Research in the last decade and a half has tested numerous aspects of the steps-to-war explanation concerning power politics and war. This chapter will review that research by looking first at the role of alliances in war, then rivalry, arms races, and lastly the impact of combining the various steps.

Chapter 5 argues that the use of power politics does not prevent war, but actually increases the risk of war. That chapter marshals evidence from the research of the time to make the case that forming alliances, engaging in repeated crises, and building up one's military all increase the probability that a crisis will emerge that escalates to war. Since 1993, each of these claims and the steps-to-war explanation proper has been extensively tested. What does this research indicate?

Alliances and War

Previous research indicated that wars tend to follow the making of alliances and certainly do not prevent war (Singer and Small, 1966; Ostrom and Hoole, 1978; Levy, 1981). The major exception to this is in the nineteenth century, when alliances are followed by peace (Levy, 1981: 598). The analysis of alliances in Chapter 5 (herein: 187) calls for the construction of a typology that would distinguish between those alliances that are followed by peace, as in the nineteenth century, and those that are followed by war.

Gibler (1997b, 2000) has been in the forefront of creating such a typology. Classical realists believe that in the face of threat, states should increase their power, and one way of doing that is by making alliances.