Electoral Violence in Democracies, 1960-2000.

The difficulty of democracy is to ensure that the defeated at the poll obeys the results of elections. This is a difficult task, since the defeated has an incentive to engage in armed conflict to modify or annul elections. This paper investigates under what conditions elections lead to armed conflicts onset by considering a decision-making problem of the defeated who chooses to wait for the next election or to engage in violent conflicts. We argue that the compliance on the electoral outcome is dependent on following two factors; (1) the type of elections (presidential and legislative democracy) and (2) amounts of natural resource rents. First, the type of democracy directly affects the stakes of elections. Since the president exercise far greater discretionary power than the executive in legislative democracies does, the losers in presidential regime are more likely to resort to violence. Second, as the rents derived from natural resource increases the amounts of resource which the state monopolizes, the defeated is more likely to challenge the government. We test these hypothesis by employing new data set which disaggregates observations from country-year to country-month. The results show that armed conflicts are more likely to take place after presidential elections, and this destabilizing effects of presidential elections are more magnified by the oil production. [ABSTRACT FROM AUTHOR]