Ned Foley has an excellent summary of my technical argument about STV repeals.
I wonder if the fractional-transfer approach would mitigate vote leakage.
The problem is that voters don’t understand it, which is what spurred the original post.
I argue in the book that there are three ways to deal with this: have a multiparty system (that can dictate rankings and/or in which it is incentive-compliant for a majority to retain the system), use list-based allocation (possibly within STV), or somehow maintain a majority that can limit its own nominations.
My hunch is that similar issues (minus ranking and limiting endorsements) also might apply to D’Hondt. Thoughts?
I think that complexity is important mostly as an issue that is raised by opponents of any recently adopted method when (a) they don’t want to talk about their real objections and/or (b) they don’t understand the real reasons why a reform isn’t working out for them — they just know they aren’t getting what they wanted.
In that context, I think D’Hondt would be an obstacle to the retention of a list system that had been adopted relatively recently. (So would almost any other seat allocation formula used in list systems.) On the other hand, once voters, activists and pundits are used to a method, the issue of complexity recedes into the background.
I’d be interested in hearing some reasons why fractional transfers might mitigate vote “leakage”. To the extent that the transfer rule affects the results of preference flows at all, I would think it might increase (not decrease) the number of ballots that count against the voter’s first choice of slates. But I don’t know that.
My hunch was that FT might reduce the probability of (perceived) majority reversals. This is because it down-weights a ballot that already has seen a candidate elected. But I could be wrong. There’s math to do, and I’m bushed.