Back to "Afghanistan Outlook" Table of Contents

Afghanistan Outlook

UNITED NATIONS -- The United Nations Office of the Co-ordinator for Afghanistan, UNSMA and the World Bank

Conflict and Peace in Afghanistan

By Barnett R. Rubin

The war in Afghanistan is a transnational war. The Afghan protagonists in the apparently "civil" war are all linked outwardly to state and non-state actors in the region and the world and inwardly to a variety of social actors in Afghanistan.

These links are strategic, ethnic, economic, religious, and opportunistic. The logic of such a war differs from both interstate wars and traditional civil wars (a government versus an insurgency fighting for revolution, autonomy, or secession). Ending such a war includes elements of three sets of problems: ending a civil war, ending an international conflict, and constructing the institutions of a national state.

The three problems

It includes the problems of ending a civil war, because the direct combatants are citizens of the same state. Peace can come only when they agree (voluntarily or otherwise) to live under the same authority. But ending such a war is more difficult than ending a classic civil war. The combatants are less likely to become exhausted, as access to external resources through both aid from states and movements and participation in economic networks is virtually open-ended, as Afghanistan has a smaller population and is poorer than the surrounding countries.

Ending such a war partakes of the problems of ending an interstate war, because regional states and great powers have competing or antagonistic interests in the conduct and outcome of the conflict. Fortunately, these do not involve territorial claims against Afghanistan (there appears to be a firm international consensus in favour of maintaining the country's juridical existence in its current boundaries), but they do involve security, economic interest, ethnic ties, and ideology.

The states involved include Iran, Pakistan, Uzbekistan, Turkmenistan, Tajikistan, Russia, the U.S., and Saudi Arabia. Others are involved but a bit more peripherally (India, Turkey), economically (United Arab Emirates), or as donors (Japan and several European countries as well as the EU). Furthermore, the region centred on Afghanistan has little multilateral or security architecture and has been transformed in the last decade by the independence of the Central Asian and Trans-Caucasian former Soviet republics.

Afghanistan is at the intersection of three regions with distinct security and economic problems: the Middle East/Persian Gulf; South Asia; and Central Asia/Caucasus. Each of these regions is also affected by US global strategy. The interaction among these three regional formations has had very damaging consequences for Afghanistan. Pakistan's security dilemma in relation to India leads it to seek connections to Central Asia through Afghanistan, putting it in conflict with Russia and Iran. The latter has seen access to Central Asia as leverage against US sanctions and its isolation in the Persian Gulf. Thus, their attempts to solve dilemmas in South Asia and the Persian Gulf and in relation to the U.S. lead Iran and Pakistan to clash over access to Central Asia in Afghanistan.

Afghanistan's relations to these three regions have also made it the centre of a regional conflict formation. Conflicts and civil unrest in Kashmir, Pakistan (Karachi and sectarian conflict), Tajikistan, and Uzbekistan (with spillover into Kyrgyzstan), as well as the domestic political struggle in Iran, are all linked to Afghanistan in various ways. It has also, of course, become a haven for members of armed groups from Algeria to the Philippines and the Comoros. As these various conflicts become more tightly linked, it becomes more difficult to resolve any of them separately. Note that the Dayton Accords had the effect of increasing the incentives for Kosovar Albanians to use violence. Such perverse regional effects can occur in any regional conflict formation.

Ending such a war partakes of the problem of constructing the institutions of a national state, because only an Afghan state with minimal capacities and legitimacy (domestic and international) will be able to provide enough security to its inhabitants and neighbours that they will forego armed struggle. Despite the consensus on maintaining Afghanistan's formal juridical integrity, most of the institutions of the nation-state built up by the royal regime have collapsed. As discussed below, this means that the protagonists in the war are not political parties who can negotiate power sharing but armed movements with differing state (or anti-state) projects. The problems of constructing a state that will be at peace with its neighbours and population requires attention not only to political and military issues but also to ethnic, economic, and social ones, including gender. The state in the region (not only in Afghanistan) competes with alternative forms of organisations such as religious networks, long-distance trading leagues (contraband and otherwise), trans-border diasporas or military groups, rogue agencies, and local solidarity networks of various types. The state lacks hegemony not only domestically but also internationally.

