Afr.j. polit. sci. (1997), Vol. 2 No. 1, 41-69 
Ethnicity In The Electoral Process: 
The 1992 General Elections In Kenya 
Walter O. Oyugi* 
Abstract 
The primary concern of this essay is to explain the nature of the 'transaction' 
between electoral politics and tribalism, especially with reference to the multiparty 
elections of 1992. It argues that the politics of the 1992 general elections 
show how ethnicity continues to be a major force influencing the behaviour of 
politicians and voters alike. What is more, where power and wealth were at stake, 
ethnic relations became conflictual. The elections also manifested how the elites 
can mobilise ethnic passions to defend and or promote what is otherwise their 
narrow sectional interests. The masses followed their leaders because of the 
lingering belief that only 'one of your own' can best serve communal interest if 
placed in a position of power. But it was also clear that ethnic ideology has its 
limitations. Intra-ethic divisions were manifest where narrow sectional interests 
came into play. The emergence of splinter parties led by members of the same 
ethnic group was the inevitable consequence of such contradictions. 
Introduction 
Many Kenyans believe that tribalism is a canker which is deeply lodged in the 
Kenyan body politic. Yet the same people are usually reluctant to make it a subject 
of discussion across ethnic boundaries because of its emotive force: it is always 
other people's problem and not ours. As a writer once put it, we feel more at ease 
discussing other people's tribalism and not our own. And that is the problem in 
Kenya today. 
The primary concern of this essay is to explain the nature of the 'transaction' 
between electoral politics and tribalism, especially with reference to the multiparty 
elections of 1992. It carries forward a discussion already started by the writer 
in an earlier study which addressed the problem of ethnic politics in Kenya 
1027-0353 © 1997 African Association of Political Science
42 Walter O. Oyugi 
generally (Oyugi 1993). Considering that ethnicity (tribalism) was a major 
influence on voters behaviour during the 1992 elections, there is need to assess its 
effects. Furthermore it is intended to use the experience of the 1992 elections to 
demonstrate the saliency of ethnicity in inter-elite political competition. 
Ethnicity (or tribalism) involves a common consciousness of being one in 
relation to the other groups (Nnoli, 1989 : 10). The ideology of tribalism defines 
"the loyalties and identification of people engaged in conflict" (Gulliver 1969). It 
implies, Gulliver adds, divisiveness and unscrupulous partisanship. What is more, 
I should add, it connotes group antipathy against others. Ethnicity manifests hatred, 
suspicion, envy and mistrust. Elsewhere I have contended: 
It is not a neutral concept merely depicting and signifying the act of 
interactive relations that is expected to take place in a multi-ethnic society. 
To speak of ethnicity is to speak of inter-ethnic interactive situations 
characterised by suspicion, competition, rivalry and often conflict as well 
(Oyugi, 1993). 
It is, as one writer puts it, a blame-pinning devise, according to which it is other 
people's tribalism that is responsible for one's own difficulties, perplexities and 
failures (Parkin, 1968). And it is generally agreed that tribalism is an ideological 
weapon often used in economic competition and political conflict (Nnoli 1978, 
1989, Gulliver 1969, Leys 1975, Horowitz 1985, Bates 1974, etc.). 
A detailed discussion of the origin of the problem of tribalism in Kenya is 
unnecessary here since it is well covered elsewhere (e.g. Oyugi 1993, Leys 1975, 
Horowitz 1985, Dirk Berg-Schlosser 1992). A brief sketch of it is however, in 
order. The origin of the problem in Kenya (and Africa in general) is colonialism. 
It is the institution of colonialism that created a common centre that all the existing 
ethnic groups in given colonial states are at once required to relate to. The emerging 
relations soon became relations of competition over access to goods and services 
associated with modernity. Ethnic consciousness was further accentuated as the 
tempo of modernization accompanied by urbanization gave rise to free movement 
and settlement of peoples in areas other than their own. The notion of 'a people's 
own area' which resulted from the formal politico-administrative regimentation of 
the colonised people into ethnic administrative enclaves was later to lead to the 
heightening of ethnic self-identity or sense of belonging. It also in the process, 
created a sense of exclusiveness which sooner or later manifested itself in the 
rejection of 'outsiders'. In the meantime, shared involvement in the colonial 
economy increased, thereby preparing the ground for the eventual conflict based 
on inter-tribal competition. Leys's much-quoted statement is apposite: 
The foundation of modern tribalism were laid when the various tribal
Ethnicity In The Electoral Process 43 
modes and relations of production began to be displaced by capitalist ones, 
giving rise to new forms of insecurity, and obliging people to compete with 
each other on a national plane for work, land and ultimately for education 
and other services seen as necessary for security (Leys 1975 : 199). 
A combination of colonial attitudes and strategies and the response to them by 
the various ethic groups were later to provide the setting for future competition and 
conflict. The colonial authorities regarded the attitudes of some groups as anti- 
'modernization' and marginalized them. Accordingly, the 'development' strategies 
devised tended inevitably to benefit some groups at the expense of others. 
'Open' areas with more missionary stations received early and relatively better 
education as the 'closed' areas (inhabited mainly by nomads) lagged behind. It has 
been observed that in the process, the Kikuyu, the Luo, the Luhya and a few other 
agricultural communities became early beneficiaries of especially modern education. 
Education was later to prove crucial as a criterion of access to gainful 
employment and other economic activities. But it has also been observed that in the 
process of colonial 'development' some groups adapted much earlier than the 
others. The Kikuyu are said to have been the first to adapt their social structure and 
culture to the capitalist mode of production (Leys, 1975 : 200). This enabled them 
to be more mobile and to adapt to different local situations outside Kikuyuland in 
search of economic opportunities, especially land and business. Many years later, 
their aggressive economic presence would become a source of resentment by their 
'hosts'. 
The Luo and the Luhya also became mobile but for a different reason. Unlike 
the Kikuyu, the primary concern of these groups was the search for wage 
employment in urban centres and on European farms. It is in this area (of wage 
employment) that their rivalry and competition against the Kikuyu was later to be 
experienced. Indeed, it can be argued that up to the time of independence, ethnic 
conflict at the national plane was confined to the struggle among the three groups 
in search of employment and access to other basic needs and services, especially 
in the urban centres. But the situation changed soon after independence. The 
struggle for the control of the new state brought with it new competitors as well as 
new areas of potential conflict. Over the years, this conflict has tended to assume 
ethnic characteristics and intruded in the political sphere. 
Background to the Elections 
Kenya has had only limited experience with competitive multi-party elections; and 
the experience is confined to the first three years of independence. At independence, 
a multi-party contest involved the Kenya African National Union (KANU) 
and Kenya African Democratic Union (KADU) — the two parties formed in 1960 
in anticipation of independence. Up to that year, countrywide political organiza-
44 Walter O. Oyugi 
tions had been proscribed following the declaration of a state of emergency in 
October 1952 and the banning of the Kenya African Union (KAU) (intended to 
contain the Mau Mau revolt). Between 1955 and early 1960 only district based 
political parties or associations were allowed in non-Mau-Mau areas. And in 
districts where they were formed, they became the organizational outlets through 
which the nationalist agitation for independence found expression. In addition, 
purely ethnic based political associations such as the Kalenjin Political Alliance, 
the Masaai United Front, were formed during this period. Later it was these 
associations — both ethnic and district based, that were required to disband in 
preference for one united national party. KANU was formed with that aspiration 
in mind. But that was never to be. The stakes were just too high. Ethnic and interest 
group calculations had led the so-called 'minority' tribes to withhold their support 
for KANU. The settlers, fearful about the security of 'their' land if such a united 
party came under Kikuyu-Luo leadership, exhorted the 'minority' tribes to have 
nothing to do with it.1 Land ownership became a major political and constitutional 
issue with the 'minority' tribes made to believe that they could only secure their 
rights under majimbo (regional) type of constitution and with a party of their own. 
Unity in the nationalist movement in the run-up to independence elections was not 
possible in those circumstances. Therefore, KADU was formed as a counterpoise 
to KANU. 
In the 'mid-term' elections held in March 1961 the nature of ethnic support for 
the two parties became evident. KANU received support from the Kikuyu, Luo, 
Meru, Embu, Kamba and Kisii. KADU on the other hand received support almost 
exclusively from the pastoral tribes: the Kalenjin, Masaai, Giriama and a few other 
minor tribes (Bennett and Rosberg 1961; Bennett 1963). The Luhya, one of the 
three major ethnic groups (after Kikuyu and Luo) were divided between KANU 
and KADU. This pattern of support would hold (except for the Kamba) during the 
1963 elections leading to independence. 
Feeling marginalised in KANU right from KANU's formation in 1960, the 
Kamba leader Paul Ngei succeeded in mobilizing his fellow Kamba elites into 
forming yet another party in 1962—the African Peoples Party. Their intention was 
obviously to create an organizational framework within which to bargain for 
inclusion in the government after the 1963 independence elections. The trick 
worked, and Ngei and other Kamba elites were appointed to ministerial posts after 
APP's eight MPs (out often Kamba MPs) had rejoined KANU. Soon after, APP 
died a natural death. The seeds of ethnic calculations in electoral politics had thus 
been sown. 
