FOREWORD

Both the landscape and the architecture of humanitarian assistance to refugees have changed dramatically in the last two decades. The proxy wars that flourished within the global struggle of the cold war have given way to internecine, factional fighting within countries. The destruction of civilian lives and livelihoods continues and, with it, large populations of refugees, internally displaced people, and recent returnees who are still dependent on external assistance provided by a bewildering array of public and private international organizations.

Afghanistan has been and continues to be a laboratory for these changes. One of the major theaters of cold war confrontation, it then became the archetypal failed state and still has no functioning central government. It does, however, have the attention of United Nations agencies, bilateral donors, and a plethora of nongovernmental relief and development agencies. Paula Newberg's paper analyzes in thoughtful detail the action and impact of international assistance to Afghanistan. She writes from the perspective of a long-time student of the region and a sometime practitioner of UN reform. Few people know the aid community, or the political context in which it operates, as well.

Dr. Newberg's paper was written as the centerpiece for one of a series of roundtables on the evolution of humanitarian response in the 1990s, convened by the International Migration Policy Program of the Carnegie Endowment. The series was generously supported by a grant from the MacArthur Foundation, and by the program support of the Ford Foundation. We are grateful to them, and to our colleagues who participated in the roundtable discussion and enriched the final paper with their insights.

Kathleen Newland
Co-director
International Migration Policy Program



INTRODUCTION

Ten years ago, Soviet troops withdrew from Afghanistan, signaling a symbolic end to a proxy war that had lasted for almost a decade. It was a moment burdened by mixed messages. The cold war retreated from Afghanistan-and, Afghans would later argue, would end entirely because of that retreat-but with it went the interest and concern of the rest of the world. When foreign soldiers left, the world's largest refugee population lived in neighboring Iran and Pakistan, and only aid agencies and nongovernmental organizations (NGOs) remained to help Afghans begin the process of rebuilding Afghanistan. Sadly, those who predicted that the worst was about to begin turned out to be correct. Ten years later, Afghanistan is still at war, but the world has become an even more complicated place in which to wage it.

This essay examines the environment, actors, and institutional choices that face the international assistance community working in Afghanistan today. The contexts in which it works and in which its collective decisions are taken are exceptionally difficult to define and reconcile, all the more so for the sheer length and accrued political histories of Afghanistan's conflicts. And if Afghanistan's travails present an array of moving targets, so too do the choices that aid organizations have taken during the past decade. UN agencies, international and local NGOs, and donor governments have offered a menu of relief and development alternatives that is often bewildering in its richness and complexity.

Assistance strategies have often been frustratingly reactive and often contradictory. The international assistance community has endeavored to correct some of its own weaknesses while still trying to serve a country and a people whose situation grows more intricate with every passing month. Almost by definition, these reform efforts are incomplete and may, in the end, be defeated by Afghanistan's domestic circumstances. Despite persistent attempts to insulate Afghans and Afghanistan from the worst excesses of fighting, negotiations among fighting factions have been limited in scope, duration, and effect. Moreover, the entanglements of security relationships between and among factions, the United Nations, and NGOs-Byzantine at the best of times-have affected the provision of aid and the content of political discussions among Afghan factions, as well as between them and their neighbors.

This essay argues for continuing institutional reform, led by the United Nations and necessitated not only by the familiar woes of the assistance world, but also by the imperatives of working in a state that has collapsed. It summarizes prior and current UN and NGO efforts, but diverges from them, in the manner of friendly critique, in three ways.

First: reform initiatives, whether based in the field or at headquarters, always assume that existing institutional mandates determine the substance of decisions, and will endure. Reform has therefore meant rearranging the way that existing organizations do their business, and is therefore necessarily oriented more toward process than substance. But existing institutional frameworks-for Afghanistan, for countries experiencing complex political emergencies, perhaps for most crisis countries-can neither deliver necessary aid nor address the critical reform tasks at hand.

Real reform requires relief and development agencies to reshape radically their responsibilities and the command structure within which they work. Failed states like Afghanistan may require the UN Secretary-General to take a far more active role than he has to date, not only in securing peace but also in organizing assistance. Until this happens, the Afghanistan experience suggests that the autonomy of UN agencies, funds and programs ultimately slows reform; by making policy problems harder to solve, this interpretation of autonomy thwarts the very change that agencies purport to support.

Second: overlapping relationships between the United Nations and NGOs, and between assistance actors and donor governments, may defeat prospects for institutional reform, as currently defined. Donor government interpretations of complex political emergencies are often at variance with each other, and their political agendas are often at odds with those for whom a broad concept of humanitarianism is paramount. Reform efforts to date have tried to accommodate all of these sensibilities, and all at once. But the process of inclusion does not lead inevitably to consistent theories, policies, or practices. Political mediation and assistance policies alike will succeed or fail on their substance, not just on the methods by which decisions are taken.

Third: the United Nations now finds itself in the business of relief and social welfare in places like Afghanistan, where its political and assistance roles are very difficult to coordinate, let alone reconcile. No amount of humanitarian assistance will by itself solve the basic problems of a failed state and ongoing military engagement; the provision of aid almost inevitably complicates the relationships between and among donor governments, military allies, and the United Nations. This in itself should not deter those who provide assistance. But in states like Afghanistan, where complex political emergencies determine the structure of assistance in unusual ways, it is necessary to reassert the primacy of politics.

The United Nations-much more than its partners among NGOs and donor governments-must decide whether its most important mission in complex political emergencies is political or philanthropic. This essay argues for the primacy of the political, and for a limited humanitarian agenda that can be followed with integrity and independence. Otherwise, in countries like Afghanistan, its many activities are likely to remain confusing and its ultimate goals elusive.

Some of these conclusions and proposals parallel those of observers to other conflicts, and all inevitably raise more questions than they answer. Perhaps even more than the grueling experiences of the Balkans or Great Lakes regions, Afghanistan's problems-and the difficulties that international aid providers encounter-provoke us to rethink the meaning of politics in and among intergovernmental organizations. Even more, they move us to reexamine the deeper meaning of politics for the citizens of countries living through political crisis.

Washington, D.C.
March 1999



AFGHANISTAN'S COMPLEX POLITICAL EMERGENCY

Few conflicts span the vast political changes of the post-World War II era as do those that have been waged in Afghanistan. Beginning with civil strife in 1973, continuing through a decade of anti-Soviet war, and re-emerging as internal conflict for the last decade-albeit with significant foreign interference-Afghanistan may have helped to end the cold war, but it is now a portrait of a state that the cold war lost. The wages of war have placed enormous burdens on the country's residents and displaced populations: millions of refugees and displaced persons, millions of land mines, millions of malnourished children and adults alike, and, seemingly paradoxically, unchecked population growth in a place almost defined by its poverty. The sheer numbers seem far too large to those distant observers for whom Afghanistan is simply a small country suffering through big troubles.

The story of Afghanistan is more than a simple tale of fighting that forgot to stop. Its cautionary narrative raises many questions about the theory and practice of international assistance and the complex relationship that aid bears to politics. It tell us about the perils of uneven development planning in the 1960s and 1970s that contributed to political backlash in the 1970s and 1980s and 1990s. It reminds us of development disputes that turned into armed battles, and of the dangers that are created when relief is designed simply to accompany political agendas. And its condition today describes the perilous place of peripheral countries during the cold war, whose proxy wars about diplomacy have now been transformed, at least in part, into proxy wars about profit. The story of Afghanistan is a parable about the hazards of outsiders' good intentions in an environment which seems unable to support them: when wars sour and emergency aid continues for too long; when political agendas change but the victims of war remain; when the viability of domestic political discourse is risked by the devastating effects of war on civil society; and when neighboring countries become protagonists in a proxy war while international assistance actors try to plan for a country whose future seems frightfully insecure.

Many of the critical issues that now face international assistance actors in Afghanistan could occur elsewhere, and often do. Afghanistan's political fragmentation and the concomitant difficulties of organizing assistance to its people echo similar problems in Somalia, and its politicized refugee population resembles those in Cambodia and Rwanda. Its environmental devastation calls to mind the ecological nightmares of war-torn societies in central America. And Afghanistan's poverty-as much a cause as a consequence of its wars-places it squarely among those poor countries that hover at the bottom of the UN human development index. Moreover, the problems that long-term assistance so often exhibit-duplication, contradiction, policy inconsistency, and the pervasive problem of institutional mandate creep-are present in goodly number across the Afghan countryside. For all these reasons, Afghanistan offers a case study of the long-term effects of complex political emergencies and the profound difficulties that they present to all its foreign interlocutors.

But Afghanistan shows us more than this. Its complex emergency-profoundly political in its origins and effects-illuminates many difficult questions about the role of international assistance in building or breaking peace, in protecting populations against domestic and external enemies, and in creating a foundation not simple for stability, but also for free political choice. The architecture of international humanitarian assistance has played a significant part in the structure of ongoing conflict in Afghanistan, even if this is often subtle and only occasionally intentional. The governance of international assistance agencies, their decisions about the nature and scope of relief and rehabilitation, and the sometimes precarious and often contradictory relationships between UN member states as relief donors and as political actors provide an unwieldy venue for securing protection or peace. Afghanistan has been an unwitting laboratory for experimentation in international assistance. By virtue of its persistent conflicts, inconvenient location, and the unfortunate indifference of the post-cold-war world community to its current plight, it has become a test case for the viability of integrated assistance planning in complex political emergencies.


AFGHANISTAN'S POLITICAL ECONOMY: THE ASSISTANCE CONTEXT


The Period to 1995

War has been a persistent feature of Afghan politics for more than twenty years. The revolutions and counter-revolutions of 1973 and 1978 left an array of disenchanted proto-politicians who found encouragement in neighboring Pakistan and in a nascent international Islamist movement that grew in resources and organization during the heady days of high oil prices in the 1970s. Some political observers date Afghanistan's conflict in its slightly more distant history: in disputes over the establishment of a central state and the powers and prerogatives of monarchy in an agrarian nation moving slowly toward urbanization, and in the contrary relationships between weak local devolution and the state-building efforts of foreign donors from both sides of the Iron Curtain. The pressures that this historical Asian crossroads confronted as the cold war divided Afghanistan from its closest neighbors in Soviet central Asia repositioned it in an uneasy axis between Iran and Pakistan. Like so many discussions about continuing conflicts in old societies, Afghan commentators date their histories to suit their politics. Regardless of party affiliation or tribal identity, however, almost every Afghan sees the last two decades as a series of challenges to national sovereignty, territorial identity and the concept of Afghan citizenship.

