[Election-Methods] Partisan Politics

[Abd ul-Rahman Lomax]

At 04:29 PM 5/11/2008, Fred Gohlke wrote:
>re: "Only on the (country independent) technical properties of the 
>"groups of three" method."
>
>"(If there are e.g. two parties, one small and one large, the 
>probability of getting two small party supporters (that would elect 
>one of them to the next higher level) in a group of three is so 
>small that in the next higher level the number of small party 
>supporters is probably lower than at this level.)"
>
>The significant word in the cited passage is the gigantic 'IF' that 
>opens it.  'IF' one assumes the entire electorate is divisible into 
>two parties, and 'IF' those two parties can be shown to embrace all 
>the interests of the people, it is easy to show that the parties 
>will achieve power in proportion to their distribution in the electorate.

No. Gohlke missed the point, entirely, and so misstated what Juho had 
stated. He wrote the *opposite* of what Juho had said. The parties 
will *not* achieve such power, with a system that consolidates at a 
low level, it's the opposite. The party which has a distributed 
majority will have essentially all the power, and the other party 
will not have representation *at all*.

>But, to say that is to say nothing, for the assumption is faulty.

The point is missed. Absolutely, there is no such party, there is 
what Gohlke goes on to state. But to the extent that there is a 
single issue which is considered important by a majority, so 
important that position on it influences their choice of continuing 
representative, the majority position becomes rapidly the *only* 
position represented. Now, if a single decision were being made, and 
deliberation were not important, this would be fine. This would be a 
method of discovering a true majority position.

But the discovery of existing positions is actually a small part of 
democracy. Democracy, of the functional and wise kind, is about 
*deliberation*, not about *aggregation*. Consensus does not exist 
such that all we have to do is rip off the blankets and there it is 
in all its naked glory; it is something that is constructed through 
the collection of evidence, the development and presentation of 
arguments, and *then* aggregation. Direct democracy fails because the 
process breaks down when there are too many direct participants, and, 
further, collective decision-making is only a small part of what the 
social organism must do; most of the parts (i.e., individual people) 
have personal lives to attend to. So we tend, when groups become 
large, to move to representative democracy of some kind. Gohlke is 
using a highly restricted and controlled kind of delegable 
representation -- not "proxy," since proxies are inherently chosen by 
a sovereign agent to be represented -- to put together a 
representative body, but he's using a purely aggregative technique 
(on the identity of the continuing representative) that loses 
minority representation very early on. It is designed to force the 
kind of compromise that is ultimately important in decision-making, 
but without allowing the deliberative process *on each issue* that is 
essential to democracy. The goal of representative democracy 
generally is that the people are represented in two ways: in 
collective deliberation, so that all the significant evidence and 
arguments are considered, and in aggregation, or voting, not only on 
final questions, but on all the intermediate process details that 
determine the exact questions asked.

There are several important aspects of my own work; delegable proxy 
is an idea that is actually pretty old, Dodgson came up with it and 
published it in 1886 as a method of proportional representation. (It 
looks like a modification of STV, but, in fact, it's delegable proxy, 
the "candidates" serve as proxies, who then elect the actual 
representatives in the parliament deliberatively, negotiation being 
an aspect of deliberation. Because each seat is created by the 
voluntary assignment of votes, the seats represent unconditional 
proxies assigned by voters to the candidates and then by the 
candidates to seats, and the deviation from pure delegable proxy only 
takes place with the dregs, the votes not used to create a seat. 
Dodgson used the Droop quota, so there are such dregs. I would 
probably use the Hare quota, because the dregs don't lose voting 
rights under systems I'd design, and so it is simpler to not assign 
their voting power to any seat at all if they have not been able to 
find a compromise.

In any case, the aspect that I'm mentioning here is that I realized 
that it was possible -- and desirable -- to separate voting rights 
from the right to participate in deliberation by other than voting. 
What causes direct democracy to break down isn't voting, it's 
deliberation, it's the noise, essentially, the redundant or crackpot 
arguments, and the people who, as a local pastor here put it when I 
was explaining DP to him, who have nothing to say and are willing to 
take a long time to say it.

