The decade from 1990 to 2000 saw a total of seventy-eight top leadership elections involving forty-three of the forty-eight sub-Saharan African countries. Of these, only twenty-one elections led to power transition from an incumbent to an opposition political party in nineteen countries. Paradoxically, even where there was such transition, authoritarian tendencies persisted. Focusing on Kenya and Zambia, this article argues and seeks to demonstrate that the limited number of transitions from an incumbent regime to an opposition party and the persistence of authoritarianism are a function of political liberalization without democratization of political institutions and rules of the political game.
The decade from 1990 to 2000 was a period of sustained political activity in Africa leading toward democratization. During this period, most of die African countries shifted from single-party authoritarianism or military dictatorship to multiparty polities. Whereas in some cases (such as Cape Verde, Benin, Madagascar, Malawi, Mali, and Zambia) this transition resulted in the ouster of old authoritarian regimes, in others (such as Cameroon, Cote dTvoire, Kenya, Togo, and Zimbabwe) the old authoritarian regimes continued to reproduce themselves in power. Even where new regimes emerged, autiioritarian tendencies persisted, as the case of Chiluba's Zambia illustrates. Focusing specifically on the centrality of electoral system design to electoral contests and their outcomes, this article argues that the limited transitions from incumbent regimes and the persistence of authoritarianism are a function of effecting political liberalization without democratizing the political systems and the rules of the game. By political liberalization I mean the mere act of legalization of opposition political parties and their freedom to contest political office. Democratization, by contrast, entails redesigning the rules of the game, especially the electoral system, to make political competition fairer and restructuring governance institutions to take account of the changed political landscape and to make them more responsive and accountable to the electorate.
The article begins with an exposition of types of electoral systems and then examines the electoral outcomes of the first multiparty elections in selected countries in Africa. It then focuses on the cases of Kenya and Zambia, and finally delineates the implications of political transition without transformation of the rules of the game. The significance of the two cases lies in the contradictory scenarios they present, despite their similarities in social and political circumstances, in regard to democratic advancement in Africa. Kenya and Zambia are former British colonies. They achieved political independence within one year of each other. In both countries the process of political consolidation saw a falling out between the president and his vice president, laying the ground for the establishment of singleparty states. By the time of the 1990s wave of democratization, both Kenya and Zambia were single-party states by law. Yet in Zambia in 1991, the opposition united against the incumbent to score the first opposition electoral victory in 1990s elections in Africa. In Kenya, the opposition was fragmented, facilitating the incumbent party's ability to retain power in both the 1992 and 1997 elections. In the third multiparty elections in the two countries, the situation was reversed; in Zambia in 2001, opposition fragmentation facilitated victory for the new incumbent, while in Kenya in 2002 the opposition united to edge out the authoritarian incumbent. These contradictory scenarios are a function of the electoral system's design in the two countries, a crucial determinant of political competitiveness and, ipso facto, of electoral outcomes.
Institutional Design for Democracy
The aspiration for democratic rule is an almost universal phenomenon, as illustrated by the ubiquitous wave of democratization across the world at the close of the last century. Indeed, the concept of democracy is so honored that all manner of political systems claim to be democracies. Even countries that have never held an election in decades, such as the former Zaire, are conveniently baptized "democratic republics," while a country such as Uganda, in which political parties were banned between 1986 and 2006, was called a "no-party democracy." However, the prerequisite for representative or republican democracy-a condition to which most of the democratizing countries in Africa aspire-is the realization of a number of institutional guarantees. First, political authority must be based on a limited mandate, with citizens reserving the right to renew it periodically in free and fair elections. second, elections must be based on universal adult suffrage and on genuine choices between alternative political parties. Third, citizens must be guaranteed the right and freedom of association, expression, and the eligibility, in principle, of any citizen to seek public office. Fourth, aspiring political leaders must be afforded the right to compete freely for support and votes, buttressed by multiple channels of political communication. Fifth and finally is the imperative need for democratically accountable governmental decision-making institutions, with elected officials free from overriding opposition from unelected officials (Walle 2003; Harbeson 1999; Gitonga 1987; Dahl 1982).
In short, the realization of democracy is contingent upon rules of the game that provide for alternative political parties competing against one another for the chance to govern within institutional systems that guarantee fairness and a genuine opportunity for alternation of power between parties. The significance of such institutional design and appropriate rules of the game is illustrated by the case of Mexico. In spite of the existence of multiple political parties in Mexico and regular electoral contests among them, it took over seven decades for Mexicans to be able to elect a president from one of the opposition parties-National Action Party's (PAN) Vicente Fox-in 2000.1 But, as Anyang' Nyong'o (2002) points out, Fox would not have been elected had not his predecessor, President Ernesto Zedillo (of the Institutional Revolutionary Party [PRI]) changed the country's electoral laws to facilitate democratic elections in four main ways. First, Zedillo granted the national electoral commission independence from the executive and the ruling party. Second, he introduced laws to guard against election rigging. Third, he made parliamentary representation fairer. Fourth and finally, he changed presidential election laws, pegging the election on an absolute majority rather than a mere plurality of the popular vote.
An electoral system is designed to do three things. First, it serves to translate votes cast into seats won in a legislative chamber. The system may give more weight to proportionality so that the disparity between a political party's share of votes and its share of seats is not great. Alternatively, it may funnel votes (however fragmented among parties) into a parliament that contains two large parties representing polarized views. According to Arend Ljphart (1998), presidential systems can have an important effect on legislative elections if presidential elections are based on a plurality or first-past-the-post (FPTP) electoral system and if legislative elections are held at the same time. In such circumstances, large parties have an advantage in presidential elections, since small parties do not have much of a chance to have one of their candidates elected. This advantage tends to carry over into legislative elections. Hence presidentialism tends to discourage multipartyism, as illustrated by the cases of the United States and Costa Rica. second, an electoral system serves as a conduit through which the people can hold their representatives accountable. In this regard, the electorate reserves the right to renew or revoke the mandate given to their representatives at regular electoral intervals based on the performance of the leaders in regard to ensuring the people's economic well being, security, and social welfare. Third, electoral systems serve the normative function of structuring the boundaries of "acceptable" political discourse and giving incentives for those competing for power, especially political parties, to couch their appeal to the electorate in distinct ways (Reynolds & Sisk 1998:19). The design of an electoral system thus determines electoral competitiveness and the possibility of alternation of power between parties.
