Lucy Mulli
Institute for Security Studies

In 1991, there was an outbreak of fighting in the Rift Valley Province of Kenya. These clashes shattered Kenya’s image as an island of peace in a sea of conflict. Studies seeking to explain the conflict have focused on how President Moi used the cover created by these clashes to manipulate the election process successfully.

This article focuses on two aspects of the Kenyan socio-political arena, and explains how these provide a useful setting for the President to manipulate the political process to his advantage. The first of these features may be described as a system of personal rule; the second as the ethnicisation of politics. These act as the context within which Moi, by instigating and exacerbating conflict, could and can control the political process. The argument is made here that, in so far as this continues to be the nature of politics in the country, Moi or any eventual successor, can choose to use the same means to retain political power virtually at will.

Ethnic clashes in Kenya can be traced back to the year when Kenya was due to have its first multiparty elections since the restoration of a pluralist system in 1991. In October 1991, a gang of youths — said to be from the Kalenjin1 ethnic group, armed with spears and machetes — attacked members of the Luo ethnic group living at Meitei farm in the south Nandi district of the Rift Valley Province.2 In these attacks, thirty houses were burnt and some 4 000 people were left homeless.3 In November and December, fighting between the Luo and Kalenjin extended to Western and Nyanza Provinces, and in the process drew in members of the Luyhia and Kikuyu ethnic groups.

During the course of the December 1992 elections, there was a lull in fighting after which conflict restarted and escalated, now encompassing the Molo, Narok, Pokot, Londiani, Elburgon and Burnt Forest areas of the Rift Valley. The perpetrators of this latest violence expanded to include the Maasai and Pokot ethnic groups. These attacks were aimed primarily at the Kikuyu. After another lull in fighting, there was renewed violence in March 1994. The Kalenjin again fought with the Kikuyu in the Rift Valley and Burnt Forest areas. This was followed by the forced eviction of Kikuyu by the Maasai in the Enoospukia region. In 1995, in the Mai Mahiu area of Naivasha, fighting broke out that left 300 000 people displaced.

The violence that characterised the first elections was to be repeated on a greater scale in 1997, the year of the second multiparty elections. The outbreak started in August 1997 when a local police station – the Likoni Ferry Police Post – was attacked by a group of raiders. Seven policemen were killed, and 44 guns and 5 000 rounds of ammunition were stolen. In the ensuing violence, homes and stalls of the local people were also attacked. Meanwhile, ethnic clashes in the coast province between ‘coastal’ and ‘inland’ ethnic groups left forty people dead and 120 000 displaced. Until this time, sporadic violence had been explained away as the result of competing traditional land claims. This particular incident, however, took on additional religious and ethnic overtones because of local resentment about outsiders settling on the coast and the exploitation of the areas by international tourism.4 Because of the seriousness of this outbreak, and the attention it drew in the foreign media, the Kenyan army was called into action to restore law and order and attempt to capture the perpetrators of the raid at the police post. Nevertheless, after several incidents where extreme force was used that resulted in fatalities, the army was withdrawn.

Whereas the 1992 elections in Kenya had seen violence limited to the Rift Valley, and the Western and Nyanza provinces, the violence in 1997 was distinctive as it had expanded to include the Coastal Province. The point of concern here is the possibility of violence encompassing the whole country in the 2005 elections.

Emerging evidence has suggested heavy government involvement in the clashes, both in 1992 and 1997. Eyewitness accounts vary in the details of each actual incident but are united in either directly or indirectly pointing a finger at the government. Three accounts will suffice as illustration.

The first account is that given by Chief Inspector Julius Ndegwa, who was in command at Songho police station during the October 1991 clashes.5 He explained that he had received a report that a group of Nandis were attacking and burning houses belonging to other communities. He gave orders to his police officers on how to handle the situation once they arrived on the scene. He told them that they were not to shoot at anyone but to try and put out the fires. The curious point is that, although quite aware that houses where being burnt while people were still inside, he did not ask the police to shoot at the offenders. When questioned about this he said, "I told my men that two wrongs do not make a right and that they should use their guns to shoot suspects found running. You see we can recover property, but not lives once lost." Following this incident, 53 people were arrested, but none of them were charged. On order of Mr Otieno, the OCPD, they were later released. These actions on the part of the administrative structure are suspect, especially if one considers the apathetic effort made to stop the attackers and the subsequent release of the offenders without any investigation, following orders ‘from above’.

Even more damning is evidence given by a village elder of the Luo community in Kericho district in Western Kenya bordering the Rift Valley Province. The man said that, in 1993, a contingent of policemen and armed Kipsigi (Kalenjin) youths, stood guard as the former Kericho District Commissioner supervised the demolition of more than 600 houses belonging to the Luo community at Thessalia. Prior to this eviction, the assistant chief had addressed a meeting at the farm and ordered a pastor to read the letter that he claimed, was ‘from above’. In the letter, Luos were given an eighteen-hour ultimatum to vacate or else, "... they would see the red beret." The same elder said that he had positively identified two Kipsigi elders who were commanding scores of heavily armed Kipsigi youths burning the houses.

