[EM] Implementing Practical Democracy with Votorola

Michael Allan mike at zelea.com
Thu Mar 26 17:28:01 PDT 2009

Fred Gohlke wrote:
> Thank you for the detailed description.  I had no problem with the 
> formatting (I use a 1970's word processor.  It doesn't have proportional 
> spacing.)  The displays were very clear.  I also visited the links you 
> supplied and have spent some time thinking about the material.  You'll 
> understand digesting the concept is a slow process (at least, for me).

It's a general facility, with many aspects, so it's difficult to get a
handle on it.  I'll try to relate it specifically in terms of
Practical Democracy.  It's interesting to see how it might be
stretched, to cover PD.

(A 1970's word processor?  That's an expensive piece of kit for its
 time.  I would keep it too!)

> I take it the basic idea is that anyone may nominate a person by supplying 
> their email address.  Then others, with an interest in the choice, show 
> their approval or disapproval by indicating their trust or distrust of the 
> person nominated, or enter another person in nomination.

Trust is only for compiling the voter list.  If a few neighbours
extend their trust to you, they are vouching that your on-line
registration is honest - that you're a real person, living at the
specified street address.  That's all.  If a site administrator
already had an authenticated voter list (somehow), then Votorola's
trust network wouldn't be needed.

> Votorola maintains a running total of the number (and strength, I think) of 
> supporters for each candidate.  This provides a dynamic reflection of 
> candidates' progress through the electoral process and gives an indication 
> of the level of approval and disapproval for each candidate.  I'm not sure 
> I understand how one determines the quality of the trust/distrust the 
> voters express.  As you mentioned in an earlier post, voters can, if they 
> wish, remove their original vote and cast it for a different candidate if 
> they feel their interests are better served by the change.

(Not strength, just number.)  Yes, that's how it works, generally.
But not for PD applications.  For PD, the vote shifts will be
restricted by triads and stages.  Here's one way to restrict them:


See the configuration script called eligibility.js?  It has a
voterBar() function that is called for every vote, before it is cast,
and once again before it is counted.  So eligibility is checked on a
vote-by-vote basis.  This means we could run PD on a vanilla release
of Votorola; no modifications would be required, only configuration.
The default configuration simply returns null from the voterBar()
function, which effectively allows any voter in the list to shift her
vote as she pleases.  But the PD site administrator could change that
by adding a triad lookup, and so forth.  Votes would then be
constrained according to PD rules.

> One of the pages I reviewed was a summary of information about a voter 
> (Voter X).  It looked something like this:
> ...
> As mentioned before, I feel it is important for the people who will be 
> voting to engage in a detailed discussion of public issues (hence, the 
> provision in Practical Democracy for the electoral authority to supply 
> triads with summaries of budgets and ordinances).  I'm not sure how that 
> would be accomplished with Votorola.  I found no 'statement of position' 
> for Voter X, so I'm not clear about how I would decide to trust or doubt 
> this voter.  Even though it would not provide the kind of depth necessary 
> for sound decisions, would it be a good idea to add a section below the 
> voter identification to give a brief recap of the voter's positions on 
> issues and people or would that open a can of worms?

We can use other channels (like mail) for delivering reading material
to the voters.  And they can use other channels (mail, phone,
face-to-face) for general discussion.  Votorola is a specialized
medium, only for the purpose of communicating assent (votes).  It has
no other purpose.  (Think of it as a polling station.  You go there
only to vote.)

> The purpose and use of a voter's neighborhood (and the voter list) appears 
> to give a recap of the support and opposition the voter has attracted.  Am 
> I correct in thinking the trust/doubt values attach when Voter X (the one I 
> looked at) votes for someone else?

No, as mentioned, the trust network is only for authenticating the
voter list.  Otherwise, you can ignore it.  (Doubt signals too.)

The only measure of political support is the quantity of votes
received.  PD requires no other measure, I gather.

> Votorola could, as shown by your post, record and report the decisions of 
> the triads.  It could also, perhaps, provide a mechanism for constituents 
> at lower levels to express their views for the guidance of those they've 
> selected to represent them.

It could certainly help in recording and reporting the electoral

Where it could help in terms of guidance, I think, is in normative
voting.  Votorola allows candidate norms to be openly proposed,
composed and voted on.  This occurs in separate polls, one for each
candidate norm.  The results might guide officials in their choices of
which laws to pass, which plans to execute, and which policies to

Michael Allan

Toronto, 647-436-4521

More information about the Election-Methods mailing list