The main electoral system choices are between plurality-majority systems and proportional representation (PR) systems. Plurality-majority systems often use single-member districts (SMD). In a plurality, or first-pastthe-post (FPTP) system, the winner is the candidate with the most votes, not necessarily an absolute majority of the votes. In a majoritarian system, the winner must garner an absolute majority (50% + 1). The underlying rationale for PR, on the other hand, is to reduce the disparity between a party's share of national votes and its share of parliamentary seats. For instance, garnering 40 percent of the votes should translate into winning 40 percent of the legislative seats (Lijphart 1998:10). The difference between the percentage of votes received by a party and the percentage of seats won in a legislature is called disproportionality. Whereas deviations of plus or minus five percentage points are within acceptable limits, the farther the index moves away from five, the more disproportional and thus the more skewed the electoral system. In addition, PR systems enhance the representation of minority groups. Pure FPTP or plurality SMD systems are mainly found in the U.K. and the countries it influenced. In Africa, Botswana, Gambia, Ghana, Kenya, Lesotho, Malawi, São Tome and Principe, Zambia, and Zimbabwe all use SMD plurality systems. List-PR is the electoral system most frequently used in democracies and is increasingly becoming common. More than twenty established democracies use some variant of list-PR with seats allocated within a number of regionally based districts. Germany, Israel, and the Netherlands allocate seats on the basis of a single nationwide district. In Africa, Benin, Burkina Faso, Burundi, Madagascar, Mozambique, and Niger follow the district-based listPR model; while Angola, Namibia, Senegal, and South Africa calculate seats at the national level (Reynolds & Sisk 1998:21).
There are a number of electoral system dimensions that have important consequences for the proportionality of electoral outcomes and for party systems. First is the electoral formula, including the majoritarian-plurality, PR, and semi-PR formulas. The second key factor is the district magnitude, by which is meant the number of representatives elected per district. "The greater the number of members it [district] elects, the more closely will the result approximate to proportionality" (Lijphart, 1998:11). This, however, applies to multimember electoral districts. With regard to SMDs, which are the norm in most African countries, the relevant factor is constituency size in terms of population. The more equal the demarcation of constituencies in terms of population, the greater the proportionality of representation. The third dimension is electoral threshold, which applies to PR electoral systems. This refers to the minimum level of support that a party needs in order to gain representation. It is usually expressed in terms of percentage of votes garnered. For instance, in South Africa a party needs to garner at least 5 percent of the votes cast in order to gain parliamentary representation. Finally, assembly size is the fourth dimension that has a strong influence on proportionality and on the degree of multipartyism. Lijphart (1998) notes that the smaller the assembly, the less proportional and the less the number of parties it promotes.
Available evidence indicates that of the four factors above, the first two-electoral formula and district magnitude-have the most significant impact on proportionality of electoral representation and the nature of the party system (see Nasong'o 2004; Lijphart 1998). Against this background, Reynolds and Sisk (1998) place analytical premium on the role of institutional design, especially on the role of elections and the systems under which they are contested in the democratization and conflict management processes. They posit that the choices about the basic rules of the game affect its outcomes. Indeed, the electoral system design determines the competitiveness of elections and hence the possibilities of either alternation of power between different parties or the entrenchment of one particular party in power. Electoral outcomes of transitional multiparty elections in a cross-section of African countries largely vindicate this contention. In most of these elections, defeat of the incumbent regime occurred largely in cases in which the elections were held following a negotiated redesigning of the rules of the political game intended to enhance fair competition.
The first example is Benin, where long-serving President Mathieu Kérékou was defeated by Nicephore Soglo in the first multiparty elections of 1991. The constitution-making process was all-encompassing and established a transitional government. Accordingly, the strategic considerations of the politically ambitious were shaped by the near-absolute uncertainty about future electoral outcomes. This led to agreement on constitutional checks and balances that made it difficult for one political institution or individual to control the entire apparatus of state (Magnusson 1999:222-23).
The second example is Malawi, where self-declared "President-for-Life" Kamuzu Banda was defeated in 1994 under circumstances similar to those of Benin. In the run-up to the first multiparty elections in 1994, the process of negotiating for new rules between the Public Affairs Committee (PAC)-composed of representatives from the Alliance for Democracy (AFORD) led by Chakufwa Chihana, Bakili Muluzi's United Democratic Front (UDF), churches, and Malawi Law Society-and the Presidential Committee on Dialogue (PCD)-composed of cabinet ministers and members of the ruling Malawi Congress Party's (MCP) National Executive Committee-culminated in Malawi's transitional National Consultative Council (NCC). The NCC reviewed electoral laws and procedures, revised electoral constituency boundaries to make representation fairer, and drafted a new constitution before the 1994 multiparty elections. As in Benin, therefore, the incumbent regime in Malawi was removed from control of the electoral process and the state apparatus in the transitional period, thus facilitating some semblance of fair competition resulting in the defeat of the entrenched MCP regime of Kamuzu Banda (see van Dijk 2000; Venter 1995).
The third example is Mali, which was fortunate in being able to write a new constitution and hold multiparty elections without the burden of continued participation by an incumbent party and head of state (Vengroff & Kone 1995:45). The long-serving military ruler, Moussa Traoré, was toppled on March 24, 1991, by Lt. Col. Amadou Toumani Touré on account of his intransigence and high-handed clamping down on prodemocracy demonstrators. The new regime pledged a return to civilian rule within a year and constituted the Comité de Transition pour le Salut du Peuple (CTSP), composed of both military and civilian representatives. The CTSP set up a transitional government that organized a national conference, which wrote a new constitution, a new electoral code, and a legal framework for establishing political parties-La Charte des Partis. The inauguration of a new constitution in Mali paved the way for municipal elections on January 19, 1992, national assembly elections in two rounds on February 24 and March 9, 1992, and presidential elections in two rounds on April 13 and 27. Both the parliamentary and presidential elections were organized on a majoritarian electoral system. According to Vengroff and Kone (1995:48), "much of this was possible because the entrenched economic and political interests identified with the Traoré regime were effectively out of power and thus unable to skew the process as has happened in some other African countries."
Fourth and finally is the case of South Africa's transition from apartheid to majority rule in 1994. Here also, the transition was facilitated by detailed negotiations on an interim constitution, which preceded the holding of the first postapartheid national elections. These negotiations also produced an interim government of national unity that reinforced the stability of the transition through the first year of African National Congress (ANC) leadership. The interim constitution was subsequently replaced by a permanent one that was ratified by the combined houses of the South African parliament serving as a constituent assembly.
