This article uses empirical data to discuss the links between ethnicity, inequality and governance in a framework that divides countries according to their levels of ethnic polarization. It makes three main arguments. First, types of diversity, not the existence of diversity per se, explain potentials for conflict or cohesion in multiethnic societies. Ethnic cleavages are configured differently in different social structures and are less conflictual in some countries than in others. Second, relative balance has been achieved in the public sectors of countries that are highly fragmented or those with ethnicity-sensitive policies, but not in those with ethnicity-blind policies. Third, the article is critical of institutional approaches to conflict management that underplay background conditions in shaping choices. Consociational arrangements may not be relevant in unipolar ethnic settings or fragmented multiethnic societies, where governments may be ethnically inclusive under democratic conditions. They seem unavoidable in ethnic settings with two or three main groups or in settings with strong ethnic/regional clusters.

Since the early 1980s, most African countries have experienced unsatisfactory rates of economic growth and profound changes in livelihood systems, which have affected the way their modern institutions function. However, when confronted with evidence of poor economic performance in countries undergoing adjustment, the international financial institutions often blame governments for their lack of political will in regulating the activities of bureaucrats and vested interests. They recommend policies aimed at restructuring public sector institutions through privatization, public expenditure cuts, retrenchment, new structures of incentives and decentralization. Despite efforts to implement these measures in a number of countries, the problems of low institutional capacity remain. Two key contradictions appear to explain why institutions have been largely ineffective in crisis economies in Africa: the growing contradiction between the interests of bureaucratic actors and the goals they are supposed to uphold; and the contradiction between the institutional set‐up itself and what goes on in the wider society. To understand how these contradictions work, it is necessary to look more closely at the set of values and relationships that anchor institutions on social systems. The issues here are social compromise and cohesion; institutional socialization and loyalties; overarching sets of values; and political authority to enforce rules and regulations. The crises in these four areas of social relations, which are linked to the ways households and groups have coped with recession and restructuring, have altered Africa's state institutions so that it has become difficult to carry out meaningful development programmes and public sector reforms without addressing the social relations themselves.

Crisis and economic reforms have changed the status of intellectuals and their relations with dominant élites and policy makers. Because of the technical and ideological nature of these reforms, policy makers have tended to rely on intellectuals as opposed to bureaucrats to shape the agenda of change. This has converted a large number of intellectuals into technocrats and undermined the fabric of academic life in universities. Nowhere is this more pronounced than in developing countries with a large middle class and an economic or financial base that ensures some degree of independence from the multilateral financial institutions. Understanding the success or failure of economic reforms in developing countries requires some analysis of the complex relations between technocrats and political leaders and the societal constraints both sets of actors face. This article explores these issues in the context of Nigeria, whose academic community, bureaucrats and professionals grew in leaps and bounds in the 1970s, following an oil‐induced boom. The crisis of the 1980s led to attempts at economic reform and a highly programmed strategy of political change in which academics played a leading role. The article examines the effectiveness of intellectuals in government, and compares the Nigeria case with technocratic experiments in Ghana, Botswana and Côte d̂Ivoire.

The brutal nine-year conflict in Sierra Leone has defied both military solutions adopted by various governments and peace accords imposed by the international community and regional powers. The latest casualty is the controversial Lome! Accord, which gave power and amnesty to the rebels in a power-sharing government. This article offers explanations for the failures by focusing on the interplay between the policy choices of decision-makers and the country's governance institutions and social structure. The policies of decision-makers can have negative consequences on societies if they are grounded on institutions that are at variance with a country's social structure. The article develops three main arguments to support this conclusion. First, the country's bipolar ethnic structure and majoritarian presidential system of government act as serious constraints on policies that seek to forge a national coalition to end the war. Second, conventional armies in ethnically polarised settings are a poor instrument for fighting rebel groups that deliberately use mass abductions and terror as war strategies. Third, peacemakers do not understand the institutional contexts in which violence-prone rebel groups can be made to reclaim their humanity and observe peace agreements. The logic of the interconnections between territory, resources and civilisation suggests that Sierra Leone's rebels are unlikely to accept disarmament and honour the democratic road to peace.

Despite the fall in global income inequality in the last two decades, levels of living among individuals in the world are still very different and are likely to remain so for a very long time. The uneven rate of decline in inequality and growth volatility in commodity‐dependent countries suggest that there is no reason to believe that global inequality will continue to fall until it reaches acceptably low levels. Global disparities in incomes and welfare, especially in social protection, are at the heart of the problem of migration and populism in Western democracies in recent years. They bring out in bold relief the lack of fit between claims of global convergence and people's perceptions on incomes and well‐being between the global North and South. In this regard, it is more realistic to talk about ‘multiple geographies of 21st century development’ than a ‘one world’ or single geography of global development. Issues of power, and the way global and national relationships and rules allocate advantages and disadvantages, or promote convergence and divergence, between and within countries, need to be front and centre in the discussion on global convergence.