This dissertation investigates how different forms of power-sharing affect democratic quality.

While power-sharing has been widely promoted as a tool for stabilizing divided societies,

particularly in post-conflict contexts, very few studies have systematically assessed its impact on

democratic quality, and none have examined the distinct effects of inclusive, dispersive, and

constraining power-sharing institutions. To address this gap, this dissertation constructed a dataset

of 155 countries between 1975 and 2015, and analyzed, with a two-way fixed effects model, how

the three types of power-sharing influence both overall democratic quality and its institutional

components. The findings of this dissertation indicate that constraining power-sharing had the

most consistent and significant positive effect on democratic quality. Dispersive power-sharing

showed modest but statistically insignificant benefits, while inclusive power-sharing was

associated with a negative, though also insignificant, relationship with democratic quality. Further

analysis demonstrated that these trends persisted regardless of the post-conflict context,

challenging assumptions that such institutional arrangements are inherently less effective in fragile

post-conflict environments. By disaggregating power-sharing and applying the largest dataset of

its kind, this dissertation contributed new theoretical, empirical, and policy insights into how

power-sharing institutions shape democratic outcomes. The findings also call into question the

prevailing emphasis on political power-sharing institutions and highlight the advantages of

constraining power-sharing, which remain considerably under-researched.