Divide and Conquer

Tournaments are well known to be vulnerable to collusion as shown by the impossibility theorem in Ishiguro (2004), which asserts that efficient effort levels are impossible to be implemented through a collusion-proof contract. However, we argue that this impossibility is a product of simple mechanisms that prevail in collusion-proof mechanism design. In this paper, we explore more sophisticated mechanisms with discrimination and asymmetric information to prevent collusion, outlining the principle of “divide and conquer”. As a result, we establish a possibility theorem of implementing efficient effort levels, and thus break down the impossibility theorem in Ishiguro (2004).