Citizenship and ethnicity: an examination of two transition moments in Kenyan politics.
In these failed transitions where ethnicity is evidently a factor in mobilizing for political change or for anticipated electoral politics, a more robust explanation for the conflict can be found by applying the theoretical framework of citizenship. While the conflict between the postcolonial state and ethnic subnationalism has been a consistent theme in studies of African politics (e.g., Diamond 1988a and b, Glickman 1995, Olorunsola 1972, Rothchild 1968a and b, Rothchild and Olorunsola 1983, Young 1976), the citizenship lens provides a more precise way of theorizing the relationship between ethnic identity, authority, and legitimacy.(2) Most research on ethnic politics, including more sophisticated analyses cognizant of the contextual and constructed nature of ethnicity, imply that ethnic mobilization and political action flow reflexively from identity. The moral and temporal underpinnings of the process that lends authority and legitimacy to ethnicity, and hence undergirds ethnic action, is not explained. In short, present work does not explain why rational individuals respond to ethnic mobilization. For instance, the well-known propensity for ethnically mobilized "census-type" elections (Bienen and Herbst 1996, 36-7; Horowitz 1985, 83-7; 1991, 91-100; 1994) is an empirical characterization that can be explained as a rational response to citizenship demands within ethnic communities.(3)
This paper demonstrates, with evidence from two transitional periods in Kenya, how ethnic politics arises due to different and competing forms of citizenship in the nation-state and in ethnic communities. This perspective enriches the explanation of why ethnic loyalties undermine national stability by pointing to not only different loyalties (the stock explanation) but also different forms of loyalty. Moreover, citizenship theory highlights how particular democratic institutions are perceived by different elites as presenting opportunities for or constraints to their drive for state power through ethnic mobilization within electoral politics. Citizenship theory is especially appropriate in reviewing ethnicity in Africa because it introduces a discourse of rights and obligations as well as of opportunity and constraint in a political community. While such a discourse has been absent in previous studies, it is at the heart of the current political evolution in Africa, propelled by actors who see democracy as essential to assuring equitable access to state power and to curbing the abuse of that power against individuals and groups. By allowing us to invoke and envision individual rights within the context of group rights in a democratic society (Kymlicka 1995), citizenship theory is especially helpful in understanding ethnic politics and claims for self-determination within emergent democracies.
Fundamentally, citizenship "defines those who are, and who are not, members of a common society" (Barbalet 1988, 1). It expresses "a set of normative expectations specifying the relationship between the nation-state and its individual members which procedurally establish the rights and obligations of members and a set of practices by which these expectations are realized" (Waters 1989, 160). Therefore, citizenship allows one to participate in a community while enjoying certain rights and being obligated to perform certain duties in return. To the extent that ethnic groups in modern nation-states constitute identifiable communities of belonging and social practice, the traditional trajectory of citizenship theory has hindered its application to ethnic politics for two reasons. First, most discussions of citizenship that draw from North American and European experiences assume a single political community - the modern state - in which individuals participate and into which those previously excluded are admitted through the expansion of civil, political, and social rights (Marshall 1965). Thus, citizenship is held to be coterminous with the modern, democratic nationstate. Second, while debate abounds about differing conceptions of citizenship - liberal, civic-republican, and libertarian - the dominant view is that only one form can or ought to exist in a modern state, viewed as a single political community.(4) A related but less rigid consensus holds liberal citizenship as the hallmark of modern democracy.
The dominant understanding of citizenship as coterminous with the nation-state has recently been challenged by an emerging recognition of other communities as venues for elaborating citizenship.(5) Thus, in an age of increased globalization, Soysal (1994) and Roche (1995) view "transnational citizenship" as a distinct reality superseding the narrow national scope in terms not only of belonging but also of rights and obligations. This rings especially true in the wake of European unification. Alternatively, Stewart seeks to validate the subnational or substate level as an arena of enacting citizenship, which he terms "democratic citizenship," defined by a "common membership of a shared and imminent community" (1995, 75, emphasis in original). Such a citizenship is claimed and defined by political action launched from small communities of shared values and experience and enacted in the public sphere. This recognition of spheres of citizen formation other than the nation-state makes possible an analysis of ethnic groups as political communities in which citizenship can be formed and sustained.
The claim of the primacy of a singular form of citizenship in one political community is countered by Peled's (1992) analysis of Israel, in which he demonstrates the existence of two different forms of citizenship for two ethnic groups within one state. For Jews - the core members - the state prescribes a civic-republican citizenship, which allows them to secure a large bundle of rights and to participate fully in the national public sphere. For Arabs, the Israeli state extends liberal citizenship, which allows them "civil and political rights but [bars them] from attending to the common good" (Peled 1992, 436). A significant example of such exclusion from duty to the "common good" is the debarment of Arab Israelis from conscription. While Peled's analysis still assumes the nation-state as the central locus of defining and sustaining citizenship, his illustration of the coexistence of liberal and republican forms of citizenship is useful to apply to ethnically divided states in Africa. When combined with the recognition of substate political communities as arenas of citizen formation, the possibility of coexisting citizenships and forms of citizenship provides a powerful instrument to explain ethnic conflict in democratic politics in Kenya.
CITIZENSHIP AND ETHNICITY
Ethnic identity in Africa is a relatively recent phenomenon whose salience is largely a product of colonial rule (Ekeh 1975, Kalinga 1985, Young 1976) and of postcolonial dynamics in which elites have continued to reify ethnic identity for political mobilization (Lentz 1994, Sithole 1985, van Binsbergen 1994). In the case of Kenya, groups viewed today as monolithic entities were in fact not coherent communities before colonial rule. The Abaluhya, for example, are an ethnic group created by colonial anthropologists, missionaries, and administrators.(6) The name Baluhya dates only to the 1920s, when it was used to profess the unity of Bantuspeaking groups to which the British referred as the North Kavirondo, an administrative aggregation of sixteen separate groups; although proximate in location and having related languages, they differed sufficiently from one another to make unity elusive (Abwunza 1993, 130-5; Bennett and Rosberg 1961, 171; Kaplan 1984, 90-1, 96-7; Ogot 1981, 8). The British classification of these groups as one was strategically adopted by elites from the subgroups to assert unity for political goals, for example, through the creation of the Baluhya Political Union to join the independence struggle. Today, the artificiality of Abaluhya identity is marked by perennial divisions, especially among politicians, interrupted by brief interludes of unity when corporate interests are threatened.(7) Despite its recent and artificial nature, however, ethnic identity is instrumental in shaping individual interests and actions in the modern Kenyan state.
A claim to the existence of ethnic citizenship must rest on evidence of identity, authority, and legitimacy for members of an ethnic group. To elaborate, this relationship can be compared with "state citizenship" (Stewart 1995). Essentially, the difference between ethnic and nation-state citizenship lies in the Weberian legal, rational, and bureaucratic frameworks that uphold identity, legitimacy, and authority in the nationstate, as opposed to the social customs, social practices, and nonbureaucratic structures that define and uphold citizenship in ethnic groups. Identity is elemental in both realms because it demarcates the boundaries of belonging. In the postcolonial state, national identity is legally circumscribed, normally endowed on all those born or naturalized within specified territorial borders (see Shklar 1991, 3-5). An individual's claim to nationality is validated in the public realm by legal and bureaucratic instruments, such as identity cards, passports, and voting cards. In contrast, ethnic identity rests on a socially, as opposed to a legally, constructed definition of belonging. As with early Athenian identity (Frost 1994), belonging in the ethnic community commonly is established by birth (or marriage or adoption) into a family and kin who consider themselves and are considered by others to belong to a community that believes in a shared history and values. While this identity may be signified by language or naming or may be literally inscribed through rituals such as circumcision, it is validated by the individual's participation in that community and by his or her interaction with the institutions of legitimate authority in the community.
In the nation-state, authority is vested in a Weberian state which enumerates rational and legal codes to govern relations among individuals in its community. In the modern democratic iteration of the nation-state, legitimacy is derived from the democratic right of citizens to affect these institutions. Yet, this is not the only historical possibility of state legitimacy, as theocratic and monarchic traditions obviously suggest. In ethnic groups, legitimacy and authority rest on nonlegal, nonbureaucratic forms of organization and cultural practice. Ethnic authority therefore is vested in the family, clan, extended kin, and beyond, to a community "imagined" as one (Anderson 1991). In the absence of a central state, authority is dispersed in the community and legitimated by customs that sanction and that are publicly enacted, beginning at the clan level, which is the basic unit of solidarity and public discourse outside the family. Ethnic authority in postcolonial state arenas begins as social practice in ethnic arenas and can be wielded by socially legitimated persons or structures, such as real or reinvented traditional authorities, ethnic associations, or charismatic leaders who gain recognition from their ability to underwrite communal practice by providing resources or a reputedly veritable memory. It is in the former capacity that politicians are able to draw upon the "pull of the tribe" (Jones 1967, 107) to advance goals in nation-state arenas.(8) Whatever the institution in which ethnic authority is vested, legitimacy is drawn from the satisfaction of certain material and nonmaterial needs of members in the identity group.
The ways in which social practice elaborates the norms of communal authority and legitimacy and the rights and obligations of citizenship within an identity group can be noted in at least two domains. One is the moral economy, whose foundation is norms of reciprocity as opposed to contract and whose bonds are strongest within the identity group. The moral economy enables individuals in various contexts to rely on nonbureaucratic mutual aid networks and to reciprocate toward those who belong to a common society. Examples include those better off helping relatives and clan members find jobs or pay school fees, as well as regular contributions to weddings and funerals, even for persons with whom face-to-face contact has never been established but who are imagined to belong to one's community. The moral economy can be formalized in ethnic organizations (such as hometown or funeral associations) that provide social insurance in the absence of state welfare programs.
The second domain is ethnic rituals that elaborate identity, legitimacy, and authority. As Comaroff and Comaroff (1993, xvi) argue, ritual is often "a vital element in the processes that make and remake social facts and collective identities. Everywhere." For instance, such initiation rites as circumcision continue to represent among most groups in Kenya the acquisition of full citizenship in the ethnic community. Among the Kikuyu, uninitiated men (rare) cannot inherit property or adjudicate in clan disputes, and parents of initiates ascend to an elder class which holds putative power in clans (Davison 1996, 231-2). Such rituals establish a hierarchy of power within the community that, when combined with other icons of power and status in the modem state, enables elites to mobilize within the ethnic community for interests in the secular state. At the same time, individuals can hold such persons accountable within the communal realm for actions in the state realm and especially for extracting resources from the state on their behalf (e.g., Ekeh 1975). An example is traditional oaths, which affirm identity and obligation within the ethnic community but which often have a purpose in the state arena, such as securing electoral victory or political succession. These oaths are recurrent motifs in national politics in Kenya.(9)
Whether "imagined" (Anderson 1991), "invented" (Vail 1989), or instigated in current memory by ethnic or cultural "entrepreneurs" (Lemarchand 1994, 5; Young 1976, 137-9), ethnic identity can become pervasive and can inspire collective action in pursuit of communal interests and privilege (Abwunza 1993, 136). Indeed, although socially constructed, ethnicity is reified for group members and outsiders by social practice, including ideologically legitimated customs, language, and histories. Furthermore, even as it is manipulated and reshaped over time, ethnicity is experienced as a concrete web of relationships with real persons and institutions in the present. Significantly, it is a source of security at particular historical moments, even when these are themselves part of the dialectic of reformulating identity.
