Kenya is a multiethnic and multiracial society. There are about 40 ethnic groups. The dominant ethnic groups are the Kikuyu, Luhya, Luo, Kalenjin and Kamba. No one group has a clear numerical majority in relation to other ethnic groups. Kenya had until recently been spared the problem of violent ethnic conflict. In 1991, however, her image of a country enjoying political stability was shattered when fighting broke out between ethnic groups in the Rift Valley province. This outbreak of violence occurred at the peak of a movement for economic and political reform that dates back to the late 1980s. The democratisation movement was led mainly by the clergy, lawyers and political leaders who had fallen out of favour with the KANU government. It was inspired by the collapse of communist states in Eastern Europe and the attendant shift in the international political agenda from ideological cold war issues to those concerning democratisation and respect for human rights.

This paper attempts and early warning indicator identification, through an examination of the causes and dynamics of conflict in the Rift Valley and Western parts of Kenya. One result of this conflict is that politico-ethnic violence has been factored into Kenyan politics in such a way that it raises concern over the potential for the current political struggles between the opposition and the KANU government to degenerate into outright civil war. The paper concludes with a brief evaluation of the conflict resolution mechanisms that should have been activated to prevent outbreak of conflict or contain it, and an analysis of the potential for renewed conflict in the region under examination.

The first outbreak of ethnic conflicts occurred at about the same time the Consultative Group (CG) meeting for Kenya took place in November, 1991. At this meeting, Western Donor governments and the World Bank made it clear to the Kenya Government (GoK) that aid disbursements would be dependent on major economic and political reforms. Even before this meeting the President had repeatedly warned that multiparty politics was not compatible with a multi-ethnic Kenya and would lead to war and chaos, similar to what was happening in neighbouring countries at the time. In October 1991, gangs of young men supposedly belonging to the Kalenjin ethnic group, armed with spears and machetes, attacked settlements belonging to members of the Luo ethnic group in the farming areas of Kericho, Nandi and Kisumu, in Western Kenya. Non-Kalenjin ethnic groups, mainly Luo, Luhya and Kikuyu, had established permanent settlements in the region as far back as 1969. Parts of these areas were traditionally inhabited by the Kalenjin. Meitei farm in South Nandi was the first place to be attacked in October 1991. In this and following attacks, hundreds were forced to flee their homes, houses were burned down and shops looted. Three people died, thirty houses were burned and about four thousand people were left homeless. In addition, acres of sugar cane plantations were destroyed by fire in the Miwani, Chemase and Kaptweta areas. Luo men mobilised to retaliate and after 11 days the fighting spilled over into the neighbouring Kisumu district, home to the Luo ethnic group.

Fighting between the Luo and the Kalenjin continued along the boundaries of the Kisumu and Kericho districts during the months of November and December 1991. It also spread to the Uasin Gishu and Nandi districts. By May 1992 as many as 2,000 people were killed and 50,000 rendered homeless. The victims of the attacks now included members of the Kikuyu and the Luhya ethnic groups who had settled in the Kericho and Nandi areas of the Rift Valley. The attackers claimed that the Luo and other ethnic groups had encroached on traditional Kalenjin lands.

Soon after the multi-party elections in December, 1992, fighting erupted once again in the Molo, Narok, Pokot, Londiani, Elburgon and Burnt Forest, areas of the Rift Valley province . The aggressors during this post-election phase of the conflict included the Maasai and the Pokot, traditional inhabitants of these areas. Their main targets were the Kikuyu, the main ethnic group settled in the area. Reports claim that land belonging to fleeing Kikuyu families was bought up by Kalenjin or confiscated and the title deeds torn up. The Rift Valley areas of Molo, Londiani, Elburgon and Burnt Forest were hit hardest by this ethnic violence. A common factor in all the areas affected by the violence is that they are inhabited by large numbers "migrant" ethnic groups (Kikuyu, Luhya, Luo, Kisii), who settled in these areas immediately after independence. Of these, the Kikuyu are the largest and also earliest immigrants to the area. These ethnic groups form the major support base of the different Oppostion parties in the country.

According to the United States Department of State report on Kenya Human Rights Practices, in 1993, ethnic violence had by December of that year, claimed 1,000 lives and displaced between 150,000 to 250,000 people. In its 1994 report on Kenya Human Rights Practices, the Department noted that renewed fighting broke out in the Burnt Forest area of the Rift Valley (between the Kalenjin and Kikuyu) during March 1994, and similarly in the coastal town of Mombasa (between coastal ethnic groups and the Luo) and north-western Kenya (between the Turkana and non-Turkana ethnic groups) during June 1994. This time the forcible eviction of the Kikuyu from Enoosupukia in Narok district (a Maasai area), attracted the attention of international human rights activists. The report concluded that, in general, ethnic violence had decreased considerably in 1994 compared to 1993. On the other hand, it had spread beyond the Rift Valley to the Coast Province where the Luo were targets of violence by coastal ethnic groups. Renewed fighting was reported in January 1995, in the Mai Mahiu area of Naivasha, a region bordering Maasai territory and inhabited by Kikuyu farmers. Ten people were reported killed and several homes destroyed during an attack on a settlement on January 10, 1995. The attack followed allegations by the government that guerrilla warfare was being planned.

The conflict drew mixed responses from both the State and Civil Society. State response ranged from verbal statements to concrete action on the part of various state actors: cabinet ministers, administrative officials, security personnel and the President all acting in the name of the government. Soon after the outbreak of each conflict, local administration officers and security personnel mobilised security to prevent further attacks, but these in general, proved to be ineffective. For instance, a few days after the first outbreak in October 1991, the General Service Unit (GSU), a paramilitary force, together with some Administration Police (AP) were deployed to quell the fighting but were apparently unable to do so. Eyewitness accounts from the Kisumu, Kericho and Nandi areas reveal that the armed police watched helplessly as people were attacked and homes burned down. Although a government curfew in February 1992, helped curb the fighting in these areas, it did not stop the attacks from being carried out. In the Narok and Molo areas of the Rift Valley, local residents reported that police were either unable or unwilling to stop the aggressors. In some cases the police were within several metres from the scenes of violence but did not seem bothered about it. Hence, security personnel were actually deployed, but were more competent in arresting those who retaliated than in preventing further attacks by the aggressors. This would characterise police response throughout the pre-election phase of the conflicts between October 1991 and December 1992, and the first year of the post-1992 General election period (1993).

Not until September, 1993, and after two years of continued ethnic fighting in various parts of the country did the government make any serious efforts to curb violence in the affected regions. The President invoked the Preservation of Public Securities Act and declared the hardest hit areas, Molo, Elburgon, Londiani, and Burnt Forest, "security zones." Consequently, a number of measures were undertaken. A ban was imposed on possession of weapons, movement of livestock at night, and publication of information on conflict in the "security zones" without government consent. Opposition Members of Parliament, human rights activists and journalists were prevented from entering these areas. Security personnel were given wide ranging powers which included authorization to shoot to kill, requisition of private vehicles, search and arrest, and also prohibition of movement of residents. Before this Act was invoked, several persons were arrested in connection with the violence but no mention is made of their prosecution. This is surprising considering the numerous incidence of violence and the nature of destruction. It begs the question of how genuine government security measures were. It also marked the beginning of the people’s loss of confidence in the ability of state institutions to protect their lives and property and also ensure security.

Responses of the local administration included the posting of additional District Officers (DOs) and security personnel to oversee resettlement of the displaced, organise reconciliation meetings of elders from the affected ethnic groups and ensure security at the local levels. In most cases these initiatives were carried out following a Presidential directive. No plans were made to compensate families of victims for damage done by the aggressors. During the pre-election phase of the conflict, government statements issued by senior government officers and speeches made by the President refer to the attacks 'land disputes' or 'land clashes'.

