JVr.j. palU. id. (1996), Vol. 1 No. 2, 202-216 Structure, Agency and Political Liberalisation in Africa EdmondJ. Keller* Abstract The purpose of this essay is to suggest an empirically based model to be used as a framework for analysis in studying contemporary political transitions in Africa. The discussion is founded on the leading assumption that the factors which catalyse regime transformation are fundamentally the same irrespective of the direction of change: social crisis intersects with structu ral conditions and particu- lar patterns of human relationships resulting in a type of change which is conditioned by political culture and the weight of history. Democratisation is only one form of regime change. The paper concludes that while there may be ample evidence that significant political liberalisation has taken place, it is not appropri- ate to celebrate the "flowering of democracy " per se.for the process is often in the direction of "pacteddemocracy" as opposed to "liberal democracy". For almost a decade, Africa has been involved in what many observers believe is a veritable "democratic revolution." [Diamond, 1988]. Throughout the continent, authoritarian regimes have recently crumbled or been pressured into liberalising their politics as well as their economies. These developments have attracted a great deal of scholarly attention. On the one hand, there are those who have chosen to focus on the concept of democracy itself, as the most appropriate unit of analysis. [Diamond Ibid.; Ronen, 1988; Wiseman 1990; Oyugi 1987; Kpunden 1992]. Others concentrate on various institutions such as the one-party system, types of electoral systems, national elections or popular movements and their relationship to the advancement of democracy. [Meyns and Nabudere 1989; Hayward 1987; Anyang' Nyongo 1987; Reynolds 1994; Horowitz 1991]. Still others focus on democracy only indirectly, preferring to cast their analysis in terms of what they perceive to be a more neutral and all encompassing concept, "governance". [Hyden 1027-0353 © 1996 African Association of Political Science Structure, Agency and Political Liberalisation in Africa 203 Oration 1991;CaiterCenter 1989,1990; Wunsch and Olowu 1990]. What seems to be missing in the burgeoning literature on democracy as well as governance is an appreciation for the dynamics of change; how and why it is initiated and what factors lead to particular outcomes and not to others. The purpose of this essay is to suggest an empirically based model to be used as a framework for analysis in studying contemporary political transitions in Africa. The discussion is founded on the leading assumption that the factors which catalyse regime transformation are fundamentally the same irrespective of the direction of change. Democratisation is only one form of regime transformation. The trend toward democratisation is nothing more than an artifact of the historical moment, and could well be reversed. However, the types of factors that stimulate change in one direction or the other are fundamentally the same: social crises intersect with structural conditions and particular patterns of human relationships resulting in a type of change which is conditioned by political culture and the weight of history. In other words, at a very fundamental level the same types of factors influence change from authoritarianism to democracy as the reverse. This can be shown in a descriptive model that represents a framework for analysing regime transfonnation. Before I tlesh out that model, let me turn briefly to some definitional issues. Key Concepts Four concepts would be useful in interpreting this framework: liberalisation; democratisation; democracy; and civil society. Liberalisation Scholars generally agree that normally the route to democracy is a gradual, staged process, rather than being abrupt and dramatic. However, there is no number of stages in the process that are commonly agreed upon. Schmitter and O'Donnell suggest two broad phases leading lo democratic outcomes: liberalisation and democratisation. The movement from authoritarianism to liberalisation to democ- racy is punctuated by different benchmarks in the transition process. Schmitter and O'Donnell refer to transition as the interval between one political regime and another. [O'Donnell and Schmiiter 1989: 6|. The onset of the transition from authoritarianism is highly uncertain, hut a master symptom is when the incumbent regime appears to be weakening or crumbling and its ideology and policies are being seriously called into question. Leaders themselves may set this process in motion by engaging in liberalisation policies. [Dratton 1994; Przeworski 1988). Liberalisation refers to a combination of loosened restrictions and expanded individual and group rights. When transition toward democratisalion sets in, it triggers a number of often unintended consequences (hat dictate the pattern and extent of change. This libcndisiiig trend can initially be manifested in the economic 204 Edmond J. Keller ooHtaU xan or a combination of