DUAL AND COMPETING CITIZENSHIP
In postcolonial Kenya,(10) the socially enacted relationship between ethnic identity, authority, and legitimacy competes with the legally sanctioned membership, authority, and legitimacy of the nation-state.(11) Since neither has been able to erase the other, most individuals assume contingent and hierarchical allegiances depending on the arena of competition in which they find themselves. As Abwunza (1993, 147) shows among the Avalogoli subgroup of the Abaluhya, individuals and elites elaborate the following hierarchy of communities of allegiance. Foremost is one's own or the closest clan among all Avalogoli; second are the Avalogoli among all Abaluhya; third are all Abaluhya among Kenyans; and fourth are all Kenyans among any others. This hierarchy of attachment underscores the lack of an "unambiguous membership" (Waters 1989, 1) for individuals in postcolonial African states. Most significantly, individuals participate in two substantive publics (Ekeh 1975) and submit to two substantive authorities (Sklar 1993) - the ethnic group and the nation-state.
This duality of citizenship is complicated further by the fact that it is conceived differently in the postcolonial state and in the ethnic communities. The traditional dichotomy of liberal and civic-republican citizenship provides a useful framework for describing each conception. Essentially, the liberal vision holds that rights inhere in individuals, exist prior to community, and are guaranteed with minimal obligation to the community. The civic-republican vision considers rights not as inherent but as acquired through civic practice that upholds obligations to the community.(12) In the communal realm, citizenship takes an active civic-republican form, while in the modern state it is defined in liberal terms deemed appropriate for (elusive) constitutional democracy. In the modern state, a republican virtue is an ideal sought but not yet achieved at the national level, although in the parochial communities it is extracted in a variety of ways. Within the postcolonial state, liberal citizenship qualifies one to participate in the inclusive national community, while in the ethnic community republican citizenship requires members to participate in the group's preservation, especially in competition against other communities and against the national community - unless the state is controlled by fellow community members.
Liberal Citizenship in the Modern State
Analysts in the modernization tradition envisioned that the newly independent African countries eventually would forge a common national identity in place of the multiplicity of ethnic identities (Azikiwe 1978, Bendix 1964; cf. Enloe 1973). Critical to this transformation was overcoming the "crises" or challenges of reorienting identity, authority, and legitimacy by substituting the modern Weberian state for parochial ethnic communities and liberal democracy for traditional authority patterns (Binder et al. 1971). The liberal, majoritarian democracies installed in the new nations projected a liberal form of citizenship, which bestows on a person the status of a citizen as an individual member of the modern state. "The liberal-individualist conception of citizenship is an essentially 'private' conception. It is part of what is meant by calling individuals sovereign and morally autonomous beings, that they choose whether or not to exercise the rights of the status of citizen in the public or more narrowly the political arena" (Oldfield 1990, 178). Such status does not demand that the citizen perform any duties to retain these rights or membership in the political community.
A similar liberal conception of citizenship was central to the democracy movement in Africa in the late 1980s and early 1990s. African democrats demanded that the illiberal, authoritarian states that had replaced the initial democracies return sovereignty and rights to the individual citizen. As is evident from the monotony of liberal, majoritarian, multiparty models demanded and instituted across the continent, African democrats of the 1990s assumed that what led to the demise of liberal democracy in the 1960s was merely its implementation, not its irreconcilable relationship with ethnic citizenship. They continued to believe that one can achieve, for example, a functional nation of 100 million Nigerians acting as "sovereign" and "morally autonomous" individuals in a secular state, not as agents of 100 separate moral and organic communities - to which allegiance is legitimated by a web of social practices in which the daily life and significant life events of individuals are embedded.
Republican Citizenship in the Ethnic Group
In the ethnic community, citizenship can best be characterized as illiberal and republican. Whereas liberalism is centered on the individual, ethnic groups are centered on the community, and a person cannot claim rights that would jeopardize group claims. Individuals are therefore not "sovereign" or "morally autonomous" but instead gain rights and deserve defense only as active members of a community. Such benefits are secured by obligations and participation necessary "to define, establish and sustain a political community of fellow citizens" (Oldfield 1990, 181). A citizen is such through participation in the public life of the community.
Curtis's (1995) study of male migrants in Nairobi in the 1980s provides a good illustration of the rights and obligations imposed by ethnic citizenship in the modern socioeconomic setting. Curtis found that migrants from all class backgrounds maintained ties with their rural home (largely ethnically homogeneous) and participated in informal but extensive mutual aid networks with others who claimed the same or proximal rural base. These networks of "obligation and opportunity" underscore reciprocity in aiding members of a common society financially, in job placement, and in spreading news from home (Curtis 1995, 113-45). Moreover, these ties and those that each family or clan maintains with a rural base (often by holding land) establish membership in a larger imagined and experienced community. That community combines the extended clan in the native region with the experiential community in the urban diaspora and its assumed kin extensions in other areas.
The Curtis study (1995, 139) quotes Grillo's (1973, 63) apt description two decades earlier of the obligations of ethnic citizenship that migrants face:
Through his rural network he keeps alive his status as a member of the rural community. . . . By making regular visits home, presenting gifts to kinsmen, marrying a locally approved girl, by assiduous attendance at funerals and other ceremonial occasions, by building a house, he is making an investment for his future return to the rural area, paying in advance for the goodwill and cooperation he may sometime require, or which through his death his family may need any moment. . . .A break. . .is tantamount to a break with kin and ethnic ties.
It is clear from the foregoing that the ethnic community draws legitimacy and authority from its ability to provide for its members in time of need (especially if the state has not institutionalized reliable social rights of state citizenship) and to extract obligation to meet such needs. It is upon these forceful ties that politicians can mobilize for interests outside the ethnic community and make their own access to power (seem) vital to individuals within the identity group. For their part, individuals who may not hold sophisticated opinions on institutional issues give merit to the claims advanced by their ethnic leaders in the national state because they can interact with and extract from them via ethnic bonds. They view the advancement of such leaders as an enrichment of their community.
Ethnic versus National Citizenship
While the preceding discussion suggests a clear dichotomy in citizenship and in the form it takes, the reality is less clear-cut. As pointed out earlier, individuals hold multiple and shifting identities and change roles depending on the situation in which they find themselves. Furthermore, people experience a convergence of demands from both of their citizenships, rather than some people experiencing a republican form, others the liberal form. Thus, individuals experience liberal citizenship demands at the national level which they may or may not find consistent with (or advantageous to) civic-republican demands within ethnic communities. In this way, the illiberal, republican citizenship prescribed by ethnic groups undermines the liberal citizenship assumed in the national community in two ways. First, even as they act in the national arena, individual citizens "form and revise their aims and ambitions" within their experiential life in parochial communities (Kymlicka 1989, 135). Thus, parochial civic-republican obligations taint one's "individual" preferences in the nation-state arena, especially since communal obligations are more effectively enforced in the parochial community. Second, for those who capture the state, the state becomes an arena in which to fulfill obligation to the subnational community. This is particularly so when one ethnic group in competition with others emerges as hegemonic. The new state elite is essentially an ethnic elite; its sphere of authority extends to the territorial limits of the state, but its sphere of obligation is limited to the subnational group able to extract obligation (prebends) from its sons and daughters (Joseph 1987).
The different articulation of citizenship in the national and subnational communities has implications for democratic transitions in multiethnic states, especially those in which there is great size disparity among groups. Members of larger ethnic groups, exercising the republican duty to vote for their community, propel their group to electoral dominance and therefore benefit from liberal democracy with majoritarian institutions, in which the individual is the presumed principal political actor. Such a situation leads to a stratified citizenship; members of groups that capture the state enjoy a larger bundle of rights and undertake a greater responsibility for the national state than do members of excluded groups, not unlike the stratified citizenship among Jews and Arabs in Israel (Peled 1992). For the minority groups, obligations of ethnic citizenship operate just as efficiently in an election, but their smaller population results in perennial defeat or exclusion from crucial participation, despite protection of individual rights and privileges.
As I demonstrate in the next two sections, in both the transition to independence and the recent transition from single-party rule in Kenya, competing visions of political community and of citizenship framed the conflict over what institutions are appropriate to structure democratic politics in a multiethnic state. While the present analysis relies on a comparison of two historical junctures in a single country, the issue of divided loyalties and their political consequences is common to many African nations. For example, countries that experienced ethnic or regional conflict at independence and/or during recent democratic transitions include Burundi, Ghana, Nigeria, South Africa, and Zaire.(13) While the axis on which citizenship is contested varies (ethnicity, religion, region, or race), these conflicts fundamentally express disagreements about identity and political community and about power and legitimacy in the national state.
TRANSITION TO INDEPENDENCE IN KENYA, 1960-64
Kenya's independence in December 1963 was preceded by three constitutional conferences held in London in 1960, 1962, and 1963. The debates within these conferences and the events between and subsequent to them provide evidence for how dual and competing citizenship led ethnic groups to coalesce into rival parties propounding radically different institutional preferences. At the first constitutional conference, in January 1960, a broadly based coalition of African elites presented the case for independence from Britain and negotiated a new constitutional framework allowing greater African representation in colonial legislative institutions. They also received a commitment to eventual African self-government "based on parliamentary institutions on the Westminister model" (Her Majesty's Stationery Office [HMSO] 1960, 6). As soon as the African delegates returned home, the united front they had forged to fight racial domination collapsed, and two rival parties with distinct ethnic compositions emerged. Ethnicity replaced race as the primary political cleavage, and subsequent debates elaborated the tension between ethnic and national citizenship as each party developed its vision of what would constitute legitimate citizenship in independent Kenya and what political institutions would fulfill its vision.
How the rival political parties formed was critical to their ethnic composition and to their reliance on activating ethnic citizenship to mobilize votes. Party formation proceeded on two related tracks. One was at the behest of a single leader who provided intellectual and financial patronage and who drew a core of founders linked more by personal ties forged in the ethnic arena than by ideological commitment. This core group relied on family, clan, and ethnicity as the initial scaffolding for the party infrastructure, because among these persons trust and resources could be exchanged on the basis of prior or assumed reciprocity and could attract others in the "community" with minimal resistance.(14) The other track stemmed from the colonial practice of restricting African political activities to local areas, which were ethnically homogeneous. In the absence of deep ideological differences, the district-level organizations tended to coalesce into ethnic blocs, and these were co-opted into the rival parties-in-formation. Thus, the nascent national parties were little more than clubs of personally connected individuals, and they immediately spread at the grassroots level (Okumu 1975, 188-9).
The Kenya African National Union (KANU) was the dominant of the two parties. It drew the bulk of its leadership, membership, and support from the Kikuyu and Luo, the two largest ethnic groups in Kenya. Launched in March 1960 at Kiambu in the Kikuyu heartland just outside Nairobi, KANU was the reincarnation of the Kenya African Union, which in turn was a nationalist mutation of the proscribed Kikuyu Central Association. KANU also subsumed proscribed Luo-led organizations, such as the Kenya Federation of Labor and the Kenya Independence Movement. This fusion of multiethnic and labor organizations made KANU an intensely anticolonial and nationalist party. Nevertheless, because of its composition, the party was inescapably representative of Kikuyu and Luo ethnic interests and of the urban working and middle class; it also established a link between the Kikuyu and Luo urban populations, especially in Nairobi, and their rural brethren (Bennett and Rosberg 1961, 37-8). For these ethnic and class groups, the integrity of the postcolonial state and a majoritarian electoral system would present new opportunities for advancement.
