned to Balkh andFaryab from Pakistan or from IDP camps in thesouth; these communities were generally viewedas more conservative and dogmatic, and conflictswould therefore arise with neighbors over socialnorms (e.g., girls’ education). This dimension of82Interview with former Afghan senior military official, Mazar-e Sharif, Balkh Province, March 19, 2009.83Interview with Afghan provincial head of national NGO, Mazar-e Sharif, Balkh Province, April 8, 2009. Also, as noted in themethodology section, since the fieldwork was conducted, the security dynamics in the north have changed. For instance, there has beenan increase in night raids, which have the potential to change perceptions of the international military to those more similar to the otherstudy areas.84The confounding of the U.S. with other Western forces by some Afghan respondents is problematic. As the Americans are perceived to bemore aggressive, there may often be an assumption that any aggressive acts are done by the U.S. forces. On the other hand, the Americansare also perceived to have more money, so it may be assumed that any big spending is by the Americans.36Feinstein International Center

insecurity obviously differs from attacks againstthe international military forces or thegovernment. Government and international staffalso expressed concern about the politicalinfluence that might be exerted by a newgeneration of religious leaders educated inPakistan, as increasingly extreme sermonsdelivered by these leaders consistently stressedthe corrupt (and un-Islamic ) character of thegovernment and were even anti-government.It should be noted that if interviews had beenheld in less secure areas more weight wouldlikely have been given to religious issues. Also,analogous to what was noted in the discussion ofpoverty above, one could cite religiousobjections to the state yet still not commitviolent acts.4.6 Conflict over scarce resourcesIn all provinces, local conflicts over scarceresources (land, water, etc.) were described ascreating insecurity. Here again, however, theoften-overlapping nature of factors meant thatwhat might otherwise have been simple disputescould escalate into more complex and violentsituations. For example, weak institutions andcorrupt officials often alienated losing claimants,and rival claimants frequently stood on eitherside of ethnic and political divides. Numerouscases were cited of land being under disputefrom people returning from Pakistan or IDPcamps, and where rival claimants had competing(bogus) documents. As noted above, in Balkh, anumber of ethnically-tinged cases were reportedwhere Pashtun refugees or IDPs had returned totheir home areas but were unable to reclaim theirBOX 3: InjusticeOne theme that cut across numerous thematic areas was injustice and zulm (cruelty)perpetrated by people or groups with power, primarily those within or allied with the state.Respondents cited specific grievances such inequitable distribution of employment, contracts,and aid; illegal seizures of land and other property; increased economic inequality; andinequitable access to justice depending on one’s wasita (connections) or ability to pay bribes.Yet, all of these were simply examples of the broader theme of injustice.Generally, grievances included the strong sense that in post-2001 Afghanistan some peopleand groups had grown very wealthy at the expense of others, and that the nexus of corruptgovernment officials, powerful political leaders, and scheming businessmen (i.e., Kabul Bank,jihadi leaders) enjoyed apparent immunity. This was especially notable in discussions of theintroduction of the market economy; it was described as contributing to injustice byallowing corrupt and powerful figures to control and manipulate the economy at theexpense of common people. Officials were said to be directly implicated either throughtaking bribes or through their linkages with businessmen and traders. Notably, one of thefew aid programs positively described by respondents was the NSP, which employs a specificformula for the equitable sharing of benefits across communities.In Afghanistan, the notion of justice is informed by and closely linked with religiousprinciples, and so the sense of a lack of justice was often framed in religious terms. While theconcentration of political and economic power might be described in the West as “elitecapture,” Afghan respondents described it in terms of injustice. Therefore, many of thecomments about corruption and poor governance was expressed in terms of an un-Islamic orunjust distribution of benefits, which is one of the main themes of Taliban propagandaBox 3 continues on next pageWinning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 37

Box 3 continued from previous pageemphasizing the illegitimacy of the state and its leaders. The accusation that the state hadfailed to play a positive role in social welfare or to at least temper the worst excesses ofpowerful groups and individuals reinforced this theme.As discussed in Section 4.8 below, insurgents have been adept at taking advantage of thissense of injustice, in part because often people have nowhere else to turn but also because theTaliban were associated with sweeping out rapacious commanders and warlords in the 1990s.land. At the same time, cases were reported ofreturning Pashtuns being given the land ofothers by government officials. In Uruzgan,conflicts over land and water were frequentlycited as important sources of insecurity. Thesewere typically set off by a land grab or thedigging of new irrigation channels withoutgoing through traditional community consensusprocesses. As elsewhere, these conflicts couldquickly escalate into ethnic or tribal conflict. 854.7 Pakistan and the other “neighbors”The geopolitical policies of Pakistan and other“near neighbors” were cited as important driversof conflict. As one government official noted,“although part of the problem is internal—tribes—the problem is also external—Pakistan.The closer you get to Pakistan the moreinsecurity there is. We will continue to haveconflict as long as our neighbors continue tointerfere.” 86 The interference was seen as takingdifferent forms. The most commonly-cited wasmilitary, logistical, and political support (e.g.,safe havens) being offered by elements inPakistan and, to a lesser extent, by other regionalneighbors in backing factions that werecontesting political power. In Paktia, Iran wasfrequently cited in part due to the perceptionthat it was supporting the Dari press at theexpense of Pashtu, and was supporting Hazaraspolitically and economically. According to onerespondent critical of both near neighbors,there are lots of publications, TV serials and dramas,music that all have Iranian influence. It is intendedto create problems between Shias and Sunnis.Pakistan is the second problem. Pakistan is usingilliterate people, and trains them and sends them hereto implement their plans against the Afghangovernment and the international community.Pakistan doesn’t want this country to develop and forpeople to become educated. What we want is for ourgovernment to have a foreign policy that encouragesIran and Pakistan that our development is theirdevelopment, what is good for us is good for them. 87Both Pakistan and Iran were also accused ofhaving a destructive economic agenda manifestprimarily through predatory pricing of importssuch as cement to undercut Afghan production,tariffs on Afghan products, and the establishmentof “beachhead” economic activities withinAfghanistan. The usurpation of Afghanistan’swater resources, the sending of unfinishedcarpets to Pakistan where they were finished andsent abroad as Pakistani products, and thesmuggling of semi-precious stones to Pakistanfor polishing and other value-adding processingactivities were also seen as evidence of thedestructive anti-Afghan economic agenda.The radical religious agenda was also attributedto Pakistan, either due to madrassas impartingconservative, alien ideas and radicalization (e.g.,“Pakistani-style preachers”) or due to theenlistment of religious ideas and personalities bythe Pakistani security establishment.85Elsewhere in Afghanistan the competition between sedentary farmers and kochi (nomad) pastoralists over use of rangeland and pasturesresults in annual outbreaks of violence that have to be quelled by Kabul. Because the farmers are Hazara and the kochis are Pashtun, theconflict has acquired a sharp ethnic edge. See Liz Alden Wily, “Rural Land Relations in Conflict: A Way Forward,” Briefing Paper(Afghanistan Research and Evaluation Unit, August 2004). More recently, see Fabrizio Foschini; “The Kuchi-Hazara Conflict, Again”(Afghanistan Analysts Network, May 27, 2010), http://aan-afghanistan.com/index.asp?id=764.86Interview with provincial head of line ministry, Uruzgan Province, July 5, 2009.87Interview with former Taliban commander and deputy of an Islamic political party, Paktia Province, January 6, 2009.38Feinstein International Center

The playing out of global politics withinAfghanistan’s borders (e.g., the conflict betweenthe U.S. and Iran) was also seen as underminingsecurity. All of this was described in the contextof historical distrust, especially of Pakistan:Most security problems are caused by groups backedby the ISI/Pakistan (e.g., Haqqani). They use thefailures of the government and the internationalcommunity to gain support—the mistakes they aremaking cause people to support Haqqani. . . .Afghanistan is at a very important and sensitive timein its history. For three decades neighboring countrieshave tried to destroy Afghanistan. Russia thinks thenorth should be under its influence. Iran wantsinfluence in Dari speaking areas. Pakistan has theDurand line issue. 88For those respondents who had either beensympathetic to the Afghan government duringthe jihad years or who had later come to regretthe destruction visited on Afghanistan by thejihadi factions, Pakistan was seen as simplycontinuing its agenda of weakening Afghanistanfor Pakistan’s own political aims. As a group oftribal elders in Paktia put it, “it’s not Afghanswho don’t want schools and roads. It’s our mainenemies—Pakistan is sending people to stopreconstruction from happening. What can wedo?” 89Additional drivers of insecurity identified byinterviewees included narcotics and criminality.Many respondents described a symbioticrelationship between insurgents and criminals;while criminal elements used ethnic ties tofurther their enterprises, Taliban elementssupported criminals in order to contribute togeneral instability. In addition, commanders’continued abusive criminal activities (e.g.,illegal taxation, extortion) as well as thesometimes-violent spillover from rivalriesbetween commanders has created resentmentfrom communities. Again, these factors overlapand interact with the factors discussedpreviously.4.8 Opportunities for insurgents to exploitgrievancesWith respect to many of the above drivers ofinsecurity (especially corruption and poorgovernance, tribal conflict, competition forscarce resources, and counter-narcotics),insurgents are said to have been adept at takingadvantage of grievances to draw marginalizedand already alienated groups away fromsupporting the government and its internationalbackers and offering them support and protectionagainst the abuses. A common theme amongrespondents was that the growing strength of theinsurgency was due not so much to its appeal,but rather to the disillusionment and alienationresulting from government officials’ cruelty andavariciousness; i.e., to “push” rather than “pull”factors. A common formulation was that it wasn’tthat the Taliban were winning, but that thegovernment was losing. In all five provinces,respondents described how those who wereoppressed by the powerful had to seek protectionoffered by the Taliban in the face of unevenpoppy eradication, unjust court rulings,confiscation of contested land and otherresources, and general harassment. For instance,in Helmand, the Taliban were able to appeal tothose who had lost land with the message thatonly under a Taliban regime would they be ableto reclaim their rights. Similarly, they were ableto exploit dissatisfaction with the poppyeradication program that disproportionatelytargeted fields belonging to Noorzai and Ishakzaiwhile largely ignoring those connected withSher Mohammad Akhundzada. In Faryab, thespeculation is that insurgent groups havestrategically “branded” themselves as related tothe Islamic Movement of Uzbekistan, anUzbekistan-based militant organization, ratherthan Taliban so as to be acceptable tocommunities that would be unwilling to supporta Pashtun-dominated movement. To someextent, the Taliban’s ability to convincecommunities that they would be better off withthem derives from their original role in the1990s of sweeping out rapacious local88Interview with head of tribal shura, Paktia Province, January 5, 2009.89Interview with tribal elders, Gardez, Paktia Province, January 9, 2009.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 39

Photo by Paul FishsteinGraves and houses, Balkhcommanders and warlords; in that sense, theymay be seen as once again picking up the causeof justice. In Uruzgan, the resurgence of theTaliban was facilitated by the many importantTaliban leaders who were from the province orhad lived there—when the Taliban began to betargeted by the local administration and theinternational military forces, reviving a networkof commanders in response was easy. Reflectingthe complexity of the situation, in some cases,local strongmen would join the Taliban to keepother Taliban groups out.In sum, respondents described a range of factorsthat contributed to insecurity, includingcorruption and poor governance; ethnic, tribal,and factional conflict; poverty andunemployment; behavior of international forces;extremist religious ideology; conflict over scarceresources; and regional neighbors. While similarfactors were cited in all five provinces, theweight given to each varied somewhat betweenthe insecure and secure areas. The maindifferences were that economic issues were givenmore importance in secure areas, whilecorruption and bad governance and the behaviorof international forces were given moreimportance in insecure areas. Due to the oftencomplex,interrelated, and overlapping nature ofthese factors, isolating the strength and influenceof each in creating insecurity was difficult. Thisin turn suggested that stabilization projects thatfocused on only one or two drivers of conflict(e.g., unemployment) would find it difficult toplay an effective role in mitigating or resolvingmost conflicts.40Feinstein International Center

5. Perceptions of aid projectsPhoto by U.S. Air Force 1st Lt. Cale Reeves, Paktya Provincial Reconstruction TeamConstruction of clinic, PaktiaOne of the most powerful assumptionsunderpinning the belief in the stabilizing effects ofaid projects is that these projects are popular andcontribute to winning the support—or “heartsand minds”—of local communities. Initially, theintent in Afghanistan was to use aid projects for“force protection,” gaining the consent andsupport of local populations for the presence offoreign troops. As broader counterinsurgencyobjectives began to take hold, however, theobjective shifted much more explicitly to usingreconstruction and development aid to win thesupport of the population away from insurgentsand for the government. The study therefore triedto look at whether and how aid projects addressedthe drivers of insecurity identified by respondents,as discussed in the previous section.The research found that, rather than generatinggood will and positive perceptions of internationalmilitary forces, development actors, or the Afghangovernment, aid and development projects wereconsistently negatively described by Afghans. Thenearly universal stated perception was that aidprojects, whether implemented by military forces,aid agencies, or the government, were performingpoorly. Indeed, the consistently negative viewsvoiced by Afghan interviewees about aid projectsand implementers suggest that projects are notwinning people over to the government side.Perceptions of the misuse and abuse of aid werefueling growing distrust of the government,creating enemies, or at least generating skepticismregarding the role of the government and aidagencies.As noted in the methodology section above, giventhe natural human tendency to focus on what onedid not get rather than what one did, aperceptions-based study is likely to paint a morenegative picture of development assistance thanmay be warranted. But given the strongassumption that aid wins hearts and minds andleads to more positive perceptions, highlightingthe overall extremely negative perceptions thatAfghans expressed—fairly or not—about aidefforts is important.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 41

While respondents’ expressed perceptions variedsomewhat, some clear patterns and areas ofagreement emerged among aid agency staff,communities, and government officials, as wellas between Afghan and internationalrespondents. The chief complaints were thatprojects were insufficient, both in terms ofquantity (not enough) and of quality (wrongkind or poorly implemented); unevenlydistributed geographically, politically, andsocially; and, above all, associated with extensivecorruption, especially those with multiple levelsof subcontracting.5.1 “Nothing, or not enough, was done”A consistent refrain in all five provinces was that“nothing had been done.” While manyrespondents acknowledged someaccomplishments, most said that these were verysmall relative to the need. A typical responsewas, “Yes, people see that something has beendone, but. . . .” Respondents often referred tothe “large amounts” of international aid that hadeither been promised or was said to have beendelivered, and said that they could not reconcilethose large amounts with the limited evidenceon the ground. There was overlap between thecomplaint that nothing had been done and thecomplaints about corruption and the wrongtypes of projects, as discussed below.These complaints were often made whilesurrounded by evidence that much had in factbeen done. One researcher related being told byAfghan colleagues that nothing had been done,as they traveled on a newly paved road, passingnew clinics, schools, and district centers, all thewhile talking on a mobile phone. The aidcommunity in Afghanistan generally agrees that,especially in the first few years after 2001,expectations were raised unrealistically high bythe public relations and communicationsstrategies of most national and internationalinstitutions. Their default position was to useinternational and national media outlets tohighlight all the positive things being done, thegenerous donor contributions, the “successstories,” the ribbon-cutting ceremonies, etc. Thisgenerated significant cognitive dissonancebetween what Afghans heard over the Dari andPashtu airwaves about what was being given tothem and done for them and the tangible resultsthey actually saw on the ground. Few effortswere made to communicate to Afghans theenormous challenges and time needed to makesignificant progress in rebuilding a country thatwas among the poorest in the world, even beforeit was devastated by three decades of conflict.Challenges were underplayed and glossed overbecause they might undercut the “good news”narrative. 90 Communications strategies thatraised rather than lowered expectations helpexplain why, despite newly paved roads, schools,and clinics, extensive cell phone networks, avibrant media, etc.—none of which existedwhen the Taliban regime was removed frompower in the fall of 2001—Afghans frequentlyexpress the view that “nothing has been done.”As one NSP Community Development Council(CDC) treasurer in Balkh put it, “a clinic is OK,but they don’t give much medicine. People inAfghanistan will always say that you didn’t giveus anything.” 915.2 Inequitable distribution: “They gotmore than we did”Another consistent and related refrain was thataid had been inequitably distributed. ManyAfghan respondents complained that otherregions, ethnic groups, tribes, villages, orindividuals had received more than their fairshare of aid resources. This complaint was oftenexpressed as a general statement, but oftenmeant simply that others had received more.The only consistent allegation was that therespondent had not received a fair share.Respondents in Balkh and Faryab complainedthat their relatively secure provinces hadincurred a “peace penalty” by virtue of havinglittle conflict, and therefore donor resources hadbeen programmed in other, less-secure areas. InFaryab, respondents in communities close to theinsecure areas receiving aid (e.g., Ghormach90For discussion of the dysfunctional effects of the aid dynamic, see Astri Suhrke, “When More is Less: Aiding Statebuilding in Afghanistan,”FRIDE Working Paper (September 2006).91Interview with CDC treasurer, Khulm District, Balkh Province, June 10, 2009. See below for discussion of CDCs and the NationalSolidarity Program.42Feinstein International Center

