Qualitative famine scholars have for three decades argued that democratic institutions are remarkably effective in reducing famine vulnerability. Yet, only two publications exist where systematized cross-country analyses have been used for empirically investigatingthe relationship. Pl umper and Neumayer (2009) find that democracies are more effective than autocracies in combating famine, while Rubin (2011) finds little evidence to support such a claim. In contrast to these studies, I disaggregate the concept of democracy, and develop a theoretical framework explaining how contestation, participation and civil liberties may all be linked to famine prevention.
A quantitative research design is developed in order to investigate the link between each of these components of democracy and famine occurrence. While there is no strong evidence that political participation and contestation are systematically related to faminevulnerability, the analysis reveals that the protection of civil liberties has a considerable impact on whether famines are likely to occur. The findings thus provide evidence for the proposed theoretical argument that governments are accountable to the public as long as citizens are able to voice their concerns. However, robustness tests reveal that certain caveats are warranted as regards the interpretation of the results. Most notably, it is hard to ascertain the conceptual precision of the civil liberties indicator, and it is therefore not possible to determine whether some factors that characterize `free societies' are more important than others in preventing famine. Yet, despite this shortcoming, the thesis demonstrates that disaggregating democracy is a viable strategy for more specifcally identifying which mechanisms that reduce famine vulnerability.