The second party, the Kenya African Democratic Union (KADU), united a diverse set of local associations that represented minority ethnic groups joined by the fear of domination by an exclusive Kikuyu-Luo postindependence government. KADU's component groups were themselves composite unions of smaller subethnic or district-level political organizations. For example, the Kalenjin Political Alliance (KPA) was formed in April 1960, in Kapkatet, by combining a number of organizations in the Rift Valley region under the leadership of Daniel arap Moi (Bennett and Rosberg 1961, 40). In June 1960, the KPA merged with the Maasai United Front, the Coast African Political Union, Kenya African People's Party, and the Somalia National Association to
form KADU in response to the emergence of KANU three months before (Bennett and Rosberg 1961, 37; Odinga 1967, 194-5).
The prospect of Kikuyu-Luo dominance was real, since the two groups were larger, more politically conscious, and better organized than the KADU groups and presumably would sweep the polls. The expectation of a census-type vote was fulfilled in the February 1961 elections (Bennett and Rosberg 1961). Of the 33 open seats, KANU won 19 with 67.4% of the vote, while KADU won 11 seats with a paltry 16.4% of the vote. These proportions roughly approximated the population distribution of the ethnic groups backing each party. For example, the Kikuyu, Luo, Embu, Meru, Kamba, and Kisii supporting KANU made up about 60% of the population, underscoring what Bennett and Rosberg (1961, 43) call the phenomenon of the "one-party tribe." Similarly, this "pull of the tribe" (Jones 1967, 107) was evident in the urban areas, where the proportion of "KANU tribes" reflected the KANU vote.
Indeed, in the areas where either party's "tribes" commanded a majority, the opposition candidates were soundly defeated, if they bothered to run at all. For example, in Machakos the winning KANU candidate polled 21,000 votes, beating five other KANU candidates, who polled 19,000, 15,000, 7,000, 1,700, and 400 votes, respectively; the sole KADU candidate received 380 votes. Polls in the KADU stronghold in the Rift Valley mirrored the situation. In Narok and Kajiado districts (Maasailand), KADU candidates sailed through unopposed; in Kipsigis the KADU candidate pulled in 56,000 votes against fewer than 200 garnered by two KANU candidates (Bennett and Rosberg 1961, 204-17). The ethnic coalitions and rural-urban divide that emerged in 1961 elections were replayed in the independence elections of May 1963. A similar cleavage and vote pattern reappeared in the transition election of December 1992 (Barkan 1993, 94-8; Muigai 1995, 186-91), which underscores the durability of polarization in Kenya.
When the two Kenyan parties returned to London for the second constitutional conference in February 1962, KADU retracted its commitment to a "Westminister" form of government and instead demanded majimboism, or regionalism. That is, independent Kenya should be divided into autonomous regions, which KADU saw as the only institutional structure that would preserve the self-determination of minority groups, since the majoritarian Westminister model would likely keep them a permanent minority. KADU was adamant and was prepared to delay independence for "another ten years" and "sound a whistle. . . declaring civil war" against the independent state if majimboism was rejected by the dominant groups (Odinga 1967, 227-8). KANU, having prevailed in the recent elections, wished to retain the majoritarian model. KADU resisted during more than two months of negotiations and forced KANU to compromise, although the latter had a sufficient majority in the Lancaster delegation to override the majimbo advocates.
The result was a new constitution under which Kenya would hold elections for African self-government - a step away from full independence. According to this agreement a central government with a two-chamber national parliament would be created. It would consist of a lower House of Representatives with 117 elected members and 12 appointed members and a Senate of 41 members representing all administrative districts and Nairobi. In addition, power would be decentralized by dividing Kenya into seven regions, each with its own legislature and executive body, thus giving effect to KADU's demands for majimboism. The powers of the regions were substantial. Each had its own police force and civil service; regions could veto certain central government appointments, such as chief justice; and each region was entitled to a fixed proportion of the central government tax revenue. Moreover, the agreement required that any constitutional change affecting these powers must have a 75% majority (90% in the case of specially entrenched powers) in both the House and the Senate (Brown 1970, Singh 1965).
This majimbo constitution was a major triumph for ethnic minorities, who would now be able to preserve autonomy in "their" regions, since the more or less ethnically defined colonial administrative units (provinces) became the regions. The minority groups were assured representation and participation in the central government through the Senate, whose electoral districts were the even more ethnically homogeneous administrative districts. Moreover, the Regional Boundaries Commission was established to collate the views of all ethnic groups regarding the region to which they wished to belong. This resulted in some boundary changes and the creation of an eighth region to accommodate the secessionist Somali population. In another instance, the Saboti of Mount Elgon wished to be included in the Riff Valley province along with their Kalenjin cousins, but the commission denied this request.(15)
Following the Lancaster agreement of 1962, elections were held under the new constitution in May 1963. KANU won an overwhelming majority in these "independence elections," taking 66 seats against KADU's 31 in the lower house and 19 seats against KADU's 16 in the Senate. In the regional assemblies, KANU showed similar strength, gaining 158 seats against KADU's 51. The regional and ethnic spread remained similar to that of the 1961 elections, with each party predominating in its "tribal areas." When the two parties returned to London in September 1963 to finalize the independence constitution, KANU demanded amendments to the 1962 agreement to reduce regional powers, the special protections for minorities, and the constraints imposed on constitutional change. KADU, having suffered an electoral setback (and the defection of the Abaluhya and Kamba leaders), insisted on retaining the 1962 agreement as the framework for the final constitution. Again, KADU threatened the integrity of the new state if protections already attained were withdrawn. For example, it would force a partition of Kenya into two different states to avoid domination by the larger ethnic groups (Brown 1970, 26).
As a result of such intransigence, KANU accepted the majimbo arrangement as the independence constitution, but a number of changes were made in its favor. The regional police forces and civil services were combined into single, national organizations under the central government. KANU also won an important concession on constitutional change: Amendment proposals that failed to receive the required majority in the House and Senate would then require a two-thirds majority in a national referendum. KANU agreed to implement the majimbo constitution despite criticizing it as "one of the most complicated constitutions ever given to a newly independent state" (Kenyatta quoted in Brown 1970, 30) and one for which the "first requirement was a skilled corps of lawyers and clerks in the center and the regions to explain to legislators what they were required, permitted, or forbidden to do" (Odinga 1967, 234). The detailed division of powers and responsibilities was cumbersome for the nationalists in KANU but necessary for the KADU minorities.
Kenya thus became independent in December 1963 under a constitution that validated ethnic citizenship in autonomous regional governments and strove for a national, liberal citizenship within central government institutions. This constitution would prove short-lived, since the majority Kikuyu-Luo alliance in KANU was impatient with the unwieldy structures that curtailed majority power and was unwilling to invest the money to make the system work or to accommodate "parochial" minority demands that got in the way of nation-building (Odinga 1967). Within the first year of independence, KANU undermined the regional governments by withholding funds, passing legislation to circumvent regional powers, and forcing major changes to the constitution by threatening and preparing to hold a referendum if the Senate - in which KADU could block the proposals - did not accede to the changes. Outnumbered, outmaneuvered, and with no prospects for enforcing the compromise constitution or, given the reality of census-type voting, for overtaking KANU at the subsequent polls, KADU willingly dissolved and joined KANU to form a single-party state in 1964. KANU, then under a Kikuyu-Luo alliance (which soon became a Kikuyu hegemony), had installed its vision of liberal citizenship and established the nation rather than the ethnic group as the preeminent political community.
ETHNIC VERSUS NATIONAL CITIZENSHIP, 1960-64
The political discourse underlying this constitutional dispute reflects the two competing citizenships: one national, the other subnational. The dispute was inevitable, given the character of the colonial Kenyan polity as an "aggregation" of ethnic groups yet to form a historically validated national union or "association" (see Rousseau 1987, 147). While ethnic identity remained intact and indeed was heightened by dealings with "others," such as through the "polarizing" elections of 1961 (Horowitz 1991, 96), individuals were expected to transfer political allegiance, which normally coincided with small-scale groups, to the new national political community. Political elites are important intermediaries in such a process. Given their leadership position in ethnic networks and depending on their own rational calculations about access to power, they propagate acceptance or instigate rejection of such transformation. While the immediately relevant conflict arises among elites, at times over such fairly complex issues as constitutional structure, the positions articulated become legitimated across the ethnic community, in which individuals view the fortunes of their leaders as linked to their own.
For the larger ethnic groups, such as the Kikuyu and Luo (potential pluralities in national electoral politics), the transfer of political allegiance to the nationalist state was less threatening; under majority rule, they and their elites would be assured dominance or at least participation - full citizenship rights - in the larger political community. This transfer of political allegiance was a much more problematic proposition for the smaller ethnic groups, such as the Kalenjin and Maasai, since they could only remotely influence the dynamics at the center, and the republican virtues validated at the ethnic level would be denied them at the national level. If the postcolonial state were divided into smaller administrative and semiautonomous arenas of citizen participation that coincided with an ethnic preponderance (i.e., majimbo), then minority groups could be assured of an acceptable level of republican participation.
The transfer of political allegiance is not unique to formative nation-states; it is one that occurs often but is eased by assurances of a balance of obligation and rights. For example, the transfer of allegiance from the family to the clan, from the clan to the village, from the village to the tribe, and so on, is secured by privileges of belonging to the larger community. The problem for the minority groups at independence was that obligations to the larger political community were not adequately balanced by the benefits of inclusion. Moreover, the transfer of allegiance to an overarching tribal or national unit is not peculiar to the larger tribes, as the experience of the Kalenjin Political Alliance (KPA) suggests. Under the instigation of political elites seeking dominance in the state, such minor groups as the Nandi, Elgeyo, and Tugen transferred their allegiance to a larger community of identity and interest (the Kalenjin) and became numerically competitive in the state arena. It is important that when such minority groups joined KADU, decision making was consensual to preserve cohesion for the sake of numerical competitiveness: "Authority was dispersed and shared between the leaders and a sense of corporate leadership clearly prevailed. In seeking unity the party placed a high value on compromise, to attract and accommodate divergent interests and peoples" (Bennett and Rosberg 1961, 42).
The majimbo constitution reflected concern about two principles: protection from majoritarian tyranny and apportionment of political power to ensure minority participation (Ghai 1967, Ghai and McAuslan 1970, Singh 1965; see Lijphart 1977). The liberal conception of national citizenship satisfied the first objective, since it guaranteed individual protections through a Bill of Rights. By providing minority groups an arena for self-regulation, the regional government constitution fulfilled the second objective. Political participation in the regions was circumscribed by "regional" citizenship. For example, in elections to the regional assembly, "only persons who had a genuine connection to the Region" could vote for representatives, and no candidate could be elected to the assembly if s/he was not a registered voter in the region (Ghai 1967; Ghai and McAuslan 1970; HMSO 1963a, 66-7). Although subnational citizenship was defined by geography (region) rather than ethnic group, the two coincided due to the work of the Regional Boundaries Commission and the colonial policy of homogeneous tribal reserves. In effect, under this constitution Kenyans had their dual citizenship recognized.
