Many post-independence rebel movements in Africa have unleashed extremely brutal forms of violence, especially against the peasantry. Such violence, which has bewildered many observers, cannot be explained by reference to African ‘culture’, nor as an expression of rational self-interest. Instead, it must be seen in the light of the essentially urban issues that have fomented rebellion, which cannot however be successfully pursued in major towns, where incumbent regimes possess a monopoly of force. Retreating to the countryside, however, rebels can rarely swim among the peasantry like Mao's fishes in the sea. The African rural setting is generally deeply inimical to liberation war, because peasants enjoy direct control over their own land, and surplus expropriation takes place through the market, rather than through an exploitative landlord class. The African situation, too, has tended to favour ‘roving’ rather than ‘stationary’ rebellions, in Olson's terms; many rebels are merely passing through the countryside, on their way to seek power in towns. Having little in common with the peasantry, and nothing to offer it, they resort to violence as the only way to control it. However incoherent their objectives, and however brutal their methods, rebellions nonetheless reflect a serious urban malaise that needs to be addressed.

Why are ‘growth coalitions’ involving business interest groups and governments so rare in Africa? How has democratisation affected the possibilities for growth coalitions? In three cases with varying degrees of democracy – Mauritius, Zambia, and Zimbabwe – we find that hypotheses about growth coalitions that place importance on the organisation of the business sector are generally borne out. Yet even when the business community is organised in an ‘ideal’ manner, growth coalitions still depend on factors within the state: leadership, ideas, and capacity. Democratisation has a mixed effect. We find that in the case of Zambia, business–state relations did not improve despite a pro-democracy stance by business and the pro-business agenda of the democratic government coming to power in 1991. In Zimbabwe, the erosion of democracy reduced business access to state elites, breaking up a growth coalition that initially showed considerable promise. In Mauritius, the strengthening of democracy has paralleled the deepening of the growth coalition, and both have been reinforced by a strong economy. Our study shows that growth coalitions are possible in Africa; the key lies in determining the conditions under which such coalitions can be sustained in Africa's fragile polities.

Citing recent case study evidence from various parts of Africa, this article argues that the income diversification efforts of most rural dwellers over the past decade have been directed at meeting daily needs amidst declining returns to commercial agriculture. Individuals and households have experimented with new forms of livelihood, expanding their non-agricultural income sources, while retaining their base in subsistence farming. Various livelihood patterns are emerging, depending on historical, geographical and agro-ecological factors at local and national levels. Livelihood experimentation has catalysed overlapping arenas of dynamic change, notably disequilibria between households and individual members, tensions between generations, the recalibration of gender power balances, and a search for new social networks. So far this surge of livelihood ‘multiplexity’ has not generated adequate overall levels of gainful employment, technical innovation, purchasing power or welfare improvement. Thus, probing the complex interplay of economic, social, cultural and political dynamics in rural Africa becomes all the more essential for effective policy formulation.

Since the restoration of traditional leaders in Uganda in 1993, the Kingdom of Buganda has developed unusually effective institutions, financing mechanisms and policy tools, re-building itself as a quasi-state. The reinforcement of Buganda's empirical statehood provides one of the farthest-reaching examples of the current trend of traditional resurgence in African politics and to some extent supports claims for the participation of traditional structures in contemporary political systems. Yet, the Buganda experiment also highlights the limits of traditional resurgence as a mode of reconfiguration of politics in Africa. First, it is unclear how the kingdom can maintain the momentum of its revival and the allegiance of its subjects in view of its fiscal pressure on the latter and the limited material benefits it provides to them. Already the monarchists are finding it difficult to translate the king's symbolic appeal into actual mobilisation for development, shedding doubts on one of the main justifications for the kingdom's rebirth. Second, Buganda's claims to political participation clash with the competing notion of sovereignty of the post-colonial state. These limits are likely to confront other similar experiments across the continent.

Numbers of women in public representative office have increased dramatically in Uganda since the introduction of the National Resistance Movement's ‘no party’ system, because affirmative action measures have been taken to reserve seats for them in Parliament and local government. This article offers an assessment of the impact of these measures on women's political effectiveness, examining how far women in Parliament have been able to advance gender equity concerns in key new legislation. The article suggests that the political value of specially created new seats has been eroded by their exploitation as currency for the NRM's patronage system, undermining women's effectiveness as representatives of women's interests once in office. This is because the gate-keepers of access to reserved political space are not the women's movement, or even women voters, but Movement elites. The women's movement in Uganda, though a beneficiary of the NRM's patronage, has become increasingly critical of the deepening authoritarianism of the NRM, pointing out that the lack of internal democracy in the Movement accounts for its failure to follow constitutional commitments to gender equity through to changes in key new pieces of legislation affecting women's rights.