Meanwhile the outbreak drew varied responses from Cabinet Ministers and Members of Parliament (MPs). These included calls for cessation of conflicts, expressions of concern at the outbreak and outright condemnation of the attacks. It was one of the rare times that KANU members disagreed with each other passionately. At first a group of KANU Members of Parliament from Kericho and Nandi districts (in the Rift Valley) condemned the attacks but in a surprising change of mind, Kalenjin MPs claimed, at a meeting in early 1992, that the clashes were calculated to alienate and tarnish the Kalenjin community. The Hon. Nicholas K. Biwott in particular, told a pre-dominantly Kalenjin audience that the clashes were an attempt to destabilise semi-arid lands and parts of Northern Kenya by outsiders. MPs whose communities were victims of the Rift Valley attacks brought the matter up in Parliament for discussion as an issue of national importance, insisting that these attacks were politically motivated and inspired by the 'majimbo' rallies organised by the Rift Valley KANU MPs in the period preceding the conflict. It is significant that the Speaker of the National Assembly at the time, Prof. Jonathan Ng'eno resisted debate on the issue four times, before a freelance motion initiated by the MP for Vihiga (Western Kenya), Mr. Bahati Semo and calling for the appointment of a committee to probe the conflict, was passed. Soon after, Parliament approved the establishment of the Parliamentary Select Committee to Investigate Ethnic Clashes in Western and Other Parts of Kenya. This committee conducted investigations in the affected areas and compiled a report which incriminated a number of MPs, Cabinet Ministers (most of whom were from the Rift Valley and close associates of the President), administrators, security officers and ordinary people. Debate on the report was suddenly suspended and it was subsequently rejected.

After the elections in 1992, members of parliament from KANU and the opposition accused each other in turn, for inciting ethnic attacks. President Moi, and a number of Cabinet Ministers and KANU MPs blamed the opposition parties for the violence and frequently referred to the incidents as proof of how multiparty politics was incompatible with a multi-ethnic Kenya. This self-fulfilling prophecy was repeated at rallies on his many tours of the country, and in government statements issued from the State House. In one such statement released after angry public reaction to the governments' insensitive response to renewed ethnic fighting in Naivasha on January 10 1995 (see section on outbreak of conflict), senior clergymen, opposition politicians and journalists are blamed for the insecurity prevailing at at that tim.

What is significant about the strong government accusation of the opposition, is that it contrasts greatly with government silence, indeed non-censuring, of inflammatory remarks made by KANU Ministers and leaders from the Rift Valley. For instance, Mr. William Ole Ntimama, the Minister for Local Government, had, since August 1990, made frequent anti-Kikuyu remarks in public. While replying to criticisms of Maasai attacks on Kikuyu in Enoosupukia area of the Rift Valley, Ntimama maintained that "The Maasai in Enoosupukia were fighting for their rights and I have no regrets about what happened there. We had to say enough is enough. I had to lead the Maasai in protecting our rights." President Moi let this and other similar remarks by KANU politicians pass without comment, despite calls from Members of Parliament and other public leaders for Mr. William Ole Ntimama's resignation. Such inflammatory remarks dating as far back as September, 1991 preceded outbreaks of outbreaks of conflict and cannot be dismissed as having had no influence on the violence. They set the stage, even if only psychologically, for political violence. The governments' silence strongly suggests either involvement in, or tacit support of the conflicts.

Increasing international donor pressure forced a reluctant government to work with the United Nations Development Programme (UNDP) on a resettlement and rehabilitation programme for displaced persons beginning early 1994. As a result of this a number of NGOs contracted by the UNDP to implement various projects, were allowed into the "security zones". Later on diplomats and journalists were granted some access to these areas. In September, 1994, the UNDP published a report by Prof. John Rogge, on the status of the internally displaced in Kenya. It covered UNDP resettlement programmes in Nandi, Nakuru, Kericho, Kisumu, Bungoma, Trans Nzoia, and Uasin Gishu districts. According to the report, about 42,500 families representing 250,000 people were affected by the clashes. By August 1994, 10,000 to 20,000 persons were displaced in camps. 30% of those displaced by the ethnic conflict were resettled to their villages, 50% were in the process of returning and 20% remained in camps set up in market centres.

In two areas, Maela (Naivasha) and Thessalia Catholic mission (Kericho), the Government frustrated UNDP efforts to resettle the displaced. As far as the government was concerned, the displaced in these camps were not victims of clashes but squatters on land designated as livestock holding area. Government frustration of resettlement efforts was not limited to the UNDP. Legal/human rights groups offering assistance to the displaced were harassed, some legitimate land claims made by the displaced were rejected, insecurity was allowed to persist and pro-KANU government ethnic groups were not prevented from taking over farms owned by the displaced. The Rogge report mentions some harassment and intimidation of returnees which was, apparently, more severe in the case of displaced Kikuyu.

Civil society response to the conflict

The attacks took victims by surprise. Most fled from their farms and sought refuge in neighbouring towns, market places, trading centres or church compounds where several makeshift camps were set up. Local administration officials appealed to residents to return to their homes and assured them of their safety but very few of the displaced were convinced and remained at the camps. Instead some would go back to their farms during the day to cultivate but would return to the camps at night for fear of being attacked. A significant action on the part of the displaced was the exchange of land along district borders. In a kind of 'self-imposed ethnic cleansing' the Kikuyu, for instance exchanged land with the Kipsigis (a Kalenjin group) along the Nakuru/Kisumu district boundary, and along the Kakamega/Uasin Gishu district boundary, the Luhya on the Uasin Gishu side exchanged land with the Kalenjin who had been in Kakamega district. One other revealing as aspect of the victims' response is the fact that it did not develop into a systematically organised and violent retaliation such as arming to attack members belonging to the Kalenjin, Maasai, Pokot or Sabaot ethnic groups. This is not to say that no retaliation took place, indeed during the first attack in October 1991, Luo men mobilised to retaliate. Most probably other affected non-Kalenjin ethnic groups did the same. Two factors may account for this. In the first place, as mentioned earlier, security personnel responded much faster to prevent attempts by victims to mobilise against the aggressors than to attacks by the aggressors. Secondly, victims used other channels such as the media, the church, non-government organisations, and local leaders to appeal to the government to prevent the attacks and restore peace.

The National Council of Churches of Kenya (NCCK), the Catholic Church, the Kenya Red Cross, Catholic Relief Services, the Church of the Province of Kenya (CPK) and ACTIONAID were among the first to provide food and shelter to the displaced. In addition, several statements made by church leaders, politicians in the opposition and human rights activists repeatedly condemned the ethnic violence throughout its duration, and called on the government to initiate peace and reconciliation. The NCCK and the Roman Catholic Bishops carried out a number of fact finding missions which came out with reports and several pastoral letters which were published throughout the two phases of the conflict. These were strongly critical of government response to the conflict and called for peace and reconciliation. The pastoral letters from the Roman Catholic Bishops, which were read at Catholic Churches nationwide, maintained that the ethnic clashes were neither land disputes nor a product of multiparty party politics. The Bishops also observed that the clashes were not 'tribal' but rather, part of a wider political strategy, and involved well trained arsonists and bandits transported to the scene. The Bishops noted the selective protection of certain ethnic groups, in particular the Kalenjin who remained largely unharmed compared to the non-Kalenjin groups. The Churches nonetheless, did support, albeit cautiously, the government decision to establish 'security zones' in the hope that this would help curb fighting. Whereas the Churches (Catholic and Protestant) came out strongly against the ethnic violence at the national level, this was not the case at the local level in some areas. In the Mt. Elgon area in 1990, a religious dispute in the CPK Nambale Diocese created tension between the Teso and Bukusu members and resulted in the former calling for the formation of their own diocese. The ethnic tensions created by church politics at the local level added to those created by the local leaders ‘ethnicised’ interpretation of multi-party elections. In this and other cases, the behaviour of local church leaders did more to encourage ethnic hostility rather than prevent it.

Like the Churches, The Forum for Restoration of Democracy (FORD) accused the Kenya African National Union (KANU) government for inciting the attackers and wondered why the government could not mobilise security. At the time of the first outbreaks of violence, the FORD, a loose coalition of opposition politicians, represented the opposition. The FORD repeatedly rejected claims by KANU ministers and members of Parliament that opposition politicians were responsible for the violence. The Forum blamed the government for using violence to weaken popular support for the transition to democratic government and also pointed out that the conflict was planned so as to fulfil President Moi's prediction that multiparty politics would result in civil war. The opposition was thus pushed into defending itself against unwarranted claims.

