[EM] Why I think IRV isn't a serious alternative 2

[Abd ul-Rahman Lomax]

At 05:56 AM 12/26/2008, Juho Laatu wrote:
>One approach that is used in practice and
>that to some extent avoids the problems of
>
>- "few random votes to random people"
>- difficulty to identify to whom the votes actually are meant
>- votes to people that do not want to be candidates
>- having too many candidates
>
>is to require people to collect an agreed
>number of names of supporters (and
>candidate's agreement) to get their
>candidate on the candidate list.


San Francisco has a write-in requirement that candidates must be 
registered for the vote to be counted. That's not a bad idea in 
write-in situations, and the registration should be possible up to 
the day of the election. I'd consider the idea that it could be post-facto.

And that if the non-candidate who wins wishes to do so, that 
candidate may reassign the votes, effectively choosing a replacement 
for himself or herself. Doesn't want the responsibility? Sorry. This 
person has been offered the power, and can use it or not. It's a 
variation on TANSTAAFL. We cannot avoid sins of omission by refusing 
to accept responsibility, the responsibility comes from being alive 
and having the power to act.

However, for very good practical reasons, preregistration is a good 
idea. It should be cheap or even free. I disagree with petition 
requirements, they make sense only if the name is to appear on the 
ballot. There is no harm to the process from solitary registration. 
And there *could* be harm from signing a petition. It is, 
effectively, a non-secret vote for the candidate.

We forget that some of the protections of secret ballot aren't 
necessary most of the time. We have them for the rare exceptions.