This article analyses the ways in which socio-political opposition is expressed by looking into the morally loaded discourse of political legitimacy in Burkina Faso that emerged after the assassination of the journalist Norbert Zongo in December 1998. Through the analysis of different political statements, newspapers and various comments from the ‘street’, it locates the struggle against impunity in a social and political undercurrent in Burkinabe society. In this context, notions of the public space are central, because the public space defines both the boundaries of public debate and the behaviour of key political actors. Two recurrent themes in Burkinabe political discourse, namely ideas of truth and courage, and the legitimacy of White people, illustrate the various ways in which socio-political opposition seeks to define the public space within which politics is to be practised and the behaviour to be observed by those acting there. But the struggle against impunity also takes place on a symbolic level at which key symbols are appropriated, interpreted and incorporated into political discourse.

In Africa, the management of border disputes varies from sub-region to sub-region. Most puzzling is the difference between West Africa and the Horn of Africa. In the latter, border disputes are much more likely to escalate into war than in the former. Seeking to solve this puzzle, this study focuses on the territorial integrity norm. It departs from existing accounts of this norm in two ways: first, it does not choose the region but the sub-region as the level of analysis. Second, it does not isolate the territorial integrity norm from its social context but analyses the interplay of the norm with the social structure in which it is embedded. It concludes that the territorial integrity norm in West Africa is part of a social structure different from that in the Horn of Africa. It is this difference that explains the different patterns of conflict management in the two sub-regions.

Though policing is widely regarded as an inherently public good, in reality non-state policing is widespread in South Africa, doing everything that the public police force does. The paper examines the extent, nature and attitudes towards non-state policing in South Africa in its different forms, that is, autonomous citizen responses, ‘responsible’ citizen responses and commercial private security. The paper concludes by assessing the social and political implications of non-state policing.

In apartheid South African society, racial and ethnic identities were institutionally imposed. The end of apartheid has brought about the need for new identities to be created among South Africans, and for South Africans to forge a new relationship with their society and country. With this objective in mind, the national government is engaged in a process of nation-building. But in post-apartheid South African society, sub-national identities are also strongly coming to the fore. This is an empirical study of established and emerging identities in the Western Cape province, and the processes whereby these are constructed. The investigation shows two parallel flows of identity construction in the Western Cape: on the one hand, political leaders in the province attempt to foster an autonomous provincial identity; on the other, residents of the province show little evidence of strong political identities linked to the Western Cape. Instead, social identities are being constructed around residents’ local neighbourhoods and long-existing ethnic, class and racial identities. Rather than the social cohesion sought by the post-apartheid South African government, these identities point to persistent social polarisation.

In this era of neo-liberal capitalist economics in Africa, has organised private capital in the form of business associations (BAs) become more active in public life or developed influence in public policy formation or implementation? This analysis examines the impact of five key factors to explain varying activity levels and influence of BAs in Ghana and Nigeria since independence: levels of capitalist development and hence size of the capitalist class; strength and autonomy of the capitalist class; strength of capitalist ideology; democratic vs. authoritarian rule; and impact of external hegemonic powers and ideologies. The paper finds that Nigeria's BAs are more highly developed, have had higher activity levels, and had more influence, however limited, than Ghana's. Externally generated economic liberalisation has stimulated higher levels of BA activity, but not necessarily the political space for BA autonomy. Political democratisation appears to increase political space, BA access to the state, and policy influence.

Refugees impose a variety of security, economic and environmental burdens on host countries, but also embody a significant flow of resources in the form of international humanitarian assistance, economic assets and human capital. These refugee resources represent an important statebuilding contribution to the host state, but security problems and other hindrances inhibit the state's ability to access and control them. This article explores the challenges and opportunities for African states arising from the double impact of refugee-generated resources and security problems. It argues that the potential benefit for the state and its citizens go beyond the burdens imposed by a mass influx. Refugee resources and security threats potentially provide long-term gains, and, by compelling the state to strengthen its grip on border areas, enable the state to ‘harden’ its presence there. However, for host states to realise the potential of refugee resources and continue hosting refugees, they must be assisted by appropriate humanitarian programmes.

A year after the establishment of the Transitional National Government (TNG), Somalia and in particular the southern part of the country remained fragmented and riven by armed confrontations between clans and faction leaders. The TNG succeeded in re-assuming Somalia's seat in the United Nations and all relevant regional organisations, while the breakaway north-western region of ‘Somaliland’ failed to gain any international recognition, despite the widely acknowledged establishment of a relatively stable administration. This paper tries to explain this apparent discrepancy by arguing that state or government recognition is an instrument of international relations, designed to influence rather than passively acknowledge political reality. In Somalia, granting or withholding recognition to one or other political leadership has been driven by specific interests, thus complicating the process of political stabilisation and national reconciliation.

In the 1970s politics in Tanzania was substantially a bureaucratic affair. Since the 1980s, however, economic liberalisation, multiparty democracy and governance reforms have on the one hand introduced measures conducive to building a legal-rational bureaucracy and a liberal civil society, and on the other accelerated political struggle for economic resources through personalised regional networks. Paraphrasing Emmanuel Terray, the first trend is described in this article as the manufacture of ‘air-conditioned’ politics, the second as the growth of ‘veranda’ politics. The article argues that donor reforms are not leading in a straight line to liberal governance, but neither is civil society simply being colonised by patrimonial networks. Rather, both ‘air-conditioned’ politics and ‘veranda’ politics are advancing simultaneously, inundating a previously bureaucratised political sphere. The dual character of this ‘re-politicisation’ makes the fate of governance reforms exceedingly difficult to predict.

