CIFeR voting. An alternative to asset voting


#21

I can’t agree that Asset should be used for anything more than the period between the election and swearing in. Giving electors the power to remove and replace currently elected officials approaches delegative/liquid democracy, which I don’t think is possible without computers and a lot of shifting in public views. I think your ideas are great, though.


#22

Counterpoint to the ‘shifting in public views’ part: recall elections.


#23

It’s too big a sell to couple recall elections and Asset all in one package for me. Perhaps some day, but this extra change also makes it harder to explain the totality of the changes made, which then makes it harder for voters to see why they should pass it.


#24

I meant that some states already have recall elections, so replacing currently elected officials when they become unpopular isn’t that radical of a concept. Though if the electors are removing the official and not the voters, that is more analogous to a no-confidence vote in a parliament. Still, it’s hardly an untested concept.


#25

We are not looking at a specific Asset implementation. However, the
thinking here seems heavily invested in fixed ideas about elections and
“officials.” Under current systems, as with the U.S. Congress,
“representatives” do not actually represent the voters, that’s obvious,
they represent districts. That makes them “elected officials,” and
they serve for a term, presumably (but not always) until the next election.

In an Asset election, seats in an Assembly would be created by a
consensus of the electors who transfer votes to them. These are far
closer to pure chosen representatives than “officials” representing a
district. They don’t represent a district at all, unless the electors
create seats that are supported by votes from a district.

Rather, they represent a defined group of public voters, “electors,”
who, in turn, represent voters who voted for them, but rather than
“vote,” how about “chose”?

This is radically different from the voting systems that were being
discussed when I got involved with election methods. Warren and some
others had proposed Asset Voting, but the potential was not recognized.
Warren’s “Asset” had voters assign values, instead of making clear
choices, and then it used transferability to make final decisions. It
was still an “election method,” but with a tweak, but without
considering the entire collective decision-making system. It was far
more complex than necessary, creating, relatively speaking, a lack of
clear delegation and direct responsibility and clarity. It could still
be done, but I would much rather see initial Asset implementations be
very simple, and vote for your choice is the best, in a system where no
votes will be wasted.

So let’s assume the system is vote-for-one, unrestricted range among
those eligible to be electors (mostly a matter of assuring that votes
can be identified, i.e., “John” written on a ballot may not be
adequate!), and the College then has means of communication to
facilitate cooperation in electing seats. Notice: “cooperation!” You
don’t like X? So what? As a voter, what you think about X is irrelevant,
unless you choose as your elector someone who will vote for X. In which
case, tough luck! You will have to wait to the next election to do
something about this. Because Asset elections could be so simple,
holding elections once a year could be very practical, they could, with
internet voting, be even more frequent, even continuous. But consider
that “down the road.” Unless . . . and I’ll give the exception below.

So, a simple implementation, the College has a means of collecting vote
transfers and a procedure that then, when a quota is reached, locks in
those votes, that seat is elected. (It is possible to leave the matter
more open than that, but we are, here, keeping it simple.)

So, then, there is a “problem.” Let’s assume the Hare quota is used.
There will be some “dregs,” votes that have not yet elected a seat. As
the number of available votes falls, the difficulty of finding
acceptable compromises will increase. There will be, then, some unfilled
seats. For equal representation, the quota actually does not matter, as
long as it is fixed and the resulting numbers work. (I used to advocate
for Hare, but Droop would probably work just as well, and /even lower
quotas/ could be used, and what this would do is to allow the size of
the elected assembly to vary some. (It is very common in large
assemblies that some seats are vacant, for various reasons. It’s
actually a feature, not a bug.)

In any case, it might take quite a while to complete all the compromises
needed to maximize representation. I see no reason to close the process
until all possible seats have been chosen. Even then, I see no reason to
“disband” the Electoral College, which might choose – or be allowed –
to continue to function until the next election.

It does not thereby become the Assembly, with legislative authority. It
simply continues to represent all the voters, unconditionally, except
for election lag. The electors are public voters, and the affair becomes
far more like Town Meeting government, where every registered voter may
show up and vote . I’ve lived in a Town Meeting town, and I never saw
any attempt made to validate voters. You attend the Town Meeting , you
may discuss and vote…

The Electoral College may advise the Assembly. We tend to think of
control and power as being the big issue. But what if there was an
accredited representative of the voters. That is what the EC would be.
Everyone represented, period. But smaller in size, so more manageable,
and it could create committees, for example. This is why I talk about
the “judicial function.” We think of courts as having power and
authority, but, as the system was designed, they advise the executive
and the legislature. They cannot actually command the executive, and
they depend on the voluntary cooperation of the other branches. They do
not have an army, and if they do, watch out! The system is being corrupted!