KANU's opposition to the majimbo constitution repudiated the second principle, that is, allowing for smaller arenas of self-regulation short of the nationstate. For KANU the guarantees of the liberal constitution were sufficient, and the regional proposals were indicative of a tribalism that endangered the integrity of the new state. For example, Tom Mboya, one of the most articulate defenders of the nationalist vision, offered liberal citizenship to attract skeptical minorities. In July 1962, speaking to white farmers who were uneasy about the prospects of a black-ruled Kenya and who were as unsure of their attachment to the new Kenyan political community as were the minority African groups, Mboya stressed a vision of citizenship informed by liberal virtues. All individuals belonging to the national political community would be equal in the rights they derived and in the obligations demanded:
There will be a Bill of Rights for all Kenya citizens which we - the political leaders - have designed and will defend. All who accept this will no longer be "members of the immigrant community," but full members of a new community which will demand all our dedication, all our attachment to morality, all our strength. . . . Your interests will be defensible in law, and of great concern to the Government, as the interest of anybody else (Mboya 1970, 45-6).
Mboya and other nationalists saw the objective as creating a nation out of the mosaic of ethnic groups: The nation was the point at which preexisting, subnational "loyalties will coalesce" into one (Mboya 1970, 47). This vision was indeed the glue that held the liberation struggle together, especially between the Kikuyu and Luo. Another KANU nationalist, Ochieng Oneko, articulated this dissolution of preexisting prejudices in the drive to nationhood.(16) At the height of the anticolonial mobilization in 1952, Oneko exhorted a Kikuyu crowd in Nyeri: "I want you to love all Africans and I want tribes to get together. I said in Kiatabu that I would marry a Kikuyu girl to show good relations amongst tribes" (Collins 1990, 164). More telling of KANU's drive for the suspension of ethnic loyalties for the sake of a single national identity was the colonial district police officer's record of the meeting in which, commenting on Oneko's specific exhortation, he reported: "Three years ago this would have been taken by the Kikuyu as an insult, for an uncircumcised Luo was the most despised thing imaginable to the Kikuyu. Today it is applauded" (Collins 1990, 164).
Proponents of the nationalist vision saw the conflict between ethnic and national citizenship as resulting from the instrumental manipulation of ethnic allegiances by politicians for individual goals. Mboya argued that "some leaders have revived these old hostilities for their own personal reasons. When a leader feels himself weak on the national platform, he begins to calculate that the only support he may have will come from his own tribe; so he starts to create an antagonism of this sort, to entrench himself as at least a leader of his tribe" (Mboya 1963, 66). Mboya's dismissal of tribalism as the bane of African politics is characteristic of the disdain with which the African middle and educated classes dismiss the force and morality of ethnic citizenship (Davidson 1992). But such "modernized" Africans are contradicted by their own lives, which are embedded in ethnic communities, and, alas, by their deaths, which are even more intimately enmeshed in the reproduction of ethnic communities (Cohen and Odhiambo 1992; see Comaroff and Comaroff 1993). The failure of Mboya and other "nationalists" wedded to the liberal conception of citizenship to consider that such ethnic commitments inescapably shape the supposed sovereign and autonomous individual - as Kymlicka (1989, 135) reminds us - is a fundamental and enduring failing among democrats confronted by a multiethnic state.(17)
The debate over political institutions and power sharing after independence was especially animated with regard to land, which in an agrarian society is the primary social, economic, and political asset on and over which citizenship is defined and contested.(18) In Kenya, citizenship was historically defined in terms of rights of access to land, which assured livelihood: "Individuals had a community share in or the right to benefit from land. . . . Members' entitlement to land was based on the fact that the individual belonged to a particular group" (Kibwana 1990, 233). Who belonged and who did not belong to the community determined who benefited and who did not. Because of disagreement as to which was the preeminent community (national or ethnic), the differences between KANU and KADU were deepest and most irreconcilable with respect to how the land previously held by white settlers and the colonial government was to be redistributed. KANU defined community as the nation to which all belonged, and all had rights to land anywhere, subject to national laws and market exchange. The KADU groups, knowing their power would be diminished in the national political community, offered the ethnic group as the relevant community of citizenship and sought to limit land rights to those with membership in native groups in each region. Indeed, the KPA was formed expressly to "make clear its prior claim to control over land in western Kenya. . . [since] much of the land in the western highlands of Kenya belonged to the Kalenjin before the Europeans arrived" (Bennett and Rosberg 1961, 40).
At independence, the disposition of all land was open to negotiation, and by far the most contentious was Scheduled Land, which included the developed white farms in the most fertile regions of the country.(19) The situation was complicated by historical factors and the scarcity of land among Africans as a result of colonial deprivation. Land was a scarce resource for all groups but particularly for the larger ethnic populations, whose number far outstripped their traditional lands and who would benefit from a freehold system. For the Kikuyu in Central Province, land scarcity had been worsened by colonialist appropriations. Many minority groups, especially the traditionally pastoralist Kalenjin and Maasai communities, whose Rift Valley pasturage had been alienated for white settler farming, now faced the loss of these lands due to the introduction of a land market that guaranteed individual tenure, which was more suited to the farming communities, especially Kikuyu and Luo. The latter were better placed economically to purchase land and would dominate the central state that would administer the land redistribution program (Njonjo 1977, Widner 1992b).
The two parties' positions on land reflected their ethnic composition and were calculated to maximize returns for their ethnic groups. KADU demanded that all land matters be vested in the respective regional governments; this would effectively have given each community primary claim to its ancestral land. KANU demanded that only communal land under the Native Reserves be entrusted to the regions, leaving the central government the say in distributing the white farms on Scheduled Land and the vast unalienated Crown holdings. In the agreement reached at the second constitutional conference in 1962, Scheduled Land was entrusted to the Central Land Board with the mandate to purchase the white farms and redistribute them to Africans. Control over Crown Land was given to the regions, while the former Native Reserves were entrusted to county councils (i.e., local authorities). This arrangement assured minority groups that their land would not be taken by the majority tribes, but it gave them little in the way of reclaiming or of limiting nonindigenous settlement on or commercialization of their ancestral land, especially in the Rift Valley. After the dissolution of KADU in 1964, control over land reverted to the central state. The subsequent dominance of KANU enabled it to apportion land, disproportionately favoring its core groups, especially the Kikuyu. This conflict over land persisted in postindependence politics and recurred alongside ethnic violence that erupted during the transition to multiparty democracy in the 1990s.
Until the next period of fundamental political transformation beginning in 1990, the discourse on ethnic and national citizenship remained muted, as KANU suppressed any alternatives to the unitary state. The unitary state itself became increasingly autocratic, thereby undermining the liberal polity promised at independence. It also was subsumed under an ethnic hegemony, first by the Kikuyu and later by the Kalenjin groups when Daniel arap Moi succeeded Kenyatta, which undermined the promise of coalescing tribes into one nation. Beginning in 1990, however, the transition from a single-party dictatorship to multiparty pluralism provided the first opportunity since independence to reconsider Kenya's political institutions. As elaborated below, the resurgent discourse on ethnic and national citizenship became and continues to be crucial to the fate of Kenya's transition to democracy.
TRANSITION TO DEMOCRACY, 1990-95
When Kenya's first president, Jomo Kenyatta, died in 1978, power passed peacefully to his long-serving vice-president Daniel arap Moi.(20) A member of the minority Kalenjin community, Moi had been appointed vicepresident in 1967 to complete the absorption of KADU into KANU, to balance symbolically Kikuyu dominance in KANU after most Luo defected with former vice-president Oginga Odinga, and to placate the Kalenjin community as Kikuyu elites received fertile land in the Rift Valley. President Moi immediately embarked on efforts to consolidate his precarious position in a state and economy dominated by the Kikuyu. Part of this consolidation entailed reinvigorating KANU as a vehicle for bringing loyal politicians on board (Widner 1992b). By 1990, the state had been transformed into a stronghold of elites from the Kalenjin and other minority ethnic groups (Throup 1987, Widner 1992b), with these groups carrying an inordinate responsibility for and drawing immense benefits from the state (Africa Confidential 1990a, 1990b, 1996).(21)
Such practices undermined opportunities for advancement that would have accrued to the two most populous and most educated groups had a more liberal, competitive system been used instead of the affirmative action that the Moi regime instituted to provide access for minority groups to, especially, higher education, administrative recruitment, and the allocation of public funds (Barkan and Chege 1989). By 1990, the middle class - "with Kikuyus the most numerous among them, alarmed over the Moi regime's deepening Kalenjin ethnic bias - evinced a growing desire for the rule of law in the face of stiffening repression" (Holmquist and Ford 1995, 177). Increasingly, the KANU government came to view the "bid for democracy [as] a movement under the control of the rival Kikuyu ethnic group" (Widner 1992a, 216).
Although agitation for democratic reform by the Forum for the Restoration of Democracy (FORD) was a multiethnic effort, its leadership was dominated by the Kikuyu and the Luo. The FORD of 1990 echoed the KANU coalition of the 1960s, not only in its leadership but also in its middle class and urban base and its successful mobilization of rural support from its "ethnic areas." FORD opposed a KANU that resembled the 1960 KADU coalition of minority groups led by largely middle- and upper-class leaders whose base remained rural and who were unable to attract an urban following. FORD and its allied groups in civil society articulated a unitary nationalist vision, with a liberal conception of citizenship, and offered majority rule institutions as appropriate to structure public affairs in post-single-party Kenya. Not surprisingly, the incumbent KANU government opposed political change; less visibly but more important, it opposed the return to a liberal, majoritarian form of democracy that favored more populous ethnic groups. Furthermore, prominent politicians, including Moi, asserted and invigorated the integrity of Kalenjin ethnicity, claiming that there was no such identity as Nandi, Tugen, or Kipsigis (the subgroups) - only Kalenjin (Africa Confidential 1996, 3). Moi also forged a wider collaboration with other minority groups through leaders willing to invoke minority ethnic identity or reformulate new identities to balance the "Kikuyu-Luo" multiparty advocates. Confronted with an unstoppable reform movement, regime politicians from the incumbent minority groups, which would lose out in presumed census-type elections under majoritarian rules, asserted the majimbo option in 1991, as they had done in 1961.
The Second Majimbo Debate
The institutional preferences of the incumbent minority group coalition were initially articulated at five mass rallies held during six weeks in 1991. The first was on September 8, 1991, at Kapsabet, and the second was two weeks later at Kapkatet - where the KPA was launched 31 years earlier - both in the Kalenjin heartland in Rift Valley Province. The next three rallies were in the strongholds of three major politicians aligned with Moi: Narok (Maasailand), Machakos (Kambaland), and Mombasa (cosmopolitan, but supposedly a Mijikenda "native" area). These meetings were conducted under the tutelage of leaders in the ethnic coalition Moi was building to fight eclipse under a multiparty system. The initial rally at Kapsabet was attended by senior KANU leaders, cabinet ministers, members of Parliament, and local councilors, all from the Rift Valley region. The presiding contingents at subsequent rallies were similar and included, in addition, the major politicians of the relevant ethnic group in the host region - a recognition that the notions of "one-party tribes" and "tribal leaders" were largely valid three decades after independence.
In these rallies and in other public statements, KANU politicians called for a majimbo system as an alternative to the multiparty proposal advocated by proponents of political pluralism. Substantively, the call for majimbo was a modification to that proposal, since it accepted political change as inevitable but sought to proclaim separate regions, where each party, presumably a "one-tribe party," would hold sway. The politicians announced they would introduce a majimbo constitution if the multiparty advocates persisted in their crusade for what the incumbent regime saw simply as majoritarian rule. A number of KANU's ideas underlined the centrality of ethnic citizenship, especially its rights and obligations and the limits placed on the rights of "others" in an ethnic homeland, notwithstanding liberal citizenship rights endowed by the state. Some of these resolutions corresponded with invocations of the republican virtue demanded by ethnic groups: They extract obligations for ethnic citizens to act to ensure the survival of their tribe (see Oldfield 1990, 181). For example, Kalenjin politicians led the mass meetings in adopting "resolutions" committing the Kalenjin to ban a host of multiparty advocates from venturing into Rift Valley Province and asserting that "Kalenjins were not cowards and that they were ready to counter attempts to relegate them from leadership . . . using any weapons at their disposal" (Republic of Kenya 1992, 10).
