||The thesis attempts to explore the relationship between government change anddemocratic breakdown, based on a cross-national empirical study of 75 countries from 1951 to 2006 across all continents. A series of event-history analyses demonstrate that the average life expectancy of a ruling party in the ten years and the likelihood of democratic breakdown are in a U-shape relationship, regardless of a regime’s institutional systems, socioeconomic conditions, international relationship or environment, as well as its historical and cultural background. Both frequent ruling party changes and lack of ruling party change are detrimental to democratic survival. Statistical evidence indicates that too frequent changes in the ruling party dampens economic development and governance capacity, and increases the likelihood of democratic breakdown. On the other hand, it is found that lack of ruling party change may lead to non-constitutional anti-government activities and abuse of power by the incumbent, which increases the likelihood of democratic breakdown.