A trade-off result for preference revelation (Q1576476)

For a social choice function, the Gibbard-Satterthwaite theorem shows that strategy-proofness implies dictatorship. The social choice function, of course, selects one and only one alternative. Now if a rule selects up to \(k\) alternatives, there exists a coalition of \(k\) individuals such that, for each profile of individual preferences, the choice set is the collection of the top-most alternatives in the orderings of the individuals in the coalition. This means that the selection process does not use the preferences of the individuals who do not belong to the coalition. The trade-off mentioned in the title of the paper is then the following: either some choice sets are very large, or most individuals never have any role in the social choice.