District, which was the focus of developmentinitiatives with stabilization objectives) wereespecially vocal in complaining. Many in thenorth expressed dissatisfaction that violentplaces were getting the bulk of the assistance,and that this funding imbalance was setting upperverse incentives. As one group of elders inBalkh noted, “We see the situation in Khost,where there is lots of aid, and wonder if weshould try to attract that with tak o took [a bit ofnoise].” 92 An additional grievance was voiced inBalkh, where the cultivation of opium poppyhad been suppressed 93 but promisedcompensation in the form of aid projects hadnot materialized.Because most development funding is going toPashtun areas in southern and easternAfghanistan, the sense of regional deprivationoverlapped with an ethnic one. Non-Pashtunrespondents in the north attributed this to apro-Pashtun bias of the central government,whereas in reality it had more to do with theprioritization of insurgency-affected areas bymajor donors like the U.S. Even in the threeinsecure provinces, respondents’ complaintsabout the inequitable allocation of aid wereprimarily on an ethnic or tribal basis. Forinstance, the ethnic Hazara areas of Uruzganwere essentially ignored, as they had fewerinsurgents, which led a group of Hazara eldersfrom Gizab District 94 to visit Tarin Kot to askthe provincial governor to pay more attentionto their areas.Of course, that a high percentage of aid wasspent in the south and east did not keep peoplein those areas from complaining that they toohad been left out. As one respondent insoutheastern Paktia noted, “I’m happy thatthere have been some projects in the north ofthe country, but nothing has been done in thesouth.” 95 Or, as one respondent noted in afactually incorrect statement, “the PRTs arethere, but the amount of aid is less than in thenorth and west.” 96Of course, an objective basis underlies some ofthe complaints that “they got more than wedid.” While the variety of sources and differinglevels of disclosure make measuring volumes ofaid extremely difficult, according to USAID’sown figures, 77 percent of its resources in2009–10 were spent in the insecure areas of thesouth, southwest, and east, with a plannedincrease to 81 percent in fiscal year 2011. 97 Itshould be noted that this does not includemilitary funding. If military funding is alsoincluded, the imbalance is even more notable. 98While in many cases, the complaint that an areahas been neglected had a ring of truth, in manyother cases, no matter how much had beendone, other areas, communities, and householdswere inevitably perceived to have receivedmore—even when it was not objectively true.This mentality is exacerbated by anenvironment in which resources are especiallyscarce (and contested) and misleadinginformation abundant. Civilian and militaryagencies with development funds at theirdisposal “to win hearts and minds” often failedto adequately recognize that in a zero-sumpolitical environment such as Afghanistan anyone individual’s or group’s gain is oftenperceived to be at the expense of another; insuch an environment, providing developmentaid in a way that is perceived to be equitableand just and that generates public support isvery difficult.92Interview with village elders, Sholgara District, Balkh Province, June 3, 2009.93As noted above, Balkh Province was declared “poppy-free“ in 2007, attributed largely to the efforts of Governor Atta Mohammad Noor.94Originally part of Uruzgan, mixed ethnicity Gizab District became part of Dai Kundi Province when it was created in 2004, then wasre-attached to Uruzgan in 2007.95Interview with head of tribal shura, Paktia Province, January 5, 2009.96Interview with businessman from Paktia, Kabul, January 5, 2009.97Majority Staff, Committee on Foreign Relations, United States Senate, “Evaluating U.S. Foreign Assistance to Afghanistan,” (Washington,DC, June 8, 2011), http://www.gpoaccess.gov/congress/index.html.98According to the Center for Global Development, only 18 percent of U.S. reconstruction funds have come through USAID, while morethan 60 percent have come through DOD. As DOD efforts are more focused on the insecure areas, this would greatly increase theproportion of spending that takes place in those areas. See Gregory Johnson, Vijaya Ramachandram, and Julie Walz, “The CommandersEmergency Response Program in Afghanistan: Refining U.S. Military Capabilities in Stability and In-Conflict Development Activities,”Working Paper 265 (Center for Global Development, September 2011), p. 7.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 43

Photo by Paul FishsteinCash for work project, Balkh5.3 CorruptionAnother persistent complaint, voiced by virtuallyall respondents, was of widespread and pervasivecorruption in the implementation of aid projects.Responses on corruption were intimately linkedwith the narratives on insufficiency or on howothers got more, as the misallocation of aidresources was seen to be largely the result of thecapture of aid by powerful or elite groups,especially high government officials. Notsurprisingly, corruption was also linked withexpressed perceptions related to ethnic and tribalbias, as discussed above. The corruption in theaid system was described as part of a broaderpolluted political system, the same system thatkept in power unresponsive and corruptgovernment officials and predatory police. It wasalso described as one reason why “nothing hasbeen done.”The narrative of elite capture of aid resourceswas strong in all five provinces. In Helmand, aidwas seen as going to the grouping of tribes(Barakzai, Popalzai, Achakzai, and Alikozai)who had come out on top in the post-2001carve-up of political power in the province.References were also made to the introductionof the market economy, which was seen asempowering those same individuals and groupsholding the reins of power. What was presentedby the international community as promotingfree market dynamism was often perceived byAfghans as another unjust way in which thepowerful and corrupt were strengthened at theexpense of the public.As noted above, respondents expressed difficultyin reconciling the rhetoric they heard in themedia about levels of investment andinternational assistance with their ownexperiences. They saw people around them44Feinstein International Center

getting rich on perceived ill-gotten gains, whichencouraged cynicism about aid projects. A verytypical response was,For the last seven or eight years we’ve been hearingannouncements on national television and radio fromdonors and government representatives promisingmore aid and quoting large sums of money to bespent in Afghanistan, but how much of this hasactually reached the people? Most of it is lost throughcorruption in the government. What is left is toolittle to really improve our lives. 99The construction sector was generally describedas the most corrupt, and evidence exists that insome places it has become highly criminalized.In Uruzgan, construction and related securityservices were described as going overwhelminglyto the Popalzai tribe in general, and the late JanMohammad Khan and his nephew MatiullahKhan in particular. In Paktia, Governor JumaKhan Hamdard, a Pashtun from Balkh Provincewith links to Hezb-i Islami (GulbuddinHekmatayar), was widely viewed as benefitingfrom aid contracts, either directly throughcontracts to his personal or family members’firms, or by getting a percentage from thecontracts awarded to other firms. In Balkh, therewere fewer open complaints about specificgroups, but the assumption was widespread thatany significant project or development in theprovince had to be “coordinated” with thepeople in Governor Atta’s office. In Faryab, aidwas seen as going to those who were allied withGeneral Dostum’s political party, Jumbish. InFaryab, emergency distributions and cash-forworkor food-for-work projects were criticizedfor being the most prone to corruption, due totheir nature (rural, quickly designed anddelivered, and relying on middlemen), butperhaps also because there were fewer large-scaleconstruction projects than in the southern andeastern provinces.The most often-cited specific complaint wasabout multiple-level subcontracting, consideredto be a corrupt practice that led to inflated costs,use of sub-standard materials, and low-qualitywork. Large, well-connected companies weresaid to obtain large contracts, which were thenprogressively sold down the line to smaller, lesswell-connected firms, with a percentageskimmed off at each level. Organizations lowerdown the “food chain” had no choice but to dolow-quality work or else to pass a project on tosomeone else, who would likely do even lowerqualitywork. The reduced financial resourcesleft at the bottom of the chain forced the finalcontractors, even those with good intentions, toeconomize on materials and process; e.g., usingless cement and more sand, rushing a processeven when weather may not be appropriate, orusing sub-standard components or material.Sub-contracting is a common and wellunderstoodphenomenon among internationalorganizations, but was described as a legalizedform of corruption by virtually all Afghanrespondents, which is consistent with thenational-level narrative on the form thatcorruption takes in Afghanistan. In fact, theevidence is substantial that contract “flipping”has gone beyond what is considered acceptable. 100While the overall international aid effort wasdescribed as corrupt, and a few respondents saidthat the international community and donorswere the most corrupt, most specific accusationsof corruption were leveled at Afghans. A numberof Afghan respondents noted that the workdirectly done by PRTs tended to be of higherquality and with less corruption because PRTshad the personnel and logistical and securitycapacity to monitor the implementation of theprojects they funded more carefully than manyother donor-funded projects. Allegations ofcorruption were frequently aimed at the Afghanengineers at the PRT, or at Afghan PRTinterpreters, who were accused of having theirown construction companies to which theycould steer PRT business as well as block accessof their rivals in a process that was regarded ashighly opaque. A number of former interpretersfor the international military had formed theirown “NGOs,” which in reality were contractingfirms, as they had good access to those at thebases who were managing contracts. In Helmandand Paktia, respondents made a much moreextensive set of accusations revolving around the99Interview with community member in Khwoja Sabz Posh, Faryab Province, June 9, 2009.100Majority Staff, Foreign Relations Committee, United States Senate.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 45

ole of PRT interpreters, the existence of fakecompanies, collusion among constructioncompanies to raise prices, and payments forsecurity to militias linked with criminalgroups. 101 In Helmand, respondents were alsomore convinced of the complicity of foreignerswithin the PRT; as one Afghan line ministryrespondent noted, “either they are stupid or theyare part of the deal.” 102 Likewise, in Paktia, aformer high official said, “some PRTcommanders were wonderful, but others werethieves. They made deals with contractors andgot kickbacks.” 103 Regardless of who was seen asmost compromised, the system itself was seen ashighly corrupt. One possible factor contributingto respondents’ emphasis on Afghan corruptionis that in insecure provinces, fewer foreignerswere directly involved in implementing aidprograms. Furthermore, the political andeconomic impact of corruption on local powerdynamics was much more visible and was feltmore directly when the perceived beneficiaries ofcorruption were Afghans. 104In sum, the widespread expressed perception ofrespondents was that everything valuable(employment, contracts, legal judgments) wasdone on the basis of bribes or wasita rather than onmerit. Other than some respondents in the aidagencies, very few distinguished between NGOsand contractors in this regard. As one Afghan UNofficial complained, “Employment is all done onthe basis of who knows whom, wasita. Mastersdegree holders are selling [mobile phone] top-upcards by the side of the road, while illiterates haveresponsibilities in high positions.” 105 As an Afghanprovincial head of a national NGO put it, “underthe current conditions, people believe that thepurpose of aid is personal enrichment.” 106 Acommon complaint was about “briefcase” NGOs,formed solely to take advantage of aid money;they were described as lacking motivation,substance, and capacity, and as often disappearingat the end of a project (or before). As one Afghancontractor noted, “people in power don’t knowhow long they’ll be in power so they try to take asmuch as possible while they can.” 1075.4 “Wrong kind” of projectsMost respondents expressed dissatisfaction withthe kinds of projects that had been implemented,saying that they were not addressingAfghanistan’s fundamental problems such asunemployment. Respondents expressed thedesire for large-scale, visible, infrastructureprojects, largely for two sets of reasons. First,such substantial projects were seen as symbolic ofdevelopment and the promise of a future, in partbecause historically this was what developmentwas about in Afghanistan. 108 For many Afghans,the glaring lack of such projects signified that theinternational community was not investing tocreate sustainable economic growth for thefuture. Some respondents complained that toomuch money was going to small-scale projectsthat the Afghans could do themselves, but thatprojects in which the international communityhad a comparative advantage (large-scaleinfrastructure) were being neglected.Second, because, as described in Section 4.1,unemployment was seen to cause insecurity, thelack of large and visible employment-creatingprojects (e.g., factories, major irrigation schemes)was seen as a major aid failure. In all fiveprovinces, but especially in Balkh Province,respondents consistently expressed the need for101For a description of how this operates in other areas of Afghanistan, see Aram Roston,“How the U.S. funds the Taliban,” The Nation,November 11, 2009, www.thenation.com/doc/20091130/roston.102Gordon, Winning Hearts and Minds?, p. 45.103Interview with former governor, Paktia Province, January 5, 2009.104Since the time of the field research, a number of U.S. citizens have been prosecuted in U.S. courts for paying bribes to PRT staff to wincontracts. See “Former U.S. Army Staff Sergeant Sentenced to 90 Months in Prison for Bribery in Afghanistan Fuel Theft Scheme,” U.S.Department of Justice press release, www.justice.gov/opa/pr/2011/January/11-crm-021.html.105Interview with Afghan official in UN agency, Mazar-e Sharif, Balkh Province, March 22, 2009.106Interview with Afghan provincial head of national NGO, Mazar-e Sharif, Balkh Province, April 8, 2009107Interview with head of Afghan contractor, Tarin Kot, Uruzgan Province, July 7, 2009.108According to a review of USAID assistance during the cold war period, “for most Afghan officials ‘progress was construction,’ bestmeasured by the level of foreign aid commitments for construction of physical facilities.“ DEVRES, “Retrospective Review of U.S.Assistance to Afghanistan: 1950–1979,” (submitted to USAID, September 30, 1988).46Feinstein International Center

“economic projects” that created jobs and builtphysical infrastructure (e.g., factories, hospitals,airports, large dams). In Balkh and Faryab, manyrespondents complained that local naturalresources (e.g., natural gas, sulfur) had thepotential to create jobs and wealth, but were notbeing developed. This was described as part ofthe “peace penalty” described above. As a triballeader in Paktia said, “much of the work that isdone is not appropriate. We need energy butagencies instead build a small road or canal. . . .We have gas and coal that aren’t used, we couldbuild factories and employ people. Then therewouldn’t be a security problem.” 109 Respondentsin Balkh drew comparisons between the raisincleaningfactories of the old days 110 and thewedding halls of the current era; the inferencewas that previously the society (and thegovernment) did constructive things, while atpresent it was all about personal consumption.One specific genre of complaint, especially inthe north, was the perceived lack of attentionpaid to rehabilitating Afghanistan’s pre-1978industrial infrastructure. Many Afghans believedthat, post-2001, the government would resumewhere it had left off in 1978 as the major actor inthe economy. This would include rehabilitatingthe previous symbols of modernization (e.g., theKod-e Barq industrial complex, Helmand-Arghandab irrigation system, Spinzar cottoncompany) through which mass employmentwould be created. The lack of direct governmentinvolvement has been largely due to theintroduction of the market economy and the lackof resources at the disposal of the government totake on such initiatives. However, some Afghanrespondents went further and said that either theU.S. was intentionally neglecting Soviet-builtprojects so that Russia is not seen in a positivelight, or else that the West is intentionallykeeping Afghanistan poor and backward. In fact,respondents in all five provinces lamented thenature of current projects in contrast to the typesof projects that the Soviets and Americans usedto do:People don’t understand why current projects are sosmall. The Kajaki dam for example back then, aswell as now, costs a lot of money, but these days theAmericans are doing all these small projects. Havethey become poor? There is a need for larger-scaleinfrastructure projects, such as dams that provideenergy through hydro power. This electricity can thenpower machinery etc. and improve production andincrease employment, but instead solar panels aredistributed that provide just enough power fortelevisions, where people can see dancing andsinging, but this isn’t going to fill their stomachs. 111Views of cash-for-work and food-for-workprojects varied. In Balkh, cash-for-work projectswere generally described in a positive way, asproviding income and creating some sort ofdignity. International aid workers noted theadditional virtues of being flexible, relativelyquickly achievable, and do-able at scale, even ifnot sustainable. In Faryab, however, such projectswere seen as the most corrupt type of project,perhaps because the government hadimplemented most cash-for-work projects andhad used them for political ends. Also, given theshort timeframes in which these projects oftenhad to be completed, donors relied heavily onmiddlemen to identify beneficiaries anddistribute payments, which easily led toaccusations of corruption. In Helmand andUruzgan, cash for work was considered a shorttermstopgap measure that was also very divisive,as some tribes were viewed as having benefitedmore than others.5.5 Poorly implemented (low quality)In all five provinces, aid was universallydescribed as fragmented, lacking coherence, andgenerally poorly implemented. This was ascribedin part to elite capture and corruption describedabove, but also to the lack of local knowledgeand to aid actors pursuing their own individualor institutional interests rather than what wasprogrammatically sound.109Interview with head of tribal shura, Paktia Province, January 5, 2009.110During the 1960s and 1970s, exports of raisins from Afghanistan made up 60 percent of the global market. For many Afghans, thedominance achieved by Afghan raisins symbolizes the economic potential of the country’s agriculture.111Interview with Afghan official in Western aid agency, Tarin Kot, Uruzgan Province, January 2, 2010.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 47

