[Election-Methods] Primary Elections using a "Top 2/Single Transferable Voting Method"

Abd ul-Rahman Lomax abd at lomaxdesign.com
Sun Dec 16 19:40:03 PST 2007

At 06:44 PM 12/16/2007, Don&Cathy Hoffard wrote:
>Primary Elections using a "Top 2/Single Transferable Voting Method"
>For more information on the STV method see 
>My objective was to find a method that would:
>  1. Narrow the field of candidates in the General Election to a 
> reasonable number.
>  2. Follow the US Supreme Courts ruling that only Party member can 
> select their Parties Candidate (the free association principle).
>  3. Allow the maximum flexibility by voters (subject to 2 above).
>  4. No vote is wasted.

The method proposed is missing quite a number of specifications. The 
title implies that it is primaries being described, and primaries, by 
definition, are political party processes that determine a party's 
nomination (usually only one per office; President is unusual in that 
a *ticket* is nominated, but the primaries only (collectively) choose 
the Presidential candidate, who then presents, I think, a choice for 
Vice-President for ratification by the party.

Rather, the method shown seems to be running a top-two STV election 
to end up with two candidates on the ballot. Who might be from the 
same party (though that would be unusual). This is *not* a primary, 
so the whole idea is off.

It looks like the first election described is also a General election 
-- that is, everyone votes in it. Now, if everyone votes in the first 
election, and if one candidate gains a majority, why even hold the second one?

Primaries exist for party use; they determine who a party will 
support in the general election. It's an essential aspect of the 
system that the party's own process chooses that candidate. Parties 
can use mechanisms other than primaries.

Definitely, the primary system is problematic. Using the method of 
assigning a state's delegates to those pledged to particular 
candidates, in the first vote, based on a plurality in a 
many-candidate election is definitely problematic. Assigning the 
delegates, however, using STV is not a bad method. Essentially, the 
number of delegates could be the state's electoral vote count, or 
some other measure.

However, that requires STV to be adopted in the state involved, which 
requires procedural modifications that are problematic. There is a 
simpler way to go: Approval Voting, and probably something like 
Proportional Approval Voting, or perhaps Reweighted Range Voting, 
which can be used for Proportional Representation. If delegates are 
being assigned single-winner for the state, like the electoral 
college vote, then simple Approval Voting is really an excellent 
method to use, because Approval can theoretically find the candidate 
with broadest acceptability, thus the party can more likely unite 
behind the winner. And it has the great benefit of being extremely 
simple to use, with essentially no implementation cost.

My own opinion is that state parties should directly elect delegates, 
not Presidential candidates. Then the delegates make the choice, at 
the convention. They can actually .... *deliberate*. What a concept!

The whole primary campaign wastes a huge amount of effort that could 
be directed to the actual presidential campaign.

More information about the Election-Methods mailing list