both. For example, religious organisations andlabor onions were instnimenial in catalysing change in such disparate places as Algeria, Kenya, Madagascar, Zambia and South Africa. Democratualion involves more man the mere extension of political rights. It also include* social and economic dimensions that operate according to democratic procedures. The implementation of liberalisation policies can set the stage for danoaatisation. For example, after the release of Nelson Mandela from prison in 1990 me South African political system began to exhibit signs of both economic ••d political liberalisation while talks proceeded on the establishment of a deamtacy tat guaranteed citizensbiprights for all.Most often, then, liberalisation cones before democratisation; however, it is not uncommon for the two stages to o v a t e. 1ms is clearly what happened in 1991 in Guinea-Bissau and Benin.1 In both of these countries, political rights and freedoms were extended even as coutitntional reforms were being made for the implementation of a competitive party system, based on the principles of liberal democracy. Atafundamental level democratisation generally refers to the developmentof more egalitarian social relations and the elimination of autocratic authority structures. The stale lessens its economic involvement, and economic policies that are more respectful of workers are introduced. This process is facilitated when grassroots, non-governmental, non-partisan, social, economic and political asso- ciations are allowed to flourish. As a consequence, citizens acquire the habits of democratic participation, and democratic leadership styles also develop. Such habits can be translated to the national political arena. The development of a widespread habiluatkn toward democratic values and procedures is essential if democracy is to stabilise at the national level. [Rustow 1970; Horowitz opdt.i2S2]. Democracy How democracy is conceptualised varies from situation to situation; nowhere is this more true than Africa. However, in almost all circumstances it is conceived to involve the guarantee of social justice, governmental accountability and human freedom. Former Nigerian bead of state, OlusegunObasanjo, has cogently spelt out what be believes to be the minimum standards of democracy, "Periodic election of political leadership through the secret ballot; popular participation of all adults in the election process; choice of programmes and personalities in the elections; an orderly succession; openness of the society, an independent judiciary; freedom of the press to include freedom of ownership; institutional pluralism; a democratic culture and democratic spirit; and fundamental human rights." 1989:34]. Structure, Agency and Political Liberalisation in Africa 205 For the purposes of this discussion, when the term "democracy" is employed, it refers to a form of liberal democracy. This involves tbe procedural minimum of contestation over political office as well as policy choices; popular participation in elections and other aspects of political decision-making; and the accountability of elected public officials according to the rule of law. All of this must take place within a culture in which fundamental human rights and political freedoms are guaranteed. To this inventory should be added military accountability to civilian authority. This is a requirement that has been suggested by Terry Karl based upon her work in Latin America, anil it seems apt to Africa as well. [Karl 1990: 2]. In Africa, the military has demonstrated that it is not adverse to stepping into tbe political arena whenever the politicians "mess up." Liberal democracy can be conceived as an ideal type, found in its purest form not even in one society. Political systems may more or less approximate this idealised form. Moreover, democracy in practice tends to have its shortcomings. For example, in most liberal democracies common citizens are far removed from the levers of governmental power, and in reality policy decisions are the domain of a select group ofelites. Also, in situations where there exist a majoritarian form of democracy, it is not uncommon for the rights of minorities to be ignored. In order for liberal democracy to approximate its ideal, political elites must be committed to more than the achievement of the procedural minimum. They also must be committed to social justice and the upholding of human and political rights for all. Noting what he contends are the limits of liberal democracy, Sklar describes liberal democracy as "democracy with tears"; and, he favors a hybrid, developmental democracy, "democracy without tears" composed of democratic procedures and a strong commitment to equality and social justice. [Sklar 1989: 27]. Civil Society The term civil society is often used to refer to autonomous organised groups bent on challenging authoritarian regimes to open up the political system. In a seminal article on this subject, Jean-Fnincois Bayart defines civil society as the political space between the household and the stale. [Bayart 