Ending a civil war

Ending a civil war is exceedingly difficult, and ending a transnational conflict of this sort is even more difficult. One of the most robust results in the study of conflict is that interstate wars are far more likely than others to end in negotiated solutions. Depending on some definitional questions, one study estimated the share of "civil wars" ending through negotiations rather than victory of one party was much lower than the comparable figure for interstate wars. Fifteen to 30 percent of civil wars end through negotiated settlements (the higher figure includes some marginal and special cases). Another study found that no civil war ended through negotiations without the presence of a third party (peacekeepers). There are several reasons adduced in the literature for these findings:

The issues at stake in civil wars tend to be indivisible, such as who is to be sovereign over a territory. These are winner-take-all issues, impossible to negotiate. (This surfaced in the Taliban's insistence after the Ashqabad talks that "leadership," i.e. the rule of Mulla Muhammad 'Umar as Amir al-Mu'minin, was non-negotiable.) Others have noted that civil wars can be over divisible and negotiable issues, such as ethnic power sharing.

Fighters agree to negotiate when they believe they cannot achieve more through war, i.e. when they reach a "hurting stalemate." In closed civil wars, the parties may in fact become exhausted. In transnational wars, however, access to external aid and international markets (drugs, gems, smuggled consumer goods) and even international volunteers or recruits from refugees, a diaspora, or allied states provide replenishable resources. The Taliban are linked to the Pakistan (and formerly Saudi) governments and intelligence agencies; they are also connected to Deobandi parties and madrasas, the Afghan Pashtun refugee diaspora, the Afghan trading network that extends from the Persian Gulf to Pakistan via Iran and Central Asia, local authorities in Pakistan, the drug trade, and some international Islamists. Massoud has a base in Tajikistan and receives Russian and Iranian aid. He is tied to different trading networks in gems, possibly to the Badakhshan drug trade, and other commerce. Dostum has commercial links to Central Asia and probably Turkey and co-operates with Karimov in training and basing militias for operations in Tajikistan. The Shia parties are linked to religious networks in Iran and receive aid from that country, but do not have an Iranian recruiting base as the Taliban have a Pakistani one.

The parties may not want to end the war. Many actors ("spoilers") can become rich and powerful by exercising violence in a lawless environment, and they will attempt to subvert any peace process. They may even oppose full-scale victory by "their own" side, since that might lead to some reassertion of legality or marginalisation of actors depending on violence. Those benefiting from the drug trade and smuggling could fall into this category.

Even when the parties sign an agreement sincerely, it is very difficult to implement without third-party guarantees of security. Any negotiated settlement to a civil war involves some disarmament and integration of formerly hostile armed forces and the participation by former combatants in a common political space (e.g. in institutions located in the capital city). That is, the two sides must undertake a transition from a situation where each is responsible for its own security in separate areas to one where that responsibility is shared or exercised by forces of the other side in all areas. Without third party guarantees (usually armed peacekeepers), no one is willing to risk such a transition. Where judicial, police, and legal systems are weak or non-existent, it is even more difficult to undertake such a transition. For instance, try to imagine Massoud taking up a position in a coalition government with the Taliban. Who is responsible for his security in Kabul?

Interstate aspects

In principle it should be easier to resolve the interstate aspects of the conflict, which in turn would affect some of the obstacles to resolution of the civil war, as mentioned above.

Unfortunately, the characteristics of the interstate system around Afghanistan fail to live up to the ideal type of interstate relations. States theoretically find it easier to reach agreements because they mutually agree on their boundaries, and they are hierarchically organised sovereign entities that can take decisions and implement them. In addition, the issues at stake in interstate wars, such as borders or control over resources, are often negotiable. Finally, international society has developed a set of norms and even laws regulating interstate relations, which, even if there is no sovereign to enforce them, strongly regulate the relations among states (e.g. the Vienna Convention). In this region, however, there are a number of obstacles to "normal" interstate relations.

