Open (as in Non-partisan) Primaries?

[jpmassar]

Thanks to the recent budget deal, in June of 2010, California voters
will be asked to decide whether to do away with California's current
system of primaries (which are done in the context of political
parties) and move to open or (at least according to Nate Silver) so called jungle primaries

Comments

[markgritter]

William Poundstone's book "Gaming the Vote" has an entertaining account of the havoc in Louisiana caused by their "jungle primary".

[jpmassar]

The model assumes that people run with party labels, and that prospective candidates won't do anything that would cause their
party to surely (or more probably) lose. As noted, the model is
extremely simplistic in this and all of its assumptions. I just
wanted to understand the basics.

If a person wants to run without a party label, they could perhaps
be thought of as a third party which only allows a single candidate.

[kalimac]

The analysis seems to me more to rely on the assumption (the factual assumption) that the parties don't get to decide who runs. If they did, they could prevent shooting themselves in the foot with the wrong number of candidates.

[kalimac]

regardless of the fact that one candidate may have gotten 50% or more of the first-stage vote, move on the the second and final stage.

That's not how it works in Louisiana, as I understand it: there if one candidate gets over 50% in the primary, they just cancel the second stage.

But if you leave it as a "there are two winners at this stage, no matter what," the overwhelming significance of the number of candidates from each party could be negated by one thing: the imposition of Single Transferable Voting, as used for the multi-member seats of the Eire parliament, with a pre-determined two winners.

[jpmassar]

I don't have a reference handy, but I know I read about the (non)
50% rule somewhere.

STV is also used to select City Council members in Cambridge, MA, where I used to live.