Trade and war in cellular automata worlds: A computer simulation of interstate interactions

This study investigated the impact of the micro-level factors on the pattern of the system-level properties in international relations, and the impact of dyadic and systemic factors on the probability of war. Considering that existing simulation models in international relations have focused on the conflict dimension, this study tried to integrate the cooperative dimension of world politics in order to see what would happen in the integrated model. Computer simulations, using the cellular automata framework and the complex systems theory, were designed to generate artificial datasets for testing the nine propositions related to major international relations theories.; Three major tasks were undertaken in this study. First, a replicated model of security-only world was compared to a newly integrated model of the security-and-trade world on three dependent variables—system endurance, balance-of-power, and state survival rate. The introduction of trade into the model made system endurance more probable than the security-only world. The probability of deterrent balance-of-power increased in the security-and-trade world, while there was little difference in defensive balance-of-power. State survival rate increased with the trade module, but system-level factors had more impact than state-level factors in the security-and-trade world.; Second, the integrated model of war and trade was also used to test the relationship between dyadic factors—economic interdependence and heterogeneity—to the probability of war. In this cellular automata world, the increase of economic interdependence reduced the probability of war. The type of dyads, measured by the heterogeneity in state power, had a great impact on the probability of war, too. Wars happened the most frequently in the minor-minor dyad, while major states were more prudent in starting war.; The third task concerned the impact of systemic factors—polarity and power concentration—on the probability of war. Multipolar worlds were the most war-prone, while unipolar worlds were the least. A similar result appeared with the number of major powers, instead polarity, as the explanatory variable. Also the power concentration among major powers was found to show an inverted U-shaped curve with the probability of war lowest when the number of great powers was two or three.