Table 1 illustrates the significance of the electoral system design-the rules of the electoral game-to electoral outcomes in regard to the first multiparty elections in selected African countries. In the table, of the sixteen cases under review, change of regime in the first multiparty elections occurred in 50 percent of the cases. Of these eight cases, 75 percent provided for an absolute majority to win the presidency, which led to a second round of elections between the top two candidates that resulted in defeat of the incumbent. Of the other eight cases that witnessed no change of regime in their first multiparty presidential contests, 87 percent provided for a simple plurality of votes to win the presidency. The only exceptions in the table are Chad, which provided for an absolute majority but the incumbent won in the first round; and Zambia, with a plurality electoral system that led to incumbent defeat. Whereas the case of Chad is the exception that proves the rule, the Zambian case is explicable in the fact that the contest was between two candidates, reminiscent of round-two elections in the other cases in which change of regime occurred.
Whereas the foregoing analysis may not prove a causal relationship between electoral system design and electoral outcomes, it does underscore a strong correlation between the two. It demonstrates that the more competitive the systems, the fairer the competition and the greater the chances for alternation of power between political parties. Barring such competitive systems, for instance, Kenya's Daniel arap Moi, Cameroon's Paul Biya, and Gabon's Omar Bongo all won in 1992 with 36 percent, 40 percent, and 49 percent, respectively, of the votes cast. With regard to Kenya, the results of the 1992 and 1997 presidential elections are shown in Table 2. Had the electoral system been redesigned in Kenya to provide for a majority to win rather than a mere plurality, it is arguable that a runoff between the top two presidential candidates would have resulted in incumbent defeat. Barring this, Moi emerged victorious with just below 37 percent and 41 percent of the votes cast in 1992 and 1997, respectively. The institutional environment was not conducive to fair competition. As Hartman (1999:477-78) argues, "it was not the disagreement between the opposition leaders but the noncompetitive setting that has to be regarded as the principal reason for their defeat." The cases of Kenya and Zambia illustrate with particular clarity the contention that political transition without transformation of the political institutions, especially the electoral system's design, is inimical to democratic advancement.
New Game, Old Rules: Kenya and Zambia
Kenya and Zambia provide contradictory scenarios in regard to political competition and electoral outcomes. Both of these countries use the single-member district first-past-the-post (SMD-FPTP) electoral system. The first multiparty elections in Zambia in 1991 saw the defeat of the incumbent regime of President Kenneth Kaunda of the United National Independence Party (UNIP) by the Movement for Multiparty Democracy (MMD) led by Frederick Chiluba. In Kenya, the incumbent KANU regime of President Daniel arap Moi won the first multiparty elections in 1992 and went on to win again in 1997 in the face of a fragmented opposition. Eventually, KANU was defeated in 2002 by a coalition of opposition parties in the name of the National Alliance Rainbow Coalition (NARC), which got 62 percent of the presidential votes against KANU's 31 percent. Interestingly, Zambia's third multiparty elections of 2001 were a replica of the Kenyan ones of 1992 and 1997; opposition to the incumbent MMD was so divided that the MMD presidential candidate, Levy Mwanawasa, faced ten opposition candidates. This facilitated Mwanawasa's victory with a mere 28 percent of the votes cast against his closest competitor, Anderson Mazoka of the United Party for National Development (UPND), who garnered 27 percent of the votes. These contradictory scenarios in the cases of Kenya and Zambia are a function of failure to redesign the electoral system in both countries to enhance political competition and thus advance democratic progress. A closer look at the two cases illustrates this position.
The Paradox of Opposition Victory in Zambia
Unlike in the illustrative cases of Benin, Malawi, Mali, and South Africa cited above, in Zambia there was no major comprehensive overhaul of the electoral system, nor a restructuring of the fundamental rules of the political game prior to the first multiparty elections of 1991. The incumbent president, Kenneth Kaunda, attempted to unilaterally rewrite the constitution, a move that was informed by a desire to enhance his democratic credentials and the political attractiveness of his party, UNIP, for purposes of securing victory in the first multiparty elections. Toward this end, Kaunda established a constitutional commission, which came up with a number of proposals. These included a proposal to create the post of vice president, appointment of ministers from outside parliament, expansion of the National Assembly from 135 to 150 seats, and the establishment of a constitutional court.2 These constitutional proposals were a major bone of contention between the UNIP and MMD beginning June 1991. Whereas Kaunda accepted most of the constitutional commission's recommendations and had them submitted to the National Assembly for ratification, the MMD opposed them because they vested too much authority in the president and too little in the National Assembly. The MMD also opposed die appointment of cabinet ministers from outside the National Assembly, arguing that this proposal was intended simply to create opportunity for patronage and for rewarding vested interests.
The MMD threatened to boycott the elections if the National Assembly ratified the recommendations; to demonstrate its resolve, the MMD refused to attend an interparty conference chaired by Kaunda in July 1991 to discuss the draft constitution. Later in July 1991, representatives of MMD, UNIP, and other parties met to discuss the constitutional commission's proposals under the chairmanship of Deputy Chief Justice Mathew Ngulube. As a result of this meeting, President Kaunda suspended the draft constitution and agreed to further negotiations. Church leaders intervened, and a joint commission was set up to review the constitutional proposals. As a result, Kaunda agreed to abandon the proposals for a constitutional court, appointment of cabinet ministers from outside the National Assembly, and presidential powers to impose marshal law. The National Assembly ratified the remaining proposals on August 2, 1991.
The very fact of forcing President Kaunda to rescind some of the proposed constitutional changes was a major victory for the MMD. This fact demonstrated the vulnerability of the Kaunda regime and signaled the beginning of the end of the UNIP government even before elections were held. The powerful Kaunda had been forced to give in on practically all opposition demands. In quick succession, Kaunda agreed to allow international observers into the country to monitor the elections, a position he had previously opposed. He also agreed to grant subsidies to all registered parties (Ihonvbere 1996:117-19). It is important to note, however, that though the opposition succeeded in forcing Kaunda to shelve some of his most important constitutional changes and to adopt those that were favored by the opposition, these changes were largely concerned with the exercise of executive powers in the run-up to the election. They had nothing to do with the electoral system design. In the elections, Zambia went to the multiparty polls in October 1991 with the same SMD-FPTP, winnertake-all system.