Because the assembly in an Asset Voting system is elected by public 
electors (the candidates who received votes in the general election), 
direct voting by these electors becomes possible; when they cast 
votes, these votes are deducted from that of the seat (or seats) they 
elected, proportionally. Normally, I expect, far too few direct votes 
from non-seated electors will be too few to change outcomes, so the 
direct vote may simply be reported separately, and in a non-computer 
system might not be exactly analyzed except when it could change the 
outcome; but with computers, now, it's possible to routinely report 
the accurate votes, with the exact vote data being available for all 
who voted, including electors. Which is important, since in the next 
general election, people may want to know how *their* elector voted. 
Otherwise accountability is lost, and the electors are directly 
accountable to those people who *chose* them. That's what is 
radically new about Asset Voting and delegable proxy: representation 
is by choice, not by contested election. It is personal, not a 
compromise. And this is unconditionally true for voting, and *almost* 
true for representation, since there is a level of compromise at the 
highest level.

>It fails to recognize that, among the people, there are an 
>agglomeration of parties ... so many they defy enumeration. 
>Therefore, it is facile to suggest the technical properties of the 
>'groups of three' method will grant dominance to one party, unless 
>one acknowledges that the party is society itself, in which case, it 
>is (or, at least, ought to be) the goal of a democratic electoral process.

Which completely misses the point, which wasn't about parties, it was 
about positions. *If* the positions are organized into parties, it 
then becomes about parties. Now, will people associate into parties? 
What in the world would make Gohlke or anyone else think that they 
would stop doing this?

Parties aren't, officially, a part of the system Gohlke proposes, nor 
are parties a part of the Asset Voting or DP systems I propose. But 
in "my" system, entities we call caucuses would still exist, and I 
presume that they would formally organize independently, and thus be 
identical to present parties. Not only should we not prevent it, it's 
a part of how the people control the system, it's the fail-safe that 
is built-in. Systems that confer direct power on parties are part of 
the problem, but if all parties can do is to faciliate *voluntary* 
association and coordination, they are absolutely necessary, or the 
individuals can do is individually protest, the status quo can easily 
crush them. And if a system has *any* exploitable flaws, I assure 
you, the viruses of corruption will evolve to discover and exploit them.

>It seems fairly common among those with a professional or passing 
>interest in politics to base their assumptions and arguments on 
>artificial delineations of human attitudes and to ignore the 
>fictitious lines they've introduced, in spite of accumulated 
>knowledge that shows such boundaries do not exist.  It has long been 
>known that people vote on the basis of bread-and-butter 
>issues.  They vote on the basis of what matters to them.

Of course. Gohlke has undertaken a *slightly* out of the box line of 
inquiry. He's in a peculiar position, recognizing the shallowness of 
thinking of many, the ways in which many are locked into certain 
assumptions, but apparently unable to recognize his own limiting 
assumptions and quick to dismiss objections as being based on 
assumptions he has, often correctly, seen as limiting.

There is the old story of the blind leading the blind. It is as if 
someone somehow manages to be able to see dimly, and starts telling 
the others that they are blind, but when someone who can actually 
see, perhaps having gone through the process he has begun, tries to 
take his hand and say, "Look at this!", he protests, "You are blind, 
you should listen to me!"

Now, to switch metaphors, in addiction recovery programs, there is a 
common saying: We are all crazy, but we are not all crazy at the same 
time and in the same way, so those who are sane at that moment can 
help those who aren't. Obviously, this takes patience, tolerance, and 
listening!

>If government is to be by the people, it must, by definition, come 
>from the people.  That does not mean telling the people what they 
>want.  It means asking them.  Any electoral process that is not 
>designed to let the people make their own decisions is not a 
>democratic process.