So, anyway, we have an EC elected, and they have, let’s say, completed
the election of the Assembly. Is there nothing more they could do?
Normally, recall elections can be created. But here we have a (probably
large) group that collectively represents the entire electorate (as of
the last election). More commonly, those elected resign or become unable
to serve. How are those seats to be filled?

When a seat becomes vacant, votes would be freed up. Could those votes
be used to create a new occupant for the seat?

What I expect is that new uses would be found for the Electoral College
that will improve the communication between the electorate and the
assembly. If those who voted for a seat no longer support the seat, and
if they are “public voters,” and we know who those electors are,
something is off if they cannot act to improve their representation.

Yes, I would incorporate hysteresis. A seat, if electors withdraw votes,
could fall below the original election quota, and there could be ways to
handle this. I would actually leave final details up to the Assembly, as
advised by the Electoral College.

Someone who accepts a seat has made a commitment to serve, may have
moved to the seat of government, if this is governmental. That’s an
argument for not allowing recall, but it goes too far, in my opinion.

I suggest that we recognize the value of government by consent of the
governed, but have never paid much attention to the problem of the
governed not actually being represented by their own choice.

Asset Voting allows the creation of a representative assembly through
unanimity.

This is radical, not merely a small tweak. (That is, each
representative is elected by the unanimity of a quota of votes, all
through voluntary choices, and completely independently of how others
vote.)

Yes, this resembles delegable proxy, which I worked on in the 90s before
realizing that Asset could be a simple practical implementation, and
almost delegable proxy. Maybe even better. (I would see the EC as
using delegable proxy for efficiency, and voluntarily. Proxies may
advise, or may actually exercise votes. I prefer advisory systems, but
also consider the assignment of voting rights to be within the natural
rights of free citizens. They are recognized in business and corporate
governance, but are mostly prohibited in public systems, and also small
organizations generally prohibit proxy voting, my sense is for
historical reasons that avoid large benefit to avoid small risks.)


#26

Notably, making asset transfers revocable severely weakens the ‘politicians who might betray you decide who wins’ argument against Asset Voting.


#27

One question: instead of quotas, would you be okay with consecutive deadlines where the candidates with the most votes get picked, regardless of quota satisfaction? To add to the recall idea, the candidates who no longer make the list of highest vote getters get taken out of power.


#28

No. This almost completely misses the point I have been making,
that Asset creates unanimous elections of representatives,
who then clearly represent the electors who voted for them and for
those who voted for the electors. No votes are wasted, except that
there may be a few that refuse to compromise and are not
represented until they do.

  As soon as you have "most votes" instead of unanimous, you have

broken the system, at least partially. As the seats are elected,
compromises may be more difficult, but this is natural and leaves
power in the hands of the electors. We have experience from an
actual Asset election. The election was designed by Clay Shentrup,
who, frankly, had no clue what he was doing and who did set up an
idea of a deadline. An asset election can be managed by the
electors, in fact. Those who are elected are in theory
disinterested in the rest of the election, and those who gave
their votes to them are also theoretically disinterested.

  What happened was that I received more votes than any other

candidate. I had more than enough votes to be elected myself, and
to give Warren a quota also. Clay had not defined the quota, as I
recall, so I kept both the Hare and Droop quota in mind. That left
another seat to be elected. However, we had two members elected
then, of the 3 person steering committee, and if Warren and I
agreed on something, it really did not matter who the third member
was. But, of course we would want the other voters represented.
Clay had been the second highest vote-getter. By most systems,
he’d have been elected. Clay decided to assign his votes to
another candidate, and then I asked the remaining candidate (there
had been five) if he approved the election. He did, and so it was
unanimous by Asset rules. There was one voter who did not like the
result. He had voted for Clay, I think, and was disappointed that
Clay transferred the vote. But he basically did not understand the
system, and many interested in voting systems don’t understand
basic democratic processes.

  That outcome with any other system would have been very unlikely.

The whole election science field has been dominated by attempting
get the best result from limited information, typically a single
ballot.

  The goal of Asset is full and proportional representation, with

each seat representing the same number of voters, so the seats
having equal voting power – it is a peer assembly – is obviously
fair. If this is not done, then representation becomes warped,
where seats elected with a full quota would have the same power as
those elected with fewer votes.