Moreover, the majimbo rallies introduced a new lexicon of difference and intimidation, especially with reference to rights of access to and settlement on land and rights of political participation in selected regions. For instance, madoadoa (spots) and kwekwe (blemishes) referred to "contamination" of the ancestral lands of the Kalenjin and Maasai in the Rift Valley region by migrant ethnic groups (especially Kikuyu). These new settlers were also termed chui (leopards), a more insidious reference to the interlopers as land-grabbers. These references suggested the need to cleanse the regions of nonindigenous groups (Republic of Kenya 1992, 48). With regard to political participation, KANU politicians declared areas in the Rift Valley where the Kalenjin and Maasai were the majority or could claim nativity as exclusive "KANU Zones." In these areas, advocates of multiparty democracy and opposition politicians were banned from campaigning, and nonnative residents were cautioned against voting for opposition politicians. The implication was that "settlers" without ethnic citizenship in the communities of the Rift Valley could not exercise their national citizenship rights in these areas. National citizenship was subordinate to ethnic citizenship, and ethnic citizens were obligated to act to preserve community. Indeed, according to the inchoate majimbo framework offered at the rallies, "outsiders in Rift Valley [province] would be required to go back to their 'motherland'" (Republic of Kenya 1992, 9).(22)
Ethnic versus National Citizenship
Unlike the independence period, when the political transition was concluded through constitutional conferences, and unlike contemporary transitions in most of Francophone Africa, in which political change has been managed through sovereign national conferences (Nzouankeu 1993, Robinson 1994), the contemporary discourse on citizenship and institutional preferences in Kenya has been restricted to forums such as the media, public rallies, and, more violently, ethnic clashes. This discourse is seemingly inchoate and haphazard; because it is conducted in different arenas, it does not suggest an obvious dialogue. A clear and coherent body of thought can be construed, however, from the sustained and consistent contributions of a number of political actors.
Within KANU, President Moi has consistently articulated his distaste for multiparty politics by asserting that it will introduce ethnic conflict. This is in contrast with his statements in the 1960s and his actions as vice-president of KADU, when he vigorously defended the rights of free association and ethnic parties. The most articulate and consistent defense of minority interests, however, has come from the leading Maasai politician and former Minister for Local Government, William ole Ntimama.(23) Yet, three organizations, which I will call the liberal democracy coalition, have led the discussion on constitutional reform more concretely than have the divided and disintegrating opposition parties. The leaders are predominantly Luo and Kikuyu, urban-based middle-class professionals and intellectuals. The Citizens Coalition for Constitutional Change (henceforth Citizens Coalition) is made up of lawyers, human rights advocates, academics, and other professionals and has produced and publicized the most well-developed model constitution offered for discussion. Jointly and separately, the Protestant clergy in the National Council of Churches of Kenya (NCCK) and the Kenya Catholic Bishops (KCB) also have participated in constitutional discussions and have led in criticizing government complacency toward or complicity in ethnic violence.
The debate between KANU and the liberal democracy coalition is not simply one between an antidemocratic incumbent regime and progressive democrats. It is more complex and involves fundamental differences regarding (1) the preeminent political community in a multiethnic state (national versus ethnic); (2) the political institutions appropriate to govern such a state and, therefore, the kind of constitutional reforms required to arrive at such institutions; and (3) the process of formulating constitutional reforms. These essential disagreements reflect positions developed from two distinct perspectives on citizenship, one nationalist and liberal, the other narrow and civic-republican. It is important that each side advances a position which holds the best prospects for its own access to power on the basis of an activated or assumed ethnic citizenship and demonstrated ethnic voting.
The liberal democracy advocates propose fundamental constitutional changes within, and indeed seek to strengthen, the unitary state as the preeminent political community and the arena in which citizenship is defined and upheld. The NCCK (1995, 1) asserts a commitment to "the establishment of a united nation with strong democratic institutions" underlain by "political co-existence, freedom and tolerance, the rule of law and respect for fundamental human rights." The three organizations avoid constructing an ethnic conception of political community, rights, or political relations and instead mask their recognition of Kenya's pluralism in euphemisms expressing the need "to find out what the deep-seated fears of each group are. . . what our common ground is. . . what we disagree about and why we disagree about it" (Mwangaza Trust 1995, 2; see also Citizens Coalition 1995a, b). When specific groups are outlined as having interests that ought to be given special voice, only associational groups (e.g., NGOs, political parties, professional groups) are mentioned, while ethnic or other ascriptive groups are excluded (NCCK 1995, 2). Moreover, the noticeably few references to ethnicity or particular ethnic groups are in the (negative) context of ethnic conflict fomented by the KANU regime or exhortations to transcend ethnicity in building the Kenyan nation.
Along with the integrity of the national political community, the liberal democracy coalition asserts a liberal notion of citizenship in this community. The present constitution (amended numerous times in nearly three decades of single-party rule) and the government it sustains are lacking "commitment to the principles of sovereignty of Kenyan citizens, democracy, human rights, participation of people, accountability, [and] rule of law" (Citizens Coalition 1995d, 7). These organizations repeatedly articulate the goal of limited government countered by an unfettered civil society and by institutional checks and balances. For example, the Citizens Coalition asserts that a new constitution must guarantee the "duties, obligations, responsibilities and privileges that citizens enjoy within the framework of individual freedoms . . . [and] facilitate the expression of the will of All [sic] by merely aiming at facilitating the exercise of sovereignty by all governed" (Citizens Coalition 1995d, 2-5, emphasis added; also KCB 1994, 1995).
In making specific proposals regarding electoral rules and institutions of apportioning power, the liberal democracy advocates prefer majoritarian means. The NCCK (1995, 5) calls for "a referendum mechanism [to be] entrenched in [the] constitution" and for electoral mechanisms that enhance "a united nation with strong local government" rather than those that give free rein to ethnic regions; and nominated members of parliament, traditionally appointed to achieve ethnic balance (Berg-Schlosser 1985), should be appointed to represent nonascriptive interest groups in civil society (NCCK 1995, 5). The KCB (1994, 1995) offers similar majoritarian institutions as the proper mechanisms to govern a democratic Kenya, although with caveats to protect individuals from state repression. In contrast, leading KANU politicians, such as those presiding over the majimbo rallies of 1991, view the unitary state within a multiparty, majoritarian democracy as problematic, since it would shut out minority ethnic groups even if individual freedoms were guaranteed.
The NCCK's view regarding the "twenty-five percent rule," a consociational mechanism introduced by the KANU regime just before the 1992 election, is indicative of the rejection by elites from larger ethnic groups of any modifications to majoritarian democracy to benefit minority groups (NCCK 1995, 8-11). This rule required that a presidential candidate must garner 25% of the votes in at least five of the eight provinces, in addition to having at least a plurality of the nationwide vote. Barring that, a run-off would be held between the two leading candidates, and the one with the majority vote would be declared president. The NCCK and its allies argue that the provinces (the autonomous regions under the majimbo constitution) are administrative not electoral units and prefer the president be elected by a simple majority - without requirements of provincial or regional (and therefore ethnic) representation. Such a mechanism would favor the larger groups, whose numbers would propel them to dominance.
In contrast to the vision of the political community and the specific institutions proffered by the liberal democracy advocates, KANU's Ntimama suggests that "the only constitution reform needed is a small one: it is the introduction of majimboism" (Citizens Coalition 1995e, 1-2). But this "small" change would have enormous implications. It would require Kenyans to exchange their current "national" citizenship for a jimbo (region) citizenship and would require "that all those who find themselves in the [regions] other than those in which their ancestors were living in 1895 when Kenya was born to return to the [regions] of their ancestors and abandon property without compensation" (Citizens Coalition 1995e, 2). Ntimama maintains that a federalist system like the one prescribed by the 1963 constitution is the best for governing a multiethnic state such as Kenya. The Citizens Coalition counters that the system proposed by KANU in the 1990s approximates those proposed for Nigeria in the 1950s by Chief Obafemi Awolowo and for apartheid South Africa's homelands. Both, they argue, were based on the goal of a "federation of ethnic nations," which implies dismembering the present Kenyan nation (Citizens Coalition 1995e, 8).
Since proponents of democracy in Kenya tend to view majimboism and multiparty democracy as two distinct and incompatible options, the discourse between the incumbent regime and the opposition is viewed as being at cross-purposes. Yet, this discourse is precisely about which community will be the preeminent arena for democratic politics and which citizenship (ethnic or national) will prevail. Ntimama's views on the process for implementing a new constitution reflect the preference of the incumbent minority groups to control political change so as not to be outmaneuvered and outvoted by the larger ethnic groups, which would institute majoritarian democracy. This fear underlies the final conflict over the process of constitutional reform.
In his 1995 New Year's message to the nation, President Moi announced he would invite foreign experts to collate Kenyan views on constitutional change and recommend reforms that parliament would then enact. For KANU, the process proposed by the president was the safest for the minority groups overrepresented in Parliament but outnumbered in the general population. Furthermore, according to majimbo proponents, after initiating the federalist constitution, the government would retain the prerogative to define the borders of the regions and administrative units and to delineate the rights and obligations that federal and jimbo citizenship would entail (Citizens Coalition 1995e, 3-5). This process would in fact be contrary to how the 1962 Regional Boundaries Commission determined both the territorial and ethnic composition of the regions under the 1963 independence constitution. In contrast, the liberal democracy coalition prefers a three-step process involving nationwide and grassroots participation. Specifically, it would rely on specialized caucuses for organizations in civil society, public forums around the country to explain proposals and elicit more views from the grassroots, and ultimately a national convention that would adopt a final constitution and subject it to a national referendum. The convention would then form an interim government to implement the new constitution (KCB 1994, 1995; Mutunga 1994). The mechanisms proposed - the convention, civil society participation, and referenda - magnify unfettered majority preferences and would advantage larger groups, the well-organized, and the urbanized middle class - in which the Kikuyu and Luo predominate.
Once again, land is a critical substantive issue on which the discourse about ethnicity and citizenship in Kenya converges. It was with regard to land that ethnic citizenship most clearly challenged national citizenship when minority groups enacted the initial majimbo provisions advocated by KADU at independence and abrogated by the Kikuyu-Luo hegemony in KANU in 1964. In the proposed model constitution, the Citizens Coalition asserts the right of any Kenyan to own land anywhere (Citizens Coalition 1995c, 11), a view repeatedly echoed by the NCCK (1993a, b) and the KCB (1994, 1995). The majimbo apologists deny a basis for the universality of this right, however, and limit land ownership to those who establish ethnic citizenship through clear and indisputable ancestral roots in a region. For example, in 1992 Minister Ntimama dismissed as "mere pieces of paper" the deeds held by nonindigenous residents of the Rift Valley for their land. Furthermore, there were moves to discourage nonindigenous communities from buying or leasing land in the Rift Valley. "Already the government has secretly instructed the local administration to ensure that nobody buys or sells land until and unless the provincial headquarters of Nyanza and Rift Valley are adequately informed . . . .[T]he government has issued a directive that there should be no leasing of land in Narok district unless the two parties are registered at the two provincial headquarters" (NCCK 1993a, 5; 1993b, 2, 6). KANU's views found forceful expression in the Rift Valley, where non-Kalenjin and non-Maasai farmers were violently ejected, while those near provincial borders were asked to "exchange" their land with Kalenjins farming in neighboring provinces (Republic of Kenya 1992).