Research from a community study in rural Malawi speaks directly to contemporary debates about civil society. Investigating the role of local churches in empowering citizens, the study found that the local Catholic church was more effectively fostering a nascent sense of political efficacy among women than were local Presbyterian churches. Explaining this finding, the article presents two issues that expose problems in the liberal understanding of civil society, and underscore important themes raised in the critical discourse. First, the study reveals that organisations characterised by decentralised authority structures and internal democracy may fail to contribute to the empowerment of marginalised citizens. Such organisations are prone to reproduce and exacerbate local inequalities and conflicts within their structures. Second, corroborating critical views, the study highlights the importance of recognising how power relations affect the character and operation of civil society organisations. The adjusting of power relations within organisations may be a prerequisite to their serving an empowering role with marginalised citizens.

The paper is concerned with the unintended consequences of aid as a relation of governance: in this case, the failure of aid agencies to improve the lot of displaced Southerners living in North Sudan after more than a decade of engagement. It is argued that aid, as a governance relation, is complicit with wider forms of oppression to which Southerners are subject. The aid-based IDP (Internally Displaced Person) identity, for example, resonates with state forms of deculturation. At the same time, developmental ideas of self-sufficiency articulate with the commercial need for cheap agricultural labour. Developmental strategies have tended to reinforce the subordination of displaced Southerners rather than enhancing their autonomy. Examples of this collateral effect are examined in relation to share-cropping, food aid, debt and asset stripping. The paper ends by calling into question the appropriateness of aid as a vehicle for a shared duty of care.

The example of Zimbabwe in 2000 shows that in a context of violent election campaigning, the role of international election observation is an ambiguous one. Unlike earlier elections organised by the Zimbabwean government, international donors wanted to observe its 2000 parliamentary election amidst a deepening crisis. They noted that the elections would not be free and fair in their view. Neighbouring countries with a more positive view joined the observation exercise. The government's discriminatory invitation and accreditation policy, the observers' emphasis on the peacefulness of the polling rather than free and fair elections, and the selective publication of their reports in various media, were affected both by the political agendas of the domestic players and by the governments which sent the observers. The difference between the Western view of the government, which had changed drastically since the 1980s and early 1990s, and the view of neighbouring governments, was crucial and may become significant elsewhere in Africa.

In 2000, Ghana's landmark elections ushered in a new era of democracy. Scholars, however, have yet to scrutinise the structural underpinnings of the country's electoral system. This article offers a detailed assessment of Ghana's bloated voters' register, patterns of voter turnout and the lingering accusation of electoral irregularities in the Volta and Ashanti Regions in the 2000 elections. Most significantly, it critically analyses the severe malapportionment of the country's 200 parliamentary seats. While the 2000 elections helped to consolidate the democratic process in Ghana, structural inequalities continue to plague the country's electoral system.

By 1980, Ethiopia was gripped in escalating civil wars. After a series of punitive expeditions had failed to suppress them, the government organised large-scale operations in the early 1980s against the insurgencies in the eastern and northern territories. The operations seemed to have been informed by what is called ‘total strategy’. Although the emphasis was on the coercive component, the state also used psychological and economic incentives. The results were mixed. The eastern rebels were defeated more easily because they were factious. The northern campaign failed because of the rebels' staunchness and the terrain's unsuitability. In a cold test of wills, the Eritrean fighters not only held the offensive to a stalemate, but also went on to win total military victory. Same strategy, different outcomes: this suggests that no single counter-insurgency strategy can always have the same results as it is influenced by numerous factors that may vary from one place to another.

Stefano Ponte's article in the first JMAS issue of 2001 brings good news: previous pessimistic accounts of the dynamics of economic life in the Uluguru Mountains, especially mine, have to be revised (Ponte 2001; van Donge 1993c). There may have been economic decline in the area, but economic liberalisation has turned that around and raised incomes. Ponte collected rich material, especially in his household survey. Whereas I found it difficult to get precise estimates of money, land, and so on from people there, he collected precise quantitative insights. This information covers not only their present situation: people recalled with considerable precision their situation back to the mid eighties when I did research there. His critique touches, however, not only changes over time, but also the assumptions about how actors behave: people are much more resourceful than I portrayed them.

This article makes two major claims. The first is that independent Botswana was able to generate and sustain a type of developmental state because of the presence of an indigenous initiator state culture that was preserved by the Protectorate state and was inherited by the post-colonial state elites. The second is that the non-emergence of the developmental state in post-colonial Zimbabwe is explained by the presence of a non-initiator indigenous state culture which was preserved by the Rhodesian colonial state and was inherited by the post-colonial state elites. The article briefly reviews the literature, analyses the Tswana and Shona pre-colonial state cultures, and shows that these were preserved by the colonial states and inherited by the nationalist politicians.