Civil War: Angola (1992-2002)

Beth K Dougherty. Civil Wars of the World: Major Conflicts since World War II. Editor: Karl DeRouen Jr. & Uk Heo. Volume 1, ABC-CLIO, 2007.

Country Background

For most of its independent existence, Angola was a Marxist-Leninist state ruled by a single party, the Popular Movement for the Liberation of Angola (MPLA). Freedom House gives it a rating of 6.0 (5.5-7 is “not free”). The MPLA government renounced Marxism-Leninism in 1990 and officially legalized other parties in May 1991 as part of the Bicesse Accords. There are currently more than one hundred twenty legal parties, with twelve controlling seats in the National Assembly. Nonetheless, the MPLA remains the dominant force in Angolan politics: It controls nearly 60 percent of the seats in the Assembly, and its long-time leader, Jose Eduardo dos Santos, has served as president since 1979. The last elections were held in 1992, and new elections have been promised for September 2006.

Despite a rich natural resource base and strong economic growth rates, Angola remains one of the world’s poorest countries. It is the second-largest oil producer in sub-Saharan Africa after Nigeria, and the world’s fourth-largest producer of rough diamonds (IMF 2005a). A net exporter of food at independence in 1975, decades of civil war have left most Angolans reliant on subsistence agriculture and humanitarian assistance for their food needs.

Angola’s economy is heavily reliant on the oil sector: Oil accounts for 90 percent of exports, 75 percent of revenues, and half of the GDP, and it is a significant source of collateral for commercial loans. Oil revenues brought in $4.5 billion in 2004 (IMF 2005a). There is virtually no integration between the oil sector and the rest of the economy, and most fields are located offshore, which protected oil production from disruption during the civil war. Angola possesses 5 billion barrels of proven reserves, and production has been steadily climbing over the last decade. In 1995, Angola produced about 650,000 barrels per day; by 2004, production had reached 1 million barrels per day and was expected to double by mid-2007 (IMF 2005a). This helped fuel an 11.4 percent real GDP growth rate in 2004. According to the U.S. State Department, Angola’s GDP/purchasing power parity is $35.1 billion, and per capita GDP/ppp is US $2,525 (United States Department of State 2005).

Transparency International ranks Angola as one of the world’s most corrupt countries. Based on International Monetary Fund (IMF) studies, some $1 billion in oil revenues in 2002 went unaccounted for. Between 1997 and 2002, a total of $4.22 billion in oil revenue appears to have been siphoned off, an amount roughly equivalent to all social spending (both public and private) in the same period (Human Rights Watch 2003a). Lack of transparency and mismanagement continue to plague the economy.

Diamonds are another significant economic resource, accounting for more than 8 percent of exports and 95 percent of non-oil exports (IMF 2005a). But the Angolan government has had trouble controlling diamond exports, and especially in the 1990s diamonds were overwhelmingly exploited by the rebel movement, the National Union for the Total Independence of Angola (UNITA) in a lucrative illegal trade. The government has steadily reasserted its control since 2002, and Angola will produce an estimated $900 million in diamonds in 2005 (United Nations Integrated Regional Information Network, 5 July 2005). Nonetheless, Angola continues to lose several hundred million dollars a year in revenue because of diamond smuggling (United Nations Integrated Regional Information Network, 2005a).

Conflict Background

With few exceptions, Angola was in a constant state of war from 1961 to 2002. From 1961 to 1975, several factions fought an anticolonial war against the Portuguese occupiers, which then continued as internecine civil war after independence in November 1975. The civil war has three distinct phases. Phase I (1975-1991) was a classic Cold War regional conflict, pitting the MPLA government, backed by 50,000 Cuban troops and massive levels of Soviet bloc military assistance, against its bitter rival, UNITA, backed by South Africa, Zaire, and the United States. Phase II (1992-1994) is bookended by the collapse of one peace accord, the Bicesse Accords, and the implementation of another, the Lusaka Protocol, which kept a relative peace for four years. Phase III (1998-2002) began with the total collapse of the Lusaka Protocol and ended with the death of rebel leader Jonas Savimbi. This entry focuses on the 1992-2002 period.

The first armed uprisings broke out in 1961, spearheaded by peasants resisting the forced cultivation of cotton and, in a separate action, the beginning of the MPLA’s campaign. The brutal Portuguese counterattack killed tens of thousands of Africans and sent hundreds of thousands into Zaire as refugees; by the time the peasant uprising had been put down, Portugal had 140,000 troops stationed in Angola. Sustained anticolonial guerrilla operations commenced in 1966. Three factions emerged: the MPLA, the National Front for the Liberation of Angola (FNLA), and UNITA. In 1974, following an army coup triggered by the draining anticolonial wars in Africa, Portugal announced its intentions to grant Angola independence on November 11, 1975. A brief period of relative calm was destroyed by early summer 1975 as the three anticolonial movements fought for control of the soon-to-be independent state.

