The Robustness of Quadratic Voting

I use heuristic analytic approximations to consider the robustness of the Quadratic Voting (QV) mechanism proposed by (Lalley and Weyl, 2015) to collusion and variations in voter behavior. I also consider some examples of aggregate uncertainty in the value distribution and common values. While these variations typically reduce the efficiency of QV, in plausible cases the impacts are small and variations in voter behavior may actually improve QV’s rate of convergence towards efficiency. I contrast these results with other (approximately) efficient mechanism proposed by economists which are highly fragile along these dimensions.