Poor project design and outcomes wereattributed in part to donors’ and implementers’lack of local knowledge, a broad category thatincluded information on geography, the web oftribal and social networks, historicalrelationships, and any promises that had beenmade by members of a previous PRT rotation.This was especially so with military-leddevelopment. Unfamiliar with the environment,without knowledge of local languages, andlimited in their ability to get “outside the wire,”the military were especially vulnerable anddependent not only on their choice ofinterlocutors at the community level, but, moremundanely, on their interpreters. This lack of“penetration” of local communities and theinability to adequately understand socialstructures and politics allowed powerfulindividuals at provincial and district levels tomislead and manipulate ISAF in their disputeswith other power brokers. The less secure thearea, the less external actors were likely to know.In Helmand, the lack of knowledge within thePRT of who the significant players were or whatthe development needs were led to ad hocprogramming of projects, with the military oftenchoosing “targets of convenience,” opportunisticallydeveloping projects based on CIMIC patrols,often mainly in the more permissive areas.Consequently, projects were often not informedby either development needs or strategicconsiderations. In addition, security restrictionsplaced on civilian staff led to a greater relianceon the military to identify and manage projects,but due to security conditions even the PRTmilitary staff had chronic difficulties inidentifying and effectively implementing aidprojects. While some of these shortcomings haveimproved over time due to somewhat extendedrotations and the increasing attention paid tocultural and other training in advance ofdeployment, making programmatic decisionswith limited information on the environment isstill a challenge for military aid personnel.A recurrent complaint from Afghans,international development workers, and evenmilitary officials themselves in all five provinceswas the military’s lack of coordination with localauthorities. In many cases, the government wasnot consulted in the selection or location ofprojects. This resulted largely from oneroussecurity restrictions and resulting lack ofmobility, but also from lack of knowledge ofAfghan institutions and, in some cases, warinessabout working with what is considered to be acumbersome and corrupt administration. Also,line ministry representatives at provincial anddistrict levels are not always empowered to makedecisions, so engaging the ministry in Kabulwould sometimes become necessary, a processthat was either virtually impossible or woulddelay the decision well beyond the timeframe ofthe officials at the PRT. Therefore, sometimeseven when PRT personnel knew the system,they got frustrated and took short cuts. Thiscould achieve immediate short-term aims, butworked against longer-term sustainableoutcomes. As a CIMIC officer noted, “We oftenundermine the systems we say we’re trying to setup.” 112 CERP-funded projects were especiallynoted for their lack of transparency and for notbeing coordinated either with Afghan authoritiesor with other aid agencies or even with othermilitary actors. This led to building schools andclinics without adequate groundwork (i.e.,finding out whether another school was nearby)or ensuring the staff and material resources tomaintain the facility after construction. In Balkhand Faryab, personnel at the PRT (includingCIMIC) were unclear about what was beingdone under CERP and said that whencommunities cannot distinguish betweendifferent men in uniform, the PRT is blamed forany shortcomings in CERP implementation.Lack of timeliness of aid activities, mostlyattributed to cumbersome and bureaucraticprocedures, was also a recurrent criticism.Examples included animal feed that wasdesperately needed during a drought but whicharrived six months late, loans for agriculturalactivities that were offered during a droughtyear, and fruit tree saplings that came after theplanting season. A number of respondents citedthe Afghan proverb about “henna after thewedding” to describe inputs or activities thatarrived too late to be useful.112Interview with CIMIC officer, Mazar-e Sharif, June 3, 2009.48Feinstein International Center

Photo by Paul FishsteinDonor plaque on school, BalkhTime pressure also had a negative effect on thequality of aid projects in a number of differentways. First, pressure originating in Washingtonand other capitals for showing quick resultsproduced unrealistic timeframes forimplementation in the field. 113 Agenciesdependent on grants and contracts often agreedto more “aggressive” implementation schedulesto gain and maintain funding. PRTs were undersimilar political pressure to deliver projects andspend money, both of which were key metrics.In all five provinces, field-level staff expressedconcern about the excessive levels of funding,although in Balkh and Faryab this was more thecase towards the end of the fieldwork period,when U.S. funds were anticipated as part of the“civilian surge.” In Uruzgan concerns were thatnew Special Forces personnel were pushing for acertain level of CERP projects to beimplemented each week and month. As an aidcontractor noted, “CERP gives somethingtangible to offer for those cooperating withforeign forces. But it’s not always optimallyimplemented, partly because there is too muchpressure to spend the money . . .we’re moreassessed on numbers than impact.” 114The time pressures caused by short-termpersonnel rotations (some as short as six or evenfour months) encouraged a short-termorientation. Each PRT rotation wanted to reportgreater success in the mission than the previousrotation, and to leave the situation better thanbefore, or at least not worse. This was partly dueto individual career considerations, but also dueto the heartfelt desire to make a difference.Therefore, personnel charged with stabilizingareas were tempted to take short-cuts that would“quiet down” the situation, despite being warnedagainst “quick fixes.” As described above, aidagency staff expressed concerns that paying offmaleks could in fact be de-stabilizing, as it wouldcreate perverse incentives (e.g., what happenedwhen you stopped paying, and what about othermaleks who wanted a share). However,institutional incentive structures as well as113For an account of the negative consequences of the pressure to spend money, see Rajiv Chandrasekaran, “U.S. Military Dismayed byDelays in 3 Key Development Projects in Afghanistan,” Washington Post (April 28, 2011), http://www.washingtonpost.com/world/us-military-dismayed-by-delays-in-3-key-development-projects-in-afghanistan/2011/04/22/AFD6jq8E_story.html.114Interview with international aid contractor, Tarin Kot, Uruzgan Province, January 2, 2010.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 49

human nature often meant concern aboutlong-term consequences was less if one wasgoing to leave in only a few months’ time. 115The rapid rotation of PRT teams contributed toprojects’ lack of continuity, a source of frustrationamong both Afghan officials and PRT personnelthemselves. In Balkh, CIMIC officers and aformer woleswal (who had served in anotherprovince) described their frustration at projectsbeing lost at handover. Some international aidofficials expressed frustration at the lack ofmonitoring and the general lack of institutionalknowledge concerning projects that hadpreviously been implemented. In some PRTs (e.g.,Balkh, Paktia) personnel were unable to locaterecords of projects implemented by theirpredecessors, some as recently as six monthsbefore. Many Afghan respondents blamed the lowquality of some construction projects on the lackof monitoring as, in the absence of oversight,contractors were under no pressure to do qualitywork. New teams either had no knowledge of, orfelt no need to follow up on, the commitments oftheir predecessors (which in some cases were notdocumented). Community perceptions of brokenpromises may have arisen due to a number ofreasons related to both volition and competence:poor translation, differing cultural interpretationson what constitutes a promise, translators andpatrols taking the path of least resistance duringvisits by holding out the promise of futureassistance, and willful deception in order toachieve short-term objectives.5.6 Non-sustainableNot surprisingly, aid agency staff as well as somePRT personnel, raised concerns aboutsustainability and creating dependency. Manyproblems described above were directlyattributed to institutional pressures to spend largesums of money quickly and to do large numbersof projects. While virtually all respondents saidthat projects of a long duration were better thanshort-term ones, many Afghans did notnecessarily envision developmentally soundprojects; rather, in many cases they were simplylooking for projects with a long duration (e.g.,cash-for-work projects that ran for a long time).BOX 4: Komak: Aid versus charityIn the Dari-speaking areas, most respondents used the word komak (help) to refer to bothdevelopment assistance (projects) and charity. Outside the circle of development practitionersand some government officials, not many Afghan respondents distinguished betweendevelopment and charity.Even when aid projects were designed to be sustainable and market-oriented,developmentally sound approaches were often viewed with resentment by local communities.For instance, respondents in one district in Balkh complained that the daily wage paid by acash-for-work project was an inadequate $4 (AFS200). Although the going local wage ratewas half that, respondents apparently saw no inconsistency in believing that aid projectsshould pay more than twice the local wage rate, a notion that is at odds with generallyaccepted principles of development practice. In the same area, complaints were made thatkunjala (a type of animal feed) had been sold to people at $5.40 (AFS270) per bag; while thiswas described as a fair price in the market, it was seen as exorbitant in the context of an aidproject. The fact that these transactions were related to aid projects resulted in their beingviewed through a different lens.Box 4 continues on next page115For an example of the dangerous conflict between short-term and long-term goals, see the discussion on the distribution of water pumpsin Helmand in Dempsey, “Is Spending the Strategy?” p. 3. See also reports that the ten French troops killed in Sarobi in August 2008 hadbeen ambushed due to the discontinuation of bribes previously paid to insurgents in the area: “French Opposition Demands Answers onBribe Claim in Sarobi Ambush,” The Times (London, October 16, 2009), http://www.timesonline.co.uk/tol/news/world/europe/article6876691.ece.50Feinstein International Center

Box 4 continued from previous pageIn fact, NGO staff reported that some communities were reluctant to engage with NGOsthat insisted on a community contribution (e.g., a 30 percent community contribution in theform of labor or paying for materials such as seed and fertilizer rather than receiving themfree), instead holding out for one that would provide free assistance. In some cases, theinsistence on community contribution created the suspicion that their NGO was corrupt,and was siphoning off money being paid by the community.A surprisingly large number of respondents in Balkh province criticized micro-finance projectsas un-Islamic because they allegedly collected excessive interest (sudh, or usury), did not helppeople out of poverty, and benefited only the implementers, who managed them as commercialenterprises rather than as a service for society. Some said that this sort of commercial approachon the part of micro-finance organizations simply reflected the government’s new free-marketstrategy. A typical comment was that “these organizations are not here for development help,but for business. This reflects the new government strategy of the free market.” 116 Some of theresentment possibly comes from having to pay for something, especially when it has beendelivered by an NGO. This area requires additional analysis.The separate standard for aid may be encouraged by projects that, for whatever reason, pay inexcess of the going rates for labor and materials. While often intended to curry favor withcommunities or individuals, they may lead to high expectations and dependency—andperhaps to alienation when the tap is turned off.As noted above, a major complaint was that aidprojects were not addressing Afghanistan’sfundamental needs. Unfavorable comparisonswere made with Soviet aid, which built factoriesand dams. Respondents in Helmand noted thatthey wanted projects like those that theAmericans had built in the 1960s and 1970s,while in Balkh the call was for projects like thosethat the Soviets made. Also as noted above, thiseasily led to accusations that the West wished tokeep Afghanistan poor or that it did not want toleave anything behind.On a related note, a number of respondentsmentioned good projects that had fallen apartwhen the implementing NGO lost funding,ceased work in the area, or was otherwise unableto follow through on planned activities. InBalkh, a district head of agriculture andcommunity members complained that the goodwork done by one NGO in introducing newvarieties of crops and developing test plots hadbeen wasted because the project fell apart whenits short-term funding ended and noarrangement was made for handover orcontinuity. Also in Balkh, in two casesrespondents mentioned that they had personallybeen left holding the bag when NGO activitieshad suddenly been terminated. In the first case,the head of an orchard owners’ association washeld responsible by his members when they wereunable to sell their pomegranates at theadvantageous price that had been negotiatedthrough the good offices of an internationalNGO. In the second case, a district official waspressured to pay several days of wages to 500workers who had been promised the opportunityto work planting pistachio trees that neverarrived. 1175.7 Positive views: Some good newsCommunities did provide some positive viewson aid projects, mainly with respect to the116Interview with Afghan NGO staff worker, Khulm District, Balkh Province, June 7, 2009.117Paul Fishstein, Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan’s Balkh Province (FeinsteinInternational Center, Tufts University, November 2010), p. 35.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 51

Photo by Paul FishsteinSign for National Solidarity Program, BalkhNational Solidarity Program (NSP), 118 somesignificant and highly visible projects, andrelationships with long-serving NGOs.The most consistently positive views in allprovinces except Uruzgan were on the NSP. Thelanguage respondents used to describe NSP wasremarkably consistent; i.e., “responsive to thecommunity,” “transparent,” “communitiesidentify their own needs,” “solved problems,”“NSP is in our own hands,” and “people came tobelieve that their vote had meaning.” People alsolargely reported being pleased with the choice ofprojects, most commonly small-scale powergeneration (“allowed us to listen to the worldnews”), bridges, and communal guest houses. Atypical comment was, “NSP has had a goodimpact because the priorities of the people aretaken into account and we know what happensto our money.” 119 Another factor, whichironically may have contributed to greatersatisfaction, was the small amounts of moneyinvolved. On average, $27,000 was disbursed percommunity. This may have been too little tointerest the large construction companies orpowerful interests, which might have broughtthe competition and corruption described inother types of projects. Also, there was a standardand publicly understood formula for allocatingfunding ($200 per household up to a maximumof $60,000 per village), which may havemitigated the perceptions of injustice that wereassociated with other aid programs. Wherecomplaints were heard, they were usually thatthe CDCs were influenced by political parties orcomposed of the same people (i.e., commanders)who had preyed on the community previously.In Uruzgan, the NSP was cancelled due to thepoor performance of the implementing partner.Still, far fewer complaints were heard about NSPcompared to other projects and programs, andCDCs appear to be less prone to corruption thanindividuals who would otherwise be the conduitfor aid, such as commanders, maleks, and arbabs. 120118The NSP is a national program that uses a community development approach to build minor infrastructure (e.g., roads, small irrigationstructures, hydro and solar power, community buildings) and in the process promotes village-level governance. Communities electcouncils that identify community needs, develop proposals, and oversee small grants under which work is done. The NSP is acollaboration between the international community, which provides funding and technical guidance, the Ministry of Rural Rehabilitationand Development, which provides general oversight, and Afghan and international NGOs, which as the “facilitating partners” interactwith communities.119Interview with CDC members, Qaisar, Faryab Province, April 11, 2009.120For similar observations, see Beath, et al., “Winning Hearts and Minds?”52Feinstein International Center