No other state in the region has decayed as much as Afghanistan, but many regional states are extremely unstable, incoherent, and insecure. The presence of Afghanistan's porous borders means, given the nature of borders, that all of its neighbours also have porous borders and exercise indirect control or influence over ill-defined zones in Afghanistan. Afghanistan as a closed buffer state was a stabilising factor for its neighbours; Afghanistan as an open failed state undermines the statehood of its neighbours. Much could be said about Russia and Uzbekistan's fears of the effect of Islamic militancy ("Wahhabism") on their internal stability, Iran's competing elites and cross-border ties to Afghan Shiites, Tajikistan's civil war and its relationship to Afghanistan, or Turkmenistan's desperate search for an outlet to the world market for its natural gas in order to avert the collapse of its fragile dictatorship. Pakistan, however, remains the core of the problem.

The role of Pakistan

Pakistan is a fundamentally and existentially insecure state that feels the need to achieve military parity with its neighbour India, with seven times its population and economy. The resultant excessive military expenditures, along with massive corruption and mismanagement, have left the state bankrupt and major political institutions discredited.

Insecurity about India has also led to Pakistan's search for "strategic depth" to the west and north, in Afghanistan and Central Asia. Since the country itself has few resources, these operations have tended to be foreign-funded, by the US and Saudi Arabia during the Cold War, by Saudi Arabia afterwards (until the crisis over Usama bin Ladin), and by other Arab contributions and drug trade and smuggling income.

There appears to be little if any hierarchical political control over Pakistan's security policy, including its policy on the ground in Afghanistan. It is questionable whether that policy is even run by the regular military or intelligence apparatus through official channels or whether it is a quasi-rogue operation that is more or less authorised but not controlled. It is not clear who if anyone would have the power to change it.

The government appears powerless to control sectarian (Sunni-Shi'a) and ethno-political (muhajirs vs. others in Karachi) violence. Much of the violence is linked to Afghanistan through groups that have bases there. Political parties such as Harakat-ul-Ansar use Afghan territory as a staging ground for operations in Kashmir, in co-operation, at least in the past, with Usama bin Ladin and his colleagues. These organisations collaborate, at least at the operational level, with the ISI in Kashmir.

Important Pakistani elites (that support political parties) have become rich on the drug trade and smuggling operations that depend for their success on continued disorder in Afghanistan and the growth of illicit transnational networks.

Paradoxically, while pressure against Pakistan is necessary, a strategy to stabilise and reform it and increase its security is essential to any peace process. Pakistan's involvement in Afghanistan could continue, but in a positive way through economic participation in reconstruction, for instance. A policy that is solely based on threats and pressure without taking into account some of Pakistan's legitimate needs is less likely to succeed. Afghanistan cannot be stabilised by destabilising Pakistan.

Iran and Russia

Iran and Russia's policies appear to be mainly reactive. They are determined to prevent Taliban consolidation in Afghanistan because of the range of threats they believe this would pose to them. Their position virtually rules out the possibility of war termination through a stable Taliban victory. Both deeply distrust Pakistan. To the extent that India makes Pakistan feel less secure, that also increases Pakistan's determination to persevere in eliminating Afghan national forces that might co-operate with India. Iran also wishes to prevent the development of access to Central Asia through Afghanistan in order to assure that the price of continued U.S. economic sanctions is continued blockage of Central Asia's access to the Indian Ocean and Persian Gulf. This is an important source of leverage for Iran. Russia similarly prefers to keep the Afghan route closed.

Collapse of the state

The collapse of the national state in Afghanistan conditions the strategies of all actors, though the actions of particular agents also brought about and perpetuated the collapse of the state. Rebuilding the state-- which is a different problem from sharing power in a government-- is the fundamental task of any peace settlement. This has consequences for what kind of peace process is possible or desirable.