1989]. It is outside the formal political arena, but it can be drawn in when there exist a political crisis. However, as Callaghy has noted, " . .. new or rcinvigoratcd autonomous voluntary associa- tions and socio-political movements do not necessarily a civil society make." [Callaghy 1994:236]. Civil society is a sub-set of society writ-large. What defines it is its agenda. Civil society is created when autonomous associations adopt and act upona civic agenda. These groups may not have been born as civic organisations, but they are moved by circumstances to engage in politics. For example, both the National Christian Council of Kenya and the Uganda Joint Christian Council are primarily church-based organisations, but over the past decade they have devel- oped political wings that make them significant players in the national politics ol 206 EdmondJ.Keller those countries. Tbe groups that comprise civil society are usually intellectuals, artists, profes- sional such as tawymaml doctors, organised labor, cburcb associations, women's and student associations. During tbe drive for independence, such groups provided the support base for nationalist parties. Crawford Young suggests that based upon the vitality of associational life in political matters, the nationalist period could be considered the "golden age" in tbe evolution of Africa's civil society. [Young 1994: 38]. After independence, autonomous civic associations were either co- opted by mainstream political organisations or repressed by autocratic regimes, and forced to bide their time, waiting for openings in tbe political opportunity structure. Sidney Tarrow has noted that rational people do not confront strong opponents when they perceive opportunities for their success to be minimal. [Tarrow 1995: 86]. However, when the risk of collective action appears to have been significantly reduced, social movements such as are represented in civil society emerge spontaneously. This may occur when political leaders choose to voluntarily open up tbe political system or when they are forced by circumstances to do so; when shifts begin to occur in elite alignments; when an incumbent autocratic regime begins lo implode; or when external pressures are applied on incumbent regimes to open up tbe political and/or economic systems. • The first signs of a resurgent civil society began to appear in Africa at about the time of the overthrow of tbe Jaffar Nimeiri regime in a popular uprising in 1985, but it was not until about 1988 that there were clear manifestations of a genuine social movement. Since then, African civil society has not only grown, it has also changed, become emboldened, and focused on the spoils of national politics. In many cases, it has been tbe decisive catalyst in regime change. [Bratton 1994:51]. Many observers claim that a vibrant and mobilised civil society is the key to the promotion of democracy in Africa. However, a few words of caution are in order. First, rarely is civil society a coherent and cohesive mass movement with a clear sense of its identity and whose members share a common sense of their objectives. Instead, civil society is often comprised of a lose collection of groups with a vaguely defined common objective that often amounts to no more than a desire to oust corrupt or incompetent political regimes. Second, and related to this is that civil society is ephemeral. It is brought into existence by political crisis, and co- opted by more institutionalised political forces such as old-line politicians opposed to incumbent regimes. Bratton argues that: ... tbe role of civil society in political transition is circumscribed toa short- lived interlude: from the time immediately before the "opening" to the convocation of competitive elections. It is during this period, which may last months rather than years, that civil society is ascendant, in the sense that civic political actors are taking the initiatives that are driving forward Structure, Agency and Political Liberalisation in Africa 207 political transition. [Ibid.: 64]. Good examples of this could be seen in the case of Zambia and South Africa where in the late 1980 and early 1990' s a vibrant civil society was clearly manifest. However, it faded from the scene as soon as its issues were tak#« op by formal political parties. The pattern has consistently been for civil society to retreat into hiding once victory has been secured or when defeat is certain, only coming out again when another crisis occurs that seems unmanageable through normal political institutions. All of this notwithstanding, in order to understand the dynamics of the changes now taking place in Africa, civil society has to be factored into the equation. Given the right circumstances, civil society in motion can be decisive in bringing about regime transformation. Toward a Model of Regime Transformation in Africa There is no specific theory of regime transformation available to help us under- stand the present process of change in Africa. In a large measure this is due to the fact that the process is highly uncertain. At (he same time, however, it would seem possible, based on broad comparative evidence, to construct an analytical model that illuminates what is now happening in Africa. Structure and Change The change from authoritarian rule towards democracy is not normally abrupt and even. It is not like the "Big Bang" that instantaneously created the universe. Instead, the process generally takes place in incremental stages. Objective condi- tions and structural relationships at a particular point in time, along with a nation's particular political culture, set the parameters of change. [Preworski op tit:. 