Threatening Extinction? Patronage, International Monitoring, and Electoral Fraud

I address why states cheat in elections despite inviting international monitors. Nearly all elections are monitored, yet in over a third of these elections monitors catch the government cheating. Evidently, monitoring does not stop cheating in many countries. I argue that the key to explain cheating is whether the ruling party risks political extinction if it loses an election. When a party fears extinction it is more likely to cheat. In transitioning states, the breadth of the ruling party's patronage networks helps determine the likelihood of cheating, with narrow patronage networks leading to more cheating and broader networks to less cheating. A party can build support in society by creating inclusive patronage networks that co-opt important groups in society. A party that depends on the state for resources and neglects building support in society is the most vulnerable to extinction after failure at the polls. It is more likely to cheat. A party that builds support in society can remain competitive even in defeat, so it is less likely to cheat. I use a mix of regression analysis and case studies to test my argument against hypotheses derived from prominent realist and constructivist theories in International Relations. My case studies of Croatia and Albania are based on field work in both countries.