  There is no reason to do this except a belief that elections must

be complete by some deadline or the sky will fall. That was, more
or less, the argument of the Supreme Court in 2000 in disallowing
the completion of vote counting in Florida. Ordinary organizations
following Roberts Rules know that this is undemocratic nonsense.
Somehow we tolerate it in public elections, and that is obviously
the reason for plurality. The idea may be that the plurality
winner is, perhaps, most likely to win if the election were
repeated, but repeated election until a true majority of voters
has clearly chosen the winner is strong tradition in democratic
organizations that meet directly (Robert’s Rules includes in the
basis for majority, including mail ballots, any ballot with a mark
on it. In other words, a ballot with NO written on it – or
anything – counts. (We imagine that putting NOTA on a ballot is
some kind of revolutionary idea. No, it is basic democratic
process, it doesn’t have to be there under Roberts’ Rules. Just
cast a ballot without a vote for any of the candidates but any
mark, and it has that effect. Blank ballots are considered so much
scrap paper. Suppose that had been the rule in Florida in 2000. .
. .

  The problem with plurality is not so much vote for one as trying

to declare a winner without a majority. Vote for one works, to a
degree, if a majority is required. All voting systems, my position
has been, should require explicit voter approval of the winner and
never complete without that. Restricting votes by reducing choices
to two looks like requiring a majority, but it’s a fake and
constrained majority. California had a system, active in San
Francisco, where write-in candidacies were allowed in runoffs.
Hardly anyone in the voting system community noticed when San
Francisco killed that. It was the possibility of a runoff
candidacy that made top-two runoff more democratic, and there was
a case in Long Beach where a write-in candidate won the runoff. It
can happen.

  I'm not suggesting Asset for single-winner elections, but it

could still be superior. We need to separate the election process
from the process of choosing the best person for the job. The
election process, for an officer position, is much better handled
by a representative body, which can select someone not even on the
ballot. But we don’t think that way.

  In my original work on delegable proxy, in the 1980 and 1990s, I

started with the idea that one could hold a national Presidential
election in less 10 days by having people meet in groups of 10,
and then each group elects the best person to represent the, tje
next day in groups of 10 representative/candidates. Then I
realized that this would disenfranchise those who did not support
the majority choice. So, what if these groups were organized
around the “candidate,” and what if candidates then carried that
number of votes forward, each transfer being freely chosen? This
became delegable proxy. Every proxy is chosen voluntarily, by
agreement on the proxy, and gains representative power according
to the exact assignments gained. Unanimous election among that
group. Don’t like it? Choose another proxy…

  Asset preserves free choice in representation. Unless there is

actual issue voting in the assembly by sub-quota electors (it’s
possible, but complicates the system), the electors must
compromise to some degree, agreeing to cooperate with other
electors to assign their votes to create quotas. What remains is a
clear process, where the actors are responsible for their actions,
and actual vote transfers are visible.

  If the idea of no votes being wasted doesn't get people excited,

I don’t know what will.

  In real organizations using real process, cooperation and

compromise is normal. If some measure cannot get a majority, it
doesn’t pass. If an officer (say the meeting chair) does not get a
majority, they are not elected, and if the members want someone
elected, they have no choice but to keep voting until they find a
majority choice. And the members of a Robert’s Rules assembly can
recall a chair at any time, it’s trivial. If you can get a
majority, done. Real organizations also realize the dangers of
making decisions by a mere majority. It can weaken the
organization, so discussion may continue, but, in the end, forcing
the majority to continue discussion when they no longer are
willing to allow it creates a kind of minority rule. So a wise
majority will attempt to compromise with minorities, to increase
overall satisfaction.

And if the majority is foolish, consider emigration.


#29

The entire concept of “politicians” is rooted in systems that
Asset could make obsolete. Once no votes are wasted, once voters
can freely chose the person they most trust from among a field
restricted only by eligibility to serve (and restrictions on that
should be minimal. If people really want to vote for Vladimir
Putin to represent them, the problem is? But being realistic,
there will probably be some restrictions.

  Revocability I consider advanced, but once Asset Voting is in

place, the electors will have real power and will, I predict, be
able to modify the system, rather efficiently. As public voters,
having real power, I expect that electors would make themselves
known and would be active in what we now call “lobbying.” They can
deliver votes.

  For initial implementations, I suggest the method being used for

the election of a representative assembly without more direct
democratic options. I suggest, though, that once the Electoral
College is set up, it will find ways to improve the system, and it
will be in a prime position to lobby the legal Assembly for
changes, with some teeth behind it.