A further incriminating report was given to the parliamentary committee set up to investigate the clashes.6 The commissioners heard from eyewitnesses that the attackers were junior military police, general service unit officers and former servicemen who were on leave when they carried out the raids. The attackers were said to have been trained in camps in Doineet, Nesuit, Marigot and Kerio in the Rift Valley.7 The report pointed out that "[t]he attackers appeared to be disciplined and obeyed instructions. These exhibited regimental discipline usually found in those who have undergone rigorous military training ..."8

Furthermore, "[e]vidence received by the committee ... indicates that the fighters were on hire and were paid sums ranging from Kshs500 ($6.50) for safe return from the clashes front, Kshs1000-2000 ($12.50-25) for killing one person or burning a grass — thatched house and Kshs10 000 ($125) per permanent house burnt."9

The most significant indicators of the government’s guilt were utterances made by ministers close to the President. William Ole Ntimama — then Minister of Local Government — in responding to attacks on Kikuyus by Maasai in Enoospukia, said that "... the Maasai in Enoospukia were fighting for their rights and ... we had to say enough. I had to lead the Maasai in protecting their rights."10 In addition, a Ntimama supporter, William Ole Aiye said in 1996: "All you Kikuyus and other tribes. I want to assure you that you voted me out during the last elections, showed disregard to our minister [Ntimama] and hold the entire Maasai community with contempt, the time has come for the Maasai to show you that you must support the minister or leave."11

From these accounts, the following picture emerges. In the period before the elections, a series of attacks were made against migrant ethnic groups in the Rift Valley, and the Western and Coastal Provinces in Kenya. These attacks were made with the full knowledge and support of the government, under the leadership of President Moi. While conventional wisdom would argue that disturbances such as these have a destabilising effect on a government’s political life and as such on the President’s power, and that it would therefore be in his interest to stop these attacks, this was not the case. Indeed, aided by the clashes, the President and his KANU party sailed through the elections in both 1992 and 1997. The causal relationship between the incidence of ‘spontaneous’ violence and a sweeping electoral victory will be explained below.

The President’s defining role in the political life of the country and his ability to manipulate events to suit his agenda is explained by the nature of the political order in Kenya. Jackson and Rosberg12 have referred to this as a ‘personal rule’ system. Only by understanding this system can one grasp why the President was and is able to manipulate the political system even to the extent of orchestrating ethnic clashes to enhance his hold on power.

A personal rule system is one in which the person of the President has a determining role in the political arena. The system is not dependent on the particular leader per se, but power is rather vested in the office of the President as an institution. Thus, any person occupying the position of President has immense control which derives from the power vested in this office. This power is overarching in nature and, despite there being a Parliament and Constitution in place, the President can easily bypass these. With reference to this phenomenon, Goldsworthy13 explains:

"Without established institutions to govern by, politicians were left with the task of governing with their personal power and authority. In this system where politics depends on the person rather then the institutions a system of politics based on the person and not institutions, has emerged. It is this system which is called personal rule. A non-institutionalised politics is a personalised politics."14

Politics in this system was and is characterised by a lack of rules and regulations. Politics is seen as ‘a dirty game’. There are no rules, and coercion is used to the players’ advantage. The political arena is likened to the royal court in Machiavelli’s The Prince, where intrigue and struggle for power are the keys and the prince (read ‘President’) has to establish his position by a combination of skills, will power and fortuna.15

Though it appears disorganised, personal rule is systematic on two levels. The first is at the formal functional level where political processes, stability and security, and the provision of goods and services and other material rewards are performed – albeit less efficiently than in a system where politics is institutionalised. The second level is, in an operational sense, a system of relations. There is a system of patron-client relationships linking politicians to their supporters (who comprise a section of civil society). Thus, through such relations in Kenya, political order has been attained without the complete suppression of politics. Because of the unstable nature of these politics, no position is ever certain. Those not in power will try to secure power, while those in power will try to suppress them. "Clientelism, factionalism, plots, purges and succession are rife."16 And because rules and regulations are not institutionalised, this instability is likely to continue. In such a system the government is the President.