The anti-Soviet war created a broad spectrum of mujahideen-based parties and fighting factions, based primarily in Iran and Pakistan. Their ideologies varied, as did their memberships. Their sponsors-the United States, Pakistan, members of the Organization of the Islamic Conference, and a host of far-flung allies and aid conduits-managed a clever contest of balance and attrition that kept each party competing with the others for weapons and funds, even while they fought against a common enemy. Money, however, was rarely a problem. The total cost of the anti-Soviet war reached hundreds of millions of dollars each year, with the sophistication of weaponry and the network of suppliers rising as the weaknesses of the Soviet military became more evident. By the time the Soviet army withdrew across the Uzbek border in February 1989, the mujahideen were prepared for a long future war.

The Geneva peace accords that facilitated Soviet withdrawal were not designed to secure peace. They created a way for the Soviet army to leave, in part by assuming that the communist government in Kabul would continue to rule-and, presumably, that the mujahideen would oppose that government. As a result, no provision was made for rebuilding Afghanistan, demobilizing fighters from any side, or organizing locally plans for post-conflict relief. At the time, the United Nations was keenly aware that the mujahideen did not include relief wings, as did fighting factions in the Horn of Africa, for example. It therefore helped to organize Afghan groups that would provide assistance in refugee camps in Pakistan and Iran, and continue cross-border assistance that had begun in the early 1980s. These nongovernmental organizations-many of which survive today, with vastly differing levels of skills, effectiveness, funding, and popular outreach-contributed to short-term relief, but more important, became part of an evolving problem about Afghan political voice and presence that continues today.

War continued between 1989 and 1992, and in 1992, a western-backed plan required the Afghan head of state, communist leader Najibullah Ahmedzai, to step down in favor of a rotating cast of mujahideen leaders whose idiosyncratic consociationalism belied both democratic idealism and administrative realism. The next years witnessed the erosion of civil security, social cohesion, economic stability, and political opportunity across Afghanistan. Fighting factions led by a changing group of commanders pillaged the countryside and destroyed the capital (which had remained relatively secure during the decade of the 1980s). Government in Kabul functioned in name alone, with each faction committed only to maintaining its title without taking responsibility for running the state.

Throughout this period, as before, international assistance was concentrated on the millions of refugees based in Iran and Pakistan. To facilitate the delivery of cross-border assistance, agencies worked on the basis of expediency rather than long-term planning. This meant that relief was provided with the help of anyone who could help-tribal leaders, local shura, military commanders-to serve the immediate needs of populations under siege. The disjunction between hope and reality was made clear in 1992, when the mujahideen government took office. The modicum of rehabilitation assistance that had been planned by the UN Development Program( UNDP) -designed on the basis of a thorough evaluation of the decayed infrastructure of war-torn Afghanistan-was effectively canceled when the United Nations relocated to Pakistan. Diminishing security scotched these prospects for rebuilding, establishing a cycle of dashed assistance wishes, unrealized desires for international legitimation by office-holders, and a captive population under the sway of fighting factions. These patterns of political and social interaction-made visible by the Taliban movement when it later gained control of Kandahar, Herat and Kabul-were in place long before the Taliban arrived in Kabul, and the Northern Alliance tried to consolidate its fragile command.

Two other trends emerged during this period. First, the geography and demography of Afghanistan became more fragmented. Many observers have suggested that Afghanistan's tribal and ethnic mix led to its previously weak federalism, and some believe that disintegration is inevitable. This is not usually an opinion offered by Afghans. But the effects of political and economic fragmentation-combined with patterns of refugee displacement that rely on familial, tribal, and village support systems-did seem to reify a projected ethnic division of the country. For the most part, however, fragmentation was the result of military movement and political outlook. The rule of commanders was almost necessarily uneven, with fiefdoms ranging from a few city blocks to entire provinces. With transport slowed by degraded infrastructure and undetected land mines, each locality assumed an autonomy that was difficult to maintain but even harder to contest. This kind of division-physical at the outset, increasingly political over time-became one to which outsiders tacitly acceded, since the distribution of goods and services required the good graces of commanders and local leaders. It is from this period that some, if limited, popular opinion arose, suggesting that the United Nations, for example, was complicit not only in commander rule but also in the future breakup of Afghanistan.

Second, the difficult distinction between war and peace became even harder to discern. Large portions of the country, while insecure, were nonetheless free from direct military engagement after 1992, and military stalemate became more common in these areas. This relative quietude led aid agencies to view Afghanistan as a post-conflict arena rather than one seized by war, and to concentrate their concerns on refugee return and the first stages of rehabilitation. The initiative, however, was political. The immediate impetus was Pakistan's desire to have 3.5 million refugees return home (Iran harbored similar hopes for the 1.5 million refugees living there); the real push may well have been to support the 1992 post-communist accord that, however shakily, had been fostered by the United States, Pakistan, and like-minded states. At no time were humanitarian decisions taken without reference to a political climate that was, for the most part, engineered (or imagined) by external powers.

UN political intercession was therefore conducted with less hope than resignation. For many Afghans, the express purposes of future negotiations-establishing a cease-fire that could lead to representative government-were illusory at best: few fighting factions seemed interested in pursuing non-military objectives. Even as the factions themselves split and weakened, they directed their attentions to military achievement rather than popular legitimation. As a result, negotiators-tireless, but often seen as quixotic by Afghan civilians-engaged in endless entreaty with gun-toting commanders. For the most part, they devoted little if any attention to the civilian population or to the popular prerequisites for peace. When UNDP proposed in 1996 to devote some of its program resources to peace-building and governance, there was little evidence that the United Nation's political negotiators had given this subject much thought. Not surprisingly, there was little, if any, discussion about what would happen within Afghanistan should a cease-fire actually occur.

As a result, the political economy was influenced by two factors, and both were outside the control of Afghans. On the one hand, international activity was based on what Afghans perceived to be foreign interference, in which international aid was simply the handmaiden for political connivance. For many Afghans-including, it later transpired, disaffected mujahideen who formed the vanguard of the Taliban movement-the vaunted neutrality and impartiality of the United Nations was sullied by its expedient pacts with local commanders. On the other hand, there were few resources in Afghanistan available for relief, rehabilitation or redevelopment. Production rates plummeted, and transit trade-which had continued through the 1980s with minimal impediment-declined. A generation-long brain drain had exhausted personnel resources, and the indirect effects of war were startlingly clear. Rampant poverty, displacement, infrastructure damage, and the devastating consequences of land mine penetration had left large portions of the local population without sanitation, potable water, basic medical care or, in some instances, food. Mortality and morbidity figures placed Afghanistan among the world's most vulnerable countries, preventable diseases were recurring at rapid rates, and internal displacement was creating one of the world's fastest rates of urbanization.

Economic fragmentation took a political toll. With international assistance primarily devoted to the immediate, human dimensions of relief-inoculations and wheat rather than roads and communications-the structural prerequisites for developing and sustaining an Afghan economy were limited indeed. Although the debate about the relationship between relief and development was effectively entered in 1991, when UNDP began its survey of the war-damaged economy, there was little to suggest that relief providers expected to be able to do much more than provide basic relief. Those in the development business either participated in relief or went out of business.

The period leading up to the entry of the Taliban as a significant actor offered ample evidence of institutional confusion among assistance actors. Humanitarian activities were colored and controlled by political interests during the anti-Soviet war, and this relationship persisted after the communist government fell in 1992. The independence, neutrality, and impartiality of UN humanitarian bodies seemed at odds with the express political nature of some of its partners, but seemed insincere when compared to some NGOs that were servicing both refugee and cross-border populations. Additionally, the UN political mission seemed to occupy a world that assistance agencies did not inhabit. In short, formal foreign policies defined the assistance environment and informal politics colored the provision of aid, although a political vocabulary was missing from the language of aid providers. Most important, however, politics was absent from Afghanistan.

Contemporary Afghanistan


With the advent of the Taliban movement and its swift entry into the Afghan military terrain, most of these trends were accentuated. Nevertheless, Taliban policies underscore the difficult calculations that Afghan civil society has faced and will continue to confront in coming months and years. By extension, these same factors continue to affect the scope and depth of international assistance, the ways that the international assistance community organizes itself, and the contexts within which it chooses to define its work.

Retrospective conventional wisdom often describes the entry of the Taliban movement into Afghanistan as almost inevitable, suggesting that a new kind of military force was required to repair the chaos and unrest within Afghanistan. At its inception as an organized movement, however, the Taliban was described as a lackey of Pakistan and the United States. Together, these two points of view offer a partial description of Afghanistan in late 1994 far more than they do of the Taliban, whose genesis predates the anarchic 1990s. Indeed, the Taliban-with its cousin parties, the Harkat-ul-Ansar (which fights in Kashmir) and the Sipah-i-Sahaba Pakistan (a militant Sunni party within Pakistan)-is an offshoot of the Pakistan-based Jamiat-i-Ulema-Islami, and is an almost natural outgrowth of religious schools based in Pakistan that catered to young refugee men who, at earlier points, formed the backbone of the mujahideen. The transnational character of the Taliban movement-ranging from the sources of its funding to the recruitment of fighters-was integral to its founding, its successes, and its weaknesses. Its origins within Pakistan, and the seeming relief with which many Afghans greeted their initial exploits in Afghanistan, are important elements of its status and behavior today.

The Taliban moved toward Kandahar in late 1994 and early 1995, when fractiousness had reached its peak in Afghanistan. Between early 1995 and September 1996, it was able to gain control of most major cities. In each region, commanders either fled into exile or joined the Taliban. The movement therefore built its network of control from the detritus of disaffected mujahideen factions as well as from its own ideological adherents. In most of the areas, the Taliban has secured a kind of repressive control over cities, relied on a looser command structure in the rural areas, and devoted its efforts to extending its military control and determining an ideological path for future rule. By the end of 1998, the Taliban had secured as much as 90 percent of Afghan territory by the deft management of manpower, including growing foreign recruitment, and materiel. Like its opponents, the Taliban movement relies on external support for arms and fuel supplies: fighting patterns often reflect the capacities of each faction to muster resources from (or through) neighboring countries and the availability of air and land access to border supply routes. The sharply contested front line traverses the Hindu Kush, dividing northern areas from the south, isolating transport routes, and endangering the availability of international assistance as well as war supplies.

Although the Taliban movement was originally thought to be exclusively Afghan and entirely Pukhtun, the accession of local commanders changed the complexion of the Taliban forces, if not its leadership, and the recruitment of foreign fighters among young men across the Muslim world has transformed its membership as well. With time, the Taliban has acquired the liabilities as well as the assets of its new recruits. This has meant that ideological consistency, once assumed, now sometimes appears imposed, as rigor enforced through a regime of edicts substitutes for full-scale governance. It also means that both popular support and underground resistance-both difficult to measure-complicate the political terrain more than the Taliban might hope. Recruitment for fighting is increasingly difficult, with villagers reportedly resisting forced conscription and often finding ways to elude, rather than confront, unpopular edicts. Although the maintenance of local order seems universally appreciated, for personal safety as well as economic opportunity, the ruling shura's puritanical edicts suggest that the rigid, conservative ideological interests of the Taliban's founders and fighters take precedence over the more practical (and therefore flexible) concerns of the civilian population. These impressions-garnered as much from Taliban statements and actions as from those of their opponents-have led observers to surmise that the governance vacuum left by former faction commanders may lead to longer-term crises of governance.