Photo by Geert GompelmanNSP irrigation project in FaryabRespondents also offered positive comments on afew significant or highly visible projects. InFaryab, these included extending the pavedportion of Afghanistan’s ring road and bringingelectricity from Turkmenistan to Maimana andAlmar, both of which were described as havingbeen transformative.Although most respondents complained about“NGOs” in general, many were appreciative ofthe work that “their” NGO was doing. In mostof these cases, the positively cited NGOs hadworked in the area for a number of years.International aid workers noted that peopletended to like the individual NGO they workedwith, but complain about NGOs. Generalcomplaints about NGOs may reflect, in part, thedrumbeat of criticism by President Karzai andother government officials, which may simplyreflect, in part, competition over resources. Asnoted above, few Afghan respondents madedistinctions between NGOs and contractors.Interestingly, while respondents wereconsistently critical of the government, donors,and NGOs, somehow NSP, a collaboration ofthose three entities, avoided the same criticism.In Paktia, the Germans were described relativelypositively, although this was likely the result of along pre-war history of German involvement inthis area in small-scale vocational training.Although the international community and theAfghan government have recognized the healthsector as a success story, respondents did notdescribe it especially positively. Many peoplecomplained about the lack of 24/7 physicianservices, ambulances, or willingness on the partof clinics to provide enough pharmaceuticals.This may be a matter of rising expectations, ascomplaints were voiced in areas that most likelydid not have even a health post until recentyears. It may also indicate a limitedunderstanding of the role of preventive care;respondents placed more importance on curativeservices than preventive ones, a near-universalphenomenon. Likewise, complaints about a lackof willingness to distribute pharmaceuticals arehard to interpret. In a society in whichphysicians massively over-prescribe medicine andoften have financial interests in pharmacies,restraint in prescribing medicine may in fact bean indicator of a higher quality of care ratherthan a lower one. 121121Research has shown that while a general correlation exists between a population’s health status and a country’s economic strength, whichis in turn correlated with a lower level of conflict, there is scant evidence that health systems development contributes directly, either inthe short- or medium-term, to reducing conflict. Rather, physical security, a functioning justice system, and employment are likely to bemore important influences. Moreover, designing health interventions with security objectives in mind are likely to result in projects (i.e.,high visibility, quick impact) which may be inconsistent with long-term health system development. See Leonard S. Rubenstein,“Post-Conflict Health Reconstruction: New Foundations for U.S. Policy,” USIP Working Paper (September 2009).Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 53

6. The stabilizing and destabilizing effects of aidU.S. military overseeing drilling of water wells, BalkhWhile the environments in the five provincesdiffered greatly, a number of consistentobservations emerged concerning theeffectiveness of aid projects in promotingstabilization objectives, both in the short termand long term. First, some military officialsreported that in some areas militaryadministeredaid projects may have had shorttermsecurity benefits, at least in the limitedsense of force protection. Some CIMIC officerscautiously reported that projects, including thosethat hired from communities near militaryinstallations, were successful in providing somelimited force-protection benefits. As one CIMICofficer put it, “in the short-term, this avoidspeople throwing rocks at patrols, so that NGOscan operate, and in turn help with long-termsecurity.” 122 In Balkh and Uruzgan, militaryrespondents reported that such projects save lives,and that community members were morewilling to report IEDs or provide usefulinformation after the implementation of smallprojects. In Balkh and Uruzgan, developmentadvisors and military officials also reportedimproved access to villages after theimplementation of small projects (e.g., “socialprojects” coordinated with police training). InFaryab, however, there was no indication that aidprojects contributed to short-term stabilization.122Interview with civil-military advisor, Mazar-e Sharif, Balkh Province, April 11, 2009.54Feinstein International Center

DOD photo by Fred W. Baker IIIPRT personnel at community meeting, PaktiaSecond, some evidence was found that aidprojects did help to build relationships. Some feltthat development projects gave the militaryaccess to the people and encouraged a positive,or at least pragmatic, view of the military. Onepolitical advisor in Balkh said that CIMIC wasvaluable for “opening doors” and legitimizingtheir presence. As a former woleswal and CDCchairman in Balkh put it, “Projects can alsobuild relationships between government, NGOs,international community.” Similarly, in PaktiaPRT staff felt that projects provided a “platform”or context for meeting with people whom theywould not otherwise be able to engage, andlegitimized their interaction. For instance, theZadran Arc Stabilization Initiative in Paktiaappears to have developed relations betweenestranged Zadran tribal elders, UNAMA, andthe government. It also got local agreement toprotect the workers of local constructioncompanies and NGO partners working on aschool and road-building project. However, noguarantees were given that NATO/ISAFconvoys would not be attacked, and whether theproject had any substantial positive influence onsecurity beyond its initial role in facilitating adialogue led by UNAMA and some Zadrantribal elders is unclear. 123Being able to deliver a project may be useful in asociety where one needs to offer or exchangesomething in order to develop relationships. Asput by an international aid official in Uruzgan,“if we need help from local leaders we need arelationship—projects help to build relationships.They provide an incentive for local leaders totake a risk and work with us. Aid projects arevery useful in building a relationship.” 124According to another official, “if they say ‘weneed a water pump,’ and we deliver, then theygain trust to do business with us.” 125 On theother hand, there is a danger that “deliveringsomething” can have immediate positive effectsbut contribute in the longer term to corrupt,transactional relationships. According to onegovernment official, taking bribes “is nownormal and accepted in society, and is gettingworse day-by-day. It started four to five years123Formulated in 2005 with the assistance of UNAMA, the Zadran Arc Stabilization Initiative in Paktia was intended to stabilize a veryinsecure area through providing employment, development activities, infrastructure, government outreach, activation of arbakai,additional security resources, oversight of madrassas, and empowering the tribes through a representative tribal shura.124Interview with international aid official, Tarin Kot, Uruzgan Province, January 31, 2010.125Interview with PRT official, Tarin Kot, Uruzgan Province, February 1, 2010.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 55

DOD photo by Fred W. Baker IIIU.S. military personnel and Afghan civilian, Paktiaago with small gifts that then created obligationsand the need for reciprocity.” 126 This dynamicmay be reinforced by international officials’short-term horizons and by the uncriticalacceptance of an assumption about Afghancorruption (culture of bakhsheesh) which suggeststhat foreign officials must pay to accomplishanything with government officials. The risk isthat short-term transactional relationships thatgave access to local communities in exchange forprojects— “you scratch my back and I’ll scratchyours”—were confused with longer-term“winning hearts and minds” work.Finally, beyond limited short-term forceprotectionobjectives, facilitation of initialinteractions between international actors andlocal leaders, and some reported cases ofintelligence gathering benefits, there was littleconcrete evidence in any of the five provincesthat aid projects were reducing unrest in thelonger term. As noted above, this may be due tothe difficulty of measuring effects and attributingcausality in an environment with so manyconfounding variables. Still, the dominantexpressed view was that, at most, a very limitedand short-term relationship existed between aidand security. Moreover, in southern and easternAfghanistan, where the volume of aid resourcesto promote stabilization objectives was muchgreater, more evidence was found that aidprojects and the entire aid architecture was doingmore to undermine rather than promotesecurity. 127The following sections explore several relatedissues, starting with whether aid is focusing onthe wrong drivers of insecurity and how aidefforts focused on stabilization are subverted byinsurgents or other malign actors. The sectionsalso discuss why many aid projects are destabilizing,due to corruption, the creation orexacerbation of competition over resources, thereinforcement of inequalities and the creation ofwinners and losers, and the aggravation ofresentments due to regional disparities.126Interview with senior provincial representative of a line ministry, Tarin Kot, Uruzgan Province, July 5, 2009.127A number of the findings were similar to those from the Horn of Africa. There, in a non-kinetic environment, military-sponsored aidactivities were successful in facilitating the U.S. military’s entry into regions of potential concern and in allowing the military to acquirelocal connections, networks, and knowledge. There was less evidence, however, that aid succeeded in changing overall attitudes towardsthe U.S. or, more important, in achieving the strategic objectives of countering terrorism, reducing conflict, or improving stability. As inthe Afghanistan study, the Horn of Africa study found that attitudes and behavior are influenced by a complex of factors—many of whichare not affected by aid projects. See https://wikis.uit.tufts.edu/confluence/pages/viewpage.action?pageId=34807224.56Feinstein International Center

6.1 Addressing the wrong drivers ofinsecurityStabilization theory as practiced in Afghanistanplaces high importance on socio-economicdrivers of conflict such as poverty, illiteracy, andlack of social services, and therefore emphasizessocio-economic solutions, including providingemployment opportunities and building schools.The best-selling author, Greg Mortenson, helpedpopularize this theory by suggesting that hisschools in Pakistan and Afghanistan werepromoting peace “one school at a time.” 128 Thecurrent study and other research, however, hasshown that the causes of insecurity are diverseand intertwined. Moreover, given the multitudeof factors at play in a given environment,establishing causality or isolating the effects ofone variable from the others is extremelydifficult. What, for instance, is the influence ofpoverty, as opposed to political grievances,exclusion, or tribal solidarity? Respondentsmentioned cases where “Taliban”-relatedinsecurity in a certain area had abated when oneof two feuding brothers returned to Pakistan orwhen a long-running land dispute was settled.Other than employment, most developmentassistance was not directly addressing the majordrivers of conflict reported by respondents.While poverty was given as an important factorin creating unrest, many related grievances relateto political or identity issues that are notaddressed by development.As one military official observed, “a diversegroup of people are creating problems for avariety of reasons (power, money, ideology,religion), and are not affected by development....We can hold the area longer if we spend moremoney, but eventually insecurity will take hold.The things working against us are not affectedby development.” 129 The conclusion drawn inHelmand was that the stabilization model usedbetween 2006 and 2008 “focused on the wrongdrivers of conflict—on the lack of developmentand government presence rather than on poorgovernance and insecurity.” 130The centrality of poverty in the assumptionsunderlying aid projects aimed at stabilizationas well as in the responses from Afghans on thedrivers of insecurity suggests the need tocritically examine to what extent and in whatway poverty and unemployment affectsecurity.Poverty and unemployment are the major focusof security-related development, which posits adirect link between poverty and unemployment,and young men joining the insurgency. This ismost starkly stated as the “ten dollar a dayTalib,” a young man who is unemployed andtherefore finds something to do to earn moneyand pass the time. While the military anddevelopment actors often hold divergent views,this view of the linkages between poverty andinsecurity is shared by both. 131 To some extent,this model relies on Western concepts ofindividual decision-making. Individualeconomic actors are often implicitly assumed,while ignoring the important roles ofhousehold, community, and tribe, in makingwhat in the West would be seen as individualdecisions. As part of a traditional society withstrong family, tribal, and social bonds, mostyoung Afghan males are unlikely to becompletely free agents who can independentlymake a major decision to join the Taliban giventhe very direct impact this could have on one’sfamily, village, and tribe. Yet many Afghansinterviewed for the study also stated the strongbelief that unemployment was a major cause ofviolence and insecurity. In the north, but alsoelsewhere, respondents talked about thesubsistence needs of the desperate. As a lastresort to get money to satisfy physical needs(e.g., food, medicine), many were said to bewilling to do anything—including acts thatviolated their own sense of right and wrong.Drawing a distinction between attitudes andaction may be useful in understanding why someand not others join the insurgency. Poverty is awidespread condition in Afghanistan, while the128The titles of Mortenson’s two best-selling books are Three Cups of Tea: One Man’s Mission to Promote Peace . . . One School At A Time, andStones into Schools: Promoting Peace with Books, Not Bombs, in Afghanistan and Pakistan.129Interview with international military official, Mazar-e Sharif, Balkh Province, May 28, 2009.130Gordon, Winning Hearts and Minds?, p. 51.131See Andrew Wilder, “Aid and Stability in Pakistan: Lessons from the 2005 Earthquake Response,” Disasters, v. 34 (2010): S406–S426.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 57

insurgency is less so. While the majority ofAfghans are very poor, and many would agreewith the negative characterization of thegovernment and the international communityoffered by the Taliban, relatively few wouldactually join the insurgents.BOX 5: MahroumWhile poverty may not be the single explanation for why young men join the insurgency,those without a job or hope for the future may be more susceptible to mobilization againstthe government. The Dari word mahroum, which translates as “deprived” or “left out,” butwhich also conveys a more profound sense of both alienation and being discriminatedagainst, was used by respondents to reflect a broader definition of poverty. Unemployed butalso lacking human and social capital—no livelihood, no bride, no prospects, and perhapsdispossessed from land or other property at the hands of a corrupt government official— thecentral issue is lack of hope for a decent future. Moreover, mahroum can also imply injustice,or being deliberately deprived at the hands of others. Thus, if one is already angry at thegovernment and the international military, one may yield more easily to immediateeconomic necessity. Conversely, if one is already dispossessed, then one may also more easilybecome estranged in response to a specific event or grievance. In this sense, poverty is acontributing but not sufficient condition. Also, to the extent that the state and theinternational community are seen as not having the interests of the population in mind andtherefore as unlikely to provide a livelihood or social services, one may be more willing tooppose them violently. This may also suggest in part why respondents focused on the lack oflarge, economic projects; the lack of large projects meant not only that people aren’t busyworking, but also no longer have the positive sense of a trajectory of development andtherefore hope for the future.Another possible factor is the interplay betweenrelative poverty and perceptions of injustice. Ifpoverty is a generally shared experience, thecondition may not generate much hostility. If, onthe other hand, poverty is not shared and someelements of society are perceived to have grownwealthy at the expense of others, this may feed asense of injustice, informed by Islamic notions ofequity, that could motivate armed opposition to astate that tolerates and even encourages this un-Islamic behavior. In this scenario, anger is bred notso much by absolute poverty but by a strong senseof exclusion and injustice. (See Box 5.) This isconsistent with the “zero-sum” outlook thatseemed to pervade so many of the complaints aboutdevelopment, but it may also reflect a response to aperceived widening social and economic gapbetween poor and rich. At present, poverty is oftenexperienced as a state of being excluded fromresources, from economic and political power, andfrom justice. In this sense, the corrupt and unjustsystem produces poverty. As noted above (see Box3), a general sense of injustice pervaded many of thespecific complaints.Even if not a direct driver of insecurity, povertyincreases vulnerability. Respondents in the northnoted that poverty increased the need to migrateto the south for work or to enroll children inPakistani madrassas—both of which wereassociated with radicalization or mobilization bythe insurgents. 132 Poverty was also associatedwith political vulnerability, in that it made onemore vulnerable to being dispossessed ofproperty by predatory actors, which in turnincreased alienation.Similarly, given the vested interests of criminalsand others in fomenting conflict and insecurity,and the lack of a clear line between criminaland political violence, development projects areunlikely to reduce unrest related to criminality.132Ladbury, “Testing Hypotheses,” makes a distinction between mobilization and radicalization.58Feinstein International Center