One would expect that both the increasing incidence of 'ethnicized' violence and evidence of government complicity, would serve to unite the opposition politicians and other civil rights groups against the regime. Instead, as Frank Holmquist and Michael Ford observe, the clashes set in motion a withdrawal into ethnic sentiment for protection. This is clearly evident with regard to the opposition movement. By mid 1992, the FORD had split into two parties the FORD-Kenya and FORD-Asili (Original). A third major party, the Democratic Party had been formed soon after the repeal of the one-party clause in the constitution to allow for the re-establishement of multiparty politics, in late 1991. The fracturing of the opposition movement into several disunited parties weakened the movements ability to respond effectively to the now ethnicized violence. It also provided the Rift Valley KANU politicians with evidence for their argument that multiparty politics would only heighten ethnic animosity and lead to war. Despite this fracturing of the opposition, a number of opposition politicians, especially from the FORD-Asili, and FORD-K parties did visit the affected areas but were either prevented from meeting victims, or arrested and charged for inciting locals to violence. Continued government harrasment and intimidation by security personnel blocked any opportunities the opposition may have had to organise for a peaceful transition to multi-party politics.

International Community response

Donor response varied from cautious and discrete pressure on the government to institute reforms to outright criticism of government inaction. Denmark, Germany, and the United States of America put greatest pressure on the Kenya Government to institute economic and political reforms. In 1993, Denmark announced further cuts in aid to Kenya, one of the reason being the government's inability to end the ethnic clashes. By 1995, Denmark and Germany were still not happy with the government performance. Germany cut its aid by two thirds while the Denmark made it clear that they would not disburse aid pledged because of setbacks in political and economic reforms, one of which included the forced transfer of the displaced from Maela camp in early 1995. American pressure came from the then American Ambassador's strong criticisms of the KANU government for its reluctance in carrying out reform, and his open support of the Opposition parties. These were given very high media publicity and drew the KANU government’s anger at what was seen as US interference in internal affairs. Although well meant, Smith Hempstone's criticisms served more to put the government on the defensive rather than open a way for dialogue, much as they did contribute to strengthening the voice of the opposition. Other donor governments (such as the British) preferred to use a more discrete way of pressurising the GoK on aid. The use of aid conditionality did not bring about immediate results before the elections, except for the decision by the November 1991 Consultative Group meeting in Paris to tie aid disbursements to political and economic reform. This decision influenced the KANU government's repeal of the constitutional amendment (Section 2(A)) legalizing one party rule in Kenya, and by so doing facilitated a return to multiparty politics but it did not influence cessation of conflict in the Rift Valley and Western Kenya, neither did it result in negotiations over a democratic Kenya. Interestingly, ethnic attacks did not feature on the CG meeting agenda and yet the meeting took place a month after the first outbreak. With regard to donor community responses, Stef Vandenginste observes that political conditionality did not result in the curbing of ethnic clashes but provided the occasion for the ethnic clashes. Political conditionality was imposed without a careful consideration of how it would be used to justify KANU's political strategy, for instance KANU’s interpretation of conditionality as foreign support for the opposition.

The most vocal external critics of the government policy towards the violence were the international human and civic rights organisations, the most prominent being the Robert F. Kennedy Memorial Centre for Human Rights whose Director Ms Kerry Kennedy-Cuomo visited affected areas. Her statements on the clashes drew harsh criticism of her visit from the government. Apart from international NGOs , Dutch Parliamentarians tried to visit the affected areas but were denied entrance to the affected areas. A number of British Parliamentarians who visited the country at the invitation of prominent members of the opposition criticised the government for the clashes. Other civic and human rights organisations that published reports strongly implicating government complicity were the Kenya Human Rights Commission (KHR), Human Rights Watch/Africa, the National Elections Monitoring Group, the Commonwealth Observer Group and the Robert F. Kennedy Memorial Centre. Any attempts by individuals or human rights groups to lobby for the rights of the displaced were frustrated or obstructed by the government

To summarize, responses to the conflict reveal a strong government involvement in the conflict, even though its exact nature is not clearly evident. Although the opposition and various groups in civil society were highly critical of the conflict, and conducted investigations that corroborated reports of government complicity, some of the rhetoric, especially that of the opposition and a section of the independent press, indicates that they too fell prey to the same ethnic sentiments and biases they blamed the KANU government for. The conflict thus heightened ethnic hostility and facilitated the interpretation of literally everything in ethnic terms. As a political strategy, the violence interfered with the electoral process by preventing those over 18 (from the targetted ethnic groups) from obtaining national identity cards which are a requirement for voter registration, registering to vote, or even returning to vote where they had registered.

Conflict early warning is the early identification of conflict potential in individual societies, with the aim of preventing an outbreak of violent conflict through various third party interventions. It involves the identification of certain actions of parties to the conflict that indicate a potential for violent conflict in a particular society. Early warning indicators ae identified through an analysis of both the background conditions that make a society prone to conflict and the conjuctural factors that lead to the outbreak, escalation and destructive management of conflict. The background conditions examined include structural factors (entitlements, territory and social cleavage); psycho-cultural aspects (cultrual identity, myths, memories and social cleavages); institutional legitimacy and mobilisation; role of third parties (external actors). From an analysis of the dynamics of a conflict certain indicators can be identified, for instance, conflict accelerators (events which fuel horizontal and vertical escalation of conflicts); signal flares (statements and actions of parties that warn of increasing confrontation); conflict triggers (decisions or actions that undermine the stability and constructive management of conflict). Their interaction reveals the conflict-proneness of a society or the nature of an existing conflict. This section examines the background conditions to the conflict in the Rift Valley and Western part of Kenya and identifies indicators of the conflict proneness of this area. The background conditions examined are democratisation and institutional legitimacy; land ownership as a key structural factor in a mainly rural area; psycho-cultural factor of ethnic hostility.

The government is ultimately responsible for the security of its citizens. Unfortunately it was unable to prevent or control the ethnic attacks until it was too late. The reason for the government’s helplessness lies partly in the political developments that were going on during the period preceding the conflicts and the KANU government interpretation of these changes. However, the KANU government attitude is best understood when placed in the context of Kenya's shift from multi-party to one-party politics to which we now turn.

From "multipartyism" to "mono-partyism": the Kenyatta era

When Kenya became independent in 1963 under a multiparty constitution with a federal (majimbo) system of government, the two main political parties at the time were the Kenya African National Union (KANU) led by Jomo Kenyatta and the Kenya African Democratic Union (KADU) headed by Ronald Ngala. Support for the KANU came from the Kikuyu (Central and Rift Valley provinces), the Embu and Meru (Eastern), Luo (Nyanza), Kamba (Eastern) and Kisii (Nyanza). The KADU party drew its support from the Luhya (Western), the Kalenjin (Rift Valley), MijiKenda (Coast) and various nomadic groups from the arid and semi-arid areas of the Rift Valley and Northeastern Provinces. At the Second Lancaster House independence negotiations in 1962, KADU successfully pushed for a regionally-based constitution granting political autonomy for the country's eight provinces. On the other hand, KANU (and Kenyatta in particular) favoured a central government, but acceded to KADU demands for the sake of the 1963 independence elections.

