Since the early 1990s a growing number of constitution building processes in countries transitioning from authoritarian rule or recovering from war or severe institutional crisis have involved public participation. This increase stems from an assumption made by many peacebuilding scholars and practitioners that public participation in constitution building will lead to higher levels of democracy. This assumption has not, however, been the subject of systematic or comprehensive analysis. Therefore, the overarching purpose of this thesis is to scrutinize the participation-hypothesis - as it is referred to in this study. The study is a two-step investigation. The first part begins with an analysis of twenty cases of participatory constitution building that have occurred in post-conflict states, transitioning states and countries that have experienced a severe institutional crisis. In order to differentiate the cases in terms of how much influence participants were granted, an analytical framework is developed and the cases are categorized as either false, symbolic, limited, consultative or substantial participation. The participation-hypothesis is then empirically investigated by comparing democracy levels prior to and after the process for each of the 20 cases. In order to further test the hypothesis, cases of constitution making in which there was no public participation are then added to the investigation. These cases are included as a point of reference – the democratic outcome in this group is compared with the democratic outcome in the twenty participatory processes. The empirical results reveal that there is no relationship between public participation in constitution building processes and higher levels of democracy. On the contrary, some cases that involved considerable influence for participants have not experienced improved levels of democracy, while cases with low levels of influence for participants have shown democratic improvement. Moreover, a majority of cases of constitution making without public participation have also experienced increases in their democracy scores. Therefore, the conclusion of the first part of the study is that the participation-hypothesis does not stand up to empirical scrutiny. Particularly challenging for the participation-hypothesis is the fact that the analysis in part one shows that similar participatory processes have been followed by democratic improvement in some countries and democratic decline in others. Two such cases are Kenya and Zimbabwe. While democracy levels have increased in Kenya since the conclusion of the process, they have steadily declined in Zimbabwe. In the second part of the study, these two countries are therefore the object of intense, systematic and comparative scrutiny in order to explore factors beyond participation in constitution building that might explain the different trajectories of democracy. The comparison shows that the actions of political elites – in particular their ability to cooperate with each other – is the major explanation as to why the two wind up on different paths. The importance of elite cooperation is well-established in the democratization literature. One major conclusion of this study is therefore that the participation-hypothesis needs to be informed by insights drawn from this literature.