[EM] Trees by Proxy

[Abd ul-Rahman Lomax]

It should be understood that my writing is often quite general. When 
I propose Delegable Proxy, for example, I am referring to the most 
generic implementation, though I may sometimes assume certain rules. 
Mr. Ketchum seems to treat DP as if the rules were fixed and quite 
particularly fixed.

At 03:27 AM 3/24/2007, Dave Ketchum wrote:
>Anyone interested in understanding what I am offering here had best 
>ignore anything Abd offers here:
>      He offers Free Associations, Asset voting, and Delegable 
> Proxy.  He may have something of value, but I also claim value for my thoughts.

What Mr. Ketchum has proposed seems to be Delegable Proxy with a 
relatively fixed structure, constraining how proxies are used. I'll 
also note, in the face of his apparent irritation, that I took him as 
inviting comment and comparison in his original post in this thread, 
the first paragraph. If he doesn't think what I'm writing to be at 
all relevant, he's free to disregard it, as is anyone.

>      I offer proxies as a way of populating a legislature.  While 
> we both got proxies from the same source, there are enough 
> differences in the way they are used that you get nothing but 
> headaches if you mix Abd's ideas with mine.

What I've proposed is apparently closer to the standard, generic, 
definition of the term proxy.

>On Sat, 24 Mar 2007 01:07:54 -0400 Abd ul-Rahman Lomax wrote:
>
>>At 09:36 PM 3/23/2007, Dave Ketchum wrote:
>>
>>>Seeing Free Associations and Trees by Proxy as different concepts:
>>>      Abd's Free Associations use proxies to create Free Associations,
>>>which decide for themselves what they are and do.
>>>      My Trees by Proxy use proxies to elect legislatures, which then are
>>>much like traditional legislatures.
>>I'd like to point out that I write about FA/DP, not just FA. DP is, 
>>of course, Delegable Proxy.
>
>I avoid the phrase "Delegable Proxy" because the usage is just 
>different enough that they had best be kept apart.

Sure. What he has proposed appears to be, however, a variant of 
Delegable Proxy, as I mentioned above. A number of restrictive rules 
have been added.

>[...]
>While a legislature is usually geographical, I do not demand that of such.

Legislatures are geographical in two ways: one is that 
representatives have a fixed district, from which they are elected. 
This, of course, eliminates representation in the legislature of all 
those who don't consider themselves represented by the winner of the 
election. (Which is not to deny that many or most legislators do 
consider themselves responsible to *all* their constituents; indeed, 
that this is frequently true is a reason why the existing system has 
not fallen on its face long ago.)

The other way is to have multiwinner districts. This imposes the 
constraint of only a few representatives per district, and thus 
leaves out groups that are thinly spread across the state. Such 
groups could actually be substantial in number, yet remain unrepresented.

I find it an amazing lacuna that we repeat the phrase of the American 
Revolution, "No taxation without representation," as if such taxation 
is obviously oppressive, but we have continued it in merely a 
different form. If it is argued that the elected reps represent all 
constituents, it could also have been argued that the governors of 
the early states, or other appointed representatives, represented the 
colonists. From the point of view of those whose selection lost the 
election, the result is the same: one is represented by someone who 
was appointed by someone else. It is representation of a kind, but 
not democratic representation.

It is more accurate that the majority -- or plurality -- is represented.

In any case, this consideration leads inexorably to the conclusion 
that fixed districts are inherently undemocratic. Note that it is 
still possible to have *relatively* tight districts, geographically, 
but with the provision that some districts are floating to some 
extent. I've detailed in quite a number of posts how to do this with 
Asset Voting. DP, of course, accomplishes the same thing. Trees by 
Proxy seems like it would, also, but with some provisions that reduce 
the powers of the voters in certain ways.

>>>Takes time for a proxy to become effective, or to lose effectiveness when
>>>giver ends it.


