[EM] voting reform effort in DENVER - PLEASE HELP

[Abd ul-Rahman Lomax]

At 03:45 PM 6/11/2006, Chris Benham wrote:
>That assumes that the interests and/or preferences of voters are 
>identical to the candidates they
>vote for.

This is an argument against representative democracy, isn't it?

Mr. Benham is assuming, first of all, that "candidates" in an Asset 
Voting situation would be the same kind of candidates as we have 
today. In prior discussions of Asset, it has been pointed out that we 
would probably see *many* more candidates, so many that ballot design 
could become an issue. In Asset, as I'd implement it, one does not 
have to be a candidate to be elected. "Candidates" in Asset are 
really candidates for the position of elector; and they *may* elect 
themselves, or another candidate, or anyone eligible for the office.

The interests and preferences of voters are certainly not identical 
to that of anyone other than the voter. But they could be similar, 
and under Asset, the voters could vote for the candidate who they 
believe will most closely represent their interests. I've entered 
this discussion many times. I don't expect politicians to advance my 
personal interests. I expect them to consider the interests and 
welfare of the whole society and to come to their best judgement and 
decision regarding it. I assume that they will be, as professionals 
dedicated to the job, generally better informed than I. Surely 
information can shift preferences.

However, I do expect them to *listen*. And it could be that I know 
something they have overlooked. When a good politician has a large 
constituency, he or she will prudently have good staff to filter 
communication. "Good" is the key word here. It means staff that is 
trustworthy and knowledgeable enough to recognize new information 
from what is redundant....


>>What if voters want to assign this power to candidates?
>And what if they don't?

Then they vote for a candidate who thinks like them, who will not 
reassign their vote. Of course, that would be a weak action, when 
other voters can take stronger action. However, remember 
candidate-list? Some candidates could publish their preference order. 
It is even possible that preference order could be written into the 
law; that is, a candidate who publishes such an order would be 
constrained to follow it, until it was exhausted.

>>Fundamentally, the position expressed here is that voters are not 
>>to be allowed this freedom.
>>
>No, the shoe is on the other foot.

Mr. Benham is clearly arguing that votes should not be delegable. The 
only reason we don't clearly see this immediately as a loss of power 
for the voters is that we are accustomed to it. We are almost 
universally disempowered in this way. Has it ever occurred to us that 
people of wealth delegate authority when they can? If vote delegation 
is a bad idea, why is it universal in business? After all, the 
interests of proxies are not identical to the interests of the 
shareholders they represent.

But proxies attempt to serve the shareholders they represent, if they 
are worthy of the position. The same is true of politicians. Voters 
can choose the politicians who accomplish this the best. And they do. 
But this same process is not allowed to take place in public 
elections (it does take place in corporate elections). Why?

I have *never* seen a cogent explanation. It's like Approval Voting. 
Overvoting is prohibited, almost universally. Why?

>>Mr. Benham has not stated why vote delegation should not be permitted.
>Because that is a completely different issue.

I must be confused. He argued against asset because it allows votes 
to be delegated to candidates. There is nothing in Asset that 
restricts "candidates" to those who will actually serve in the 
office. Indeed, Asset allows *all* candidates who receive votes to 
serve a function; they are, quite precisely, electors.

>>Note that, ultimately, power *is* delegated, with any election 
>>method. To the winner or winners of the election. So why not during 
>>the election process itself?

>That doesn't make any sense at all.

To Mr. Benham. He's right in that sense.

>   The voters have little power, occasionally getting to elect
>"representatives"/office holders with no (or next- to- no) control 
>over them between elections.
>The power to choose these office-holders  (directly, by their votes) 
>is obviously greater than
>the power to choose who chooses.

It may be obvious to Mr. Benham, but it is not obvious to me, and, 
indeed, it is incorrect. The power to choose who chooses is greater, 
in general, than the power to choose. The power to choose, to be real 
power, is dependent on many factors that are quite often absent. 
Deliberative process is required for genuine choice to be exercised.

Asset can be used single-winner, and vote delegation is only relevant 
when a candidate does not gain a majority. If a majority of voters 
choose a candidate, that candidate wins, quite simply. (The simplest 
form of Asset is actually standard single-vote with revoting allowed. 
This could be done with any election equipment. It is only the more 
complex forms of asset that require more complex counting. FAAV, 
Fractional Approval Asset, would use existing voting equipment but 
does require more complex calculation when the votes are summed. It 
would be best, I'd suggest, if full ballot data is transmitted.)