In Kenya, President Moi’s power has thrived under the personal rule system. Through a process of ‘cronyism’, Moi is able to retain power by using his ethnic group as a support base. In exchange for their support, the Kalenjin have been rewarded with resources from the public sector. These include senior positions in parastatal organisations and the administration, as well as actual monetary benefits in the form of government loans. Thus, it comes as no surprise that the accession to power by Moi saw the simultaneous entry of Kalenjins into top positions, and the corresponding exit of Kikuyus.17

Against this background, it can be understood how the multiparty debate took on ethnic overtones. With the advent of multiparty politics, Moi identified the threat of losing power, as well as the resources he had used to maintain his power base. Consequently, Moi reinforced his power by fuelling the flame of ethnic identity that had been rekindled by the multiparty debate. He did this by instigating ethnic violence. Through such violence, a false crisis would be created and in the resulting atmosphere of uncertainty and fear, his actions (such as rigging the elections) would not be challenged. Moi did what Machiavelli suggests in The Prince: "... a wise prince when he gets the chance craftily fosters some enmity against himself, that by crushing it he may make his power greater."

When the multiparty debate was reintroduced into the political scene, President Moi said that the country was too young for this kind of politics. He argued that Kenyans would vote along ethnic lines and by so doing divide the country. This point was contested by the proponents of multipartyism, who argued that Kenyans were much too mature for this kind of politics. Indeed, pro-multiparty rhetoric at the time was indicative of a certain degree of unity among political actors. Without wishing to be seen as utopian, it seems that there was some willingness among ethnic groups to work with one another in a common desire to remove Moi. By bringing the ethnic debate into the picture through clashes, opposition parties played into Moi’s hand and their lack of unity ensured his success.

Thus argues Enloe,18 ethnicity is not just a problem, it is also a resource which regimes may manipulate to their advantage, whether to legitimise their authority, enhance their power, strengthen state security, or promote national unity. The leader does this by employing a tactic of divide and rule and the mobilisation of subordinate classes behind their respective ‘ethnic patrons’, in this case Moi.

The reality of Kenya’s politics is such that the ethnic factor is significant in the political arena. By instigating ethnic violence, Moi fanned the fires of ethnic distrust which underlie relations among ethnic groups. By stimulating this distrust, the voting pattern, especially for the presidential vote, was split and Moi was able to win despite failing to gain an outright majority. The voting pattern of the 1992 elections will serve for illustrative purposes (see table 1).

From the table, it becomes clear that, had the opposition been united enough to field one candidate, Moi would not have won even by rigging the elections. However, under the tense conditions brought on by the elections, what little hope there was of such a coalition was lost.

The question raised by this analysis is, under these conditions, what the prospects are for the removal of Moi as President by the ballot in the 2005 elections. The chances of this happening are very slim. The reason why Moi keeps ‘winning’ the elections is due to conditions inherent in the political system of Kenya. Personal rule, the use of ethnic groups as a support base by leaders and the looming fear of more ethnic clashes render the hope of Moi legally exiting futile.

4 The violence centred around Likoni (a Mombasa district). In this area, most of the inhabitants are from the other parts of the country. They are mainly Christian and from the Luo, Kamba, Luyhia and Kikuyu ethnic groups. Politically, they voted for the opposition in the last elections. The locals, on the other hand, are mainly from the Digo ethnic group and mainly Muslim. The gangs attacked mainly bars and graffiti on the wall read, ‘3 days to go or you die.’ This was repeated on several occasions. Most people heeded this warning and left. In refugee camps where most people moved to, they were also attacked. Eyewitness accounts say that this were actually not gangs, but trained militia. Some of the gangs were calling themselves the Kaya Bombo after the forest where local people say they have witnessed groups of young men undergoing weapons training. Similar activities around the Similaini caves were also reported.

5 Akiwumi inquiry told of killings, arson in Nandi, Daily Nation, 17 March 1999.

6 Report of the parliamentary select committee to investigate ethnic clashes in Western parts of Kenya, 1992, National Assembly, Republic of Kenya, Nairobi, Kenya, September 1992.

7 Killing the vote — State sponsored violence and flawed elections in Kenya, Human Rights Report, KHRC, Kenya, 1998.

8 Ibid., p. 15.

9 Ibid., p. 17.

10 B Amisi, A crises in the making — Conflict in the Rift Valley and Western Kenya, Occasional Paper Series, Joan B Kroc Institute for International Peace Studies, University of Notre Dame, Notre Dame, Indiana, 1986, p. 17.

11 Human Rights Report, op. cit., p. 28.

12 D Goldsworthy, Civilian control of the military in black Africa, African Affairs, January 1986.

13 Ibid., p. 101.

14 H Cobie, Persistence and fragility of civilian rule in Kenya, in C Danopolous, Civilian control of the military in the developing world: Democracy on the march?, Westview, Boulder, 1992.

15 Ibid.

16 Ibid.

17 Former President Kenyatta was a Kikuyu and, during his period of presidency, Kikuyus were in dominant positions in the social, political and economic arenas.

18 C Enloe, Police, military and ethnicity: Foundations of state power, Transactions, London, 1980, quoted in D Brown, Ethnic revival: Perspective on state and society, Third World Quarterly, 11(4), October 1989.