>>We understand proxies as they are used in common law; they are 
>>revocable at any time by the client. Obviously, there can be 
>>latency, but the same is true with attorneys-in-fact. If I'm at the 
>>bank executing a document on your behalf, you might revoke the 
>>proxy, but it will still be binding until involved parties have 
>>notice of the revocation.
>
>
>Here we have Abd talking without understanding the topic.  I write 
>of time BECAUSE it is not appropriate for changes to take effect 
>before a deliberate delay after specifying such.

Ketchum can be puzzling. I understand latency. What Ketchum has 
stated here is that some deliberate delay is appropriate. He hasn't 
stated why. I gave an example of the latency that is normal with 
proxies. If there is a term of office, which to some extent fixed 
delays introduce, there is a corresponding distance between the voter 
and the proxy. Combined with a disallowance of direct voting, it then 
becomes much more possible for proxies to be abused. By the time 
voter proxy changes can become effective, the damage will have been done.

>>But a legislative system, I'd assume, would have a process whereby 
>>you could revoke your proxy, effectively immediately. The 
>>revocation might not affect a vote in the process of being expanded.
>>Note that generally proxies are exercising rights that the client, 
>>if present, could exercise directly. We have been assuming Direct 
>>Democracy as far as voting is concerned, so you could *immediately* 
>>cancel your proxy, effectively, by voting on any pending issue.
>
>Another difference - my proxies have a monopoly on performing their tasks.

But that can also be true with DP as I've described it. Depends on 
what tasks are involved. Ketchum is specifically referring to a 
monopoly on voting. He has not given any reason why this is 
necessary. What I've assumed is the minimum necessary restriction to 
allow scalability. The right to vote does not impact scalability, 
under most conditions. The right to deliberate does. So I assume that 
delegable proxy assemblies will set rules that permit the relevent 
population to be as fully represented as possible, without making the 
assembly unwieldy. With DP, the assembly can be much smaller than 
would be necessary with a peer assembly, for a given degree of representation.

I've often considered that participation rules might not be fixed. 
For example, the number of proxies that must be held -- if that is 
used as the participation criterion -- might not be fixed; rather the 
size of the assembly might be fixed to a number that has been found 
to be efficient. Thus at any time a particular proxy might or might 
not have floor rights. But he or she could always vote, and that 
includes voting on such matters as rules. (And, in practice, getting 
the floor would merely be a matter of getting permission; the 
relevant request would be presented by a proxy with floor rights, 
and, of course, if a proxy abused the privilege, the proxy's floor 
rights could be revoked.

Essentially, I've considered that, just as voters and proxies are 
free, so too are meetings, which can set their own rules. This 
actually is standard Robert's Rules.

>[I described how direct voting might function in a large institution.]
>Another lack of thinking to ignore.

If there is a lack of thinking, there is likewise a lack of common 
courtesy. It is normal to give or refer to some *explanation* as to 
why one thinks a proposal or concept is defective. Mr. Ketchum's 
comment here is something on the order of a blank "You're wrong," 
with no explanation of what, exactly, is wrong. It's his privilege, 
but so too is it mine to point out that much of what he writes is 
like this. Laconic, highly critical, and, in fact, usually ignorant. 
The latter is, of course, just my opinion....



>>>      Holding too many is discouraged by limiting the number that can be
>>>counted as giving a legislator power.
>
>And ignore what follows.

If it is to be ignored, why does Mr. Ketchum quote it?

Because he *usually* quotes far more than what he responds to. Again, 
we could consider that rude. I do think it inconsiderate. In what I 
wrote, I considered the question of limiting the number of proxies 
which could be exercised by an individual. Without in any way 
detailing *why* one might wish to limit this, Mr. Ketchum seems to 
dismiss it as irrelevant, though it was he who had proposed limiting 
them. I won't however, quote myself. It's there in the record.

Now, from his original post on "Trees by Proxy":

>Guidelines:
>        Tailor numbers as further thought dictates - I am just trying for
>ideas.