In the words of one UN official, “developmentprojects will have no effect on criminal activity,as criminals seek to maintain the status quo[insecurity]. The source of this sort ofinstability is more due to poor governance andpolice; improving the police would help toimprove this part of the situation.” 133In addition, the weight given by respondents ininsecure areas to the international militaryforces as a cause of insecurity would suggest aninternal contradiction in COIN. In small ruralcommunities where almost everyone is relatedthrough tribal or other networks, militaryactions taken against suspected Taliban actors orsympathizers, well-targeted or not, will almostby definition alienate parts of that community.In Paktia, in reference to the work done by thePRT, one respondent noted “120 kilometers ofroad constructed in Khost, 183 schoolsconstructed in Khost, but whatever is done . . .then they kill two kids and everything isdestroyed.” 134 Another respondent in Paktiaadded “the PRT’s aid efforts are not effective.How can it be when you have a gun in onehand and a piece of bread in the other?” 135Similarly, if aid is delivered in association witha government that is distrusted for not beingneutral or attentive to the welfare of the people,it is less likely to achieve the COIN objective ofwinning the population over to the side of thegovernment. One respondent cited a Pashtuproverb, explaining, “you can milk a cow andget a bucket of delicious milk. But if the cowthen kicks over the bucket is it a good or badcow? PRTs are a good cow with lots of milk,but then the government appoints Juma KhanHamdard [the governor of Paktia] and thebucket is kicked over.” 136 With high levels ofantipathy towards the government, due largelyto the corrupt and predatory actions of localofficials, the narrative of governmentcorruption may easily overwhelm theattractions of engaging with the governmentand its actors. In most provinces channeling aidthrough the government had not evidentlyincreased its legitimacy; on the contrary, theperceived corruption associated with aid hadhurt or reinforced the bad reputation of theprovincial administration. In Uruzgan, forexample, officials explicitly questioned whetherextending the reach of the government wasstabilizing, as shura members did not wish toattend meetings in which government lineministries participated. According tointernational officials, “the reputation of thegovernor is so corrupt, and the reputation of theprovincial government so bad, that goodprojects don’t improve the reputation of thegovernment.” 137Some analysts contend that in Afghanistan,legitimacy rests on national identity andIslam. 138 To the extent that this is correct, theattempt to create legitimacy through servicedelivery and governance is both alien andproblematic, as they imply expectations fortangible action, whereas the traditional bases forlegitimacy are ideational. Moreover,development agencies can only provide money,materials, and technical assistance, and havelittle ability to address issues connected withnational identity or Islam.One Afghan academic and former senior officialnoted the challenges against which developmentprojects were working: “bad governance, theculturally insensitive work of human rightsorganizations, the work of fundamentalistgroups, tribal elders who are maintainingdifferent networks with different people, thosewho feel they have not benefitted from the newsystem and new government—all play a role inworking against the good work of PRTs.” 139133Interview with UN official, Mazar-e Sharif, Balkh Province, April 6, 2009.134Interview with former Taliban commander and deputy of an Islamic political party, Paktia Province, January 6, 2009.135Interview with local community leader, Paktia Province, January 10, 2008.136Interview with academic and former Ministry of Education high official, Paktia Province, January 6, 2009.137Interview with international development officials, Tarin Kot, Uruzgan Province, January 31, 2010.138See Astri Suhrke, “Exogenous State-Building: The Contradictions of the International Project in Afghanistan,” in The Rule of Law inAfghanistan: Missing in Inaction, ed. Whit Mason (Cambridge: Cambridge University Press, 2011), pp. 225–48.139Interview with academic and former Ministry of Education high official, Paktia Province, January 6, 2009.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 59

6.2 Subverted by insurgentsAnother related factor limiting the positiveeffects of development assistance was that, mostlythrough physical threats and social intimidation,insurgents actively subverted overt expressions ofcommitment or engagement by communities.Here, fundamental differences in perceptions andattitudes were found between the secure andinsecure areas. To some extent, in the north theinternational military was seen as on the side ofsecurity (or at least there was a pragmaticconfluence of interest), whereas in the south andeast it was seen by many as a dangerouslythreatening presence. Especially in Balkh, wherelevels of violence were low, the internationalmilitary was seen as targeting a minority whothreatened the well-being of the majority. InHelmand, Uruzgan, and Paktia, on the otherhand, the population’s experience of violenceand conflict—both directly at the hands of theinternational military, and from the Taliban as aconsequence of their presence—meant thatcommunities were often too intimidated toactively cooperate. This was not a question ofprinciple; rather, the lack of cooperation wasdriven by a very pragmatic sense of security. Nodoubt communities would have likeddevelopment money and projects, but feared theconsequences of accepting it (or being seen toaccept it). “A sense pervaded that the net benefitsof ISAF were strongly outweighed by theperception that their presence exacerbatedinter-communal conflict as well as with theTaliban.” 140 This was the case in Helmand, wherethe inability of ISAF and the government toprovide security discouraged the populationfrom cooperating with or even engaging withdevelopment activities. Communities weresimply too frightened to participate in activitiesthat could result in being attacked by insurgents.As an aid official noted, “development activitiescan contribute to stability, but if the communityhas no confidence in the government and insecurity, development cannot solve problems.People need to feel secure.” 141 In the insecureareas, communities would only engage if theycould “buy” security from the Taliban or elsewere powerful enough to assure their ownprotection. In Balkh, on the other hand, withouta context of conflict, most communities werequite willing to engage with the internationalmilitary on development projects, as the risk thatinsurgents would later show up to ask questionswas almost nil. This is in line with observationsmade by Zürcher and his colleagues, who foundthat attitudes towards foreign forces were drivenby communities’ perceptions of their ownsecurity rather than by aid projects, and that theability of aid to influence attitudes depended onsecurity and threat perceptions; in the relativelysecure areas of the north, people largely felt that“foreign forces are still mainly seen as positivelycontributing to security.” 142Some evidence exists in Helmand that thelong-term presence of government orinternational forces, which signaled a likelylonger-term security benefit, changed theseperceptions somewhat. However, whereoperations were mainly “drive-through,”communities had no incentive to engage withpeople who were not going to be there a weekfrom now to provide the protection that theywere promising. As was noted in Uruzgan,people and communities might stand up to theTaliban, but they might not have anything tolink to (e.g., government, security forces) oncethey stood up, especially in areas distant from themain population centers. As one Afghan NGOworker in Uruzgan noted, “in remote areas theTaliban are very strong, and where they arestrong the people cooperate with them.” 143 Manyrespondents noted that unrest and intimidationmay be created by small groups, although theymay not represent the views and wishes of thelarger community; two or three men with gunscan control a village.Moreover, some evidence exists that insurgents,perhaps for tactical reasons and to not alienatelocal communities, may tolerate developmentprojects as long as they fall within certainboundaries. Military officials noted that140Gordon, Winning Hearts and Minds?, p. 46.141Interview with international development official, Mazar-e Sharif, April 8, 2009.142Böhnke et al., “Assessing the Impact,” p. 12.143Interview with official at Afghan NGO, Tarin Kot, Uruzgan Province, July 6, 2009.60Feinstein International Center

communities could be willing to acceptassistance that would make people’s lives better,yet still work against the institutions providingthat assistance; as one put it, “they could acceptwhat we bring, and still be working againstus.” 144 Another military official noted that “theTaliban may allow you to build a well if theyknow that one is needed in the area, but thisdoesn’t mean that you’ve won hearts andminds.” 145Even when not actively subverted by Talibanpropaganda, the environment is fertile forconspiracy theories that undercut messages ofgood intentions that aid projects are intended toconvey. It has been widely observed that manyAfghans believe (or claim to believe) that theU.S. and NATO are actually supporting theTaliban to justify a continued presence inAfghanistan. This is inspired by disbelief that themost powerful military in the world has beenunable to eliminate what was a defeated, rag-taggroup of rebels. On the development side, theperceived lack of investment and exploitation ofAfghanistan’s natural resources created furtherdoubt on exactly what the internationalcommunity’s intentions were. According to thisline of reasoning, if the international communitywas serious about developing Afghanistan, thenit would have invested more in large projects thatmade Afghans self-sufficient (and employed).This naturally leads to speculation on thealternative motives for the intervention andopen-ended occupation in Afghanistan, whichinclude revenge for British defeats in thenineteenth century or against Afghanistan forgiving shelter to al Qaeda during the Talibanera, eyes on Afghanistan’s mineral resources, orgeo-strategic presence (e.g., proximity to Iran).These suspicions are reinforced by the belief,among some, that foreign intervention hasresulted in the moral pollution of Afghan societythrough conspicuous consumption (e.g., weddinghalls) and the introduction of liberal habits. Onerespondent noted that small-scale powergeneration was enough to watch serials andsinging and dancing on television, but it was notenough to create jobs. Even one of the mostsignificant gains of post-2001 Afghanistan, theexplosion of media and connectivity, was oftenregarded with suspicion, especially due to theprevalence (and widespread popularity) offoreign imported or inspired content such asfilms, television serials, and other programswhich were viewed by many as beinginconsistent with and undermining of Afghantraditional culture and values.6.3 Poor quality of implementationEven discounting people’s natural tendencies tocomplain, discussions with a range ofrespondents clearly indicated that aid deliveryitself was a source of dissatisfaction. This suggeststhat even if the “hearts and minds” model isaccepted as plausible, the poor quality ofimplementation would have further limited theability of aid projects to produce stability. Thefactors that affect the international community’sability to effectively and accountably deliver aid(lack of local knowledge, time pressure, etc.)were discussed in Section 5.5.6.4 Destabilizing influences of aidThe research findings from this study suggestthat while the stabilizing effects of aid were atbest modest, the destabilizing effects of largeamounts of development assistance being spentrapidly in conflict zones could be considerable.Particularly destabilizing was the role of the waraideconomy in fueling corruption, which inturn reduced the government’s legitimacy. Thetensions and conflicts caused by aid projects andresources that were perceived to reinforceinequalities and create winners and losers are alsodestabilizing.6.4.1 CorruptionThe main way in which aid was found to bedestabilizing was by fueling the corruption thathad such a corrosive effect on the legitimacy ofthe government. The assumption thatdevelopment aid had stabilizing effects meant144Interview with international development official, Mazar-e Sharif, Balkh Province, April 8, 2009.145Interview with military official, Mazar-e Sharif, Balkh Province, January 14, 2010.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 61

DOD photo by Fred W. Baker IIITrucks along “K-G” Road, Paktia-Khostthat as security deteriorated, more and more aidresources were poured into the most insecureareas. Spending too much money too quickly inenvironments where there was very littleimplementation capacity, and even less capacityto provide oversight over the implementers,inevitably fueled corruption. While thecorruption problem was widely recognized, andfrequently condemned by international officials,many noted that institutional priorities andincentive structures resulted in amounts of aidthat exceeded both the implementation andabsorption capacities of aid agencies, the Afghangovernment, and local communities.Institutional incentive structures to get and spendmoney were generally not matched by incentivesto spend money effectively and accountably. Onetenet of the COIN model was to use money todeliver services that would help legitimize thegovernment. However, a fundamental andperhaps fatal flaw of the COIN strategy was thatAfghanistan’s leaders never seemed to actuallyshare or buy into this legitimization strategy ofwinning the population away from theinsurgents through services. Instead, the strategyadopted by the country’s leadership often seemedto be to consolidate their power by using aidmoney to strengthen patronage networks—which had the effect of reducing thegovernment’s legitimacy by fueling perceptionsof corruption, inequality, and injustice.6.4.2 Competition over resources: The war-aideconomy and perverse incentivesCompelling evidence shows that competitionover aid-related financial and material resourcescan lead to conflict and violence. This wasobserved most clearly in Paktia, where aidproject-relatedresources fuelled violent conflictamong competitors. Violence attributed to theTaliban was in fact often due to contractorwars. 146 As noted by an informed Afghan inPaktia, “if an insurgent group comes and makesa deal with an engineer not to attack, thentomorrow someone else will come and make adeal. When that project ends, then another dealis necessary. Each construction company has itsown security. If they lose out on a large bid, theytry to make the environment insecure for theircompetitors.” 147 One respondent related a specificcase where a Talib lost in the bidding for a road146See also Aram Roston (op cit) .147Interview with local community leader, Paktia Province, January 10, 2008.62Feinstein International Center

project, then killed one of the winning bidder’sengineers and burned vehicles belonging to aforeign construction company. The winningbidder subsequently gave up, and requested thatthe Talib do the work on the road instead.” 148Clearly, some political and economic actorsdirectly benefit from the lack of security, andsecurity providers have themselves createdinsecurity. As has been documented in themedia, providing security has become a bigbusiness for both Taliban and local commanders.In Uruzgan, Matiullah (nephew and closeassociate of former governor Jan Mohammad)earns a steady and lucrative income for securingsafe passage of convoys between Kandahar andTarin Kot, and for providing security for roadconstruction projects. Matiullah reportedlyensures that the road does not become safeenough to allow travel without his services.Similarly, in Paktia, those responsible forproviding security on a road project stagedattacks in order to elicit higher payments. 149 Thedynamic was summed up succinctly by anAfghan UN official: “create a problem and thenget paid to solve the problem.” 150This sort of conflict may be more likely in placesand at times where no one force is dominant andwhere power, influence, and resources arecontested. This may explain why apparentlyfewer of these types of conflicts occur in Balkh,where the political and security situation is moretightly controlled by one individual or group.Even where contractors are not themselvescompeting over resources, projects can createconflict between communities, especially wherethe social structure is fragmented and power iscontested. Road projects were notorious forcreating tension over the distribution of benefits.In Uruzgan and Balkh, local communitiesforcibly stopped work on road projects whenworkers were brought from outside the arearather than hired locally. In another case, inUruzgan, conflict arose when local tribal leadersclaimed to have been told that they would get acontract from Matiullah to provide security oneighteen miles of road but received a contract foronly eight. In Helmand, while several groups ofelders in Babaji (twenty-five kilometers north ofLashkar Gah) approved the route, design, andcontracts to build a road in the area, several othercommunities contested all three and expresseddiscontent with the PRT and ISAF. Similarly, inBalkh, Helmand, and Uruzgan, communitiesopposed the construction of bridges, in partbecause property ownership issues had not beenresolved. During the field research, cases werecited where new water projects led toantagonism between communities. Othersources have noted the longer-term conflictproduced by irrigation projects due to the lack ofconfidence in institutions charged withmanaging them. In Chora District, Uruzgan, forexample, a planned project to rehabilitate anirrigation canal built by NGOs in the 1990sexacerbated tribal animosity. In the last decade,98 people have reportedly been killed.Views were mixed on the role that developmentprojects play in providing incentives (ordisincentives) to communities to create anatmosphere conducive to security. While thedominant narrative in Balkh, for instance, is thatthe insecure areas of the south and east are beingrewarded with development projects, somerespondents observed that if an area is insecure,projects will not come, and that this providedcommunities with (positive) incentives to createsecurity. Some officials said that somecommunities may intentionally be creatingproblems to prevent the monitoring of existingprojects. Many respondents in Balkh and Faryab,including a district police chief, half-jokinglyoffered that a good strategy for attracting aidprojects would be to create a little noise tosuggest that insurgents were operating in the areaand that a little “hearts and minds” activity wasneeded. Some international aid officials claimedthat some communities are using the “mantra of‘bring us development’” as a threat. Manyrespondents, especially in the north, speculatedthat regional disparities might encourage148Ibid. For a media report on this dynamic, see Alissa J. Rubin and James Risen, “Costly Afghanistan Road Project is Marred by UnsavoryAlliances,” New York Times (May 1, 2011).149Ibid.150Interview with Afghan UN official, Tarin Kot, Uruzgan Province, February 2, 2010.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 63