The heavy involvement of the United States and the Soviet Union as the sponsors of opposing movements helped turn Angola into a major Cold War battleground. Cuban troops began arriving in late summer in support of the MPLA, and South African and Zairean troops intervened on behalf of the FNLA and UNITA. The Cuban intervention proved decisive, and the MPLA routed the FNLA, which all but disintegrated as it retreated into Zaire. The MPLA declared itself the Government of Angola (GOA) at independence on November 11, 1975. UNITA, the smallest of the three movements, refused to accept the legitimacy of the MPLA government. It continued a guerrilla campaign in the southern areas, backed by Zaire and South Africa. American assistance to UNITA was legally cut off by the 1976 Clark Amendment and would not be restored until 1985. In the early 1980s, the Reagan administration tied the independence of Nambia, illegally occupied by South Africa, to the withdrawal of the 50,000 Cuban troops in Angola. Fighting between the Angolan government and UNITA dragged on for thirteen years after independence.

As the United States and the Soviet Union began to end their Cold War, and as apartheid South Africa moved toward democracy, diplomatic efforts mounted to solve the international element of the Angolan war. Soviet leader Mikhail Gorbachev wanted to extricate the Soviet Union from its costly third-world involvements, Cuba could not afford to maintain its forces in Angola without Soviet financial assistance, and South Africa was staggering under the combined weight of sanctions, the war in Angola, and colonial control of Namibia. Heavy fighting in 1987-1988 produced a mutually hurting stalemate as all parties recognized that military victory was not possible and that further fighting was economically prohibitive.

On December 22, 1988, the Brazzaville Protocol was signed at the UN by Angola, Cuba, and South Africa. It featured a phased withdrawal of Cuban troops, to be completed by July 1991, and the implementation of UN Resolution 435, which governed Namibia’s transition to independence. A seventy-person UN military observer force (United Nations Angola Verification Mission [UNAVEM I]), verified the Cuban withdrawal. The diplomatic momentum carried over into an effort to solve the internal dimension of Angola’s civil war as the Americans and the Soviets both pressured their clients to settle politically. The GOA announced its commitment to a political solution in January 1989, and over the next two years a variety of mediators brokered negotiations.

Joint American, Soviet, and Portuguese mediation over six rounds of talks in 1990 produced the 1991 Bicesse Accords. The agreement’s provisions included a cease-fire, effective May 15, to be followed by cantonment and disarmament; a multiparty system, which allowed for UNITA’s participation in politics; a commitment to hold free and fair elections under international supervision; the recognition of the MPLA government by UNITA until the elections; the extension of central administration to all parts of the country; the termination of outside military assistance by the United States and the Soviet Union; the creation of a unified national military force; and a UN peacekeeping mission (UNAVEM II) to support the parties. UNAVEM II had a very weak mandate—its role was simply to verify that the two parties were following the cease-fire, and thus it deployed only 350 military observers. The Joint Political and Military Commission, the body charged with the actual implementation of the accord, comprised the two parties plus the United States and the Soviet Union; the UN’s special representative to the secretary-general was an observer.

The elections were held in September 1992 despite clear evidence that both parties had failed to demobilize or disarm. Three weeks before the elections, less than half of the government’s forces had demobilized, and not even 25 percent of UNITA’s forces had done so (Jett 1999, 104). Nonetheless, the elections went forward, and 92 percent of registered voters participated. The MPLA won nearly 54 percent of the vote for the national assembly (129 seats) and UNITA 34 percent (seventy seats). President dos Santos received 49.57 percent of the vote for president and Savimbi 40.6 percent. The failure of any candidate to achieve a majority should have triggered a second round of voting, but Savimbi rejected the results of the elections.

Savimbi’s decision plunged the country into the worst violence yet seen in the long civil war. An estimated 300,000 people died in the 1992-1994 period (Human Rights Watch 1999, 15). UNITA was well positioned for a return to war. Its forces were now three times larger than those of the Forcas Armadas Angolanas (FAA), thanks to the government’s crash effort to demobilize its forces in the weeks surrounding the election (Jett 1999, 104). Unsurprisingly, UNITA rapidly seized over three-quarters of the country, including several provincial capitals, before the FAA began to reverse those gains. Several factors contributed to the FAA’s gradual resurgence. In 1993, the GOA hired Executive Outcomes, a private security firm whose personnel were former members of the South African Defense Forces (SADF), to train and assist the FAA. UNITA lost its outside supporters, with the exception of Zaire, because it had rejected the election results. It was forced to turn to diamond smuggling to buy the weapons previously provided by the United States and South Africa. The UN imposed sanctions on UNITA in September 1993, forbidding sales of fuel and arms to the rebels. By late 1994, the FAA had retaken most of the cities captured by UNITA, and the rebels were under heavy military pressure.

Diplomatic efforts produced the Lusaka Protocol in November 1994. Lusaka built on the Bicesse framework but directly addressed the previous agreement’s weaknesses, adding power sharing, direct UN oversight of the peace process backed by a 7,000-person peacekeeping force, and a new round of elections to take place after the DDR (disarmament, demobilization, and reintegration) process was completed. The protocol’s effectiveness was cast into doubt almost immediately when Savimbi refused to attend the signing ceremony. Consequently, dos Santos did not sign for the GOA. For the first time in three years, dos Santos and Savimbi met face to face in May 1995, the first of four such meetings in 1995 and 1996. Savimbi’s role in the government had yet to be determined and was a major point of contention. In June 1995, the GOA offered Savimbi the largely ceremonial post of vice president. When it did not receive any reply, it made the offer again in March 1996. Savimbi finally rejected the post in August, claiming that UNITA did not want him to accept it. Savimbi also never went to Luanda, although the UNITA deputies elected in 1992 took their assembly seats, and UNITA members were appointed to the Government of Unity and National Reconciliation (GURN).

