The re-emergence of multi-party elections in Africa has often not been taken seriously, especially by many area specialists (less so by scholars with a general political science background). ‘Electoralism’ is the derogatory term used to belittle the significance of multi-party elections in Africa in general and their meaning for democratic development in particular. Because elections do not change the ‘real’ power structures in Africa, the argument goes, they have little impact on the fate of ordinary people and the prevalence of poverty.

Stefan I. Lindberg's extremely detailed work on elections in Africa since 1989 repudiates this conventional scepticism. Based on a systematic statistical analysis of data from 232 contested elections in 44 countries between 1989 and 2003, he comes to a number of well-argued conclusions about the power of elections – even in Africa. Elections are not democracy, of course, but without free and fair elections you cannot have a democratically legitimized government and, hence, representative democracy.

His first point is that elections are not the completion of any democratic transition in Africa; rather, they foster liberalization of the political regime. The second, more important point is that elections have a ‘self-reinforcing power’ that ‘promotes increased democracy’ (p. 96). This means that the quality of elections tends to get better, more democratic, with repetition; and after three elections regime breakdown is, statistically, a rare occurrence. Third, elections support the institutionalization, and even the ‘deepening’, of civil liberties – what he calls the ‘self-improving quality’ of elections. And finally, elections are a causal variable for democratization. Essentially, Lindberg argues that what has often been termed ‘mere procedures’ has over time gained a value and a quality of its own.

However, there is one crucial weak point in Lindberg's argument which is related to the link between democratic elections and the consolidation and quality of democracy. Of course, he is aware of the fact that elections are only one of the variables that explain the quality of democracy, and consolidation is more than free and fair elections. The simple question is: what happens during the four or five years between elections? Thus, his direct conclusion that improved democratic elections lead to an improvement in the quality of democracy or its consolidation may be methodological fallacy. The missing link here is a causal chain or mechanism of how the many institutions involved are intertwined into a self-reinforcing sequence.

However, jumping this causal chain does not make Lindberg's well-written argument about the power of elections as an institution any less convincing. In fact, in a wider context the book provides strong evidence for institutionalists' dictum that institutions matter – they constrain, shape and enable the behaviour of political actors. More importantly, it provides material for reasoned doubt for all those, like Goran Hyden, whose understanding of African politics is based on the ‘truism’ that ‘abstract constitutions and formal institutions exist on paper, but they do not shape the conduct of individual actors, especially those in power’.Footnote ¹ Without ignoring the duality of African politics, captured in the concept of neopatrimonialism (a mixture of patrimonial and legal-rational bureaucratic domination), in the light of Lindberg's analysis we have to consider how far the acceptance and observation of formal institutions is, albeit slowly, gaining ground in African politics. Too often Africanists have been one-sidedly obsessed with the informal dimension of politics, with the ‘politics of the belly’ (Jean-François Bayart) or ‘disorder as political instrument’ (Patrick Chabal and Jean-Pascal Daloz).

From a methodological perspective, Lindberg also provides good arguments for scholars in the historical institutionalism tradition who are interested in path-dependent developments. So his approach may also be seen as a stimulus to apply the instruments of historical institutionalism to Africa. By now, the depth of historical research may well be adequate for such an undertaking. But even apart from such specialized fields, this book is essential reading for all people interested in elections and democracy in Africa.