[EM] Trees and single-winner methods

[Abd ul-Rahman Lomax]

At 04:07 PM 3/15/2007, Juho wrote:
>On Mar 15, 2007, at 17:18 , Abd ul-Rahman Lomax wrote:
> > One person's horse trade is another's sensible compromise.
>
>I agree that the horse trading has also some potential to lead to
>good results. It however carries the risk that some of the negotiated
>topics might be related to the very personal needs of the candidate.

This risk is intrinsic in representative democracy. What Asset Voting 
does is to provide voters with the freedom to delegate the candidate 
selection process. They are not necessarily required to do so; a 
ranked ballot, for example, counted as IRV or Condorcet, with the 
provision that if the ballot is exhausted (because the voter 
truncated). the vote becomes an Asset of the first-ranked candidate, 
would qualify as Asset Voting, but it allows the voter the additional 
freedom to specify vote reassignments as deeply as desired.

However, it's a freedom that I would personally avoid. The rigidity 
of such is precisely what can lead to bad results in IRV, for example.

>This potential is of course in some form present everywhere where the
>representatives can make major decisions. In this case they however
>play with their personal interests (to become elected) and the
>candidates are expected to trade (no clear ideological driver that
>would force them to vote in some predetermined way).

There is, again, an assumption that the "candidates" are seeking 
office, and that they personally desire to serve (as distinct from 
being willing to serve). Under Asset, "candidates" are electors. It 
happens that they can vote for themselves, and the system may be 
designed such that if they win at the outset, they are simply 
elected, but it would be simple to assign them the votes and allow 
them to cast them as they choose.

The point is that I can choose whomever I trust, without needing to 
consider electability or other complex and vexing issues. I can vote, 
presumably, for one candidate only, or I can create a virtual 
committee (under Fractional Approval Asset Voting).

>I also think that the following scenario is not nice. Based on the
>ballots some candidate seems to win the election, and all the
>supporters are already celebrating - until some of the candidates
>announces his withdrawal in order to get some candidate that suits
>him (personally) better elected.
>
> > Suppose that an election is held and there are three candidates.
> > Let's suppose that the rules are approval and/or rull ranked,
> > perhaps Condorcet, but, as alleged by some could happen, everyone
> > bullet votes. And the three candidates have equal support. What
> > would you do with this election? Elect the candidate with the most
> > votes, even though that would mean electing a candidate who was
> > only approved by one-third of the voters? Or if there was an exact
> > tie, choose the winner by lot, which still has the same result -- a
> > minority-approved winner. And it could be a lot worse if there were
> > more than three roughly equal candidates!

First of all, a candidate can withdraw, generally, under any election 
method that does not involve enslaving the candidate. It's rare, but 
they do it. Under current law, this gives the candidate no power 
whatever in choosing his or her replacement.

However, what is missed by many who consider this is that when we 
cincerely vote for a representative to serve in an assembly, or an 
officer to hold an office, we presumably trust that person more than 
the other possibilities to serve properly or beneficially in the 
office. And typically this involves a lot of delegation of 
responsibility. Persons who can't delegate well, who can't choose 
others to do the necessary tasks, usually make very poor choices for 
high office. An essential skill for high office, to summarize, is the 
ability to choose well in delegating responsibilities.

Asset Voting simply uses this; it assumes that if we would vote for 
someone for the office, we would trust that person to choose 
reasonably well a replacement for himself or herself if he or she is 
unable to serve for whatever reason. If actually elected, this is 
really what is going to happen with respect to much that is covered 
under the duties of high office.

And given that voting under Asset becomes totally free of the need 
for strategic considerations: just vote for the candidate you most 
trust! -- I should be able to focus entirely on candidate 
qualifications. Given, again, that there is no need that the 
"candidate" actually be elected or electable, I can choose a 
candidate whom I personally know. I expect the numbers of candidates 
to blossom if Asset is adopted. And the result will be much closer to 
what a hiring search would produce.

I've often pointed out how silly it is to elect officers to serve us 
based on their campaigns. No company would hire in this way! (Or, 
certainly, they would heavily discount such campaigning, realizing 
that if someone is trying so hard, they probably have a personal 
agenda, they expect to greatly profit if they are hired. It might not 
just by through salary....)

What we would do is to *interview* candidates. But we can't do that 
because of the scale. Asset, like its close relation Delegable Proxy, 
reduces the scale. That is, votes are reassigned by people holding 
many more than one vote, so it becomes increasingly possible for 
those recasting votes to interview those for whom they are considering voting.

