Measure 65, The Open Primary
The Open Primary is a ballot measure in Oregon whereby everyone can vote, in the primary, for anyone they want. The top two vote receivers will move on to the general.
The Open Primary supporters claim the measure is about reducing extremist and supporting moderates. One difficulty with this is that it's much easier to tell extremists apart than moderates. Moderates will split the vote, or they will have to control who runs to make sure that they don't. Rather than getting a good choice between two moderates, as we sometimes do in the general, we'll get a desperate choice between a moderate and an extremist, or worse, a terrible choice between two extremists. Citizens will have to think hard about whether to run for office based on whether, when doing so, they will split the vote -- that's taking the worst case for running (spending money, having public humiliation, and then losing) and making it worse (now you do all that, and you end up preventing someone you'd have preferred over the other choices from winning either). Nobody wants to vote strategically -- A lot of folks are painfully undecided on whether to vote for Measure 57, which is bad policy, but is better than Measure 61. A system that encourages it is a bad system.
As for what we should do instead? Well, ranked preference voting would make this system work if it were simple enough, but we shouldn't vote for this until we have ranked preference. Letting anyone choose a partisan ballot at election time gets the benefits of enfranchisement without quite as much strategic voting (though there is some, and it's not an ideal system -- in Michigan's primary this year, Democrats picked up Republican ballots to vote for Romney and cause mischief. This has happened in such states before.)
The Open Primary supporters claim the measure is about reducing extremist and supporting moderates. One difficulty with this is that it's much easier to tell extremists apart than moderates. Moderates will split the vote, or they will have to control who runs to make sure that they don't. Rather than getting a good choice between two moderates, as we sometimes do in the general, we'll get a desperate choice between a moderate and an extremist, or worse, a terrible choice between two extremists. Citizens will have to think hard about whether to run for office based on whether, when doing so, they will split the vote -- that's taking the worst case for running (spending money, having public humiliation, and then losing) and making it worse (now you do all that, and you end up preventing someone you'd have preferred over the other choices from winning either). Nobody wants to vote strategically -- A lot of folks are painfully undecided on whether to vote for Measure 57, which is bad policy, but is better than Measure 61. A system that encourages it is a bad system.
As for what we should do instead? Well, ranked preference voting would make this system work if it were simple enough, but we shouldn't vote for this until we have ranked preference. Letting anyone choose a partisan ballot at election time gets the benefits of enfranchisement without quite as much strategic voting (though there is some, and it's not an ideal system -- in Michigan's primary this year, Democrats picked up Republican ballots to vote for Romney and cause mischief. This has happened in such states before.)