The polarisation into KANU and KADU followers did not last long, for in 
October 1964 KADU was 'voluntarily' disbanded and Kenyaemerged as adefacto 
one party state.2 But the unity was not to last long; for in 1966 an open split in 
KANU between the 'moderates' and the 'radicals' over the control of the party and
Ethnicity In The Electoral Process 45 
government, led to the removal of the radicals from the party and government and 
the formation of Kenya Peoples Union (KPU) by Oginga Odinga. The by-elections 
held in 27 constituencies in that year revealed two influences in the electoral 
process. Firstly, the regime was determined to portray the opposition as a Luo tribal 
formation by ensuring its defeat in non-Luo constituencies through electoral 
malpractices.3 Secondly finding themselves on the defensive, the Luo became 
victims of ethnic ideology by perceiving the contest as one between them and 
their adversaries in KANU: they returned all the KPU candidates in Luoland to 
office. 
Three years of sustained marginalisation and isolation of the Luo led to 
increased tension between KPU and the KANU regime, which later exploded into 
violence during Kenyatta's visit to Kisumu in December 1969. Following the 
incident, KPU leaders were arrested and detained and the party banned. Kenya 
once again became a de facto one party state. It was to remain so until 1982 when 
the de facto situation was given legal basis through a constitutional amendment. 
Therefore, between 1969-1991, the five general elections held during the period 
(1969, 1974, 1979, 1983, 1988) were 'KANU-only' elections. 
The threat posed earlier by an ethnically based party (KPU) was to lead the 
KANU regime to resort to a strategy of denial and exclusion. Thereafter, any 
politician whose loyalty was doubted or who was considered as a potential threat 
to the regime was at once denied the chance to stand in an election. The ruling party 
introduced a mechanism through which all aspiring candidates were subjected to 
a clearance process. And all cleared candidates had to pledge loyalty to the (state) 
president (who incidentally was also the party president), the government, and the 
ruling party — KANU — as a condition for nomination. 
The strong-arm tactics employed by the regime thus destroyed the prospect for 
the formation of a viable opposition; and the ruling party encounted no organised 
opposition throughout the 1969-91 period. In the circumstances no presidential 
contest was possible. The leader of the sole party was always assured of automatic 
nomination, and since intra-party challenge for that office was also inconceivable 
then, the president was always 're-elected' unopposed. The multiparty movement 
of 1990-92 should be seen against this background. 
Up to the late 80s, any serious challenge to the authority of the regime was a sure 
way of earning a term in detention or frequent harassment by the political police. 
Fear was by and large institutionalised. Many Kenyans would pick up courage only 
from the late 80s. The collapse of the Soviet empire had much to do with the new 
development. The enunciation of glasnost andperestroika by Gorbachev, and their 
application in the Soviet Union from 1986, the extension of the same to Soviet 
Eastern Europe, and the eventual collapse of the Soviet empire at once led to a 
global attack on centralism and authoritarianism of the left and right. African 
authoritarian regimes (military or one party) became natural targets of the West.
46 Walter O. Oyugi 
The USA, the Scandinavian countries, Germany, the World Bank and the IMF 
especially led the way. But even reluctant world powers like France and Britain 
were forced to also apply pressure later when the movement gained momentum 
from within individual African states. 
In the case of Kenya, the stimulus given by the changes in the external 
environment was to be reinforced by a number of domestic events. The most 
important ones were the introduction of queue voting in 1988 which was aggressively 
opposed by many civic groups — especially the Church. Then came the 
massive rigging of the (1988) elections where candidates with long queues were 
declared loosers and those with short ones declared winners in the full view of the 
electorate. The outcry that followed these events was sustained both by the external 
and domestic mood for change. The critics thus felt increasingly more secure as 
they challenged the legitimacy of the regime. Much sooner than had been expected, 
a crack developed in the regime. The first cabinet resignation in more than twenty 
years occurred over the rigged elections. The resignation of Ken Matiba provided 
a golden opportunity for the political opposition, some of whom began to see him 
as a rallying point for their cause.4 
The much-talked-about Luo-Kikuyu alliance began to show signs of picking up 
again after 1993; for early in 1990, several meetings were reportedly held between 
Matiba and the Odingas (Oginga and his son Raila) before Matiba, Charles Rubia 
(both of whom had then been expelled from KANU for their opposition to the 
regime) and Raila were detained following the abortive pro-democracy rally on 7 
July 1990 which ended in rioting in Nairobi and the surrounding towns in 
Kikuyuland. 
Scared of the manoeuvres by Matiba, Rubia and the Odingas, the regime 
mobilised MPs to pass a resolution rejecting the idea of allowing multiparty 
politics contending that such a system would encourage tribalism (Nation 29.3.90). 
A few months later, as Moi himself hit at critics of the one-party system and 
threatened them with detention (Weekly Review 18.5.90), pro-single-party rallies 
addressed by KANU MPs were conducted in non-Kikuyu and non-Luo areas (Ibid). 
The detention of Matiba, Rubia and Raila instead of nipping in the bud what had 
by then assumed the proportion of an opposition movement, helped to intensify the 
attack on the regime by the civic and political opposition. As the movement 
gathered momentum the idea of launching an opposition party became real, when 
Odinga announced the formation of his National Democratic Party (Daily Nation 
3.1.91) which was immediately denied registration. This response by the state 
forced some elements in the opposition movement to return to the drawing board. 
The result was the birth of an opposition alliance (not a party this time) to mobilize 
the populace for change. Behind the alliance was a group of 'young' professionals 
and intellectuals in the movement — later to be referred to collectively as the 
'young turks'. Prominent actors included: Gitobu Imanyara (lawyer-journalist)
Ethnicity In The Electoral Process 47 
Kituy i (Social anthropologist); Paul Muite (lawyer, then also Chairman of the Law 
Society of Kenya); Anyang-Nyong'o (political scientist); James Orengo (lawyer); 
Raila Odinga (Engineer, former political detainee) etc. It is this group that turned 
to the more experienced and widely known politicians to lead the alliance. The 
interim leadership went to Oginga Odinga, a Luo (Chairman) Masinde Muliro, a 
Luhya (Vice-Chairman); Martin Shikuku, a Luhya (Secretary General); George 
Nthenge, a Kamba (Treasurer). Others were Gachoka, a Kikuyu, and Bamaritz, a 
Kenyan of Arab descent. 
Founded in July 1991, the Forum forthe Restoration of Democracy [FORD] had 
by December of that year emerged as a national movement whose support base cut 
across major ethnic boundaries — Kikuyu, Luo, Luhya, Kamba, Kisii, Meru, 
Embu especially. By then it was only the Kalenjin (and kindred tribes) and the 
Masaai that were still hanging-on to KANU. 
With the formation of FORD, the opposition to the regime now assumed an 
organizational form. Donors were now also in a position to determine whether 
indeed the quest for change had popular support. Apparently convinced that the 
need for change had wide popular support, donors (both bilateral and multilateral) 
became openly involved in the movement for multiparty policies. The following 
examples are illustrative. Barely a month after the formation of FORD, a visiting 
IMF official made it clear in Nairobi that the Fund preferred to do business with 
regimes which practised participatory democracy (Standard 15.8.91). The involvement 
of the USA ambassador, Smith Hempstone, became increasingly open. 
On several occasions he had to intervene on behalf of arrested opposition 
politicians (see e.g. Standard 26.9.91). And when a number of opposition politicians 
were arrested in mid-November 1991, fourteen US Congressmen sharply 
criticised the arrests (Nation 20.11.91). By this time even Britain, which is usually 
cautious, had began to take a tough stance against the regime. In a BBC interview, 
Overseas Development Minister, Lynda Chalker, gave warning that Western 
donor countries would take tough action (at the Paris donors meeting soon to be 
held) to ensure that President Moi ended one party rule and respected human rights 
(Sunday Nation 24.11.91). Indeed, when the meeting was held, donors decided to 
withhold financial assistance to Kenya pending political reforms. It is against this 
background that the relevant organs of KANU met in early December 1991 to scrap 
Section 2(a) of the Constitution of Kenya, thereby effectively ending a decade of 
de jure one party rule. 
But the opposition unity that had influenced the change was not to last. With the 
country's constitution amended to allow for the restoration of multiparty politics, 
FORD was registered as a political party. At once personal and ethnic considerations 
began to influence the decisions and actions in the party. A month or so later, 
two new parties were launched both with an ethnic agenda. By the time of the 1992 
elections up to eleven parties had been formed — none of which could claim to be
48 Walter O. Oyugi 
national on the basis of the structure of popular support. It is against this 
background that the 1992 elections were held. 