underserved areas to create insecurity to elicitaid, although no evidence was found that thishad been done.Finally, resources obtained in one area maycreate spillover effects in another. For instance,the governor of Paktia, Joma Khan Hamdard,was widely believed to be using money from hiscontracting business to de-stabilize his homearea in northern Afghanistan. 1516.4.3 Reinforcing inequalities and creatingperceived winners and losersAid can also be de-stabilizing if ethnic, tribal, orpolitical groups (some of which can be highlycorrelated, viz. Jumbish and Uzbeks in Faryab)capture the bulk of the aid at local levels, as thiscan reinforce existing inequalities and lead toconflict. This may occur on two different butrelated levels. First, the winner-loser dynamicsbetween powerful tribal and political elites at alocal level can lead directly to instability andviolence in the area. Second, the perceivedinjustice due to inequalities at the mass publiclevel shapes pervasive national narratives thatreduce the government’s legitimacy. The mainevidence for the destabilizing effects of aid camefrom the insecure provinces of Helmand, Paktia,and Uruzgan. In Helmand and Uruzgan, asdiscussed above, the capture of aid by one tribalgroup aligned with President Hamid Karzai’sPopalzai tribe as a result of its political andeconomic power created resentment among thegroups losing out. Also as discussed above, theinternational community’s reliance on andsupport for local and regional strongmen hasexacerbated rivalries. 152 While aid did not itselfcreate these historic rivalries, it often had thedestabilizing effect of reinforcing andexacerbating existing grievances and tensions.According to an international official, “it is moreabout tribal issues of haves and have-nots. Werisk becoming a party to conflicts with ourmoney.” 153 Evidence shows that in some cases,such as in Uruzgan, the losers sought redress inpart by aligning themselves with Taliban groups.6.4.4 Regional disparitiesAs noted above, regional disparities in levels ofaid have aggravated resentments in areas wherethe inhabitants feel underserved. This has largelybeen a north-south split (as noted above, during2009–10, 77 percent of USAID resources wereallocated to the insecure areas of the south,southwest, and east, with a planned increase to81 percent in fiscal year 2011), but the disparitiesalso exacerbate ethnic tensions, as the bulk of theaid has gone to Pashtun-majority areas. 154Disparities also have the potential to undercutthe perceived effectiveness and legitimacy oflocal officials if they are seen as not having beenable to bring resources to their provinces ordistricts. In Balkh, this has led to tensionsbetween the PRT and the governor, as the latterhas, largely for political reasons, complainedpublicly about the lack of support from the PRT,noting the huge discrepancy between the $89million available to the U.S. PRT for NangarharProvince and the $495 thousand available to theSwedish-Finnish PRT for Balkh, Jawzjan,Samangan, and Sar-e Pol Provinces. 155Dissatisfaction may be further fed by the Afghanofficial and private media, which often show thegreat developments happening in Nangarhar;respondents in Balkh referred to the U.S.-fundedconstruction of parks, roads, and buildings thatthey had seen on television.In sum, the study did not find evidence that aidprojects were making a significant contributionto stability in Afghanistan. Similarly, given the151See Kate Clark, “Wikileaks and the Paktia Governor” (Afghanistan Analysts Network, June 12, 2010). http://aan-afghanistan.com/index.asp?id=1367.152For a discussion of how support for local leaders can ultimately be destabilizing, see Dempsey, “Is Spending the Strategy?” See alsoMatthieu Aikins, “Our Man in Kandahar,” Atlantic Magazine (November 2011).153Interview with international official, Tarin Kot, Uruzgan Province, January 31, 2010.154This dynamic has become even more important in the last couple of years, when the government has put so much emphasis onreconciliation with insurgent groups. Aid can be perceived by some groups as another example of excessive accommodation of thepopulation in the south and east.155These are 2009 figures and for military-related projects only. Although they are only one part of the funding picture, they are seen asemblematic of the regional discrepancies.64Feinstein International Center

largely negative perceptions of the aid effort byAfghans, there was little evidence that the largeamount of money spent was helping to winhearts and minds. Several interviewees didsuggest, however, that while Afghan hearts andminds were not being won, these projects didplay an important role in helping to win thehearts and minds of both troops and the publicin troop-contributing nations. While not thefocus of this study, the morale of many troopswas likely boosted by their efforts to improve thelot of Afghans through activities like buildingand supporting schools. Furthermore, the role ofinternational military forces in directlyimplementing and supporting reconstructionactivities to rebuild Afghanistan plausibly helpedgenerate domestic support for the NATO/ISAFmission in troop-contributing nations. This waslikely particularly true for many of the U.S.’sNATO allies who justified the deployment oftroops to Afghanistan to their publics not asfighting a war on terror, but as supporting apeacebuilding operation. It was thereforeimportant that their troops be perceived asfocusing on primarily non-kinetic activities suchas supporting reconstruction efforts.The overall findings of this study are largelycorroborated by other studies and evaluationsmentioned in Section 1.3. As part of an on-goingstudy of whether development cooperationpositively influenced attitudes in northeasternTakhar and Kunduz Provinces, ChristianZürcher and his colleagues concluded thatbetween 2005 and 2009, the ability of aid toinfluence attitudes depended on security andthreat perceptions (i.e., communities that feltmore secure were more likely to feel positiveabout aid); communities’ attitudes towardsforeign forces were driven by perceptions ofsecurity and not by aid projects; positive effectsof aid on the population’s attitudes towards thestate were “short-term” and “non-cumulative”(e.g., based on an attitude of “what have youdone for me lately”); and, acceptance ofdevelopment actors by communities is notreduced by their collaboration with the military.The study also concluded that between 2007 and2009, at a time when aid resources wereincreasing, Afghans in the two provinces hadbecome more distrustful of developmentorganizations; the proportion of communitymembers who felt threatened by foreign forcesincreased; and, the perceived usefulness offoreign forces decreased. 156Research conducted in 2009 by Sarah Ladburyin Kandahar, Wardak, and Kabul Provincesconcluded that young men joined the insurgencyfor a complex combination of reasons, somepersonal and some related to broader grievances.Factors included resonance with religiousenticements due to government corruption andpresence of foreign forces; reaction to the state’sfailure to provide security and justice, andinstead being corrupt and predatory; and as ameans of earning an income. This researchdistinguished between being mobilized andbeing radicalized, with young men possiblybeing mobilized for various reasons withoutnecessarily being radicalized. 157In a study of the NSP conducted between 2007and 2011 by Andrew Beath and his colleagues,research in villages outside of the main conflictareas found that overall, the NSP had positiveeffects on communities’ perceptions of economicwell-being, all levels of government (except thepolice), and of NGOs, the security situation, and(weakly) international forces. This study did not,however, find measurable improvements inactual security, although it looked only atshort-term outcomes. The investigators leaveopen whether the more positive view ofgovernment in NSP communities mightinfluence the extent of support for it, andeventually decrease levels of insurgentviolence. 158A quantitative study supported by the U.S. Armyassessed the relationship between developmentprojects and both security and public opinion.Using district-level database records ondevelopment projects and security incidents forJanuary 2002 through December 2010,supplemented by ISAF-funded public opinion156Böhnke, et al., “Assessing the Impact.”157Ladbury, “Testing Hypotheses.”158Beath, et al., “Winning Hearts and Minds?”Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 65

polling executed since 2008, it found that whileprojects can affect the number of securityincidents in a district, in most cases theirinfluence is so small as to not justify them as aconflict-mitigation tool. In the report’s words,“development exerts too varied and weak aninfluence upon security in active conflictenvironments for it to deserve seriousconsideration as a means of countering oraddressing insurgency.” 159Finally, a report prepared in June 2011 by theMajority Staff for the use of the U.S. SenateForeign Relations Committee concluded, afterexamining the evidence from several studies(including the present one and those mentionedabove), that “the evidence that stabilizationprograms promote stability in Afghanistan islimited . . . the unintended consequences ofpumping large amounts of money into a warzone cannot be underestimated.” 160 Thisconclusion was supported by a March 2010conference, co-sponsored by the FeinsteinInternational Center and Wilton Park, on aidand stabilization, where a range of civilian andmilitary actors concluded that “there was aconsensus that imperatives such as pushingsignificant amounts of funding through thesystem over a short period of time is not aproductive way to stabilize Afghanistan.” 161Photo by Geert GompelmanBridge over Maimana River, Faryab159Wheeler and Stolkowski, “Development as Counterinsurgency.”160Majority Staff, Foreign Relations Committee, United States Senate, “Evaluating U.S. Foreign Assistance to Afghanistan,” (June 8, 2011),p. 2, http://www.gpoaccess.gov/congress/index.html.161Wilton Park, “Winning ‘Hearts and Minds,’” p. 20.66Feinstein International Center

7. Summary of conclusions and recommendationsPhoto by © Kate Holt/IRINAfghan civilians and water pump, HelmandThe following is a summary of the conclusionsdrawn from the research, along with policyrecommendations.As noted above, since the field research wasconcluded, the international community,especially the U.S. and ISAF, has acknowledgeda number of the issues described above, and havetaken some policy and management steps tomitigate them. This includes, for USAID, severalaccountability initiatives, the devolution ofauthority closer to the field, and more attentionto the assessment of local drivers of insecurity.For the U.S. government more broadly, itincludes increased spending through the Afghangovernment and a policy that CERP projectsmust be signed off by local Afghan officials.Despite these policy changes, however, theresearchers feel that the followingrecommendations remain valid and relevant, inlarge part because many of the institutionalincentives for spending aid funds remainunchanged.1. Primacy of political over economicdrivers of conflictThe research findings from the more insurgencyaffectedprovinces in southern and easternAfghanistan suggest that many of the complexand overlapping factors that may have causedpeople to oppose the government and to supportviolence against it were related most frequentlyto political factors such as the corrupt andpredatory behavior of government actors. Yet,most stabilization initiatives have placedimportance on the economic drivers ofconflict—focusing on poverty, unemployment,illiteracy, social services, and infrastructure. Thissuggests that aid and development, even ifappreciated, may not be effective at addressingthe fundamental issues and drivers of insecurity.The Taliban claim to legitimacy based on Islam,justice, and security may be more resonant inmany unstable areas than the promises ofdevelopment assistance by the government andits international partners. On the other hand, inareas not affected by insurgency where povertyWinning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 67

or conflict over land and water were named asmore important drivers of conflict, welldelivered,conflict-sensitive aid interventionsmight have had a more stabilizing effect. An areafor additional research would be to test thehypothesis that aid may be more effective inconsolidating stability in relatively stable areasthan in stabilizing insecure areas.There was little concrete evidence from this orother studies that aid led to stability inAfghanistan. However, given the pervasivenegative attitudes towards aid projects and actors,whether aid did not lead to stability due toineffective delivery or because it simply did notadequately address the main drivers of conflict isdifficult to definitively determine. If aid hadbeen delivered more effectively, in smalleramounts, with more community participation,with more realistic timeframes, and with betterknowledge of local political and tribaldynamics—it possibly would have had a morestabilizing effect.On the other hand, one of the fundamental flawsof the population-centered COIN approach fromthe beginning was its reliance on winning thepopulation away from insurgents and over to thegovernment. This strategy struggled to gaintraction in part because the government’sleadership never seemed to accept it as theirstrategy and instead often pursued a patronagebasedapproach to buy the support of localstrongmen over winning the population over.Another problem was that the U.S. and many ofits NATO/ISAF allies actually had contradictorystrategies, simultaneously wanting to provideservices and good government to win over thepopulation, but also supporting local strongmenwhose predatory behavior alienated the localpopulation. Aid delivered by or associated withcorrupt officials or strongmen who were in manycases responsible for alienating people in the firstplace has not surprisingly proven to be anineffective way of winning people over to thegovernment. Much more acknowledgement isneeded of the fact that lack of progress ongovernance has not primarily been due to lack ofmoney, but to a lack of political will or a sharedstrategy on the part of the government and theinternational community to push a consistentreform agenda.Recommendations:• Focus more on identifying the drivers ofconflict and alienation, and if these areprimarily political in nature, do not assumethey can effectively be addressed throughprimarily socio-economic activities. Given thestrong perception amongst Afghans that badgovernance and the lack of rule of law aremajor causes of alienation from thegovernment and of insecurity, more attentionshould be given to addressing these issues.• The international community should take abetter-coordinated and more forceful stand oncertain key issues that would help promotebetter governance (e.g., merit-basedappointments to key national and sub-nationalpositions, more rigorous anti-corruptionmeasures including better monitoring ofexpenditures, avoiding alliances withnotorious strongmen known for corrupt andpredatory behavior).2. Spending too much too quickly can becounterproductive—less can be moreEvidence from this study has shown that thepressure to spend too much money too quickly iswasteful and undermines both security anddevelopment objectives. Funding fewer projectsthat are better planned, implemented, andmonitored will have more positive effects thanspending more money in insecure areas that havealready exceeded their absorptive capacity toeffectively and accountably receive aid money.A powerful institutional incentive is to spend asmuch money as quickly as possible, whichrewards quantity over quality. This concept wascaptured in the pressure many internationalcivilian and military officials noted to maintainhigh levels of expenditure (“burn rates”), in partso that their agencies are not penalized in thenext budget cycle (the “use it or lose it”problem). Strikingly, one of the few programsthat Afghans positively describe is the NSP,which on average disburses $27,000 percommunity. The experience of the NSP andsome other development projects suggests that,in terms of development, quality is moreimportant than quantity, as is the sense thatbenefits are equitably shared. The research68Feinstein International Center

suggests that in terms of potential stabilizingbenefits, the process of development, especiallyin terms of building and sustaining relationships,is as important as the product of development.Taking more time and spending less money oftentranslates into more effective programs withgreater impact and better overall outcomes. Theexpressed concerns of field-level developmentworkers appear at odds with the amount ofmoney being programmed by those above them.There are few incentives for spending less moneybut spending it more effectively over time. “Lessis more” can never be a reality when “more ismore” is rewarded.Also, in an uncertain, competitive, and highlyresource-scarce environment with weakinstitutions, the attempt to spend too muchmoney too quickly virtually guarantees failure inachieving development outcomes, and greatlyincreases the odds that aid will in fact fuelcorruption and instability. With the increasinguncertainty about what will happen during andafter the post-2014 transition, in some quartersthe sharp increase in spending since 2009 mayfunction almost as a “last call” at the pub.Recommendations:• Provide incentives for quality and impact ofaid spending over quantity. Aid money shouldonly be committed when it can be spent in aneffective and accountable manner.• Establish multi-year, predictable funding (asrecommended by the Majority Committeestaff report to the Senate Foreign RelationsCommittee) and/or a longer-term fund thatcould be drawn down (as recommended by theWilton Park conference). These approacheswould reduce the current institutionalincentives and negative effects of spending toomuch too fast, while also conveying a sense oflong-term international commitment toAfghanistan.3. Insufficient attention has been paid tothe political economy of aid inAfghanistanAn important consequence of the pressure tospend so quickly has been inadequateconsideration of the incentive structures facingpolicy makers, donors, implementers, andcommunities. Evidence from this as well as otherstudies indicates that the way in which aid hasbeen delivered has contributed to instabilitythrough reinforcing uneven and oppressivepower relationships, favoring or being perceivedto favor one community or individual overothers, and providing a valuable resource foractors to fight over. The most destabilizingaspect of the war- aid economy in Afghanistan,however, has been its role in fueling corruption,which reduces the legitimacy of both thegovernment and the international community.Not only are foreign aid projects unlikely tomake either the Afghan government or itsinternational backers more popular, butreconstruction assistance seems to be losing—rather than winning—hearts and minds.This study suggests that institutional incentivestructures and the planning process work againstthe changes in the allocation of aid that would besuggested by research-based evidence. From localstrongmen-aligned militias to aid agencies tomilitary and international aid agency officials inWestern capitals, personal and institutionalinterests play a significant role in continuing aidprograms that the evidence suggests areinappropriate, too large, or otherwisedysfunctional. Under the current status quo ofweak institutions and insecurity, some actors aredoing quite well, and so have little incentive topush for change.If improperly designed aid programs are creatinginstability, then the current structures andincentive systems of political-aid-militarybureaucracies are themselves drivers of conflict.Discussions with individual field-level actors aswell as senior officials confirm that the problemis often not that we do not know what needs tobe done, but rather that institutional incentivesgenerally reward getting and spending moneyquickly rather than carefully and effectively.Recommendations:• Invest more in understanding the politicaleconomy of aid, including local conflictdynamics, the impact of the war-aid economyon these dynamics, the perceived winners andlosers of aid programs, and the role of theseWinning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 69

programs in legitimizing (or delegitimizing)the government.• Give more attention to understanding theincentive structures of national andinternational civilian and military institutions,and the impact of these incentive structures onthe effective delivery of developmentassistance.4. Insecurity rather than security isrewardedThe primary objective of U.S. aid to countriessuch as Afghanistan is not development for itsown sake, but rather to promote security. As aresult, funding for insecure areas takes priorityover secure areas. Since 2001, the bulk ofUSAID’s assistance program in Afghanistan hasbeen spent in insurgency-affected provinces inthe south and east. The last several years haveseen an even greater prioritization of insecureareas despite the lack of evidence that the aidfunds being spent are promoting stability orimproving attitudes towards the Afghangovernment and the international community.The findings from this study and other researchsuggest that aid is more effectively spent insecure regions where good development practiceand stronger oversight is more feasible, and lessmoney has to be spent on security. The researchalso suggests that in areas where insecurityremained chronic and governance structuresbroken, resources (e.g., for road building) havetended to fuel corruption (both perceived andreal), inter-communal strife, and competitionamong local powerbrokers. Not surprisingly, theprioritization of insecure over secure areas isbitterly criticized by Afghans living in morestable areas, who feel they are being penalizedfor being peaceful. While no evidence was foundthat relatively peaceful communities wereactually creating “a little bit of noise” to attractmore aid resources, there was evidence that ininsecure areas local strongmen with militias thatwere being paid to provide security recognizedthe need to perpetuate insecurity—i.e., to“create a problem to solve a problem.”Recommendation:• Reward security, not insecurity, by reversingthe current policy of rewarding insecure areaswith extensive aid while effectively penalizingsecure areas where aid money could be spentmore effectively and accountably. Invest insecure areas and, except for humanitarianassistance, make aid in insecure areas morecontingent on security. While this study didnot specifically examine this issue, this couldhave a demonstration effect and provide anincentive for communities to seek improvedsecurity—rather than the reverse.5. Accountability and the measurement ofimpact have been undervaluedThe political need for “quick impact” along withinstitutional imperatives to spend money have inmany cases reduced the incentives for carefulevaluation of project impact. Currently it is noteven possible to get a complete list of the projectsPRTs have implemented since they were formed,let alone an indication of what the impact hasbeen of the approximately $2.64 billion in CERPfunds appropriated for Afghanistan between 2004and 2010. The study’s findings have beenreinforced by the increasing number of media andU.S. agency reports noted above (e.g., SpecialInspector General for Afghanistan Reconstruction[SIGAR], Office of the Inspector General [OIG])on funds that have been wastefully spent with no(or negative) impact.While in an environment with little reliablequantitative data, with numerous independentvariables that make determining correlation (notto mention causality) virtually impossible, andwhere Western-style public opinion pollingmethodologies may not be reliable, thedetermination of impact may have to be more artthan science. Nevertheless, much more focusshould be given to measuring the impact andconsequences of aid projects than has been doneto date. Recent initiatives by SIGAR, OIG, andstaff at the Senate Foreign Relations Committeeare positive, but they come late in the game. Inaddition to the waste of taxpayer resources andnegative consequences on the ground, thediscrediting of all programs for Afghanistan maybe collateral damage if aid resources are not spentin a more accountable and effective manner.70Feinstein International Center

