[EM] average over time proportionality election method

[Abd ul-Rahman Lomax]

At 09:00 PM 3/6/2006, Raphael Ryan wrote:
>Abd ul-Rahman Lomax <abd at lomaxdesign.com> wrote:
>I see your point about asset voting.  A negotiation step can allow 
>the interactions to happen between candidates that would require a 
>complex voting procedure to model.  However, you lose the 
>anonymous/secret ballot effect at the negotiation stage.

That is correct. However, when you are willing to accept public 
office, you are generally making a decision that your actions will be 
public and that you accept the consequences. Electors are public. So 
are the assembly members. Indeed, it is essential that vote transfers 
subject to the initial poll be public, or else there will be no 
accountability. Voters must know what is being done with their votes.

>   There are also issues for hold outs and deadlocks and "chicken" 
> effects.  A deterministic voting system avoids all that.

Sure, and is practically brain-dead as a result. Asset Voting turns 
election, normally an aggregative process in the language of 
political science, into a deliberative one, the kind that is really 
used in small organizations, even when there is some formal process 
using some election method or other.

How would a deadlock occur? As an example, suppose that all the votes 
have been reassigned, and winners determined, except that there are 
two holdouts, and both of them have approximately the same number of 
votes, half of that necessary to elect a seat. And they cannot find 
some compromise. Note that I'd allow the reassignment of votes 
outside the circle of candidates, so this failure to find a 
compromise means that there is no common trusted person. I could make 
an argument that such people should be shut out of government, but I 
would not favor any mechanism which does so except the natural one: 
inability to compromise may shut you out of influence.

The voters would decide if they were willing to accept failing to 
elect a representative, and they would demonstrate their opinion 
about this in the next election.

There are other possibilities for how to handle holdouts, such as a 
restricted status for candidates below the quota, the ability to 
reassign the votes much later, i.e., no deadline (except the next 
election), but I'd prefer to keep it simple. I think it quite 
unlikely that more than a very few seats would remain vacant due to 
holdouts. The exclusion from participation of minorities unable to 
reach a quota is rigid in existing systems.

The time proportional system removes the problem for such minorities, 
thus giving them reason to not compromise; instead they continue to 
hold out until they get their day. But if they are in such a 
minority, what do they gain by the delay? Not much, I would suggest, 
that they would not gain anyway immediately by compromising on a 
representative in an asset system who was at least willing to 
consider their ideas, since that is really all they are going to get 
if they are awarded victory by a proprotional system.

(Okay, we will give you our votes if you will agree to present our 
views and arguments for them, in a fair way. Asset Voting creates a 
negotiator who can accomplish the task. Most election methods attempt 
to be deterministic -- it is a standard election criterion -- and 
thus there is no room for give and take. In my opinion, this is one 
big cause of the unsatisfactory results from present election 
systems, the other one being representation failure, which 
time-proportional also attempts to address in a deterministic way.)

The delay due to quantization noise, in my opinion, could be fatal, 
unless somehow election update frequency could be very short. Much 
shorter than present election cycles.

>Also, I was reading Warren Smith's paper.  I think that the 
>mechanism for elimination he proposes still suffers from the middle 
>squeeze effect of IRV.  Perhaps a better rule would be to allow a 
>candidate to voluntarily exclude himself.  However, I guess a 
>candidate can do that by just transferring all his votes to another person.

Exactly. I don't think there is any middle squeeze effect in Asset 
Voting. Quite simply, the best voting strategy is to vote for the 
candidate or candidates you most trust. Given that much of what a 
legislator does is to find widely-acceptable compromises, and also a 
legislator (or officer) must be able to delegate, to find trustworthy 
people, plus, under Asset Voting, votes are transferred directly to 
people with whom the transferor will likely have direct 
communication, I think that pretty much the same qualifications exist 
for legislator and for legislative elector. Someone good at being a 
legislator would probably also be good at electing one (or more).

Once again, I think time proportionality is an interesting idea, but 
I think its goals, and more, can be accomplished much more simply. 
Further, as I mentioned, time proportionality has a significant time 
lag; if the electorate is rapidly changing, there will be a 
substantial delay. *One election cycle is substantial.* Some might 
think of this as a feature, but the conservation of political 
momentum is not a happy phenomenon under some circumstances; for 
example, it can easily happen that the public comes to recognize that 
a candidate, even one who was a favorite, is a loose cannon and 
dangerous. All existing election methods allow the dumping of this 
official at the next election, plus there is impeachment or recall. 
What would you do with recall in time proportionality? Pretty much, 
in order not to loose minority representation, you would have to 
eliminate it. Yes, in a proportional assembly with sufficient 
numbers, the problem is reduced, but it is not necessary to have the 
problem at all, beyond the degree to which the problem is caused by 
terms of office.

(And delegable proxy can eliminate even that. There are no terms of 
office in delegable proxy, beyond the current status analysis delay, 
which might be a day. Would this potential rapidity of change cause 
instability? I think not. I don't think that proxy assignments would 
change rapidly; indeed, once a citizen has found a reliable proxy, I 
would not expect that to change until the proxy retires or dies or 
becomes incapacitated, mentally or otherwise. There *will* be a 
conservatism in a delegable proxy system, it is the conservatism of 
opinion itself. Delegable proxy will not be particularly vulnerable 
to the "majority instability," where a few votes can radically shift 
the outcome of an election. Rather, DP will naturally tend toward the 
development of intelligent compromise and consensus, and I would 
expect something like that for Asset Voting as well, which is almost 
as good, in my opinion, just not as rapid and flexible.)