The greatest strength of the opposition lay in its united front, which reduced the electoral contest to one between the incumbent UNIP and the opposition MMD, with Kaunda pitted against a single opposition presidential candidate, Frederick Chiluba. Realizing his vulnerability, the jittery Kaunda introduced last-minute cabinet changes in the hope of demonstrating a stronger resolve to provide a democratically credible and accountable leadership. Nonetheless, the united opposition achieved an impressive victory over Kaunda and UNIP. John Wiseman (1995:9) observes that although Kaunda was not averse to using his control of the state to try to create electoral advantages for UNIP, the very strength of the opposition resulted in this having limited effect and was clearly insufficient in preventing an electoral defeat of considerable magnitude. The MMD won 125 of the 150 parliamentary seats, with UNIP managing a paltry twenty-five seats, and the Zambian elections of 1991 produced the first transfer of power through the ballot box of any mainland African state.3
Following Chiluba's landslide victory over Kaunda, a Constitutional Commission was established to revise the constitution along democratic lines (Ihonvbere 1996; Mphaisha 2000). Curiously, however, the MMD government resisted strong domestic and donor pressure to publish the Constitutional Commission's draft document for public debate prior to editing by the government. In addition, the government insisted that the document be ratified by the Zambian parliament in which MMD enjoyed an 83 percent majority, and not by a national referendum. Arguably, in order for a constitution to be broadly acceptable and therefore acquire the potential to endure the test of time, the constitution-making process must engage all relevant population groups and seek to secure a voluntary and mutually beneficial consensual social contract (see Mbaku 1997). In one of its most serious departures from democratic practice, the MMD revised the constitution in the run-up to the 1996 elections to restrict eligibility to run for president to third-generation Zambians. This provision targeted Kaunda and explicidy curtailed his much-awaited rematch with Chiluba in the 1996 elections. The amendments also required traditional chiefs to abdicate their chieftaincy before becoming eligible for elected office, a provision that also targeted Kaunda's vice presidential running mate, Senior Chief Inyambo Yeta, who was sure to draw wide support for Kaunda from the Lozi of Western Province.
In effect, the MMD government disregarded the recommendations of the Mwanakatwe Constitutional Commission, Western donors, and civil society that the new constitution be ratified by a national referendum or a constituent assembly rather than by parliament. Now that the MMD was in power, it was no longer in the political elites' interest to allow for a democratic constitutional review process that would have had the effect of diluting their power. Instead, they used their positions to consolidate their power and maximize their chances for reelection in 1996. Having been so blatantly targeted by the constitutional amendments, UNIP, the biggest opposition party at the time, boycotted the 1996 elections, leaving the MMD to have a field day. It is for this reason that John Harbeson (1999:49-50) rightly notes that far from strengthening and creating momentum for democracy, the 1996 Zambian elections had the opposite effect. Instead of advancing the cause of democracy, the constitution-making process ended up serving the strategic interests of the new incumbent government by way of hedging and entrenching it in power. In essence, the resounding defeat of Kaunda marked a shift from a single-party state under UNIP to a dominant-party state under MMD. The MMD controlled more than three-quarters of the seats in parliament (125 of 150 seats), which rendered the UNIP opposition a voiceless minority that struggled with overcoming the burden of a discredited past and an unprecedented electoral humiliation. As a consequence, a decade after the return to multipartyism in Zambia and the defeat of the incumbent regime, it was within the ruling party rather than between contending parties that political competition was to be found (see Donge 1995).
The inequity of the Zambian electoral system was clearly demonstrated in the country's third multiparty elections of 2001. Here, Chiluba's handpicked successor in MMD, Levy Mwanawasa, faced ten opposition challengers for the presidency. Four of these were serious contenders, including Anderson Mazoka, Christon Tembo, Tilyenji Kaunda, and Godfrey Miyanda, as shown by the election results in table 3. Despite winning a combined total of over 71 percent of the votes cast, the opposition votes were shared among the ten opposition candidates, such that Mwanawasa scraped through to victory with a mere 28 percent of the votes cast. Had the electoral system required an absolute majority to win the presidency, a run-off between Mwanawasa and his closest rival, Anderson Mazoka of UPND, is likely to have produced an upset for MMD. This is particularly so given the disenchantment with the Chiluba regime and the high-handed manner in which he picked his successor within MMD. Indeed, through the vote counting, Mazoka maintained a lead over Mwanawasa until returns from the final twenty constituencies gave Mwanawasa a modest lead.4 This turn of events prompted accusations of rigging and resulted in widespread street demonstrations over the final result, which Mazoka unsuccessfully contested in court.
The extent of the disproportionality of the Zambian SMD-FPTP electoral system is illustrated even more clearly by the 2001 parliamentary election results as tabulated in table 4. The figures show that MMD won a total of 28 percent of the parliamentary vote, yet it secured 46 percent of the seats in the National Assembly. Similarly, UPND garnered 23 percent of the votes but secured 32 percent of the seats in parliament. The indices of disproportionality are eighteen in favor of MMD and nine in favor of UPND. The rest of the parties were disadvantaged in the sense that they received a lesser proportion of the seats than their proportion of the popular vote. The most disadvantaged in this regard was the Forum for Democracy and Development (FDD), with a disproportionality index of minus eight, followed by the Heritage Party (HP) with an index of minus five. This disproportionality is great, given that ideally, 28 percent of the votes should result in 28 percent of parliamentary seats, as is normally the case in PR electoral systems. In such cases, the index of disproportionality is normally below five. The implication here is that there are far more electoral constituencies in areas that support MMD, followed by those that support UPND. In essence, these areas are overrepresented in the National Assembly, a situation that negates the democratic principle of equal representation and equality of the vote.
The Protracted Transition in Kenya
In Kenya, as in Zambia, there was no redesigning of the electoral system before the first multiparty elections in December 1992. Internal forces for political change in the country culminated in violent riots in urban areas (see Murunga 1999). This highly charged political atmosphere was coupled with an aid crunch by international financial institutions to force the Kenya government to repeal section 2(A) of the constitution, which made the country a one-party state by law. At this moment the KANU government, like its UNIP counterpart in Zambia, was quite vulnerable and could have been pushed into reengineering a new constitution. However, the forces of democratization in the country did not find it necessary to push the incumbent regime toward this end. It would seem that the Moi regime had been so oppressive that the mere legalization of opposition political parties was received as a major act of political liberation. Accordingly, emergent political parties were more than willing to go to the polls within the rubric of the single-party electoral system design.
The electoral system's design, however, was such that the odds were stacked against the opposition. First, the provincial administration was not delinked from the electoral process and was thus used by the incumbent party to harass and intimidate the opposition. Opposition political parties found it difficult to campaign, as tiiey were denied permits for political rallies and their rallies were routinely disrupted or canceled "on security grounds." Second, because of violence, intimidation, and arrests of opposition politicians and their supporters and the declaration of so-called KANU zones in Moi's stronghold of the Rift Valley, 41 percent (18 out of 44) of the KANU candidates from the Rift Valley were returned to parliament unopposed.5 Third, the opposition was denied access to the government-controlled television and radio, the only mass media with a national reach, a fact that curtailed their capacity to effectively reach the electorate with their political message. Fourth and most important, the opposition had no say in the appointment of the electoral commission, which was single-handedly appointed by President Moi. This reality has serious implications for electoral outcomes in the country, given the role of the Electoral Commission of Kenya (ECK). The ECK is responsible for preparing and supervising elections, maintaining and revising voter registers, promoting voter education, and determining the number and boundaries of constituencies based on the most recent census. According to Foeken and Dietz (2000), it is the determination of the number of constituencies (the degree of representation) that was exploited to unfairly benefit KANU.