Recommendation:• Reinforce at all levels the message and cultureof accountability and effectiveness. This is nota recommendation to add several morebureaucratic levels of cumbersome nationaland international oversight mechanisms tooversee inputs, but rather to invest more inmeasuring outcomes. Establish incentivestructures for quality work and carefulassessments of effectiveness and not just forspending money.6. Development is a good in and of itselfThere is considerable evidence that developmentassistance in Afghanistan during the past decadehas directly contributed to some very positivedevelopment benefits, including decreases ininfant and maternal mortality, dramatic increasesin school enrollment for boys and girls, a mediarevolution, major improvements in roads andinfrastructure, and greater connectivity throughtelecommunication networks. One consequenceof viewing aid resources first and foremost as astabilization tool or “a weapons system” is thatthese major development gains have often beenunder-appreciated because they did not translateinto tangible security gains. U.S. developmentassistance in Afghanistan has been justified onthe grounds that it is promoting COIN orstabilization objectives rather than developmentobjectives. While in the short term this has ledto much higher levels of development assistancein Afghanistan, the failure of these resources toimprove the security situation is now leadingmany policymakers to question the value ofdevelopment assistance.Recommendation:• Value development as a good in and of itself.Program development aid first and foremost topromote development objectives, where thereis evidence of impact and effectiveness, ratherthan to promote stabilization and securityobjectives, where this research suggests there islittle evidence of effectiveness.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 71

Annex A. Research MethodologyThe objective of the Afghanistan aid and securityresearch project has been to better understandthe effectiveness of aid in “winning hearts andminds” and promoting stabilization and securityobjectives. The following section describes thedefinitions and research methodology used toachieve this objective.DefinitionsThis study has largely used U.S. militarydefinitions for terms such as “stabilization,”“stability operations,” and “winning hearts andminds.” This was done because the U.S. isdeploying most of the military and non-militaryaid intended to promote stability objectives inAfghanistan, and U.S. military doctrine(especially COIN doctrine) is driving thestabilization agenda in Afghanistan. Therefore,using the U.S. military’s own definitions todetermine the effectiveness of efforts to use aidto promote stability objectives seemed mostappropriate.The foreword to the U.S. Army’s StabilityOperations manual (FM 3-07) states that “thegreatest threat to our national security comes notin the form of terrorism or ambitious powers, butfrom fragile states either unable or unwilling toprovide for the most basic needs of their people.” 1The manual, which resulted from DOD Directive3000.05 (2005) that directed that stabilityoperations “be given priority comparable tocombat operations,” 2 defines “stability operations”and “stabilization” as follows:Stability Operations—“Various militarymissions, tasks, and activities conducted outsidethe United States in coordination with otherinstruments of national power to maintain orreestablish a safe and secure environment,provide essential governmental services,emergency infrastructure reconstruction, andhumanitarian relief.” 3Stabilization—“The process by whichunderlying tensions that might lead toresurgence in violence and a breakdown inlaw and order are managed and reduced,while efforts are made to supportpreconditions for successful long-termdevelopment.” 4Winning hearts and minds—The conceptof “winning hearts and minds” is moredifficult to define precisely as, even withinthe U.S. military, different actors use theterm differently. Unlike “stabilityoperations” or “stabilization,” no one precisedefinition of the term “hearts and minds”exists. Rather, it has been used as a sort ofshorthand and, in the translation fromdoctrine to field-level vernacular, has beenmuch abused. The U.S. Army’sCounterinsurgency manual (FM 3-24) explainsthe phrase as follows:Once the unit settles into the AO [area ofoperations], its next task is to build trustednetworks. This is the true meaning of the phrase‘hearts and minds,’ which comprises twoseparate components. ‘Hearts’ means persuadingpeople that their best interests are served byCOIN success. ‘Minds’ means convincing themthat the force can protect them and that resistingit is pointless. Note that neither concernswhether people like Soldiers and Marines.Calculated self-interest, not emotion, is whatcounts. Over time, successful trusted networksgrow like roots into the populace. They displaceenemy networks, which forces enemies into theopen, letting military forces seize the initiativeand destroy the insurgents. 5Despite the cautionary note that winninghearts and minds is not about getting peopleto like military forces, many of theinternational military personnel interviewed1Department of the Army, Stability Operations, FM 3-07 (October 2008), p. vi.2DOD, Directive 3000.05: Military Support for Stability, Security, Transition, and Reconstruction (SSTR) Operations” (November28, 2005), p. 2.3Department of the Army, Stability Operations, p. vi.4Ibid, Glossary-10.5Department of the Army, Counterinsurgency, FM 3-24/MCWP (2006), p. A-5.72Feinstein International Center

for this study did perceive this to be an importantobjective of their aid efforts. 6 Even morecommon, however, was the view that the primaryobjective of aid projects was to make thepopulation like and support the Afghangovernment. Not surprisingly, many of themilitary and civilian PRT officials who wereinterviewed had this view, given that the primaryobjective of NATO/ISAF PRTs is to “assist theIslamic Republic of Afghanistan to extend itsauthority.” 7 This study did not adopt any onedefinition for the phrase “winning hearts andminds,” but rather tried to explore how differentactors understood and used the term, and tounderstand its effectiveness—whether in terms ofbuilding trusted networks, generating consent andsupport for the presence of foreign troops, orlegitimizing the Afghan government.Field Research MethodologyThe research team conducted field research inKabul and five provinces—Balkh, Faryab,Helmand, Paktia, and Uruzgan. In theseprovinces, as in nearly all of Afghanistan’sthirty-four provinces, international civilian andmilitary actors are making efforts to usehumanitarian, reconstruction, and developmentaid to promote greater stability and security.However, notable differences between the fiveprovinces provided opportunities to examine thedevelopment-security nexus in very differentcontexts. For example, Balkh and FaryabProvinces in the north were much more securethan Helmand, Uruzgan, and Paktia Provincesin the south and southeast where the Taliban-ledinsurgency was much more active. In the twonorthern provinces, the Pashtun were a minorityethnic group, whereas in the south and southeastthey comprised the overwhelming majority.Another significant difference was the variationsin approach, budgetary resources, and characterof the different NATO/ISAF nations heading thePRTs in each province.The study team used a relatively consistentmethodology in four of the five provincial studyareas (Helmand being the exception 8 ), bearing inmind that the varied security and otherconditions allowed or required somewhatdifferent approaches in different areas. Fieldbasedinterviews with Afghan and internationalrespondents provided the primary data source forthis study. These were conducted between June2008 and February 2010 during multiple visits toBalkh, Faryab, Paktia and Uruzgan Provinces.As detailed in Table 1 at the end of this annex,574 people were interviewed, including 340Afghan and 234 international respondents. Theseprimary data were supplemented by informationfrom secondary sources, including existingdatabases (e.g., the Central Statistics Office’sNational Risk and Vulnerability Assessment, theNATO/ISAF Afghanistan Country StabilityPicture, donor project lists), surveys, publicopinion polls, media articles, and a wide varietyof published and unpublished reports.Approval from the Tufts University InstitutionalReview Board was obtained in advance of thecommunity-level fieldwork. In accordance withstandard procedures for informed consent,respondents were told orally that theirparticipation was voluntary, that their responseswould be confidential, and that they couldterminate the interview at any point. In somecases (i.e., with staff of international agencies andaid contractors) this information, along withbackground material on the study, was providedby email in advance of the interview.6Press accounts from Afghanistan often quote military forces suggesting that their aid projects are intended to generate good willamong local populations for their presence. For example, “If [soldiers] can spread the message that, ‘Hey, coalition forces builtnew toilets,’ it makes us seem that much more legitimate, and makes them more willing to work with us,” said Zambarda, ofthe 2-12 Infantry, Dagger Company, as quoted by Bradley Blackburn in, “‘Warrior-Diplomats’ on the Front Lines inAfghanistan: US Forces on a Dual Mission to Fight the Enemy and Reach Out to Him,” ABC News, May 12, 2010.7ISAF, ISAF PRT Handbook, Edition 3 (2007), p. 2.8The Helmand case study focused specifically on whether the UK government’s Quick Impact Projects in Helmand between2006 and 2008 were demonstrating impact. The methodology consisted of analysis of qualitative data from focus groups,quantitative data taken from polling data drawn from communities and provided by the PRT, and interviews with keyinformants (e.g., PRT staff, Afghan government officials). These primary data were supplemented by extensive review ofsecondary sources.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 73

The original plan had been to gather qualitativedata through focus group discussions withcommunity members and semi-structured keyinformant interviews with Afghan andinternational officials. However, during the firstround of field research in Paktia and BalkhProvinces in June and July 2008, it became clearthat semi-structured interviews with individuals(or on occasion small groups) at the communitylevel generated more fine-grained and nuancedinformation than focus group discussions.Afghan social hierarchy may discouragewillingness to talk openly or express ideas thatviolate social norms, or may encourage a sort ofgroupthink. This is likely to be especially truefor sensitive topics such as the influence of localpower holders or the characteristics of thegovernment. At the same time, while theresearch teams tried to obtain individualinterviews, social protocols (i.e., that it isconsidered rude to ask people to leave a room)sometimes required that interviews take place ina group setting.The semi-structured interviews were conductedusing two different questionnaires—one forAfghan and international officials and one forcommunity-based respondents. Thequestionnaires were developed by the principalinvestigator (PI) and field tested during a June-July 2008 visit to Paktia and Balkh Provinces.The semi-structured nature of the interviewsmeant that not all questions were asked of allrespondents, and issues were discussed indiffering levels of detail depending on thebackgrounds of the respondents and the timeavailable for interviewing. The interviews withkey informants included current and formergovernment officials, donors, diplomats, militaryofficials, PRT personnel, journalists, and UNand aid agency staff. The community-levelresearch included interviews with tribal andreligious leaders, local government officials,members of civil society organizations, tradersand shopkeepers, beneficiaries of specificreconstruction and development projects, andcommunity members more generally.The semi-structured interviews followed astrategic structure of clusters linked to specificthemes such as actors, aid effectiveness, andsecurity. The order reflected the degree ofpotential sensitivity, from an initial request forstraightforward information and progressing topersonal views. The structuring strategy wasused to develop trust before potentially sensitivequestions about security were asked. To initiatetrust and rapport prior to each interview, auniform method was employed to briefly andinformally introduce the reasons for the researchand how the information would be used whilestressing and demonstrating confidentiality. Eachinterview concluded by asking for furthercomments and questions.Most of the interviews with Afghans wereconducted in Dari or Pashtu, although someinterviews with senior government and NGOofficials were conducted in English. In northernAfghanistan nearly all the interviews withAfghans were conducted in Dari. The twointernational researchers leading the fieldresearch in Balkh and Faryab Provinces wereexcellent Dari speakers, and could directlyinterview Afghan respondents. They wereassisted in setting up and conducting interviews,as well as in note taking and analysis, by Afghanresearch assistants. In Faryab, a small number ofinterviews were conducted in Uzbeki, whichwas immediately translated orally to theresearcher and research assistant. As respondentswere able to understand Dari, they were able tointervene if their answers had been incorrectlytranslated. In Paktia Province most interviewswere conducted in Pashtu, with the help of aresearch assistant translator, although some of theinterviews with government and aid agencyofficials were conducted in Dari or English. InUruzgan Province one of the internationalresearchers could conduct some interviewsdirectly in Dari, although a translator was usedfor interviews where respondents only spokePashto. The interviews varied in lengthdepending on circumstances, but generally theylasted one to two hours (although some went onfor more than four hours).The field research initially was designed to beimplemented in partnership with theAfghanistan Research and Evaluation Unit(AREU), an independent policy researchorganization based in Kabul, which would takeprimary responsibility for conducting thecommunity-based field research. However,74Feinstein International Center

following the deterioration in the securityenvironment—which made conductingcommunity-based research in the south andsoutheast increasingly difficult and dangerous forboth researchers and research subjects—and thedifficulty in finding and retaining qualifiedresearchers to lead the community research, itwas decided that the community research wouldbe scaled back and that the FeinsteinInternational Center (FIC) PI, researchconsultants, and research assistants wouldconduct all the field research. The most negativeconsequence of this decision was that withoutAREU’s male and female research teams, theability to interview women in the culturallyconservative and gender-segregated contexts atthe community level was greatly constrained.While the FIC researchers were able to interviewa number of women, these were mostly theAfghan staff of NGOs and international agenciesand some government officials; they includedvery few women at the community level.CaveatsAny research in Afghanistan, and particularlyresearch that looks at the types of sensitive issuesraised in this study, requires a number of caveats.Specific to the present study, the relationshipsbetween aid and security and the notion of“winning hearts and minds” are hard to define,much less measure. This difficulty iscompounded by the insecure context in whichmuch of the field research was conducted, whichdemands that special consideration be given toensuring that both researchers and researchsubjects are not put at risk.In general, field research benefits a great dealfrom establishing trust and understanding amongrespondents, especially before posing questionsabout the role of the international military andother powerful actors. While the concept ofobjective research is obscure in rural and evenurban Afghanistan, the concept of the meddlingoutsider is not, and visitors asking sensitivequestions may raise suspicions and inhibitresponses. Likewise, the phenomenon of the“survey” has become common in recent years,and community members may interpret visits toask questions about aid projects as yet another“survey.” This can both raise hopes and generatefrustration, and respondents may try to outdoeach other (and nearby communities) indescribing the devastation and neglect of theirarea in order to attract development projects.Aside from the hope of getting something out ofthe transaction, people like to highlight theirproblems and, given the opportunity to do so,may overstate negative attitudes. On the otherhand, the Afghan notion of hospitality towardsguests may inhibit some respondents from tellingtruths that they perceive will offend a (foreign)visitor, including those about what people reallyfeel about the foreign military and theinternational community. Afghan socialhierarchy, especially in a group setting, willoften result in the voices of the elders and thepowerful being heard, while others lower downon the social scale are expected to keep quiet anddefer. Moreover, given the separation of homeand public spaces, interactions with outsidersoften occur in the public space, and the resultinglack of privacy means that even carefullyorganized and planned private interviews caneasily become public focus groups. Finally, evenin the relatively peaceful northern areas, securityand mobility limitations constrain researchersfrom moving about at will, restricting theirchoice of fieldwork areas and even with whomthey can interact. As respondents’ perceptionsdepend largely on their own situation, and onwhether or not they have benefited from aidprojects and processes, restrictions on mobilityobviously affect the ability to triangulateinformation provided by respondents and to findthe “truth” about what actually happened incertain projects.The study relied primarily on the statedperceptions of the wide range of respondentsmentioned above, and, where relevant, thediscussion differentiates the perspectives ofdifferent types of actors. The researchersacknowledge the need for caution when basingfindings on the stated perceptions of respondents,as respondents’ statements may not alwaysaccurately present their perceptions and inaddition may not match behavior. The study didnot aim to measure causality, as this was simplytoo ambitious in an environment with so manyconfounding variables. Still, because aid projectsexplicitly aim to change attitudes, perceptions (ifcaptured accurately) are relevant. Moreover,however imperfect, the research team believedWinning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 75

that in the Afghan context the qualitative datagathered in in-depth interviews provided abetter and more nuanced data source and gaugeof perceptions than most data collected usingquantitative methodologies, such as publicopinion polling.Despite the above caveats, the methodologyoffered a number of advantages: repeat visits tofollow up on observations, flexible semistructuredinterviews that allowed spontaneousresponses, and triangulation of responses amongexperienced team members who had all spentsignificant amounts of time in the field.Confidence in the methodology was borne outby the remarkably consistent core findings acrossall five provinces as well as across informants.Since the field research was completed in early2010, a number of the issues raised in thefindings have been acknowledged by the U.S.and by ISAF, and measures taken to mitigatethem, as discussed in the text. Securityconditions have also changed somewhat. WhileBalkh and Faryab are still relatively secure,insecurity in the north in has widened ingeneral, and the troubled districts in theprovinces of Balkh and Faryab have becomemuch more so. On the other hand, security inareas of Helmand has improved since the time ofthe research. Nevertheless, based on more recentvisits and discussions as well as the analysis ofothers, the researchers feel that the broadconclusions and concerns remain valid and verypolicy-relevant.76Feinstein International Center

