From the Archives: Post-Cold War State Disintegration: The Failure of International Conflict Resolution in Afghanistan

This article, by Barnett R. Rubin*, first appeared in the Journal of International Affairs, Vol. 66, No. 2 in the Winter of 1993. Rubin discusses the political destabilization in Afghanistan in the aftermath of the Cold War, as well as the role that the international community has, and should continue, to play in the region. 

Barnett R. Rubin
October 18, 2018

On the evening of 18 March 1992, President Najibullah of Afghanistan interrupted normal radio and television broadcasts with a dramatic speech. “I agree,” he announced, “that once an understanding is reached through the United Nations process for the establishment of an interim government in Kabul, all powers and executive authority will be transferred to the interim government as of the first day of the transition period.[1] By announcing his intention to step aside, the former secret police chief — who had headed the Soviet-backed Afghan regime since 1986 — seemed to clear the way for the implementation of a laboriously prepared international peace plan for resolving one of the last Cold War-era conflicts.

Less than a month later, however, President Najibullah was in hiding, under the unacknowledged protection of the U.N. office in Kabul. His Watan (Homeland) Party had split, with different factions allying with their former mujahidin (Islamic resistance fighter) opponents along ethnic lines; one coalition of government rebels and mujahidin had prevented him from leaving the country. The leading resistance commander and spokesman for the alliance that overthrew Najibullah, Ahmad Shah Massoud, told the United Nations that the mujahidin would form an interim government, which assumed authority on 29 April 1992 and established the Islamic State of Afghanistan.

Before that government arrived in Kabul, however, fighting had erupted in the streets of the capital as guerrillas belonging to rival parties, factions and ethnic groups battled for control of the city. By August 1992, according to the United Nations, food supplies in Kabul were scarce, shops were closed and over 500,000 people had fled the city in all directions.[2] By November, some observers estimated that fighting in the city had killed at least 4,000.[3] In the Afghan hinterland, tribal and ethnic coalitions took control of the major regional garrisons, with mujahidin and former government commanders forming autonomous power centers, paying little allegiance to the new interim government.

Afghanistan had rarely appeared on the world's television screens, but the disaster did not result from simple neglect. In the 1980s, the Soviet Union intervened militarily to prop up its Afghan government clients, whose efforts to impose a Marxist-Leninist political system on the country were failing. This intervention, coupled with the military response from Pakistan, the United States, Saudi Arabia and China fueled a war that killed over a million of the country's estimated 15.5 million people — and drove over 5 million into exile. U.N.-mediated talks, begun in 1981, slowly elaborated the framework for a Soviet withdrawal. After Soviet leader Mikhail Gorbachev's decision to seek a global rapprochement with the United States and its allies, this mechanism led to the April 1988 signing of the Geneva Accords, which provided for the withdrawal of Soviet troops from Afghanistan by the middle of February 1989.

The accords established monitoring and implementation roles for the United Nations, and involved it in the provision of humanitarian assistance for refugee repatriation and national reconstruction. Subsequently, U.N. Secretary-General Javier Pérez de Cuellar became deeply involved in efforts to find a negotiated solution to the domestic conflict, through the establishment of an internationally supported interim government that would hold elections — as the international community was also trying to do in Nicaragua, Angola, Namibia and Cambodia.

In some of these other countries, the U.N. Security Council provided armed forces or administrative assistance to oversee the disarmament of hostile forces, the administration of the transition and the conduct of elections. In Afghanistan, however, rather than invoke the powers of the Security Council, the U.N. General Assembly merely passed a resolution asking the Secretary-General to use his good offices to promote a negotiated solution.

The best efforts of Pérez de Cuellar and his successor, Boutros Boutros-Ghali, proved unable to halt further civil war. According to the U.N.’s interim government model, free and fair elections should replace the pursuit of politics through armed conflict. Yet in Afghanistan, the decline of communism and the overthrow of the Soviet-backed government did not leave “democracy unchallenged as the supreme principle of political legitimacy.”[4] Rather, Islam was the rallying cry of the Afghan mujahidin, and the Islamic prohibition on collaborating with so- called apostates — as some saw the communists — complicated the formation of an interim government. Disagreement among Islamic activists and scholars about who was qualified to select or serve in any Islamic government made it nearly impossible for any form of representation to legitimate such a government.

Further, democratization requires an effective law-bound state, which had not existed in Afghanistan since at least the beginning of the conflict. Where basic institutions such as a unified armed force, administration and legal system break down — as has happened in Afghanistan — no one can govern, democratically or otherwise. Yossi Shain and Juan Linz emphasize the importance of the “legality” of the state in establishing an interim government.[5] Despite a change in top authorities, the military, police, administration and courts must continue to carry out legally issued orders. Political stability requires not only the formation of a legitimate government through an electoral process, but also the construction of law-bound state institutions — a lengthier and more difficult effort.

Finally, while the end of global superpower rivalry removed the incentives that had led the United States and the Soviet Union to intervene in the Afghan conflict, it also reduced their incentive to devote scarce resources to seeking a peaceful resolution. This reality has constrained the effectiveness of U.N. mediation. While the Security Council's permanent members now reach agreement far more frequently, the United Nations still has little institutional autonomy from its member-states and is dependent on their material and political capital. The breakup of the Soviet Union, while undercutting the Najibullah government, also reduced the total supply of great-power leverage in the conflict. Regional powers like Pakistan and Iran, with their own strategic goals, were thus able to gain greater latitude to intervene.

State and Society in Afghanistan

Understanding the failure of conflict resolution efforts in Afghanistan requires a grasp of both the country's integration into the modern state system, and its domestic bases of political organization. The civil war in Afghanistan exploded into a conflagration when local conflicts, loosely related to ethnic divisions, became linked to the Cold War. Rival superpowers and their regional allies provided competing Afghan elites with access to ideologies, organizational models and financial and military resources. To guarantee this external support, Afghan political elites publicly adopted the ideological discourses favored by their respective patrons in Moscow, Washington, Islamabad and Riyadh — rather than the discourse of clan, tribe and ethnicity that actually dominated life in the mountain valleys and desert plains of Afghanistan. As international agreements decreased or ended the aid flows, these indigenous patterns of political action reemerged — although in new organizational forms.