Ethnic clashes that began in October 1991 in the Rift Valley and spread to the Coast, Nyanza, and western provinces were an elaboration of the discourse between national and ethnic citizenship in the area of land rights. The clashes began at Meteitei Farm, a cooperative not unlike thousands of others in the Rift Valley that resulted from the redistribution of white farms to Africans after independence. Located in Nandi District, Meteitel was owned jointly by 310 Kalenjin and 280 non-Kalenjin farmers. On October 29, a dispute arose between these two groups. With the apparent collusion of local administrators and politicians, the Kalenjin members of the cooperative claimed sole ownership of the land and expelled the non-Kalenjins (Kikuyu, Abaluhya, Kisii, and Luo). Those who resisted had their houses and property destroyed. This was barely two months after the first majimbo rally in nearby Kapsabet. Between October 1991 and September 1992, when a parliamentary committee investigating the violence completed its report, more than 700 lives had been lost; 54,000 people were displaced, 9,400 houses destroyed, and more than 3,000 head of livestock stolen. The overall damage was estimated at 200 million Kenya shillings (about U.S. $5 million) (Republic of Kenya 1992).
This communal violence peaked during the multiparty election in 1992, abated thereafter, and erupted sporadically in 1995. These land clashes were viewed as a strategy by the Kalenjin-dominated government to force the majimbo alternative by creating "a situation on the ground for a possible political bargain in the system about the government in future Kenya" (NCCK 1992, 1) and by "intimidat[ing] and terroriz[ing] the ethnic groups that seem to be in support of multiparty democracy" (NCCK 1992, 3). For example, in one incident reported by the press (and not without precedent) a clash victim, "Mrs. Odondi. . . said police and administration officials ridiculed them with two finger salutes [an opposition symbol] and said 'let FORD help you'" (NCCK 1992, 7). The ethnic clashes, therefore, were an extension of the discourse on citizenship and democracy in Kenya, with land once again an arena as well as the object of contest.
Ethnicity is a fundamental force in Kenyan politics, a fault line along which elites mobilized and competed for power within incipient democratic institutions at independence, in the authoritarian interim, and in the recent return to party pluralism. The significance of ethnic cleavage in the recent transition and in subsequent electoral politics was clear in the 1990s reenactment of the independence-eve conflict over national identity and representational institutions. Moreover, political parties today are as narrowly based on ethnic coalitions and organized under putative ethnic leaders as were those at independence. Yet, most analysts have laid undue emphasis on the intransigence of the incumbent regime in explaining the protracted nature of the Kenyan transition. Increasingly, they view the conflict as one between a "democratic" opposition and an "antidemocratic" incumbent. Instead, this paper gives ethnic cleavage prominence, not merely as a tool of frustrated politicians seeking to stay in power but as a foundation upon which deliberate calculations about representation and access to power in a democratic context are made, leading to fundamental disagreements about institutions.
I argue that individuals in a postcolonial state like Kenya operate with dual citizenship, one in an ethnic community and the other in the nation-state. This dual citizenship differs in form, especially in the balance of rights and obligations that each type grants the individual. Within the ethnic community, on the one hand, citizenship takes a civic-republican form that subordinates individual rights and demands certain actions in the public arena to preserve and advance the community. The postcolonial state, on the other hand, grants all its members a liberal citizenship that emphasizes individual rights and does not or is unable to extract obligations. I argue that the transition to democracy in Kenya is protracted because the vision of democracy assumed, propagated, and installed is the liberal majoritarian variety, and its presumption of autonomous individual actors is at odds with the reality of individuals fulfilling republican obligations to their subnational community. This condition affects both the "democratic" opposition and the "antidemocratic" incumbents, as is evident from the reliance of ethnic and subethnic or regional networks on the formation of political parties and mobilization of votes leading to patterns that approximate an ethnic census.
The comparative analysis of the two junctures of political transition in Kenya illustrates the consequences of this duality of citizenship on electoral calculations and, in particular, on institutional preferences. Individuals formulate opinions and actions within an "imagined community" (Anderson 1991) whose experiential component has pervasive sanction in their lives and extracts obligations to advance its corporate interest. Politicians, in particular, relying on the electoral behavior of ethnically mobilized voters, perceive opportunities and constraints in different institutions and, thus, propound institutional preferences that advantage their access to power. As demonstrated here, the current transition in Kenya has stalled because the incumbent coalition of minority ethnic groups proposes federalist structures to limit the danger of domination by larger groups, who advocate liberal majoritarian democracy in a unitary state. The assumption of census voting is practically a reality, since this has been the character of the three democratic elections in Kenya's independence history.
The explication of ethnicity in terms of citizenship is not meant to lend ethnic politics an ideological or philosophical veneer - it remains a calculated, maximizing strategy. It does, however, enable us to move beyond the simplistic view of the Kenyan transition as retarded by conflict between antidemocratic incumbents and a progressive opposition. Such an explanation glosses over the fundamental debate about the appropriate form of democratic institutions for a multiethnic Kenya. By exposing the moral and temporal foundations that lend ethnicity saliency (especially the relationship among identity, legitimacy, and authority), citizenship theory provides a more elegant and parsimonious explanation for why and how ethnicity continues to plague democratic politics in Kenya. Conceiving ethnicity in terms of citizenship clarifies why the institutional preferences of the two major ethnic coalitions were consistent in both transition periods, even when the monopoly over state power was reversed. By recognizing, indeed legitimizing, subnational communities as arenas of claiming citizenship and as relevant aggregations of interest in democratic politics, the citizenship perspective allows us to incorporate issues of rights and obligations into the design of democratic institutions, such as alternative electoral rules and constitutions.
The dual and competing citizenships cannot be fully reconciled, but the tension inherent in their coexistence can be mitigated in two ways. First, electoral mechanisms can be employed that both respond to the reality of ethnic voting founded upon ethnic citizenship and offer incentives to politicians to seek votes outside narrow communities of accountability. Typically, federalism, limited electoral engineering, and consociationalism are the options proposed to accommodate ethnicity in democratic politics. As usually conceived, these options are of limited utility because they focus more on cushioning ethnically riven electoral outcomes and less on addressing issues of rights and obligations that propel individual action. Federalism's main advantage is to secure a local arena for political practice (which may coincide with mobilized subnational groups) and to assure participation at the center for all groups. In practice, the tendency has been for regional entities to emerge as vehicles for pursuing dominance at the center and for denying liberal rights to nonnatives of the local region (Diamond 1988b). Limited electoral engineering, as proposed by Horowitz (1991, 1994) and as attempted in Kenya (i.e., the 25% rule), seeks to provide electoral incentives to candidates to expand their community of accountability beyond a narrow ethnic or regional base. In practice, such designs are unsustainable since temporary electoral coalitions are unlikely to supersede the ethnic coalitions in which individuals' lives are embedded beyond electoral cycles. Consociational mechanisms combine elements of the previous two (Lijphart 1977). They allow for smaller arenas of political practice in which to enact ethnic citizenship and for representation at the center, and they provide institutional mechanisms at the national level to moderate elites exploiting or acting on ethnic obligations or citizens reacting to subnational mobilization.
By themselves, however, electoral mechanisms are insufficient to reconcile the destabilizing demands of dual citizenship. For instance, of the three options, consociationalism is most appealing, but it also falls short in that its overriding concern and justification is moderating or safely channeling electoral behavior and outcomes as opposed to resolving underlying citizenship dynamics of rights and obligations that propel such behavior. (This explains the shortcomings in the options noted above.) Therefore, the second way to mitigate tensions is to resolve outstanding conflicts between demands made by subnational communities for group rights and those made by national states for individual rights, even within groups. Thus, a strong constitutional articulation of group and individual rights can minimize some of the strains on democracy emanating from the dual and competing citizenships. In particular, it is necessary to specify the relationship and relative status between the dual citizenships. Such an articulation is obviously best negotiated in constitutional debates and constitution making, and ultimately it is underscored by political practice through time.(24)
Finally, this analysis contributes more generally to citizenship theory, which at present takes liberal and republican citizenship as coterminous and as compatible with democracy in a single political community - the modern nation-state. While replicating Peled's (1992) claim of the possibility of liberal and republican citizenship coexisting in a multiethnic state, my analysis goes farther by suggesting that these forms of citizenship are not necessarily coterminous. Citizens in a single state can be formed in two substantial political communities, one bestowing liberal citizenship, the other extracting republican citizenship. Moreover, these two distinct forms are not necessarily compatible; they can be contradictory and can undermine liberal democracy. In the case of Kenya, the national political community (nation-state) lacks the capacity to extract obligation but grants rights freely through a liberal citizenship. The ethnic community grants rights balanced by obligations enforced by social practice. Because the latter's authority and legitimacy are more pervasive in an individual's life, it predominates in dealings with others in the national arena (especially in questions of delegating authority) and often leads to conflict between different groups. Recognizing separate arenas of citizen formation and practice as well as the possibility of conflict between forms of citizenship forged in distinct arenas is a useful increment to current citizenship theory. It advances the possibility of understanding the persistent conflict between subnational and national loyalties, especially where the state has yet to form a capacity to extract obligation as strong as that of its subnational communities, whether ethnic, religious, or regional.
Earlier versions of this paper were presented at the annual meetings of the African Studies Association in Orlando, Florida, in November 1995, and the American Political Science Association in San Francisco, California, in September 1996. I am especially indebted to C. R. D. Halisi for inspiring my interest in citizenship studies and for his innumerable critiques. I am also grateful to Scott Gerber, Bruce Heilman, Paul Kaiser, John Lucas, Paul Mbatia, Richard Joseph, Michael Schatzberg, Alemante Selassie, Roger Smith, and Brian Winchester for comments on earlier drafts, and to Meg Mahoney for research assistance.
1 African countries that have debated or adopted institutional alternatives to accommodate subnationalism include Burundi, where power sharing has been attempted with mixed results (Lemarchand 1995); Tanzania, where the union between mainland Tanganyika and Zanzibar has come under intense discussion (Kaiser 1996); Ethiopia, where the overthrow of the Marxist regime led to regionalism and independence for Eritrea; Uganda, where the current no-party grassroots democracy is an attempt to avoid ethnic and denominational divisions inherent in organized parties; Nigeria, with its perennial creation of new federal states to accommodate claims to self-determination and its recent attempt to legislate only two political parties; and post-apartheid South Africa, with its transitional government of national unity and negotiated constitution (Sisk 1995). See also Selassie 1992, more generally.
2 I am grateful to an anonymous reviewer for suggesting this precise formulation.
3 A census-type or polarizing election (Horowitz 1985, 83-7, 326; 1991, 97-8) is one in which voting is largely along ascriptive lines, leading to an outcome that is akin to a census of the different groups in the body politic and that gives the largest group a lock on power.
4 For now, it will suffice to note that the liberal vision of citizenship holds that rights are inherent in individuals, are prior to community, and are guaranteed with minimal obligations to the community, while the civic-republican view sees rights not as inherent but as acquired through civic practice that upholds obligations to the community. The libertarian vision, which will not be discussed in this article, considers citizenship a contractual exchange into which rational individuals enter in order to gain access to public goods (Miller 1995, 433).
5 For other exceptions at the frontier of redefining arenas of citizenship practice, see Miller 1995 and, in particular, Kymlicka 1995.