The availability of military support, the desire for victory, the tenacity of ideological contest, and the lure of future rule and profit have all led to continued military engagement. The Taliban believes that victory is possible, and its opponents believe that continued fighting will secure their place at a future bargaining table. Although the primary contest is therefore a political one, almost no one has, until now, felt the need to suspend military activity in favor of serious political negotiations. Occasional parleys between the Taliban and its opponents-including a heralded meeting in Turkmenistan in March 1999-have flirted with the idea of peace, but have generally succumbed to the habits of war. As a result, UN political negotiations have concentrated on persuading Afghanistan's neighbors-and primary arms suppliers-to withhold support for these factions. Changing political sensibilities in the neighborhood-the recent security split between Uzbekistan and Russia, Turkmenistan's continuing faith in prospects for a trans-Afghan gas pipeline, tentative political change in Iran, and economic crisis in Pakistan-may ultimately help the political to triumph over the military. The nature of the political compact that may result, however, has been discussed in only vague terms.

For years, the civilian population has been held hostage to the direct and indirect consequences of war. Political negotiators have been drawn into a handful of important humanitarian problems-securing access to isolated and vulnerable populations, securing agreement from the Taliban to respect the privileges and immunities of UN aid agencies-as a coda to their primary responsibilities. These activities have brought the political and assistance wings of the United Nations together more closely than have other, more formally choreographed efforts in the past, but they have highlighted misgivings on several sides.

Those doubts are well founded. The interplay of international politics and domestic confusion that the Taliban has both encountered and instigated is one that the international assistance community does not fully understand. The Taliban has become well known for the zeal with which it pursues its ideological goals and the cloak of religion that covers its political strategies. Its human rights record has become its public signature, even though its predecessors also abused rights with impunity. The edicts that its governing shura has issued since 1995 range from the personal to the political-regulating the length of beards, female attire and access to public services and public life, and the availability of news and entertainment. The Taliban has tried to control all aspects of Afghan life in an effort to impose an order that its leaders find comfortable and believe to be ideologically consistent. In turn, it has become a lightening rod for issues and policies that the world hitherto ignored.

These actions, and the general isolationist nature of Taliban behavior, underscore two different problems for Afghan citizens and international aid workers. First, Taliban policies affect the provision of assistance. If and when females are excluded from relief, if and when access to populations under siege is restricted, the aid enterprise is defeated. Only if aid is distributed according to a strictly numerical, utilitarian calculus can agencies justify providing assistance only to men and boys. If the whole of the society is taken into account-with the knowledge that large portions of relief are useless unless provided with women and families as conduits-then discrimination against women delivers a deadly blow to assistance. To date, the Taliban have not only restricted access to Afghan communities, but its leaders have also interfered with relief agencies themselves, violating long-standing agreements to respect the privileges and immunities of the United Nations and its partner organizations. This behavior began when the Taliban first reached Kandahar, and has increased in frequency, breadth and depth during the past three years. In response, large portions of international assistance has been withdrawn or limited, provoking endless rounds of discussions among UN, NGO, and Taliban officials that, to date, have still not produced agreement on the basic terms of engagement for assistance. As a byproduct or this contentious atmosphere, aid groups have not been able to spend available funds or to establish a stable source of funds until these disagreements are settled.

Second, human rights abuses under the Taliban have finally focused some attention on pervasive violations within Afghanistan and, at the same time, have underscored the inconsistencies and inadequacies of international response to those violations. Although the world now pays heed to gender discrimination, its attentions have been superficial and fleeting, addressing the symptoms of gender bias more than the basic absence of free choice within Afghanistan. It is an axiom of international assistance that the spheres of humanitarian action and human rights scrutiny and response are distinct. Many assistance providers argue that this separation is absolutely necessary to maintain impartiality and the perception of neutrality. Others, however, suggest that this division is artificial and that so-called humanitarian imperatives-providing assistance, with and through de facto authorities, regardless of their repressive behavior-lead to rights violations, allow rights violations, and insulate rights violators from direct sanction. In a country in which most aid was overtly political for many years, it is particularly difficult to establish a new a posture of impartiality and neutrality that many Afghans-including the Taliban leadership-already believe to be myth.

THE POLITICAL ECONOMY OF ASSISTANCE


As an economic entity, Afghanistan is currently in tatters, despite discrete aid interventions designed to provide a foundation for reconstruction. Some priority areas have been cleared of land mines, and demining-perhaps the most effective nongovernmental activity sponsored by the United Nations and the nongovernmental community-promises future success, provided deminers are able to retain control of the conditions under which they work. This means that primary transport routes and agricultural areas are, or can be made to be, accessible for economic activity. However, the effects of disease and malnutrition are severe, and are compounded by lagging-and with the restrictions on female participation in public life, often closed-access to education. Together, these indicators anticipate an under-served economy, untrained labor force, segregated labor market, and under-developed resources. The open border between Pakistan and Afghanistan, which creates a simple safety valve for Afghans and virtual Afghan suburbs for Pakistani cities, means that Afghanistan could well drag into the next century with few developed domestic resources.

These indicators are minimally offset by two counterpoised trends that do not augur well for a fully-grounded, legitimate economy. First, transit trade has resumed with some robustness, but it is a derivative economic activity that requires little but roadways to provide a source of customs revenue to presumptive authorities. The state need not invest in order to derive profit. Second, narcotics cultivation, production, and trade offer a route to relatively easy money that is, by definition, a cornerstone of a black market economy. Not only does this make Afghanistan something of a pariah in the international economy, but it threatens borders and security among neighboring states whose governments have yet to control the narcotics traffic that comes their way. For minimal investment, narcotics producers have stepped into the vacuum created by a non-existent banking sector-no financial services operate just now in Afghanistan-and provide agricultural credit for small farmers who are otherwise not served by local markets or the aid community. This fact, more than the assumed use of narcotics profits for military purposes, has led various leaders to proclaim the eradication of a poppy crop that, for the moment, they have no interest in controlling. It is easier to let things lie than stir up trouble, particularly when constituency support services have yet to insinuate themselves onto the Taliban agenda.

Even in strictly economic sectors, the relations between citizens and those in power are primarily punitive. The Taliban rarely responds to citizen needs and when it does, it is within the sphere of edicts and punishments. It can halt impulsive price hikes, for example, by threatening to punish tradesmen, but has yet to plan to restore production. Similarly, it can control security on trade routes but has done little to rehabilitate communities or provide village-to-market access. It maintains shariah courts, but uses them primarily to restore properties expropriated by the communist government, and to enforce recent edicts. The Taliban does not collect taxes as such, and uses its limited revenues for military rather than civilian purposes. In effect as well as in formal status, the Taliban has yet to take on the attributes of government.

Cumulatively, this means that Afghanistan is terribly dependent on its neighborhood for economic survival. Afghanistan's economy is subject to the parries and thrusts of neighboring economies. A 1997 United Nations/World Bank survey showed what every Afghan already knew: economic ties among Afghanistan's regions are far weaker than those between each region and its closest cross-border neighbor. This is a symptom of vulnerability: border closings can spell disaster, and Afghanistan's small, open economy is exposed to changes in the currency supply, devaluation, trade surpluses, and domestic disputes of its neighbors. Psychologically, this is a continuing blow to Afghanistan's sense of its sovereignty; politically, it fosters a form of indirect dependency that humanitarian assistance cannot begin to repair. These factors contribute to a sense on the part of many Afghans that dependency, however indirect, is a close cousin to foreign interference.

But international assistance forms the basis for-and some would argue, the only source of funds for-relief and rehabilitation. As a conduit for resource transfers, international aid stands alone in its efforts to fulfill basic needs. International assistance has nonetheless been accompanied by many conditions, most of which Afghan citizens have not helped to formulate or control, and the implementation of those conditions has been contradictory, haphazard and sometimes (unintentionally) disrespectful to Afghan society. The singularity of this endeavor is troubling, however, as it is in other crisis countries. There have been few formal local relief initiatives, which the international community interprets as a lack of ownership, broadly conceived, in the project of rehabilitation. When the small private sector does engage in rehabilitation, it often does not dovetail with international aid. Even so, mendicants and merchants cannot by themselves sustain a national economy.

It is the failure of the Afghan state, however, that provides the most important context within which assistance is designed and distributed-and it is the same failure that has created the most troubling context for aid recipients as well.

GOVERNANCE, STATE FAILURE, AND ASSISTANCE

War affects governance in a multitude of ways. It defines the present: territorial control; ownership, accessibility, and use of resources; and relationships between ideology and political practice. And it defines the future: the kinds of citizen-state interactions that are possible, and the nature and prospects for civil society. This is true in any war-torn society, and is particularly so in Afghanistan.

When asked to list those aspects of Afghanistan that most affected the provision of assistance in 1996, the heads of most UN agencies and major NGOs replied that the "failed state" was their biggest problem. By this they meant not only that government relief funds were unavailable, but also that there were few if any government interlocutors who could set the stage for relief and sustain rehabilitation. For international aid organizations, state failure means the absence of government. It is for this reason that they, like their diplomatic counterparts, initially greeted the Taliban with more warmth than skepticism. The potential success of one fighting faction meant the prospect of a single government voice and the possible resumption of orderly, if poverty-stricken, government.

But for Afghanistan and Afghans, including fighting factions, state failure means much more, for it is as a political entity that Afghanistan has sustained the most serious blows in recent years. Citizen-state relations have diminished, surely, but so, too, have the citizen-citizen relationships that are the foundation of communities and of the state. When observers speak of the diminishing resilience of civil society, in large measure they are talking about the absence of political society and all the attributes that accompany the concept. Without a functioning state, these have been terribly weakened. Any faction or alliance that ultimately assumes control of the capital and of government will therefore be faced with little apparent local legitimacy, sporadic validation from outsiders who must deal with someone (if not anyone), and minimal outreach in the country as a whole. The effects of this political vacuum are evident in perfunctory political negotiations and, by extension, the inconclusive discussions that typify faction relationships with the assistance sector.

What remains are the towns and villages that dot the Afghan landscape, where many assistance actors now prefer to work. Commentators are quick to note that this is the backbone of Afghanistan and of the local, tribal democracy that it has sporadically fostered. Aid workers, who are understandably quick to look for ways and places where aid might work, have increasingly shunned major cities, where Taliban patrols are more vehement. They have tried to move their activities to the rural areas, with the hope that local, informal governance-generally a combination of ulema, tribal elders, and Taliban overseers-might triumph over imposed ideology and provide a more welcoming, unfettered environment for relief.

