| Publication |
2011.
|
| Summary/Abstract |
Although the statistical literature on conflict studies has generated strong and consistent findings on the relationship of political irrelevance and dyadic democracy to conflict, scholars have paid scant attention to the interesting theoretical issue of how they matter. The authors argue that additive controls and dropping irrelevant dyads constitute misspecifications of their effects. There are theoretical reasons to believe that the impact of distance on conflict is not sufficiently severe to justify the practice of simply dropping irrelevant dyads. Moreover, they argue that political irrelevance and dyadic democracy, rather than subtracting some constant quantity, interact to impose an upper bound on the probability of conflict initiation. They find both of these arguments to be supported in a reanalysis of a prominent study of dispute initiation.
|