KANU and KADU disagreement over what government system to adopt, arose out of a conflict of interests concerning what Joel Barkan calls distributive issues. The split in the nationalist movement was thus over economic interests and chiefly land. KANU's support base included three of the five dominant ethnic groups in Kenya: the Kikuyu, Luo, and Kamba. Members of these ethnic groups (particularly the Kikuyu) were more educated, more urbanized and more politically mobilised than those from the ethnic groups supporting KADU. In explaining the socio-economic advantage the Kikuyu had at the time of independenc, Cherry Gertzel observes that they were already settled in most parts of the country even before independence and were easily identified as a cohesive group in the places they settled, some of which are in the Rift Valley, home to the KADU supporting Kalenjin ethnic group. Their socio-economic advantage was to a great extent a consequence of the colonial experience which was more intense among them than any other ethnic group and also because of their geographical proximity to settler communities and the capital city, Nairobi. Despite losing much of their land to European settlers, and experiencing inter-ethnic divisions over land and economic differentiation, the Kikuyu stood to benefit greatly from the socio-economic development undertaken primarily for the settler community in their region. Their presence in the Rift Valley, however, aroused the resentment of the Kalenjin who, like the Kikuyu claimed ownership of land in the European Highlands in their own traditional home areas. The Kalenjin feared that with a Kikuyu dominated KANU, the Kikuyu would have the upper advantage on claims over the land. These fears (whether real or imagined) and the tensions they produced resulted in the KADU's demand for a federal type of government (majimbo) with a bicameral legislature in which the Senate would represent district interests and regional authorities would have autonomy over Trust land.

At the 1961 and 1963 pre-independence elections the KANU defeated the KADU and went on to form the new government in independent Kenya while KADU became the opposition. Kenyatta's main concern at the time was to tap into the support bases of both parties to strengthen the legitimacy of his government. With Kenyatta's influence the KANU government amended the constitution eighteen months after independence to provide for a unitary system of government in place of the regional (majimbo) arrangement. In what Barkan calls a combination of carrots and sticks, several leaders of KADU were persuaded to join KANU in 1964. The KADU deputy chairman Daniel Arap Moi was later rewarded with the post of Vice President of KANU and of Kenya in 1964, for his instrumental role in helping diffuse the Kikuyu land problem Kenyatta found himself with, by ensuring Kalenjin support for government land settlement policies in the Rift Valley. It is evident from the foregoing that the ethnic element qua 'ethnicization' of issues played a significant role in Kenyan politics because of the nexus between ethnic and economic interests, the colonial legacy of unequal socio-economic development in the country, and political support and mobilisation along ethnic lines. Yet, the fears and tensions especially over land claims did not break out into the kind of violence and expulsion of non-Kalenjin ethnic groups settled in the Rift Valley experienced between 1991 and 1995.

For the next thirteen years after 1964, Kenya became a de facto one party state. However, Kenyatta, progressively diluted party power by developing his own patron-client hierarchical network of ethnic regional leaders, and making full use of the central administration to consolidate and maintain his political hold. Once again the ethnic element in politics, in this case the 'ethnicization' of state institutions, is manifest. In the first place, the Africanization of the Provincial Administration positions favoured the Kikuyu more than other ethnic groups. This allowed Kenyatta to exercise great influence in the country and it strengthened the Presidency. Second, Kenyatta increasingly relied on a Cabinet and inner circle of advisors who were mostly Kikuyu to rule the country. Despite this concentration of power in institutions dominated by the Kikuyu (particularly the Kiambu Kikuyu), a number of factors diffused much of the resentment other ethnic groups felt. Kenyatta's charismatic appeal and stature as "Mzee" (the Old One) earned him nationwide support. His ethno-regional patron-client network, though more favourable to the Kikuyu, included leaders from non-Kikuyu ethnic groups who ensured support of their people for the Kenyatta regime. In addition, ethnic organizations formed during the colonial period to promote and defend group interests were allowed to continue, the Harambee (self help) movement was encouraged, and Members of Parliament enjoyed a measure of freedom to criticize government policies. All this would change during Moi's regime.

When Moi assumed power after Kenyatta's death his first task was to build a strong political power base. He established control over the Civil Service, especially the Provincial Administration, by replacing Kenyatta's appointees with his own. Moi's appointees were from ethnic groups that made up the old KADU party, but most were from his own ethnic group, the Kalenjin. To keep officials from acquiring influence with time, Moi reduced the tenure of office to two or three years and rotated personnel frequently. He went on to build his own patron-client networks even as he neutralized those Kenyatta built. Whereas Kenyatta had by-passed the KANU, Moi revitalized and mainstreamed it, using it as the institution through which his networks would be built. By so doing he undercut the power of established ethno-regional political leaders and made the party an instrument of personal control.

In 1982 fearing opposition from those affected by his reorganization of political power, Moi sponsored through KANU, a constitutional amendment making Kenya a de jure one party state. Critics of this amendment were detained and when an attempted coup took place in August 1982, Moi's government became even more repressive. Any critic of government policies was branded disloyal. Those leaders whose loyalty was in question were expelled from KANU. Parliament lost the status of 'friendly critic' it enjoyed during Kenyatta's time because members feared being branded disloyal. Moi began to rely more on a small circle of advisers, loyalists from former KADU strongholds in the Rift Valley. Political power had been removed from Kenyatta's close associates (mainly Kiambu Kikuyu), and redistributed among former KADU supporters, especially the Kalenjin. Finally though elections were carried out regularly, candidacy depended on who was seen to be loyal to Moi. This was more evident in the 1988 General Elections when several accounts of rigging in favour of candidates loyal to Moi were reported.

Moi's fear of real or imagined threats to his political control extended to civil society, where independent centres of power such as ethnic welfare organizations were banned. The independent Harambee movement was submitted to the regulation of the Provincial Administration and soon became an instrument of personal rule. Interest groups like the Kenya Farmers Association (KFA), the Kenya Planters Cooperative Union (KPCU) and the Coffee Board of Kenya (CBK), were either dissolved and replaced with others, or restaffed with Kalenjin in place of Kikuyu. Ironically, the economic and political disempowerment of the pro-Kenyatta Kikuyu elite had the adverse effect of interfering with the important technical support the above mentioned organizations offered farmers country wide.

Moi's redistribution of resources away from ethnic groups that supported the original KANU was not limited to the political realm. During Kenyatta's regime, public expenditure for roads, health, water and education to a large extent favoured the Central Province home to the Kikuyu. Moi reversed this situation by shifting resources to other areas, giving priority to the Rift Valley areas inhabited by the Kalenjin. Agricultural policy was changed so that it favoured tea and cereal farmers in the Western and Rift Valley regions. A quota system was introduced for enrolment in secondary and tertiary educational institutions to facilitate increase in the number of students from unrepresented or under represented areas. State owned parastatal corporations were staffed with Kalenjin in place of the Kikuyu and other ethnic groups. The growing African business class was frustrated as attempts were made to undermine their business activities. This group was predominantly Kikuyu as it had gained more access to opportunities, under Kenyatta, to acquire credit. By 1990 Moi had to a great extent managed to deconstruct the Kenyatta legacy and construct his own political and economic power bases but at a great cost. The economic impact of his 'ethnic balance' policy began to be felt in the late 80s and early 90s as reports of economic mismanagement, corruption and financial scandals involving members of the pro-Moi Kalenjin elite close to the President were made public.

Return to multi-partyism

Although these policies generated a great deal of discontent, an evolving culture of sycophancy and fear kept politicians and members of the public from mobilising to protest. Anyone who attempted criticism was detained or harassed by security forces. Except for the clergy. Indeed the Church (especially the Roman Catholic church and the Protestant Churches under the umbrella of the National Council of Churches in Kenya - NCCK) played an important role in speaking out against corruption, administrative incompetence, injustice, disregard for the rule of law, nepotism, and ethnic favouritism. The Church thus took on the role of an opposition party. This did not appear to threaten the Moi government since it had consolidated itself and was adept at silencing criticism. But, the situation changed in 1990, when soon after the collapse of communist regimes, a number of well known politicians critical of the regime joined the clergy and the Law Society of Kenya (LSK) in calling for the return to constitutional governance, multi parties, and rule of law. Moi's government responded to this resurgence of opposition by arresting its critics and detaining them under the Public Security Act. At the same time Moi pointed out that advocates of multiparty politics were out to foment ethnic conflict.