The consequences of this rather practical choice exhibit the many problems that expedience creates for reconstruction and development. First, it tends to convince aid actors that the mission of assistance-to help the needy-means that short-term output trumps long-term impact. Assistance planning is sidelined in favor of a catch-as-catch-can mentality-a practice that places not only assistance but also beneficiary populations in jeopardy. This is-and should be-a contestable proposition, anywhere and anytime. But one of the most important decisions to be made in failed states is whether and when to stop assistance, or to change its provision radically-knowing that the decision to end aid, or the consequences of continuing aid, may both compromise long-term governance. Under such circumstances, a disposition toward short-term fixes tends to skew these calculations: what are the effects of ignoring urban areas on the recovery of the country? Who should decide how urban and rural areas interact in an aid regime? In Afghanistan, decisions about stopping or reducing aid have been taken half-heartedly, incompletely, and without adequate analysis or preparation. Political emergencies deserve far more political analysis than they have thus far received.

Second, such a choice appears to emphasize the flexibility of aid. In fact, it allows assistance priorities to be determined by those who presume to rule rather than those who live under imposed power. This offers the appearance of evading, rather than confronting, the ideological precepts that, for better or for worse, identify the ruling faction and distinguish the use of its power. Moreover, it further disenfranchises a population from fundamental decisions about its future. In this sense, it is a choice fraught with contradictions.

Last, it reduces all aid to the category of emergency aid-do what you can, when you can. Doing so, however unintentionally, ignores the critical differences between assisting the victims of natural disasters and helping those caught among the competing interests that complex political emergency has left at Afghan doorsteps.

The difference between complex political emergencies and other disasters that befall states and their peoples is not simply the pervasiveness of the emergency, but its effects on the future of politics, sovereignty and governance. In the absence of workable government institutions, and particularly in the absence of an understood framework for formal political choice, informal politics defines the Afghan assistance environment. Bargains are struck, knowingly and often unhappily, with unofficial counterparts who may or may not be able to facilitate the provision of aid. This can leave the delivery of assistance out of the control of its providers or its beneficiaries.

In the absence of a political framework, three models of interaction have existed side by side. For those who believe in the primacy of the aid enterprise, any interlocutor can be a worthy counterpart; they have dealt with commanders, ulema, tribal leaders, and corrupt middlemen without distinction so that aid can be available for needy populations. This practical view has typified aid to Afghanistan for many years, and although the Taliban professes to disdain it-particularly where its rule is challenged-it has found it difficult to eschew such practices.

Others profess a self-described principled view that proscribes dealing with any presumptive authorities, and particularly those who impose conditions on the delivery of assistance that challenge the mandates of aid providers. Still other practitioners, convinced that all assistance requires a counterpart, seek the closest approximation to a government institution in a state that has none. Parsing language to suit circumstance, they distinguish between political actors (mayors) and technicians (engineers) in order to deal with the remnants of government or pretenders to power to facilitate aid delivery.

Differences between them notwithstanding, the practical and the principled approaches have met with opprobrium from Taliban officials. The third approach has been adopted by many UN agencies and NGOs because it seems to split the difference. But by confusing practice and policy with principle, it has created more problems than it has solved.

Of course, all assistance, in every country, affects governance. But the complex relationships among governance, local ownership, and capacity building in a failed state are close, inevitably ticklish, and often baffling for Afghan citizens. They also set the context for policy and institutional decisions that political and assistance actors confront on a daily basis.

POLICY CONTEXT FOR ASSISTANCE REFORM

History accumulates in many ways. The institutional context of aid to Afghanistan draws its sources equally from political and humanitarian relationships that have developed over time and in many places, and from the ways that organizations have coped with a wide range of exigencies. These include: the wary ways that political and humanitarian decision-making interact (or do not); the conundrum of coordination and the weaknesses of institutional hierarchies and chains of command; the remarkable tenacity of indirection, acquiescence, and proclivities toward the familiar in policy making; a marked chasm between the rhetoric of local ownership and financial practices that require that fiscal control remain outside affected communities; and an irresistible penchant for conducting business in crisis countries as if they were peaceful. All these tendencies denote international organizations, much as custom and ritual demarcate competing ancient societies; all are well burnished by earnest intentions, good will, and occasional self-imposed blindness toward the political consequences of humanitarian choices.

To assist Afghans and Afghanistan means seeking to alleviate suffering while providing an environment that encourages future peace. Both goals require aid actors to juggle within several competing environments, all at once.

The Political-Military Environment


The political-military environment poses a complex mix of difficulties, including: (a) the primacy of military prerogatives, which leaves assistance agencies in almost as much flux as civilian populations each time fighting resumes or new civilian populations are held under siege; (b) incomplete political consolidation on the part of each fighting faction, and the concomitant absence of civilian-based efforts at national governance, reconciliation, or peace-building; (c) the challenge to maintain Afghanistan's sovereignty in an environment of physical fragmentation; (d) the challenge to maintain the sovereignty of Afghan citizenship during a profound crisis of governance; (e) the absence of means to arbitrate disputes that develop over the use of international aid resources, particularly under conditions of contested command;
(f) pervasive human rights problems in all regions, which are both a cause and consequence of the governance crisis; and (g) a local and global problem of narcotics cultivation, production, trade, and use that challenges governance across south, west, and central Asia.

The problems that make daily news headlines-the sequestration of women, sectarian strain, and the specific content of ideological division-are Afghanistan-specific. The absence of political consolidation, the fragmentation of governance and resulting social disorder and human rights abuses are all characteristics of long conflicts, often in poor countries. In Afghanistan, generic difficulties are too often put aside precisely because the singularity of its ideological condition so easily influences daily decisions. One basic challenge for institutional reform is to be able to understand the particular and the general, together.

The Humanitarian Context

Afghanistan's humanitarian context provides a background of tangled, competing priorities. Its degraded human and physical environments provide a backdrop for the difficult tasks of building capacity. These are complicated, in turn, by confusions about the relationships between relief and development in a country that requires both. Crises of governance and rights are compounded by relationships with local leaders and presumptive authorities, and accompany an ever-present debate-among Afghans as well as aid workers and political intermediaries-about the relationship between human rights and humanitarian imperatives.

Afghanistan is, at once, a country engaged in war, living with the direct effects of long war, and recovering from war. An April 1996 UN/DHA mission reported that Afghanistan was no longer experiencing a humanitarian crisis, but by the time this assessment was published in December 1996, the situation seemed already to have changed for the worse. Epidemiological crises join physical insecurity to create a climate of profound uncertainty. Active fighting and implied collaboration of civilian and military populations in some areas underscore the uncertainties of the humanitarian environment. Prolonged political emergencies exacerbate the provision of relief in natural disaster: Afghanistan's frequent droughts, floods, and earthquakes would challenge the relief community under normal circumstances, but in the context of a failed state, do so all the more.

The sheer length of the war means that assistance actors constantly question the role that humanitarian aid plays in perpetuating or relieving conflict. Because no local funds are used for relief or rehabilitation, do international resources offering faction leaders a way to ignore their people? Are aid agencies interceding in the relationship between factions and their would-be constituents? Is aid, indeed, helping to keep politics out of Afghanistan? These questions are raised in virtually all war-riven states. In Afghanistan, they are harder to answer because there is no one to speak for the population, and no mechanisms for taking collective decisions.

This vacuum has led to intricate policy decisions that are often taken with inadequate knowledge. Do explicit policies-most important, a decision taken by donors and agencies to encourage repatriation, which has resulted in the return of some four million Afghans in the past four years-burden the country more than it, or donor agencies, can support? Even more, to what ends should aid strive? Under the circumstances that prevail in Afghanistan, should aid strive to "do no harm," or does this simply save profound problems for a later date? Is it possible to take difficult decisions about the mix of aid resources and programs and still "do no harm"?

Relief workers and development actors calculate their responses to these questions quite differently. The logic of development is long-term, even when development decisions require short-term pain for longer term gain. This is why development is usually undertaken by governments-who speak for their people-rather than solely by external agents. The logic of emergency assistance, however, generally precludes such choices: aid counters emergencies, long-term gain almost always takes second seat to alleviating short-term pain. For affected populations, these choices are extremely important, and in a failed state-with no way to gauge public opinion, and no one to take responsibility for hard choices-the responsibilities of aid workers become far broader than their job descriptions might initially suggest.

Many aid workers-whether engaged in relief or development-question whether these issues can be resolved empirically. How accurately can we define the contours of humanitarian emergency in prolonged war, and what distinguishes it from a crisis of development? Donor governments-however keen they may be to promote peace-building-offer most of their Afghanistan funds from emergency budgets, as they do in most crisis countries. Does an institutional change from emergency relief to development assistance-counting beans differently, or perhaps counting different beans-signal a significant change in Afghanistan itself? How much are our discussions about Afghanistan, and how much are they about organizational divisions that emerge from the corridors of distant bureaucracies? For many Afghans, the crisis of a failed state is precisely a crisis of development, and the technical distinctions devised by aid agencies mean far less than the brutal fact that a non-functioning state cannot spawn, encourage, or accommodate development.

The distinctive concatenation of emergency and rehabilitation assistance applied to crisis countries is hard to plan, almost impossible to coordinate, and perversely difficult to execute. Even so, institutional divisions barely reflect ground realities. Almost all aid to Afghanistan is rehabilitative, and while it is axiomatic that all aid should include elements of development assistance-particularly the manner in which it is organized and the way that local communities participate in its design and implementation-practice rarely makes perfect. But even when governance crises require choices to be made-what to stop, what to limit, what not to do-almost all assistance providers demand to be seen to be providing emergency relief. As a consequence, assistance is offered in whatever form it comes. Such are the perils of good intentions.

Relief and Development

To the degree that there is a debate about the differences between relief and development in Afghanistan, it has crystallized around the relationship between governance and capacity building. Development agencies traditionally build capacity to enhance the capabilities of government, and consider this to be a defining element of development aid itself. Relief agencies, on the other hand, build capacity to facilitate the distribution of emergency aid, for which helping existing institutions is a useful by-product. Confusion about the difference-who is the target, who is the by-product-pervades both endeavors in crisis countries. In Afghanistan, the political environment makes these knotty choices all the more difficult. Efforts that are normally complementary-providing food and developing food security (the respective mandates of the World Food Program and the Food and Agriculture Organization), providing direct medical aid and supporting medical training (the respective mandates of the International Committee of the Red Cross, and the World Health Organization)-can conflict when they are distinguished by the relationship that aid bears to de facto officials. When aid directly benefits presumptive authorities, its effects on governance are serious and enduring.

