Actually, Arrow still applies to rated voting systems. Any rating induces a linear ordering on the candidates and Arrow can be applied. So, rated voting systems cannot be perfect using the same criteria as voting systems based on ordering of candidates. I would suspect that are other criteria that are better suited for rated systems. In particular, it would be nice to replace IIRC (independence of irrelevant alternatives) with a property better suited to rated voting systems.
Range voting has many nice properties that are very appealing. However, there are 3 major properties of voting systems that it fails to meet:
1) Majority Property: If over 50% of voters prefer a single candidate over all others than that candidate should win the election.
2) Condorcet Winners Criterion: If a candidate would win any head-to-head election then that candidate should win the election.
3) Condorcet Losers Criterion: If a candidate would lose every head-to-head election then that candidate should not win the election.
Arrow's theorem implies that EVERY voting system has MAJOR flaws. This includes range voting, instant runoff, etc.
However, I have to say that I do like range voting (in particular its reduction of regret). But it should not be considered a panacea for alls the problems with voting methods.
As a CS professor at a large university I run into many roadblocks in getting open source alternatives considered. Our administration only wants to buy commercial products and spend millions because they think they will get better support. They equate open source with having to do their own support and not getting the professional help they need. To overcome these misconceptions, the open source community needs to do a better job educated these institutions in the support models for OSS. In particular, these institutions need to be open to hiring outside consultants to provide the necessary training and support that they are used to. For most large open source projects, its not hard to find consultanting groups that can provide the help you need. And hopefully, with some of the money they save, they can contribute to the foundations funding the OSS projects they use.
Slashdot Mirror
Actually, Arrow still applies to rated voting systems. Any rating induces a linear ordering on the candidates and Arrow can be applied. So, rated voting systems cannot be perfect using the same criteria as voting systems based on ordering of candidates. I would suspect that are other criteria that are better suited for rated systems. In particular, it would be nice to replace IIRC (independence of irrelevant alternatives) with a property better suited to rated voting systems.
Range voting has many nice properties that are very appealing. However, there are 3 major properties of voting systems that it fails to meet:
1) Majority Property: If over 50% of voters prefer a single candidate over all others than that candidate should win the election.
2) Condorcet Winners Criterion: If a candidate would win any head-to-head election then that candidate should win the election.
3) Condorcet Losers Criterion: If a candidate would lose every head-to-head election then that candidate should not win the election.
Arrow's theorem implies that EVERY voting system has MAJOR flaws. This includes range voting, instant runoff, etc.
However, I have to say that I do like range voting (in particular its reduction of regret). But it should not be considered a panacea for alls the problems with voting methods.
As a CS professor at a large university I run into many roadblocks in getting open source alternatives considered. Our administration only wants to buy commercial products and spend millions because they think they will get better support. They equate open source with having to do their own support and not getting the professional help they need. To overcome these misconceptions, the open source community needs to do a better job educated these institutions in the support models for OSS. In particular, these institutions need to be open to hiring outside consultants to provide the necessary training and support that they are used to. For most large open source projects, its not hard to find consultanting groups that can provide the help you need. And hopefully, with some of the money they save, they can contribute to the foundations funding the OSS projects they use.