  Asset could eliminate the need for money to be elected; in fact,

with Asset in place, I expect that someone spending a lot of money
to get elected will be viewed with high suspicion. As should be
normal . . . except that in the present system, high spending is
necessary. We make massive fundraising necessary and then blame
politicians for doing what is practically required to get elected.


#30

It would be only a small difference to do it as quota versus consecutive deadlines. As you said, people who refuse to compromise would be left unrepresented, putting pressure on them, while there would still be a fairly large amount of representation within the first deadline, only increasing every deadline after until unanimity. I ask this because I’m interested in the more practical implementations of Asset, and I’m not sure voters would agree to a theoretical quota to decide who is represented and who isn’t.


#31

The deadlines would be artificial, and I see no benefit at all.
If it takes the electors still not fully represented six months to
agree, and if this does not prevent normal operation of the
Assembly pending, the problem is? What problem is worth, in order
to solve, the damage to the concept of full and equal
representation?

  The quotas are determined by the size of the Assembly desired.

Given votes V, and seats to be elected N, the Hare quota
(Wikipedia) is V/N, the Droop quota
(Wikipedia), V/(N+1) (approximately, there is more than one method
of calculating a similar quota). Neither is arbitrary, the Hare
quota being slightly easier to understand.

  What I see is that in Asset, using the Droop quota does no harm

if it remains theoretically possible for an extra seat to be
elected. It is also possible to elect a seat with tiebreaking
power, by a majority vote of the remainng unassigned electoral
votes. That seat could have deliberative privileges, the right to
address the assembly, etc., but would not normally vote except on
questions of privilege.

  Again, quota depends on the target number of seats to be elected.

The principle is very easy to understand.

  There is no harm from variation in the quota level, as long as it

remains possible for all voters to end up with equal
representation, with only natural lack of representation because
of inability to compromise.

  Thus the quota should not be larger than the Hare quota. It could

be smaller than the Droop quota, with an increased risk of
creating more seats, which is probably mostly harmless. I kind of
like the Droop, using it as described unless all the electors get
it together.

  The idea of forcing compromise through "pressure," is

antidemocratic as with any voting system practiced that attempts
to coerce voters (such as mandatory full ranking in Australian
STV, a system which then pretends that these lower ranked votes,
coerced, can form part of a majority.)

  It's quite enough that there be natural consequences. After all,

the present system requires massive compromise just to get a party
nomination and get on the ballot.

  Allowing election by a majority of remaining votes harms the

principle that each seat is equal in voting power represented. The
suggestion I make above ameliorates that, by using the voting
power only to break ties, in which case it is like 1/2 vote, which
could not create a tie, but could break one. Seems fair to me. If
those voters get it together to agree, then that seat could get a
full vote. Maybe if they reach some threshold, like 2/3 pf a
quota, that could be allowed.

  I think it is important to get Asset systems into use and that

initial systems be quite simple. The Asset system itself, through
the power that electors will obtain, can reform itself. It will
should have the clout to do that


#32

A final question: do you like the idea of weighted congressmen (legislators with unequal voting weight)?


#33

Politicians need SOME time when they are not directly preparing for an upcoming election. (Politicians should be held accountable for their actions, but they need a break from campaigning to be able to function properly.)


The thing about Asset being “radical” confuses me. I am used to “Asset” meaning a multiwinner voting system where candidates, in real time, can transfer their votes among themselves. I am seeing elements being discussed that are not part of my Asset definition.


Droop quota does not allow an extra seat to be elected. In a 5-district, only 5 people can have 0.166668 of the vote because 0.166668 * 6 is greater than 1. That being said, I might want to rule that a Hare quota is required for all but the final seat (although that may mean 3 rival factions that all hate each other can throw the election).


Weighted Representatives… Well, you need SOME method of filtering candidates, because a congress of 97,228 people is just unworkable. Most of them will not be able to give more input than their vote even if they tried. And that raises questions like should they get unequal pay, staff, security, etc?


#34

There’d still be a limited number of representatives under Weighted Congressmen. I wouldn’t support unequal pay or anything like that, just the social privilege that naturally comes from such a system.


#35

In general, no. I proposed in a previous mail that a single
member could elected with, essentially, a half vote, until and
unless enough votes are put together to make it full. It’s just an
idea and would only be one member, the assembly would be otherwise
fully peer, equal members. Factions with more than that (if we
consider factions, unlike most PR systems, Asset does not depend
on parties or factions) would have more than one seat.

