Patterns of Late colonialism and Democratization in Africa: using V-Dem to measure the long heritage of the colonial state

Between 1989 and 1995 the third (or fourth?) wave of democratization hit Sub-Saharan Africa. Protests and demonstrations in the streets, combined with a renewed international pressure, (re)introduced elections in almost all African countries. By the end of the decade, most resulting political transitions were either reverted or revealed as merely cosmetic. Still, almost a quarter of a century latter, some of the successful democratizations still endure. These different outcomes pose a challenge for most contemporary theories of democratization as they fail to explain this variation. So how can we explain democratization in this region? This paper will readdress this problem by re-introducing a variable that has not been fully explored: the late colonial period. During this period, that lasted from the 1930s through independence, colonial empires pressured by their own needs, the changing international arena, and by struggles in the colonies themselves had to reform their rule. In some cases these reforms resulted in the introduction of elections with universal suffrage and local parliaments, two institutions that were quickly adopted and used by the liberation movements to achieve independence. We argue that this legacy of democratic experience is a crucial factor determining a successful democratization by the end of the 20th century. Two different routes will be taken to test our hypothesis. First, we will sketch a brief historical comparison of different cases from the British and French colonial empire. Through this exercise we seek to understand if these differences did occur, how significant they were and what mechanisms might link them to a process that occurred almost three decades later. In a second moment we will use the Varieties of Democracy (V-Dem) database and see how it shows the differences during late colonialism and we will use it to test statistically our hypothesis. This unique dataset presents two major advantages to this study. First, it comprises data during colonialism, which is seldom the case in these databases. Secondly, because different dimensions of democracy were took into consideration while creating the data, it allows for a more fine and granular analysis of how changes in some crucial components were more significant than others. During this exercise we hope show how the V-Dem data can be used not only for macro-quantitative testing, but also can be used by more qualitative driven, regionally and historically focused studies.