The democratic qualities of competitive elections: participation, competition and legitimacy in Africa.

This article analyses the democratic qualities of core institutions of representative democracy: multiparty elections. Focusing on the three basic democratic values participation, competition and legitimacy, the empirical examination of Africa compares both over time and between founding, second, third and following elections. The results, based on 203 observations of presidential and parliamentary elections, show that there are significant improvements of democratic qualities in Africa and breakdowns typically occur only after founding elections. The core institutions of representative democracy may have a future in Africa. Hence, there is a cotinuing case for demo-optimism on the continent.

**********

The literature on the 'third wave' of democracy (1) has produced an increasing number of comparative accounts over the last decade. From an initial focus on transitions and pathways to democracy, and on democratic consolidation, concerns over the quality of democracy have come to the fore. It has also been argued that democratic quality is more fruitfully analysed in terms of partial regimes, rather than in terms of ambiguous multidimensional concepts of democracy, or various inconclusive categorisations along the lines of semi-democracy and non-democratic regimes. (2) This article analyses the democratic qualities of a partial regime at the heart of representative democracy: competitive elections. Using 11 indicators of the three intrinsic democratic values of elections--participation, competition and legitimacy--the empirical analysis covers 203 cases of election in Africa (3) from 1989 to 2001. Multiparty elections alone do not make a democracy, but the study of the democratic quality of competitive elections serves as one piece in the puzzle.

Previous studies of competitive elections in Africa carried out by Bratton and colleagues (4) have been pessimistic. Elections have helped some new regimes to survive but little more than that, and the frequency and quality of elections are on the decline. Late founding elections (held after 1995) as well as second elections in general were worse than earlier attempts, in terms of participation, competition and legitimacy. Finally, fewer turnovers resulted from second elections, and in this regard African politics is said to have returned to the 'normality' of dominant parties and personalities in the form of 'big man' politics. The results presented here challenge these earlier findings on several counts. There is no general negative trend in either frequency or the democratic quality of multiparty elections. Previous results were a consequence of these authors' choice of periodisation and the limited number of years included in the analysis. To the contrary, analysing trends over founding, second and third elections there is a significant improvement in democratic quality, in particular as countries hold third elections. Moreover, while the cited authors noted that turnovers were uncommon in second elections, current evidence shows that alternations in power have been as common among third elections as in founding ones. (5) In conclusion, it seems that there is a case for demo-optimism (6) with regard to the increasing democratic qualities of elections in Africa.

ELECTIONS AND DEMOCRACY

This article does not aim to enter the debate on substantive versus procedural conceptualisations of democracy, or the dispute as to whether the non-democracy and democracy divide is better understood as a dichotomous or a graded phenomenon. (7) Suffice it to say that I agree with Dahl that even the narrowest procedural definition of democracy entails integral substantive rights and freedoms. Hence, the distinction between procedural and substantive definitions of democracy collapses. The right to self-government, as Dahl reminds us, is neither a trivial nor merely a procedural right. (8) Thus, at democracy's very core the electoral process enacts an important kind of distributive justice with regard to power and authority. In order to rule, the people must have some way of ruling, some procedure for making decisions that are binding for the members of the political community. To this effect, every modern vision of representative democracy regardless of definition entails the notion of elections as the primary means of selection of political decision-makers. As such, elections are a crucial procedure providing for substantive values. In any election of high democratic quality the right to self-government, for example, has de facto been effectively enforced. Hence, the centrality of studying elections, even if not to measure democracy writ large but as a partial regime. Yet an approach that sets out to measure the democratic quality of elections at the minimum must specify how such democratic qualities are defined. Accordingly, the specific justification for the focus on participation, competition and legitimacy of elections is discussed further below.

There are at least two additional reasons to study the democratic quality of elections. First, authoritarian regimes dominated the political landscape of Africa until the end of the 1980s. At the outset of the political changes in the early 1990s, most Africanist political scientists were bolstering demo-optimism. For many observers, however, political liberalisation lost its sheen within a few years. A new demo-pessimism developed among authors who, like Joseph, saw 'virtual democracies' rather than true democratization. (9) Other scholars argued that things quickly returned to normal 'big man', neopatrimonial, clientelist, informalised and disordered politics of the continent. (10) A few dissenting voices, like Wiseman, have taken accounts such as these to task for being excessively pessimistic. (11) The debate on democratisation in Africa came to resemble a 'dialogue of the deaf'. (12) Consequently, the literature has been flooded with contradictory hypotheses based on disparate approaches using various conceptualisations of democracy. Political scientists working on Africa have indeed paid too little attention to careful conceptualisation of the dependent variable(s), clear delimitation of hypotheses about the relationships between causes and effects, and rigorous measurement and compilation of comparable data. The outcome is that not enough cumulative work has been done. (13) One strategy to address this state of affairs is to begin to conceptualise, and collect comparative data, on partial regimes rather than the 'bundled wholes' (14) of overarching concepts like democracy and neopatrimonialism. Competitive elections constitute one such partial regime. An election is a phenomenon that can be conceptualised and measured in relatively unambiguous terms, indicators are of high validity, and data are available to formulate comparative indices. Its theoretical significance feeds into many areas of application. In sum, the study of the characteristics of African multiparty elections as a partial regime serves a necessary data-gathering, variable-creating and puzzle-solving function in our field. This article seeks to provide a further step in this regard.