[Election-Methods] Partisan Politics

[Abd ul-Rahman Lomax]

At 01:00 PM 5/18/2008, Jonathan Lundell wrote:

>>Should I infer that there is a basis for opposing regional
>>proportionality?  I ask because it never occurred to me to question
>>the wisdom of "forcing all voters to vote at their home region".
>>Indeed, even the idea of "force" never occurred to me.  I am of the
>>opinion that voting is a right and that one's home region is the
>>most logical place to exercise that right.
>
>The objection is to "spending" all of our opportunity for
>proportionality on regional proportionality; we're looking at the
>fundamental argument for PR.

One of the problems with Mr. Gohlke's approach is that "logic" isn't 
directly applicable unless a series of assumptions have been made. 
Something may easily seem "logical" when enough considerations have 
been set aside. If we think of a representative, we'd want that 
representative to be someone local, sure. When state legislatures 
were being put together, and when the federal system was put 
together, the old feudal system of regions being represented rather 
than people seemed logical. And, in fact, it worked, more or less. 
California's two Senators are the "gentlepeople from California," and 
they represent California, not the voters of California; they are 
chosen by plurality vote. Gohlke discards some of the logic of this, 
but not all.

In fact, though, it is possible that we could have both regional 
representation and representation of people. There are federal 
systems that do this, with complex schemes that assign some seats 
according to one formula, and other seats according to another.

But there is an extraordinarily simple system that cuts to the chase. 
Turns out, by my analysis, that if you represent people through their 
free choice, you will *probably* get regional representation as well. 
After all, won't most people choose someone local? If you are a 
voter, which would you prefer to have, someone who represents your 
location, or someone who represents your views? With Asset Voting, 
you can vote for both. Normally, there will be someone local who 
represents your views, or whom you otherwise trust to be fair and to 
act properly and intelligently in office. Asset Voting places 
practically no constraints on you.

Until recently, we thought that this was invented by Warren Smith, 
one busy little bee. However, when I first heard of Single 
Transferable Vote, and didn't know how it worked, I actually though 
it worked using the Asset Voting idea. In other words, if someone 
came in without the prejudices of knowing how things work, that 
person just might come up with Asset Voting. I didn't tell anyone, 
though, as far as I recall. But another out-of-the-box thinker did. 
In 1884. Charles Dodgson, better known as Lewis Carroll. It was, for 
him, simply a method of proportional representation. Only it truly 
represents the voters, though much might depend on implementation of 
details. In a party system, parties might be represented, but the 
method does not directly consider parties. It's up to the voters and 
the candidates.

Asset Voting is also similar to the very old proxy system that is 
used for corporate governance. I've remarked many times that proxy 
voting is what we get when people who have means can demand what they 
desire, or leave. Conditions have changed somewhat with broad 
ownership of stock, but, centuries ago, corporations wanted to 
attract capital and they found it necessary to cede proportional 
control to investors. Investors did not want to have to attend 
tedious and inconvenient shareholder meetings, and, by common law and 
the practice of corporations, they could name proxies to serve. 
Proxies are "elected," technically, but through any kind of 
opposition or contested election. They are chosen.

Perhaps without realizing the similarity with proxy voting, and just 
thinking about how to handle proportional representation -- he was 
certainly familiar with STV -- he realized that voters could simply 
vote for the candidate they most trust, and *that candidate* could, 
then, represent them in the process of electing a parliament. He used 
the Droop quota, I think, and any candidate with a Droop quota of 
votes was elected (I've noted that this candidate could also decide 
not to personally serve, but could instead election someone of his or 
her choice to serve instead). Just as with STV, if there were excess 
votes, those were still distributable, just as with candidates who 
did not receive a quota. This is really like STV, except that the 
reassignments aren't based on a list on the ballot, but on the 
judgement and action of candidates trusted by the voters. (There are 
also proposals whereby candidates publish a list of vote transfers, 
which might or might not be binding. But, personally, I prefer the 
much simpler system. Vote for the person you most trust, period. 
Situations and conditions change, and if your candidate ends up 
holding a few thousand unused votes, conversations might occur with 
otherwise obscure possible winners. In fact, there need be no 
requirement that the winners even be on the ballot.)

It is so totally simple, it matches common practice in other areas 
(economic life, in particular), yet ... it's extraordinarily 
difficult to get any attention about it. It's the same basic idea as 
delegable proxy, but with implementation details that create 
institutions that more closely resemble already existing ones, such 
as a peer assembly, where every member has the same voting power.

>J S Mill makes the case better than I can: 
>http://etext.library.adelaide.edu.au/m/mill/john_stuart/m645r/chapter7.html 
>[John Stuart Mill: "Of True and False Democracy; Representation of
>All, and Representation of the Majority only", Chapter 7 of
>Considerations on Representative Democracy (1861)]

Yes. The problem has long been known, and many countries have 
implemented systems that deal, at least partially, with it. But Asset 
Voting is, in fact, a total solution.

Does it have a snowball's chance in Hell?