Ethnic Influences On Electoral Politics 
According to Lakemann elections in a democracy are justified on the principle that 
"the rulers whom the people are expected to obey should not only rule in their 
interests, but also rule according to their wishes; and that the rulers should be 
accepted by the ruled — this acceptance being no mere acquiescence but conscious 
choice ..." Thus, it has become increasingly accepted, he continues, that the 
governed should elect their rulers, and that they should have the power to renew 
at intervals the authority they give to such rulers (Lakemann, 1994). In the 
'developing' ethnically plural societies of which Kenya is one, the situation is 
however different. Elections in such societies are seen as an opportunity to compete 
for control of the state, precisely because of the discriminative use to which the 
state is usually put by the group that happens to control it. Larry Diamond observes 
that in "ethnically divided societies... elections... become not only the vehicle for 
protecting the general process of capitalist accumulation but also for promoting 
accumulation by one cultural section of the dominant class in competition with 
others. Thus, they become a major expression of ethnic conflict" (Diamond 1986). 
His further contention is that because manipulation of mass ethnic feelings is often 
the surest instrument of electoral success, democratic participation (in elections) 
serves to fan ethnic conflict at the mass level as well. The ethnic conflict in the Rift 
Valley during the 1992 elections in Kenya as well as the many clashes in the urban 
constituencies such as Langata in Nairobi and several others elsewhere e.g. 
Mombasa and Nakuru, are good examples. 
To be specific, during the 1992 elections, the Kalenjin (the ethnic group of 
President Moi) believed that the capture of the state by the opposition would at once 
mean the loss of economic privilege which they had enjoyed for over a decade. 
Similarly, every major ethnic actor believed that their party's victory would end 
their relative deprivation. These perceptions were functionally conflictual. Indeed, 
what Wolpe and Melson said of the Nigerian experience is equally true of Kenya, 
namely that competition for material goods and hence for control of the state which 
governs access to them accelerate the ethnicization of society.5 
The 1992 multiparty elections in Kenya were a living manifestation of these 
fears. Indeed, never before had ethnic considerations so directly influenced the 
electoral process. It was the first time in the country's history that the post of 
president was openly contested and by implication, the first time the electorate 
were given a chance to determine who and consequently which group would 
control State House, together with the benefits that go with it; for after the creation 
in 1964 of the institution of executive presidency under a one party system (initially 
de facto, and later de jure), the president of the ruling party (Kenya African
Ethnicity In The Electoral Process 49 
National Union — KANU) always received automatic nomination for the position 
of State President and was formally pronounced elected at the nomination stage in 
all the subsequent elections (1969, 1974, 1979, 1983 and 1988). As such general 
elections had no bearing on who would control state power and patronage. 
The 1992 contest was perceived differently by key ethnic actors. For the 
incumbent ruling coalition — the Kalenjin and their kinsmen in the Rift Valley, it 
involved the defence of the existing structure of privilege, on the one hand, and the 
possibility of deprivation, on the other. For the former rulers — the Kikuyu — it 
presented an opportunity for politico-economic 'resurrection'. For the others (the 
Luo and the Luhya especially), 1992 was their turn "to eat also". The Luhya elites 
were quick to remind the nation of their tribe's numerical strength (the second 
largest group after the Kikuyu since 1969) and therefore their right to vie for the 
highest political office in the land. The Luo who have never hidden their 'legitimate' 
claim to leadership once the Kikuyu whom they regard as their equal 
adversary had had their chance, expected the opposition alliance to support their 
bid this time without any challenge. At the same time, the possibil ity of controlling 
State House which the 1992 elections presented to the various communities acted 
as a disincentive to rational behaviour. Intra-elite contest within the opposition 
movement soon assumed ethnic dimensions as each ethnic group increasingly 
coalesced around one of their own. Contrary to the wishes of ordinary opposition 
supporters, reaching a consensus about one opposition candidate to challenge the 
incumbent president because increasingly remote with the passage of time. None 
was ultimately agreed upon. 
The Struggle for FORD 
The moment the prospects for removing Moi and KANU through democratic 
elections became real, the struggle for FORD also began in earnest. At once the 
political elite began to calculate their fortunes within the movement, wondering if 
indeed staying in it would serve their own interests individually and collectively. 
The structure of leadership of FORD became a subject of great interest and 
concern. Of the six interim leaders, only Odinga, Muliro and Shikuku were well 
known nationally. The rest were not. Therefore, as one would expect, the first 
group to develop a sense of concern and denial were the Kikuyu. Their concern was 
based on the fear that the Luo and the Luhya in the interim leadership of FORD 
might take advantage of the situation to dig themselves in, thereby preempting the 
leadership contest when the party was registered and party elections called. 
The sense of urgency among the various contenders regarding the leadership of 
FORD was heightened when in January 1992, FORD was registered with the 
former interim officers as its substantive leaders. This writer was reliably informed 
of the manoeuvre by some powerful former GEMA functionaries who felt their 
interest could not be well served in the existing leadership structure of FORD. It
50 Walter O. Oyugi 
is these people who had earlier on prevailed on Mwai Kibaki (then Minister of 
Health) to resign from the government to lead an alternative party. Hence within 
a few weeks of the formation and registration of FORD, the Democratic Party of 
Kenya was launched with Mwai Kibaki as its interim chairman. There were others 
who also felt they should not be left behind; for within the same month, the Makau, 
MP for Mbooni in Kambaland, quit KANU to form the Social Democratic Party. 
At the time, Makau was a fairly popular MP and might have hoped to use his 
popularity and his new party to mobilise support from his Kamba people and 
ultimately use that support, as Paul Ngei had done on the eve of independence with 
his APP, as a bargaining weapon for a bigger stake. Launched in the same month 
also was the Islamic Party of Kenya which at once gave the Muslim fundamentalists 
especially in Mombasa, an organizational framework for letting out their 
hitherto suppressed political passions. 
With the formation of the DP, the Kikuyu masses began to wonder whether it 
was a signal for them to begin withdrawing their traditional support for FORD. But 
the signals they received at the time were rather mixed. Some of their senior 
politicians who had defected to FORD were still staying on. These included 
Kenneth Matiba (a former cabinet minister who had been detained during the early 
stages of multi-party movement (mid-1990), released from detention after suffering 
a major stroke, and was in early 1992 still recuperating in London after being 
hospitalized there). Former vice president J. Karanja, former MPs Charles Rubia, 
Andrew Ngumba, Maina Wanjigi, Matu Wamae etc. and others such as Mungai 
Njoroge were on the other hand moving back to KANU after receiving a cold 
reception in the opposition movement. 
What were the costs and benefits of joining this or that party? This appears to 
have been the dilemma which the ordinary Kikuyu in particular faced at a time 
when their leaders were dispersed among the three main parties — KANU, FORD 
and DP. For the highly politicised urban Kikuyu masses, Matiba appeared to be 
their man. He was regarded as a brave fighter in remembrance of his decision to 
resign his cabinet post on a matter of principle when he disagreed with the 
government over the conduct of the 1988 elections. He was also associated in their 
minds with the 7th July 1990 mass civil disobedience in Nairobi which had resulted 
in the so-called saba saba riots of that year. On the other hand, Kibaki, the interim 
leader of DP had for a long time been portrayed as a cowardly and hesitant leader 
who had failed during his tenure as vice-president to protect Kikuyu interests when 
the same came under attack.6 In the circumstances, and with KANU virtually 
deserted by the Kikuyu masses, most of the Kikuyu elites chose to remain in FORD, 
hoping to capture it and use it as a vehicle for making a political come-back at the 
centre. To succeed in this regard, the Kikuyu needed to identify one of their own 
with a national stature. No such a person was readily available in the then existing 
leadership structure of FORD. And even outside the leadership structure of FORD,
Ethnicity In The Electoral Process 51 
some of the possible Kikuyu contenders such as former vice president Karanja 
lacked popular appeal nationally in comparison with non-Kikuyu contenders such 
as Odinga, Muliro and even Shikuku. Hence the decision to turn to Matiba in spite 
of his poor state of health. 
In the meantime, the strategy was to push the incumbent FORD leaders into 
taking some precipitous action that would tarnish their public standing. One such 
action involved the instigation of a national strike in April 1992 in demand for the 
release of all political prisoners. The strike was poorly organized and it failed 
miserably. Another call for a boycott of Madaraka festivities on 1 st of June 1992 
was also ignored by Kenyans. It now appeared as if FORD and the entire opposition 
movement was beginning to loose ground to KANU. These developments encouraged 
the Matiba supporters in FORD who now saw him as the only man that could 
put the opposition back on track again. 
Revelations were later made of secret delegations to Matiba by some of the 
Kikuyu elites still in FORD. The Week End Mail (18.2.93) named some of the 
persons involved as Nyanja, Kimani Nyoike, Charles Rubia, Matu Wamae and 
Bedan Mbugua. Their intention was to block Odinga from becoming both 
chairman of FORD and its presidential candidate. Matu Wamae on returning from 
one of his many visits to Matiba declared at a public rally in Muranga town that he 
had recently travelled to London at the behest of a section within FORD to see 
Matiba, and that he had found him fit and healthy (Weekly Review 21.2.92: 8). This 
statement was made at a time when Matiba could neither read nor write, according 
to subsequent revelations by one of the principal actors, Nyanja (Kenya Times 
30.9.92 : 2) who revealed this in Kiambu District following disagreements 
between himself and Matiba over FORD-Asili leadership. He revealed further that 
the new party was in fact formed after a split in the original FORD which occurred 
following Matiba's return from London.7 
For the Kikuyu, the struggle for FORD involved targeting Odinga; for earlier 
in the year he had declared his interest in the (state) presidency if Kenyans gave him 
a chance (Standard 23.1.92). He later reaffirmed his interest on the same day that 
Matiba was being accorded a hero's welcome by his supporters on his return from 
London (Daily Nation 2.5.92). Odinga and his Luo backers especially were on 
record as having reminded the Kikuyu that without Odinga's support, Kenyatta 
might not have been released from detention let alone become president.8 This 
historical fact was now to be denied by some Kikuyu politicians such as Njenga 
Mungai and Wa Nyoike. Odinga's Luoness also became the subject of derision. 
