[EM] Re: simplcity of range v condorcet

Andrew Myers andru at cs.cornell.edu
Sat Aug 13 22:18:23 PDT 2005


The problem with range voting and other methods that attempt to capture voter
utility is that voters have no incentive not to lie by amplifying their claimed
utility to the maximum extent allowed, causing the method to become approval in
most cases.

If you want voters who understand the system to be more sincere, you have to
provide them that incentive.  Of course, there is already a small incentive
because if you don't vote, you effectively cast a 0 ballot but spare yourself
the trouble of voting. One can imagine more effective incentive systems.
For example:

* Give voters points to be spent on their ballot in each election, where unspent
  points carry over from election to election to be spent when the voter
  really cares about an issue.

* Charge voters money proportional to the size of their vote.

* Pay voters not to vote at all.

-- Andrew



More information about the Election-Methods mailing list