48]. These conditions might be found in the domestic or international environment or both. For example, the position of African countries in the world economic system has been declining since the oil crisis of the early 1970s, and over the years, rather than abate, the continent-wide economic crisis has deepened. The high price of petroleum, coupled with declining commodity prices for the items produced in many of Africa's mono-crop economies, has had disastrous consequences. Unem- ployment, rising consumer prices and drastic declines in the purchasing power of the average citizen are some of the manifestations of the domestic economic crisis of governance. Since the independence period, there have been two distinct phases of regime transformation. The first occurred in the immediate post-independence period when ascendant political leaders tended to turn to authoritarian practices in an effort to enhance their political control and governmental effectiveness. Many of them argued that long-term development could only be achieved if society was disciplined. In an effort to impose this discipline, some used raw force in the form of military rule; others governed through de jure or de facto single party mecha- 208 EdmondJ.Keller T «rt-r case, ideology was often used to rationalise statist rule. For • farms of African socialism were used to justify the one-party ^i populace to be willing to sacrifice in the interest of ^ O ^ t t ! e t o r S ^ ^ es of the independence era, most African leaders ««n«l more concerned with asserting their power, authority and hegemony over S u b j e c ts than with enhancing their legitimacy. In a context where there were fa.TnoUtical and economic resources that could allow leaders to purchase legiti- n*cvthe tendency was for them to try and establish the security, control and autonomy of their regimes. Consequently, in many places, politics as well as economics came to be tightly controlled by the state, and the state class - ruling nolitical authorities, the central bureaucracy and its regional funcuonaries, the top echelons of the military and members of the dominant political party, where it exists was shielded from popular demands for public accountability. In this situation, corruption and prebendalism became the order of the day. Politics became more like economics, as the surest route to economic power was political power. A common practice was for potential opponents of the state to be co-opted into the state class and allowed to enrich themselves through public office. Consequently, African public bureaucracies grew much larger than was necessary, and bureaucratic corruption and inefficiency predominated. Over the last decade, this pattern has caused strains in the relationships between African governments and bilateral and multi-lateral aid agencies. For example, the World Bank and the IMF have exerted considerable pressure on prospective aid recipient countries to trim the size of their public bureaucracies, and to abandon their statist economic strategies. [Gordon 1993: 90-129]. Efforts to implement structural adjustment measures have caused further strains in the domestic economies of African countries, and this has had a ripple effect in many places, exposing the political failings of regimes. In some places, strains grew into crises, and this provided at least one spark to ignite popular discontent, and calls for democraiisaiion. I shall return to this point below. Authoritarianism only masked the weakness of African regimes that did not have the capacities to adequately cope wilh the uncertainties of ch;inge, and lacked the moral authority to ensure, on a consistent basis, citizen compliance with rules, regulations and policies. Structural constraints were manifested in such ever-present threats to political stability as politicised ethnicity, regionalism, sectarianism, persistent poverty, undcrdcvelopmem, inequality and class conflict, all combined wilh exogenous structural factors such as world-wide economic conditions, shifts in the relationships between Structure, Agency and Political Liberalisation in Africa 209 donor and recipient countries, the end of the Cold War, and the contagious effects of changes taking place in other parts of the world, to create favourable structural conditions for the most recent wave of political transitions. Agency and Change As important as structural variables are, they alone cannot trigger the