The arrests and detentions only strengthened criticism of the government, as lawyers began highlighting these government measures as human rights violations. For most of 1990 and much of 1991, Moi rejected calls to repeal the one party state amendment and vowed to crush the opposition. The situation changed in favour of advocates for democratization when the CG met with the Kenyan delegation in Paris to negotiate extended development funding. This time the CG decided to withhold about $350 million in fast-disbursing aid until the government initiated political and economic reforms that allowed for accountability and transparency. Because the Kenyan economy relies a great deal on economic aid from the West and the CG decision was a major blow for the KANU government. About two weeks after the CG meeting in November, 1991, the KANU Governing Council met and voted unanimously for the repeal of the amendment making Kenya a One-Party State. In addition the President announced that multiparty elections would take place in 1992. The way was now open for the formation of other political parties. In December 1991, the Forum for the Restoration of Democracy (FORD) was registered as a political party, followed by several others.

Studies in democratisation reveal that the process can either generate or diffuse tensions in multiethnic states. The movement for multiparty politics and democratic reform entailed negotiating over the nature of future political and economic institutions, a process that would result in the restructuring of existing institutions and dismantling of the KANU regime's political power base. On that account a number of KANU leaders were not willing to consider or accommodate these changes. Renée de Nevers argues that, "For democratization to reduce ethnic tension, the inclusion of all relevant groups in the negotiating process is required : in addition, there must be a willingness by all parties to work for, and then accept, a mutually beneficial agreement." Unfortunately, this did not happen in Kenya. From the outset the government adopted a non-negotiable attitude, rejecting demands for political reform and even dialogue with civic groups, the church, and the opposition. Any attempt by the latter to mobilise support for democratisation or dialogue with the government were obstructed and activists suffered intimidation by security forces. On the other hand KANU government initiatives were criticised by the opposition and civic groups. To date there has not been a national referendum bringing together all parties to negotiate over a democratic Kenya.

Powerful KANU leaders, most from the Rift Valley did play an influential role in obstructing the democratisation process. De Nevers observes that, "More often than not......group leaders feel that either their personal political aims or those of the group they represent will be better served by......exacerbating tensions or raising the level of violence." The Rift Valley leaders did this when they interpreted the democratic concept of majority rule in raw majority terms: rule by the ethnic group (or groups) with the largest number of people. Thus, prospects of a Rift Valley with a large number of Kikuyu and other non-Kalenjin groups most likely to vote for the opposition presented neither a comfortable nor exciting prospect for the Rift Valley leaders. There was the possibility that the opposition (read dominant ethnic groups) could win a number of seats large enough to put them in power if elections were called. This 'ethnicised raw majoritarian' view of multiparty elections was communicated to those who are natives of the Rift Valley region as one that should be countered by all means necessary to prevent it from becoming a reality. Thus the majimbo rallies and the instrumental nature of the ensuing conflicts.

The clashes most probably served to frighten and intimidate the non-Kalenjin in the area into supporting the regime. It worked to a great extent. Large numbers of non-Kalenjin ethnic groups were displaced from the Rift Valley before the elections. This interfered with the issuing of identity cards to youth aged 18 and above, a crucial exercise that allows them to register as voters in the region. It also interfered with the voter registration exercise since many of the displaced who had fled from the province or were in camps in the province and feared return to their homes, did not register to vote. Those who had registered but feared for their lives if they returned to vote were denied the right to vote. The ensuing ethnic tensions and hostilities worked against the opposition parties which found it difficult to campaign in this province. On the other hand, the fracturing of the Forum for the Restoration of Democracy (FORD) into two parties ( FORD-Kenya and FORD-Asili), and the registration of numerous other parties added to the fears of ethnic domination. This development in opposition politics ironically played into the hands of the Moi regime. Thanks to the increasing fratricidal nature of opposition politics, the opposition ceased to be a united movement for democracy, appearing more and more like a collection of numerous disunited groups with ethnic and personal interests. To a great extent, the opposition, perhaps unwittingly, fulfilled Moi's prophecy concerning multiparty politics. This, in addition to the clashes and an aggressive national campaign facilitated KANU's victory in the Rift Valley and Moi's re-election as President.

Kenya is an agricultural based economy with 80% percent of people engaged in an agriculturally related activity whether farming or ranching. Land is thus a very important economic asset, if not the most important. Apart from its economic value, land has an identity value because of the way people specify their origin terms of the geographical area their ethnic group is located. These two competing claims on land (identity and welfare), make it a highly sensitive and political issue, one that can be easily manipulated to fuel conflict. As we saw in an earlier section, the government initially attributed the violence to ethnic disputes over land ownership. Yet, disputes over land ownership in the Rift Valley are not recent phenomena. They have their roots in the colonial period, when the colonial government's policy of encouraging European settlers, alienated Africans from their land in the eastern, western and southern areas of the Rift and parts of the Central province. Although a number of measures were undertaken to address the ensuing conflicts these did not resolve the land question.

The Maasai were the first to experience this alienation. Beginning in 1904, with the disputable Maasai agreement, they were progressively dispossessed of their land. As a result of this agreement and subsequent colonial policies, the Maasai lost valuable fertile land to European settlers and were moved to the southern more arid areas of the Rift Valley. The land from which the Maasai were alienated did not immediately attract as many Europeans settlers. Most preferred to settle in Kiambu and the areas between Nairobi and Kisumu, along the railway line. This policy, together with the post-World War I and II Soldiers Settlement schemes and other land policies, further alienated the Kikuyu, Nandi, Kipsigis, Sabaot and Machakos Kamba. To keep indigenous people from this 'European' only land (White Highlands), the colonial government established Native Reserves to which the Africans were restricted. By 1960, over seven and a half million acres in the Highlands of the Rift Valley, had been alienated for use by about 4,000 settler farmers.

The colonial land and agricultural policies created a number of problems. The pastoral economies of the nomadic peoples such as the Maasai, Turkana, Sabaot and Kalenjin which depend on the availability of vast amounts of grazing land were adversely affected by the shrinking land and its poor quality. For the Maasai this spelled the beginning of what Hillman calls their pauperization, a trend that has continued to date. In addition, thousands of Africans, mainly Kikuyu found themselves squatters on their own land. They constituted 'free' labour for the European settler. These early squatters were soon joined by others fleeing the harsh life in the Native reserves. At first the squatter system served both groups well. The squatter had land for use while the settler gained from both the use of the squatters labour and produce. The situation changed when the settlers shifted to mixed farming which required more land and increased labour input. The demands arising from this shift conflicted with those of an increasing population of squatters (more land for cultivation and more time to cultivate it) and resulted numerous evictions of squatters from settler farms.

The colonial government tried to solve the problem by purchasing land from the Maasai to settle the Kikuyu. This land, however, was of poor quality and the Kikuyu refused to move. According to Zeleza Tiyambe "... this was not to be real solution. Not only was the land very poor in quality, but the Kikuyu, who constituted the largest portion of the evicted squatter population, refused to move to these lands. With over 30,000 evicted and landless squatters by 1939, the settler economy had sowed the seeds of rural discontent and the eventual demise of settler power itself." This rural discontent to a great extent contributed to the outbreak of the Mau Mau war. The Kikuyu because of their close proximity to the urban centres and the colonialists were the most affected and first to mobilise politically and call for the return of their land. The Mau Mau resistance thrust the land question to the fore and was instrumental in paving the way for changes that would lead to Kenya's independence.

Colonial government attempts to resolve the land question were either too late or created another set of problems. Thus at independence the Kenya government inherited not only alien land and agricultural policies, but also the problems created by their implementation : the landless Kikuyu, the land starved Maasai and other nomadic groups and agricultural groups like some of the Luhya, the aggrieved Kalenjin groups alienated from their land. After independence in 1963, President Jomo Kenyatta was confronted with the serious land problems facing not only his people the Kikuyu but also other ethnic groups. To solve the squatter issue, the government facilitated the settlement of thousands of Kikuyu on land purchased from European settlers in the Highlands east and west of the Rift Valley, traditionally home to the Kikuyu, Kalenjin and Maasai. Many Kikuyu, however, were settled in the Rift Valley in areas historically inhabited by the Maasai and the Kalenjin. The Kikuyu, though the main beneficiaries of the settlement schemes in the Rift Valley, were not the only ethnic group to settle there. Many Luo, Luhya and Kisii purchased land here too. Thus Kenyatta solved a problem that posed a threat to his legitimacy but at the same time, left unresolved one that would be politicised to serve the interests of the Moi regime several years later: the socio-economic impact of modern land tenure systems on pastoral communities.