The Kenya constitution (1998), Chapter III section 42(3), requires that all constituencies contain as nearly equal numbers of inhabitants as appears reasonably desirable to the Commission. However, the Commission may depart from this principle in order to take account of population density, demographic trends, means of communication, geographical features, communities of interest, and the boundaries of existing administrative areas. Nonetheless, the number of inhabitants is the constitutional principle for determining the size of constituencies. The otiier factors are exceptions that the ECK may consider to justify departure from the principle of equality (see Mungai 2002). Unfortunately, as Kibe Mungai (2002) argues, since independence-particularly after 1986, when Kenya's constituencies were increased from 158 to 188, and then to 210 in the run-up to the 1997 elections-it has been ethnic considerations and boundaries of existing administrative areas, as opposed to the number of inhabitants, that have been the major factors in constituency demarcation. Exceptions to the rule have effectively been elevated above the rule itself. Table 5 illustrates a clear case of malapportionment whereby electoral constituencies have substantially unequal voting populations. Provinces in which KANU has traditionally the strongest support, including the Rift Valley, Nordieastern, and, to some extent, Coast, are overrepresented in the sense that they have more constituencies per capita than other provinces. Table 5 indicates that whereas the ideal number of voters per constituency is 49,976, Northeastern Province has an average of 19,914 voters per constituency, the Rift Valley has an average of 26,389, and Coast an average of 42,330. This unfavorably compares with the situation in Central Province, with an average of 54,000 voters per constituency, and Western Province, with an average of 50,296 voters per constituency. Such malapportionment in SMD-FPTP electoral systems like the Kenyan one systematically favors specific political parties and dius contributes to electoral disproportionality.
Against this background, though KANU had fewer overall votes in the 1992 parliamentary elections than those of the combined opposition, it nonetheless won one hundred seats compared to eighty-eight for the opposition. This is a result of the fact that KANU-dominated areas are overrepresented in parliament while opposition strongholds are underrepresented. It is arguable, further, that without this case of malapportionment, KANU would have won less than the sixty-four seats it won in the 2002 parliamentary election. Since democracy is founded on the principle of equal representation and the one-person-one-vote tenet, this principle is grossly violated in situations in which some constituencies have 26,389 voters and others 54,000 yet their respective representatives have an equal voice in the legislature. Indeed, in May 2002, two High Court judges, Justices Mbogholi Msagha and J. V. O. Juma, sitting as a constitutional court, delivered a landmark ruling with the potential of having significant implications for the development of democracy in Kenya. In the case of John N. Michuki v. Attorney General and Electoral Commission of Kenya, the justices took issue with the fact that the size of constituencies manifested serious imbalances in representation, violating the democratic principle of equal representation. Their ruling, however, fell short of making clear orders on the issue (Mungai 2002).
The KANU government continued to play the game of musical chairs with the much-needed constitutional reform necessary for advancing democracy. Against this reality, and given the SMD-FPTP Kenyan electoral system, the incumbent President Moi secured victory with 37 percent of the votes cast in the 1992 general elections compared to a combined opposition tally of 63 percent. Similarly, KANU won one hundred of the parliamentary seats to the combined opposition's eighty-eight seats. These results were replicated in the 1997 general elections when President Moi once again secured victory with 41 percent of the votes cast against 59 percent for the combined opposition. The inequity of the Kenyan electoral system was further manifested in the fact that KANU received a higher proportion of parliamentary seats (51 percent) than its proportion of the popular vote (43 percent). The disproportionality index was eight in favor of KANU (see Barkanet. al. 2001).
Finally, the concerted efforts to push the government toward constitutional reengineering bore fruit in 2000 when, through an Act of Parliament, the Constitution of Kenya Review Commission (CKRC) was constituted, paving the way for the review to commence. However, as if to underscore the KANU regime's ambivalence toward the review process, the 2002 elections were called while the review process was still in progress, albeit near completion. Nevertheless, the key opposition politicians, namely Mwai Kibaki, Michael Wamalwa, and Charity Ngilu, had found common ground and were now united under the National Alliance Party of Kenya (NAK). The position of these "Big Three" was buttressed by the fallout in KANU following President Moi's decision to single-handedly pick his successor in the party. Key politicians, including Raila Odinga, Moody Awori, Kalonzo Musyoka, and George Saitoti, decamped from KANU, took over the little-known Liberal Democratic Party (LPD), and joined forces with NAK in a coalition now renamed National Alliance Rainbow Coalition (NARC). It was within this conjuncture that KANU was trounced in the December 2002 polls, as shown in table 6. It was a scenario reminiscent of the MMD victory over UNIP in Zambia in 1991.
The decision by the opposition to present a united front in the 2002 Kenyan elections resulted in a majoritarian electoral system by default (as table 6 illustrates), resembling the Zambian scenario in 1991. In this event, the electoral contest for the presidency was largely between NARC's Mwai Kibaki and KANU's Uhuru Kenyatta. The opposition's Mwai Kibaki garnered an impressive victory of 62 percent of the votes cast, compared to KANU's Uhuru Kenyatta's 31 percent. The other three fringe candidates had litde impact on the electoral outcome. It should be noted, however, that the scenario presented in table 6 is somewhat inaccurate because NARC, unlike the other four parties, is not a single political party, but a coalition of parties. First, Mwai Kibaki's Democratic Party (DP), Michael Wamalwa's Forum for Restoration of Democracy in Kenya (FORD-K), and Charity Ngilu's National Party of Kenya (NPK) joined forces to create the National Alliance (Party) of Kenya (NAK) to which each of the parties was a corporate member. Then, following the fallout in KANU, diose who defected took over LDP and linked up with NAK, creating the new grand coalition of NARC. All the named individuals-Mwai Kibaki, Michael Wamalwa, Charity Ngilu, and Raila Odinga-contested the 1997 presidential election. Obviously, had each led his or her respective party to contest the 2002 elections, they would have splintered the opposition and the KANU candidate would have emerged victorious with his 31 percent of the total valid votes cast.