process of regime change. Human actions are essential before change is actually initiated. This might involve governing or non-governing elites, or it might involve factions within the regime as well as groups outside of it What is crucial is the perception on the part of individuals and groups about the opportunities and constraints presented by particular structural patterns. In other words, how individuals and groups perceive the opportunity structures they confront at a given historical moment serves to catalyse the change process. In some cases incidents of mass social mobilisation have a contagious effect For instance, groups in one country might simply be inspired by the successful political mobilisation of similar groups in other countries. Most often, however, groups become emboldened when they perceive that the risks of collective action are less than they had been in the past, and at the same time, they feel that prospect for such activities yielding adesired outcome are improved. For example, when the incumbent regime loses its cohesion or is otherwise weakened, this may. send a signal to potential opponents of tberegime that diey can form an alliance with "soft- liners" in the government, and thereby undermine authoritarian rule. Similar developments in 1990-91 led to the marginalisation and eventual ouster of the Kerekou regime in Benin, and the Sasso-Nguesso regime in the Congo; and, to toe dramatic, if temporary, loss of authority around the same time of the regime of Mobutu Sese Seko in Zaire. Under such circumstances, clear vision and good timing are everything. Should opposition groups be disorganised or misread the situation, their movements could fail to achieve their objectives. Opposition groups must not only be well-organised, and focused, they must correctly perceive when the opportunity structure is open. For example, in the summer of 1991, the Islamic Salvation Front [FIS] in Algeria, led by Abassi Madani and his deputy Ali Belhadj, perceived itself strong enough to force the hand of the government of President Chadli Benjedid, which had on its own initiated a liberalisation process in 1988. The FIS accused the government of attempting to rig upcoming multi-party parliamentary elections, and took to the streets in violent protest The regime, however, was determined to take a cautious and measured approach to political liberalisation, and when the protest turned violent, the government was quick to suppress the movement and to deal with Us leaders harshly. Moderates within the FIS accused Madani and Belhadj of having unnecessarily turned to violence. [Ibrahim 1991]. 210 EdmondJ.Keller In Kenya in 1990, calls for tbe reintroductjon of a multi-party system not only by disgruntled political opponents of President Daniel arap Moi, but also by orgtnised groups such as lawyers, students, intellectu;ils. and church leaders was greeted with political murders, unlawful imprisonments, and only a promise to try and make the only legal party, the Kenya African National Union, work better. [Hiltzik 1990]. Within less than a year, however, a combination of pressures from within and without forced Moi to agree to a return to multiparty ism. The Paris Club came to require political liberalisation as a precondition for foreign assistance, and tbe United States Ambassador, Smith Hempstone, led a personal crusade in support of Kenya's civil society. The result was multi-party elections in December 1992. However, tbe opposition proved to be too divided to be able to signal the death knell of the Moi regime. Accordingly, the regime could continue to structure political outcomes in its favour. By contrast, in Ghana [1969 and 1979], Sudan [1964 and 1985], and Benin [1990-91], popular movements and the recognition on the part of the political leadership that they were incompetent as managers, forced authoritarian regimes out and replaced them with civilian regimes through multi-party elections. [Chazan 1988: 102; Woodward 1989; Neavoll 1991: 30-42]. Authoritarian regimes do not have to possess legitimacy to survive. They merely have to be cohesive, and able to rely upon the loyalty of a capable and subservient military. When a regime becomes weakened for whatever reason, it has the option of either considering liberalising its policies or "digging in its heels," and responding to popular protest with repression. In other words, the perceptions and actions of governing elites are crucial in determining the rate and pattern of political transition. Indeed, statecraft could be used to preempt or neutralise even a highly mobilised civil society. For example, in Ghana, Head of State, Flight Lieutenant Jerry Rawlings in the early 1990's resisted calls for a rapid return io mulli-partyism, insisting that he would noi be forced intoa chaotic democralisation process at the expense of Ghana's resurgent economic progress. He was deter- mined to have a say as to how Ghana once again traversed the road to democracy/ In such divergent places as Guinea-Bissau, Madagascar, Gabon, Burkina laso, Angola and