UNITA largely failed to live up to its side of Lusaka, pursuing “a policy of procrastination with concessions only when unavoidable” (MacQueen 1998, 409). It dragged its feet on the quartering of its troops and never handed in its heavy weaponry. It refused to allow state administration into many of the areas under its control. In early 1998, nearly one year into the process, UNITA had turned over only 239 of the 344 localities specified in Lusaka (Human Rights Watch 1999, 23). With the billions it gained from illegal diamond sales between 1993 and 1998, UNITA continued to purchase large amounts of weaponry, in contravention of the UN sanctions. It stockpiled weapons and materiel in neighboring countries. UNITA’s foot-dragging constantly pushed back the timetable for completing the various components of the Lusaka process.

The government also utilized the post-Lusaka period to rearm. Military expenditures as a share of GDP had been between 6 and 12 percent from 1990 to 1993; that figure averaged nearly 20 percent between 1994 and 1999 (Omitoogun, 2001, 4). In 1996, for example, the GOA purchased $350 million dollars worth of Russian attack helicopters and fighter jets (Reno 1999, 65). Localized fighting broke out in 1995 and 1996, and cease-fire violations multiplied throughout 1997. The UN imposed additional sanctions against UNITA because of its failure to meet its commitments under Lusaka in August 1997 and June 1998, including a ban on sales of diamonds from Angola not accompanied by a certificate of origin. Simultaneously, the UN was downsizing its peacekeeping mission; by June 1997, only an observer force of roughly 1,500 troops with 345 civilian police officers was left. UNAVEM III transitioned to the UN Observer Mission in Angola (MONUA), which had an authorized strength of 2,000 personnel.

With a return to war looking increasingly likely, the GOA moved to cut UNITA off from its outside support network. Zaire’s leader, Mobutu Sese Seko, was UNITA’s longest running, and by this point its most essential, patron. Savimbi had stashed a large quantity of weapons and supplies in Zaire and had also maintained a “secret army of perhaps 15,000 soldiers,” both deliberately hidden from the Lusaka process (Turner 2002, 81). When the rebellion against Mobutu flared in 1996, the GOA recognized an opportunity both to sever UNITA’s Zairean network and to punish Mobutu for years of meddling in Angola. Initially, it accepted Zairean promises to cut off UNITA in return for Angola staying out of the burgeoning rebellion. But when such promises were not fulfilled—several Zairean generals sold UNITA tons of arms even as their own army was crumbling in the face of the rebel advance—the GOA sent some 5,000-7,000 troops to support rebel leader Laurent Kabila (Turner 2002, 82). Kabila’s capture of Kinshasa in 1997 severely curtailed UNITA’s ability to use Zairean territory and forced Savimbi to establish alternative supply networks and sanctuaries.

Savimbi turned to Pascal Lissouba, the embattled president of Congo-Brazzaville, itself on the brink of civil war. Thousands of UNITA fighters who fled Zaire went to Congo-Brazzaville, where Lissouba incorporated them into his militia. When war broke out in June 1997, the GOA sent 3,500 troops to help Lissouba’s rival, Denis Sassou-Nguesso (Hewitt 1999). With Angola’s help, Sassou-Nguesso seized power in Congo-Brazzaville in October, once more cutting UNITA off from an important external base. The Angolan troops stayed in Congo-Brazzaville until 2001.

To preserve its success in denying UNITA access to the Democratic Republic of the Congo (DRC, the former Zaire), the GOA sent additional troops to support Kabila when his Rwandan and Ugandan allies turned against him in August 1998. The UNITA fighters remaining in the DRC now supported their former enemies, Rwanda and Uganda. By this time, though, UNITA officials had visited both states and were transshipping guns and diamonds through their airports with official sanction. This earned Rwanda and Uganda the GOA’s enmity, and further strengthened Angola’s reasons to back Kabila.

The government’s patience with UNITA finally ran out in December 1998. President dos Santos asked the remaining UN observers to withdraw, suspended the Lusaka Protocol, and ordered a massive military offensive. This round of fighting was markedly more brutal than before, as both sides committed egregious violations against civilians. Between 1998 and 2002, the number of refugees nearly doubled, and the number of internally displaced persons (IDPs) quadrupled (Human Rights Watch 2003b, 6). The government began to forcibly resettle people in government-controlled areas, which included camps with insufficient assistance and security, and pursued scorched-earth tactics. UNITA, hard hit by the offensive and reeling from the loss of its outside networks, could no longer mount sustained conventional opposition, resorting to hit-and-run guerrilla attacks that took a heavy toll on civilians. In one instance in August 2001, a UNITA-laid land mine caused a train derailment that killed more than 250 people.