Afghanistan entered the modern state system after the Second Anglo-Afghan War (1878 to 1880), as a buffer between the British and Russian empires, which drew the borders that persist to this day. Consolidation of a central state in this sparsely populated, largely tribal country depended on foreign aid, mainly from Great Britain.[6]

During the 1950s, Afghan Prime Minister Muhammad Daud, cousin of King Zahir Shah, played on the country’s renewed status as a buffer —now between the Soviet Union and the U.S.- sponsored Baghdad Pact — to build an expanded state apparatus with foreign aid from both the Cold War antagonists.[7] The largest donor, the neighboring Soviet Union, supplied and trained the military and sponsored some state economic projects, while the United States engaged in teacher training, agriculture and engineering.

Ethnic Groups of Afghanistan

The structure of the Afghan state imposed a pattern of ethnic stratification on the diverse local societies. The head of state (king until 1973, president from 1973 to 1978) was a member of one clan of the Durrani confederation, one of three major groups of Pashtun tribes. The state claimed to represent the national identity of Pashtuns, who constituted 40 to 45 percent of the population. The official religion of the state was Sunni Islam.

The royal clan topped the ethnic hierarchy. Below them came the other Pashtuns, with some preference for Durranis. The Shi’a — about 15 percent of the population with most belonging to the Hazara ethnic group — occupied the bottom of the social hierarchy. Between the Pashtuns and the Hazaras were the other pre- dominantly Sunni ethnic groups. Of these, the Tajiks, Persian-speakers of the Northeast, served as junior partners of the Pashtuns in ruling the country. Other groups of intermediate status included the Turkic speakers of the North, mostly Uzbeks. For further information on the ethnic distribution of Afghanistan, please refer to the map on the previous page.

Origins of Conflict

Between 1963 and 1973, Afghanistan was governed by a form of constitutional rule known as “New Democracy.” The government held two national elections for a consultative parliament. Although political parties were not permitted to compete in the elections, various factions of the intelligentsia began to organize politically. Few of these groups made procedural democracy or the rule of law a central concern; most were nationalist, Marxist or Islamist.

Until the 1970s, the United States and the Soviet Union mainly competed for influence over a regime they both supported, rather than supporting political factions. In 1973, however, Daud over threw Zahir Shah in a coup, abolished the monarchy and pro- claimed himself president. Both superpowers, as well as Pakistan, feared a future succession crisis. Under the Nixon doctrine, Washington encouraged the Shah of Iran to use his oil wealth to draw Afghanistan into an Iranian-dominated regional grouping.[8] The Brezhnev government, believing that the “correlation of forces" in the world was favorable to the expansion of Soviet influence in the Third World, expanded its support for the Afghan communists.[9]

The principal Soviet-backed communist organization in Afghanistan was the People’s Democratic Party of Afghanistan (PDPA), founded in 1965. In 1967, the party split into two factions: Khalq, led by Nur Muhammad Taraki and Hafizullah Amin, and Parcham, led by Babrak Karmal.[10] Parcham and Khalq constituted distinct political and social groups. Parcham recruited from the middle and upper ranks of the Persian-speaking urban elite; Pashtuns were a minority among them. Khalq, on the other hand, recruited from the newly educated of rural background, mainly tribal Pashtuns.

Also in the 1970s, Pakistan increased its support of Islamist groups.[11] In 1973, the Islamist movement at Kabul University organized as Jamiat-i Islami (Islamic Society), led by Burhanuddin Rabbani, a lecturer at the shari’a (Islamic law) faculty. That same year, Daud sought to suppress the movement, and its leaders fled to Peshawar, capital of the mainly Pashtun Northwest Frontier Province of Pakistan.

In exile, Jamiat split, into the Jamiat-i Islami, still led by Rabbani, and the Hizb-i Islami (Islamic Party), led by Gulbuddin Hikmatyar. Both groups recruited from the newly educated of rural background, as did Khalq. Jamiat’s leaders included Islamic scholars from different ethnic backgrounds, but its cadres largely consisted of Tajiks with secular or state religious educations from northeast Afghanistan. Hizb, the more radical of the two, mainly attracted detribalized Pashtuns with secular educations.

Neither the communist nor the Islamist factions were ethnically or linguistically homogeneous, and all denied that ethnicity played any role in their politics. Nonetheless, Pashto was the main language within Hizb and Khalq, while Persian was the main language within Parcham and Jamiat. Within each movement the rivalry between the two main factions developed into a bitter feud, with strong ethnic overtones, that ultimately overrode their ideological differences.

Other Afghan ethnic groups remained aloof from communist and Islamist factions. The Uzbeks participated little in national politics, while radical Hazara youth joined either Maoist or sepa- rate Shi’a Islamist organizations, most of which united into a single Iranian-sponsored party, the Hizb-i Wahdat-i Islami (Islamic Unity Party) in 1990.

In April 1978, the PDPA — which had been nominally reunited by the Soviets a year earlier — seized power in a military coup in which Prime Minister Daud was killed. The PDPA established the Democratic Republic of Afghanistan, which quickly became de- pendent on Soviet aid. Within a few months, the ruling Khalqis had begun a bloody campaign to eliminate any would-be political opponents, expelled the more moderate Parchamis and announced a revolutionary program — which they attempted to impose on the whole of Afghan society.

In response to the Khalqi program, revolts broke out in various parts of the country, usually without any link to national political groups. From their Pakistani exile, the Islamists declared a jihad (holy war) against the communists. They were soon joined by representatives of the conservative Muslim clergy and the elites of the Daud and Zahir Shah eras. The Afghan army and administration, weakened by mutinies and defections, seemed headed for collapse.

