Mansfield: The Distribution of Wars Over Time

Mansfield, E., 1988. The Distribution of Wars Over Time. World Politics.

“Political scientists and other scholars have displayed a lively interest in the distribution of wars over time. A variety of research has been conducted on the subject, and the implications of these studies are important: only through empirical research can we hope to identify patterns, trends, and possible causes of warfare…In this paper, I shall try to determine the extent to which various well-known empirical findings concerning the distribution of war over time are sensitive to the particular data set employed. I then will present preliminary tests of some additional hypotheses concerning the factors influencing the probability of war, and ascertain whether—and, if so, how—the results depend on the inclusion of all wars or only of wars involving major powers” (21).

Mansfield makes the point that different data sets about war come to different conclusions about the causes of war and the conditions of peace based on what is coded as being a war and what is not coded as being a war. This clearly affects the results of different statistical analyses. The causes for this discrepancies lies with three things: how to define war, the level at which war is analyzed and finally how wars are dated (27-8).

He fits the data of the various authors to a Poisson distribution and finds that they mostly fit nicely, though de Mesquita’s did not.

Mansfield also compares Kondratieff cycles and the onset of great-power wars. Goldstein argues that there is no correlation between either upswings or downswings in the economy and the onset of war. Other authors disagree and argue that there is a correlation between upswings and war. Mansfield argues (briefly) that there is a correlation between long-range economic cycles and wars.

He tests the relationship between trade interdependence and war. Mansfield finds that there is an increased likelihood of war when a country moves either from being more open vis-à-vis trade or less open and conflict.

He tests the relationship between hegemony and war and finds that there is a correlation between the existence of a hegemon and the increased likelihood of war, though his overall finding is that this is not significant and that the existence of a hegemon nether necessitates peace or war.