The government inflicted heavy losses on UNITA throughout 1999. UNITA was unable to import the fuel, spare parts, and munitions it needed to counter the FAA offensive, and its morale plunged. The FAA seized Andula and Bailundo, the military and political headquarters of UNITA, in September, and in late December captured Jamba, UNITA’s headquarters until 1994. The GOA persuaded Namibia to allow its forces to attack UNITA positions from Namibian territory in late 1999, and fighting that also involved Namibian forces flared along the border. In 2000, “UNITA was defeated as a conventional military force” (Turner 2002, 86).

Despite the government’s string of victories, the fighting dragged on through 2000 and 2001. On February 22, 2002, UNITA leader Jonas Savimbi was killed by government commandos in a raid. A cease-fire was in place by April 4, and peace was officially declared on August 2. All UN sanctions were lifted in December.

The Insurgents

UNITA was founded in 1966, when Jonas Savimbi left the FNLA to form his own group. Initially based in Zambia, UNITA was expelled in 1967 after an attack on a strategic rail link between Zambia and Angola. Inside Angola, the group grew stronger, but it remained the smallest of the nationalist groups until independence. It counted only some 3,000 fighters in January 1975.

Following the collapse of the FNLA, UNITA became the main armed opposition force to the MPLA government. UNITA drew its support almost entirely from the Ovimbundu, who constitute roughly a third of Angola’s population. Yet the war, especially after 1992, was not about ethnicity; it was about Savimbi’s personal ambition to be the leader of Angola. Savimbi’s control over UNITA cadres was a “product partly of his personal charisma and genuine leadership qualities, but it [was] reinforced by a fearsome security apparatus, a culture of zero tolerance for dissent and a personality cult that has parallels with those of Mao Tse-Tung and Kim Il-Sung” (Hodges 2001, 18).

Sources: Doyle and Sambanis 2000.
War: UNITA vs. the government
Dates: September 1992-April 2002
Casualties: 500,000-700,000
Regime type prior to war: -7 (ranging from-10 [authoritarian] to 10 [democracy])
Regime type after war: -3
GDP/per capita year war began: US $586 in 1992 (1995 dollars)
GDP/per capita 5 years after war: US $632 in 2003 (1995 dollars)
Insurgents: UNITA (National Union for the Total Independence of Angola)
Issue: Control of central government and mineral resources
Rebel funding: Diamonds
Role of geography: Limited
Role of resources: Revenue from oil and diamonds funded the fighting
Immediate outcome: Government victory due to death of rebel leader
Outcome after 5 years: Peace, elections scheduled for 2006
Role of UN: Facilitated peace talks; five peacekeeping missions; humanitarian assistance
Role of regional organization: None
Refugees: 435,000; all should be repatriated by mid-2005
Prospects for peace: Favorable
Table 1: Civil War in Angola

Throughout the war, UNITA captured “substantial quantities” of weapons from the government, but in the 1990s especially it actively purchased weapons on the black market (UN 2000a, para. 39). Thanks to several UN investigative reports, a great deal is known about UNITA’s funding and arms acquisitions.

UNITA became a major participant in the blood diamond trade following its seizure of the Cuango Valley in November 1992. According to the UN, UNITA cooperated with a foreign consortium to run industrial diamond mines in Cuango Valley, close to the DRC border. Diamonds were sold directly to diamond cutters and their intermediaries, through tenders issued in third countries, and through South Africa’s open market (UN 2000b, para. 176). The governments of Burkino Faso, Zaire, Uganda, and Rwanda facilitated UNITA’s illegal diamond selling. Significant quantities of diamonds also moved through Namibia, South Africa, and Zambia, although not with official sanction. Diamond sales between 1993 and 1998 brought in an estimated $3 billion to UNITA and its partners (UN 2000b, para. 153). The peak year was 1996, when UNITA produced roughly $800 million in diamonds (UN 2000b, para. 152).

Diamonds allowed UNITA to fund the purchase of large quantities of equipment, including “mechanized vehicles such as tanks and APCs, mines and explosives, a variety of small arms and light antiaircraft weapons, and a variety of artillery pieces” (UN 2000a, para. 48). Until the overthrow of Mobutu in 1997, Zaire was UNITA’s main backer. From as early as 1997, Bulgaria was the source for the majority of UNITA’s weapons purchases, although the Bulgarians believed the end-user certificates presented from Togo were authentic (UN 2000b, para. 50).

Despite multiple rounds of UN sanctions, UNITA continued to find suppliers, seeking out arms dealers willing to barter for diamonds. The Fowler Report identified four reasons for UNITA’s success in circumventing the sanctions: the willingness of certain African countries to provide UNITA with end-user certificates and to facilitate the movement of arms through their territory; the willingness of some countries officially or unofficially to sell arms with little or no regard for their final destination; the eagerness of international arms brokers and air transport companies to act as intermediaries between UNITA and its suppliers; and UNITA’s ability to pay for what it wanted, thanks to its diamond operations (UN 2000a, para. 51).