With regard to parliamentary elections, the 2002 Kenyan election illustrates some more modest disproportionality compared to the Zambian case (see table 4). In diis case, as shown in Table 7, NARC garnered 46 percent of the total valid votes cast in the parliamentary election but received over 59 percent of the seats in the National Assembly, demonstrating a disproportionality index of fourteen in favor of NARC. The disproportionality indices for KANU, FORD-P, and Safina are within acceptable limits. Three other parties-FORD-A, Sisi Kwa Sisi, and Shirikisho-roughly balanced the proportion of their total votes with the proportion of their seats. Nonetheless, twenty-five other minor political parties cumulatively received 12 percent of the total votes cast but none of them secured representation in parliament. The significance of this is that the SMD-FPTP electoral system tends to encourage the proliferation of minor political parties that have the consequence of scattering votes but have little real chance of earning representation in the National Assembly. What is interesting from table 7 is the shift in position in regard to the beneficiaries of disproportionality. Whereas in 1997 and 1992 it was KANU that benefited, in 2002, it was the opposition coalition, NARC. The implication here is that the disproportional nature of the SDM-FPTP electoral system naturally favors the dominant party of the moment.
Consequences of Transition without Transformation
As observed at the beginning of this article, in countries like Benin, Mali, Malawi, and South Africa, political transition was preceded by an all-encompassing renegotiation of the rules of the game and restructuring of political institutions. First, incumbent regimes were removed from complete control of tins process and the strategic considerations of the politically minded were thus shaped by the uncertainty of future political contests. This facilitated broad agreement on new rules of the political game in these countries. By contrast, Kenya and Zambia, among odier countries, went through political transition to multipartyism without broad-based renegotiation of the rules of the game. In these countries, any subsequent efforts at constitution-making were either skewed in favor of the new incumbent regimes, as was the case in Chiluba's Zambia, or stonewalled and torpedoed altogether, as happened in Kenya under Kibaki.
Second, in spite of their reform campaign agendas, both the MMD and NARC did not undertake state restructuring to reflect the interests of nonbourgeois forces (Nyang'oro 1997:115). The new regimes retained in large measure the ministers, institutions, and policies of the previous regimes and even resorted to authoritarian tactics to muzzle opposition and critical viewpoints. For instance, under Chiluba, NGOs, journalists, and public figures critical of the government or its leaders were regularly accused of being "un-Zambian" or "fronts for foreign interests." Several political opponents were deported or threatened with expulsion by immigration authorities following government allegations that they were not, in fact, true Zambians (Bratton & Posner 1999:394). Chiluba went as far as declaring a state of emergency in 1993 when UNIP threatened to make the country ungovernable. Further, the 1993 Criminal Procedure Code (Amendment) Act was used to deny bail to individuals accused of certain crimes. The government used the National Assembly speaker, an MMD nominee, to sanction critical parliamentarians and even, in one incident, to imprison journalists who criticized government spokespersons. When the Zambia Independent Monitoring Team and the Committee for a Clean Campaign called a press conference to announce their opinion that the 1996 elections had been flawed, their offices were raided, their bank accounts were frozen, and their chairmen were detained. President Chiluba threatened diese groups further by announcing his intentions to more closely scrutinize NGOs that received foreign funds.
Similarly, the NARC regime in Kenya increasingly became intolerant of freedom of association and the free media became critical of its mode of governance. Following the stalling of the constitutional review process, a pressure group named Bomas Katiba Watch was constituted to pressure for the completion of the process. In a manner reminiscent of the old KANU days, the NARC government banned the pressure group's public rallies in July 2004 and unleashed police violence against Kenyans demonstrating peacefully for the enactment of a new constitution in Kisumu city. In addition, journalists writing reports critical of the government and some of its key cabinet members were arrested, detained, and arraigned in court under repressive laws that had rarely been applied even during the singleparty days. The targeting of the Standard journalists and editors and of the Citizen Weekly publisher, Tom Alwak, led a seasoned columnist and Standard executive editor to assert, "I have been around for a while, I have seen plenty but I feel there is a deliberate plan to intimidate and browbeat journalists and editors and the pressure to do so is coming from some ministers. I think that is dangerous and anti-democratic" (Ombati 2005). Indeed, one cabinet minister, Musikari Kombo, felt obliged to defend the media. He protested what he termed "suppression of press freedom" in Kenya, arguing diat journalists should be left alone to pursue their fundamental duties of reporting and commenting on issues of national importance. He went further to praise the press for exposing corrupt activities in the country (The Standard, Nairobi, Jan.18, 2005).
Third, abuse of office and conflict of interest on the part of the transition governments in Kenya and Zambia illustrate the paradox of transition without transformation of the strategic environment of political engagement. In the case of Chiluba's Zambia, as a result of implementation of the Ministerial Code of Conduct following concerted donor and local pressure, evidence of abuse of office emerged in 1997. Under the code, ministers and members of parliament were forced to declare their assets and liabilities. Some ministers, among them Finance Minister Ronald Penza and Works and Supply Minister Kelly Walubita, declared as their assets, among other things, the numerous contracts they had entered into with government ministries and departments. This was a clear case of abuse of office. As former President Kaunda observed, "While a number of current leaders' businesses were virtually folding before coming to power and they were very poor, their businesses had since flourished and the politicians had become instant billionaires" (Chikulo 2000:165). In particular, Finance Minister Penza, who had declared that his companies had more than twenty-five contracts with government institutions, was labeled a "thief in parliament by bodi MMD and opposition MPs and was singled out as the most corrupt minister in the cabinet. As the Post noted:
A country in which ministers are the most prosperous businessmen surely is a strange one. This is a recipe for corruption It is a negative and extremely pernicious phenomenon which is corrupting politics in our country and undermining the ideal of service among those charged with the responsibility of managing the affairs of state and government. This clearly corrupt way of doing business with government should be eradicated with the same urgency one fights plagues. (Quoted in Chikulo 2000:166)
With regard to Kenya, one such case of abuse of office and conflict of interest is the award of a KSh45 million (USf 600,000) insurance contract to a company run by the son of a cabinet minister under whose portfolio the contract fell (Irungu 2003). The others involve the Ministers for Finance, David Mwiraria, and the Minister for Cooperative Development, Peter Njeru Ndwiga, respectively. According to Mutiga and Makali (2005), Kinondo Holdings Limited, a family enterprise owned by Minister Ndwiga, acquired a prime parcel of land in a horticultural farming zone in Embu, measuring 73.5 hectares (182 acres) in December 2004. Finance Minister Mwiraria, who is in charge of revenue collection, waived a mandatory stamp duty for Ndwiga in the amount of KSh6 million (US$80,000) under circumstances that suggested abuse of power. What is more, the title for the parcel of land was transferred to Kinondo Holdings on New Year's Eve, the same day the farm was used to guarantee a loan in the amount of KSh40 million (US$533,330) from the Cooperative Bank. Interestingly, the bank falls under Ndwiga's ministry. As if this were not enough, another Ndwiga company, secular Insurance Brokers was awarded a KSh6.5 million insurance services tender by the Kenya Cooperative Creameries (KCC) despite the fact that it quoted higher than some firms that also bid for the contract. Again, the cash-strapped KCC is a public dairy company that falls under Minister Ndwiga's docket (Mutua 2005). Ndwiga's case is emblematic of the kind of abuse of office that has become the modus operandi of the NARC regime, facilitated by the all-powerful institutions inherited from the previous KANU regime.