State-like structures are not completely absent in Afghanistan. The Taliban have reconstructed some rudiments of a centralised military and administration in the territories under their control. They have disarmed populations (relatively) and reintroduced the distinction between military and policing functions in the cities under their control. They are reportedly about to introduce a written constitution. Massoud appears to have done something similar in the territories he controls directly, but, despite several plans, there is no United Front ("Northern Alliance") centralised administration, military, or legal structure. One hears of administrative offices and courts functioning fragmentarily in various parts of the country on the basis of laws and regulations left over from various old regimes. Despite much talk of "revolution," it appears that the basic functioning of state offices, where these continued to exist, did not change much under either the PDPA-Watan Party regime or the Islamic State of Afghanistan. I heard such talk, for instance, in Farah in 1998 and in Taluqan in 1996 but could not verify it. UNDP carried out a rough survey of such institutions in 1997, but I have not seen it.

Nonetheless, the basic elements of a state are absent. There is no professional army with even a modicum of professionalism or autonomy from a particular political leadership. There is no civil service. There is no national police. There is no constitution. There is no legal code (shari'a is not a code, though it can be codified). There is no national revenue system. Various taxes and tributes are collected, but there is no budgeting, or at least the major expenditures (the war) have nothing to do with any budget. Mulla 'Umar reportedly keeps a chest near his charpai that he treats like a seventh-century bait al-mal, doling out wads of cash to those who find favour in his eyes. The Taliban tax the drug trade and the transit trade from Dubai to Pakistan. They are, of course, still hoping for the Turkmenistan-Pakistan gas pipeline, though prospects for financing seem slim. Massoud is developing gem exports with the help of a Polish company.

Strangely enough, the sense of common nationhood does not seem to have eroded. I do not know through what means it has been maintained (BBC?-- the usual explanation for the unexplained), but I hear not much more talk of ethnic separatism than I did 15 years ago, when I started working on Afghanistan. Hazaras are demanding autonomy, but they also want a share in central power, not separation from it. Despite the ethnic origin of the term, Afghans of all ethnic groups still insist, as far as I can tell, that they are Afghans and that the solution or the war will take place within an Afghan framework.

Obstacles to the state

While there are many obstacles to the construction of an Afghan state, let me list just a few of the principal ones.

Domestic legitimacy

There is no consensus on the source of domestic legitimacy of state power. The Taliban have a rather circular definition: it comes from the selection of (technically bay'a to) the amir by ahl al-hal-u-'aqd, which they define as a group of ulama selected and recognised by the amir. I am not sure if this is a Deobandi idea, a Salafi practice, or of some other origin, but it differs from both more traditional Hanafi practice of recognising any de facto Muslim ruler who enforces shari'a and the modern Islamist ideologies, which do not accord such a role to ulama (in Iran the clergy-based power of the leader, courts, and Council of Experts is balanced by the elected president and parliament).

In insisting on ruling directly themselves, the Taliban also differ from the traditional Pashtun conjunction of charismatic ulama legitimating a jihad carried out by tribal forces, whose victory they subsequently consecrate. Their legitimacy formula allows no room for political participation by Afghans who are not Taliban-approved ulama, which also virtually precludes political negotiations.

The other groups seem less concerned with legitimacy. Rabbani also experimented with an appointed ahl al-hal-u-'aqd while he was in power, but the discourse of a national-Islamic legitimisation of power Loya Jirga plus shari'a seems to be gaining ground in the public statements of Massoud and others in the UF. The Loya Jirga, though a theoretically attractive way to legitimate power by popular (or at least national) sovereignty without demanding impossible elections, was always a state institution, not a spontaneous emanation of an "Afghan society" that never existed. It depended on national leadership to convene it, which does not now exist. With belated US support, Zahir Shah is now trying to revive the tradition of the emergency Loya Jirga for situations when the state is contested or illegitimate.

International legitimacy

International legitimacy would also be required for a reconstructed Afghan state. Pakistan is the only significant state now actively advocating recognition of the Taliban. Taliban practices such as banning (or severely limiting) women's education, full access to health care, and employment on principle preclude such legitimacy. The state will not receive de jure recognition and aid unless certain minimal conditions are fulfilled. Those have not, moreover, been clearly defined by the international community, and the gradual increase in pragmatic dealing with the Taliban could start to accord them a kind of de facto legitimacy.

That phenomenon, of course, results from some of the Taliban's genuine achievements, namely the provision of more public security and less corruption than most of the other groups, making humanitarian access and economic activities somewhat easier.