Table 1Comprehensive List of Respondents by Province and TypeRespondent Category Afghan International TotalBalkhGovernment (civilians) 15 24 39Military 1 27 28UN Agencies 2 13 15Aid Agencies (NGOs, contractors) 15 14 29Community Members 57 0 57Others (journalists, analysts, etc.) 3 0 3Balkh Sub-total 93 78 171FaryabGovernment (civilians) 26 0 26Military 0 7 7UN Agencies 5 8 13Aid Agencies (NGOs, contractors) 8 6 14Community Members 79 0 79Others (journalists, analysts, etc.) 0 0 0Faryab Sub-total 118 21 139PaktiaGovernment (civilians) 21 6 27Military 0 16 16UN Agencies 3 8 11Aid Agencies (NGOs, contractors) 5 4 9Community Members 35 0 35Others (journalists, analysts, etc.) 1 1 2Paktia Sub-total 65 35 100UruzganGovernment (civilians) 11 29 40Military 11 26 37UN Agencies 5 1 6Aid Agencies (NGOs, contractors) 11 7 18Community Members 14 0 14Others (journalists, analysts, etc.) 2 3 5Uruzgan Sub-total 54 66 120KabulGovernment (civilians) 1 7 8Military 1 9 10UN Agencies 1 5 6Aid Agencies (NGOs, contractors) 5 10 15Community Members 0 0 0Others (journalists, analysts, etc.) 2 3 5Kabul Sub-total 10 34 44TOTALGovernment (civilians) 74 66 140Military 13 85 98UN Agencies 16 35 51Aid Agencies (NGOs, contractors) 44 41 85Community Members 185 0 185Others (journalists, analysts, etc.) 8 7 15TOTAL 340 234 574Notes: 1. Kabul interviews were conducted by research team members as input to allprovincial case studies.2. As the Helmand case study used a different methodology, the number ofrespondents is not given here.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 77

BibliographyAikins, Matthieu. 2011. “Our Man in Kandahar.” Atlantic Magazine. November.Barfield, Thomas. 2010. “Afghanistan: A Cultural and Political History.” PrincetonUniversity Press.BBC News. 2011. “Hamid Karzai says Afghanistan aid teams must go.” February 8. http://www.bbc.co.uk/news/world-south-asia-12400045.Beath, Andrew, Fotini Christia, and Ruben Enikolopov. Undated. “‘Winning Hearts andMinds,’ Evidence from a Field Experiment in Afghanistan.” Working Paper No. 2011-14.Massachusetts Institute of Technology, Political Science Department.Bennett, Jon, et al. 2009. Country Programme Evaluation, Afghanistan. DFID, EvaluationReport EV696. May.Blackburn, Bradley. 2010. “‘Warrior-Diplomats’ on the Front Lines in Afghanistan: USForces on a Dual Mission to Fight the Enemy and Reach Out to Him.” ABC News.May 12.Böhnke, J., J Koehler, and C. Zürcher. 2010. “Assessing the Impact of DevelopmentCooperation in North East Afghanistan 2005–2009: Final Report.” Evaluation Reports 049.Bonn: Bundesministerium für wirtschaftliche Zusammen arbeit und Entwicklung.Bradbury, Mark, and Michael Kleinman. 2010. Winning Hearts and Minds? Examining theRelationship between Aid and Security in Kenya. Feinstein International Center, TuftsUniversity. April.Bremner, Charles, and Marie Tourres. 2009. “French Opposition Demands Answers onBribe Claim in Sarobi Ambush.” The Times (London). October 16. http://www.timesonline.co.uk/tol/news/world/europe/ article6876691.ece.Chandrasekaran, Rajiv. 2011. “U.S. Military Dismayed by Delays in 3 Key DevelopmentProjects in Afghanistan.” Washington Post. April 28. http://www.washingtonpost.com/world/ us-military-dismayed-by-delays-in-3-key-development-projects-inafghanistan/2011/04/22/AFD6jq8E_story.html.Clark, Kate. 2010. “Wikileaks and the Paktia Governor.” Afghanistan Analysts Network.June 12. http://aan-afghanistan.com/ index.asp?id=1367.Cloake, John. 1985. “Templer: Tiger of Malaya: The Life of Field Marshal Sir GeraldTempler.” Harrap.Committee on Foreign Relations, United States Senate. 2011. “Evaluating U.S. ForeignAssistance to Afghanistan.” Majority Staff Report. Washington, DC. June 8. http://www.gpoaccess.gov/congress/index.html.78Feinstein International Center

Dempsey, Scott. 2011. “Is Spending the Strategy?” Small Wars Journal. May 4. http://smallwarsjournal.com/jrnl/art/ is-spending-the-strategy.DEVRES. 1988. “Retrospective Review of US Assistance to Afghanistan: 1950–1979.”Submitted to USAID. September 30.Duffield, Mark. 2001. “Governing the Borderlands: Decoding the Power of Aid”. Disasters.Volume 25, Issue 4. December.Duffield, Mark. 2001. “Global Governance and the New Wars: The Merging ofDevelopment and Security.” Zed Books.Duffield, Mark. 2007, “Development, Security and Unending War: Governing the World ofPeoples.” Polity Press.Eide, Espen Barth, Anja Therese Kaspersen, Randolph Kent, and Karen von Hippel. 2005.“Report on Integrated Missions: Practical Perspectives and Recommendations.” IndependentStudy for the Expanded UN ECHA Core Group. May.Eronen, Oskari. 2008. PRT Models in Afghanistan: Approaches to Civil-MilitaryIntegration. Crisis Management Center Finland, Civilian Crisis Management Studies, Vol. 1:No. 5/2008.Fishstein, Paul. 2009. “Afghans Need to Find a New Model of Democracy.” Financial Times.October 28. www.ft.com/ cms/s/0/32d3b638-c362-11de-8eca-00144feab49a.html.Fishstein, Paul. 2010. Winning Hearts and Minds? Examining the Relationship between Aidand Security in Afghanistan’s Balkh Province. Feinstein International Center, TuftsUniversity; November.Foschini, Fabrizio. 2010. “The Kuchi-Hazara Conflict, Again.” Afghan Analysts Network.May 27. http://aan-afghanistan.com/index.asp?id=764.Gates, Robert M. 2007. Testimony to U.S. Senate Committee on Appropriations. February.http://usacac.army. mil/cac2/call/docs/10-10/ch-8.asp.Gates, Robert M. 2008. Speech given at U.S. Global Leadership Campaign, Washington,DC. July 15. http://www.defenselink.mil/speeches/speech.aspx?speechid=1262.Gompelman, Geert. 2011. Winning Hearts and Minds? Examining the Relationshipbetween Aid and Security in Afghanistan’s Faryab Province. Feinstein International Center,Tufts University. January.Gordon, Stuart. 2011. Winning Hearts and Minds? Examining the Relationship betweenAid and Security in Afghanistan’s Helmand Province. Feinstein International Center, TuftsUniversity. April.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 79

International Security Assistance Force. 2009. “ISAF Provincial Reconstruction Team(PRT) Handbook, Edition 4.” March.Johnson, Gregory, Vijaya Ramachandram, and Julie Walz. 2011. “The CommandersEmergency Response Program in Afghanistan: Refining U.S. Military Capabilities inStability and In-Conflict Development Activities.” Working Paper 265. Center for GlobalDevelopment. September.Jones, Bruce, and Feryal Cherif. 2003. “Evolving Models of Peacekeeping: PolicyImplications and Responses.” Center on International Cooperation, New York University.September. http://www.peacekeepingbestpractices. unlb.org/.Ladbury, Sarah, in collaboration with Cooperation for Peace and Unity (CPAU). 2009.“Testing Hypotheses on Radicalisation in Afghanistan: Why Do Men Join the Taliban andHizb-i Islami? How Much Do Local Communities Support Them?” Independent Report forthe Department of International Development. August.Larson, Anna. 2009. “Toward an Afghan Democracy? Exploring Perceptions ofDemocratisation in Afghanistan.” Discussion Paper. Afghanistan Research and EvaluationUnit. September.Leader, Nick, and P. Colenso. 2010. “Working Effectively in Conflict-Affected and FragileSituations,” Briefing Paper E: Aligning with Local Priorities, DFID Practice Paper. March.Malkasian, Carter, and Gerald Meyerle. 2009. “Provincial Reconstruction Teams: How DoWe Know that They Work?” March. http://www.strategicstudiesinstitute.army.mil/pubs/display.cfm?pubID=911.Mukhopadhyay, Dipali. 2009. “Warlords as Bureaucrats: The Afghan Experience.”Washington, DC, Carnegie Endowment for International Peace. August.Nesbitt, Francis Njubi. 2009. “Hearts and Minds and Empire.” Foreign Policy in Focus. March20. http://www.fpif.org/articles/hearts_and_minds_and_empire.Nye, Joseph. 2004. “The Benefits of Soft Power.” Working Knowledge. Harvard BusinessSchool. August . http://hbswk.hbs.edu/ archive/4290.html.OECD. 2006. “DAC Peer Review: Main Findings and Recommendations, Review of theDevelopment Co-operation Policies and Programmes of United States.” https://www.oecd.org/document/27/0,2340,en_2649_34603_37829787_1_1_1_1,00.html.Open Society Foundations and The Liaison Office. 2011. “The Cost of Kill/Capture: Impactof the Night Raid Surge on Afghan Civilians.” September.Patrick, Stewart, and Kaysie Brown. 2007. “The Pentagon and Global Development: MakingSense of the DOD’s Expanding Role.” Working Paper Number 131. November.Ploch, Lauren. 2011. “Africa Command: U.S. Strategic Interests and the Role of the U.S.Military in Africa.” Congressional Research Service. July 22. http://opencrs.com/document/RL34003/.80Feinstein International Center

Roston, Aram. 2009. “How the US Funds the Taliban.” The Nation, 11 November. www.thenation.com/doc/20091130/roston.Rubenstein, Leonard S. 2009. “Post-Conflict Health Reconstruction: New Foundations forU.S. Policy.” United States Institute of Peace Working Paper. September.Rubin, Alissa J., and James Risen. 2011. “Costly Afghanistan Road Project is Marred byUnsavory Alliances.” New York Times. May 1.Special Inspector General for Afghanistan Reconstruction. 2011. “Quarterly Report to theUnited States Congress.” January.Stein, Michelle M. “Lawmakers Question CERP Funds in Afghanistan.” Military Times.August 8. http://medilldc.net/2011/08/ lawmakers-question-cerp-funds-inafghanistan/.Strand, Arne, and Organisation for Sustainable Development and Research (OSDR). 2009.“Faryab Survey: Comparison of Findings from Maymane 2006 and 2009.” ChristianMichelsen Institute.Suhrke, Astri. 2006. “When More is Less: Aiding Statebuilding in Afghanistan.” FRIDEWorking Paper. September.Suhrke, Astri . 2011. “Exogenous State-Building: The Contradictions of the InternationalProject in Afghanistan,” in The Rule of Law in Afghanistan: Missing in Inaction, ed. WhitMason . Cambridge University Press.Tarnoff, Curt. 2010. “Afghanistan: U.S. Foreign Assistance.” Congressional ResearchService. August 12.Tribal Analysis Center. 2009. “The Quetta Shura: A Tribal Analysis.” October. http://www.tribalanalysiscenter.com/PDF-TAC/Quetta%20Shura.pdf.United Kingdom Ministry of Defense. 2010. British Army Field Manual: Countering Insurgency,Vol. 1, Part 10. January.United Nations. 2005. “In Larger Freedom: Towards Development, Security and HumanRights for All.” A/59/2005. March 21.United Nations Assistance Mission in Afghanistan and the Afghanistan Independent HumanRights Commission. 2011. “Afghanistan: Annual Report on Protection of Civilians inArmed Conflict 2010.” March.United Nations. Undated. “Namibia – UNTAG Background.” http://www.un.org/en/peacekeeping/missions/past/untagFT.htm.U.S. Department of Defense. 2005. “DOD, Directive 3000.05: Military Support forStability, Security, Transition, and Reconstruction (SSTR) Operations.” November 28.U.S. Department of Defense. 2009. “Money as a Weapon System-Afghanistan (MAAWS-A).”USFOR-A J8 Publication 1-06. May.Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan 81

U.S. Department of Defense. 2010. The Quadrennial Defense Review Report. February.U.S. Department of Justice. 2011. “Former U.S. Army Staff Sergeant Sentenced to 90Months in Prison for Bribery in Afghanistan Fuel Theft Scheme.” Press release. January.www.justice.gov/opa/pr/2011/January/11-crm-021.html.U.S. Department of the Army. 2006. “Counterinsurgency.” FM 3-24/MCWP.U.S. Department of the Army. 2008. “Stability Operations.” FM 3-07. October.U.S. Department of the Army. 2009. “Tactics in Counterinsurgency.” FM 3-24.2. April.Wheeler, Schaun, and Daniel Stolkowski. 2011. “Development as Counterinsurgency inAfghanistan.” Unpublished manuscript. Draft dated June 17.Wilder, Andrew. 2010. “Aid and Stability in Pakistan: Lessons from the 2005 EarthquakeResponse”. Disasters.Wilton Park. 2010. “Winning ‘Hearts and Minds’ In Afghanistan: Assessing theEffectiveness of Development Aid in COIN Operations.” Report on Wilton ParkConference 1022, held March 11–14, 2010. July 22. http://www.wiltonpark.org.uk/resources/en/pdf/22290903/22291297/wp1022-reportWily, Liz Alden. 2004. “Rural Land Relations in Conflict: A Way Forward.” Briefing Paper,Afghanistan Research and Evaluation Unit. August.82Feinstein International Center

Feinstein International CenterTufts University200 Boston Ave., Suite 4800Medford, MA 02155USAtel: +1 617.627.3423fax: +1 617.627.3428fic.tufts.edu