When everything is needed, to choose between these alternatives seems to accept a devil's bargain. Both forms of engagement sometimes confuse means and ends: aid workers are prone to interpret their missions solely as building bridges or digging tube wells, rather than helping communities to choose their own pace and form of rehabilitation. And yet, the differences in impact, whether in the near term or at a later stage, are critical.

In Afghanistan, government is either a relic of a distant past, or an idea cushioned in dreams of future recovery. Most professionals left the country long ago, and they are generally not counted among returnee populations. To build capacity by assisting "government," as defined anywhere in Afghanistan, is therefore an overtly political act that most (though not all) aid actors have discouraged. However, many of their actions are seen as helping to entrench power holders whose rule has yet to be validated by Afghans themselves.

Those who have nonetheless tried to build capacity, generally at the local level, have found that a failed state does not mean a rule-free environment. Faction fighters, commanders, traditional leaders, and Taliban appointees all want a piece of the assistance pie to solidify their political and social status and to expand their discretionary funds. "Feeding the commanders," a theme of an earlier period, can easily be transmuted into "feeding the factions." Aid workers and beneficiary communities generally agree that while the effect of this habit is hard to measure, it is usually undesirable.

It is clear, however, that traditional development work is difficult to conceive or deliver in Afghanistan. With the participation of local NGOs and other UN agencies, the UN Development Program recast its programs to include district development plans, the formation of rehabilitation shura, and, starting in 1997, an umbrella program to eradicate poverty and empower communities (P.E.A.C.E.). These efforts tacitly acknowledged a conundrum in contemporary Afghan politics. They attempted to bypass the worst obstacles through a form of shadow development that creates alternate venues for local decisions, attempts to empower local leaders and their communities, and provides the first building blocks for post-war Afghanistan.

Taliban authorities quickly realized that rehabilitation shura offer funds that are too attractive to resist-leaving shura members at risk if they do not comply with Taliban directives, and risking as well the integrity and effectiveness of the development enterprise. In part for this reason, the P.E.A.C.E. program places only limited authority within communities, often reinforces traditional leaders, and does not repose final responsibility for the disposition of funds in community hands. These practices may not define good development, and are certainly incomplete and potentially faulty. However, these same characteristics may have saved a number of individuals from dangerous conflicts with de facto authorities. The political limits on capacity building at the outset directly constrain the developmental reach of assistance in each community.

Under these circumstances, building capacity-and the concept of sustainability on which it rests-suffers from sharply reduced expectations. The burdens on aid actors and their local partners have grown in almost direct proportion to the interference of presumptive authorities, and the costs of working in a volatile environment increase with each crisis. Experience in Afghanistan may well call into question the viability and propriety of working toward even minimal development goals in a failed state.

Governance and Capacity Building

Local governance crises have been complicated by the national aspirations of presumptive authorities and the seeming impossibility of maintaining national assistance programs under conditions of profound fragmentation. Local programming seems to presage-however unintentionally-the physical division of the country, and the contradictory practices of presumptive authorities undercut the sustainability of relief and development efforts. History plays a part in this: during the anti-Soviet war, aid agencies became accustomed to dealing with virtually anyone who could facilitate the delivery of assistance, and breaking the practice seems to some to symbolize a broken bond with the Afghan people. Aid actors therefore work hard to fit square pegs into round holes, having to confront long-term abnormality as if it were a version of normality.

The systematic absence of cooperation by presumptive authorities-de facto official counterparts, as some agencies call them-jeopardizes even this effort. The necessary interactions between assistance organizations and de facto officials offer an unintended veneer of validity to presumptive authorities whose rule has not been validated by Afghan citizens in any way. Moreover, the minimum security required to work in conflict areas has been purchased in Afghanistan at the price of community participation-which itself jeopardizes not only capacity building but human rights guarantees.

Perhaps most strikingly, otherwise reasonable disagreements about the extent of engagement with presumptive authorities have had high costs. Because each NGO and agency often acts on its own view of rectitude, divisiveness has risked the integrity of the assistance endeavor, and has placed at risk those individuals and agencies who do not pay adequate obeisance to those who hold power. Disputes within the assistance community about just these issues led to confrontations with presumptive authorities in Kandahar in the spring of 1998, leading the United Nations to withdraw international staff from the region and suspend UN programs for a period of time. Similar problems led to bitter recriminations between the UN and nongovernmental organizations, among NGOs, and with the Taliban leadership in Kabul in the summer of 1998.

Two important prerequisites for assistance planning are critical to the success of even limited humanitarian interventions in Afghanistan, as they would be in any other country: a willingness on the part of aid organizations to seek similar goals and agree on the means to achieve them, and a similar commitment to find ways for Afghans to participate in seminal aid decisions. The absence of a working, legitimate government can be an invitation for imaginative programming and planning, but only if the assistance community agrees on the parameters of its actions in synchrony with the political policies that, however distantly, guide their presence in Afghanistan. In particular, the resolutions of the UN Security Council and General Assembly-which seem alternately to offer the prospect of a seat in the General Assembly as a carrot and a stick, with little discussion of popular validation of imposed rule-have complicated the operations of UN aid agencies.

The Problem of Principles

The most contentious problem for assistance actors is one that is most difficult to solve: the relationship between human rights and humanitarian principles. The regimen of edicts promulgated by the Taliban has focused international attention on the swift effects of human rights violations on the provision of humanitarian assistance. The spotlight has focused on limitations on women's participation in public life, and the corollary effects of gender discrimination on family life. But the wide range of human rights violations in Afghanistan has the effect of limiting free choice for everyone, and limiting the possibilities for open political debate so that free choice can result. Cumulatively, they test the resolve of the international community to give substance to United Nations founding principles. Although rights advocates aver that ignoring violations in order to provide aid may serve short-term purposes but jeopardizes prospects for renewed political life, many aid workers believe that ignoring basic humanitarian needs risks the future well-being of the citizenry. Afghans are caught in the middle of this debate.

These dilemmas were brought to the fore by aid workers, and echoed by human rights commentators precisely because neither humanitarian principles nor human rights instruments helped resolve the problems raised by Taliban restrictions on Afghan-and later, all Muslim-women. A principle-centered approach to gender and capacity building was proposed by the UN Emergency Committee on Humanitarian Affairs after a two-year debate about how to respond to restrictions on the activities of women and girls. The guidelines were to coordinate aid activities that otherwise ranged from closing specific programs (as did UNICEF and Save the Children U.K. in Herat in 1996) to redesigning activities to take women's rights into account, to ignoring violations in an effort to continue basic programs. The principle-centered approach was contrasted with an assistance-based approach that kept rights issues to the side so that the country as a whole would not suffer further from the depredations forced by de facto officials.

Both approaches assumed that humanitarianism dictates the undisturbed continuation of aid, and that a human rights perspective necessarily requires the imposition of sanctions or conditions on the delivery of aid. Neither assumption is correct: the creative pursuit of humanitarianism can, and should, incorporate human rights protections into assistance; and rights protection does not require the imposition of punitive conditions in order to work. Elasticity can be both an invitation and a prerequisite for success.

Nevertheless, the assistance problem is a human rights problem. Some aid officials and some Afghans prefer to treat any guidelines as aspirations rather than instructions; others view them as the beginning of a new era in coordination that, by extension, establishes a moral basis for assistance. Still others remain confident that human rights as such do not belong in assistance discussions, and that they should be treated no differently than any other obstacles to the provision of aid. The sheer diversity of Afghan opinion-to the limited degree that it can be expressed and heard-is sometimes used as an excuse for inaction. Over time, however, frustration over unrelenting Taliban restrictions against women has reinforced a reaction against them and a concomitant fear that a future Taliban government would close Afghanistan to assistance. The response to other human rights abuses-including the arrest, detention, and disappearance of aid workers-affects the security environment in which aid can be provided, and brings home the most direct effects of rights abuses.

Afghanistan's humanitarian environment poses problems that appear to be generic to crisis countries and unique to its political and ideological situation. Once again, the common element among the five major humanitarian problems is the absence of a workable state. However unusual the ideological dictates of the Taliban, for example, it is the context of their promulgation that should be important to outside interlocutors-the reasons for humanitarian disasters, the nature of restricted choice-rather than their specificity. To the degree that outside agencies appear to impose solutions, they cannot help but seem paternalistic and insensitive to local opinion. The degree to which these agencies appear consumed by their own organizational difficulties reflects in part-although certainly not wholly-the absence of the disciplining authority that host governments might otherwise provide. The reform of institutional assistance is necessary to cope with the pressing needs of crisis countries, and equally imperative to cope with the failure of the Afghan state.

Who Owns Aid?

If a basic precept of relief and development is the integral involvement of beneficiaries, then aid to Afghanistan counts few successes. With the rise of fighting factions, the decline of the state, and the persistence of only a handful of small civil society organizations in exile, Afghans barely participate in aid. The few mechanisms that existed to secure their involvement in the design, execution, or management of aid projects have been removed. Afghan participation, for the moment, is restricted to employment by aid organizations, recruitment for temporary rehabilitation shura, and, on rare occasion, community organizations. Despite almost two decades of international humanitarian assistance, the roots of aid in Afghan communities remain shallow.

For a host of reasons the aid community has resisted involving Afghans in their own recovery. Bureaucratic inefficiencies, the arrogance of outsiders, difficulties ceding control to others, inflexible national partners and the structural dependency of foreign-initiated local NGOs, the fear that continuing war risks investments, as well as ignorance and occasional stubbornness-all these factors conspire to keep aid decisions away from those for whom it exists. These weaknesses are illuminated, but not caused by, Taliban restrictions on women and the attempted co-optation of rehabilitation shura. The reluctance of the UN political mission to deal with civilians has reinforced a sense of alienation. As a result, the important message that aid should bring-that the purpose of assistance is to help Afghans rebuild Afghanistan-is often lost in the din of competing ideologies.

It is too late simply to change the administration of aid to bring in more Afghans. (In any event, this practice cannot substitute for real capacity building.) Instead, new structures of assistance will be required that assume the primacy of Afghan participation in assistance. The political context(s) for such change suggest that this will be a terribly complex undertaking and if Afghanistan's recent history is any guide, will not proceed linearly. Two examples illuminate future difficulties. On the one hand, until some form of validated authority takes charge of government, involving communities without the sanction of presumptive authorities is difficult. On the other hand, until the international community clarifies its adherence to its own principles, it will be quite difficult to come to agreement with any Afghan authorities on the role that these principles must play in the organization of assistance. This question alone is likely to occupy the international community as it tries to determine when to confer formal recognition, with all the fiscal and political obligations that this implies.

The time to think about these questions, however, is before the outlines of new government are solidified, and while it is possible to craft new terms of political and assistance engagement. This may mean offering control of aid resources to localities in ways that aid professionals generally prefer to avoid, and that complicate the interactions that political negotiators have with Afghan society and future political leaders. The risks of failure are considerable. The risks of demurral, however, are far greater to the safety of the region and to Afghanistan's sovereignty.