George Nyanja (later to become MP for Limuru) declared publicly: "Odinga 
cannot lead anybody because he is not circumcised.1* It is these forces that 
apparently prevailed on Matiba to declare his candidacy for the presidency in 
February 1992 when he was still recuperating in London (Standard 5.2.92). 
Indeed, the manoeuvres by the advocates of ethnic interests were to deal a final
52 Walter O. Oyugi 
blow to the fragile unity in FORD. Odinga and Shikuku and their followers were 
equally guilty. Odinga for one began to surround himself with his kinsmen in a 
manner that openly gave signals of things to expect in the event that he became 
president. It is not a trivial matter that he opted to make his daughter-in-law his 
personal assistant through whom all those seeking to see him had to pass. She 
remained in that position until after Odinga's death. As if that was not enough, 
Odinga proceeded to influence the appointment of a fellow Luo, Professor Peter 
Anyang Nyong'o to the key position of Executive Director of FORD. Some of the 
FORD insiders later complained that the director had usurped all the functions of 
the Secretary General Martin Shikuku10 thus precipitating the belligerence of 
Shikuku's supporters and also his threat to quit FORD before the final show-down. 
The alleged active participation of Odinga's son in party affairs at the headquarters 
where he did not hold any position before the party elections was also often cited 
as a source of friction. Indeed, Martin Shikuku openly complained that the 
headquarters was being run by Odinga as a family affair. As a result of such open 
Luo manoeuvers Shikuku and Muliro also came under great pressure from the 
Luhya to be more assertive in their quest for leadership positions in FORD. Elijah 
M wangale, for example, criticised Muliro and Shikuku (whom he claimed were the 
initiators of the crusade for multipartyism) for betraying the Luhya by allowing 
themselves to be relegated to obscurity when they were supposed to have assumed 
the leadership of the largest opposition party (Daily Nation 29.1.92). But the two 
were known not to be on good political terms. Each one of them struggled for a 
prominent position in FORD against the other: for they knew that it could 
ultimately be either of them who would succeed. Obviously, Muliro had an edge 
over Shikuku and that might explain why Shikuku chose to seek an alliance with 
the Matiba faction of the party even though their political values were known to be 
poles apart." 
The final split in FORD came within two months of Matiba's return to Kenya. 
His return had been organised by the Kikuyu elites in FORD to give him maximum 
publicity. And it was hoped that the occasion would be used as a launching pad for 
Matibas leadership challenge. The other factions in FORD would have nothing to 
do with the arrangements and Odinga for example chose to be out of Nairobi on the 
material day. 
Matiba's handlers must have given him the impression that the party needed him 
more than he needed it, for soon after his return, he chose to go solo. Within two 
months he had, with Shikuku's support, managed to lure the majority of the six 
interim leaders of the party to his side and therefore felt strong enough to announce 
the suspension of Odinga on 11.8.92 (Weekly Review 14.8.92) and to convene a 
meeting of the National Executive Committee of the party to ratify the decision as 
well as designate himself as the new party leader. The move did not succeed as 
many bonafide members of the Committee ignored the call. But the damage was
Ethnicity In The Electoral Process 53 
irreparable. The party split into two factions and were later registered as Ford- 
Kenya and Ford-Asili under Odinga's and Matiba's chairmanship, respectively. 
The split marked the end of the only movement since pre-independence that had 
succeeded in bringing most of the Kenyan masses into an enthusiastic political 
collaboration. 
The Grand Ethnic Strategy 
Up to around April 1992, there was very little hope of KANU ever returning to 
power. But with the return of Matiba from London in May and the subsequent split 
in FORD in July, some life seems to have returned to KANU. By then KANU had 
been abandoned by the GEMA (Kikuyu, Embu, Meru) communities, the Luo, the 
Luhya, Kisii, Coastal people, the Somali and to a large extent the Kamba. By the 
middle of 1992, therefore, KANU was essentially a Kalenjin-Masaai, 
Turkana, Samburu (KAMATUSA) party.12 The prominent politicians from GEMA, 
Luo and Luhya who chose to stick with KANU had done so primarily after realising 
that defection could not save them from eventual defeat during intraparty 
nominations. Some of them were later to engage in the kind of desperate 
and divisive political campaigns reminiscent only of the majimbo elections of 
1963. 
Earlier on, that kind of campaign had been started by the KAMATUSA coalition 
in KANU. Expecting at the time to be humiliated at the polls, they got together and 
decided that those ethnic groups that had 'betrayed' them should be taught a lesson. 
The lesson in question involved their expulsion from especially "Kalenjin-Masaai 
lands" in the Rift Valley. Apart from this being a good lesson for betrayal, it would 
also rid the Rift Valley of anti-KANU, anti-Moi voters. The series of meetings held 
by the Rift Valley politicians during 1991 were intended to achieve those 
objectives by mobilizing the Kalenjin and their political allies against other 
communities living in the Rift Valley. At a meeting held in Kapsabet in Nandi 
district, the notion of multi-party ism was condemned and FORD was portrayed as 
an anti-Moi and anti-Kalenjin movement (Weekly Review 13.9.91 : 6). At the said 
meeting, two Kalenjin ministers reportedly began their speeches by claiming that 
the Kikuyu represented the main opposition to Moi. Successive speakers threatened 
that if majimboism were introduced, all the Kikuyu who had settled in the Rift 
Valley would have to pack and leave. Others asked all government critics in 
Kalenjin areas to move back to their "motherland". Besides the GEMA, these 
threats were aimed especially at the Luo, Luhya and Kisii. At subsequent meetings, 
the need for Kalenjin unity in the face of opposition threat was stressed (Weekly 
Review 27.9.91 : 5). Earlier on (Nation 9.9.91) at a meeting attended by Kalenjin 
MPs and other politicians the need to protect the Moi presidency and the Kalenjin 
land was emphasised. The Minister of State in the Office of the President, Kipkalya 
Kones, for example, called for the appointment of more Kalenjin staff in the
54 Walter O. Oyugi 
Ministry of Lands in order to save their land from being grabbed by 'outsiders' and 
also warned that the Kalenjin would fight to the last man to protect Moi's 
government. The call for political pluralism was thus perceived by the Kalenjin 
leaders as a call for the end of the Kalenjin leadership. And on 21.9.91, 19 Rift 
Valley MPs led by Biwot, Ministers Cheruiyot and Mibei and assistant minister 
Paul Chepkok pledged to counter any attempt to relegate them from leadership; 
Chepkok on his part urged the people to arm themselves with sticks, bows and 
arrows and destroy any FORD member on sight. Such rallies and speeches 
continued throughout the period leading to the December 1992 elections. For 
instance as late as a few weeks before the elections, a leading Kalenjin pastor — 
the Reverend Elija Kiprotich Yego (he had at one time tried to become the Anglican 
Bishop of Eldoret Diocese) did on 2.12.92 give the non-Kalenjin communities 
resident in Nandi District 26 days ultimatum to vacate the area if they should fail 
to vote for Moi and KANU on December 29, 1992. He told an audience of about 
4000 at Kapsabet: If the president looses, make sure there are no more (of these 
'foreigners') in our Nandi land. (Daily Nation 3.12.92). It is againstthe background 
of such inflammatory speeches that the ethnic clashes which began in late 1991 and 
have continued sporadically since then should be seen. As intended, they did have 
the effect of driving many registered and potential non-Kalenjin voters out of the 
Rift Valley during the elections.13 
Elsewhere in the Rift Valley, similar threats were issued against "foreigners", 
especially by one Masaai Cabinet Minister, Ole Ntimama (Daily Nation 24.12.92; 
25.12.92). The politicians were later to be joined by some local Kalenjin administrators 
in threatening non-Kalenjin voters (Daily Nation 23.12.92 : 5). 
Apart from this strategy of riding the Rift Valley of opposition voters, Moi also 
took advantage of the disunity within the opposition movement to build new 
coalitions in divided opposition areas. Four areas were central to this strategy: 
Kisii, Luhyaland, Ukambani and Meru. It is believed that a vice presidential offer 
was made to prominent politicians from these areas as a way of persuading them 
to stick with KANU. And aware of the vulnerability of KANU, some of these actors 
chose to exploit the situation to their personal advantage by resorting to the use of 
ethnic ideology. Former cabinet ministers, Elijah Mwangale and Burudi Nabwera 
from Luhyaland in particular merit mention here. In numerous rallies during the 
campaign, they stressed separately the need for Luhya support for KANU and Moi 
as the surest way to be rewarded (Daily Nation 25.11.92 : 5; 2.12.92:4; 29.12.92 
: 2). The vice presidency was often mentioned; and so was the presidency itself 
after Moi's departure. For instance, at a rally in Ikolomani in Kakamega district on 
29.9.92, Mwangale reportedly declared: 
Luhya must be total in demanding for the second top post. We will resist any 
attempt to stop a Luhya from ascending to the top. The community will bid
Ethnicity hi The Electoral Process 55 
for the Vice Presidency when President Moi and KANU form the next 
government (Kenya Times 30.9.92 : 2). 