So while it's possible to provide a vote reassignment list ("ranked 
ballot") with Asset, I personally would avoid using it. I'd rather 
have the flexibility of intelligent reassignment based on the 
presumably superior knowledge of someone with more opportunity to 
investigate directly the candidates.

This first level vote would not necessarily be for one who I consider 
qualified for the office. If an office requires quick response, for 
example, someone who was slow and methodical might not be a good 
choice. But such a person might make an excellent elector. If the 
person was intelligent and trustworthy and would know not to elect 
himself or herself.

And, of course, because Asset reduces the scale, I can vote for 
someone that I either personally knew already, or I can meet as part 
of the process.

Asset really does carry many of the characteristics and potential 
benefits of Delegable Proxy.

> >
> > Now, suppose that a candidate can reassign his or her votes to
> > another. If any two of these candidates can agree on who should
> > win, we'd have a winner who, for two-thirds of the voters, it is
> > true that they either chose the winner or the winner was chosen by
> > someone who they preferred as the winner. That seems to me to be
> > *far* better than choosing without such a reassignment.
>
>Note that the votes could have been:
>17: A>B>C
>17: A>C>B
>17: B>A>C
>17: B>C>A
>16: C>A>B
>16: C>B>A
>
>Candidates A and B agree than the winner is A. The "B>C>A" voters may
>be very angry to B.

This is assuming that all voters have voted a ranked ballot. I 
consider that highly unlikely, actually.

There is an anomaly here. The idea is that very many voters who 
prefer B more than any other, apparently distrust the judgement of B. 
Why would they vote for B?

Perhaps B was dishonest. But Asset makes it possible to vote for 
candidates you would know fairly well.

If A and B agree upon the winner, I have no problem whatever with this outcome.

Consider if this single-winner election were converted to a 
three-winner proportional representation scheme. It would elect A, B, 
and C. And this would actually be quite accurate.

Now, if A and B are congenial, then A and B will be a voting block 
with twice the power of the C "block." If there were only three 
members in this "assembly," A and B, where they agree, could make 
whatever decision they like.

Asset is only taking this concept and using it to elect a single 
winner. If this were an electoral college, quite clearly, A and B 
together (or any other pair) could clearly choose the winner.

The problems with the U.S. electoral college have to do with how 
electors are chosen, through unfair state-level rules, plus some 
degree of inequity in the numbers of electors. Vote delegation isn't 
really the problem, we wouldn't be debating the elimination of the 
college if it were, for example, a good PR system. Such as Asset.

> > And it could get even better if the candidates holding
> > redistributable votes are not limited to candidates who were in the
> > original election. That is, two of the original candidates could
> > possibly agree on a *different* candidate who, had this candidate
> > been in the election, would actually have gained a majority. Or at
> > least the two think that this is a good compromise.
>
>Note that voters that would have voted "A>B>C>D", "A>B>D>C" etc.
>could be very angry if A and B agreed to elect D.

And the A voters described, if A is elected and votes for something 
they detest, are similarly going to be upset.

*We are delegating voting power, or other forms of authority, no 
matter what.* That's why I find this argument particularly frustrating.

I described Delegable Proxy to a neighbor, one time. She commented, 
"Oh, I could never trust anyone with my vote." As if she was not 
already doing that, only with very little choice about whom to trust!

We strain at a gnat and swallow a camel. I'd suggest that someone who 
is worried about delegating voting authority should be *really* 
worried about representative democracy in general. For that is 
exactly what it involves.

And when the scale gets large, direct democracy is generally 
considered to become impossibly unwieldy.

However, there is a solution that is part of the BeyondPolitics.org 
proposals: leave voting rights with the members, and simply *allow* 
them to be delegated; restrict only *deliberation* rights, i.e., the 
right to speak to the assembly and to enter motions.

With such a system, *generally* those voting would be those who were 
participating in deliberation (assuming that the assignment of 
deliberation rights goes with, say, the number of people represented 
directly or indirectly), preserving the principle that people should 
generally be informed about the subject of their votes, while 
reserving the ultimate judgement as to competence to the voter.

>In places where direct democracy or electing the representatives
>(directly) is not wanted but more indirect decision making structure
>is needed, then a true electoral college is an option. I think such
>need of indirectness in democracy is not that common. Bargaining will
>happen anyway already between the elected representatives. The
>electoral college level might be an unnecessary additional level.