Voters may have sufficient information to make a primary choice as to 
preferred candidate. But being required to make intelligent choices 
about a whole series of candidates requires much more information and 
analysis. My own opinion is that if voters are allowed to make a 
choice of the single candidate or candidates whom they would trust to 
either serve in the office, or, if they are not going to serve, to 
help choose who will, they will make quite good choices, especially 
if they are not constrained to "major candidates." Asset allows this.

>In my view the aim should be to increase the power of ordinary 
>voters, not decrease it at the
>expense of candidates.

I suggested, above, a means whereby Asset would *clearly* increase 
the power of voters, allowing voters to choose candidates who had 
published preference lists. It is also possible that ballots could 
allow ranking. I'd think the way it would work would be that if a 
voter votes for only one at first rank, that one vote would be 
revotable by that candidate. But if the voter adds ranks, the revotes 
would be in that sequence; however, if the ranks were exhausted in 
the counting process, I'd have the vote revert for revoting to the 
first rank choice(s).

But, frankly, it's a complication that I think would not add to the 
actual power of the voters. It is already true that winners of 
elections are going to exercise delegated power, and we assume that 
this is better than direct democracy, when the scale is large. Now, I 
have proposed and am actively working for a method of dealing with 
the problem of scale in democracy, but this is outside the scope of 
this post. Assuming that we are going to have elected 
representatives, Asset Voting allows the same deliberative process 
that we trust for actual government process to be used in the election process.

Aggregative methods, i.e., voting, are not particularly good for 
making optimum choices. The best method among the election methods 
for this is Range; in my own opinion, there is a problem with culture 
and Range. Range will work best when people simply make honest 
appraisals of the relative value of the election of candidates; but 
the culture is to distort and exaggerate preferences in order to 
"win." Of course, this distortion causes election results that do 
*not* optimize value to the society; and the whole process is quite 
easily influenced by the corrupting power of special interests with 
relatively substantial resources.

I'm a voter, and I'm claiming that my power is restricted because I 
cannot delegate my voting power. Nothing says that I *must* delegate 
it. An Asset Voter could write in his own name. Why not? But most 
would not, I think, even if they understood that it was allowed. It 
would generally be a bad idea, unless one is an expert, one who knows 
the candidates and the issues well.

To me, the core of democracy is consent of the governed. It is not 
essential to democracy that decisions be made directly by the 
electorate, but consent is essential. Election methods, generally, 
except for very fine-grained PR, don't allow the individual voters 
true representation, unless they happen to be in the majority.

Consider this: is the interest of the voter identical to the interest 
of the majority? I think it is inexorably true that majoritarian 
elections result in government contrary to consent.

>Approval is much more vulnerable to disinformation than any 
>reasonable ranked-ballot method,
>including IRV.  Because of disinformation, it is possible that the 
>sincere strict favourite of more
>than half the voters can lose.

Sure. Theoretically. And the important qualification is made here to 
make the statement true: "Strict" means that *any* difference in 
preference is in effect.

>In Approval, voters can feel at sea without some "information" about 
>the viability of the candidates,
>but with the (ok- to-good)  ranked-ballot methods they can fall back 
>on just ranking sincerely.

Voters who don't have any information about the viability of 
candidates ... I question how informed these voters are. Under Asset, 
they only have to know *one* candidate, that they trust that 
candidate. As I've written, they could choose anyone they trust. And 
if they don't trust anyone, they can choose themselves.

But choosing themselves, if they only represent themselves, is not 
likely to have much impact on the outcome of the election.

>>Asset Voting, in fact, is the least vulnerable, since votes end up 
>>being ultimately distributed according to the decisions of trusted 
>>candidates who have better access to information than the general 
>>voter. With Asset Voting, a voter need make no strategic decisions; 
>>it is enough to find a single candidate that one trusts; but Asset 
>>also allows voters to distribute the trust among a set of candidates.
>I thought that one of the problems of US democracy is that 
>candidates are generally not very
>trustworthy.

Not in my view. The problem is the system, not the candidates. The 
system largely requires candidates to pander to perceived interests 
and knee-jerk reactions when they are running for office and often 
more or less continuously. Many manage to do a good job in spite of 
this. But when they vote for what they actually see as best, they 
will be viewed by this or that part of their constituency as having 
betrayed their "promises."

The real problem is that they are expected to make promises in the first place.

(Well, there are lots of real problems. But that politicians break 
their promises is probably one of the saving graces of the system.)

My own policy and suggestion is to vote for candidates judged as 
trustworthy, not candidates who promise to give the voter exactly 
what the voter wants. I'd assume that a trustworthy candidate would 
either give you what you want, or would explain to you why it's not a 
good idea. A "politician" -- the kind which is all too familiar -- 
would never do the latter, too risky.