Mozambique, political leaders in the early 1990's attempted, with varying degrees of success, to rush to the head of the democratisaiion movement, in the hope of not only preserving their own place in history, but also in an effort to protect their own class interest. Huntington has argued that one of the defining features of what he calls The Third Wave of Democratisaiion [Huiuinglon 1991: 85-100) is the profound influence of governments and institutions external to a given country in support of democralisation. Not only wasextemal pressure instrumental in pushing the Soviet Union to liberalise its politics; Huntington also credits the U.S. with a key role in dismantling apartheid in South Africa. Despite the decisive potential of pressures Structure, Agency and Political Liberalisation in Africa 211 for democratisation emanating from outside a country, such pressures will only have the desired effect when the external actor has some leverage over the country in question. As is demonstrated by the ability of Mobutu in Zaire, Eyadema in Togo, Bashir in Sudan, and Abacha in Nigeria, to cling to power in spite of considerable pressures being exercised by the international community, without real leverage external pressures will not matter all that much. As long as deter- mined autocrats are able to maintain their core base of support and the internal cohesion of the regime they are likely to be able to resist or finesse those who wish for a speedy and complete turn to democracy. For example, despite having secured multi-party elections, opposition parties, even with support from influential external actors, have been unable to guarantee democracy for Kenya. In November 1994, the government of President Omar Bongo and Gabonese opposition groups signed, with the blessing of the French government, the "Paris Accords." This was an agreement which, in principle, committed Bongo to establishing an independent electoral council to prepare the way for democratic elections. However, as a delaying tactic, Bongo proposed a referendum on the Accords. Critics contended that no referendum was needed since me President can enact the Accords by decree. Since 1990, Bongo has been employing his consid- erable political acumen to delay democratisation while publicly professing to be its staunchest advocate. Civil society, or for that matter mainstream opposition parties, are unlikely to successfully challenge autocratic regimes unless the regime is weak or divided against itself. When an authoritarian regime is weak, divided within itself, or both, it is vulnerable either to successful popular uprising, or to violent revolutionary change. In Benin, the Congo, and Zambia, for example, popular movements forced authoritarian regimes that eschewed excessive repression, to open up their political systems, agree ultimately to constitutional reforms, and to submit to multiparty elections. On the other hand, the authoritarian regimes of Samuel K. Doe in Liberia, Obote and Okello in Uganda, Siad Barre in Somalia, and Mengistu Haile Mariam in Ethiopia were brought to a violent end by armed opposition groups. The regimes in these countries remained cohesive, and made only token gestures toward liberalisation; and their opponents were left with no alternative but to resort to violence as a form of political opposition. The anti-state struggles of opposition forces in the realm of legitimate politics, given the right circumstances, can result in revolutionary movements. However, it remains to be seen if democracy will grow out of such revolutionary outcomes. The Meaning of Political Transitions There is little doubt that significant political changes are now taking place in Africa. However, the question remains, how deep and how durable are these changes; and, what do they represent? Some observers suggest that it is highly 212 Edmond J.Keller unlikely that this current wave of change will result in the "flowering" of democracy in Africa. The lack of a democratic past is the most common explana- tion given for such an argument At the same time, it seems reasonable to suggest that an equally important explanation might be found in the fact that in the interest of its own survival, the tendency will be for the state class to politically liberalise without being wholeheartedly committed to full democracy. Granting that there are bound to be regional differences, what is more likely in Africa is new varieties of corporatism. Philippe Schmitter has identified two formsof corporatism: societal corporatism and state corporatism. [Schmitter 1974]. Societal corporatism is a pattern of institutional relationships in which the officially sanctioned sectoral interest organisation, while collaborating with each other and state policy-makers in pursuit of a commonly accepted national interest, quite autonomously represent the interest of their membership. Their primary responsibility is the representation of the interests of their membership through their dealings with the state. On the other hand, state corporatism is an institutional