Presidentialism, Parliamentarism, and Democracy Reconsidered

The purpose of this dissertation is to add to the understanding of democratic consolidation, and to address a debate within this topic: Is presidentialism harmful to democratic consolidation? I argue that presidentialism induces higher levels of political violence (attitudinally and behaviorally). Unlike parliamentary and semi-presidential systems, which offer mechanisms to alter the incumbent government through legislative responsibility, such as a vote of no confidence or a government reshuffle, when there exist mismatched policy expectations between the public and the government, or when the public dissatisfaction with the government is high, presidential systems do not have this mechanism to change the government composition and the president is empowered to govern until the next election. Even in the case that the public's discontent toward the president is high, there exist almost no mechanisms except for her own resignation and an impeachment to remove her from her office. However, a voluntary resignation and an impeachment are rarely occurred in the history, and thus, the expectation of the public regarding whether their grievances can be resolved and addressed is more difficult to be fulfilled in presidential democracies. Therefore, using and considering violence as a mean to address their political and social problems becomes a more viable option in presidential democracies. But by doing so, political stability will decrease and democratic consolidation will be hindered. I employ the World Value Survey and the Asian Barometer Survey to find support for this argument.

To further extend this argument and to address the debate, I argue that democratic breakdown must be considered a two-step process. For a democracy to break down, the presence of a democratic crisis that presents a significant likelihood of overthrowing the current democratic regime is necessary. Specifically, I argue that presidentialism generates political instability through its institutions, which are associated with a greater likelihood of the emergence of a democratic crisis, but political instability does not further contribute from democratic crisis to democratic breakdown. Using data covering all democratic regimes from 1946 to 2008, I demonstrate that presidential democracies are more likely to encounter crises than either parliamentary or semi-presidential systems. However, once a crisis occurs, presidentialism does not lead to a higher likelihood of breakdown. Thus, presidentialism is associated with a higher likelihood of democratic breakdown, but only by affecting half of the process.