Zaire, Togo, and Burkina Faso all helped UNITA procure weapons. In the early 1990s, Mobutu arranged for weapons to reach UNITA by providing false end-user certificates and facilitating transport (often by air) into UNITA-held areas (UN 2000a, para. 21). In return, Mobutu received cash and diamonds as well as the assistance of several thousand UNITA troops against the Kabila rebellion. Savimbi forged a close relationship with Gnassingbe Eyadema of Togo in the early 1990s, when Togo agreed to help UNITA purchase weapons and other military equipment in return for a 20-percent cash or in-kind cut of the shipment’s value (UN 2000a, para. 33). With Mobutu’s position increasingly threatened by the rebellion in Zaire, Savimbi in early 1997 turned again to Togo, which gave UNITA a genuine end-user certificate that UNITA then forged and used repeatedly. Burkina Faso also supplied end-user certificates; Blaise Campaore and Savimbi enjoyed a close personal relationship in addition to their financial ties (UN 2000a, para. 103). Eyadema and Campaore received diamonds in payment for the false documents and the assistance in storing and transporting the weapons.

The FAA offensive in 1999 made UNITA’s illegal diamond trading and arms acquisitions extremely difficult. The UN embargo on air travel into UNITA-held areas and the GOA’s interventions in the DRC and Congo-Brazzaville also cut off UNITA from its outside supply network. As its resources were choked off, UNITA found it increasingly difficult to prosecute the war.


Savimbi established his headquarters in Jamba, in the farthest reaches of the sparsely populated Cuando Cubango province, an area the Portuguese had called the “end of the world.” This isolation allowed the group to survive once the MPLA took power in 1975, and Jamba’s close proximity to the Namibian border meant South African assistance, including SADF troops, could flow in easily in Phase I of the war. In the early 1980s, UNITA expanded the areas under its control from some southern and central provinces toward the Zairean border. By the mid-1980s, UNITA’s reach extended to areas bordering Zaire, allowing it to use Zaire as a transshipment point, supply depot, and sanctuary. Zaire gave UNITA strategic depth, which helped the movement to pose a serious challenge to the government’s ability to rule for more than two decades.

Most of Angola’s diamonds are in alluvial deposits, which are much easier to exploit than kimberlite deposits. Thus, UNITA’s mining operations were labor intensive and did not require a great deal of mining equipment. The location of the richest deposits in areas easily accessible to Zaire/DRC also eased UNITA’s ability to move diamonds out of Angola and to bring weapons, fuel, food, and equipment in from Zaire. The location of the major oil fields offshore ensured government control of this resource and protected it from disruption during the fighting.


For most of its existence, UNITA was a classic rural guerrilla insurgency. In the 1992-1994 period, it evolved into a semiconventional force that was capable of laying siege to and capturing major provincial capitals. These tactics took a heavy humanitarian toll. UNITA’s twenty-one-month bombardment of Kuito, which it failed to capture, killed an estimated 20,000-30,000 people and all but leveled the town; its shelling of Huambo killed 10,000 people (Human Rights Watch 1994). UNITA rained 1,000 shells per day on these cities. Such sieges also produced severe food shortages. UNITA captured and held six provincial capitals: Caxito, Huambo, M’banza Kongo, Ndalatando, and Uige (Hodges 2001, 15). It also moved its capital from Jamba to Bailundo in the central plateau province of Huambo during this period. Under the Lusaka Protocol, UNITA was to turn the areas under its control over to the central administration, but when fighting resumed in 1998 it still held sixty localities (Human Rights Watch 1999, 23).

Government successes in 1999 forced UNITA into hit-and-run ambushes and terrorist tactics for the remainder of the war. UNITA began to target expatriate workers, government and traditional leaders, and humanitarian workers, and its abuses against civilians (such as forced recruitment and mutilation) increased dramatically. Freedom of movement was denied in UNITA areas, and no humanitarian assistance reached those populations for several years. Southern Bie province, for example, was inaccessible from 1998 until the end of the war. The FAA also engaged in widespread forced displacement, moving people from rural areas into major population centers or pushing them into refugee camps. The humanitarian emergency that emerged when the war ended was a direct consequence of the warring parties’ tactics.

UNITA’s ability to wage anything more than a guerrilla campaign was dictated by the number of men it had under arms, the resources it controlled, and the size of the FAA. UNITA’s major gains came at a time when the FAA had disproportionately demobilized, giving UNITA a large numerical advantage. Accurate estimates of UNITA’s size are difficult to make, but it was to have quartered 62,500 combatants as part of the Lusaka process and 105,000 combatants went through DDR at the war’s end (Hewitt, 1999 ; UN 2003, 3). The FAA was the largest army in sub-Saharan Africa by the late 1990s, with roughly 120,000 men. Over the final two years of fighting, it grew to 146,000 (Stockholm International Peace Research Institute 2001).

Most UNITA forces were light infantry units, although UNITA was able to field several armored and mechanized units. It captured some T-54/55 tanks from government forces and purchased four or five others (UN 2000a, para. 49). It was not until December 1998 that UNITA first used tank divisions. UNITA also had artillery and antitank and air defense units, but the mainstay of its weapons inventory was small arms and light weapons (UN 2000a, para. 47). Like the government, UNITA made heavy use of land mines.