Intervention and Resistance

In December 1979, the Brezhnev politburo feared that the United States would exploit the disorder in Afghanistan — and recoup its losses in the previous February’s Iranian Revolution — by installing a pro-American government in Kabul with Pakistani assistance.[12] The Soviets sent a “limited contingent” of troops to seize the Afghan government from the unreliable and brutal Khalqi leader, Hafizullah Amin, who was killed during the take-over.[13] The Soviets then forced the Parcham and Khalq factions to reunite under Babrak Karmal, the leader of Parcham. Moscow sought to prop up its PDPA clients by trying to defeat the mujahidin militarily, increasing its presence to about105,000troops by 1981. For the next six years, Moscow’s and Kabul's forces carried out massive counterinsurgency operations in the Afghan co try- side, which included indiscriminate bombing of rural areas.[14]

The Soviets soon found that, far from stabilizing the situation, their troops provoked national resistance and worldwide condemnation. The United States, Saudi Arabia and China began programs of aid to the mujahidin through Pakistan, whose Directorate of Inter-Services Intelligence (ISI) allocated the aid and tried to direct the resistance movement.[15] In 1981 Pakistani authorities recognized seven mujahidin parties, including the Islamist Hizb and Jamiat. Henceforth all refugees and mujahidin commanders had to join one of these parties to receive assistance.[16] While the resistance harshly denounced the repressive and authoritarian policies of the PDPA regime, Islam — and not democracy — was the value system that mobilized their heterogeneous movement. Neither the PDPA nor the Afghan resistance was able to field a single conventionally controlled military force. Rather as the war progressed, state administration collapsed throughout Afghani- stan, and all segments of ssciety became increasingly well-armed. Local power devolved into the hands of mujahidin and govern- ment field commanders, who used aid from Pakistan or Kabul to build autonomous power bases — in some cases including large military organizations and civil administrations — grounded in ethnic or religious ties. These regional politico-military organizations became the primary vehicles for the bloody ethnic politics that emerged following the Soviet withdrawal  from  the  country in 1989.

In northeast Afghanistan, for example, the Tajik Jamiat commander Massoud established a base area under his “Supervisory Council of the North.” In the largely Uzbek areas of the North, government militia leader General Abdul Rashid Dostam used Soviet aid to build up the most powerful military force of the Kabul regime. The Hazara areas in central Afghanistan became virtually independent, and the Shi'a mujahidin there developed strong links to Iran. The heavily armed Pashtun tribes of the East were split between allegiance to various mujahidin parties and the government. In the areas of Pakistan inhabited by Afghan refugees, Hizb leader Hikmatyar used the aid he received from the ISI to recruit a conventional army from the refugee camps.

Military Stalemate and the Geneva Accords

To most international observers, the war in Afghanistan appeared to have reached a stalemate by the mid-1980s. Under Gorbachev, Moscow could no longer tolerate an open-ended commitment to a regime that could not sustain itself independently; by 1986 the Soviets were looking for a way to withdraw. First they tried domestic reform, replacing Babrak Karmal — a symbol of the Soviet invasion — with Najibullah, his rival within Parcham. This move created a split within Parcham between supporters of Karmal — mostly non-Pashtuns — and Najibullah, of Parcham’s Pashtun minority. Gorbachev pressed Najibullah and the PDPA to share power with local mujahidin leaders and field commanders under a program called “National Reconciliation.” The stigma attached to cooperating with the agents of the Soviet occupation, anti-Islamic forces and Najibullah, the hated former secret police chief, however, prevented all significant re- sistance leaders from accepting his offer. By late 1986, increased international aid to the mujahidin confronted Soviet military planners with a stark choice of withdrawing or facing an unending war.

Gorbachev then sought to use the United Nations to forge a negotiated settlement that would enable him to withdraw his troops, with some guarantees for the Afghan government. The U.N.-sponsored Geneva Talks among Washington, Moscow and Islamabad, that had dragged on fitfully for six years, were reinvigorated — and led to the Geneva Accords of 14 April 1988. The accords provided the legal framework under which the Soviets were able to withdraw their troops, who were gone by 15 February 1989.[17] Reflecting the political realities of the early 1980s, when the negotiating agenda had been set, these accords dealt only with “external” aspects of the conflict such as Soviet troop with- drawal, the end of international military aid to the muJahidin, refugee repatriation and international guarantees. They did not link the end of international involvement to a domestic political settlement based on the formation of an interim government or internationally monitored elections, as did later agreements in Cambodia, Nicaragua and Namibia.

Ultimately, the accords amounted to little more than a U.N. cover for the Soviet decision to withdraw from Afghanistan. Pakistan and the United States had originally agreed to link ending aid to the resistance to the Soviet withdrawal. Although this arrangement benefited Kabul by allowing Soviet forces to continue assisting the PDPA government during the withdrawal process, Washington and Islamabad then believed that this ar- rangement was the only way to get Moscow out of Afghanistan. By 1988, however, the United States realized that the Soviets would withdraw their troops regardless of continuing external assistance to the mu/añidin. Therefore, via a still-secret exchange of letters that the United States regards as part of the Geneva Accords, Washington reserved the right to support the mujahidin as long as the Soviets supported the Afghan government. Mos- cow and Kabul protested furiously, but in the end the withdrawal proceeded as required by the accords without Washington and Islamabad fulfilling their original obligation to end military aid to the resistance.

After the Soviet Withdrawal

During and after the Soviet withdrawal, Najibullah proceeded with various unilateral reforms designed at least to give the appearance of democratization. Both he and his Soviet supporters at times advocated what they called the Nicaraguan model, in which the government would organize internationally monitored elections open to all elements of the opposition. In December 1987, Najibullah summoned a Loya Jirga — the traditional Great Council of Afghanistan — to approve a new constitution providing for an elected government and to elect him as president with virtually unlimited powers. During parliamentary elections held in April 1988, immediately after the signing of the Geneva Ac- cords, Najibullah reserved seats in the national assembly for the resistance, invited seven field commanders of different mujahidin parties and ethnic groups to join his government, and unsuccess- fully tried to persuade Jamiat commander Massoud to become defense minister. The concentration of powers in the presidency, however, nullified any effect elections might have had.

