POLITICAL VIOLENCE AND TIME IN POWER

With this article we hope to add something new to explanations for vio lence within polities. We examine the consequences of time in power fo r political violence. We are interested in the length of time a leader is in power and the time when political violence occurs. We start wit h a simple hypothesis: the longer a leader is in power, the less likel y it is that there will be occurrences of large-scale political violen ce. We are not aware of any theories of violence that have taken accou nt of leadership duration. We do not argue below that leadership durat ion and the timing of violence explain all variation in violent politi cal outcomes. Rather, it is our assertion that theories of violence ha ve ignored an important factor: the length of time that a leader has b een in power. Throughout the analyses presented in this article we hav e found a relationship between the number of years a leader has been i n power and the probability and volume of political violence that the nation that they are governing is likely to experience. As the length of a leader's time in power increases, the probability of political vi olence declines. This relationship has been demonstrated to be signifi cant in both statistical and substantive terms in a number of differen t statistical settings that make a variety of assumptions about the qu ality of the data we are working with and the underlying relationships we are investigating.