Taking advantage of the tension between the Kikuyu and the Kalenjin, M wangale 
and Nabwera considered that an all out attack against the Kikuyu would be the best 
way to demonstrate their loyalty to the regime. They exploited popular prejudices 
and stereotypes and condemned the alleged Kikuyu domination of business in 
Luhyaland. At one rally, Nabwera 'reminded' the Kikuyu that time had come for 
them to leave the Western Province to the Luhya to run business and raise their 
living standards (Daily Nation 2.12.92 : 4). 
The idea of electoral alliances also emerged in speeches by other politicians. For 
instance at a rally in Nyamira, the Kisii bureaucrat turned politician, Simeon 
Nyachae, reportedly said that the Kisii had agreed to vole for President Moi and 
KANU "on condition that the community was given its rightful share in the 
country's leadership". He then added that because they had been promised a 
number of things, they should unite and vote for Moi and KANU (Standard 
27.10.92 : 4). These alliances must have been entered into in furtherance of 
especially the requirement that a presidential winner must, besides having the 
highest number of votes, win 25 percent of the votes cast in at least five provinces. 
Sure of losing the Luo votes in Nyanza and the Kikuyu votes in Central Province, 
Moi must have considered these alliances as being critical to his success at the 
polls; and indeed they turned out to be. 
Meanwhile, Moi did not waste any time in exploiting the simmering rift within 
the opposition movement. His decision to invite FORD Secretary General Martin 
Shikuku (before the fateful split in FORD in July 1992) fordinner was a calculated 
move because of its timing. At the time, there was a lot of feud between Shikuku 
and Odinga over the management of party affairs and on the method of electing a 
presidential candidate on FORD ticket. Moi must have calculated that his move 
would at once bring the simmering rift within FORD to the boiling point and 
thereby cripple it. Although the disclosure of the visit did not lead to an immediate 
split, it exacerbated the crisis and eventually sealed the split a few months later. 
Immediately before and during the campaign, Moi's supporters resorted to what 
essentially amounted to ethnic cleansing of the Rift Valley. Non-Kalenjin candidates 
or would-be candidates were openly harassed. Those with debts were 
arraigned before the courts. The case of one Stephen Ngali Waimaithori who had 
declared his intention to contest the Eldoret South seat illustrates a trend. (Standard 
5.11.92:4). The grand strategy also involved declaring all KAMATUSA constituencies 
as KANU zones. Those who offered themselves as opposition candidates 
were harassed regardless of their ethnic affiliation. For instance in Tinderet (in 
Nandi District) Kimaiyo Arap Sego, a Kalenjin, narrowly escaped death when a 
group of political thugs attacked him in order to prevent him from presenting his
56 Walter O. Oyugi 
nomination papers (Standard 11.12.92 : 5). In Londiani (Kericho District) the 
houses of DP council candidates were set on fire (Daily Nation 12.12.92 : 4). And 
on 28.12.92, DP candidates for Eldoret South, Charles Murgor and Uasin Gishu 
(DP Chairman) were reportedly attacked by KANU supporters (Daily Nation 
1.12.92 : 17). Many of such cases were reported before and during the campaign 
period. And on nomination day, opposition candidates were physically prevented 
from presenting their papers as a result of which seventeen candidates (in mainly 
Kalenjin and kindred group areas) were returned "unopposed" (Daily Nation & 
Standard 11.12.92). In the meantime, tribal clashes continued, thus forcing many 
prospective non-Kalenjin voters to flee their respective constituencies. In the end, 
the strategy of declaring the Rift Valley a KANU zone seems to have worked. Of 
all the Presidential votes cast in the province, the opposition received only 29 
percent. The rest went to Moi. 
Persistent Disunity of the Opposition 
For the opposition movement, the formation of the DP and the subsequent split in 
FORD had put into question the chance of dislodging KANU. This realization led 
to many frantic efforts at unity by some organizations sympathetic to the opposition 
such as the Church and Professor Wangari Mathai's Middle Ground group. 
But a combination of ethnic calculations and obstinacy of the party leaders 
rendered these efforts futile. 
Of the three leading opposition candidates, Kenneth Matiba (Ford-Asili), Mwai 
Kibaki (DP) and Ajuma Oginga Odinga (Ford-Kenya), Kibaki was the only 
accommodating leader. However, his efforts were frustrated by Matiba and Odinga 
for whom unity without them as leader was inconceivable. Indeed, both were 
prisoners of their own ethnic constituencies. There were those in Ford-Asili, 
mainly the Kikuyu, who believed that without Matiba as a presidential candidate, 
their own electoral chances would come to naught. And Matiba himself not being 
in a position to assess the political situation independently, had been persuaded by 
his advisers that he was the strongest of the opposition candidates. Accordingly he 
increasingly became belligerent and obstinate. Similarly, Odinga refused to 
believe that none other than himself deserved to lead the opposition. At a rally in 
Busia in May, he reminded the nation that he could not be expected to surrender 
the presidency again having done so at independence (Nation 3.5.92). Odinga and 
the Luo believed that this was their turn. They had supported the Kikuyu in 1963. 
It was now the turn of the Kikuyu to pay back their 'debt'. But the Kikuyu would 
have none of this: some senior Kikuyu politicians such as Kimani Nyoike and 
Njenga Mungai chose to publicly deny that Odinga ever relinguished the 
presidency in favour of Kenyatta (Daily Nation & Weekly Review May 1992). To 
the Luo elites, and masses these denials only helped to strengthen their support for 
one of their own as they accused the other side of selfishness. What Ojuang
Ethnicity In The Electoral Process 57 
Kombudo, then MP for Nyakach, said in parliament on 16.9.92 captures the mood 
of the Luo: 
Luo have been blocked by Kikuyus who have been cheating them for a long 
time. Kikuyus have misled Luos for many years and Luos should know this 
and take action. 
He then advised the Luo to block the Kikuyu from taking over Kenya's leadership 
by rejoining KANU (Kenya Times 17.9.92). The Luo did not rejoin KANU: they 
chose to remain in FORD-Kenya and prepare for any eventuality. 
What further complicated the situation in the opposition movement was the 
struggle for the Kikuyu votes between Matiba and Kibaki. It is this struggle that 
made Kibaki hesitant in coming out decisively for an alliance with Odinga. He 
must have feared that such an alliance might erode his ethnic power base. More 
importantly, his party candidates from GEMA territory are said to have strongly 
cautioned against it, fearing the consequences of such a pact for their own electoral 
chances. A further irony is that Kibaki's options were quite limited. He could not 
enter into an alliance with Matiba or vice versa without the dreaded GEMA ghost 
being brought to the fore. The '25 percent formula' made that kind of a pact highly 
impolitic, because of its potential for scaring other tribes away. What followed was 
a vicious struggle for the Kikuyu votes between Matiba and Kibaki. By allowing 
themselves to be forced into that situation, the Kikuyu political elites betrayed the 
interests of the Kikuyu masses that they were supposedly out to promote. In the end, 
the division within the opposition movement played into the hands of Moi and his 
KANU party. Moi astutely exploited these divisions to the advantage of his 
presidential bid as well as that of his party's parliamentary candidates. The result 
was the KANU victory in both the presidential and parliamentary contests. 
The Election Results: 
In this section, the results of the presidential and parliamentary contests are 
explained in terms of how they were affected by ethnic loyalties. For reasons to be 
explained later, the focus excludes the minor presidential candidates. 
Presidential Contest 
For the four major presidential candidates (Moi, Matiba, Kibaki and Odinga), the 
results of the presidential contest show a strong correlation between a candidate's 
ethnic origin and the number of votes obtained in his ethnic region and in other 
regions. In every case, a candidate's ethnic affiliation secured him overwhelming 
support from his own group. In assessing the degree of ethnic support, the coverage 
has been restricted to rural districts usually predominantly inhabited by a single 
dominant ethnic group. A number of districts so inhabited constitute in our analysis
58 Walter O. Oyugi 
the 'heartland' of a given group. 