Following the Soviet withdrawal, the U.S. Central Intelligence Agency and the Pakistani ISI predicted that the Afghan government would not long outlast the Soviet withdrawal. Rather than foster a political settlement with Kabul, Washington and Islam-abad concentrated their efforts on achieving the seemingly inevitable mujahidin military victory. At Washington and Islamabad's insistence, the seven recognized Sunni mujahidin parties chose an “Interim Islamic Government of Afghanistan” (IIGA) at a shura (council) held in Pakistan in February 1989 — as the last Soviet troops left Afghanistan. The IIGA, however, was a government in exile rather than an interim government. Moreover, it was composed almost entirely of exiled Eastern Pashtun leaders and had a narrow political base that undercut its credibility. There was no participation by officials of the former Daud and Zahir Shah governments who had fled to the West, the Shi’a parties or mujahidin commanders inside Afghanistan. The IIGA never succeeded in establishing itself in Afghanistan or organizing the elections it promised.[18]

Further, the mujahidin proved incapable of defeating Afghan government forces following the Soviet withdrawal. In March 1989, the Pakistani ISI organized a mujahidin offensive intended to install the IIGA in the eastern Afghan city of Jalalabad. The offensive failed because of a lack of effective coordination among the parties, as well as unexpectedly stiff government resistance.[19] Conflicts between mujahidin groups increased. In the summer of 1989, an attempt by Jamiat commander Massoud to capture the important northern center of Kunduz had to be called off when Hikmatyar’s Hizb-i Islami forces massacred 10 of Massoud's senior commanders.

Conflict between Hizb and Jamiat reflected a larger fragmentation of both government and resistance forces as a result of the military and political stalemate that emerged in the wake of the Soviet withdrawal. Kabul lost much of its ideological motivation in resisting the mujahidin, and Najibullah sought to redefine the PDPA, renouncing Marxism and embracing Islam, democracy and market economics — and changing its name to the Watan party. More importantly, while the continued bipolar flows of aid maintained a surface appearance of polarization, for many mujahidin the withdrawal of Soviet troops meant the end of the jihad. Instead of a struggle pitting Muslims against non-Muslim aggressors, a many-sided conflict over power emerged, structured along ethnic and tribal lines.

The U.S.-Soviet Dialogue of 1989-1990

Following the failure of the mujahidin to secure a military victory, the Bush administration began discussions with the Soviets on the framework for a settlement in Afghanistan. This dialogue was part of a larger rethinking of U.S. involvement in Afghanistan. With the Soviets gone, Washington began to be concerned about the possibility of a takeover by anti-American Islamic radicals in the muJa6idin. A secret decision in the fall of 1989 defined the goal of U.S. policy after the Soviet withdrawal as “sidelining extremists” on all sides, including both Najibullah and Hikmatyar. According to the new policy, no weapons paid for by U.S. funds would be given to Hikmatyar's Hizb group. This policy made little practical difference, however, since Saudi Arabia — an important source of the resistance’s financial sup- port — saw Hizb as the strongest Sunni counterweight to Iranian influence and made up the difference.

Discussions between Moscow and Washington complemented ongoing U.N. mediation activities. All parties to the Geneva agreement had agreed to grant Perez de Cuellar the authority to use his good offices to promote the formation of a “broad-based” government in Afghanistan. The accords had also established a U.N. monitoring group in Afghanistan and Pakistan whose political officer, Benon Sevan, shuttled constantly between Kabul, Islamabad and Tehran.

In the United Nations' view, the external dependence of all political forces in Afghanistan was so great that they would never reach agreement until their sponsors developed an “international consensus” on such issues as the nature of a transitional period and the cessation of military aid to Afghan parties. These issues were a key part of U.S.-Soviet discussions. In September 1989, both sides agreed that “a transition period is required, as well as an appropriate mechanism to establish a broad-based government.”[20] Talks continued throughout 1990, and were increasingly focused on the holding of elections as a means to resolve the conflict.[21]

The United States initially insisted that a prerequisite for any transition was the departure of Najibullah.[22] Further, Washington argued that to assure that the government could not use its control of the security forces and mass media to affect the out- come of an election, the transition mechanism should be an interim government with full powers. For their part, the Soviets insisted that the transition mechanism should be either an election organized by the Najibullah government (the Nicaraguan model) or an independent commission that would organize the election while the government remained in power. Negotiations between these two positions focused on the precise form of power sharing.[23]

Moscow insisted on both a cease-fire —which Washington could not deliver — and a simultaneous end to military aid by all parties, including Pakistan and Saudi Arabia. By Spring 1990, Washington came back to supporting the concept of “negative symmetry,” by which the United States and the Soviet Union would simultaneously terminate military aid to all parties in Afghanistan, and to which Islamabad and Riyadh could later accede. Washington began to insist that Moscow set a “date certain” for negative symmetry.

The United States and the Soviet Union appeared to reach an agreement by 11 December 1990, when Soviet foreign minister Shevardnadze met U.S. secretary of state Baker in Houston. Both sides agreed to support the establishment of a U.N.-sponsored transitional organ that would replace the current government, to end all weapons supplies and to leave the precise structure of the transition to U.N. consultations with the Afghan parties. At the last minute, however, Shevardnadze’s efforts were blocked by hard-line officials in the Soviet government, and he was forced to refuse to agree to a “date certain” for negative symmetry.[24] Upon his return to Moscow he resigned, warning the Soviet public against these “reactionaries.”

Effects of the Gulf War

The rising power of the Soviet hard-liners in early 1991stymied further progress in the U.S.-Soviet dialogue, but the Gulf War helped settle some of the outstanding issues. Both Pakistan and Saudi Arabia began to look more favorably on a political settlement as those opposed to it were discredited. The radical elements of the Afghan mujahidin — in particular Hizb, and its Pakistani supporters — joined the international Islamist opposition to the U.S.-led coalition. Pakistan's conservative civilian leadership, along with the nationalists and moderates among the mujahidin, supported the U.S.-Saudi position. As a result, the Saudis at least temporarily suspended funding to Hizb.