Right. So why, I may ask, does Gohlke design a process that, 
practically every step of the way, tells the people exactly what they 
must do? No. You may *not* be represented unless you agree with at 
least one other of the three of you who are randomly assigned to a 
group, and unless the one you choose similarly agrees, and so forth 
through a fairly long chain. And if you don't decide in the specific 
period, you are not represented, period. And you can't revoke your 
decision until the next round. Shut up and go back to work, unit 342987598!

(Naturally, the "unit* wouldn't be told that, directly. But the 
*system* is doing that, and if the individual wants to do something 
other than what is assigned, tough. The Universal Voter 
Representation Act of 2099 has determined exactly what is and what is 
not Fair Democratic Representation, and you are lucky that we don't 
punish failure to compromise! In fact, next year that's on the 
agenda. Too many people are complaining that they were unable to 
agree with another in their group and hence weren't represented, and 
some have even begun to disrupt our democracy by charging that the 
group assignments are corrupt. How could they be corrupt, when the 
Randomizing Unit Assignment System is absolutely secure and 
unauthorized access is totally forbidden?)

As an *aggregation* process, with one specific issue being 
considered, it would work, though it would be colossally inefficient 
for that. But it seems that the "one issue" is the choice of 
representation at a high level. That is not one issue, that is many 
issues, and democratic process learned long ago that, for democracy 
to be real, issues being decided must be confined to Yes/No 
questions, which are themselves formed through a series of Yes/No 
questions, which are suggested through the interactive process of 
deliberation. Nearly all the various paradoxes and problems of 
elections methods arise when attempts are made to short-circuit this 
process and resolve a complex (not binary) question in a single poll. 
The latter works, often, when the question being asked is 
sufficiently well-understood at the outset, but it breaks down badly 
under many circumstances.

>Having said all this, and recognizing your preference for 
>party-based solutions, I wonder if we have reached the point where 
>we will be best served by acknowledging that we have irreconcilable 
>differences.  I have genuinely enjoyed our exchanges and the 
>challenges you have posed, but I've no wish to harangue you with the 
>repetitious assertion of views inimical to your beliefs.

Gohlke, in short, you are a fool. Juho is a very open-minded person 
who has actually been willing to consider your ideas.

"Repetitious assertion of views inimical to your beliefs" and 
agreement to disagree -- i.e., a conclusion that discussion is 
useless -- are not your only options, unless you are utterly 
incapable of considering and addressing the flaws in your concepts, 
or of at least correctly hearing the nature of the objections and 
views of others. Instead, it seems, you file the objections according 
to your own preconceived notions of what others understand. And you 
aren't correct about that.

Juho does not have "a preference for party-based solutions." Rather, 
parties stand for collections of people who share some position or 
general philosophy or simply some decision to act collectively. 
Parties can be abused and taken over by special interests, it's 
common. Good electoral solutions are not, in my view, party-based, 
and I've seen no sign that Juho thinks they should be. "Party" in his 
writing simply means a faction or subset of the voters, aligned in 
some way, *not* a specific organization. And when we wrote about a 
"minority party," we mean, essentially, a "minority view." Given that 
the question the system of 3s addresses is that of representation, 
the "view" would be a way of thinking about what makes an acceptable 
representative. To take it to some extreme, does the gender of the 
representative make a difference.

If there is *some* effect in that direction, and if all the groups of 
threes are making the decision independently, we can then expect that 
effect to be amplified through the system until it dominates at the 
top. Through a series of apparently independent decisions, made 
according to the rules that have been specified, a small effect at 
the beginning ("a narrow majority" on that issue) has come to 
dominate the result. One could even document that, for example, only 
a small margin, surely inconsequential, of gender bias existed in the 
individual group decision, surely, with such a small bias, it's fair!

And the same would be true for racial markers, social acceptability 
of ideas and ways of thinking, position on some scale like 
progressive/conservative or libertarian/authoritarian, economic 
status, and so on. The minority will tend to be suppressed the higher 
we go in the system. That's what the math was showing, and which you 
essentially avoided examining by pretending it was based on a 
prejudice about, a "belief" in "party-based solutions."