Beck and King: Improving Quantitative Studies of International Conflict

Nathaniel Beck and Gary King, “Improving quantitative studies of international conflict: A conjecture,” American Political Science Review 94, no. 1 (2000): 21.

“We address a well-known but infrequently discussed problem in the quantitative study of international conflict: despite immense data collections, prestigious journals, and sophisticated analyses, empirical findings in the literature on international conflict are often unsatisfying. Many statistical results change from article to article and specification to specification. Accurate forecasts are nonexistent. In this article we offer a conjecture about one source of this problem: The causes of conflict , theorized to be important but often found to be small or ephemeral, are indeed tiny for the vast majority of dyads, but they are large, stable, and replicable wherever the ex ante probability of conflict is large” (21).

“In short, we conjecture that many quantitative international conflict studies lack robustness because they look not only for the effects of variables averaged over all dyads, whereas in reality the effects vary enormously over dyads and are only substantively large for those already at relatively high risk of conflict” (22).

“According to our idea, international conflict data differ from other rare events data sets in two ways. The effect of any single explanatory variable changes markedly as a function of changes in the other explanatory variables…and the dependent variables are, in principle, powerful enough to predict whether conflict occurs if the appropriate model is used” (23).

“…neural networks are sometimes treated as a black box for classifyfing very complex data patterns in the absence of theory…In contrast, we hypothesize hat for international conflict data there are massive nonlinear interactive effects, and only the confluence of many causal factors leads to a nontrivial increase in the probability of war. This allows us to interpret the output of the model in a way that is useful for the international relations scholar, not simply as a black box that does a good job of classifying which observations are more or less likely to be conflictual” (27).