Politics and the Conditions of Uncertainty

Humanitarian agencies tend to believe (or act as if they believe) that their decisions are, by definition, taken in the best interests of recipient populations and as a corollary, that they are the best representatives of besieged and beleaguered populations. They often speak on behalf of those who, they believe, cannot speak for themselves. Moreover, the long tenure of a UN political mission, working in parallel to assistance agencies but generally talking with different Afghan constituencies, creates confusion: who speaks for whom, and how is dissent about aid and politics reflected? In a war environment in which popular allegiance is at best uncertain, should the political consequences of aid influence the way that assistance is provided? Should political actors, whether local or international, take account of the nature of assistance in their deliberations and decisions?

These questions range far beyond the usual distinctions between aid as a carrot-increasing incentives to perform or behave as aid providers would like-and aid as a stick-conditioned or withdrawn when behavior does not change. Of course, political and humanitarian timetables are quite different: finding and taking advantage of opportunities takes on different meanings for both. But neither aid recipients nor de facto authorities distinguish among international interlocutors, except to the extent that a bit of manipulation may secure a bit of leverage. In this equation-aid is aid is aid-the UN, bilateral donors, NGOs, political negotiators and external spoilers are all part of one big pot of outside money and favor. When one assistance provider suspends its aid, all are affected; when one provider or agency collaborates with de facto authorities while others do not, all suffer. And when the aid and political branches of the United Nations do not appear to act in tandem, both endeavors suffer.

The specific forms these problems take in Afghanistan directly affect recipient populations. The UN Special Political Mission for Afghanistan (UNSMA) has long been in the business of encouraging cease-fire agreements among fighting factions to lay a basis for future peace negotiations. The Taliban's ascendancy has not fundamentally changed this effort, except to the degree that its military success has persuaded its leaders to pursue military as a means to political control. Battlefield successes thus far have been translated into ideological dominance, further complicating UNSMA's tasks. For example, Taliban practices ensure that human rights issues are on the agenda of security discussions with aid organizations and are also on the agenda for political negotiations among factions. It also means that belated plans to recruit UN human rights monitors raises the status of rights and principles in discussions about the future apportionment of authority in Afghanistan. Even if monitors are attached to the UN political mission rather than to aid organizations, in the eyes of Afghanistan, the entire international assistance community will bear the responsibility for their work. The need for monitoring is clear, but neither the aid agencies nor the political mission has been able to answer pressing questions that have been raised in other crisis countries. Can monitors protect rights, or simply note their abuse? Can aid protect rights? Does monitoring diminish assistance access, or increase it? Which doors are opened, which are closed, and to whom?

All such policy decisions have consequences that are direct, visible and far-reaching-whether in the delivery of goods, the availability of donor funding, the stature of fighting factions in the eyes of the diplomatic community, or the sources and breadth of foreign military assistance. As conditionality has become a more frequent part of assistance vocabulary-largely in response to rights abuses and discrimination-the content of assistance has become part of political discussions, and the political impact of assistance has become a frequent accompaniment to discussions about aid. Political negotiators have been involved in decisions about the suspension of assistance, due to their effect on the security of UN personnel, and have participated in negotiations about future assistance commitments with the Taliban. For the Taliban, discussions about assistance with multilateral organizations and bilateral donors have been a deliberate part of its efforts to secure de facto control and de jure recognition as the government of Afghanistan.

Although many recent negotiations have taken place in the face of crisis, the joined forces of assistance agencies, UN political negotiators and NGOs have by implication called into question the presumed impartiality of humanitarian assistance. Even so, many assistance actors, frustrated with the slow pace of rehabilitation and the slower end to war, have sought closer cooperation with the political mission. The distinction between formal politics and negotiations on the one hand, and on the other, informal, indirect politics across the country, is not always evident, but both affect matters of principle and practice. If the ends of the political and humanitarian enterprises-where do each lead, and do they get to the same place, together?-are not necessarily or always the same, will this make a difference to Afghans?

This question highlights the close, if anxious, relationship that binds aid to politics. Regardless of the time it takes for fighting to stop, and the forms of governance that will be available when it happens, a host of political problems will become, or remain, assistance problems. The next time a peace treaty appears on the horizon, several issues will be paramount: the role of former fighters and the nature of demobilization; the territorial home for winners and losers, the different development paths they may choose, and the tension between formal recognition and informal acquiescence to local power arrangements ; the effects of ideology on the principles guiding international assistance, if and when governance is restored; the capacity to return several million more refugees, their willingness to return, and the pressures that neighboring states will exert to force that return ; the role of transnational political parties in helping or hindering cross-border rehabilitation efforts; and the difficult problem of relocating assistance agencies to Afghanistan, if and when security conditions improve.

These questions were neither raised effectively nor answered properly ten years ago, when the Geneva Accords were signed. Silence then led to chaos later. All of these issues, and others, now join assistance and political negotiations at the hip-and require a new, probably more formal, reworking of assistance administration.

REFORMING INTERNATIONAL ASSISTANCE TO AFGHANISTAN

When assistance practitioners in Afghanistan describe their work, they often complain about familiar woes and inadequacies. Some typify international assistance anywhere, some are exacerbated by a vacant government and absent state. All, whether generic to the aid enterprise or specific to Afghanistan, contribute on occasion to confusion when clarity should prevail. These include: (a) confusion between policy discipline-which requires direction-and operational autonomy, leading to contradictory urges toward devolution (at headquarters and in the field) when policy direction is necessary-whether in response to conditions in Afghanistan, or as a natural part of institutional planning; (b) a sadly predictable but often lethal combination of mandate protection, mandate jealousy, and mandate creep that leads to fractious policies among agencies and NGOs, and offers ample opportunity for fighting factions to capitalize on these divisions; (c) rudimentary coordination among NGOs, often unwilling coordination among UN agencies, and indecisive coordination among donor governments, on pressing matters of policy; (d) predictable but pernicious policy divides between field operations and headquarters; (e) contradictory policies among UN agencies and their executive bodies, and between donor government bilateral policies and their policies as members of executive boards of UN agencies, funds and programs.

These tendencies wax and wane. Donor governments, for example, have been quite willing to take coordinated stands at some points, and then later-on the same issues-alter their stances when the external political environment changes. UN agencies have been known to act coordinately in the face of crisis, but splinter almost immediately thereafter. Decisions about policy, principles, priorities, and personnel have been difficult to conclude and even harder to maintain. Policy commitments are often more rhetorical than material, and even self-evident needs-for example, for simple assistance guidelines-are quite difficult to achieve.

In and for Afghanistan, as elsewhere, money drives the urge toward institutional autonomy and limits the capacity of coordination (and coordinators) to respond to enduring crisis. Those agencies or NGOs with adequate, autonomously derived funds are reluctant to be coordinated: the World Food Program, UNICEF, and Save the Children have been able to take independent decisions that, while heralded outside Afghanistan, affected other agencies and political negotiators without their prior knowledge or consent. Similarly, donor governments are able to, and often do, offer contrary instructions to UN agencies on the one hand, and to NGOs on the other, and they often encourage multilateral decisions that are not supported by their own bilateral policies. Although the substance of coordinated decisions may be at issue, too often it is the fact of coordination and consistency that is contested.

Such issues, as much, if not more than, the failures of the Afghan state, prompted a relatively thorough review and attempted reorganization of assistance to Afghanistan, starting in 1996. The progress of that endeavor illuminates the limits of assistance and the constraints that international assistance actors place on their own capacities for reform.

Reform from Within

The first efforts to restructure aid to Afghanistan were administrative. In the period that followed the Geneva Accords, UN assistance was propelled by two organizations-one humanitarian and one developmental, although the humanitarian coordinator held greater seniority and, at times, functioned as lead political interlocutor as well. UN agencies functioned as they now do in many crisis countries, with relative autonomy, minimal coordination, and an annual appeal to raise funds for cross-sectoral and coordinating needs, and to fill gaps in funding. Nongovernmental organizations were coordinated through representative bodies that had informal consultative relationships with the United Nations.

Political change in Afghanistan taxed this administrative structure. The frustrations-some would argue, failures-of UN efforts to craft a cease-fire suggested that war might continue forever. The costs of conducting regional assistance programs grew as the terrain of fighting factions changed. Moreover, the physical distance of the United Nations and most NGO headquarters in Pakistan seemed artificially to keep Afghanistan at bay, and the temporary nature of employment in the aid sector eroded the capacities of almost all assistance organizations to sustain their activities.

The creation of the UN Department of Humanitarian Affairs, to which UNOCHA was subsequently attached, seemed to elevate the profile of the Afghanistan dossier within the United Nations. But, despite sporadic peace in Afghanistan that raised hopes by early 1996 that rehabilitation and development might assume the assistance center stage, international inattention strained dwindling resources. This, in turn, affected the viability of NGOs, whose budgets were supported by international sources and UN contracting. These three reasons-the relation of humanitarian to development assistance, administrative changes within the United Nations, and the scarcity (and concomitant inflexibility) of resources-elicited renewed interest in restructuring assistance. At the same time, UNDP (like others) was struggling to find ways to understand the relationship of political change to development opportunities-a conceptual activity-and the relationship of its own mission to DHA-a bureaucratic matter. The two concerns were joined, almost by chance and rather uneasily, in efforts to craft a strategic framework for assistance.

The concept of strategic frameworks, as developed by UNDP, was straightforward in concept and complicated in practice. A framework was to help existing actors organize their efforts coordinately, and most important, to join their understanding of the political environment with humanitarian and, where possible, development needs. As with almost all UN-led efforts, however, persuading agencies to do in fact what they already agree to do in principle is extremely difficult. Once this endeavor was seen as a UNDP initiative, a quiet battle was joined at the headquarters level that led ultimately to an effort that diverged quite markedly from more practical, field-based activities.

From September 1996 through September 1997, several coordinate efforts were initiated in the field. They began by consolidating the UN presence by merging the leadership of UNOCHA and UNDP, and then reinforcing nascent local consultative institutions (the Afghanistan Task Force and the Afghanistan Support Group) that began more consciously to join donor governments, UN agencies, funds and programs, and NGOs in a common endeavor. Using the consolidated appeal as a mechanism to draw the assistance community into the strategic framework endeavor, the assistance community then began to employ the appeal process as a tool for assistance transition and reform. At the same time, UNDP reorganized its small budget to promote a cross-sectoral community development project explicitly designed to lead to community-based peace-building. To facilitate these efforts, it was necessary to understand the country that Afghanistan had become, and was becoming. Therefore, the United Nations began to develop and refine the empirical and policy foundation for programmatic reform in the fields of governance and political economy, gender, data collection and distribution, and communications. Although some of these activities began independently, they were all intended to involve the entire assistance community and, however tentatively, to develop common goals for the community.

