Preliminary Draft

Artemov, Kunimoto, and Serrano (2009) show that, in quasilinear envi-ronments with incomplete information where interim preferences of types are diverse, any incentive compatible social choice function is robustly virtually im-plementable in iteratively undominated strategies. This paper offers two main results: First, I restrict attention to complete information and show that when there are at least three agents, any social choice function is robustly virtually implementable. This is a robust analogue of Abreu and Matsushima (1992a) that only applies to “non-robust ” environments, i.e., complete information. Therefore, I also show that robust virtual implementation is unchanged even under “almost ” complete information.