While public attention has focused on the stature of Nelson Mandela, there has been at a deeper level in South Africa since 1990 a steep decline in state capacity, and a marked deterioration in democratic practice. The participatory democracy which had so characterised the decade of the 1980s was brought to a sharp end after the return of the nationalist leaders, and the workings of even a liberal, representative democracy have also suffered under the rise since 1994 of a predominant party system and élitism. The latter features are present too in Namibia, with similar consequences. Democracy which is understood merely as electoralism, as Botswana earlier had shown, has few defences against predominance. The voters' brief electoral act is wide open to manipulation and containment. Power is shared by élites, while popular participation is rendered moribund, and concern for justice and equality ceases.

There are two remarkable features of post-independence economic policy-making in Zimbabwe: the very limited nature of the changes made by the new government in 1980, and the complete reversal of policy announced in 1990. It was surprising that a more radical transformation had not been introduced soon after independence, since this had been achieved by a civil war prompted not only by the denial of even basic rights to the majority of the population, but also by an extremely inequitable distribution of economic resources. The volte-face in 1990 was also unexpected, because it required a repudiation of governmental rhetoric at a time when the economy was by no means in a state of crisis, even though under stress. This article attempts to understand these policy shifts.

How can Africa's societies reorient and rebuild state apparatuses where predatory, neo-patrimonial governance has held sway? This emerges as a key question if one accepts that dysfunctional institutions, and not just poor initial conditions, a hostile international environment, external shocks, or policy errors, have impeded Africa's economic recovery. Although analysts and organisations place differing weights on these various causes, few now dissent from the consensus that political-institutional reforms are a necessary condition for African development. Yet recognising the need for such change is one thing; knowing how to bring it about is quite another. Few guidelines exist to steer this complex and long-term task. Democratisation is now often cited as one means to foster institutional renewal, and this article draws on the Ghanaian experience to reflect upon the efficacy of this route.

This is an attempt to assess Eritrea's democratic prospects in light of its political experience since 1993 as a sovereign state. I readily concede that it is too early to make a definitive assessment; yet, just as a child's first few years are extraordinarily critical in helping to determine its subsequent intellectual, physical, and moral development, so can a new nation cumulate experiences that help set it firmly on its feet and give it the capacity for yet more positive growth. Further, there is no denying the force of genetic and environmental factors in the development of both children and states: what they inherit from the past, as well as the circumstances of their creation, and experiences of infancy, all contribute to what they will become. I have no wish to stretch the metaphor further; my point is that a detached analysis of Eritrea's political inheritance and the forces and key events since its formal independence can give, at least, a preliminary reading on its prospects. Thus, I intend to argue that these first five years have been marked by pragmatic policy-formation and decision-making which, taken as a whole, warrant a cautious optimism about the country's democratic future.

The Nile flows for 6,700 kilometres through ten countries in north-eastern Africa – Rwanda, Burundi, Zaïre/Congo, Tanzania, Kenya, Uganda, Eritrea, Ethiopia, the Sudan, and Egypt – before reaching the Mediterranean, and is the longest international river system in the world – see Map 1. Its two main tributaries converge at Khartoum: the White Nile, which originates from Burundi and flows through the Equatorial Lakes, provides a small but steady flow that is fed by the eternal snows of the Ruwenzori (the ‘rain giver’) mountains, while the Blue Nile, which suffers from high seasonal fluctuations, descends from the lofty Ethiopian ‘water tower’ highlands. They provide 86 per cent of the waters of the Nile – Blue Nile 59 per cent, Baro-Akobo (Sobat) 14 per cent, Tekesse (Atbara) 13 per cent – while the contribution from the Equatorial Lakes region is only 14 per cent.

Today Burundi resembles a torn-apart patchwork of Batutsi and Bahutu colonies by being divided territorially into almost isolated enclaves, ethnically purified to a high degree, and more or less closed off from their surrounding areas. Most Batutsi believe that they can only live in security if sheltered by the army in various urban agglomerations, including the capital, Bujumbura. By way of contrast, most Bahutu are to be found in the hills of the countryside in an often entrenched and hidden way, and close to economic self-sufficiency. Such an unusual situation is undoubtedly a big challenge to the relations that need to be established between state authorities and citizens in rural as well as urban areas.

Since 1990, more than 30 sub-Saharan states have introduced multi-party electoral systems to choose their national governments. This development, normally understood as part of the post-cold war ‘wave of democratisation’, has led many to think that there is something intrinsically new about such a competitive process in Africa, which is not the case. This article looks at some elections that have been held in various states and studied, both before and after their independence.