Yes. But the path there is equally obscure to most. The path is to 
set up, outside of government, voluntary associations that operate 
using the delegable proxy principles. (And an Asset-created Assembly 
is an option that the members might choose, to me, it's a detail). If 
the theory of democracy is correct, that the power of a society 
increases with the freedom of its members, provided that they are 
efficiently organized, these organizations, once formed, will be 
successful, which will attract attention, more participation, etc.

The fact is that the primitive organizational, plurality or 
party-dependent systems we already have, if the people were 
independently organized, would be quite adequate! But we have, in 
relatively democratic societies, thought of the government as the 
organization of the people, which, in fact, is practically an 
oxymoron. It works that way sometimes, indeed often, especially in 
small communities, but with increasing rarity in large ones. In small 
communities, people tend to be cooperative rather than confrontive 
and divisive. ("Tend," there is no guarantee.) What is presented to 
voters has often been worked out with wide communication, informally. 
Elections are often unopposed, and political party affiliations often 
aren't even stated or known. But with increasing scale, this breaks 
down, badly.

There is a basic problem that was also long ago realized, and that 
realization was on the minds of the foungers of the United States: 
the problem of the corruption of judgment by power. Give judges 
executive power, and they *will* be corrupted. Allow the executives 
to judge (aside from normal discretion) and they *will* judge in a 
biased way. So we separated the powers of judgement from the 
executive, and likewise we separated it from the legislative power. 
In theory, we reserved for the people the power of the purse; if the 
people didn't like what the executive was doing, they could withhold 
funding. Unfortunately, we thought of the House as representing the 
people, when, as it was designed, it represented only the majority.

Imagine what the House would be like if appropriations were 
voluntary. I.e., the chosen -- not elected -- representatives of the 
people each had discretion to spend their quota of taxes. And, of 
course, they could set tax rates, subject to rules they developed. 
(I'm assuming here that *taxes* aren't voluntary, that they are still 
decided by majority vote; more deeply libertarian concepts are 
possible, but I'm not going there today.) They would have a choice: 
cooperate on projects of overall importance, or spend for their own 
constituents. My opinion is that such a House would generally vote 
for minimal taxes, just enough for what is needed for cooperative 
activities, i.e., true federal spending, as distinct from pork. Pork 
is inefficient, and that's why it would disappear, I believe. Why 
take away and return? Just leave it!

(But some districts pay more taxes than others, it would be noticed. 
Some of what the government does is to redistribute. And that's true, 
and redistribution is a central function or decision. In a pure 
libertarian system, the spending of each district would be 
proportional to what that district collects, and redistribution of 
wealth would be purely voluntary. But I'm not a libertarian, as such. 
Even though what I'm proposing makes possible libertarian structures 
on a scale never before seen.)

Back to the original question: districting. Asset Voting seats would, 
if the voters and the candidates desire (the candidates making the 
actual decisions), allow each seat to have an associated district. 
The districts are determined from the votes, and, by definition, each 
district is the same size in population. Gerrymandered precisely by 
the candidates putting together a seat. There is no central decision 
regarding districts. The vote is reported in precincts and, since I 
think most would perceive an advantage to having your local 
Libertarian seat rather than a couple, all of them for the whole 
state, that is how I expect votes will be transferred: mostly in 
precinct blocks. All the votes for Joe, who happens to be a serious 
Libertarian, from this defined set of precincts. There will be a 
small amount of slop, less than one precinct worth, and I'd suggest 
the slop be distributed among the districts being assigned so that, 
rather than some voters being 100% represented by Joe's choice, and 
some with no clear assignment, all would be represented 99% of Joe's 
choice. It doesn't really matter, it is just a device for making the 
voter's connection with the seats more personal. Connecting people 
with government.

And Joe himself, in the Asset systems I see, would, if he gets enough 
votes, continue to be a political player even if he does not get a 
seat: he'd be an elector, a public voter, and putting together his 
votes to create seats is only one possible function. He could, for 
example, be allowed to directly vote, so, if that's the way he wants 
it, his assignment of votes for a seat is only to gain representation 
in deliberation, i.e., for the presentation of motions and for 
debate. He doesn't have to choose someone who exactly agrees with 
him, but someone who will fairly represent his views. (Thus quite 
small parties could cooperate on someone whom they trust to see that 
their positions are represented, even if they would vote differently.)

What I see is a delegable proxy system for coordinating vote 
assignments and for actually casting non-seat votes (so that it is 
efficient); but I think that most votes, in practice, would be cast 
by seats, and a shift in outcome due to non-seat votes would be rare.

But I'm not proposing Asset Voting for public implementation. It is 
merely a vision of what is possible. What I'm proposing *now* is the 
formation of the voluntary associations that would facilitate 
communication, cooperation, and coordination among people interested 
in some topic. Any topic. It doesn't matter, for once people 
experience how this works, I suspect, it will see wider application.