Immediately after the Gulf War, at the end of March 1991, the Pakistani ISI threw all its resources — including many Pakistani advisers on the ground — into the battle for the Afghan garrison town of Khost. The garrison fell, but the cost in Pakistani effort and the inability of the Pashtun tribal mujahidin to establish a government even in this small town — instead they pillaged it — only reinforced the failure of the military option. The Pakistani president and prime minister decided that it was time to promote a political settlement; the Saudis agreed.

The U.N.-Mediated Settlement

In May 1991, as a result of the U.S.-Soviet dialogue in 1990 and the change in the positions of Pakistan and Saudi Arabia after the Gulf War, Perez de Cuellar was able to issue a statement summarizing an “international consensus” on Afghanistan in five general points. The political settlement would begin with the establishment of a “transition mechanism” in Afghanistan. In conjunction with the beginning of the transition, all external par- ties would stop supplying weapons to Afghanistan, and all internal parties should adopt a cease-fire. The interim authority would organize “free and fair elections, in accordance with Afghan traditions,” to choose a “broad-based government.[25]

The remaining disagreements between Washington and Moscow were resolved after the aborted August 1991 coup against President Gorbachev. On 13 September 1991, the new Soviet foreign minister Boris Pankin and Secretary Baker agreed to a “date certain” for negative symmetry. Moscow agreed to Washington’s proposal that the transitional authority should re- place rather than operate alongside the government of President Najibullah. Both sides would end all weapons supplies at the end of the year; Pakistan and Saudi Arabia quietly indicated they would comply. Soon after, Najibullah told the United Nations he would agree to leave at the beginning of the transitional period.[26]

This consensus, however, included only a limited-liability involvement by the international community in Afghanistan. The United States and the Soviet Union agreed to try to remove sophisticated weapons and weapons of mass destruction from Afghanistan, but they made no provision for disarming or merging the various armed forces in the country.[27] Given the high militarization of the Afghan countryside and the lack of political order, no party gave serious consideration to the United Nations administering the transition itself, or disarming the opposing sides. Privately, U.N. officials also said they did not want to try to monitor elections in Afghanistan. Rather, the plan relied on the belief that the various Afghan parties and armed forces would all have to look to the internationally sponsored interim government for patronage once external assistance had ended.

Installing the Transition Mechanism

By the end of November 1991, Pakistan, Saudi Arabia and Iran indicated their support for the U.S.-Soviet agreement. In January 1992, the new Secretary-General, Boutros Boutros-Ghali, announced a plan under which all Afghan parties would submit to his office lists of candidates for an “Afghan gathering,” intended to lead to a nationwide meeting to decide on an interim government and the holding of elections. This unique and cumbersome procedure was necessary because the mujahidin groups still re- fused to meet or cooperate openly with Najibullah, his Watan Party or the existing government. Further, any Afghan who tried to convene such a gathering, however prominent or neutral, would be suspected of promoting his own power; any non-Afghan (including the Secretary-General) would lack legitimacy.

The three traditionalist-nationalist mujahidin parties submitted a joint list of proposed participants in the gathering, as did the Iran-based Hizb-i Wahdat. Former President Zahir Shah and his advisers, however, refused to submit a list of representatives to a gathering that gave such weight to the resistance parties. The Sunni Islamist parties also refused, suspecting that the U.N. plan was an effort to prevent an Islamist victory. Najibullah also postponed giving his list, which would have been a decisive signal that he was prepared to depart.

By March, Islamabad and Washington had informed Sevan that they needed Najibullah’s explicit and public commitment to depart, in order to pressure the “rejectionist” mujahidin groups that had not named representatives. After long sessions in Kabul with Sevan, Najibullah presented his list of participants and, on 18 March 1992, publicly stated his intent to leave office.

From Conflict Resolution to Ethnic Conflict

The U.N. plan aimed at finding a solution to a stalemate be- tween ideologically opposed parties; only when no side can win will they agree to a negotiated solution. Ending military assistance to the mujahidin and Kabul was supposed to deprive them of the means and motive to continue the conflict; the U.N. plan was an opportunity to resolve it.

With the end of aid from Washington and Moscow, and Najibullah's promise to depart, however, both the stalemate and the ideological conflict were definitely over. The termination of aid redefined rather than ended the conflict. The Afghan government and the resistance alliance, whose leaders had been the United Nations' main interlocutors, fragmented along largely ethnic lines; now, power was in the hands of the local commanders with their autonomous military base areas around the country. The persistence of the Soviet-supported state apparatus centered in Kabul had concealed a profound realignment of demographic and political power within the country. Although Najibullah and his closest collaborators were Pashtuns — like the leaders of previous governments and dynasties — the war had weakened the Pashtuns; they were the great majority of the five to six million Afghan refugees in Pakistan and Iran. By 1992, Persian-speakers were probably the largest group within Afghanistan and the largest mujahidin force was Jamiat’s, mainly Tajik, Supervisory Council of the North, led by Massoud.

The war had also empowered new groups. The Uzbeks had benefited from heavy Soviet investment in the development of non-Pashtun northern Afghanistan, and — to counter his enemies in Khalq — Najibullah had built up General Dostam's Uzbek militia to be the most powerful military unit in the country. The Shi'a Hazaras, who had been conquered and enslaved a century before, now had their mujahidin alliance, Hizb-i Wahdat, and enjoyed the support of a resurgent Iran. In the battles for power that emerged with the end of foreign military aid, the non-Pashtuns asserted their new power, which Pashtun mujahidin and government forces refused to accept.

The new ethnic alliances had been in the making for some time. All leaders of the Watan party and the military in Kabul had been seeking to assure their own survival by finding allies among the mujahidin. In March 1990, a Pashtun alliance — backed by the Pakistani ISI, which still sought a military victory for its Hizb clients — formed when Defense Minister Shahnawaz Tanai and other Khalqi officers allied with Hikmatyar and nearly overthrew Najibullah; ISI and Saudi intelligence pressured the other mujahidin leaders to join. With U.S. support, they resisted.