6 For an example of the kind of anthropological work that imbued groups with a coherence that may not have existed in reality or was enforced by colonial administrative practices, see Middleton and Kershaw's (1965) ethnographic work on the Kikuyu and the Kamba.
7 Ethnic groups similarly created or consolidated include the Kikuyu and others in central Kenya (Ambler 1988), the Giriama and other Mijikenda groups (Brantley 1981), and the Kalenjin (Ogot 1981, 84). For instance, the Kikuyu are sewn together from peoples who continue to be divided by region (Nyeri, Muranga, and Kiambu) and, more recently, by class, which is especially salient among emigrants to settlements in the Rift Valley province. Indeed, many Kenyan subgroups continue to be divided by clan, region, and dialect.
8 I use "tribe" advisedly and interchangeably with "ethnic group" because it remains a legitimate reference in Kenyan political discourse.
9 Another example is more directly related to national electoral politics. In the 1992 presidential campaign, politicians from communities that practice male circumcision as initiation into adulthood used this ethnically valid measure of full citizenship to denounce opponents from communities that do not circumcise men, suggesting that they were unfit presidential contenders since they were not yet "men." Given that a majority of Kenyan ethnic groups practice male circumcision, this was a potent charge against a minority. This powerful exclusionary tactic was invoked in the national arena but drew its potency from ethnic customs and assumptions.
10 I am indebted to C. R. D. Halisi (1997) in crystallizing the following formulation, which draws on his analysis of citizenship in South Africa.
11 This competition repeatedly has been noted across Africa (e.g., Rothchild and Olorunsola 1983). An important exception is Miles and Rochefort 1991, which shows that individuals from an ethnic group split between two countries exhibit closer affinity to citizens of their state than to fellow ethnics in the other country. This conclusion, based on a study of Hausas in Niger and Nigeria, does not necessarily undermine my argument, since the two Hausa groups orient themselves toward different states and, therefore, compete and reformulate their identities in two separate contexts. Also, their identities have been molded within separate colonial histories (French and British, respectively). In effect, they constitute two different communities in different countries. This is similar to European Jews before World War II, who saw themselves as citizens of Germany, Italy, or Denmark, for example, even as they identified culturally or religiously with Jews across the world. Within these countries, they were often treated as a distinct group. I am indebted to Richard Joseph for calling my attention to the Miles and Rochefort study and to Roger Smith for clarification on Jews in Europe. The interpretation is wholly mine.
12 For concise discussions of the distinction between liberal and republican citizenship, see Oldfield 1990 and Smith 1988. For longer treatments, see, for example, Flathman 1989; Kymlicka 1989, 1995; Moon 1993; and Walzer 1970.
13 A comparative analysis is beyond the scope of this study, since it would require a deep immersion in the political history of each country and contextual reading of all the evidence to explicate the theoretical argument. The use of single-country case studies to demonstrate, elaborate, or challenge theory has produced critical advances in citizenship theory, as demonstrated by Peled's (1992) work on Israel; various works on the United States, such as Shklar (1991) and Smith (1988); and, more recently, Halisi's (1997) work on South Africa.
14 A similar trend in formation and structure was evident in the parties that emerged in the transition from single-party rule in Kenya in 1990. See Barkan 1993 and Muigai 1995.
15 This denial proved instrumental in the ethnic clashes that erupted in this area during the transition to multiparty politics in 1990-92, as Saboti expelled non-Kalenjin (especially the Abaluhya), who had come to dominate the area politically and economically and who allied with the opposition parties against KANU (Republic of Kenya 1992, 22-4).
16 This is a common tactic used by dominant groups to discourage ethnic countermobilization, as Lemarchand (1994, 31) shows in the case of the Tutsi elite in Burundi.
17 Indeed, despite his dismissal of "tribalism," Mboya accepted the inescapable nature of the commitment to tribe: "Even African leaders and heads of African states have not succeeded in transforming themselves completely into individual personalities: They are still to some degree communal because their background and their relatives remain tribal, and so they themselves cannot afford to change at the risk of offending their family" (Mboya 1963, 64).
18 While land is critical to citizenship claims in Kenya, other icons of identity may provoke conflict more powerfully in other countries. Regardless of the material object that defines identity or animates citizenship conflicts, the basic dynamic is one of competition for command over or access to resources and political power. Thus, in South Africa, language (especially among Afrikaners) is a critical issue in defining citizenship and self-determination; in Liberia, race is a salient divide among the Americo-Liberians, "white" immigrants such as the Lebanese, and indigenous groups (Konneh 1996); while among the secessionist Katangans in the then Congo, regionalism and control of mineral wealth were important to citizenship claims.
19 Land in colonial Kenya was divided into three categories: Scheduled Land was fertile land alienated and occupied by Europeans in the highlands and in the Rift Valley. Native Reserves land was reserved for Africans and was under communal ownership, without title. Crown Land was all unalienated land (forests, parks, etc.), which was considered state land.
20 The seemingly smooth transition belied potentially catastrophic machinations launched by a faction within the ruling elite intent on retaining power in Kikuyu hands after Kenyatta's death (see Karimi and Ochieng 1980).
21 Significantly, most critical positions (e.g., state security) were occupied by members of the Tugen, Elgeyo, and Marakwet subgroups of the Kalenjin, from which Moi and his closest allies hail (see Africa Confidential 1990a, 1). This is consistent with the hierarchy of allegiances (from clan, to subgroup, to ethnic group) discussed earlier. It suggests the possibility of having multiple and shifting citizenships, especially within an ethnic group, which, as pointed out, is not monolithic. I focus here on the conflict between ethnic and national identity (rather than on subgroups, for instance) because that is the point at which the distinction is most relevant to national politics and to democratic transitions and institutions.
22 As recently as December 1995, prominent KANU politicians were making similar demands that "nonindigenous" settlers in Maasailand (in this case Kikuyu) be excluded from participating in elections and from voter rolls. See Daily Nation, December 2, 1995.
23 At present he is Minister for Home Affairs.
24 Under current circumstances it is perhaps unrealistic to seek to transform ethnic citizenship into a liberal form or, alternatively, to transform national citizenship into a strongly civic-republican form.
Abwunza, Judith M. 1993. "Ethnonationalism and Nationalism Strategies: The Case of the Avalogoli in Western Kenya." In Ethnicity and Aboriginality, ed. Michael D. Levin. Toronto: University of Toronto Press.
Africa Confidential. 1990a. 31 (June 1):1-2.
Africa Confidential. 1990b. 31(October 26):1-3.
Africa Confidential. 1996. 37(July 19):3.
Ambler, Charles H. 1988. Kenyan Communities in the Age of Imperialism: The Central Region in the Age of Imperialism. New Haven, CT: Yale University Press.
Anderson, Benedict. 1991. Imagined Communities: Reflections on the Origin and Spread of Nationalism. Rev. ed. New York: Verso.
Azikiwe, Nnamdi. 1978. "From Tribe to Nation: The Case of Nigeria." In Themes in African Social and Political Thought, ed. Onigu Otite. Enugu, Nigeria: Fourth Dimension.
Barbalet, J. M. 1988. Citizenship. Minneapolis: University of Minnesota Press.
Barkan, Joel D. 1993. "Kenya: Lessons from a Flawed Election." Journal of Democracy 4(July):85-99.
Barkan, Joel D., and Michael Chege. 1989. "Decentralizing the State: District Focus and the Politics of Reallocation in Kenya." Journal of Modern African Studies 27(September):438-46.
Bates, Robert H. 1974. "Ethnic Competition and Modernization in Contemporary Africa." Comparative Political Studies 6(January): 457-84.
Bendix, Reinhard. 1964. Nation-building and Citizenship. New York: John Wiley.
Bennett, George, and Carl G. Rosberg. 1961. The Kenyatta Election: Kenya 1960-1961. London: Oxford University Press.
Berg-Schlosser, Dirk. 1985. "Elements of Consociational Democracy in Kenya." European Journal of Political Research 13(March):95-109.
Bienen, Henry, and Jeffrey Herbst. 1996. "The Relationship between Political and Economic Reform in Africa." Comparative Politics 29(October):23-42.
Binder, Leonard, L. Pye, J. Coleman, S. Verba, J. Lapalombara, and M. Weiner. 1971. Crises and Sequences in Political Development. Princeton: Princeton University Press.
Brantley, Cynthia. 1981. The Giriama and Colonial Resistance in Kenya, 1800-1920. Berkeley: University of California Press.
Brown, Brack. 1970. A Guide to the Constitutional Development of Kenya. Nairobi: Government Printer.
Citizens Coalition for Constitutional Change. 1995a. "Press Release." Issued in Nairobi on April 8. Typescript.
Citizens Coalition for Constitutional Change. 1995b. "Press Statement of Steering Committee." Issued in Nairobi on January 6. Typescript.
Citizens Coalition for Constitutional Change. 1995c. "Report on the Seminar
on Labor and Economic Rights." Seminar held at the Milimani Hotel, Nairobi, on March 23.
Citizens Coalition for Constitutional Change. circa 1995d. Challenges to Change: The Way Forward. Nairobi: the Coalition.
Citizens Coalition for Constitutional Change. circa 1995e. Summary of the Report of Caucus I. Nairobi: the Coalition.
Cohen, David W., and E. S. A. Odhiambo. 1992. Burying SM: The Politics of Knowledge and the Sociology of Power in Africa. London: James Currey.
Collins, Robert O., ed. 1990. Eastern African History Texts and Readings. New York: Markus Wiener.
Comaroff, Jean, and John Comaroff, eds. 1993. Modernity and Its Malcontents. Chicago: University of Chicago Press.
Curtis, John W. 1995. Opportunity and Obligation in Nairobi: Social Networks and Differentiation in the Political Economy of Kenya. Munster, Hamburg: Lit Verlag.
Daily Nation. 1995. December 2.
Davidson, Basil. 1992. Black Man's Burden: Africa and the Curse of the Nation-State. New York: Random House.
Davison, Jean. 1996. Voices from Mutira. Boulder, CO: Lynne Rienner.
Diamond, Larry. 1987. "Issues in the Constitutional Design of a Third Nigerian Republic." African Affairs 86(April):209-26.
Diamond, Larry. 1988a. "Nigeria: Pluralism, Statism, and the Struggle for Democracy." In Democracy in Developing Countries: Africa, Volume Two, ed. Larry Diamond, Juan J. Linz, and Seymour Martin Lipset. Boulder, CO: Lynne Rienner.
Diamond, Larry. 1988b. Class, Ethnicity and Democracy in Nigeria. Syracuse: Syracuse University Press.
Diamond, Larry, and Marc F. Plattner. 1994. "Introduction." In Nationalism, Ethnic Conflict, and Democracy, ed. Larry Diamond and Marc F. Plattner. Baltimore, MD: Johns Hopkins University Press.
Ekeh, Peter P. 1975. "Colonialism and the Two Publics in Africa: A Theoretical Statement." Comparative Studies in Society and History 17(January):91-112.
Ekeh, Peter P. 1978. "Colonialism and the Development of Citizenship in Africa: A Study of Ideologies of Legitimation." In Themes in African Social and Political Thought, ed. Onigu Otite. Enugu, Nigeria: Fourth Dimension.
Enloe, Cynthia H. 1973. Ethnic Conflict and Political Development. Boston: Little, Brown.
Flathman, Richard E. 1989. Toward a Liberalism, Ithaca, NY: Cornell University Press.
