There are two prevalent themes in literature and general observation about democracy in Africa. The first is heavily influenced by modernisation theory and assumes a teleological development of the state toward liberal democracy. It is observed that the liberal democracies of Europe, North America, Japan, Australia and New Zealand (commonly shorthanded to 'the West') developed gradually over time from feudalism through industrialisation and slowly into the flowering beds of freedom, liberty and equality they are today. The countries of Africa, indeed of all the Second and Third Worlds, will soon and slowly but surely develop into the same kinds of liberal democracy we see in 'the West': strong protection of individual freedom, elected (for the most part - see UK's House of Lords) leaders, along with social and economic liberty. The connection between liberalism and democracy is 'natural' and the presence of both is inherently good: so good, in fact, that Western governments are justified in killing thousands of civilians abroad in the name of protecting and spreading these values. Adherents of this theory used to say that all African countries needed were elections and they'd naturally vote for more liberty (both social and economic) and before long there'd be a bunch of lovely, stable liberal democracies. When the so-called democratisation of Africa began after the Cold War, there were a huge number of elections in 1993. Did these bring peace and stability to Africa? No. The reasons why, many liberalists concluded, revolved around lack of economic growth and lack of civil liberties. As Bratton and van de Walle argued in 1997: elections were not enough in themselves to demonstrate democracy: they must be conducted within a matrix of civil liberties.
As you may have guessed, I am not totally convinced by this teleological view of liberal democracy. I do agree that, as far as political regimes go, the one I live in is more or less as good as it gets: I love my freedoms, I love that I can write a blog post about how much I disagree with David Cameron's policies and that he would proudly uphold my right to do so, I love that I can have a say in the policies of my government through the ballot box. What I disagree with is that the situation I am in is somehow 'natural' or 'inevitable'; that humans objectively desire freedoms and liberties for other people or even for themselves; that there couldn't possibly exist a better political system than a liberal democracy in which I live. I do think it is possible to hold these doubts about liberal democracy whilst also seeing the oppression of other people and wanting it to end. What I want is that those oppressed people can create their own futures - free from what their dictator tells them or what a Western aid agency or government thinks they should want.
This is not to say that I fall into the other dangerous camp of literature on democracy in Africa: the racist one. The one that says Africa cannot have liberal democracy because of its 'culture'; because it is naturally traditional, patrimonial or any other political scientific euphemism for 'backward' you want to throw in. Which isn't to say that the only adherents of this view see it is a necessarily bad thing: there are plenty of African romanticists who view the African system as naturally different from 'the West', as naturally 'traditional' and adverse to modernisation.
The two problems common to both of these theories is that they posit a 'naturalness' or 'objectivity' about a particular political reality which is in fact situated in a highly volatile and fluid historical context. Because one can posit similarities between different regimes or societies across time and space does not make them objectively natural - it makes them comparable and in some way useful to study but does not render them inevitable. The second problem with both of these views of democracy in Africa is that they allow no room for variability within the continent. This is the most common problem I find with literature about politics in Africa and in fact I am wholly unconvinced that there are any generalisations that can be made about the continent. Swathes of articles will make sweeping generalisations based on ten, five or even one country. That there are 57 countries on the continent gives one the sense that the scholarly act of making generalisations should not be taken lightly. This is especially true of democracy where, in Africa, there exist countries which have pretty well-entrenched multi-party democracies (Benin, Ghana, Botswana) while there are also others whose elections have seemed, if anything, to worsen ethnic tensions, violence, and corruption (Democratic Republic of Congo, Kenya, Zimbabwe).
The answer to these issues for scholars is to analyse carefully historical, political and socioeconomic factors of a given state or society, using comparative analysis where appropriate but not as a solve-all tool. The answer for international policy-makers is a whole lot more complicated, but it has to start by shifting their focus from the central and single importance of elections; ending the characterisation of the states of the world as either 'friends' or 'enemies' with the former to be appeased no matter what ills they cause (e.g. Yoweri Museveni in Uganda or Joseph Kabila in DRC); and placing greater responsibility on the corporations who pay taxes to their states while making money off the backs of death, destruction and corruption in authoritarian regimes.