arrangement in which the state seeks to co-opt or control major sectoral interest organisations, usually by estab- lishing rules that govern their very creation as well as their behavior. Tendencies toward state corporatism exist in Africa; and what is likely to result in years to come is a particularly African variant of this organisational form. It will most likely be closer to the state corporatism found in parts of Latin America than to the societal corporatism that predominates in Europe. However, given the vigor of civil society in some parts of Africa, it is quite conceivable that a form of societal corporatism could become the rule in some places. ,To the extent that progress toward democracy might be made in Africa, it is likely to be grounded in the formation of political pacts. Political pacts usually involve formal agreements among most if not all competing political elites and their organisations over the rules of the political game, as well as over rules relating to the distribution and redistribution of material benefits. Terry Karl has described this process as "democracy through undemocratic means," in that it involves elite compromises instead of genuine popular involvement in determining the polices of elected officials. [Karl op cit.: 11]. However, the only place on the African continent where this pattern appears to be somewhat established is the Indian Ocean micro-state of Mauritius and South Africa Between 1990 and 1994, political elites, representing all significant political organisations engaged in a process of negotiations that resulted in a power-sharing formula for governing South Africa overafive year period whileanewconstitulion was formulated. What was remarkable about this process was that the leaders of the various parties all agreed to some contingent compromises in order to move the process along. What resulted was the surprisingly easy dismantling of apartheid. The potential for the formation of a type of pacled democracy was also evident Structure, Agency and Political Liberalisation in Africa 213 in Ethiopia, where a newly installed revolutionary government, headed by the Ethiopian People's Revolutionary Democratic Front [EPRDF], initially in 1991 attempted to include almost all relevant political interests and ethnic groups. Politics proceeded according to rules accepted by all constituent parties at a charter conference, and there was at least tbe theoretical possibility of the main executive positions circulating among the parties. The signatories to the Charter agreed to a two-year transitional period, to be followed by the inauguration of Ethiopia's first democratic constitution, and a multi-party electoral system. However, within a year, tbe coalition had narrowed so much that it now included only the EPRDF and other parties that supported its program and parties that had been created by the EPRDF as a counter-weight to various ethnically based parties. Despite tbe fact that tbe transitional period had to be extended for two additional years, the EPRDF regime forged ahead according to its own design to lay the foundation for what it claimed would be a new, democratic Ethiopia. A new constitution was enacted in 1994, and a year later, national elections were held, establishing the Federal Democratic Republic of Ethiopia. Most major opposition groups boycotted the elections, but tbe new regime seemed benton implementing procedural democracy even if this meant a less than perfect pact. Whether the trend is toward democracy or toward authoritarian retrenchment, it is likely that Africa will witness in the near future growing tendencies toward some form of corporatism. Democratically inclined regimes might use this approach to state-society relations to enhance their authority in the face of vocal opposition. For example, in 1995 President Nelson Mandela of South Africa systematically attempted to enlist the cooperation of traditional chiefs, particularly in the troubled Zulu areas, so as to undercut the base of support of one of his main opponent, Chief Mangosuthu Buthelezi. Mandela utilised the considerable re- sources of tbe state treasury to provide tbe traditional chiefs with perks of office that B uthelezi was unable to provide. In this way he was cultivating loyalty towards tbe national government and away from tbe separatist tendencies of Buthelezi. Rather than creating new corporate groups, Mandela merely took advantage of significant ones that already existed. By contrast, in Ethiopia, the EPRDF seemed compelled to create new corporate groups with which to cooperate. Conclusion In the short run, if the current wave of political transition in Africa is to result in some form of democratic outcome, it seems that it will, in most places, have to go through the channels of corporatism and political pacts. At the same time, no matter how limited the gains of political liberalisation at this time may be, authoritarian 214 EdmondJ. Keller regimes are gradually being weakened; and this represents astep in the democratic direction. Notes * 1. Director, J.S