The history of democratization in Sub-Saharan Africa has demonstrated that competitive elections are always important historical moments-they have ushered in change, rebirth and renewal. Yet while elections are the bedrock events of any democracy, they are also the political institution most open to manipulation. The electoral moment is fraught with fragility. As often as elections in Africa have been positive events, they have also precipitated moments of crisis, which exacerbate ethnic conflict, political breakdown, and related social disequilibrium. To understand whether elections are likely to bring progress or instability, one must consider whether the political sphere is ready for competitive elections and how the elections themselves are designed and operated. The electoral system, or how the votes cast are actually translated into seats, has a huge impact not just on inclusion and exclusion but also on the tone of the entire political system. The system will also craft the space for corruption and vote rigging-it will not eliminate the space for malfeasance, but it can limit it. For these reasons, the crafting of appropriate electoral systems is one of the key factors shaping democratization and political conflict on the continent.
It has become de rigueur to dispute the emphasis that international aid missions and scholars place on elections as the holy grail of democratic transitions in post conflict states. In War, Guns, and Votes, Paul Collier argues that when elections are superficial-corrupt, flawed, and "ethnic"-political violence is heightened and the resulting governance is worse than it would have been if elections did not determine the government at all. Despots are seemingly "legitimized" through stolen elections and substantive participation never penetrates beyond the existing elites.1
In this line of reasoning, blaming elections for causing political conflict is similar to suggesting that in order to avoid getting wet from a leaky roof, one should live outside to stop getting rained on. In reality, the metaphorical roof must simple be constructed in a way that fits the house. Fortunately Collier's broader argument is more sophisticated than stating that all elections in poor places are bad. His analysis is that elections tend to work better in societies that have larger populations and fewer ethnic divisions. They also tend to work better in polities with checks and balances on the power of government, and in particular where the elections are properly conducted. In a sense, he argues a modern version of John Stuart Mill's belief that free institutions are next to impossible in a country made up of different nationalities. And while Collier has the kernels of a point, he fails to build in the crucial variable of institutional design, which my research demonstrates does contribute greatly to whether elections promote conflict or accommodation.
Clearly, one successful election does not prove the establishment of democracy, nor should foreign assistance and attention atrophy soon after the first votes are counted. The Sudanese scholar Francis Deng says it well: "Democracy is a concept that advocates popular participation in the political, economic, social, and cultural life of the country through ongoing, sustained comprehensive reforms and measures, not just at the election polls."2 But while accepting that elections without democracy impoverish and often retard progress, elections themselves remain a focal point for elites and lay the foundation for what comes after, whether good or bad. Elections are the first stage of public involvement in the process of ratifying or crafting the governance element of a peace process. This is partly symbolic, but it is also substantive and the only way popular control can be born, as the state must ask for consent from the people for its rule. The American declaration of independence notes that "legitimate governments derive their just powers from the consent of the governed," but it is also practical in the sense that elections select the leaders who will then share the spoils.3 This is especially meaningful if society and voters are ethnically segmented. The election becomes the guiding principle for a just sharing of the state's resources. It provides a census of not just parties and ideologies but of the component parts of society. Elections legitimize some leaders while delegitimizing others. Therefore, without national elections, there can be no democratization or any chance of an enduring self-sustaining peace process. As Diamond notes, "while free and fair elections are only one component of democracy, they are the most indispensable one."4 O'Donnell also stresses the cathartic moment that elections represent in transitions from authoritarian rule. "This is not because such elections will necessarily lead to wonderful outcomes. It is because these elections...mark a crucial departure from the arbitrariness of authoritarian rule. When some fundamental political freedoms are respected, this means great progress in relation to authoritarian rules and gives us ample reason to defend and promote fair elections."5
Elections are not only transition points; they are repeatable moments that become critical to democratic development. As Bunce points out, democratic breakdowns often happen in reaction to national elections, the repetition of elections seems to enhance democratic quality, and aid for elections is the only area of foreign support that is clearly correlated with improved democratic performance.6 Over time, she says, good elections sustain democratic investment: "People are more likely to express support for democracy when they see it working to provide genuine political competition, including alternation of power, and when it has at least some effect in controlling corruption, limiting abuse of power, and ensuring rule of law."7
When elections in a democratizing state do not work well, the architecture of democracy has no foundation to stand on. In the last decade, Lesotho, Kenya, and Zimbabwe have exemplified three ways in which failed elections have damaged the basis of the democratic system. The Lesotho elections of 1998, despite a bizarre result were not won and lost by fraud, but failed because the electoral system produced an untenable exclusion of all other voices beyond the government. With 61 percent of the popular vote the Lesotho Congress for Democracy won 78 of the 79 seats up for grabs. In Kenya multi-party elections have become increasingly delegitimized since 1992. The long-standing malapportionment of seats to the parliamentary districts, combined with the proliferation of fraudulent practices in Presidential elections, precipitated a break down of law and order and inter-ethnic peace in the aftermath of the disputed 2007 elections.
