[Prev] [Next]   [Index]   [Thread Index]

02671: Re: [WDDM] Democracy axioms

From: Joshua Petersen <joshupetersen(at)gmail.com>
Date: Fri, 13 Aug 2010 18:57:44 -0500
Subject: Re: [WDDM] Democracy axioms

My knowledge of this comes more from classes in how mass media (which covered how to manipulate polls). Unfortunately, I have no statistics for that portion, I'm sorry. Though corruption is obviously a thing that isn't documented as much as it should be, so getting worthwhile statistics on such a topic would be tricky. And you are right, there is no attempt to control for distorting factors in self-selecting polls. But that's the beauty, you don't really need to. The topic here is how to decide when the people would have to vote on a legislation. The final vote, not the poll, is the true measure if people actually want the legislation or not.

Whether or not it goes up for vote should be more dependent on how pressing of an issue it is. That's how come the self-selected polls would work better. If people have an issue with the way things are going, or want things to improve in a certain way, more active individuals would seek out and add their support to the potential legislation, which is a good measure of whether it should be brought to a mass vote.

By taking only the most-supported legislations, that makes it also so that time is more taken into account. People will react more to what they think are urgent issues, thereby getting important topics voted on first. If each poll had to be randomly given out (which is a lot of work, I might add) there's little inherently involved to make voting on what to name a new post office as compared to voting on whether or not we should retaliate from a recent terrorist attack.

And yes, lack of single-entity control is an advantage. That's what makes a direct democracy a preferable alternative to a monarchy. It keeps the power in the hands of the people, rather than in the hands of a few individuals.

And self-selecting polls ARE good for determining interest in a topic. It's a logical conclusion: The more people that have interest in a topic, the more people will express interest in that topic. Since people have to go out of their way to participate in self-selected polls.

I should point out, there is lots of dislike for self-selected polls, and with very good reason. Traditionally, most polls exist in a sort vacuum. The poll by itself says the point of the topic. "52% of people support option A, 48% support option B." If a traditional poll like this were self-selected, it would be a very bad idea.  However, what I'm suggesting is nothing like that.

The point is to have a public forum where all the 'polls' are available. (In a number of ways, 'polls' is actually a bad word here, as we're not trying to judge typical preference on just a single topic, but interest when compared to all available topics, which are two extremely different things.)

Here's a somewhat extreme hypothetical situation to help illustrate the point:
Using traditional polling methods:
Two polls may be presented to the populace:
Should Sally May's pet cat get a street named after her?
Should we declare war on a 3rd world communist country that just declared it was going to develop nuclear arsenal?

Using traditional polls, the Sally May's cat bill gets sent out. Only a dozen people get sampled, but most don't see a problem with it, and the vote goes out: 8 for, 3 against. aka 75% for, 25% against
The war poll also goes out. Thousands of people get it, and it's a harsh issue, very divisive. We'll say it's 10,000 for, 10,000 against. it's roughly 50% for, 50% against.

Going by traditional polls, the Sally May's Cat bill appears as a more pressing issue. Granted, the polls do accurately show to some degree how the public will vote on the topics.

However, using a self-selected poll, we'll go with the similar numbers (although for the sake of argument, skew them due to people seeking out polls they liked.)

The Sally May's Cat legislation gets 11 poll vote for getting a street name, and 1 poll vote against. (notice the lack of focus on percents, they're pointless for this kind of poll.)
The war poll goes out, and many agitated patriots vote on it, and a number of activists against war also go at it a bit later, as the topic's covered in the news, and everyone's talking about it.
It comes out 15,000 for, and 5,000 against.
In this situation, it means that there are 10 votes for bringing up the issue of Sally May's cat getting a street name vs. 10,000 votes of whether or not we need to vote on whether or not we go to war.

The poll doesn't decide which side will win, again that's the purpose of the vote itself. What it does decide is when a topic becomes important enough to vote on. For that purpose, the self-selected polls actually do serve quite usefully, more so than traditional polls. It's because they're purpose isn't to discern what's preferred, as a typical poll, but what's important. For a situation like this, whether or not people are going out of their way to take part in the poll is even more important than the results of the poll itself.

Also, this allows for something traditional polling doesn't allow for as easily: alternative versions. If anyone can start a poll (again, akin to youtube starting a video), people will initially vote up a single idea. As the idea gains steam, different people will submit other versions. These other versions will also get voted up and down. At start, those who the issue is important to will more than likely vote for all of them, but as different versions of the legislation begin rising in the ranks, so to speak, the versions will naturally begin competing with each other for the top spots. And supporters of one version will likely vote down another version. This results in a very organic and socially natural ways for legislation to evolve before finally going to vote.

Yes, in traditional normal situations, a random poll is insanely more reliable than a self-selected poll. However, a government fully by the people is anything but a traditional normal situation.


On Fri, Aug 13, 2010 at 1:47 PM, <Joshua N Pritikin> wrote:
On Fri, Aug 13, 2010 at 12:38:13PM -0500, Joshua Petersen wrote:
> In some ways, initially, yes as only people interested in a topic will seek
> it out at the beginning. However, as the topic comes more to the forefront,
> people will seek it out from both sides. Further, the self-selected polls
> are harder to corrupt.

Have you studied any statistics?

> Many politicians and corporations have done 'random' polls designed to
> get specific results. By controlling the way in which the people are
> randomly contacted skews results (such as a poll on the street is more
> likely to hit city people - in the U.S. that'd mean a higher
> concentration of democrats, phone polls which are more likely to be
> answered by people who have time on their hands - more less productive
> members of society, etc., etc.)

Yes, obviously these factors matter. And in self-selected polls there is
basically no attempt to control for these distorting factors.

> The strength of self-selected polling is that at no point does their
> have to be a single entity controlling it.

Strength? You call that a strength?

> Self-selected polling, however, is very good for determining the
> *interest* in a topic.

According to who? That's your assertion. I am highly skeptical.

> So when it comes to legislature and final votes, it shows that the
> legislation being voted on is what's important enough to the people to
> bring to all of their attention, which is really the whole point of
> this stage of the process.

Yes

> On Fri, Aug 13, 2010 at 11:21 AM, <Joshua N Pritikin>wrote:
> > Self-selected opinion polls are vastly inferior to randomly selected
> > opinion polls.


[Prev] [Next]   [Index]   [Thread Index]