Fiona McGillivray and Alastair Smith. 2005. The Impact of Leadership Turnover and Domestic Institutions on International Cooperation

Fiona McGillivray and Alastair Smith, “The Impact of Leadership Turnover and Domestic Institutions on International Cooperation,” Journal of Conflict Resolution 49 (5) (October 2005), pp. 639-660.

The differences in cooperation between two nations can be attributed to the domestic institutions of those nations if actors within a nation employ leader-specific punishment strategies. The leader of Nation A can punish the leader of Nation B for Nation B's non-cooperative policies by refusing to cooperate with Nation B until the leader is replaced. The citizens of Nation B will lose out on cooperative benefits and have an incentive to remove the non-cooperative leader. Then in cases where leader removal is easy, as in the case of large winning coalition systems such as democracies, leaders will have reduced incentives to defect and states can cooperate deeply. Instances of cooperation breakdown between such nations are rare. In contrast, when one of the nations attempting to reach agreement has a high cost for leader removal, cooperation is shallow. Instances of cheating lead to prolonged periods of non-cooperation.

Author: Fiona McGillivray and Alastair Smith

The Point: The logic of the theory derived here relies solely on the institutional differences between different regime types. The differences in cooperation between two nations can be attributed to the domestic institutions of those nations.

Lit Review: Audience cost theories have also been proposed to account for the higher level of cooperation between democratic states, in which electoral accountability forces democratic leaders to pay higher domestic political costs for breaking commitments than autocrats (Leeds 1999; Bueno de Mesquita and Lalman 1992; Fearon 1994; Schultz 1998, 1999, 2001, 2002; Smith 1998). But these arguments have no theoretical origin.

Core Analogy: International cooperation is like an iterated prisoner’s dilemma game.

Argument: Cooperation can be encouraged by leader-specific punishments. If the leader of Nation A only cooperates with the leader of Nation B until the leader of Nation B defects and vice versa, then the citizens of the defecting nation lose out on T rounds of cooperative benefits until the leader who defected is removed. It should be easier to cooperate with a nation whose leader is easier to replace and prolonged periods of punishment following cheating should be only likely when leader replacement is difficult.

Variables:

P_i = Protection of Nation i (an indicator of cooperation in which P=100 is full non-cooperation and P=0 is full cooperation)

Q_i = the protection level of Nation i as perceived/observed by nation j

H_j = the threshold at which Qi must rise above in order for Nation j to stop cooperating with Nation i

Model Design: The model is a modification of Downs and Rocke’s (1995) continuous-choice prisoner’s dilemma in which all the actors in one nation have identical preferences with respect to policy outcomes and receive the same payoff for policy outcomes. The payoffs for policy outcomes between two nations are symmetric.  

Conclusions: Leader-specific punishment strategies imply different dynamics depending on the cost of leader removal. When domestic political institutions are such that the cost of leader-replacement is low, as in the case of large winning coalition systems such as democracies, then states can cooperate deeply. Instances of cooperation breakdown between such nations are rare. In contrast, when one of the nations attempting to reach agreement has a high cost for leader removal, cooperation is shallow. Instances of cheating lead to prolonged periods of non-cooperation.