[EM] CR style ballots for Ranked Preferences

[Anthony Simmons]

>> From: Dave Ketchum <davek at clarityconnect.com>
>> Subject: Re: [EM] CR style ballots for Ranked Preferences

>> On Mon, 24 Sep 2001 06:14:41 Anthony Simmons wrote, in part:
>> >
>> > We have an interesting institution in the U.S., which
>> > illustrates the importance of arbitrary boundaries:  The
>> > electoral college.  California gets a certain number of
>> > electors in the presidential election.  If, as some people
>> > would like, California were to split into two states, the
>> > total number of electors for California would be increased by
>> > two.  Same people, same territory, two more electors.
>> >
>> > Likewise, small states get more electors than they would have
>> > if they were combined into a single state.  The effect is to
>> > make boundaries important not just because they determine
>> > which people will interact politically as groups.  They have
>> > political signicance in their own right.

>> Actually, the above assumes something that is not in the
>> US Constitution:  Presidential elections:
>>      Each state is authorized 2 electors for being a
>> state, and 1 for each Representative its population
>> assigns it.  Very few restrictions (e.g., Representatives
>> cannot serve here).
>>      Each state has a Legislature (presumably elected by
>> the people, but no words here on this topic).
>>      The Legislature in each state decides how electors
>> get picked for that state.
>>      NOTHING in there about whether the people get
>> involved in picking electors, or how the legislature might
>> decide they get involved.
>>      LOTS of room for any state to get into proportional
>> representation or other fancy election procedures.

Actually, I wasn't talking about whether the voters directly
choose the electors, but rather about the fact that how many
electors are assigned to a given patch of ground depends on
how it is divided into states.  The more states that
terrritory is made into, the more electors it gets.  Same
territory, same voters, more electors.

>> In defense of what IS in the Constitution:  This clearly
>> was agreed to when the Constitution was written -
>> presumably to give small states a reasonable amount of
>> muscle.  ...

Actually, Alexander Hamilton spoke to that.  I don't recall
whether it was about the Senate or the method of electing the
President, but he wrote in the Federalist that basically, it
sucks but if we don't do it, a lot of states won't sign up
... 

>> ...  Splitting CA WOULD create a couple extra electors
>> AND a couple extra Senators - and would thus be permitted
>> only if it was worth this pain (if I remember right, Texas
>> asked for and got permission to split into as many as 5
>> states if Texas ever thought that to be desirable).

Actually, accident of history has made California and Texas
so large, and Delaware so small, and that has partly
determined how much power they have per person.

>> > There are other similar effects.  Districts are designed in
>> > order to improve the prospects of the party that designs
>> > them.  Large minority concentrations have more political
>> > clout than the same members would have if spread out, because
>> > they have representation.
>> >
>> > It seems to be a fact that we live with, that how we divide
>> > up the populace determines group decisions when the election
>> > takes districting into account in some fashion.  In the case
>> > of elections that extend across boundaries, it would be nice
>> > to know that the choice of boundaries does not influence the
>> > outcome of the election.

>> Actually there are two classes of districting involved
>> here:
>>      Whoever designs Congressional districts, and even
>> town, city, and ward boundaries, does their thing.
>>      Election districts (precincts) normally do not cross
>> any of the above lines and, where that rule would permit
>> too large a precinct, may divide up potential single
>> precincts.
>>      Election methods had BETTER get the same results for
>> electing a mayor or governor as would be expected if all
>> the voters managed to vote in a single precinct in that
>> election.

Actually, I was talking about elections that cross boundaries
-- i.e. people in different political units voting in the
same election -- e.g. Senate elections.

>> Actually, I see that as another black mark against IRV.  I
>> expect a dozen candidacies for NY governor in 2002 (we had
>> 12 in 1998).  That means a LOT of potential vote patterns,
>> and patterns are significant in IRV.

>> BTW - candidacies is tricky.  In NY with plurality voting
>> multiple parties may nominate the same candidate.  The
>> votes are kept separate to decide if a party retains
>> ballot status; summed (fusion) to decide which candidate
>> gets to be governor.