The anti-Najibullah Parchamis in the Afghan government, mainly non-Pashtuns, had also begun to negotiate with Shi'a, Tajik and Uzbek mujahidin. Many Afghans believe that the Iranians were promoting an alliance among them. If so, Tehran was likely seeking to balance Saudi and Pakistani influence among the Pashtun Sunni mujahidin.[28] In 1991, Tehran had sponsored an accord on cultural cooperation among Persian speakers, signed by Iran, Tajikistan, Hizb-i Wahdat and Jamiat. This accord signaled Iran's intention to form a special relation with the north- ern, non-Pashtun part of Afghanistan, which constitutes a land bridge between Iran and the newly assertive republics of Uzbekistan and Tajikistan.

Najibullah faced open revolt from his own non-Pashtun forces in northern Afghanistan in January 1992, shortly after the end of both Soviet aid and the Soviet Union itself. The rebels, led by Uzbek general Dostam, enjoyed the support of Parchami elements within the armed forces and the Watan party and formed an alliance with commanders of the Hizb-i Wahdat and Jamiat. The day after Najibullah announced his departure, this coalition seized control of the northern capital of Mazar-i Sharif, which gave them control of all military forces in northern Afghanistan, from the former Soviet border to the gates of Kabul. Jamiat's Massoud emerged as political spokesman for the alliance.

The U.N. Plan Blocked

As the anti-Pashtun revolt threatened to topple the government and preempt the lengthy transition process, the United Nations streamlined its mechanism for setting up an Afghan meeting to choose an interim government. On 10 April, Sevan met in Geneva with Boutros-Ghali, who approved a new version. A “pre-transition council composed of impartial personalities” chosen from the lists submitted to the United Nations in early 1992 would take over “all powers and executive authority” from the current government.[29] This council would then convene a shura to choose the interim government.

The United Nations hurriedly canvassed support for the plan, and under severe pressure from Washington and Islamabad, the major leaders on all sides accepted it. Those members of the proposed interim authority who were outside Afghanistan assembled in Peshawar, Pakistan. Both the resistance and the Afghan government agreed that on the night of 15 April 1992, a U.N. plane would fly the members of the interim government into Kabul, where Najibullah would transfer power to them at the airport and leave on the same plane.

During the day of 15 April, however, two of the mujahidin parties expressed misgivings about the pre-transition council. Jamiat leader Rabbani, whose forces seemed on the verge of taking Kabul with the support of Dostam's fighters, wavered. Another mujahidin official suggested forming their own government instead. The United Nations asked them to submit a list of members of such a government, but the mujahidin leaders were unable to agree on a list that day. Sevan flew to Kabul in the early hours of 16 April.

There, however, the Dostam-Jamiat coalition prevented Najibullah from escaping, seizing control of the airport with 750 to 1,000 troops flown from northern military bases.[30] Najibullah fled to the U.N. office, which found itself in the awkward situation of granting asylum in his own country to a deposed head of state accused of serious human rights violations.

The Islamic State of Afghanistan

The Parchami rebels who now controlled Kabul denounced Najibullah as a hated dictator and secretly asked Massoud to enter the capital as head of state; the Parchamis reportedly in- tended to use Massoud as a figurehead who would continue to depend on them.[31] Massoud rejected this offer, however, and asked the mujahidin leaders in Peshawar to accelerate their efforts to form an interim government. Despite intense Pakistani pressure, the leaders in Peshawar argued for 10 days over arrangements for a transitional government.

Meanwhile, Pashtuns in the Afghan military reacted to the threatened takeover of Kabul by the Massoud-Dostam coalition. These officers infiltrated unarmed fighters of Hikmatyar’s Hizb-i Islami into the city, where they received arms from their co-ethnics in the interior ministry. Hizb's conventional military force also crossed over the border from Pakistan and moved within artillery range of Kabul. Despite Islamabad's official neutrality and support for a U.N.-sanctioned interim government, Hikmatyar was able to continue to recruit fighters and transfer weapons from Pakistan with the assistance of the Pakistani ISI — which had become a virtual law unto itself.[32] Some Afghans speculate that Pakistan and Saudi Arabia wanted to strengthen Hikmatyar to counter Iranian influence in the Massoud-Dostam coalition.

On 26 April, before the Peshawar leaders had reached an agreement, Massoud’s and Dostam’s forces — already in control of the airport — entered the rest of the city to preempt a Hizb takeover. Independently, the non-Pashtun Parchamis — assisted by the Iranian embassy — had also armed Kabul-based Shi’a mujahidin.[33] After violent battles at the interior ministry and the presidential palace, Massoud and Dostam's forces — together with these groups — expelled the Khalqi-Hizbi Pashtun forces. Other mujahidin groups were able to enter the city, however, and engage in looting.

That same day, mujahidin leaders in Peshawar finally reached agreement on a transitional arrangement, known as the Peshawar Accords, which claimed to establish a provisional revolutionary government composed of mujahidin. The interim government arrived in Kabul from Peshawar on 29 April. Massoud became minister of defense. For two months, Sibghatullah Mujaddidi — a moderate but weak mujahidin leader — would be acting president, to be followed by Rabbani for four months. After the six-month interim period, the government was to hold a shura to choose an interim government for 18 months, at the end of which elections would be held. The acting president would answer to a “leadership council,” composed of the leaders of the mujahidin parties.

Fragmentation and New Civil Wars

Massoud and Dostam’s seizure of Kabul did not arrest the disintegration of the Afghan state. From May through August 1992, Hikmatyar's forces rocketed Kabul, killing thousands and destroying both the city’s electrical grid and the water system, which shut down the hospitals; the United Nations and many diplomatic missions withdrew personnel. The long-standing rivalry between Hizb and Jamiat — now overlain with the feud between Khalq and Parcham — took on the dimensions of a battle for the control of Afghanistan between Pashtuns and non- Pashtuns, or even between Iran and Hizb's Pakistani and Saudi backers. By the end of August, however, a cease-fire was brokered — although Massoud would have preferred to continue the offensive and use the Afghan air force to wipe out Hikmatyar's forces.