Coheman African Studies Centre, California, USA. "Guinea-Bissau: New Paths to Old Goals," Africa News Vol. 34, No. 8 (May 20,1991); and Vivian Lowery Derryck, "Benin: The Velvet Revolution," Africa Report Vol. 36, No. 1 (January-February 1991). In general usage, the term "prebend" refers to offices of feudal states that could be obtained in recognition of services rendered to a noble person, or through outright purchase, and then utilised to generate income from the bolder of such offices. In the African context, it refers to patterns of political behavior based on the assumpiion that the offices of the existing state may be competed for and then used lor the personal benefit of the officeholders and their supporters. This condition is characterised by the intense struggle among various segments of society to control and exploit offices of the state. Graft and corruption are part and parcel of everyday political lifeat all levels. See, Richard Joseph (1987) Democracy and Prehendal Politics in Nigeria: The Rise and Fall of the SecondRepublic. Cambridge: ( amhridpc I Iniversity Press. "Interview - Flt.-Lt. Jerry Rawlinjis: Construct ing a New Constitutional Order, "Africa Report (May-June 1WI). 2. 3. Structure, Agency and Political Liberalisation in Africa 215 4. "Gabon: Saying Yes and No," Africa Confidential Vol. 36, No. 13, July 23, 1995. References Anyang' Nyang'o, P. (ed.) (1987) Popular Struggles for Democracy in Africa. London: Zed Press. Bayart, J-F. (1988) "Civil Society in Africa," in Patrick Chabal, Political Domina- tion in Africa. Bratton, M. (1994) "Civil Society and Political Transitions in Africa," in John W. Harbeson, Donald Rothchild, and Naomi Chazan, (eds.), Civil Society and The State in Africa. Boulder Lynne Rienner. Callaghy, T. (1994) "Civil Society, Democracy, and Economic Change in Africa: A dissenting opinion about resurgent societies," in John W. Harbeson, et al. (eds.), Civil Society and the State in Africa. Carter Centre (1989) "Beyond Autocracy in Africa," Working Papers for The Inaugural Seminar of The Governance in Africa Program (February 17-18). Carter Centre (1990) Working Papers From The Second Annual Seminar of The African Governance Programme, March 23-25,1990. Chazan, N. (1988) "Ghana", in Larry Diamond, Juan Linz, and Seymour Martin Lipset (eds.), Democracy in Developing Countries, Vol. II: Africa. Boulder Lynne Rienner. Diamond, L. (1988) "Preface," in Larry Diamond, Juan Linz, and Seymour Martin Lipset (eds.). Democracy in Developing Countries, Vol. II: Africa. Boulder Lynne Rienner. Gordon, D. F. (1993) "Debt, Conditionality, and Reform: The international relations of economic policy restructuring in Sub-Saharan Africa," in Tho- mas M. Callaghy and John Ravenhill (eds.). Hemmed in: Responses to Africa's Economic Decline. New York: Columbia University Press. Hayward, F. M. (ed.) (1987) Elections in Independent Africa. Boulder: Westview. Hiltzik, M. A. (1990) "Kenya bans leading dissident," Los Angeles Times, (September 30,1990). Horowitz.D. (1991) A Democratic South Africa? Constitutional Engineering in A Divided Society. Berkeley: University of California Press. Huntington, S. P. (1991) The Third Wave: Democratisation in The Late Twentieth Century. Norman: University of Oklahoma Press. Hyden, G. and Michael, B. (eds.) (1991) Politics and Governance in Africa. Boulder: Lynne Rienner. Ibrahim, Y. M. (1991) "Algeria says 2 opposition chiefs face trial on conspiracy charge," The New York Times, (July 2, 1991). Kpundeh, S. (ed.) (1992) Democratisation in Africa: African Views and African Voices. Washington, D. C: National Academy Press. Karl, Terry L. (1990) "Dilemmas of democraiisation in Latin America," Compara- Meyns, P. and D. W. Nabudcre, (eds.), Democracy and the One-Party State in Africa. Hamburg: Instilut lur Africa-Kunde. Neavoll, G. (1991) "Benin: A victory for democracy," Africa Report, Vol. 36, No 3, (May-June 1991): 39-42. O'Donnell, G. and P. Schmiiter (1989) Transitions From Authoritarian Rule: Tentative Conclusions About Uncertain Democracies. Baltimore: The Johns Hopkins Press. Obasanjo, O. (1989) "Remarks to the conference on the democratic revolution " Proceedings of a Conference Sponsored by The National Endowment for Democracy, May 1-2,1989, Washington, D. C. Oyugi, W. et al. (eds.) (1987) Democratic Theory and Practice in Africa. London: James Carey. 216 EdmondJ.Keller tive Politics, Vol. 23, No. 1. Pizeworski, A. (1988) "Some problems in the study of transitions to democracy " inG. O'Donnell, P. C. Schmitter, and L. Whitehead (eds.), Transitionsfrom Authoritarian Rule: Comparative Perspectives. Baltimore: The Johns Hopkins University Press. Reynolds, A. (ed.) (1994) Election '94: South Africa - The Campaigns, Results and Future Prospects. New York: St. Martin's Press. Rouen, D. (ed) (1988) Democracy and Pluralism in Africa. Boulder: Lynne Rienner. Rustow, D. (1970) Transitions to democracy: Towards a Dynamic Model" Comparative Politics, Vol. 2, No. 3 (April 1970). Schmitter, P.C. (1974) "Still the Century of Corporatism7'/tevj