Zimbabwe is the more stereotypical case where the system has been rigged in a variety of ways to stop the opposition from gaining a foothold, thus making it impossible for them to receive the majority. Initially, the electoral system was used to curtail political opposition: Mugabe switched from proportional representation to first past the post (FPTP) and in the 1990s, ZANU-PF managed to wrap up almost every seat in parliament. However, when the Movement for Democratic Change harnessed urban Shona votes alongside their Ndebele core from 2000 onward, they were able to break through the FPTP barrier. This led the ZANU state apparatus to turn to a more systematic program of political intimidation, harassment, murder and electoral manipulation.
The growing electToral experience in Aafrica
Multi-party elections are a growth business throughout the world, but especially in Africa. Rakner and van de Walle counted 137 legislative elections in 41 African nations over the 18 years after Namibia's independence in 1989.8 Before that only Botswana and Mauritius had regular multi-party elections. Despite Collier's doomsday review, some of these elections have gone well. In Ghana in 2008, elections brought the opposition into power peacefully. The ruling party presidential candidate won 49 percent of the first round votes, only to lose to John Atta Mills of the NDC in the run-off. The administrative quality and institutional legitimacy of this election was very high, scoring at 79/100 using the Elklit-Reynolds election quality assessment methodology.9 After the catastrophic events of 1998, Lesotho ushered in a new electoral system, a new government, and a new calm in 2002. This year's election scored 90/100 on the Elklit-Reynolds scale. South Africa's first democratic elections of 1994 were clearly instrumental in defining future stability, and since then the country has held four elections, with election quality improving from 76 to 85 between 1994 and 2004.
But, as Collier suggests, flawed elections do feed a vicious cycle of instability. It is fairly certain that if there is socio-political chaos after an African election, the legitimacy of the election has been deeply compromised. Zimbabwe's parliamentary election of 2002 scored 44/100-the lowest election ever rated on the Elklit-Reynolds scale, and in 2008 the overall legitimacy of the elections were even worse (yet to be scored using the methodology). Kenya's 2007 parliamentary and presidential elections engendered a devastating ethnic conflict that spread like wild fire across the Rift Valley, and while the degree of outright vote fraud is disputed, these elections scored a disappointing 51/100 on the quality of elections scale overall.
As noted, elections are high stakes games (even when democracy is inchoate) and the rules of the game are the election system. Electoral system design is a crucial variable in democratic stability because it is the fault line where the inclusion of political parties and marginalized communities is either assured or defeated. As Collier notes, "if there are no limits on the power of the winner, the election becomes a matter of life and death. If this life-and-death struggle is not itself subject to rules of conduct, the contestants are driven to extremes."10 But the elections, and indeed politics overall, need not necessarily be winner-take-all, or life-and-death. That political system is an inheritance of British colonial influence and may not necessarily be the only, or correct, way to organize democratic politics in a fragile African state.
The election system is the key driver of power sharing or majoritarian constructs of government. As Norris shows, the core sequential steps of power-sharing theory start with the inclusion of minority and majority groups through election systems that are either proportional or encompass positive action mechanisms. This theory states that this creates incentives for legislative and executive compromise and builds trust between hostile groups. As political stability is strengthened and violent conflict is reduced, community members become more invested in supporting democracy.11
In a globally comparative book I am currently finishing on the role of political institutions, I find that electoral systems are the most consequential of political institutions when it comes to promoting democratic participation. Among a variety of institutional, social, historical and economic factors, electoral systems are significant predictors of state stability (the type of violence, death, riots, assassinations and strikes indicators that Collier uses). Furthermore, they predict the executive constraints and political competition elements of democracy. The more proportional an election system for the national legislature, the more stability and political competition was enhanced, although proportionality was correlated with weakened constraints on the executive. When it comes to the inclusion of marginalized ethnic groups in legislatures, proportional systems (and list proportional representation in particular) include more representatives of minority groups than majoritarian systems. That inclusion then has a positive effect on public policy and levels of social tolerance.
This data confirms and builds upon Pippa Norris' findings that nations using proportional representation (PR) systems are more successful democracies when holding other variables constant. And when the analysis is limited to just plural societies the results are even stronger. My analysis goes even further in associating PR not just with democracy but stability as well. But, as noted above, I do add the rider that PR systems appear to weaken the executive constraint dimension.
There are alternative views of how electoral systems can best promote democracy and stability but all successful electoral system design revolves around a series of core principles: (i) that the system is perceived to be fair (usually because all major interests are represented in parliament); (ii) that it avoids anomalies in the results which would lead to illegitimacy; (iii) that space is created for cross-cutting multi-ethnic parties; (iv) that women and minorities are given fair access to representation; (v) that internal party democratization is promoted; and (vi) the connection between a voter and her representative is maintained in the interests of accountability.