The US has set forth a list of conditions on which ceasing support for terrorism occupies first place, and formation of a broad-based government is part of the 6+2 formula. Measures to reduce production and sale of opium are also usually included on this list, though this has to be treated in large measure as a development problem rather than one of law enforcement. Iran, Russia, Uzbekistan, and Tajikistan have more direct interests in opposing recognition of the Taliban. Turkmenistan and the UAE have strong economic interests in improving the Taliban's international profile.

The army

A unified state will require a unified army. At present in Afghanistan and the surrounding region, there are a range of armed groups at various points along the spectrum from political to criminal. Each is closely linked to a particular leadership. State reconstruction will require an army linked to state power and recruited nationally.

There are also issues to be settled concerning the size and doctrine of such an army. The doctrine of the army is closely linked to the structure of the state. The highly centralised model favoured by the Taliban (similar to that of Amir Abd al-Rahman Khan) requires a large army with a mission focused on domestic control. A more decentralised model (see below) would require a smaller army and more focus on local police and self-defence forces.

Centralisation

A reconstructed state will have to define the degree of centralisation, the relationship of administration to representation, and in particular the role of regions, localities, solidarity groups, and ethnic groups in governance. The pre-war Afghan state was weak but highly centralised, with all local officials appointed from the centre and consultative representative bodies convened only at the national level. Amanullah Khan's constitution, however, called for shuras at all levels of government, and this model appears to be popular with the Afghan public and even some Taliban. In the past twenty years local solidarity groups and regional-ethnic coalitions have become armed and mobilised (though the Taliban have disarmed and demobilised some). The alternatives are conquest, disarming, and centralisation (the Taliban model) or integration, disarming, and decentralisation

Afghan Pashtun and Tajik political-military leaders tend to oppose schemes for decentralisation. When they were in power Massoud and Rabbani espoused a centralised model but de facto recognised the local power-holders in Jalalabad, Herat, and Mazar. Massoud has spoken of an officially decentralised system since losing Kabul. The Taliban and some others see decentralisation or federalism as legitimating warlordism (or at limiting their power once they control the government). Uzbek and Hazara leaders, who know they will never control the government, favour decentralisation, by which, unfortunately, many of them seem to mean warlordism. True decentralisation, of course, must be part of a state building strategy, not a form of resistance to it.

Civil service

A state will require a civil service and other trained cadres, who are mostly dead or in exile and, in any case, enjoy little status or autonomy (not to mention pay) under any of the groups. A revived education system would be needed to rebuild such a group.

Local power

The war and political relations have become entangled with many local conflicts among sub-ethnic solidarity groups (qawm), always the building blocks of Afghan politics. These involve conflicts over land, pasturage, trade, and other matters. Any change of power in Kabul, now as always, involves a shift in local power structures as the new power tries to root itself in local alliances that may bear little if any relation to the ostensible legitimisation formula of the regime (but may reveal much about the real source of its power). The Taliban have also intervened far more than recent rulers in relations between the sexes as well as relations among qawms. Re-establishing stable state power and public order will also require finding a way to settle these conflicts. At present no single actor appears to know much about them; such knowledge is dispersed among many actors if it exists at all, and it would be worth an attempt to try to gather it on the district level, for instance.

The citizenry

In general, the concept of Afghan people as citizens of Afghanistan has been lost. This is the case de jure for women under Taliban rule, but it is true for anyone who is not commanding an armed group. Reconstruction of the state will have to include reconstructing a basis on which Afghans can participate in that state, which will require some form of legitimacy linked to the concept of an Afghan nation and gender and ethnic legal equality. A peace process must address each of these sets of issues: ending the civil war, ending the international conflict, and rebuilding the state and state-society relations. Up until now, the means deployed have been far less than what would be required. The intensity, depth, scope, and breadth of the conflict will require a commensurate peace process, involving the entire society of Afghanistan, neighbouring states and societies, and other international actors. The UN has an important political role to play, but the UN's other roles (humanitarian aid, development, human rights) are also involved. Other actors are also needed, including powerful states, donor states, regional states, international financial institutions, and NGOs of various types.

top of page