Land disputes are a major problem and not just between ethnic groups. Most of the disputes over ownership are either intra-ethnic or have a class dimension to them. Family and clan conflicts over ownership and use of land fall into the intra-ethnic disputes category. The 'class' disputes involve wealthy and politically influential leaders or directors of land buying companies who have been accused of cheating or grabbing land from ordinary Kenyans, thus depriving them of opportunity to purchase it or having user-access to it. The latter phenomenon has been on the increase beginning the early 80s and is frequently reported in the press. In one such report, a powerful Rift Valley KANU politician was embroiled in a dispute over land which several local residents in his constituency claimed belonged to them. The class dimension to the land disputes, involving the wealthy and politically powerful on the one hand, and the poor and ordinary Kenyan on the other, is probably more representative of reality than the inter-ethnic dispute rationale.

It is significant that during the one-party rule, none of these disputes resulted in an outbreak of violence such as that which witnessed between 1991 and 1995. A number of factors partly explain this. In the first place, the intra-elite political bargaining that went on during the initial years of independence, especially Moi's brokering of the land issue between the Kikuyu and the Kalenjin helped ensure a peaceful if not acceptable settlement of the Kikuyu in the Rift Valley. Secondly, any tension over land threatening to break into violent conflict was swiftly dealt with by government in both the Kenyatta and the Moi regimes. In one such incident between the Luhya and the Nandi (members of the Kalenjin ethnic group) in Tindinyo, Kapsabet, in the mid 80s, security responded swiftly to stop the fighting and the President visited the area to address a public rally in which he ordered that peace be restored. Finally, the President has on numerous occasions, personally intervened to settle disputes within land buying companies and organised for the allocation of land to members most of whom have saved for years to purchase the land. Although this may have served to promote his political interests, the action on the part of the President did much to diffuse tensions over land. On that account it is ironical that at the peak of the movement for democratic reform and return to multiparty politics, land disputes should trigger off conflict in the Rift Valley, when ample evidence exists for alternative ways of dealing with the dispute.

The ethnic element has always been, and continues to be, one of the salient features of Kenyan politics. A combination of factors account for this: the existence of " relatively ethnically homogeneous geographical spaces" in pre-colonial Kenya; the colonial policy of divide and rule which intensified ethnic separation by establishing administrative jurisdictions and land tenure along ethnic and racial lines; and, two ethnically biased post-independent regimes that have legitimized ethnic representation. Some of these factors have promoted certain ethnic cleavages, for instance the `ethnicization' of land which is a legacy of the settlement patterns pursued during the colonial period. In addition, the division of administrative jurisdictions along ethnic lines, and the concentration of economic development to areas inhabited by Europeans during the colonial period, together with the dependence of Kenyan political leaders on ethnic political support bases resulted in the `ethnicization' of economic and political resources. It is against this background that the KANU politicians inflammatory statements and war of words with the opposition gain significance with regard to the conflict in the Rift Valley.

The ethnically biased inflammatory statements made by KANU politicians from the Rift Valley and the revival of the majimbo (regional) debate are largely responsible for creating ethnic tension in the months preceding the outbreak of conflict in the Rift Valley. The majimbo debate was initiated when the MP for Eldoret South at that time, Dr. Joseph Misoi called a press conference at the parliament buildings in August, 1991, to announce that a bill on the majimbo system of government would soon be tabled in the National Assembly for discussion. Soon after the announcement a KANU MP, Mr. Joash wa Mangoli while addressing a fundraising meeting in his constituency said that "We in the leadership have been disturbed for a long time now by some few individuals clamouring for pluralism.... To silence them, a federal government should be introduced under the leadership of President Daniel Arap Moi." The proposal, which was supported by politicians from the Rift Valley, generated heated debate and even threatened to divide KANU politicians. A substantial number of them strongly support a unitary system of government. At that point, President Moi intervened by calling for an end to both majimbo and multiparty debates because they were dividing the country rather than encouraging peace and national unity.

The debate was shelved for most of 1992 while politicians concentrated on campaigns for the December 1992 elections. It was revived in 1993, this time in response to an attempt to revive the Gikuyu, Embu and Meru Association (GEMA), one of the ethnic based welfare organisations that existed during the Kenyatta regime but was banned soon after Moi came to power. On examining the rationale for the revival of majimboisms it is evident that this revival was more of a campaign aimed at neutralizing the movement for democratisation and multiparty politics, and a tactic for uniting Rift Valley leaders against activities of the opposition whose existence Rift Valley leaders interpreted as divisive politics. One of the reasons given for its revival was that the 'small tribes' needed to unite against the domination of the 'big tribes' who were advocating for multiparty politics, just as they had done in the early years of independence. To fan support for this revival, the Rift Valley politicians exhumed KADU grievances, long buried after the crossover of the party to KANU in 1964. These were repackaged for use as framework for interpreting the return of multiparty politics and ongoing calls for democratic reform, so as to mobilise 'small tribes' against these new developments.

This anti-multiparty-politics-democracy attitude and its ethnically loaded character was evident at a series of public meetings convened in the Rift Valley province during the months of September and October, 1991. At the Kapsabet and Kericho rallies, organised by leaders from the Kalenjin community, speakers called for the reintroduction of a majimbo system, warned non-indigenous communities resident in the region to support the KANU or risk eviction, cautioned individual multiparty activists not to set foot in the province, and called on youths in the Rift Valley to arm and drive out non-Kalenjin, non-Maasai and non-Pokot. One such rally in Narok took on a more combative character. Several cabinet ministers, assistant ministers, members of parliament and district KANU chairmen from areas other than the Rift Valley attended and gave speeches. Members of the FORD were warned that they would be `chased out' of areas in which KANU branches were located. The convener of the meeting, Mr. William Ole Ntimama vowed to mobilise local Maasai and other residents of the Rift Valley to scatter FORD members. Earlier on Mr. Ntimama had called on the Maasai to arm themselves with spears and rungus (clubs) and kick out advocates of multiparty politics in Narok. Although the rally had been called to show support for the President it turned out to be not unlike a hate campaign against the advocates of democratisation and multiparty politics movement and by implication the ethnic groups to which these advocates belonged.

The speeches are crucial for what they communicated to the people who attended the rallies, particularly how they interpreted the ongoing reforms. By interpreting the political reforms as an anti-government, anti-President Moi and thus anti-Kalenjin campaign they `ethnicised' political discourse on the reforms, creating a situation of `them' against `us'. According to the Kiliku report, the majimbo rallies propagated the theory that the Rift Valley was for Kalenjin and those in political parties other than KANU should leave. The report also notes that the Kalenjin were made to believe that anyone who was anti-KANU was against them and must be resisted at all costs. This ethnicisation is more evident in the majimbo debate which was seen as the only way to counteract the threats posed by the opposition. Yet, its appeal to the ethnic 'purity' of communities and geographical space is more of a fiction and incongruous with the historical fact that for more than two decades, different ethnic groups have been living in the Rift Valley and parts of Western Kenya and have even intermarried. For even a longer time 'interface communities' have existed on the boundaries of areas inhabited by different ethnic groups (for instance, Kisii/Kipsigis, Maasai/Kikuyu, Luhya/Nandi). In these boundary areas, people from different ethnic groups have intermarried and traded with each other (despite the periodic wars that took place) even during pre-colonial times. Most, if not all ethnic groups in Kenya are, in historical terms, recent migrants to the areas they now inhabit. Unfortunately, the majimbo debate did not highlight this choosing instead to appeal to ethnic bias and by so doing, intensified ethnic suspicions, hostility and facilitated polarisation of ethnic communities.