As the embattled government fought for control of its own capital, the rest of the country remained divided, without any central administration. Throughout the Afghan interior, mujahidin negotiated the surrender of local government garrisons. Regional shares, some including commanders formerly on opposite sides, formed along ethnic and tribal lines. Tehran sought diplomatic advantage from the breakdown of the Afghan state and, while continuing to support the Shi'a Hizb-i Wahdat, also developed direct relations with the Jamiat shura in western Herat, and Dostam's administration in the north.

Other elements of the Afghan population, particularly the Pashtuns in eastern and southwestern Afghanistan, maneuvered between the two power centers, more interested in local autonomy and opium cultivation than the fate of the would-be central government in Kabul. No one had any reason to secede from such an impotent state, and all the regional states — most aggressively Pakistan — let the various Afghan commanders know that they would not countenance any formal division of Afghanistan. In fact, however, the regionally based ethnic coalitions became autonomous, with their own armed forces and sources of revenue. All major customs posts, which were formerly the government's principal source of revenue, were under the control of regional shuras who kept the funds for themselves.

Six months after assuming authority, the would-be interim government in Kabul — whose control over the city was still tenuous — had made little progress in building a national consensus. In late October, the interim government’s leadership council voted to extend Rabbani's mandate for 45 days. After that period, the government summoned a nationwide shura to elect a president for the next 18 months. Somewhat belatedly, Rabbani managed to convoke a shura of 1,335 men on 29 December; over a tenth of the members were drawn from Dostam’s Uzbeks and the rest were mostly Jamiat. Most areas of the country had some representation, but the other mujahidin parties generally boycotted the shura, charging that it was manipulated by the victor, Rabbani. Fighting soon resumed both within and outside Kabul.

Preliminary Conclusion: Interim Governments Without A State

The failure of the U.N. plan has attracted considerable international attention because of the way in which it approached success and then dramatically collapsed. Virtually every conceivable form of interim government has been proposed in Afghanistan, yet none has reestablished political order — to say nothing of democracy. The Jamiat-dominated interim government formed under the Peshawar agreement may ultimately evolve into a forum within which new power-sharing arrangements are worked out among the various groups, or it may give way to another equally shaky arrangement. The process will likely involve a mixture of warfare and negotiation; ballots alone will not transfer power to anyone in Afghanistan for a long time.

While the model of an international interim government has many attractive features, the case of Afghanistan suggests that its requirements for success — viable state structures, a consensus on the nature of the political community and some degree of social order — will rarely be present in conflicts in the less developed world. Many in Afghanistan found the U.N. plan mystifying or, in the words of a former PDPA Central Committee member, “a joke.” He asked, “How could these people, outsiders, rule Afghanistan, when they did not command any military forces?”[34] For Afghanistan — as elsewhere — the end of the Cold War meant an unraveling of a state apparatus that was previously supported by foreign aid. When such state institutions collapse, and armed factions emerge as the main form of collective action, interim governments offer no quick solution to the problem of political order.

The high level of violence seen in contemporary Afghan society is not simply a product of local society or culture. Rather, it results from the billions of dollars of weapons poured into the country by outside powers. Afghan tribal societies had a tradition of using violence in social conflict, but they also had traditions to resolve conflicts and limit violence. Yet such norms of conflict resolution are of little use when the instruments of violence include armed forces driven by ideological or ethnic agendas, equipped with SCUD surface-to-surface missiles and armored helicopters.

Inter-ethnic conflict as seen in Afghanistan does not derive from immemorial hatreds, but rather from power struggles, which are worsened by the international system's definition of states as political communities with fixed borders. Afghanistan's current borders derive from the political requirements of Britain and Russia at the end of the nineteenth century, defined in relation to the balance of power that existed in Afghanistan at that time. A Pashtun dynasty had asserted loose control over most of the areas between the British and Russian empires; foreign aid consolidated that control by enabling the dynasty and its successors to conquer and rule the peoples within the boundaries.

Yet the breakdown of the state and the arming of virtually the entire population of Afghanistan has led to a distribution of power radically different from that which prevailed when the borders were drawn. Before the advent of the contemporary state system, the Uzbek, Pashtun, Tajik and Hazara military leaders might have established new “emirates” or “khanates” in their own territories. The international recognition of Afghanistan, however, virtually obligates these forces to battle for control of Kabul. The international system is thus deeply implicated, and will be deeply affected by these continuing struggles. As yet, however, the present system does not appear to have the means to resolve them.


[1] Edward A. Gargan, “Afghan President Agrees to Step Down,” New York Times, 19 March 1992, p. A3.

[2] United Nations Office for the Coordination of Humanitarian and Economic Assistance Programmes relating to Afghanistan (UNOCA), Press Release, "Immediate Humanitarian Needs in Afghanistan Resulting from the Current Hostilities," 23 August 1992; United Nations Department of Humanitarian Affairs, "Note on Winter Emergency Needs in Afghanistan," 1 November 1992, p.2

[3] Patrice Piquard, "Pourquoi le chaos afghan peut faire exploser l'Asie central," L'événement du jeudi, 7-13 January 1993, p. 48.

[4] Yossi Shain and Juan Linz, "The Role of Interim Governments," Journal of Democracy, 3, no. 1 (January 1992) p.75. See also Francis Fukuyama, The End of History and the Last Man (New York: Free Press, 1992).

[5] Shain and Linz, p.74.