That's fine, though I prefer, myself, to avoid going into too much 
specific detail, preferring to leave the details to those actually 
working on implementation. The numbers Mr. Ketchum gives are totally 
arbitrary and would be more likely to be suboptimal than not.

>        Juho's village, town, etc. are nominal goals for sizes - given 350
>people they should be around 3 or 4 "villages".

On what is this based? 5 representatives from a village of 100 is 
pretty inefficient. With DP and total freedom and independence of the 
voters and villages, I'd expect the norm to be much less than that.

Mr. Ketchum, it may appear, has spent a few minutes or maybe a few 
hours looking at these concepts. I've been working on them for twenty 
years, with many thousands of hours put in recently. It may show. I 
wouldn't mention this -- it's a bit rude -- if not for his constant 
refrain of how ignorant or shallow my thinking is....

>        Borrow proxies fresh from corporate stockholder usage.  Their
>effectiveness starts at midnight 10 days after filing; ends 10 days after
>a replacement is filed or signer dies.

That is a particular rule which may have been set for some 
corporations. It's a restriction on the rights of shareholders, and 
many such rules have arisen due to boards and management conspiring 
to take rights away form shareholders wherever they can. In a 
corporate with a relatively small number of knowledgeable 
shareholders, they would never get away with this. What this does is 
to prevent shareholders, as a proxy fight or other controversy 
arises, from revoking proxies that they may have routinely granted to 
management, or as suggested by management, which is a very bad 
practice that should probably be outlawed. Indeed, the failure of 
proxy democracy in the U.S. corporate system might be traced to 
precisely this abuse. It seems hardly desirable to replicate the 
error in political proxy systems.

>        Representatives, such as Juho's 5 from a village in a town
>government,  have power according to how many effective voter proxies they
>hold, directly or indirectly:

This, of course, is delegable proxy. That is why it is justified to 
call Ketchum's system "Delegable Proxy." But he has tacked on quite a 
few rules that he apparently pulled out of a hat, and the effect of 
these rules appears to be to dilute the democracy involved, to a 
greater or lesser degree.

>             Must hold 1% of a legislature's proxies to be able to vote there.

Why the limitation at all? Unstated. And, since he *has* 
discriminated between voting and floor rights -- which, as far as I 
know, was first proposed by me -- he seems to realize that there is a 
difference, that floor rights must be more restricted than voting 
rights. Yet he doesn't simply eliminate the direct voting rights, and 
then, of course, he will need to limit proxy powers to avoid the 
abuses that become much more possible.

>             Must hold 2% of a legislature's proxies to have full
>capabilities of being a legislator - offering bills, debating, etc.
>             Limit on voting power is 40% of proxies voted in any vote - no
>czars allowed.

Why not limit it to less than a majority? Why 40%? I'd ask, why limit 
it at all? -- with certain caveats. Essentially a superproxy (which 
could be defined as someone who represents everyone, or as someone 
who represents a majority) shouldn't be able to pull a binding 
decision out of a hat without notice. And, I pointed out and Ketchum 
"ignored," deliberative bodies become far more efficient when a chair 
senses the consensus of the assembly and proceeds, effectively, as if 
he or she was a superproxy, providing that any member can interrupt 
the proceedings. So as long as there is appeal, I see no need to 
restrict the number of proxies which can be exercised by an 
individual. Technical details, though, might make it necessary.

>             Sideways proxy - possible for representatives to be too weak
>above.  Such can pass what they hold to others for legislature
>participation.  This does not release anyone from the above limit, nor
>does it affect what anyone passes up to others via proxies.

"Sideways proxy," if I've understood it, is simply further delegation 
of proxy. If not for it, restrictions on participation would be a 
severe problem (though far less abusable if voting rights are 
retained and exercisable freely). But then Ketchum states that it 
does not "affect what anyone passes up to others via proxies." I'm 
totally unclear on what this means. How could one person's 
disposition of a proxy affect others?