Abstract
Two data sets for interstate wars between 1495 and 1815, and between 1816 and 1990, were examined to determine how a nation's tendency to initiate a war depended on its history of winning or losing previous wars. In both data sets, the proportion of wars initiated by a nation increased with successive wins and decreased with successive losses, and the time to initiate a war after a previous win was shorter than after a previous loss. Pooling the data sets revealed that the slowing effect of a loss was offset by one or more wins before the loss, and that losers had a high probability of initiating a war 20-24 years after the loss. Warmaking is interpreted as a cultural practice that is selected by victory or defeat, much as individual behavior is selected by reinforcement and punishment. One possible mechanism for this selection process involves the combined effects of national strength based on prior wins, military influence within a nation (also based on prior wins), and military preparations that raise the odds that a dispute will escalate to war.
Get full access to this article
View all access options for this article.