Many activities were driven by equally by political events and managerial concerns. By late 1996, the Taliban movement had broadened its territorial control, strengthened is drive for ideological consistency, and had begun to seek international recognition. Its actions summoned concerns about the international community's policies, and helped to drive discussions about the goals of assistance. As was to be expected, however, those policies and activities that provided resources for the assistance community were generally applauded; those that proposed policy consistency, or required additional effort, were often resisted.

Much of the field-based program was pursued consciously as an adjunct to a headquarters-driven strategic framework process, although some reform was essential just to keep aid in business. During the same period, a series of headquarters decisions-led by the Administrative Committee on Coordination (ACC)-led to an interagency strategic framework mission in October 1997 whose work contributed to a document that was finalized by the Deputy Secretary-General in September of 1998. The change of responsibility within the United Nations-from UNDP through the ACC to temporary responsibility vested in the Department of Political Affairs, and then to the Deputy Secretary General-reflected discomfort among agency heads about the degree to which a framework for assistance would force their compliance with policies outside their control. At some point during this process-pushed, to a degree, by events in Afghanistan-the exercise broadened to become, it seemed erroneously, a virtual constitution for assistance agencies.

Some agency reluctance therefore reflected traditional interests in maintaining control of their own budgets and programs, and some reflected sharp divisions among the agencies about substantive policies including, notably, human rights and gender discrimination. Similar schisms developed among NGOs and between NGOs and the United Nations. Not surprisingly, the greatest stumbling block to a broadly-based strategic framework was the notion of broadly-based agreement at all, the commitments this would require, and the erosion of agency power that it would imply. Throughout, political leadership within the United Nations was lacking-the Secretary-General did not exercise authority in order to resolve these disputes.

Headquarters activities were therefore rather separate from a field-based reform process that developed an assistance strategy for Afghanistan, drafted in consultation with all field-based assistance actors and validated by the Afghanistan Support Group. This strategy offered a first step toward programmatic consistency. Consistency in principle, however, did not translate to consistency in action, and an initial drive to create a common program for Afghanistan, with a common fund for financing, was quickly replaced by an effort to engage in common programming. (In the bureaucracy of the assistance business, gerunds trump nouns.)

A habit of commonality, however, was gradually absorbed within the assistance community, even if the mechanisms for creating and enforcing it remained rudimentary. Common programming, common services, and efforts to develop common assessments of policy problems have begun to permeate the assistance community's assumptions about itself, although common responses to policy crises have yet to be commonplace. Cooperation bore early fruit: participatory emergency management mechanisms to cope with two major earthquakes in 1998 led to a reevaluation of common programming as a tool to prepare for natural disasters. But more profoundly, attempts to set a unitary policy on capacity building and governance-how should agencies deal with de facto authorities?-have been greeted with great reluctance, even in the face of the obvious, deleterious consequences of reactive fragmentation. Indeed-reflecting a failure to persuade headquarters to accept mechanisms developed in the field-significant negotiations to establish terms of engagement with the Taliban have been undertaken outside the consultative mechanisms for common programming.

The field-based strategy to move toward common programming concentrated on four distinct elements: (a) focusing on donor governments to solidify their fiscal commitments and to push agencies and NGOs beyond parochial interests toward shared goals ; (b) developing decisions by consensus and moving decision-making, as much as possible, to the field; (c) reducing ad hoc temporizing in favor of long-term planning, based on assessments of the long-term impact of aid rather than its short-term outputs; and, perhaps most important, (d) compensating for the absence of government and the weaknesses of the Afghan state by establishing self-regulating standards for the assistance community itself.

Each element of this strategy has achieved some successes. The end result, however-the effective design and delivery of assistance as a step toward achieving peace in Afghanistan-has yet to be realized. Institutional loyalties are quite difficult to change and thus far, the structure of rewards for bilateral as well as multilateral actors remains tied to old practices.

The headquarters process established a different foundation for change. The strategic framework document published by the Deputy Secretary-General, although motivated in part by assistance challenges in Afghanistan, bypasses and goes beyond the Afghanistan experience. In one sense, this was inevitable: by the time it was completed, a two-year reform process was in progress in the field. Although the strategic framework acknowledges a number of critical elements of the field-based assistance strategy, and pays homage to the initial UNDP effort to join the political and assistance enterprises, its greatest contribution lies elsewhere. It explicitly recognizes UN leadership in responding to complex emergencies and facilitating the coordinate goals of peace, reconciliation, and recovery. Its even more explicit bow to the concept of principled assistance reiterates a fundamental UN responsibility, and locates its assistance role firmly in its founding documents and principles. For Afghanistan-battered by war and the best intentions of the assistance community-a clear commitment to principled assistance provides a foundation for future institutional change.

If the United Nations follows its own leadership, the framework process will have been useful. If it envelopes itself in familiar disputes over turf and terrain, it can easily sully its hard-won progress and risk its relationships with donor governments, assistance partners, and aid beneficiaries alike. The field-based process moved forward in part because it could be described to all its participants as an element of UN reform as a whole. But as public, headquarters-based reform seemed to wane, the bona fides of the project paled as well. For this reason, the accession of donor governments to the goals and implementing mechanisms outlined in the strategic framework document and in the local assistance strategy is essential if principled assistance is to move from a dusty office shelf to the dusty plains of Afghanistan.

Making Reform Work

Assistance reform suffers from weaknesses that affect the entire international community.

Money and voice. Fiscal structures seem almost immune to change, and this inertia undermines the progress of all consensus-building and cooperative enterprise. As long as money drives decisions-and funds are raised and administered by specialized agencies, autonomous funds and NGOs, and quasi-independent programs-consultation within the United Nations will be limited and collaboration will not work. UN and NGO decisions are taken coordinately only when they represent few real costs, or when desperation sets in (regarding, for example, negotiations with presumptive authorities about security). Even more decisively, donor governments retain rights and influence that defy the principles of partnership. Money talks, but it rarely listens: the two roles that donor governments play-as benefactors and as bilateral political actors in an unusually fluid and dangerous environment-often contradict each other. Similarly, NGOs act as independent actors but also work as contractors for the United Nations. Multiple, overlapping roles have caused confusions of voice and representation, and multiple sources of funding feed this confusion.

In its own small way, the strategic framework has substituted for a housecleaning that has yet to be done. To correct this problem requires action that goes far beyond the dossier of a strategic framework or local assistance strategy. The accumulated histories and habits that pervade the operations of UN agencies, funds and programs have created a governance structure that defies logic. Each pursues its own agendas, and some have created their own foreign policies. Interagency disputes sometimes overshadow the wars whose victims these agencies should protect. The fact that the humanitarian enterprise is riven with disputes-however understandable-sullies the reputation, presence, and work that the United Nations should otherwise be able to achieve.

By extension, this means that the fiscal underpinning of agency autonomy must be revised. Decisions should drive money: raising and spending funds should be undertaken to support common policies. Only when such discipline is clearly understood within the United Nations can serious partnerships with donor governments and nongovernmental organizations be useful.

When the United Nations speaks, it must be able to do so with one voice, and certainly with one humanitarian voice-and that voice must belong to the Secretary-General. This is important in any emergency but is critical in failed states.

Political consistency. A framework for peace and reconstruction was to prepare for post-war governance in Afghanistan. This meant that preparing for peace-through political negotiations and through assistance-would be undertaken coordinately. The reluctance of the international community to engage in reconstruction in Afghanistan has been detailed here and elsewhere. But its hesitancy to join-conceptually and politically-the political enterprise with assistance is a major failing. Whether in developing an understanding of political sensibilities within Afghanistan, reworking an understanding of the region's role in prolonging or settling crisis, or deciding when external actors should remove themselves from difficult situations-to date, the international community has let each actor work as it wishes, leaving Afghans to live with the political consequences of disjointed aid. Without serious thinking that systematically dissects the political impacts of aid decisions, the aid impacts of political decisions, and the costs and consequences of taking decisions about them together, it is impossible to judge whether a de facto common program-as refugee repatriation has been for several years-is appropriate.

The framework process does not provide a way to judge whether the surrounding political environment, including the policies and governance of neighboring states, supports or deters international assistance. This question is important not only because Iran and Pakistan are host and conduit countries, but also because UN political strategy has been premised on the cooperation of neighboring states to end war. The strategic framework has not found a way to combine these functions in a common pursuit.

Indeed, although the strategic framework is now viewed as a mechanism for dealing with the problems of collapsed states, it is too modest a venture for this task. The steps that the strategic framework proposes for joining the political and humanitarian enterprises are too tentative. Failed states like Afghanistan are besieged by humanitarian crisis, but their primary problem is political. There is no question that political engagements-negotiating with fighting factions and with those who support them-may appear to tarnish the name of humanitarianism. But as this discussion details, no endeavor is pure in conditions of political failure, institutional collapse, and humanitarian crisis. There are situations-and Afghanistan is one of them-where the long-term goal of resolving political crisis may be more pressing than the shorter-term goals of relief and rehabilitation.

Minimalism. The environment in Afghanistan argues for a minimalism that is missing from the assistance vocabulary. The complexity of assistance-defined variously as emergency aid, relief, rehabilitation, development and reconstruction-has almost paralyzed the humanitarian and political environments. Negotiating the terms of assistance has, on occasion, substituted for determining conditions for peace; providing aid as a means to build peace has, on occasion, become just another way to provide aid during war. The basic preparations for peace, however, lie in those tools that allow Afghans to regain control of their lives, however minimally: removing land mines, providing immunization against communicable diseases, and securing access to food.

Even this limited agenda of aid runs afoul of Taliban restrictions, and it does not ensure that the twinned obstacles of war and state failure can be overcome. Critically, however, a minimalist strategy accomplishes two objectives: it maintains developmental consistency by providing only that assistance that can be controlled by beneficiary communities and in ways that allow them to expand their capacities, and it removes assistance-related complications that have stymied political progress.

Were the United Nations able to agree on a minimalist program, it would be unlikely to secure the support of NGOs for similar constraints on their programs. Such potential divergence is not fatal: NGOs are not-and should not be-fully implicated in the political agenda of the United Nations. For now, the strategic framework and other existing consultative mechanisms offers opportunities for debate without the means for taking broad decisions-particularly about removing aid from Afghanistan. Taking such decisions, however, is crucial for the success of any humanitarian endeavor in states like Afghanistan, and is critical to the survival of the state itself.

Knowledge. The analytical framework and the resources that accompany assistance are radically inadequate. No personnel or institutions provide the consistent and wide-ranging analysis that common services require. Without this knowledge, however, principled assistance-whether comprehensive or minimalist-remains only an aspiration. Even the simplest wordings elicit probing questions: what does it mean to provide assistance as part of an overall effort to achieve peace? Can aid be provided solely on the basis of need, and without distinction by gender-or does this principle preclude aid in Taliban-controlled areas? When does aid contribute to war rather than peace, and how should assistance actors determine this? To interpret the principled provision of assistance should mean-first and foremost-that the entire assistance community, including political counterparts, partakes of a common understanding of the situations in which its members labor. Little in the strategic framework or the field-based assistance strategy for Afghanistan mandates this kind of work. Little, however, can be accomplished without it.