The Kiliku report attributes the success of these 'ethnic entrepreneurs' to the ordinary Kenyan's ignorance of the newly introduced multiparty politics. So did the President's 'predictions' that multiparty politics would lead to war. It comes as no surprise that the violence and resultant increase in ethnic tension has been blamed on these rallies. Even then, it is important to distinguish between two dimensions of the conflict: the rhetoric or 'war of words' (inflammatory statements, rumours, warnings on risk of violence which inflamed public opinion) and real (physical violence - attacks, killings, burning and looting of property). This distinction helps in identifying whose war it was. At the rhetorical level (which preceded the real war), the conflict was mostly confined to the political elite both in the opposition and in the KANU and was most intense. Recourse to ethnic sentiments, which fanned the flames of hostility and suspicion, is understandable when analysed within the framework of the ethnic nature of the patron-client network of most African regimes, where political support bases are often a combination of the ethnic and economic. At this level the war was for the political elite fearful of losing the power and influential positions they held at both national and local levels.

On the other hand the physical violence, as mentioned elsewhere was not spontaneous in character, that is, it did not automatically translate into a logical conclusion of the rhetorical 'war but appeared to 'authenticate' it. Unlike the rhetorical conflict, this physical conflict did not seem to have a life of its own, was episodic, petered out just before the elections, took place soon after inflammatory statements in self-fulfilment of earlier 'prophecies' and had more of a punitive nature. The Kiliku report notes that numerous investigations carried out by the Church and human rights groups, and eyewitness accounts reported in the independent press provide strong evidence of this. According to this and other independent reports, the aggressors were armed youth gangs between 14 and 30 years of age, well trained and armed with bows and arrows, led by army personnel on leave, wore similar attire, were not from the local population or familiar to locals and were transported to the area by vehicles belonging to the government or influential KANU politicians. The emerging picture is that of a planned attack on non-Kalenjin ethnic groups that was camouflaged as an ethnic conflict. It therefore begs the question of to what extent the conflict can be labelled 'ethnic' and real.

EARLY WARNING INDICATORS AND CONFLICT PREVENTION MECHANISMS

Our analysis of the background conditions to the conflict and its denouement reveal that the loudest warnings of a possible conflict in the Rift Valley and Western Kenya were the calls by Rift Valley KANU politicians and MPs for the revival of a majimbo government, inflamatory ethno-centric statements made at rallies in the Rift Valley and Moi's argument that multiparty politics would lead to ethnic violence. Janie Leatherman and Raimo Vayrynen observe that although psychocultural tensions are difficult to measure in cross-cultural investigations, their escalatory potential can be gauged by the existence of historical memories which capture the essence of the conflict and keep the fear and animosity alive. Much as ethnicity is a potent political resource for the ethnic entrepreneur, its existence alone does not, automatically, translate into conflict. This depends on the degree of ethnic tension and ability of ethnic entrepreneurs to mobilise the group against another. The KANU campaign discourse framed the movement for democratic reform and multiparty politics in mutually exclusive terms as a movement of "them" (non-Kalenjin-big tribes) against "us" (minority tribes). It is difficult to quantify the degree of ethnic tension and level of ethnic polarisation prevailing in Kenya as a result of the ethnic clashes and ethnicised political discourse. What the ethnic clashes have proved is that although it is easier now than before to appeal to ethnic grievances and mobilise some ethnic groups against others, the situation will, however, differ from one ethnic community to another and will depend on the level of intra-ethnic rivalry. The more intense the intra-ethnic rivalry (at sub-ethnic or clan levels), the more difficult it is for ethnic entrepreneurs to mobilise the community against other ethnic groups.

One of the structural factors that increases conflict proneness in society is the distribution of economic and political resources. In agricultural societies like Kenya, in general, and the Rift Valley, in particular, the distribution of land and adequate access to it is an important indicator of conflict potential. The economic and identity value of land in Kenya make not only a vital economic resource but also a politically volatile issue because of its identity (ethnic) value. The history of land disputes is at the same time a history of inequities in socio-economic development of different ethnic groups in Kenya that began with the imposition of colonial rule. These inequities have intensified during independent Kenya because of the high population growth, competition for the scarce and marginal pastoral land, and the retention and expansion of colonial agrarian laws and institutions which have progressively marginalised the pastoral communities from the mainstream economy. The land question in Kenya evokes feelings of grievances over ownership and access. These grievances are not unique to a particular ethnic group but are experienced by almost all ethnic groups which place a high value on land ownership, with varying degrees of intensity. These grievances are an easily available political resource for ethnic entrepreneurs to consolidate their position in the threatening world of multiparty party politics and democratic reform. They were manipulated to trigger off conflict and subsequently developed into a campaign by pastoral communities, especially the Maasai, for the return of their land and a reform of current agrarian laws and institutions. The land question remains a volatile issue and the government's neglect in addressing it is an indicator of potential for further conflict not only in the Rift Valley but also in other areas, especially the pastoral areas of northern and eastern Kenya and the urban slum areas.

Psychocultural and structural indicators of conflict can be tolerated as long as social and political institutions are perceived to be legitimate and just. The breakdown of political and institutional legitimacy is an indicator of society's slide into turmoil as was the case in Uganda and Somalia during the 70s and 80s respectively. In Kenya, the 'orchestrated' nature of the conflict in the Rift Valley exposes the extent to which these political institutions could be manipulated by political leaders and their proper function distorted. It also reveals how existing mechanisms for conflict prevention can be reversed to serve the interests of ethnic entrepreneurs. As argued elsewhere, the conflict potential of unequal land access and land disputes, ethnic hostility has been decreased by a number of domestic conflict prevention measures (local administration and security institutions, the Public Security Act, traditional elders) and the skilful brokering of the President. These conflict prevention institutions did not break down during the 1991 -1995 period but were reversed to serve the interests of the KANU regime. The local population's confidence in the capacity of the local administration, security, and courts to prevent, manage and resolve the conflicts was seriously eroded. They could not be viewed as representative of all Kenyan citizens regardless of ethnic identity. To restore this confidence will require at the same time the restoration of the independence of these institutions at the national and local level. None the less, this erosion of legitimacy has been tempered by a measure of tolerance since the loss of confidence is not complete nor is it universal but differs from region to region with the most affected areas being those where conflict took place. The potential for a total breakdown of social and political institutions will become real if there is a general and complete loss of confidence in the population.

Apart from use of state institutions, other preventive measures were the surveys and fact-finding carried out by the Parliamentary Committee to investigate the conflict in Western and other parts of Kenya, and the Churches (National Council of Churches in Kenya-NCCK, and the Catholic Bishops). These reports, unfortunately did not result in concrete action to resolve the conflicts since debate on the findings did not take place, nor were recommendations implemented. The non-negotiable and intransigent position of the government did not allow for this. The reports, however, publicised the state of the situation in the affected areas drawing international and national attention to what was happening and also put some (like the UNDP Rogge report) pressure on the government to respond positively. Although the government bears much of the blame for not facilitating an environment conducive to negotiations and debate, one cannot help but wonder whether the human rights activists, opposition leaders, and the church bear some blame for this. In other words, to what extent did their response contribute to this intransigence? A critical self-analysis may throw light on this.

One interesting development on the use of conflict prevention mechanisms is the victim's use of the Church, and international and local human rights groups (and also the church) to appeal to the government to restore peace. The Church, in particular, assumed a significant role of speaking on behalf of the community, one that dwarfs that of elders (whom the President called upon, much later, to initiate peace and reconciliation meetings in the affected regions in conjunction with the local administration). The churches through the NCCK, not only condemned government response to the conflict, but also petitioned the President and undertook peace-building initiatives in areas affected by the conflict. The role of the Church is significant because of it represents a new or 'modern' community in a number of migrant community areas, especially where traditional influence of elders is weak. Perhaps the relatively 'weak' role of the traditional elders in migrant community areas could be explained by the heterogeneous socio-cultural environment the migrant ethnic groups find themselves in, which is different from that of their ancestral homes. Another reason may be the gradual of substitution of the traditional authority structure with that of the Church and local administration. And yet, the moral authority of the Church was abused when some local church leaders gave tacit support to the ethnic conflict.