In medical terms, proportional electoral systems are akin to antibiotics since, in the vast majority of cases, they are helpful aids in fighting the social ailments of minority exclusion and inter-ethnic hostility. They tend to contribute in helping to stabilize unhealthy states and can bring a different and more constructive tone to political competition, if used appropriately. But just like antibiotics, PR election systems will not be successful in every case. Sometimes the society is antibiotic resistant, while in other cases the body politic has learned how to counter the effects over time. Indeed, politicians and voters may find new ways of evading the incentives for accommodation and inclusion that proportional systems offer.
Sub-Saharan Africa is a perfect laboratory for the study of the consequences of election systems because the region demonstrates great variety. As of 2009, 18 (or 42 percent) African states use First Past the Post systems, 14 (33 percent) use List Proportional Representation, eight (18 percent) use mixed systems (all of which but Lesotho's are parallel, meaning the district and list elections are unconnected) and three (7 percent) continue to use the French Two Round Systems. Colonial inheritance still dominates the explanation of how African nations came to have their election systems. Only Benin, Lesotho, South Africa, Sierra Leone, Niger, Namibia, and Sudan, significantly depart from the election system that was imposed under colonization, or indeed the election system of their colonizers.
Seven states have changed their system over the last twenty years, with Liberia alternating back and forth between FPTP and list PR. All that changed, bar Liberia in 2005, was that these states moved from a majoritarian system to a more proportional one. In all cases the reforms were driven by conflict, crisis, and a resulting peace settlement. It seems unimaginable that any of these cases could have survived post-conflict with a majoritarian election system, one that put the power of the State solely in the hands of one group or region. Perhaps Liberia will be able to buck the trend of majoritarian failure with its steady and calming new President, Ellen Johnson-Shirleaf, and a polity that is defined more by elite maneuvering than ethnic polarization.
The fragile Sudan
Sudan is hoping to hold national elections in the spring of next year (although they have already been postponed twice).12 These will be an important test of the ability of a power sharing agreement and a new mixed electoral system to promote accommodation and political stability. If there is one state in Africa that does not need increased tension and competition between ethnic groups right now, it is Sudan.
The elections for the national parliament will be held under a mixed (parallel) system of both districts and party lists. This was the product of the negotiated peace settlement between the Sudan People's Liberation Movement (SPLM) of the South and President Omar al-Bashir's Khartoum National Congress Party (NCP) regime in the North. 270 (60 percent) of the seats will be single member districts, with the ten states of the South electing 57 and the fifteen Northern states electing 223. There will be 68 (15 percent) seats awarded from lists on the basis of party votes in the 25 state districts of the country (53 in the North and 15 in the South). The threshold to be entitled to seats under the PR allocation is 4 percent. Another 112 (25 percent) seats are reserved for women, awarded on a party list basis (88 in the North, 24 in the South).
But the reality on the ground makes these seat allocations far less advantageous to the Northerners than it first appears. Nine states within the North are likely to vote overwhelmingly for Northern and Arab parties and candidates (even if not Bashir's NCP), but the other six are almost certain to reject the Northern elites overwhelmingly. The disputed territories of Nuba Mountains, Abyei, and the Blue Nile are geographically within the North but are likely to vote en masse with the South, if, of course, there is anything approaching a free election. The three large states that make up the Darfur region are, to say the least, unlikely to vote in large numbers for the government of Khartoum (but again this assumes the people of Darfur will be able to vote at all in the April elections). Last, the Beja people living in the northeastern state of the Red Sea are the sleeping "Darfur." The paramilitary Beja Congress and Rashaida Free Lions have been skirmishing with the Khartoum government for years and feel just as marginalized from power as the Darfurians ever did.
President Bashir's National Congress Party (NCP) is not only unlikely to do well outside of the nine core states that he controls, but a variety of Arab and Islamic groups may strongly challenge him in his heartland. Bashir has never been tested in a free election and again, if there is any space for free choice, parties and voters may fragment. The NCP has a history of fracturing between moderates and hard liners. A splinter group, the Popular Congress Party of Hassan al-Turabi, could do very well in popular elections. The Umma party of al-Mahdi also has little love for the existing Khartoum regime and has formed a common cause with Southerners in the past (and more recently with the rebel Darfur Justice and Equality Movement [JEM]). Bashir's NCP may well use its considerable carrots and sticks to build alliances with potential competitors in the North, but the success of such a project is far from assured.