In Kalenjin heartland, comprising Kericho, Bomct. Nandi. Elgeyo-Marakwet 
and Baringo districts, Moi obtained 96 percent or 794.806 of the 829,367 votes 
cast. This represented 41.2 percent of the total votes he received throughout the 
country. In the Kikuyu heartland, which consists of the Central Province districts 
of Kiambu, Murang'a, Nyeri, Kirinyaga and Nyandarua, the votes were shared 
between Matiba and Kibaki who together received 95 percent of the votes cast 
there, representing about 40.2 percent of the votes the two together received 
countrywide. And in the Luo heartland - Siaya, Kisumu, Migori and Homa-Bay 
districts - Odinga received 549,464 or 95 percent of the 577,946 votes cast, 
representing 58.2 percent of the total votes he received nationwide. What emerges 
from these figures is that the Kalenjin, the Kikuyu and the Luo voted mainly on 
ethnic lines. In the process, the parties to which their ethnic candidtates belonged 
were exposed as ethnic parties. Consequently, such candidates performed poorly 
in one another' territory as the following data indicate: 
Moi: received about 0.4 percent of the votes cast in Luo heartland and 02 percent 
in Kikuyu heartland. 
Odinga:received 1.1 percent in Kalenjin heartland and 1 percent in Kikuyu 
heartland. 
Matiba and Kibaki: together received 1.35 percent in Kalenjin heartland and 0.8 
in Luo heartland. 
Table 1: % of Votes cast for Principal Presidential Candidates per Province 
Nyanza Western Coast Central Nairobi R Valley Eastern NE 
39 15 62 16 71 02 Moi 37 72 
44 Matiba 38 01 11 11 16 10 60 
02 06 Kibaki 07 10 18 50 05 35 
Odinga 14 75 17 20 05 01 02 08 
Table 2: Hypothetical Cases 
Coast Western Nairobi Nyanza NE Central Rift Valley Eastern 
21 62 07 40 60 16 23 95 Matiba & 
Kibaki 
Matiba, 
Kibaki & 35 82 82 57 24 28 62 96 
Odinga
Ethnicity In The Electoral Process 59 
A focus on the regional configuration of the votes tell more or less the same 
story. In Kenya as is in many African countries, regional divisions coincide with 
ethnic divisions. The central province is Kikuyu, Nyanza predominantly Luo, the 
Rift Valley is occupied predominantly by Kalenjin and kindred groups (Masaai. 
Turkana, Samburu); the Western is Luhya; and the North Eastern predominantly 
Somali. Only three regions are multi-ethnic in any significance sense: Nairobi, 
Coast and Eastern provinces. An attempt is made in Table 1 to analyze the pattern 
of regional support received by the respective candidates in the election. 
The significance of regional support has added dimension because of the 
constitutional requirement that to be elected as president, the winner besides 
receiving the majority or plurality of the votes cast, must receive at least 25 percent 
of the votes cast in at least five of the eight provinces. From Table 1, Moi received 
25 percent or more in five provinces, Matiba in three, Kibaki in two and Odinga 
in one. 
We shall explain the structure of regional support of the unsuccessful candidates 
first. For Kenneth Matiba of Ford- Asili the two provinces where he received better 
than 40 percent are predominantly Kikuyu. The 38% of the vote he received in the 
Western Province is attributed largely to Shikuku who was his running mate. 
Shikuku was perceived as the effective number two to Matiba in the party and his 
Luhya supporters were made to believe that in the event of a Matiba victory, he 
would become his Vice-President. In the remaining five provinces he received only 
16%, or less. 
Apart from the Central Province, Kibaki secured the 25% vote only in the 
Eastern Province where, as the GEMA torch bearer (it was believed at the time that 
the DP was financially backed by wealthy GEMA business interests), he pulled 
behind him the two GEMA communities of Mem and Embu, and also managed to 
do extremely well in Ukambani, especially in the old Kitui district where he had 
very strong catholic grassroots support14 in the hope of putting a catholic in State 
House. The Kamba support was possible perhaps only because there was no 
Kamba presidential candidate. 
Of the four serious presidential candidates, Odinga turned out to be the least 
likely to be elected. He got 25% only in his home province of Nyanza even though 
his party was the only one to receive at least a parliamentary representation in all 
the eight provinces. At the age of 81, Odinga was generally considered to be too 
frail for the demanding task of the presidency. His frailness was quite manifest 
during the campaign. Therefore, his potential non-Luo supporters seemed to have 
had some doubts about his ability to effectively cope with the demands of the office 
he sought to occupy. Added to that is, of course, the 'social distance' between the 
Luo and many ethnic groups in Kenya (Berg-Schlosser 1992) which appeared to 
have worked against Odinga. 
For the victor, Daniel arap Moi, a number of factors worked in his favour at a
60 Walter O. Oyugi 
time he had almost lost any hope of being re-elected. The division in the original 
FORD (for which he was partly responsible) drastically increased his chances at 
the polls. Taking advantage of his incumbency, he reportedly made offers of Vice 
Presidential appointment in the event of victory to the Luhya (Mudavadi), Kisii 
(Nyachae), Meru (M'mkindia), Kamba (Kalonzo Musyoka), and possibly to 
others as well (Weekly Review). These assured him of votes in those communities. 
He also used the bureaucracy (the Provincial Administration) to ensure that his 
adversaries were denied license to address meetings in certain areas for alleged 
"security" reasons (Weekly Review January 1, 1993 : 27). This was the case 
especially in North Eastern Province and in some parts of the Coast Province.15 
Opposition candidates were also cajoled and lured to pull out of the race in return 
for material benefits (Ibid) at the 11 th hour thereby further increasing the electoral 
chances of Moi and his party. The psychological impact of these tactics was 
immediate and rewarding. Having passed the 25 percent mark in five provinces, 
all he now needed was a plurality which he managed to get with 36.4% of the votes 
for himself, 25.6% forMatiba, 19.6% forKibaki, 17.1% forOdinga, 0.3% forChibule 
wa Tsuma, 0.3% George Anyona, 0.16% Mukaru Nganga and 0.12% J. Harun. 
Allegations of vote rigging were to be heard from losers and winners alike; but none 
could be proved. 
The other presidential candidates — Anyona, Chibule wa Tsuma, John Harun 
and Mukaru Nganga were generally regarded as mere spoilers if anything. With the 
possible exception of Anyona they all lacked a national stature and block ethnic 
support necessary to put them in power. But even Anyona, a Kisii and once 
regarded as a radical politician was on the eve of the elections, a political loner 
pursuing essentially a personal agenda. He had also been compromised after 
coming out of detention by allegedly consorting with some of the opposition 
people to whom he had earned his radical image — establishment figures such as 
his kinsman Simeon Nyachae, and Chibule wa Tsuma. His presidential candidature 
was incidental and came about with the last minute defection of the leader of his 
party. Therefore, he was not regarded as a serious candidate. John Harun Mwau 
was also unknown — not even among his Kamba people (his political base was 
Nairobi where he even failed to win a Parliamentary seat). The other candidates — 
Mukaru Ng'ang'a, like Tsuma, Anyona and Harun entered the race simply because 
he had his own party to represent. But the party had no following anywhere in the 
country not withstanding his Kikuyu origin. The four were merely 'the also ran' 
presidential candidates. 
Some Hypothetical Cases 
I wish now to address two hypothetical cases. The first is about what would have 
happened had the Kikuyu fielded one candidate. The second is about the prospects 
of a united opposition against Moi. (See Table 2.)
Ethnicity In The Electoral Process 61 
The Kikuyu have been heard to lament after the elections that had they put up 
only one candidate, the result would have been a Kikuyu president. This lamentation 
is based on the combined strength of the presidential votes for Matiba and 
Kibaki: 1,354,856 + 1,035,507 or 45.2%. According to this scenario, Moi with his 
36.4% would have come second! What were the realities? 
The merger between Ford-Asili and DP under Matiba and Kibaki respectively 
would have completely changed the voting equation. Such an alliance would have 
been seen as a Kikuyu or GEMA alliance and would have scared away many non- 
GEMA voters who still felt it was too soon for another Kikuyu president. Shikuku 
for example would have found it difficult to remain in the group, for it is unlikely 
that he would have remained the only vice presidential hopeful in such an alliance. 
It is therefore unlikely that the Luhya supporters of Shikuku who voted for Matiba 
and Ford-Asili in large numbers (16% of Matiba's total votes) would have done so. 
Therefore, the voting equation would have been entirely different. Furthermore, it 
should be recalled that a combination of Matiba and Kibaki votes passed the 25 
percent mark only in Central Province, Eastern , Western and Nairobi, and would 
certainly have been reduced to three provinces with the loss of Shikuku on the 
ticket. In the event, even if they received a 'winning' plurality, the 25% requirement 
in 5 provinces would have either forced a run-off with all sorts of possible 
realignments thereafter. For the same reason it could be argued that if such 
realignment took place, it would have most probably not worked in favour of the 
Kikuyu. 
A second hypothetical case involved an alliance between Matiba, Kibaki and 
Odinga — in other words Kikuyu-Luo alliance. Such an alliance would have, 
according to the results, assured them victory both in terms of plurality of votes and 
secured 25% of the votes in seven of the eight provinces. However, the chances are 
that such an alliance would have scared most of the so-called minority ethnic 
groups. In the event, the alliance would most likely have lost the Western and 
possibly Rift Valley provinces (by loosing Luhya votes in the Rift Valley 
especially Trans Nzoia and Uasin Gishu) in addition to the North Eastern province. 