Particularly relevant here, I established the Election Methods 
Interest Group as an FA/DP (Free Association with Delegable Proxy) 
organization, and it is, to my knowledge, the largest FA/DP 
organization so far. Which isn't very large, but it doesn't have to 
be large to do some useful work. And I'm about to put it to work, for 
the purpose of functioning as a peer review process for articles on 
election methods. EMIG isn't going to publish articles on anything 
controversial. That would violate the FA principles. Rather, it is 
simply going to facilitate the orderly discussion ("orderly" isn't 
intrinsic, but people actually want it; the "order" isn't fascist, it 
is voluntary) of, say, a draft article, bringing in as well material 
from, say, this Election Methods mailing list, attempt to develop a 
consensus article, if possible, and, all along the way, measure the 
consensus found. What EMIG "publishes" is the transcripts of 
discussions and the results of polls, probably reported directly and 
expanded using the proxy table.

Who is going to do the actual publication? I just said it wouldn't be 
EMIG, which frees EMIG from any responsibility for making a single 
coherent decision, which would disenfranchise the minority. FAs are 
designed to *encourage* broad participation, you don't harm your 
cause, ever, by participating (unless your goal is truly to falsely 
pretend that your position has support, in which case you might want 
to stay away and claim that the process was biased and you didn't 
participate for that reason. But the process, in fact, won't be 
biased, and that will be quite visible in the transcripts, so such a 
position could be political suicide.)

I'll originate a post on what I'm actually going to do, and all 
interested may join EMIG and the committee that I'll form to work on 
it this first article. Any EMIG member may form an EMIG committee and 
announce it, at this point, on the top-level EMIG mailing list. That 
top level list isn't for debate, generally, at least that's my strong 
suggestion. It's for coordinating committees and receiving committee 
reports and announcing EMIG-wide polls, which may be quite 
restricted. (There is an EMIG process committee that would, in my 
vision, deal with disputes over process. Like all EMIG committees, 
anyone may join, by default.)


>But of course I'll take my own shot at it, through example.

[I've deleted most of this, not because of any deficiency of it. But 
I will point out one problem, a minor one.)

>California has an 80-seat state assembly, with 80 somewhat
>gerrymandered single-seat districts. Ignoring the subtleties of quotas
>and the mathematics of PR, let's say for convenience that each seat
>represents 1/80 of the voters of the state.

Mr. Lindell showed by his qualification about omission that he 
understands the two quotas: the Hare Quota (80 seats, each seat 
represents 1/80th of the population, and that is the quota, except 
the last seat elected may represent less.) and the Droop Quota, where 
with 80 seats, the threshold for election would be 1/81 of the votes, 
and some votes are not assigned to a seat.

Using the Droop quota under Asset (which is what Dodgson proposed) 
would leave some unassigned votes going to no seat. Since I consider 
that highly undesirable (and truly problematic if direct voting of 
electors is to be allowed), I would use the Hare Quota, precisely 
(fractional vote). In some forms of Asset Voting, candidates might be 
holding fractional votes anyway, so votes would be calculated to a 
certain precision (theoretically, they could be exact, but I'm not 
sure it would be worth the complication). In order to elect an 
80-seat assembly, then, each seat would be 1/80 the total votes. 
There could be some unassigned votes that went to electors who 
couldn't agree on a seat. In a direct voting allowed assembly, this 
is actually not a problem. And if by some miracle every single vote 
got assigned, well, we'd have 81 seats. Big whoop!

(Literally! That would be a major accomplishment, worth celebration 
and the expense of an extra seat.)

Using the Hare quota means that the dregs still have their direct 
voting power. If ever "absolute majority" is used for a result, it 
would include the dregs (i.e, that one vote). So the basis for an 
absolute majority would be 81; thus an absolute majority would be 
40.5+ votes, i.e, any vote count higher than 40.5. Thinking in full 
votes, the majority is the same, 41 votes, as with an 80-seat 
assembly, but when direct votes are cast, they count for 1/Q of a 
vote for each vote the elector voting holds, and the fractional votes 
would be preserved.

One worker promoting delegable proxy found proposals in the early 
twentieth century for members of a city council to exercise, in the 
work of the council, a number of votes equal to the number of votes 
they received in the election. Nothing new under the sun. None of 
those proposals went to implementation, as far as I know, but they 
did receive some attention at the time....



>  As a voter, I'd like to be
>able to form a voting coalition with enough like-minded voters to
>elect a representative. Depending on how strongly I feel about which
>issues, how likely is it that I'll find enough like-minded voters
>within my district to send a representative to Sacramento? Not very
>likely, unless my some stroke of luck my interests happen to be
>aligned with the major party with a (probably gerrymandered) majority
>in my district.
>
>A Republican voter in San Francisco has no chance of direct
>representation in Sacramento, nor does a Democrat in Redding. Nor does
>a Green or Libertarian anywhere in the state, even though both parties
>have in aggregate enough members to justify 1/80 seats.
>
>A typical STV proposal for the California assembly has multimember
>districts of 5-10 seats, preserving a degree of geographic locality at
>the expense of raising the threshold for minority coalitions. Notice,
>though, that if the state were treated as a single 80-seat district,
>there'd be nothing under an STV system to prevent voters from forming
>geographically (vs party or issue) based coalitions. The difference
>with that these geographic coalitions become voluntary, based on
>common geographically based interests; they're not imposed (forced) on
>the voters by the district system.
>
>So, "forced" in that respect.
>
>----
>Election-Methods mailing list - see http://electorama.com/em for list info