Thinking About Puzzles in the Study of International War

Abstract

One of the axioms of political analysis is that office-holders desire to continue to hold office and behave accordingly. A key implication of this view is that if the policies chosen by leaders are selected because leaders believe the policies will enhance their position, then the wars we see in history do not represent a random sample of all possible conflicts, but rather are a biased sample of wars that were selected by the leader of the initiator state because he or she had the expectation of a favorable outcome that would enhance their position. This leads to selection effects in the occurrence of conflict. The theoretical and empirical implications of selection effects are explored in seeking a better understanding of two prominent puzzles in the study of international conflict: (1) the non-expansion of war and (2) the question of democratic war participation.