All the same, that would have left them with the five required provinces 
to romp home. Indeed, had the original FORD remained intact as a national 
movement, it could have got the 25% of the votes required in all the provinces. But 
that was never to be. Personal aggrandisement and ethnic calculations combined 
to ensure the collapse of the united opposition movement on the eve of the 
elections and thereby guaranteed the victory of Daniel arap Moi in the presidential 
elections. 
Parliamentary Results 
Kenya's electoral law provides for single member constituencies where the 
candidates with the highest number of votes regardless of the percentage scored is
62 Walter O. Oyugi 
Table 3: Parliamentary Seats won by Individual Parties Per Province 
Eastern NBI Central NE Coast 
1 Kenya African National Union 21 17 
1 1 2 
0
9
(}
1 
6
0
0
0 
0
1 
0
0 
0 0 
0
1 
14 
II) 
II 
0
0
I) 0 0 
Forum tor the Restoration 
of Democracy (FORD-KENYA) 
Forum for the Restoration 
of Democracy (FORD-AS1LI) 
Democratic Party 
Kenya Social Congress 
Kenya National Congress 
Parly of Independent 
Candidates 
Kenya National Democratic 
Alliance 
8
1 
0
0
0
0
1 
0 
TOTAL 25 32 1(1 8 
Source: Daily Newspapers 
n
0 
20 
declared the winner. The operative electoral practice is one of the many colonial 
imports left behind by Britain at independence. 
In 1992, a total of eight parties contested the elections. The results are presented 
in Table 3. A number of factors combined to explain the defeat of the opposition 
in the parliamentary elections as well. First was the assumed strengths of the parties 
according to region. It was assumed that KANU candidates would generally do 
well in the former KADU areas (Rift Valley, Coast and to some extent Western 
Province) and in N. Eastern where the KANU government had denied the 
opposition the right to campaign. KANU's best performance in both presidential 
and parliamentary contests were to emerge in those provinces (see Table 1 for 
presidential contest and Table 3 for the parliamentary). Of the 100 parliamentary 
seats won by KANU, 71 % came from these regions. Nairobi, Nyanza and Central 
were regarded as opposition bastions and indeed the opposition received from 
these provinces 55 or62.5% of the total number of seats they won. Eastern province 
that was expected to be closely contested (this was also true of Western) was 
captured by KANU when it won 21 out of the 32 seats or 66%. It should be noted 
however that most of the seats came from non GEM A districts of Ukambani, Isiolo 
and Marsabit (18 out of 21 or 85.7%). In the GEM A districts of Embu and the old 
Meru, the opposition, (i.e. DP (6), Ford-K (1), KNC (1) obtained 8 out of 11 or 
72.7%. It is noticeable from the statistics that there is a close relationship between 
presidential and parliamentary performance by region and ethnic group. In the 
Parliamentary and Presidential elections KANU led in Rift Valley, North Eastern 
and Coast Provinces. It lost Nairobi, Central Province and Nyanza to the opposition 
(especially Ford-K and Ford-Asili), shared Western in both Parliamentary and 
Presidential contests and lost Eastern to the opposition in the Presidential votes but 
won in the Parliamentary — mainly from Ukambani where it was alleged by the 
RV Total Western Nyanza 
III 100 
3 31 
7 31 
23 
01 
01 
01 
36
2
4
2
0
0
0
I) 0 
7 
20 
0
1
1 
0
0
1) 
0
0
0
0
0 
20 44 188 29
opposition that the elections had been rigged in favour of KANU by especially 
senior Civil Servants from the district.16 
Again it is evident that there was direct relationship between ethnicity and 
voting in the parliamentary contest as well. All the 24 central province seats went 
to Ford-Asili and DP. All the 19 seats from Luo Nyan/.a went to Ford-Kenya, and 
all the 17 seats from Kalenjin heartland went to KANU. And in Nakuru district 
which is predominantly settled by Kikuyu but is in R. Valley, three out of the four 
seats went to Matiba's Ford Asili. 
The votes cast for the various parties compare well both in presidential and 
parliamentary contests. A glance at the statistics (Table 4) indicates a difference of 
about 8% between presidential and parliamentary votes in favour of KANU. That 
difference could however be explained by the excessive "high voter turn out" in the 
Kalenjin heartland in the presidential contest. 
Table 4: Total votes in Belt The Presidential and Parliamentary Elections 
Presidential Party 
1,962,862 KANU 
1,404,266 FORD-A 
1,050,617 DP 
944,197 FORD-K 
15,393 KNC* 
6,449 PICK* 
14,253 KSC* 
8,527 KENDA* 
— SDP* 
5,406,664 
Source: WR 1.1.93 P.5 and other Dailies. 
* approximations for the minor parties. 
During the elections and before, there was a lot of talk by some politicians to the 
effect that the distribution of seats were disproportionate to the population and 
tended to favour certain areas. But available evidence tends to suggest only 
marginal differentials that could not be said to favour any group or region. 
According to these statistics, the regions (provinces) which show underrepresentation 
in relation to the others are Central 02, Eastern 01, Nairobi 05, 
Nyanza 01, Western 03. Those showing over-representation are Coast 04, R. 
Valley (01), and North Eastern (07). What is evident here is that Nairobi has been 
Ethnicity In The Electoral Process 63 
Parliamentaary % % 
1,400,549 28.4 36.3 
27.5 1,356,065 26.0 
22.9 1,127,489 19.4 
18.2 894,952 17.5 
1.6 77,844 0.28 
1.0 51,505 0.12 
16,370 0.26 0.3 
771 0.02 0.16 
105 0.00 — 
4,925,650 99.9=100 100
64 Walter O. Oyugi 
Table 5: National Assembly seats as a % of Provincial Population 
Province 
Centra] 
Coast 
Eastern 
North 
Eastern 
Nairobi 
Nyanza 
Rift Valley 
Western 
Total 
Source: Population see figures in column 2 — 1994/96 National Development 
Plan, Nairobi: Government Printer, 1994, p. 27. 
under-represented and North Eastern has been over-represented. These figures are 
only indicative and suggestive. They suggest that were the seats to be distributed 
according to population size, 7 seats would have gone to the opposition (Nairobi 
and Central). The six seats in Nyanza, Western and Eastern would have gone either 
way as the results of the election reflect. On the other hand, on the overrepresentation 
side, the 12 seats for Coast, North Eastern and Rift Valley would 
have led to a loss for KANU — leaving KANU with 88 (100-12) as opposed to 
oppositions 95 (88+7). The result would then have been determined by the 5 seats 
for Nyanza, Western and Eastern. Hypothetically, it would have been too close to 
call. But as we all know, there is no country in the world where demarcation of 
constituency boundaries is only determined by population factor. That might 
explain the apparent over-representation of regions that are largely arid or semiarid 
i.e North Eastern and Coast provinces. 
Indeed, it would be wrong to analyse the electoral results in terms of the 
performance of KANU versus the opposition. Opposition parties were fragmented 
and in being so, put up candidates against one another thereby giving the ruling 
party a margin of advantage in contested areas (i.e areas not under the influence of 
Seats Population % 
25 14.5 3,626,000 
20 2,155,000 8.6 
32 17.4 4,334,000 
10 1.6 408,000 
8 1,678,000 6.7 
29 4,041,000 16.2 
44 5,690,000 22.8 
20 3,035,000 12.2 
188 24,967,00 100 
%of 
Total 
% Under 
or over 
Representation 
- 1.2 13.3 
+ 2 10.6 
-0.4 17.0 
+ 3.7 5.3 
-2.4 4.3 
-0.8 15.4 
+ 0.6 23.4 
- 1.6 10.6 
— 106
a particular ethnic group such as Eastern and Coast; and Western which did not 
have a presidential candidate from the area). In one case, there was even intraethnic 
contest (i.e Matiba's Ford-Asili versus Kibaki's DP in Central and in the 
Kikuyu diaspora in Rift Valley). Table 6 shows the number of candidates put by 
each party per province and nationally. 
Table 6: Distribution of Candidates per Party per Province 
PARTY 
KANU 
Ford-A 
DP 
Ford-K 
KNC 
Pick 
KSC 
Kenda 
SDP 
Source: Daily Newspapers and Weekly Review 1.1.93. 
Several malpractices in the electoral process arising from the absence of a level 
playing ground also worked against the opposition. In certain areas, the Provincial 
Administration in collusion with the ruling party directly interfered with the 
conduct of the campaign, for example denying opposition candidates the freedom 
to move freely and to meet the voters in the process, insisting that they could only 
address 'licensed gatherings'. But the same Administration turned a blind eye 
when KANU candidates were involved. Furthermore, the state electronic media 
gave full coverage of the activities of the ruling party and would mention the 
activities of the opposition parties only where doing so served the interest of the 
ruling party.17 And in the so-called KANU zones, especially in Kalenjin heartland, 
opposition candidates were prevented from presenting their nomination papers in 
full view of electoral officials and security forces. As a result, all the seats in the 
Kalenjin heartland except one in Kericho with mixed population of settlers (i.e 
Chepalungu) were uncontested. A total of 17 candidates were elected unopposed 
(Weekly Review 1.1.93). 