In dramatic contrast, the forceful strategy of the SPLM over forty years has delivered them a dominant role throughout the South. Because Nuer leaders like Riek Machar (Vice President of the South) and Shilluk leaders like Pagan Amum (General Secretary of the SPLM) have allied with the majority Dinka ethnic group. The SPLM is the only major player in town for the time being. While the National Assembly seats of the North will be shared among a multitude of players, the SPLM is likely to wipe out all electoral opposition in the South. They will win the FPTP seats and the women's seats and no doubt do well in the disputed territories, large parts of Darfur and in the Southern Sudanese enclaves of Khartoum. The electoral system actually reinforces this one party dominance because there are not enough PR seats in the system to ensure proportionality in the South. Each southern state will have either one or two party list seats and thus any group wishing to challenge the SPLM will need more than a third of the state wide votes to win a seat-a very high bar considering there is no serious opposition to the SPLM in the South.13 However, in the North some states will have many PR seats-Khartoum (9), Southern Darfur (7), Gezira (6)-and in these districts minority parties will be able to steal seats away from the government.
All this is little more than educated guesswork, but it suggests that if elections approximate anything close to a free vote, the SPLM, along with its allies, could be the largest party in the new National Assembly of Sudan, perhaps with as much as 45 percent of the legislature. Not only would they be able to seize independence for the South; they could take control over the entirety of the state. But what would that predict for stability in the Sudan as a whole? Would the Southerners administer the North? One cannot imagine a scenario more likely to lead to renewed bloodshed. Of course, there will be national presidential elections at the same time, which may be tougher for any SPLM candidate to win.14 Even with support from the Beja, Southern Darfurians, and those in the disputed territories, the votes of the South may not be enough to challenge the Arab majority. The Comprehensive Peace Agreement of 2005 does ensure that a national President from the South must have a Northerner as their Vice-President. But even if a Northern candidate does ride a tide of Arab unity to the Presidency, they may be facing a legislature controlled by the other side.
The election system to be used in Sudan in 2010 is not finely tuned enough to respond to the tensions that elections will undoubtedly bring to that nation. But the elections are merely one cog in a wheel of many institutional arrangements that need to work together to stop conflicts being processed by the bullet rather than the ballot. Nevertheless, it has been proven beyond a doubt than authoritarian rule does not lead to stability in Sudan. The roof may still be leaky, but at least now there is a roof to mend.
1. Paul Collier, Wars, Guns, and Votes (New York: Harper, 2009).
2. Francis M. Deng, Identity, Diversity, and Constitutionalism in Africa (Washington D.C.: USIP Press, 2008): 132.
3. As cited in Gerald Pomper, "The Wisdom of the Masses: A Philosophical Case for Elections," Harvard International Review XXIX, no. 1 (Spring 2008): 34-38.
4. Larry Diamond, The Spirit of Democracy: The Struggle to Build Free Societies Throughout the World (New York: Times Books, 2008): 26.
5. Guillermo A. O'Donnell in ed. Thomas Carothers, Critical Mission: Essays on Democracy Promotion (Washington D.C.: Carnegie Endowment for International Peace, 2004): 188.
6. Valerie Bunce, "Reflections on Elections," APSA-CP Newsletter 19, no. 2 (Summer 2008): 1-5.
7. Diamond: 161.
8. Lise Rakner and Nicolas van de Walle, "Opposition Weakness in Africa," Journal of Democracy 20, no. 3 (2009).
9. University of Aarhus Department of Political Science, "Comparative Democratic Assessment: The Quality of Elections and Civil Liberties in the 21st Century," http://www.democracy-assessment.dk/start/page.asp?page=1.
10. Collier: 15.
11. See Pippa Norris's useful flow chart in Driving Democracy, (Cambridge: Cambridge UP, 2008): 105.
12. In fact Sudan is planning to hold six different elections at the same time: the Presidency of Sudan, Presidency of Southern Sudan (GOSS), the National Assembly, the GOSS Assembly, State Governorships, and State Legislatures.
13. In June 2009 the former SPLM nominated Sudanese Foreign Minister, Lam Akol, set up a rival SPLM-DC party but there is scant evidence to show any nascent electoral strength for the new party.
14. To date it is unclear who will contest the elections for national and GOSS Presidential elections on behalf of the SPLM. Would Salva Kiir run at the national level leaving Riek Machar to win the Presidency of the South? Such a scenario appears unlikely. See Jort Hemmer, "Ticking the Box: Elections in Sudan," Clingendael Institute, Netherlands, September 2009.
University of North Carolina at Chapel Hill
Andrew Reynolds is an Associate Professor of Political Science and the Chair of International and Area Studies at the University of North Carolina, Chapel Hill. His work focuses on electoral politics, democratization, and constitutional politics. He is currently completing a book titled, Designing Democracy in a Dangerous World: Applying the Lessons of Medicine to the Art of Constitutional Design.…