At the international level, the diplomatic and donor community engaged in some form of preventive action. This included government aid conditionality for political and economic reform (Consultative Group, Scandinavian governments), diplomatic encouragement (the most vocal being the American Ambassador Smith Hempstone) and the monitoring and condemnation of violence by international human rights groups working with local Kenyan counterparts. As argued elsewhere, the aid conditionality worked with the non-negotiating stand of the government to fuel conflict rather than resolving it by providing the occasion for the ethnic clashes.

The above preventive actions were in many ways 'adhocist' and at the same time reveal the frustrating task conflict prevention is with regard to ethnic conflicts. Traditional instruments of conflict prevention such as direct negotiations, fact-finding, third-party negotiation, conciliation, and peaceful settlement of disputes appear to be ineffective in a political context (ethno-patron-client system) such as was existing at the time of the conflict. Creative ways of applying these preventive and management mechanisms are required with regard to 'ethnic' conflicts in Africa. However, before this can be done, a thorough understanding of the socio-political and economic dynamics of the conflict context and identification of early warning indicators is necessary.

Africa Watch.

1993. Divide and Rule: State Sponsored Ethnic Violence. London, November.

Apter, David E. And Carl G. Rosberg, eds.

1994. Political Development and the New Realism in Sub-Saharan Africa. Charlottesville: University Press of Virginia.

Barkan, Joel D, ed.

1994a. Beyond Capitalism vs. Socialism in Kenya and Tanzania. Boulder: Lynne Reinner Publishers.

1994b "Divergence and Convergence in Kenya and Tanzania: Pressures for Reform," in Barkan 1994a:1-45.

1992. "The Rise and Fall of a Governance Realm in Kenya," in Hyden 1992:167-192.

1984. Politics and Public Policy In Kenya and Tanzania. New York: Praeger Publishers.

Barkan, Joel D and Michael Chege.

1989. "Decentralising the State: District Focus and the Politics of Reallocation in Kenya," in The Journal of Modern African Studies, 27, 3, pp. 431-453.

Bienen, Henry.

1974. Kenya: The Politics of Participation and Control. Princeton, New Jersey: Princeton University Press.

British Broadcasting Corporation.

Var. Summary of World Broadcasts, The British Broadcasting Corporation, London.

Brittain, Victoria.

1993. Kenyan Election Not Free or Fair. The Guardian, 25 January.

Chege, Michael.

1994. "The Return of Multiparty Politics," in Barkan 1994:47-74.

Clayton, Jonathan.

1992. Kenya sets multiparty elections for December 7. The Reuter Library Report. 3 November.

Daily Nation (Nairobi).

Var. 1992-1995.

Dorff, Patricia Lee.

1992. America and the world 1992/1993. Foreign Affairs, December.

EcoNews Africa.

1994a. Vol.3, No.18, p.1-3.

1994b. Vol.3, No.5, p.5-7.

1993a. Vol.2, No. 17, p.3,5.

Economist.

1993. The Economist. The Economist Newspaper, Ltd., 20 November.

Facts on File, Inc.

1992. Facts on File World News Digest.

Gertzel, Cherry, Maure Goldschmidt and Don Rothschild, eds.

1969. Government and Politics in Kenya. Nairobi: East African Publishing House.

Gertzel, Cherry.

1970. The Politics of Independent Kenya. Evanston: Northwestern University Press.

Gordon, David F.

1986. Decolonization and the State in Kenya. Boulder, Colorado: Westview Press.

Gurr, Ted Robert, and Barbara Harff.

1994. Ethnic Conflict in World Politics. Boulder: Westview Press.

Harbeson, John W., Donald Rothchild and Naomi Chazan, eds.

1994. Civil Society and the State in Africa. Boulder, Colorado: Lynne Reinner Publishers.

Harbeson, John W. and Donald Rothchild, eds.

1995. Africa in World Politics: Post-Cold War Challenges. Boulder: Westview Press.

Haugerud, Angelique.

1995. The culture of politics in modern Kenya. Cambridge: Cambridge University Press.

Hillman, Eugene.

1994. "The Progressive Pauperization of the Maasai in Kenya," in Africa Today, Vol.41, No.4, pp. 57-65.

Holmquist, Frank and Michael Ford.

1994. "Kenya: State and Civil Society the first year after the election," in Africa Today, Vol. 41, No.4, pp. 5-25.

1992. "Kenya: Slouching Toward Democracy," in Africa Today, Vol.39, No. 3, pp.97-111.

Human Rights Watch/Africa.

1994. "Kenya: Multipartyism Betrayed in Kenya." Vol.6, No.5.

Hyden Goran and Michael Bratton, eds.

1992. Governance and Politics in Africa. Boulder, Colorado: Lynne Rienner Publishers.

Hyden, Goran.

1994. "Party, State, and Civil Society: Control versus Openness," in Barkan, 1994:75-95.

Indian Ocean Newsletter.

1993. The Indian Ocean Newsletter. Indigo Publications, 30 October.

Jackson, J. Robert and Carl G. Rosberg.

1982. Personal Rule in Black Africa: Prince, Autocrat, Prophet , Tyrant.

Berkeley: University of California Press.

Kahoso, Raphael.

1992. "The main causes of the ethnic clashes," The Standard, Nairobi, March 23, p.16 and 17.

Kenya Human Rights Commission.

1994. Independence Without Freedom: The Legitimization of Repressive Laws and Practices in Kenya. A Kenya Human Rights Commission Report, Nairobi, February.

Leatherman, Janie and Raimo Vayrynen.

1995. Structure, Culture and Territory: Three sets of Early Warning Indicators.

Paper presented at the International Studies Association (ISA) Convention, Chicago, 21-25 February 1995.

Lemarchand, René.

1994. Burundi: Ethnocide as discourse and practice. New York: Woodrow Wilson Centre Press and Cambridge University Press.

Lentz, Carola.

1995. " `Tribalism' and ethnicity in Africa," Cahiers des Sciences Humaines: Identités et Appartenances dans les Sociétés Sahéliennes, Vol.31, No.2.

Miller, Norman and Rodger Yeager, eds.

1984. Kenya: The Quest for Prosperity. Boulder, Colorado: Westview Press.

Moss, Todd J.

1995. "US Policy and Democratisation in Africa: the limits of Liberal Universalism," in The Journal of Modern African Studies, Vol.33, No. 2, pp189-209.

Mutua, Makau.

1994. "Human Rights and State Despotism in Kenya: Institutional Problems," in AfricaToday, Vol. 41, No.4, pp 50-56.

Ochieng, William R.

1990. Themes in Kenyan History. Nairobi: Heinemann.

Ochieng, William R. and R.M. Maxon. eds.

1992.An Economic History of Kenya. Nairobi: East African Publishers Ltd.

Okoth-Ogendo, H.W.O.

1991. Tenants of the Crown: Evolution of Agrarian Law and Institutions in Kenya. Nairobi: African Centre for Technological Studies.

Opala, Ken and Gitau Warigi.

1995. Ntimama: Is he too hot to tame? Sunday Nation, 15 January, p.7.

Selassie, Bereket H.

1974. The Executive in African Governments. London: Heinemann.

Schatzberg, Michael, ed.

1987. The Political Economy of Kenya. New York: Praeger.

The Reuter Library Report.

Var. Reuters News Reports.

Throup, David W.

1987. "The Construction and Destruction of the Kenyatta State," in Schatzberg, 1987:48-53.

U. S. Department of State.

1991. 1990 Human Rights Report: Kenya.

1992. 1991 Human Rights Report: Kenya.

1994. 1993 Human Rights Report: Kenya Human Rights Practices, 1993.

1995. 1994 Human Rights Report: Kenya Human Rights Practices, 1994.

Vandeginste, Stef.

1994."Development Co-operation: A tool for the Promotion of Human Rights and Democracy? A case study of Kenya," Draft version of a research report, University of Antwerp.

Vayrynen, Raimo.

1994. Towards a Theory Of Ethnic Conflicts and their Resolution. Notre Dame: Joan B.Kroc Institute for International Peace Studies.

Weekly Review.

1995. The Weekly Review, 13 January.

Widener, Jennifer A.

1992. The Rise of a Party-State in Kenya: From "Harambee" to "Nyayo!" Berkely: University of California Press.