[EM] Re: parties working to throw out top-two primary in Washington State

[James Green-Armytage]

Hi Russ,

	I don't know the details of the Washington State situation, so please
correct me if I've misunderstood the new rule.
	As far as I can tell, the changes that we are talking about essentially
create a two round runoff instead of a plurality election... except that
the second round is mandatory even if a candidate receives a majority in
the first round. 
	I don't really buy the argument that third parties are excluded by this
system. If they want to win, they need to compete with the Dems and Reps
in the primary. If they can get the votes there, they have a real chance
of being elected in the general election. Compare that to a plurality
system where even popular third party candidates have a very hard time
convincing voters that they should be considered viable and hence that
they should be voted for in the general election. Winning a spot in the
runoff guarantees that they will be given a fair shot at the office.
	I also don't buy the 'unconstitutionally preventing parties from choosing
their own candidates' argument. That has more to do with regulating the
use of party names on ballots than it has to do with the voting system
itself. Basically, I think that in a two round runoff system, the parties
should be free to hold their own primaries before the first round. Then,
party members can commit to vote in the first round for whoever wins their
primary.
	Russ, as to your description of strategic incursion (basically a kind of
'push-over' strategy (see Blake's page at Condorcet.org)), it is possible
in theory, although it might be somewhat more difficult to coordinate than
you are imagining. Let me go with your Clinton-Dole-Buchanan example. If
all the Democrats deviously voted for Buchanan in an attempt to keep Dole
out of the general election, then Clinton wouldn't make it to the runoff
(general election) at all; probably the runoff would be between Dole and
Buchanan, and the Democrats would be kicking themselves in the butt pretty
hard for their bright idea. 
	So, the Democrats need to retain at least 33.34% of the vote to be safe.
And then, yes, they have maybe 10% or 15% left to throw around.
Theoretically, they could use that 10% to help some non-viable candidate
to get into the general election. But how do they decide which 34% should
vote for Clinton and which 10% should vote for Buchanan? Do they have all
of their members generate a random number between 0 and 44, somehow, and
instruct those who draw numbers higher than 34 to vote for Buchanan? This
is the way to do it in theory land, but in practice it's not very likely
to happen. Also, even if perfectly applied, that 10% might not be enough
to cover the margin between Dole and Buchanan.
	In a way, you can think of the runoff threshold as a kind of rough Droop
quota (votes/(seats+1))+1, where there are two seats to be filled. Hence,
if 100 votes, 100/3 = 33.34. Any party with 33.34% of the vote should get
a candidate into the runoff. Any party with 66.67% of the vote should get
two candidates into the runoff.
	In conclusion, I think that this two-round runoff ("blanket primary"?)
system is better than plurality, although it is probably still worse than
IRV.

Sincerely,
James 

here is a part of your previous message:
>I don't blame the parties for not wanting a blanket primary. My 
>understanding of a blanket primary is that any voter can vote for any 
>candidate of any party for any race. That's a terrible idea. If one 
>party has a non-competitive race due to an unopposed or weakly opposed 
>incumbent, voters of that party will be free to attempt to sabotage or 
>co-opt the other party's primary.
>By "sabotage," I mean vote for a fringe candidate that has little chance 
>in the general election (e.g., Republicans voting for Jesse Jackson in 
>the Democratic primary). By "co-opt," I mean vote for their own idea of 
>who should represent the party (as when Democrats crossed over and voted 
>for John McCain in the 2000 Republican primary).
>For clear examples of how a blanket primary can be grossly unfair, 
>imagine that one was in effect during the 1984 election. Reagan ran 
>unopposed in the Republican primary if I recall correctly (or had no 
>significant competition anyway). Imagine what fun the Republicans could 
>have had with the Democratic primary (perhaps nominating Jesse Jackson). 
>Ditto for the Democrats in 1996, when Clinton ran unopposed (perhaps 
>nominating Pat Buchanan). Can you imagine the groundswell of outrage 
>over something like that?