Causes of the War

Although earlier phases of the war were fueled by anticolonialism or Cold War ideological battles, the 1992-2002 fighting was largely fueled by the ambitions of one man, Jonas Savimbi. His refusal to accept the election results of 1992 was the direct trigger for a return to war, and his recalcitrance with respect to Lusaka was largely responsible for its failure. UNITA served as his personal vehicle, as evidenced by the rapid end of the war after his death.


Conflict Status

The war ended in 2002 as a result of the death of the rebel leader, Jonas Savimbi. The two sides signed an addendum to the Lusaka Protocol, the Luena Memorandum of Understanding (MOU). The MOU dealt with military issues still outstanding from the implementation of Lusaka, such as DDR and the integration of UNITA personnel into the FAA and the national police.

Despite an end to the fighting, Angola continues to face severe resettlement problems. Demobilization created a new humanitarian crisis, as the UNITA excombatants waited in camps for the government to deliver on its promises of aid. A number of factors hindered the DDR process, including a “lack of adequate facilities, inaccessible roads, mine infestation and inadequately prepared resettlement areas” (UN 2003, 3). Excombatants were promised farming tools, household goods, zinc roofs, clothes, and $100 stipends, but the GOA did not universally provide such benefits, claiming it faced financial difficulties in paying for the DDR process. In September 2002, the World Food Program began supplying food to the excombatants because the GOA had not (Cauvin 2002). Six months after the cease-fire, most fighters were still in the camps, and a year later 15,000 fighters were still waiting to enter the reception areas (UN 2003, 3). Potentially problematic is that UNITA turned over the equivalent of one weapon for every three soldiers in the disarmament process (Human Rights Watch 2003a). Given the staggering unemployment rate and the government’s inability to provide even basic services, there are continuing concerns that ex-UNITA fighters will resort to banditry to sustain themselves and their families (ICG 2003a, 6). The UN reported in 2003 that 90,000 UNITA excombatants went through the resettlement process out of 105,000 total (each with average of six dependents) (UN 2003, 3).

When the fighting stopped, half a million people living in UNITA-held areas previously inaccessible to humanitarian assistance proved to be in dire need of aid; the UN Office for the Coordination of Humanitarian Affairs (OCHA) said that malnutrition among those living in combat zones was among the worst seen in Africa in a decade; some 3 million Angolans required food assistance to survive by mid-2002 (ICG 2003a, 4). Relief efforts were hampered by bad (or nonexistent) roads, land mines, and seasonal rains, which combined to place 445,000 people beyond the reach of assistance by early 2003, with an additional 200,000 people yet to be reached for the first time; a year after the war ended, mortality rates were still at emergency levels, especially in rural areas (UN 2003, 4).

By the war’s end, roughly 435,000 refugees had fled, mainly to the DRC, Zambia, and Namibia; another 4 million persons—one-third of the population—were internally displaced (United Nations High Commission for Refugees 2002, 2004). Large numbers of persons returned to their homes in 2002 and 2003, often outside of the formal resettlement process. OCHA reported that 3.8 million Angolans returned or resettled in 2003; some 70 percent did so without any assistance from local authorities or humanitarian organizations (United Nations Office for the Coordination of Humanitarian Affairs 2004, 1). Land mines made this especially hazardous. Only 30 percent of designated return sites had basic services in place for the returnees, and 900,000 people returned to areas without any basic services (UN 2003, 5). The United Nations High Commission for Refugees (UNHCR) reported that by mid-2005, some 320,000 refugees had voluntarily repatriated (185,000 with UNHCR assistance), and it expected to repatriate another 55,000 refugees in the later half of the year before its program ended (United Nations Integrated Regional Information Network, 2005c). Nearly all of the ex-combatants and IDPs had returned home or had decided to resettle in new communities by 2005.

Duration Tactics

The Angolan conflict dragged on for four decades largely because the parties to the conflict were not committed to a peaceful solution. The parties’ deep distrust of one another (a distrust amply deserved in both cases) repeatedly meant that, rather than fulfill their commitments as specified in Bicesse and later Lusaka, they constantly hedged their bets. Both sides held back forces in the various DDR processes, stockpiled weapons, and purchased huge amounts of equipment during periods of cease-fire. Savimbi in particular deserves blame for undermining the various peace plans. His intentions were never clear, as he routinely failed to show up for meetings or signings, refused to move to Luanda, and was unwilling to accept a lesser role in a unity government.

UNITA dragged its feet at every possible juncture, and even when it was cooperating its efforts, fell well short of full compliance. For example, during the DDR process associated with Lusaka, many of the “soldiers” UNITA sent to quartering areas were conscripted solely for that purpose, were war wounded, or were under eighteen. It handed in few weapons, many of them old or broken, and never surrendered its heavy weaponry.

Outside actors also share some responsibility for Angola’s prolonged war. In Phase I, third-party intervention by the superpowers and regional actors provided both sides with the military, financial, and political resources necessary to continue the fighting for thirteen years. In Phases II and III, key states such as Russia and Portugal continued to sell weapons to the government, and a variety of states actively connived at sanctions-busting to aid and trade with UNITA.

