[EM] Sociological issues of elections

[Kristofer Munsterhjelm]

On 09/06/2013 03:30 AM, Fred Gohlke wrote:
> re: "Arguments against direct democracy usually go that
>       the public is too short-sighted or that it doesn't
>       have enough specialized knowledge."
>
> My personal opposition to direct democracy is the susceptibility of the
> public to the influence of behavioral psychology, a tool used in
> partisan politics to persuade the people to favor one point of view or
> another.  It is much too easy to concoct fictions, particularly to
> frighten the people.  To reduce the force of the manipulations that
> engulf us, the people need an an electoral process that allows and
> encourages them to deliberate.  That would occur during the election
> stage of the hybrid process.

I was thinking of a direct democracy on the small scale. An argument for 
representative democracy goes like this:

- If we were just one town, then we could hold a meeting and decide by 
ourselves (direct democracy).
- But it's way too hard to discuss with everybody else when "everybody 
else" numbers in the millions, if not tens thereof.
- If forced to do so, people usually compensate by acting like a mass. 
But masses can be manipulated, and we have seen very ugly consequences 
of particularly charismatic demagogues (from Robespierre to Stalin).
- So brute force, "everybody speaks to everybody else" direct democracy 
on a large scale can't be done.
- So we need somebody to act for us, but few enough that they can still 
govern by deliberation.
- Hence, representative, elected democracy.

In that view, representation (the election of representatives) is one 
way of distilling the public opinion into something manageable where 
deliberation can work. The hybrid system is another, and your system is 
yet another: the assembly becomes that distillation, and the assembly 
deliberates.

(Deliberation doesn't have to be only on the assembly level, though. 
Delegable proxy has the deliberation happen "on the outside" as well. 
And economic systems, e.g. prediction markets, can work without any sort 
of *visible* deliberation going on. Unfortunately, market failure and 
manipulation is quite common, but it does show what I mean[1].)

> re: "Then the argument against the "average person" is really
>       a claim by those whose opinions are more to the left on
>       that line that the public can't govern on its own."
>
> I'm not sure where my views may appear on that line because I rarely
> think in those terms.  A friend recently suggested my approach leaned
> toward 'virtue ethics', in contrast to an approach which emphasizes
> duties or rules (deontology) or which emphasizes the consequences of
> actions (consequentialism).  I've no idea whether that would be leftish
> or rightish.  To be absolutely frank, my lack of an academic background
> hinders me in this regard.  I'd never heard the term 'virtue ethics'
> before and had to look it up.  From what I read, it seems an excellent
> evaluation of my belief about electoral systems.

I used deontology as an example where consequentialism no longer 
applies, or doesn't apply as strongly. Any sort of non-consequentialist 
ethics would work.

Whether it is leftish or rightish depends on the particular virtues you 
prefer. Those that align with a hierarchical organization of society 
would be more left, while those that align with a decentralized one 
would be more right.

> re: "But all things equal, we'd prefer something to the right,
>       because we know that concentrated unaccountable rule can
>       become corrupt ..."
>
> Whether right or left, wouldn't the hybrid approach eliminate
> 'concentrated unaccountability' because of the inflow of fresh faces
> after each election cycle?  Although I may be alone in this, it seems to
> me party-based systems are the most susceptible to becoming
> oligarchical.  They wind up both concentrated and unaccountable.

That's correct. I'm probably to the right of middle, myself, and I think 
that one of the advantages of a distributed solution is that the checks 
on power are inherent to the system.

If you have a town-hall democracy, then the people check each other by 
virtue of knowing each other. You don't need to install an apparatus to 
balance the power of that democracy. In a worse case where the majority 
starts oppressing the minority, you can suggest a reorganization of the 
rules of order to a more consensus-favoring approach.

But on the other extreme, consider you have an emperor (or an elected 
king). If you suspect corruption from power is a real problem, then you 
have to set up some form of oversight. The oversight doesn't 
automatically arise from the system itself, but rather has to be 
separately implemented. And if it isn't, then the emperor (or one of his 
successors) sooner or later goes insane enough to make some very unwise 
decisions.