It was never implemented in a way that would allow its real function 
and intention to show, it became, more or less, a rubber stamp for 
decisions made elsewhere.

If an electoral college were not limited to the existing candidates, 
if, indeed, it were completely independent of the candidates and 
could search for and select a winner or winners, with complete 
discretion, we would see a drastic uplift in the quality of winners.

And suppose there was a DP electoral college. In the end, you could 
vote directly, if you wanted to.

Remember, though, that the BeyondPolitics vision is not to try to 
implement this ideas immediately in poltical structures, but rather 
in Free Associations where power remains with the members. The 
expectation is that we would learn how it works -- and how it does 
not -- before trying out DP in political structures.

And, the benefit: such Free Associations could conceivably become 
effective enough to be able to implement political changes. FA/DP 
organizations create networks of trust, and, as such, could mobilize 
very substantial voting resources. By design, the overall FA doesn't 
take controversial positions, *but* the DP structure creates what we 
call "natural caucuses." These caucuses, in an FA, are naturally 
motivated to seek as broad consensus as is possible within the FA 
before attempting to act independently. When they act independently 
they are not restricted by the FA rules, which do not bind them.

I continue to be amazed at how thoroughly Bill Wilson worked out some 
of this when he wrote the Traditions for Alcoholics Anonymous: "For 
our group purpose we ought never be organized, but we may create 
service boards or committees directly responsible to those they 
serve." This can take a little translation. "Organized," in this 
context, refers to the traditional mechanisms of power: strong 
central treasury with substantial discretionary reserves, centralized 
decision-making and control. There is no "AA Inc." There is only 
AAWS, Alcoholics Anonymous World Services, Inc., which is a service 
corporation, it does not control AA, control is totally 
decentralized, every meeting is independent. Most of the money 
collected by passing the hat at AA meetings pays the rent for the 
meetings, the coffee, and for publications to be distributed, 
typically without charge. A small amount of that goes to Intergroup, 
typically, for local organizational expenses, mostly of maintaining a 
meeting list and in a few places a local office. And a small 
percentage of that goes on to AAWS to help with similar functions at 
the national and international levels. And all of these close 
organizations depend totally on the immediate and continued support 
of the members. They are not allowed to amass more funds, for 
example, beyond a "prudent reserve." Which for a corporation like 
AAWS pretty much means what the law requires: enough money to satisfy 
all debts if they needed to shut down.

But AA does spawn organizations that do (1) take controversial 
positions and (2) amass resources. For an example of the first, AA 
members often are the major participants in Alcoholism Councils which 
lobby for legislation and support of the needs of the alcoholic 
population. And for the second, AA wouldn't operate a formal 
treatment center. But, again, AA members can and do start such 
organizations, including quite large ones. They are not AA, AA is not 
responsible for them and they are not controlled by AA. But if not 
for AA, they would not have existed.

It's really a libertarian model, if you look at it.

The DP structure in an FA/DP organization sets up certain people who 
will have large networks of trust. It pretty much doesn't matter what 
the original purpose of the organization is, these networks could be 
quite interesting in how they would function.... where a few people 
represent many, and *really* represent them, having been freely 
chosen, it becomes possible to negotiate and deliberate on behalf of 
a large population. If voting remains a right of the general 
membership, we would have succeeded in creating what has been thought 
impossible: direct democracy on a large scale. The trick is in using 
representatives for *deliberation*, which is where direct democracy 
breaks down when the organizations get large. Then, only members who 
consider themselves qualified to vote, and who have the time to do 
it, actually vote. The rest simply allow their proxies to do it for them.

Now, how come I'm usually about the only person writing about this? 
Is it really such a bad idea?

(There are now a few interesting people on board, but, still most of 
them aren't very active yet.)

> > I disagree with those who refuse to consider deliberative process
> > as election methods. [...]
>
>Maybe the most common way to implement a deliberative process is to
>use a some proportional method to elect multiple winners and then let
>those representatives negotiate and decide.

Yes. It is called Asset Voting. It is *fully* proportional because 
the delegated voting power remains undiluted, without any loss or averaging.

But it also then can be used to create a PR Assembly that is not at 
all party-based, nor does it have fixed districts, though, as I've 
noted, it would probably result in most voters, if they vote for 
someone local, having a local representative *whom they chose or who 
was chosen by someone they chose*.

This assembly could then elect officers. I.e., a parliamentary 
system, with the assembly acting as a peer electoral college.