Local participation. The essential element of all assistance should be the participation of Afghan citizens in the full range of political and assistance activities. This idea is present in the strategic framework only by implication, and although it is mooted in the assistance strategy for Afghanistan, it has proven quite difficult to fulfill. The message has also been lost in the long process of negotiations about the status and protections required to maintain a small expatriate UN staff in Afghanistan, although the majority of aid workers in the field are Afghans.

Experience elsewhere tells us that local participation is integral to aid, particularly as a tool for building peace. Were the efforts of the international assistance community to be directed toward one activity in the coming years, this is where it should turn. All major decisions-how to deal with de facto officials and invalidated governments, how to craft peace agreements that reach the village level, how to include the broadest range of Afghan citizens in the quest for peace and reconstruction-must be taken with the representative participation of the Afghan people. Far too little effort has been devoted to finding ways to engage Afghans in decisions about aid, but these are critical to the survival of a useful aid regime. These are decision that Afghans, not outsiders, should take-particularly when personal sacrifice may be the price for participation. Paying only rhetorical respect to the idea of local ownership is misleading, and may be worse than receiving no mention at all.

Transition toward reform. The reformulation of the role of the United Nations in Afghanistan-and a reorientation of its relationships with partner organizations and the people of Afghanistan-relies on the capacity of the United Nations to reform itself, from headquarters and for its headquarters. The global governance problems that UN agencies so often exemplify are not covered in the strategic framework: indirect allusions to cranky recalcitrance do not, in themselves, offer a transition toward effective governance. For a process that has relied so much on the concept of process, at the end we are faced with a missing transition process.

The process that led to the strategic framework and field-level reform suggests that coordination is an inadequate methodology for reforming aid to failed states, and certainly offers a mixed metaphor of international assistance. The best that reform has offered is an umbrella of coordination rather than direction. This managerial weakness undercuts the thoughtful directives that the strategic framework proposes. To date, like-minded actors-donor governments, UN agencies, and NGOs-may have administrative reason to act similarly, but not to act together. Together, however, stakeholder groups-characterized more by their eponymous capacity to stake out territory than to cede it-may not be possible to coordinate, even minimally, without a system of rewards and sanctions, command and control. These are not available, however, to UN coordinators and would not, in any case, serve to organize an entire assistance community. The alternative-clear authority, with clear operational protocols to underscore it-makes more sense, but contravenes accepted practice. Nevertheless, it is the best tool for transition toward more effective assistance in failed states and equally more effective political direction to resolve persistent political crisis.

A drive to centralize the control of assistance seems to be at odds with the need to cede the control of assistance to Afghans. Present circumstances-those within the assistance community, and those of Afghanistan-suggest that there is no way to move from a fragmented assistance structure in a fragmented state toward peace and reconstruction. Centralizing authority, at the very least within the United Nations on issues of policy, may be the only way to transcend the particularities of the assistance condition. In the end-whether the end of a transition process, or when peace finally comes-the ways that assistance interacts with Afghanistan could change dramatically.

For now, the structure of international assistance is internally contradictory: democratic decentralization in the field, fiscal dictation at headquarters, and the hidden hand of donor injunction offers an interesting but self-defeating lesson in the didactics of complex organizations. It is the return of politics in Afghanistan, not the rule of foreign funds, that should guide change. Reform thus far has meant cleaning house, when a new house is needed. This is what it will mean to rebuild Afghanistan, so that the architecture of assistance can begin to serve the country and its people.

NOTES

1. Antonio Donini, Eric Dudley, Ron Ockwell, Afghanistan: Coordination in a Fragmented State, UN/DHA, December 1996.

2. See, e.g., Asger Christensen, Aiding Afghanistan: The Background and Prospects for Reconstruction in a Fragmented Society, Swedish International Development Agency (SIDA).

3. Kjell Ostrom, Understanding the Economy of Afghanistan: An Exploratory Study, SIDA, January 1997.

4. See, e.g., Peter Marsden, The Taliban: War, Religion and the New Order in Afghanistan, Oxford (Karachi) and Zed (London), 1998; and William Maley, ed., Fundamentalism Reborn? Afghanistan and the Taliban, Vanguard (Lahore) 1998.

5. See Angela Kearney, "Why we closed in Herat," Aina, Vol. 2, September 1996; and, for an alternate view, press release from the Swedish Committee for Afghanistan with regard to girls' education in Taliban controlled areas, 1997

6. Memorandum of Understanding between the Islamic Emirate of Afghanistan and the United Nations, 13 May 1998

7. See Human Rights Watch/Asia, Afghanistan: The Massacre in Mazar-i-Sharif, November 1998; and Amnesty International, Afghanistan: Detention and Killing of Political Personalities, 11 March 1999.

8. Paula R. Newberg, et al., Political Economy in Afghanistan: District Profiles, UNDP/Afghanistan, 10 May 1998.

9. Political economy in Afghanistan.

10. This conclusion is implied in Claude Bruderlein, Report of the DHA Mission to Afghanistan, UN/DHA 1997.

11. The Interim Operation Strategy Note for the United Nations in Afghanistan for the 1996-97 Biennium cites governance and stability, and human rights as the primary inter-sectoral issues critical to Afghanistan's recovery.

12. According to the Afghanistan director for the Swedish Committee for Afghanistan, rural populations are at once more conservative than urban dwellers, and unlikely to take dictation in ideology or religion from the Taliban. See "Difficulties and Opportunities: Challenges of Aid in Afghanistan," Stockholm, 24 February 1999.

13. For policy, see Afghanistan P.E.A.C.E. Initiative, Six-Month Program Review Meeting, 29-31 October 1997; for discussion, see Paula R. Newberg, "Principles, Capacity Building and Gender in Afghanistan," Report to the Afghanistan Support Group, 1 May 1998.

14. Donini, et al., Coordination in a Fragmented State

15. Paula R. Newberg and Michael Dalton, Governance in Afghanistan: Report of the UNDP Mission, April 1997.

16. See Ken Menhkaus, "Good Governance and Foreign Assistance," in Horn of Africa Bulletin, July-August 1997, for a discussion of local ownership in designing similar programs in Somalia.

17. United Nations Office of the Special Adviser on Gender Issues and Advancement of Women, Report of the United Nations Interagency Gender Mission to Afghanistan, 12-24 November 1997.

18. For a fuller analysis, see Newberg, "Principles, Capacity Building and Gender in Afghanistan," Report to the Afghanistan Support Group, 1 May 1997.

19. See Menkhaus for a discussion of the risks in Somalia.

20. See Sergio Vieira de Mello, Mission Report: Afghanistan, 20-27 February 1998.

21. Talks among factions hosted by the Government of Turkmenistan and facilitated by UNSMA brought these questions to the fore in March 1999.

22. In March 1999, the Government of Pakistan announced a new policy to return all Afghan refugees to camps, and to limit Afghan entrepreneurial activities, including those in the social service sector.

23. Negotiations between the assistance community and the Taliban led to the hesitant return of a small number of expatriate aid workers to Kabul. They had been absent for more than six months after a UN official was killed during disturbances following the American cruise missile attack on Afghanistan in August 1998.

24. A "Common Donor Position on Aid (to Afghanistan), October 1996," was proposed and agreed to within ten days of the Taliban's entry into Kabul; by the spring of 1998, donor coordination was far more tentative. Similarly, donor applause for a common assistance strategy at the Afghanistan Support Group meeting in December 1997-including a decision to vote to support its principles at Executive Board meetings-was not implemented with notable speed.

25. Agencies initially supported the suspension of aid in the southwestern region, but almost immediately lobbied to return, well before the conditions for return were articulated.

26. For a full description, see Donini et al., Coordination in a Fragmented State

27. See Briefing to Donors on Afghanistan: Longer-term Rehabilitation Assistance and its Relationship with Humanitarian Programs, Stockholm, 1-2 June 1995.

28. Within a short time, UNDP had created its own literature on the subject. See, e.g. "United Nations Development Program, Afghanistan Program: Formulation of Strategic Framework, 26 February 1997" ; "The Design and Formulation of a Strategic Framework for Recovery" ; and Hugh Cholmondeley, "The Role of the United Nations System in Post-Conflict Recovery," August 1996.

29. This was to be undertaken for all war-torn societies in which UNDP worked, including those formally labeled "post-conflict". UNDP chose seven crisis countries as test cases for strategic frameworks. Later disputes about locution and labels led to the appellation of "strategic framework" to be applied only to Afghanistan-where it seemed to work-and Mozambique-where it was not pursued.

30. Once again, Afghanistan itself defeated this goal. Pervasive discrimination, continuing war, and competing crises deterred donors, who, as of March 1999, had contributed less than 10 percent of the funds solicited in the 1999 annual appeal.

31. Strategic Framework for Afghanistan: Towards a Principled Approach to Peace and Reconstruction, 12 September 1998.

32. Agreements to return expatriate workers to Kabul in March 1999 were undercut when the United States and the United Kingdom cited security threats in their refusals to allow their citizens to return.

33. See Donini, et al., Coordination in a Fragmented State, for a discussion of the usual progression of aid relationships, which assumes that, in the end, a single state entity will interact with multiple UN agencies. This need not be the result of transition: pluralism may exist with the state, while unity is encouraged within the assistance community.

ACKNOWLEDGMENTS

From the summer of 1996 through the autumn of 1998, I served as Special Advisor to the United Nations Resident Coordinator in Afghanistan. During that time, my primary responsibility was to help the assistance community rethink its role and functions. This was an unusual undertaking, sustained by a wide-ranging and welcoming aid community that has endured years of experimentation with grace and dedication. My thanks to Erling Dessau for inviting me to come to Afghanistan during his brief tenure as UNDP Resident Representative; to Maurice de Wulf for his indefatigable imagination as UNDP Deputy Resident Representative; and to Michael Keating, who had the delicate job of facilitating a strategic framework initiative without losing anyone to the cause. Lakhdar Brahimi, the UN Secretary-General's Special Envoy for Afghanistan, helped to bridge the distance between politics and assistance that has long troubled Afghanistan's most ardent helpers. My primary thanks, however, go to Alfredo Witschi-Cestari, UN Resident Coordinator, whose untiring efforts to make assistance work for Afghanistan created an uncommon environment in which constructive dissent became an integral part of the assistance community's collective endeavors. Although the analysis and conclusions detailed here are mine alone, I hope this discussion reflects our shared concerns for Afghanistan's future.

Paula Newberg
Washington, D.C.