Other Factors 
Elsewhere especially in North Eastern Province and parts of Coast Province such 
Ethnicity- In The Electoral Process 65 
Eastern NE Nairobi Coast Central 
(32) (10) (20) (8) (25) 
32 8 10 20 25 
29 8 9 17 24 
32 9 19 8 24 
27 17 8 8 14 
8 6 10 2 12 
4 1 1 5 5 
-> 1 2 — — 
— — — — 04 
01 — — — — 
Total R.V NYA Western 
188 (44) (29) (20) 
188 29 44 20 
142 22 18 15 
153 27 18 16 
136 17 29 16 
46 02 01 5
5 29 03 5 
09 4 — — 
04 — — — 
01 — — —
66 Walter O. Oyugi 
as TANA River District, 'security' reasons were used to deny opposition parties 
access to the electorate. As one Weekly put it with regard to North Eastern, 
While KANU appeared to hold rallies at will in the province, opposition 
parties permits for similar functions were always invariably cancelled at the 
last minute because of what were termed as 'security reasons'. There were 
also several incidents when trips to the province organised by opposition 
figures were marred by lack of clearance from the police department for 
aircraft meant to be used by the party leaders (Weekly Review 1.1.93). 
These moves were calculated to work to the advantage of the ruling party; and 
indeed they did. With the opposition virtually locked out of the electoral process, 
it was quite easy for KANU to manipulate proceedings to its advantage. This and 
the general frustration the opposition candidates faced at the hands of the state 
forced many opposition candidates to defect to the ruling party at the eleventh hour 
thereby denying the opposition parties the opportunity to field replacements. These 
tactics were responsible for the so-called 'massive' victory of KANU in arid and 
semi-arid areas of the country. Eight out of the ten seats in North Eastern for 
instance were won by KANU. By virtue of these tactics KANU also won all the 17 
seats in the other districts generally characterised by insecurity Tana River (3), 
Turkana (3), West Pokot (3), Samburu (2), Isiolo (2), Marsabit (4). 
The behaviour of the Electoral Commission (appointed by the President shortly 
before the opposition parties were legalised in December 1991) also left a lot to be 
desired. Even where electoral malpractices were committed in their full view (e.g 
prevention of opposition candidate from presenting their nomination papers by 
sections of the security forces and KANU supporters) they looked on helplessly. 
A case in point was when six of those prevented from presenting their nomination 
papers petitioned the Courts and the judges ordered that no election should take 
place in the six constituencies concerned. The order was ignored, elections held 
and KANU candidates declared or elected unopposed (Weekly Review 1.1.93 :47). 
At issue also was the public standing and credibility of the Commission's 
Chairman, a man who had been twice retired as a judge under compromising 
circumstances and therefore was not morally strong enough to resist pressures 
from the ruling party.18 
Summary and Conclusion 
The politics of the 1992 general elections in Kenya, demonstrates that ethnicity 
continues to be a major force influencing the behaviour of politicians and voters 
alike. What is more, where power and wealth were at stake, ethnic relations became 
conflictual. The 1992 elections involved the choice of power holder(s) and in the 
process, the determination of the structure of access to state patronage — the major
Ethnicity In The Electoral Process 67 
source of wealth in Kenya. The ethnic conflicts in various parts of the country 
during the elections should be viewed in this light. The elections also manifested 
how the elites can mobilise ethnic passions to defend and or promote what is 
otherwise their narrow sectional interests. The masses followed their leaders 
because of the lingering belief that only 'one of your own' can best serve communal 
interest if placed in a position of power. But it was also clear that ethnic ideology 
has its limitations. Intra-ethic divisions were manifest where parochial interests 
came into play. The emergence of splinter parties led by members of the same 
ethnic group was the inevitable consequence of such contradictions. Ethnic 
solidary nevertheless still remained a major factor influencing both presidential 
and parliamentary elections. The ethnic groups that were regarded as being in 
opposition remained loyal to their ethnic leaders just as those perceived to be in the 
ruling party did. 
Notes 
* Professor of Government at the School of Humanities and Social Services, 
University of Nairobi, Kenya. 
1. See also Bennett 1963 for a different viewpoint. 
2. SeeGertzel 1970; Okoth-Ogendo 1972; Oyugi 1994 for a discussion of the 
woes of KADU before it was disbanded. 
3. For more on this see Mueller 1984 and Gertzel 1970. 
4. As if by fate, Robert Ouko, a popular Minister of Foreign Affairs from 
Luoland was killed in cold blood after being taken away from his home in 
February 1990 and the Commission set up by the Government to probe the 
killing increasingly received evidence suggesting complicity by individuals 
well placed in the regime. The regime's moral authority was again put under 
scrutiny. 
5. Quoted in Josseph. (1987). 
6. The problems which the Kikuyu-controlled financial institutions faced in the 
late 80s have often been cited in this regard. 
7. Indeed, Matiba's inability to read and write was later to be confirmed during 
the swearing-in of the MPs after the elections when it turned out that he could 
not actually read the prescribed text. 
8. This was in reference to Odinga's call for Kenyatta's release on the floor of 
the Legislative Assembly in June 1958 at a time when no politician dared to 
do so. Soon after, the release of Kenyatta became a national movement with 
Odinga in the forefront. 
9. Weekly Review 9 October 1992 quoting statement made early in the year. 
10. Personal communication. 
11. Matiba being a wealthy politician who for a long time had been associated
68 Walter O. Oyugi 
with 'establishment' politics and Shikuku being a populist politician known 
for his anti-establishment crusade. 
12. The acronym was coined by the press in the mid-90s as a collective name for 
the four ethnic groups during their (initially) clandastine peace meetings with 
a select group of GEMA business and political elites aimed (supposedly) at 
finding a lasting solution to the problems of (Kikuyu) victims of the clashes 
in the Rift Valley. 
13. See NEMU report on 1992 elections and Daily papers. 
14. Kibaki is a practising Catholic and many catholic bishops are reported to 
have openly campaigned for DP. 
15. Note that it was in these two provinces that he received the highest percentage 
of votes after Rift Valley. 
16. Ukambani was one of the areas where vote counting and announcement were 
inordinately delayed (Weekly Review 1.1.93 quoting election monitors) to 
justify such allegations. 
17. It should be noted in this respect that Kenyan rural voters rely mostly on the 
radio for news which is fully controlled by the State communication network. 
Therefore such biase in the media was bound to influence the electoral 
choices of rural voters. 
18. His appointment had been seriously challenged by the opposition leaders on 
that ground. 
References 
Bates, R. (1974) "Ethnic Competition and Modernization in Contemporary 
Africa" Comparative Political Studies Vol. 6 No. 4. 
Berg-Schlosser, Dirk "Ethnicity, Social Classes and Political Process in Kenya" 
in Oyugi, (1994). 
Bennet, George and Rosberg, Carl (1961) The Kenyatta Election Kenya 1960- 
1961 (London: Oxford University Press. 
Bennett George (1963) Kenya: A Political History — The Colonial Period 
Diamond, L. (1986) Class, Ethnicity and Democracy in Nigeria (Syracuse: 
Gertzel Cherry (1970) Politics of Independent Kenya (Nairobi: East African 
Gulliver P.H. (ed.) (1969) Transition and Tradition in East Africa (Berkeley & L. 
Horowitz, D.L. (1985) Ethnic Groups in Conflict (Berkeley & L. Angeles: 
Lake mann E. (1974) How Democracies Vote: A study of Electoral Systems 
(London: Oxford University Press). 
Syracuse University Press). 
Publishing House). 
Angeles: University of California, Press). 
University of California, Press). 
(London: Faber and Faber).
Ethnicity In The Electoral Process 69 
Leys, Colin T. (1974) Underdevelopment in Kenya: The Policies ofNeo-Colonialism 
(Nairobi, London: Heinemann). 
Mueller, Susan (1984) "Government and Opposition in Kenya" Journal of Modern 
African Studies Vol. 22 No. 3 pp. 399-427. 
Nnoli, Okwudiba (1978) Ethnic Politics in Nigeria (Enugu: Forth Dimension). 
Nnoli, Okwudiba (1989) Ethnic Politics in Africa (Ibadan: Vantage Publishers). 
Okoth-Ogendo, H.W.O. (1972) 'The Politics of Constitutional Change in Kenya 
Since Independence 1963-69' African Affairs Vol. 71 No. 282. 
Oyugi, Walter, O. (1993) "Ethnic Politics in Kenya" in Nnoli, Okwudiba (ed) 
Ethnic Conflict in Africa CODESRIA (forthcoming). 
Oyugi, Walter, O. (ed.) (1994) "Uneasy Alliance: Party State Relations in Kenya" 
in Oyugi, W.O. (ed.) Politics and Administration in East Africa (Nairobi: 
East African Educational Publishers). 
Joseph, Richard A. (1987) Democracy and Prebendal Politics in Nigeria 
(Cambridge: Cambridge University Press). 
Newspapers/News Magazines 
Daily Nation (Nairobi) 
Kenya Times (Nairobi) 
The Standard (Nairobi) 
Week End Mail (Nairobi) (Has since been out of circulation). 
Weekly Review (Nairobi).