[EM] range versus condorcet & others; practical purposes

[Abd ulRahman Lomax]

At 11:24 AM 8/14/2005, Warren Smith wrote:
>Well, in our real-world-voter study of range & approval:  USA voters
>by statistically clear margins, told us they wanted to stay with plurality and
>NOT switch to either range or approval voting.

I'd suggest that the answers may have depended on how the question was 
posed. Approval *is* Plurality with one tiny change: overvotes are not 
discarded. Range and Approval are related (Approval is effectively Range 
with only votes of 0 or 1 allowed).

Always, I have seen, Approval is presented as if it were some new method. 
It is not. As I have pointed out, mostly to deaf ears, Approval is standard 
under most rules for show-of-hands elections. It only becomes an issue when 
there is secret ballot, and the most-commonly-presented argument against 
allowing overvoting is clearly in error: the idea that allowing overvotes 
violates the one-person, one-vote rule.

So if you are asking the average Joe about Approval, it would be necessary 
for some education to take place first. Otherwise the person is quite 
likely to give an answer simply based on ignorance.

>   This makes it sound
>(correctly) like range will have a hard time getting adopted!

Much harder than Approval, I'd suggest. Approval is really trivial to 
implement; I've been pushing to try to change the debate from "Why should 
we allow Approval" to "Why should overvotes spoil ballots?" I think we can 
win the latter debate rather easily.

>Now as far as I and the other pollsters could imperfectly see from 
>listening to them,
>the top reason the USA voters felt this way, was complexity.  They felt 
>range and
>approval were too complicated.  But range voting is actually quite simple.

Approval is not complicated *at all*. The ballot is identical to standard 
Plurality (though the instructions might be changed a little) and the 
ballots are counted identically. So if respondents thought that Approval 
was complicated, they had been misled. Range is complicated (not too 
complicated to understand, but much more complicated to count).

> >From this my coauthor Doug Greene concluded that
>discussing methods significantly MORE complicated than range was
>just "mental masturbation" with no hope of actual political success.

I'd agree.

>That to me is a big reason to go with range:  it seems to me to be
>by far the simplest method out there that much improves upon
>plurality voting.  Approval voting is even simpler than range (though not 
>as simple
>as plurality).

Why???!!! Approval is exactly the same as standard plurality, it is even 
easier to count. You just count the votes and total them. You don't have to 
look for and discard overvotes. It is standard plurality that is more 
complicated!

By making it seem that Plurality and Approval are different methods, the 
poll obviously misled the voters. Approval is Plurality, *except* that an 
unjust and strange rule restricting the freedom of the voter was added 
somewhere. I've been unable, so far, to find the history of the 
no-overvoting rules....

The no-overvoting rules tell the voter that if you vote for more than one, 
your ballot will be discarded. This is an example of something that can 
look reasonable at first glance. Why should any voter get more than one 
vote? But voting for more than one in a single winner election is *not* 
getting more than one vote. In the end, one will have either voted for the 
winner, or one will not. In no case will two votes count toward a  winner.

If I were at a meeting, and an election were being held by show-of-hands, 
and I've seen many such elections, nothing would happen if someone voted 
for more than one. Both votes would be counted. Another kind of 
multi-choice vote is often held: on what day should we schedule our special 
event? Nobody thinks that allowing someone to vote for more than one day is 
giving that person an extra vote! It simply allows the vote to determine 
the most acceptable date....

>   So that may be a reason to go with approval.  However,
>my feeling is that
>(1) despite the simplicity advantage of approval, USA third parties
>would be foolish to support it because range experimentally gives them HUGELY
>more votes than approval.

Range, then, is a fairly simple modification of Approval. But there is a 
better method, and surely Warren knows it, since he invented it! And I 
modified it to make it just about identical to Approval: FAAV: Fractional 
Approval Asset Voting. The ballot is an Approval ballot.

A possible twist is that the initial count is done as a simple Approval 
ballot. But because of the nature of the full count system, many people 
might vote for only one candidate. FAAV is Asset voting with vote values of 
0 and 1, just as Approval is Range with vote values of 0 and 1. The 
difference is what happens if no candidate gains a majority: and for that, 
I'd refer readers to the writings on Asset voting.

The key in FAAV is that, quite simply, the best strategy is to decide whom 
to trust: do you trust a single candidate the most, or would you prefer to 
trust a group of candidates? The simplest strategy is to decide which 
candidate you would most trust in the office, and if it were necessary, to 
name a successor to serve out the term if the candidate could not do it 
himself or herself. And then vote for that candidate, for Asset Voting 
quite simply does this: it delegates the voter's voting power to the 
trusted candidate or candidates.

Instead of voting for one, one can vote for a slate: the vote will be 
divided among them (in the recount if needed). The slate then decides -- 
each member freely and equally as far as that voter's vote is concerned -- 
where to place the vote to create a winner or winners.

There is absolutely no reason or need for "strategic" voting in FAAV. It 
would, simply, be stupid....

I really disagree with Mr. Smith that the electorate is going to have 
trouble understanding Approval. And once one has Approval, it is actually a 
fairly small step to move to Asset voting, which, in the FAAV variant, does 
not require any ballot change. In FAAV, tabulation is a little more 
complex, but only if there is no Approval winner in the first ballot. Then, 
the votes are recounted, assigning fractional votes wherever a voter voted 
for more than one. My guess is that few would do that, actually.

Then, consider this possibility: with Plurality, implement single-vote 
Asset voting. If there is a majority winner on the first ballot, that's it. 
If not, then vote-getters may reassign their vote.

You know, when I first heard "Single Transferable Vote," I thought that was 
what it meant....

Suddenly all the need for strategic voting would disappear, since your 
sincere vote would *not* be wasted. Nader could have approached the other 
two major candidates in 2000 and said: "Okay, what are you offering us?"

The electoral college was probably intended to work something like this!