Frost, Frank J. 1994. "Aspects of Early Athenian Citizenship." In Athenian Identity and Civic Ideology. ed. Alan L. Boegehold and Adele C. Scafuro. Baltimore, MD: Johns Hopkins University Press.
Ghai, Y. P. 1967. "Independence and Safeguards in Kenya." East African Law Journal 3(2):177-217.
Ghai, Y. P., and J.P. W. B. McAuslan. 1970. Public Law and Political Change in Kenya. London: Oxford University Press.
Glickman, Harvey, ed. 1995. Ethnic Conflict and Democratization in Africa. Atlanta, GA: African Studies Association Press.
Grillo, Ralph. 1973. African Railwaymen. Cambridge: Cambridge University Press.
Halisi, C. R. D. 1997. "From Liberation to Citizenship: Identity and Citizenship in Black South African Political Thought." Comparative Studies in Society and History 39(January):66-85.
Her Majesty's Stationery Office. 1960. Report of the Kenya Constitutional Conference held in London in January and February. London: HMSO.
Her Majesty's Stationery Office. 1963a. Kenya Independence Order in Council 1963 - Kenya Constitution. London: HMSO.
Her Majesty's Stationery Office. 1963b. The Kenya Order in Council 1963. London: HMSO.
Holmquist, Frank, and Michael Ford. 1995. "Stalling Political Change: Moi's Way in Kenya." Current History 94(April):177-81.
Horowitz, Donald L. 1985. Ethnic Groups in Conflict. Berkeley: University of California Press.
Horowitz, Donald L. 1991. A Democratic South Africa? Berkeley: University of California Press.
Horowitz, Donald L. 1994. "Democracy in Divided Societies." In Nationalism, Ethnic Conflict, and Democracy, ed. Larry Diamond and Marc F. Plattner. Baltimore, MD: Johns Hopkins University Press.
Jones, N. S. Carey. 1967. The Anatomy of Uhuru. New York: Praeger Publishers.
Joseph, Richard. 1987. Democracy and Prebendal Politics in Nigeria. Cambridge: Cambridge University Press.
Kaiser, Paul J. 1996. "Structural Adjustment and the Fragile Nation: The Demise of Social Unity in Tanzania." Journal of Modern African Studies 34(June):227-37.
Kalinga, Owen J. M. 1985. "Colonial Rule, Missionaries and Ethnicity in the North Nyasa District, 1891-1938."African Studies Review 28(March):57-72.
Kaplan, Irving. 1984. "The Society and Its Environment." In Kenya, A Country Study, ed. Harold D. Nelson. Washington, DC: Department of the Army.
Karimi, Joseph, and Phillip Ochieng. 1980. The Kenyatta Succession. Nairobi: Transafrica Book Distributors.
Kenya Catholic Bishops. 1994. "On the Road to Democracy." Pastoral Letter. Nairobi.
Kenya Catholic Bishops. 1995. "A Call to a Change of Heart." Pastoral Letter. Nairobi.
Kibwana, Kivutha. 1990. "Land Tenure." In Themes in Kenyan History, ed. William R. Ochieng'. Nairobi: Heinemann.
Konneh, Augustine. 1996. "Citizenship at the Margins: Status, Ambiguity, and the Mandingo of Liberia." African Studies Review 39(September):141-54.
Kymlicka, Will. 1989. Liberalism, Community and Culture. Oxford: Clarendon Press.
Kymlicka, Will. 1995. Multicultural Citizenship. Oxford: Clarendon Press.
Lemarchand, Rene. 1994. Burundi: Ethnocide as Discourse and Practice. New York: Cambridge University Press.
Lemarchand, Rene. 1995. "Four Models for Resolving Ethnic Conflict." African Voices 4(Summer):l-3, 7.
Lentz, Carola. 1994. "'They Must Be Dagaba First and Any Other Thing Second. . .' The Colonial and Post-Colonial Creation of Ethnic Identities in North-western Ghana." African Studies 52(2): 57-91.
Lijphart, Arend. 1977. Democracy in Plural Societies. New Haven, CT: Yale University Press.
Marshall, T. H. 1965. Class Citizenship and Social Development. Garden City, NY: Anchor Books.
Mboya, Tom. 1963. Freedom and After. Boston: Little, Brown.
Mboya, Tom. 1970. The Challenge of Nationhood. Nairobi: Heinemann.
Melson, Robert, and Howard Wolpe. 1970. "Modernization and the Politics of Communalism: A Theoretical Perspective." American Political Science Review 64(December):1112-30.
Middleton, John, and Greet Kershaw. 1965. The Central Tribes of the North-Eastern Bantu. London: International African Institute.
Miles, William F. S., and David A. Rochefort. 1991. "Nationalism versus Ethnic Identity in SubSaharan Africa." American Political Science Review 85(June):393-403.
Miller, David. 1995. "Citizenship and Pluralism." Political Studies 43(September):432-50.
Moon, J. Donald. 1993. Constructing Community. Princeton, N J: Princeton University Press.
Muigai, Githu. 1995. "Ethnicity and the Renewal of Competitive Politics in Kenya," In Ethnic Conflict and Democratization in Africa, ed. Harvey Glickman. Atlanta: African Studies Association Press.
Mutunga, Willy. 1994. "Opening Remarks on a Proposal for a Model Constitution." Speech delivered in Nairobi on December 9. Typescript.
Mwangaza Trust. 1995. "These Are the Ground Rules for a New Constitution." Press statement issued in Nairobi on January 1. Typescript.
National Council of Churches of Kenya. 1992. The Cursed Arrow: A Report on Organized Violence against Democracy in Kenya. Nairobi: NCCK.
National Council of Churches of Kenya. 1993a. The Clashes Update 2(August 18):1-8.
National Council of Churches of Kenya. 1993b. The Clashes Update 2(August 31):1-5.
National Council of Churches of Kenya. 1995. Kenya: Our Common Future. Nairobi: the Council.
National Council of Churches of Kenya. N. d. Report of Symposium I: Task Force on Ethnic Clashes. Nairobi: NCCK. Forthcoming.
Njonjo, Apollo. 1977. "The Africanization of the 'White Highlands': A Study in Agrarian Class Struggle in Kenya. 1950-1975." Ph.D. diss. Princeton University.
Nzouankeu, Jacques Mariel. 1993. "The Role of the National Conference in the Transition to Democracy in Africa: The Cases of Benin and Mali." Issue 21(1-2):44-50.
Odinga, Oginga. 1967. Not Yet Uhuru. New York: Hill and Wang.
Ogot, Bethwell A. 1981. Historical Dictionary of Kenya. Metuchen: Scarecrow Press.
Okumu, John J. 1975. "The Problem of Tribalism in Kenya." In Race and Ethnicity in Africa, ed. Pierre L. van den Berghe. Nairobi: East African Publishing House.
Oldfield, Adrian. 1990. "Citizenship: An Unnatural Practice." Political Quarterly 61(April-June):177-87
Olorunsola, Victor A. ed. 1972. The Politics of Cultural Sub-nationalism in Africa. Garden City, NY: Anchor Books.
Ottaway, Marina. 1994. Democratization and Ethnic Nationalism: African and Eastern European Experiences. Washington, DC: Overseas Development Council.
Peled, Your. 1992. "Ethnic Democracy and the Legal Construction of Citizenship: Arab Citizens of the Jewish State." American Political Science Review 86(June):432-43.
Republic of Kenya. 1992. Report of the Parliamentary Select Committee to Investigate Ethnic Clashes in Western and Other Parts of Kenya, 1992. Nairobi: Government Printer.
Robinson, Pearl T. 1994. "The National Conference Phenomenon in Francophone Africa." Comparative Studies in Society and History 36(July):575-610.
Roche, Maurice. 1995. "Citizenship and Modernity." British Journal of Sociology 46(December):715-33.
Rothchild, Donald. 1964. Safeguarding Nigeria's Minorities. Pittsburgh, PA: Duquesne University Press.
Rothchild, Donald. 1968a. "Majimbo Schemes in Kenya and Uganda." In Boston University Papers on Africa: Transition in African Politics, ed. Jeffrey Butler and A. A. Castagno. New York: Praeger Publishers.
Rothchild, Donald. 1968b. Politics of Integration. Nairobi: East African Publishing House.
Rothchild, Donald. 1973. Racial Bargaining in Independent Kenya. London: Oxford University Press.
Rothchild, Donald, and Victor A. Olorunsola, eds. 1983. State versus Ethnic Claims. Boulder, CO: Westview Press.
Rousseau, Jean-Jacques. 1987. The Basic Political Writings. Trans. Donald A. Cress. Indianapolis: Hackett.
Selassie, Alemante G. 1992. "Ethnic Identity and Constitutional Design for Africa." Stanford Journal of International Law 29(1):1-56.
Shklar, Judith N. 1991. American Citizenship. Cambridge, MA: Harvard University Press.
Singh, Chanan. 1965. "The Republican Constitution of Kenya: Historical Background and Analysis." International and Comparative Law Quarterly 14(July):878-949.
Sisk, Timothy D. 1995. Democratization in South Africa. Princeton, NJ: Princeton University Press.
Sithole, Masipula. 1985. "The Salience of Ethnicity in African Politics: The Case of Zimbabwe." Journal of Asian and African Studies 20(December):181-91.
Sklar, Richard. 1993. "The African Frontier for Political Science." In Africa and the Disciplines, ed. Robert Bates, V. Y. Mudimbe, and Jean O'Barr. Chicago: University of Chicago Press.
Smith, Rogers M. 1988. "The American Creed and American Identity: The Limits of Liberal Citizenship in the United States." Western Political Science Quarterly 41 (June):225-51.
Soysal, Yasemin Nuhoglu. 1994. Limits of Citizenship. Chicago: University of Chicago Press.
Stewart, Angus. 1995. "Two Conceptions of Citizenship." British Journal of Sociology 46(March):63 - 78.
Throup, David W. 1987. "The Construction and Destruction of the Kenya State." In The Political Economy of Kenya, ed. Michael Schatzberg. New York: Praeger Publishers.
Vail, Leroy, ed. 1989. The Creation of Tribalism in Southern Africa. Berkeley: University of California Press.
van Binsbergen, Wim M. J. 1994. "The Kazanga Festival: Ethnicity as Cultural Mediation and Transformation in Central Western Zambia." African Studies 52(2):92-125.
van den Berghe, Pierre L., ed. 1975. Race and Ethnicity in Africa. Nairobi: East African Publishing House.
Walzer, Michael. 1970. Obligations. Cambridge, MA: Harvard University Press.
Waters, Malcolm. 1989. "Citizenship and the Constitution of Structures Social Inequality." International Journal of Comparative Sociology 30(December):159-80.
Widner, Jennifer A. 1992a. "Kenya's Slow Progress toward Multiparty Politics." Current History, 91(May):214-8.
Widner, Jennifer A. 1992b. The Rise of a Party State in Kenya. Berkeley: University of California Press.
Young, Crawford. 1976. The Politics of Cultural Pluralism. Madison: University of Wisconsin Press.
Stephen N. Ndegwa is Assistant Professor of Government, The College of William and Mary, Williamsburg, VA 23187-8795.
|Printer friendly Cite/link Email Feedback|
|Author:||Ndegwa, Stephen N.|
|Publication:||American Political Science Review|
|Date:||Sep 1, 1997|
|Previous Article:||Voting correctly.|
|Next Article:||Echoes from the past: the relationship between satisfaction with economic reforms and voting behavior in Poland.|