  Each seat represents a defined number of voters, the quota. The

Electoral College (EC) votes, if any, are weighted, but the EC
business is relatively simple.

  I prefer to write about assemblies rather than "Congress" or

“legislators,” because we have only very limited experience with
Asset Voting. Under those conditions, there is zero possibility
for adoption in, say, the U.S., in public elections. (A
jurisdiction already using STV might be more possible, but still
very difficult, I’d think. Political parties will be opposed .
Period. Asset could make them unnecessary. (I expect political
interest groups would persist, they merely become not as essential
as they are in more primitive voting systems. Asset creates a
layer of representation between the general membership (the
public, citizens) and actual decision-making bodies.

  Hence activist focus should focus on NGO adoption. Starting with

CES. Asset could be used for the election of the Board, and then,
between elections, the EC could be a general advisory council.

  (The known public delegable proxy experiment (Demoex, in Sweden)

created and elected, through advised votes, a seat pledged to vote
according to a political party process that used DP. That turned
democratic traditions – and the basic concept of proxies, people
who can make decisions representing someone else, but based on
their own best understanding --, on its head, outraging the rest
of the City Council, and I didn’t wonder at that. It was a very
Bad Idea. We need to elect, for public office, people who are
wise, knowledgeable, and prudent, not rubber stamps. Demoex
represented a minority constituency. I would have advised them to
become nonpartisan and to create advice, not control.

  How can one negotiate, a basic principle of deliberative

democracy, with someone who says, “I’m not voting for my opinion,
but for what I pledged to vote for”?

On a large scale, direct democracy sucks . It works better
than anything else on a small scale. Because of the latter, we may
think it ideal for the former. Direct democracy has created many
disasters, and the general public, occupied with life, can easily
be manipulated through media … or literally by mobs. The
classical solution is representative democracy, where the
representatives are trusted. Asset allows total freedom for the
general public to choose representation, and any citizen (if the
rules are as I’d have them) can volunteer to be an elector, a
public voter. Electors, by the way, could solicit financial
support from the public. That could be used to pay for security,
if there are security issues. Under difficult conditions, I would
then expect, from the economics, that there would be relatively
few electors, compared to easy conditions, where there may be
many. Electors would not be paid, in basic systems, other than
voluntarily by the public.

  Seats would be paid, in some assemblies, as they are now in

legislative bodies.


#36

Continuing the discussion from CIFeR voting. An alternative to asset voting:

Scenario 1: Three stockholders have 45, 40, and 15 shares respectively. They all have 1/3 effective power.

Scenario 2: Four stockholders have 42, 40, 15, and 3. The one with 3 can never sway any decision. Should that person have the same soft privileges even if his vote has no meaning?

Not to mention that with weighted reps, figuring out when you have a majority (I.e. The votes to pass a law!) is complex and difficult and hard to communicate with observers. (With the extra 83,441 votes from Jones, the bill has enough votes to pass! vs With the extra vote from Jones, the bill has 218 votes and can pass the House!)


#37

You might not have these conundrums if the candidates were more diligent in exchanging votes, and if the legislators had a “secret” vote to determine the agenda of bills to be voted on, thus ensuring that only bills with a decent change of passing would reach the floor.


#38

I’m not surprised to find confusion about how radical Asset is.
It departs from many assumptions about what a voting system is,
and it is not simply a voting system. It is a system allowing complete
representation of an electorate in a decision-making process, that
builds representation from the ground up, directly, through simple
voter choices.

  Asset, invented by Dodgson as a tweak on STV (he was a major STV

theorist), allowed candidates receiving votes to transfer them.
Actually, when I first encountered STV that’s what I thought it
would be! Dodgson realized that most people would not have the
sophisticated political knowledge to be able to fully rank
candidates, and if they didn’t fully rank, they would lose voting
power and end up unrepresented. But . . . they would be likely to
know who their favorite was! This, then, could create strategy
free voting, simply vote for the person you most trust.

  (The concept simply extends what any officer already does,

delegate authority, to delegating authority to someone who
actually takes on the office. The skills involved for each task
are likely to correlate.)


#39

When there are empty seats, would a majority vote in the Assembly need a majority of all seats or just a majority of filled seats? In the case of the former, if each seat took a Hare quota to fill, then requiring more than a half quota to fill the last seat would ensure that all majority votes represented a majority of all of the assets. However, that won’t work for votes that require a supermajority to pass.


#40

I am just confused because I saw different rules and terminology that I am not used to. Care to define your version of Asset right here just so we are on the same wavelength?