[6] For overviews of state formation in Afghanistan, see M. Nazif Shahran, “State Building and Social Fragmentation in Afghanistan: A Historical Perspective,” in Ali Banuazizi and Myron Weiner, eds., The State, Religion, and Ethnic Politics: Afghanistan, Iran, and Pakistan (Syracuse, NY: Syracuse University Press, 1986) pp. 23-74; Ashraf Ghani, "Afghanistan XI Administration," in Ehsan Yarshater, ed., Encyclopedia Iranica, 1 (London: Routledge and Kegan Paul, 1982) pp. 558-64; and Barnett R. Rubin, "Lineages of the State in Afghanistan," Asian Survey, 28, no. 11 (November 1988) pp.1188-1209.

[7] The Baghdad Pact also included the United States, the United Kingdom, Iran, Turkey, Iraq and Pakistan.

[8] Selig Harrison, “Dateline Afghanistan: Exit through Finland?” Foreign Policy, no.41 (Winter 1980-81) pp.163-87.

[9] On these groups see Olivier Roy, Islam and Resistance in Afghanistan (Cambridge, UK: Cambridge University Press, 1986); Anthony Arnold, Afghanistan’s Two-Party Communism: Parcham and Khalq (Stanford, CA: Hoover Institution, 1983); and Barnett R. Rubin, “Political Elites in Afghanistan: Rentier State Building, Rentier State Wrecking,” International Journal of Middle East Studies, 24, no.1 (February 1992) pp.77-99.

[10] Khalq (the masses) and Parcham (the flag) were the names of the factions’ newspapers.

[11] I follow Roy in using the term Islamist for those groups often described as fundamentalists. These Islamic revolutionaries treat Islam as the political ideology of a movement to seize state power and reshape society. Religious leaders who preach a return to the shari'a and religious observance would more properly be called "fundamentalists."

[12] Interviews with officials of the United Nations, Iran and the former Soviet Union.

[13] See the account in Raja Anwar, The Tragedy of Afghanistan: A First-hand Account (London: Verso, 1988).

[14] Jen Laber and Barnett R. Rubin, "A Nation is Dying": Afghanistan Under the Soviets, 1979-1988 (Evanston, IL: Northwestern University Press, 1988).

[15] Brigadier Mohammad Yousaf and Major Mark Adkin, The Bear Trap: Afghanistan's Untold Story (London: Leo Cooper, 1992).

[16] The traditionalist/nationalist mujahidin groups included: the National Islamic Front of Afghanistan, the Afghanistan National Liberation Front and the Movement of the Islamic Revolution/Uprising of Afghanistan. The other two Islamist groups were the Islamic Party of Afghanistan (Khalis faction), and the Saudi-backed Islamic Union for the Liberation of Afghanistan.

[17] On the negotiations, see Riaz M. Khan, Untying the Afghan Knot: Negotiating Soviet Withdrawal (Durham, NC: Duke University Press, 1991); Selig S. Harrison, "Inside the Afghan Talks," Foreign Policy, 72 (Fall 1988) pp. 31-60; and Barnett R. Rubin, "Afghanistan: The Next Round," Orbis, 33, no. 1 (Winter 1989) pp.57-72.

[18] Barnett R. Rubin, "Afghanistan: Political Exiles in Search of a State," Journal of Political Science, 18 (Spring 1990) pp. 63-93.

[19] For an analysis of the defeat of the mujahidin at Jalalabad see Barnett R. Rubin, "The Fragmentation of Afghanistan," Foreign Affairs, 68, no. 5 (Winter 1989-90) pp. 150-68.

[20] Communiqué of U.S. Secretary of State James Baker and Soviet Foreign Minister Eduard Shevardnadze, Jackson Hole, WY, September 1989.  

[21] Speech by Under-Secretary of State Robert Kimmitt, Asia Society (Washington, DC: 18 April 1990).

[22] As Kimmitt stated: "A stable political settlement is not achievable so long as the Najib regime remains in power. This is not a U.S. demand; it is a statement of Afghan reality." ibid., p. 8.

[23] See the rather circumspect discussion of this by Baker and Shevardnadze at their press conference in Irkutsk, 2 August 1990.

[24] At his joint press conference with Secretary Baker on 11 December 1990, Shevardnadze stated that the two sides had agreed on the idea of free elections and the formation of a transitional body that would supervise those free elections. We also support a cease-fire and a cutoff in arms deliveries. As to the exact date, that is still a subject of further consideration."

[25] "Statement by Secretary-General Javier Pérez de Cuellar," U.N. Department Public of Information, 21 May 1991.

[26] Pérez de Cuellar stated in his October 1991 report on the situation in Afghani have been given assurances that some of the controversial personalities co would not insist on their personal participation, either in the intra-Afghan or in the transition mechanism." The Situation in Afghanistan and its Implications for International Peace and Security: Report of the Secretary-General, U.N. General Assembly and Security Council, U.N. Doc. A/46/577, S/23146 (New York: United Nations, 12 October 1992) p.12.

[27] In December 1988 Gorbachev had proposed a U.N. peacekeeping force as part of a settlement, but neither Washington nor the U.N. Secretariat ever seriously considered such a proposal, which was regarded by the mujahidin as an attempt to protect the Kabul regime.

[28] Interview with Afghan diplomat, New York, 23 November 1992.

[29] “Statement by the Secretary-General on Afghanistan,” 10 April 1992.

[30] Interviews with U.N. and Afghan diplomats.

[31] Interviews with Parchami former Central Committee members exiled in the United States and Central Asia, New York, 23 April 1992; Alma-Ata, 8 October 1992.

[32] These reports come from interviews with Afghans and Pakistanis who were in Afghanistan at the time. While I do not have conclusive evidence to verify their accounts, these reports are consistent with patterns that have developed over the years. Omitting such uncertain reporting would, in my judgment, constitute a greater distortion of the truth.

[33] Interview with exiled Parchami Central Committee member, Alma-Ata, 8 October 1992.

[34] Interviewed in Alma-Ata, 8 October 1992.

*I would like to acknowledge my debt to officials of the United Nations, the United States, the Soviet Union and Russia, Iran, and various parts of the government of Afghanistan, without whose assistance I would not have been able to prepare this paper. Promises of confidentiality prevent me from citing these sources for many of the statements in this article.