The UN record in Angola in Phases II and III is a poor one. UNAVEM II was undersized and underresourced and could not prevent the collapse of the Bicesse Accords. It must be recognized, though, that the limited scope of the UN role under Bicesse was a reflection of the wishes of the GOA and UNITA. Allowing the parties themselves to police implementation of the peace accord was identified as a key weakness of Bicesse, and thus Lusaka provided for a much more robust UN role. But UNAVEM III actually downsized as cease-fire violations increased in the mid-1990s, and failed to halt the steady slide toward renewed war. Although the UN passed successive sets of sanctions against UNITA, it often turned a blind eye toward their violation. To its credit, once the UN commissioned the Panel of Experts in 1999, its reports on sanctions-busting were tough and detailed, even naming names of complicit heads of state.

Finally, the political economy of war helped keep the conflict going after 1991. The government and UNITA relied on oil and diamonds, respectively, to fund their war efforts, and as long as states and companies were willing to deal with them, the parties had access to tremendous sums of money.

External Military Intervention

UNITA’s support in the 1974-1975 period came from South Africa, Zaire, and the United States. U.S. assistance was forbidden by Congress under the 1976 Clark Amendment, but South Africa and Zaire remained major patrons. The United States restored military assistance to UNITA in 1985 as well as supporting it diplomatically. The GOA had the backing of Cuba and the Soviet bloc.

During Phase I of the Angolan civil war, Cuban, South African, and Zairean troops were all involved to varying degrees. The 1988 Namibia Protocol brought about the withdrawal of Cuban troops as well as the end of SADF involvement. The 1991 Bicesse Accords provided that the United States and the Soviet Union would cease supplying their respective allies in the civil war. South African support ended as well with the implementation of Bicesse and the end of apartheid.

However, once fighting resumed in 1992, Russia once more supplied Angola with weapons on favorable terms, becoming the GOA’s largest supplier from 1992 to 2002. Portugal, Belarus, Brazil, Bulgaria, China, and others also supplied the government with weapons. UNITA captured its weaponry from the FAA or bought it on the black market.

Under heavy pressure from UNITA forces after the fighting resumed in 1992, the GOA hired Executive Outcomes (EO). The private security firm committed about 550 personnel to the Angola mission. “Defense strategists generally credit Executive Outcomes (EO) with greatly assisting the MPLA to turn back the resurgent UNITA” (Howe 1998, 312). Initially, EO protected oil installations, trained FAA troops and pilots, and helped plan military operations. From mid-1994, EO also deployed its own forces to retake and then hold onto several important diamond mining areas. The Lusaka Protocol called for the repatriation of all mercenaries, but the GOA did not completely sever its ties with EO until January 1996 at the insistence of the United States and UN. EO received $40 million annually for its services (Howe 1998, 318).

Conflict Management Efforts

The UN played a prominent role in efforts to bring peace to Angola, authorizing five peacekeeping missions:

UNAVEM I (UN Angola Verification Mission): January 1989-June 1991
UNAVEM II: June 1991-February 1995
UNAVEM III: February 1995-June 1997
MONUA (UN Observer Mission in Angola): July 1997-February 1999
UNMA (UN Mission in Angola): August 2002-February 2003

UNAVEM I, established as part of the 1988 Brazzaville Protocol, was a traditional peacekeeping operation: a small observer force mandated to verify the withdrawal of Cuban troops, created with the consent of the parties. UNAVEM I “achieved its limited goals with very limited resources precisely because the parties cooperated and wanted the UN to succeed in its mission (Jett 1999, 117).

UNAVEM II confronted a much more difficult task. The Bicesse Accords, which the UN had played little role in negotiating, called for the establishment of a 350-person UN observer force to supervise the cease-fire between the FAA and UNITA. Expectations for what this small mission could accomplish were immense, and in fairness UNAVEM II did succeed in organizing the 1992 elections on a very compressed schedule. But the mission’s mandate was poorly defined, and its resources were “hopelessly inadequate for the magnitude of the task” (Hodges 2001, 14). Flaws in Bicesse, including the lack of a power-sharing requirement and the failure to make the elections conditional on the fulfillment of the military goals, were magnified by the lack of cooperation on the part of the parties themselves. “UNAVEM II had neither adequate resources nor the authority to help the parties find peace, even if they had been sincere about wanting it” (Jett 1999, 162).

Following the collapse of Bicesse and the return to war, the UN continued to search for a negotiated settlement. It sponsored several rounds of talks in 1993, and in September adopted UN Resolution (UNR) 864, which prohibited all sales or supply to UNITA of arms and related materiel and military assistance as well as all petroleum and petroleum products. The threat of further sanctions helped push Savimbi to accept the validity of the 1992 elections.

The Lusaka Protocol attempted to correct the flaws of UNAVEM II by providing the new UN mission with a much more robust mandate. It was charged with overall supervision, control, and verification of the cease-fire; chairing the Joint Commission that oversaw implementation of the Lusaka provisions; and directing the DDR process for UNITA forces. UNAVEM III’s authorized strength was 7,000 troops, 350 military observers, and 260 police observers, and its budget was roughly $1 million a day (MacQueen 1998, 416).