Fourth, the clearest manifestation of the fact that the change of regime in Zambia did not mark a departure from the established mode of political behavior was the incidence of outright corruption. If corruption was rampant under Kaunda, as Ihovbere (1996a) shows, it became even more endemic under Chiluba. Bornwell Chikulo (2000:164-71) writes that contrary to general expectations that political and economic liberalization would reduce the level of corruption, under the Chiluba regime drug trafficking and corruption became not just widespread, but systemic, especially among cabinet ministers and other officials, as the following cases illustrate. Three prominent MMD ministers singled out as being heavily involved in international drug trafficking were Foreign Affairs Minister Vernon Mwaanga, Community Development Minister Miriam Nakatindi Wina, and her husband, Deputy Speaker of the National Assembly Sikota Wina. These three were the most prominent figures but not the only ones. Former Vice President Godfrey Miyanda admitted, "All political parties, including the MMD, have drug dealers" (Chikulo 2000:166). Systemic corruption also involved the bleeding of nascent commercial banks in order to enrich the political elite and their close associates. In the first quarter of 1998 alone, the First Merchant Bank, Manifold Investment Bank, and the Credit Africa Bank collapsed. A couple of banks were suspected of having had links to drug money or laundering activities, but most banks collapsed because of looting by the political elite. As Chikulo (2000:167-68) notes, "Prudence Bank is a classic example of how politicians from the ruling MMD bled the bank to bankruptcy by borrowing at levels they knew the bank simply could not sustain... Ironically, central bank governor, Dominic Mulaisho was sacked from the post for insisting on effectively applying the Banking Act in order to stop the looting of banks and thus protect the interests of depositors."
In similar manner, Kenya under President Kibaki seems to be a replica of Chiluba's Zambia. The transition from Moi/KANU to Kibaki/NARC in Kenya, just like the shift of power from Kaunda/UNIP to Chiluba/MMD in Zambia, did not occasion much change byway of a shift in the mode of politics. In spite of a much publicized war on corruption on the part of the Kibaki government, allegations of corruption against cabinet ministers and senior government officials have caused donors to fear a return to what they brand "bad old habits." Expressing concern about corruption, donors cited such cases as irregular tender awards and the Kiptoon Report on "cowboy contractors" (Irungu, 2003). Similarly, some of President Kibaki's closest advisers have been linked to doubtful dealings involving billions of shillings with shadowy companies. Failure to act decisively on ministers linked to the twin Anglo Leasing and Finance Ltd. scams that cost the government nearly KSh7 billion (US$93,333,300) has cast a shadow on the war on graft. The scams involved the controversial KSh4 billion deal for the construction of a Police Forensic Laboratory for which the government paid Sh241 million (US$3,213,330) without any work having been done. They were compounded by the Kibaki government's decision to expand a KSh800 million (US$10,666,667) passport-issuing equipment contract to one costing KSh2.7 billion (US$36,000,000). The new contract was issued to the shadowy Anglo Leasing and Finance Ltd. without competitive bidding, and KSh.900 million (US$12,000,000) was paid as "commitment fee" (see Irungu, 2003).
Despite the transition to multiparty politics in both Kenya and Zambia, therefore, state apparatuses in both countries are still viewed as the principal instruments for personal enrichment and accumulation by top government officials. The transfer of power from incumbent authoritarian regimes to hitherto opposition parties did not change the mode of politics in both countries. The situation was compounded by the fact that in both cases, the former ruling parties failed to provide credible opposition because of their discredited past and humiliating electoral defeat. The mood of despair in view of this is captured by V. G. Simiyu, who observes in regard to Kenya that
what has annoyed people is that this [NARC] leadership, elected on a democratic platform... [and]... perceived as the torchbearers of democracy. .. have just sunk to the same disease of edinicity, of personality cult, and hardened drunkenness with power. It is fantastic how people can turn around. I know power has no logic, power has no emotions, but it is baffling that... all of a sudden we... [can] see [these people] in their true colors [and] that even the democratic agenda was only a bogey to get to political power. It is really unfortunate. (Interview with V. G. Simiyu, Professor of History, University of Nairobi, Kenya, August 15, 2003)
The foregoing analysis points to two crucial issues in the politics of democratization in Africa. The first is that the design of the electoral system is of crucial importance to the cause of promoting and consolidating democracy. As the evidence demonstrates, the type of electoral system under which elections are held has far-reaching implications for the competitiveness of the political game and for the quality and fairness of representation. African countries that reverted to multiparty politics following a comprehensive review of their constitutions enhanced electoral competitiveness while providing, in the process, circumstances under which incumbent authoritarian regimes could be defeated. In the quest for redesigning the strategic environment of political engagement in democratizing countries, therefore, no exercise is as important as designing an electoral system that enhances fair competition and just representation and, ipso facto, serves the function of conflict mitigation in multiethnic societies.
Second is the paradox of political liberalization without a fundamental restructuring of political processes and institutions to positively reshape the strategic context of governance. In countries like Kenya and Zambia, where this happened, it is arguable that the predatory use to which the elite have turned the state is under no serious threat of deconstruction. Opposition politicians in both countries seem not to have been committed democrats but strategically lurched onto the democratization bandwagon as a way of facilitating their own stab at power. Once in power, they reverted to their true colors. It may thus be argued that the social struggles for democracy across time and place are fundamentally aimed at serving sectional interests, though they may at times, dirough a stroke of luck, serve the interests of the larger society in general. Accordingly, there is a growing sense both in academic and policy circles that the promotion of rapid democratization in Africa at the beginning of the 1990s, with its emphasis on multiparty elections without much attention to electoral system designs, was misguided at best, a fundamental mistake at worst (Harbeson 1999; Reynolds & Sisk 1998:11). However, this may be so in countries such as Kenya and Zambia, where opposition forces squandered the chances for forcing incumbents into renegotiating new rules of the game. Other countries such as Benin, Malawi, Mali, and South Africa, which started off with constitutional reengineering, have brighter prospects for democratic consolidation.
I would like to thank the ASR editors and the three anonymous reviewers for their critical and insightful comments that contributed immensely to reshaping and strengthening the focus and thesis of this paper. The usual caveats apply.
1. Prior to this, the Institutional Revolutionary Party (Partido Revucianario Institucional-PRI) ruled uninterrupted for seventy-one years beginning in 1929, in spite of the fact that elections were regularly held in Mexico during the period.
