4 ms·
You missed a criterion: All voters' preferences are expressed through a ranking. As an impractical example, if you could get every voter to tell you exactly ho
by rspeer 10y ago
You missed a criterion: All voters' preferences are expressed through a ranking.
As an impractical example, if you could get every voter to tell you exactly how many utils a particular candidate is worth to them (maybe through complex incentives), as a quantity and not as a ranking, you would end up with a perfect voting system.
As a more practical example, approval and range voting don't fall victim to Arrow's theorem.
- tzs 10y ago> As an impractical example, if you could get every voter to tell you exactly how many utils a particular candidate is worth to them (maybe through complex incentives), as a quantity and not as a ranking, you would end up with a perfect voting system. Would end up with a perfect system, or could end up with a perfect system? I assume that latter, because I can easily design a system that uses that extra data and is far from perfect. Assuming could, not would, my question then is that known? Or is it just known that Arrow's theorem does not apply to such systems, and so we cannot rule out that such a system could be perfect?