So systems that are more to the right have greater checks by their 
nature, as you've noticed. The hybrid system eliminates concentrated 
unaccountability because the corruption can't keep its shape: the 
corrupted members are scattered after each cycle. On the other hand, in 
a two-party democracy, the corruption can linger and power become more 
concentrated. (In a multiparty democracy, the different parties are 
supposed to check one another, and in a coalition system, the opposition 
is supposed to check the position, where the respective coalitions may 
change from time to time, again lessening the chance that corruption can 
set in.)

> re: "One possible way would be that parties would reorganize
>       as advisory organizations surrounding the legislators.
>       If a party had drawn up a plan like the above, the members
>       would try to convince the members of the legislature to
>       go with it, and the members might or might not decide to
>       do so."
>
> Considering alternatives to the status quo and integrating them to the
> extent they are appropriate is vital for a vibrant, evolving society.
> Using random selection makes it difficult to include the best proponents
> of non-standard points of view.  That is a major drawback, to the hybrid
> approach.  Having parties function as advisory organizations might work,
> but it might be more effective if their best advocates participated in
> the election phase.

Perhaps a bicameral approach could work. Use a general election for one 
of the houses and the hybrid method for the other. Or for that matter, 
use your triad method (with declared "parties" as we've discussed) for 
one and the hybrid for the other.

One of the houses discusses alternatives and so provide the party 
consistency mechanism beyond what a simple advisory approach would. The 
other reviews the suggestions and may make countering proposals. In a 
way, you get the advisory organization system, but it's formalized into 
one of the houses and it does more than just advise.

I'm imagining the election method for the hybrid to be proportional, 
also, so that if 10 of the 500 think that advocates for position X 
should be on the legislature, then 3 (same proportion) will, assuming 
they vote according to that opinion.

> re: "... any given representative will most likely only serve
>       one term, therefore he won't feel accountable. Thus he
>       would, either consciously or subconsciously, favor his
>       own particular interests. So the system might lead to what
>       one might call 'random pork'."
>
> To do so, the rogue needs the support of a majority of the legislature
> to enact the 'pork' law.  Since the 'pork' is for the benefit of the
> rogue, such support would be difficult to enlist.  Time works against
> such an enterprise.  Corruption takes time.  Blatant announcement of
> roguish intent will alienate more people than it attracts.  In the
> present system, incumbents tend to be re-elected (at least, in the
> United States).  They have multiple terms to corrupt and be corrupted.
> That is unlikely in the hybrid system.  In addition, in partisan
> systems, legislators are subject to pressure from the party 'whip'.  If
> there is no party, there is no whip.

That is a good point. Any system is a balance between attracting your 
core (as it were) and not alienating those further away. My impression 
now that I think about it, is that a two-party system cares most about 
the former. A representative system is more careful to not alienate 
others, and a representative system with ranked balloting even more so 
(because you would like even your opponents to not rank you too low) as 
long as the ranked system makes use of lower preferences. And a random 
system is even harder because, as you say, there is no whip and no time 
to concentrate corruption.

A system can be pushed more towards "not alienating those further away" 
by increasing the threshold for action (e.g. supermajority rule), and 
that's what I noticed.

So random pork is only really a problem if enough of the people are 
corruptible that they'll say "if you let me have my special interests, 
I'll let you have yours". So in a culture of corruption, you might want 
to push the system further towards "not alienating". But a culture of 
corruption would probably not instate such a system, because if the 
people see corruption as inevitable, they'd rather want to have their 
share of the pie than throw the whole pie away.

I agree. Random pork is not a problem where roguish intent is unusual. 
And where it is the usual state of things, fixes we'd suggest would 
probably not be implemented.

-

[1] In a way, markets implement a particular algorithm to solve a 
particular policy question: who produces what for whom. Being a policy 
question, it could in principle be decided by other means as well.