2. Zambia had operated without the position of vice president since the inception of de jure single-party rule in 1972 when the position was abolished and one of nonexecutive prime minister (appointed by the president) was created.
3. Mauritius was the first country in Africa to go through such an electoral transfer of power when the incumbent regime of Sir seewoosagur Ramgoolam was handed electoral defeat by the opposition in 1982.
4. Anderson Mazoka maintained a lead over Levy Mwanawasa from the time the vote counting began dirough the results from 130 of Zambia's 150 constituencies. It was only with returns from the twenty final constituencies that Mwanawasa opened a 16,298 votes lead, with 456,308 votes against Mazoka's 440,010 votes.
5. The additional five seats for the Rift Valley shown in table 5 were created in the run-up to the 1997 general elections.
Barkan, J. D., et al. 2001. "Designing Better Electoral Systems for Emerging Democracies." www.uiowa.edu/~elecdis/.
Bratton, M., and D. N. Posner. 1999. "A First Look at Second Elections in Africa, with Illustrations from Zambia." In R. Joseph, ed. State, Conflict, and Democracy in Africa. Boulder, Colo.: Lynne Rienner.
Chikulo, B. C. 2000. "Corruption and Accumulation in Zambia." In K. R. Hope and B. C. Chikulo, eds., Corruption and Development in Africa: Lessons from Country Case Studies. London: Macmillan.
Dahl, R. 1982. Dilemmas of Pluralist Democracy. New Haven: Yale University Press.
Donge, J. K. 1995. "Kaunda and Chiluba: Enduring Patterns of Political Culture." In John Wiseman, ed., Democracy and Political Change in Sub-Saharan Africa. London: Routledge.
Foeken, D., and T. Dietz. 2000. "Of Ethnicity, Manipulation and Observation: The 1992 and 1997 Elections in Kenya." In J. Abbink and G. Hesseling, eds., Election Observation and Democratization in Africa. London: Macmillan.
Gitonga, A. K 1987. "The Meaning and Foundations of Democracy." In W. O. Oyugi et al., Democratic Theory and Practice in Africa. Nairobi: Heinemann.
Harbeson, J. W. 1999. "Rethinking Democratic Transitions: Lessons from Eastern and Southern Africa." In Richard Joseph, ed., State, Conflict, and Democracy in Africa. Boulder, Colo.: Lynne Rienner.
Hartman, D. 1999. "Kenya." In D. Nohlen et al., Elections in Africa: A Data Handbook. Oxford: Oxford University Press.
Irungu, G. 2003. "Donors Alarm as Graft Smears Kibaki's Team." Daily Nation, Nairobi, November 27.
Ihonvbere, J. O. 1996. Economic Crisis, Civil Society, and Democratization: The case of Zambia. Trenton, N.J.: Africa World Press.
Kenya Government. 1998. The Constitution of Kenya. Nairobi: Government Printer.
Lijphart, A. 1998. Electoral Systems and Party Systems: A Study of Twenty-Seven Democracies, 1945-1990. New York: Oxford University Press.
Magnusson, B. A. 1999. "Testing Democracy in Benin: Experiments in Institutional Reform." In Richard Joseph, ed., State, Conflict, andDemocracy in Africa. Boulder, Colo.: Lynne Rienner.
Mbaku, J. M. 1997. "Effective Constitutional Discourse as an Important First Step to Democratization in Africa." In E. I. Udogu, ed., Democracy and Democratization in Africa. Leiden: Brill.
Mphaisha, C. J. J. 2000. "The State of Democratization in Zambia." Commonwealth and Comparative Politics 38 (3): 131-46.
Mungai, K 2002. "Distribution of Parliamentary Seat Unfair." East African Standard, December 25.
Murunga, G. R. 1999. "Urban Violence in Kenya's Transition to Pluralist Politics, 1982-1992." Africa Development 23 (1 & 2).
Mutiga, M., and D. Makali. 2005. "Ministers Conspire to Dodge Lad Tax." The Standard, Nairobi, January 16.
Mutua, M. 2005. "Storm as KCC Tender goes to Ndwiga's Firm." The Standard, Nairobi, January 18.
Nasong'o, S. W. Forthcoming. "Negotiating New Rules of the Game: Civil Society and the Kenyan Transition." In G. R. Murunga and S. W. Nasong'o, eds., Kenya: The Struggle for Democracy. London: Zed Books
Nohlen, D., et al. 1999. Elections in Africa: A Data Handbook. Oxford: Oxford University Press.
Nyang'oro, J. 1997. "Critical Notes on Political Liberalization in Africa." In E. I. Udogu, ed., Democracy and Democratization in Africa. Leiden: Brill.
Nyong'o, A. 2002. "Hazards of Transition from Authoritarian Governments." Sunday Nation, Nairobi, March 31.
Ombati, C. 2005. "Standard Editor Grilled over Story." The Standard, Nairobi, January 12.
Reynolds, A., and Timothy C. Sisk. 1998. "Elections and Electoral Systems: Implications for Conflict Management." In T. C. Sisk and A. Reynolds, eds., Elections and Conflict Management in Africa. Washington, D.C.: United States Institute of Peace.
Vengroff, R., and M. Kone. 1995. "Mali: Democracy and Political Change." In John A. Wiseman, ed., Democracy and Political Change in Sub-Saharan Africa. London: Roudedge.
Venter, D. 1995. "Malawi: The Transition to Multiparty Politics." In John A. Wiseman, ed., Democracy and Political Change in Sub-Saharan Africa. London: Routledge.
van de Walle, N. 2003. "Presidentialism and Clientilism in Africa's Emerging Party Systems." The Journal ofModern African Studies 41 (2): 297-321.
van Dijk, R. 2000. "secret Worlds, Democratization and Election Observation in Malawi." In J. Abbink and G. Hesseling, eds., Election Observation and Democratization in Africa. London: Macmillan Press.
Wiseman, J. A. 1995. "The Movement Towards Democracy: Global, Continental, and State Perspectives." In John A. Wiseman, ed., Democracy and Political Change in Sub-Saharan Africa. London: Roudedge.
African Studies Review, Volume 50, Number 1 (April 2007), pp. 83-107
Shadrack Wanjala Nasong'o is an assistant professor of international studies at Rhodes College in Memphis. He teaches courses in African politics, comparative politics, and international relations. His research interests include Africa's political development, the politics of nationalism, and conflict studies. He is the author of Contending Political Paradigms in Africa: Rationality and the Politics of Democratization in Kenya and Zambia (New York: Routledge, 2005) and co-editor of Kenya: The Struggle for Democracy (London: Zed Books, 2007).