Although progress was made in implementing Lusaka, UNITA’s procrastination repeatedly meant that target dates in the Lusaka timetable were not met. UN pressure often brought at least some measure of compliance from UNITA, but by the fall of 1996 the UN’s patience was all but exhausted. It threatened new sanctions if UNITA failed to deliver its fighters to the quartering areas. By early 1997, 70,660 UNITA troops had registered, and UNITA declared it had completed the quartering of its troops. The UN remained concerned, however; of that number, 22,686 troops had deserted or temporarily absented themselves from the camps (UN 1997, para. 10). UN threats also helped push UNITA to agree to the formation of the Government of Unity and National reconciliation in April 1997, although Savimbi refused to attend the swearing-in ceremony.

With nearly all UNITA troops supposedly in the quartering areas and the GURN in place, the UN extended UNAVEM III’s mandate in April for a final six months, planning to replace it with a smaller observer mission. Yet, as the UN transitioned into MONUA, there were already abundant signs that peace was precarious. UNITA still had not met several key obligations under Lusaka, Zaire was being consumed by a civil war that would pull in both UNITA and the FAA, and serious cease-fire violations were multiplying. The intense fighting that broke out in northern Angola in the summer of 1997 “demonstrated not only that UNITA still controlled major portions of the country but that it had hidden a major military capacity from the UN” (Jett 1999, 163).

Like its predecessor UNAVEM II, MONUA found that its small size and limited resources were dwarfed by the tremendous challenges that faced it. With Lusaka disintegrating and MONUA outmatched, the UN finally imposed the oft-threatened sanctions against UNITA in October 1997. UN Resolution 1127 placed an international travel ban on UNITA officials, closed UNITA offices abroad, and prohibited flights into Angolan territory without approval by the government. Further sanctions were passed in June 1998, targeting UNITA’s diamond-trading network. UNR 1173 prevented the direct or indirect import from Angola of diamonds not controlled by a certificate of origin and banned the export of equipment for mining and mining services to UNITA-controlled territory. But it was too little, too late. The GOA asked the UN to withdraw MONUA preparatory to a return to full-scale war in December 1998.

From UNAVEM I through MONUA, the UN spent $1.5 billion and lost nearly sixty peacekeepers.

Following Savimbi’s death, the UN authorized its final mission to Angola (UNMA). UNMA was tasked with assisting the parties in consolidating the peace; its mandate included support for the reintegration of excombatants, facilitation and coordination of humanitarian assistance, technical support for mine action, and the protection and promotion of human rights (UN 2003, para. 11). The mission operated from August 2002 to February 2003.


Decades of civil war combined with corruption and poor economic management have left Angola desperately poor. According to the 2004 Human Development Index, Angola ranks 166 out of 177 and has among the world’s worst infant, child, and maternal mortality rates (United Nations Development Program 2004, 142). Life expectancy at birth is 40.1 years. More than half the population is malnourished, and the vast majority lack access to basic health care or clean water. Angola is one of the world’s most heavily mined countries, and one in every 415 Angolans is an amputee (International Campaign to Ban Landmines 2002, 74). Infrastructure has been shattered, and the endemic land mine problem has vastly complicated transport, agriculture, the movement of people, and the delivery of humanitarian relief assistance. Unemployment is estimated at over 50 percent (Timberg 2004).

The plight of children, who constitute 60 percent of the population, is dire. One quarter of Angolan children die before their fifth birthday; 45 percent are chronically malnourished; and nearly half of school age children are not in school. Eleven percent of children lost one or both parents in the war (United Nations Children’s Fund 2003).

Decades of war drove many people from rural areas into the cities, and many do not wish to return given the miserable and dangerous conditions in the hinterlands. Luanda’s population “has swelled from several hundred thousand to more than 3 million, overwhelming the city’s infrastructure. Other cities have grown as well” (Timberg 2004). Despite several years of robust economic growth, Angola is plagued by “a sizeable debt, a swollen public sector payroll, and largely unaccountable state institutions that dominate critical areas of the economy” (IMF 2005b). There is a yawning chasm between rich and poor, which often equates to those who work for the government, including former UNITA officials, and everyone else.

A return to war is unlikely given the total defeat of UNITA and the exhaustion of the population. Angola nonetheless faces enormous challenges. State building has been distorted by the war that consumed nearly every year of Angola’s independent existence. Its coercive and extractive capacities are outsized, while it is virtually unable to meet even the most basic needs of its population. It has been plagued by dismal political leaders, men committed to their own power and enrichment regardless of the cost to Angolans. And it has been betrayed by an international community that contributed, through sins of omission and commission, to the ability of the MPLA government and the UNITA rebels to sustain nearly thirty years of war.

If Angola is to become anything more than a failed state, its leaders must commit themselves to genuine political and economic reform—to ending cronyism, to professional and transparent management of the country’s natural resources, to accountable government, to respect for democratic freedoms and rule of law, to significant investment in the country’s broken infrastructure, to massive social sector spending. For Angola, state building and postconflict peace building are mutually reinforcing processes.