Peter P. Jones wrote:
> Does the model (for a voting advice trust network) go beyond what
> is happening at sites like http://www.opensecrets.org/
> ?
God, yes. Their site is devoted to political solutions--campaign finance
reform and the like, all of which are good things. But when the dealers
are writing the rules, they're not about to change the rules in ways that
make them lose. (01)
As I've tried to indicate in this post, by hitting the highlights, the goal
of this system is to use technology in such a way that money becomes
virtually /irrelevant/ to politics. (02)
This is an issue I've been wrestling with since 2002, or slightly earlier.
I wracked my brain for years, open to any solution that would work.
It can be done, if we use existing technology appropriately. (03)
Unfortunately, /any/ alternative that attempts to reform the system from
within has the problem of being a likely non-starter or, even worse, a
lip-service issue that produces cosmetic reforms with no real changes. (04)
It's a basis tenet of game theory--if you're losing, /change the game/.
Appropriate use of technology changes the game so that decisions are
no longer /made/ in the polling booth, they're only /recorded/ there--
much as the electoral college used to be a decision making body,
but is now only a recording device, except in exceedingly rare
circumstances. (05)
>
> Eric Armstrong wrote:
>> For folks who aren't subscribed to my highly sporadic weblog,
>> here's a post I made this morning about the goodness of Ruby:
>> http://www.artima.com/weblogs/viewpost.jsp?thread=152464
>>
>> The reason for mentioning here is that understanding the
>> technology made it possible to see how to implement a trust
>> network for voting advice using RSS feeds.
>>
>> What follows is an outline of the value propositions and the
>> design. For those who want to know more about how it will
>> make money irrelevant to politics, see
>> http://www.citizensAdvisory.org
>>
>> Note:
>> I say "candidate or issue" a lot in the following. I need a
>> short term that means either one. I could probably go
>> with "ballot choice". But the ontology needs a single word
>> for the concept, ideally.
>>
>> Value Propositions
>> =============
>> The value proposition for the user is this:
>>
>> * There is one place you can go to get voting advice from people
>> and organizations you trust, on every candidate and issue you're
>> eligible to vote on.
>>
>> * Before you go to the polls, you look at your recommendations.
>> For many issues, all of your advisors agree. Your choice is
>> a given. For some issues, there may be disagreement. You
>> go to the pages the choices link to, where your advisors give
>> their reasons. You make your choice.
>>
>> * You print out a list of voting recommendations that you can
>> take to the polls.
>>
>> * You now are ready to make an informed choice about
>> every candidate and issue you're eligible to vote on, in the
>> same way that you get your car fixed and your plumbing
>> fixed--by depending on people you trust.
>>
>> The value proposition for the advisor is this:
>>
>> * You can easily reach every person who has ever decided to
>> trust you at any time in the past.
>>
>> * You can do so at little or no cost. So money is no longer a
>> barrier to outreach. Rather, it depends on the clarity of your
>> explanations, the quality of your recommendations, and how
>> long you've been making them.
>>
>> Overview of the Design
>> ================
>> The basic mechanism looks like this:
>>
>> * Build a data-aware aggregator for voting advice.
>>
>> * Define a content schema for the message body.
>> (Tags might also work, but I'm not sure.)
>>
>> * The schema defines constructs like state=xx, region=yy.
>> (Start with the national schema, which is easiest. Even
>> then, there are senate and representative regions to
>> consider. Progressively refine the system over time
>> until it gets down to the local dog-catcher level.)
>>
>> * Users subscribe to RSS feeds for voting advice.
>>
>> * The advice comes with a candidate name or ballot choice,
>> and a link to a reason, encoded in the XML
>>
>> * The aggregator listens to the feeds, and displays any
>> choices that match the user's stored information. So
>> the aggregator gets the broadcasts for all 50 states, but
>> only displays them for the state(s) users have selected.
>>
>> * Create region-specific versions of the configuration
>> files that identify state, senate district, and so on.
>>
>> * Team with the League of Women Voters to identify
>> which configuration file to use, based on user's address.
>>
>> Implementation Notes:
>>
>> * If the League of Women Voters or some similar
>> organization makes it possible to access their database,
>> use that. Otherwise, crawl their web pages to gather
>> the information as needed.
>>
>> * Ruby makes XML parsing easy, and it includes the
>> TK windowing system, among others. So a reference
>> implementation can be created with minimal effort.
>> Ruby plugins are available for browsers, too. And there
>> are server side frameworks like Rails, so there are a variety
>> of ways to dice the cat.
>>
>> Open Issues:
>>
>> * As soon as we get below the state level, naming becomes
>> a problem. For states, we can use two-character land
>> mail addressing codes. That makes it easy to ensure that
>> advice is sent out with a code that the receiver will
>> recognize. Below that, there will be probably District 11
>> in some states, Parish 33 in others. So the schema names
>> (parish or district) and data values (1..11) will need to be
>> incrementally refined at each level.
>>
>> * Wiki pages can be set up for people and organizations
>> to give advice. Each person or organization would have it's
>> own web. The RSS feed would most likely be for the person
>> or organization, rather than for individual pages. To work, each
>> page posted to the Wiki needs to be tagged with identifying info
>> (state, district, etc), all of which needs to be driven by the same
>> schema(s) that advice aggregators use.
>>
>> * A search capability for advisors would be a cool thing. Ideally,
>> you would have the option of making your voting history
>> available so that the system would be able to find highly
>> compatible advisors. (But it should therefore be possible to
>> revise the voting history to retract a vote for someone or
>> some thing that turned out to be a mistake.)
>>
>> What the System Leaves Out
>> ====================
>> A couple of highly beneficial features described at Citizens' Advisory
>> depend on the ability to ensure that a user is only registered with the
>> system one time, that the user is a real, live person, and that their
>> anonymity is preserved.
>>
>> The technological foundation for that capability doesn't fully exist,
>> yet,
>> so in the interests of making progress, we have to do without it for
>> now.
>> In consequences, some highly desirable features of the full system have
>> to be left out:
>>
>> * *Counting* RSS feeds are nice, but I wish I knew how many
>> people are actually listening. Mailing lists are good for that.
>> Web counts for static HTML pages were decent, too. But what
>> I'd really like to know is how many people are listening, and is the
>> number growing or shrinking? That number translates into
>> political influence, which empowers advisors with the ability to
>> influence policy /between/ elections, so it's important.
>>
>> * *Multi-party Coalitions* If several organizations can easily
>> pool their numbers /and eliminate the overlap/ -- which is hard
>> to do outside of a system like this -- then they can find regions
>> where they have a good chance of determining the outcome of
>> an election, and focus their efforts there.
>>
>> A perfectly implemented system therefore has the capacity to organize
>> effective multi-party coalitions in cyberspace, even if there is a
>> two-party
>> system on the ground. That's important, because right now no one can
>> afford to vote for the Green party candidate unless it's clear
>> they're going
>> to win. That's never a given, so the votes never come--and we never
>> get a true measure of how much support they really have.
>>
>> Instant Runoff Voting (IRV) would change that, so that's another measure
>> I favor. But the voting advice trust network could also change that
>> reality by making it possible to determine, for example, that in a given
>> region a coalition of the Green party, Greenpeace, and the Consolidated
>> Federation of Inveterate Birdwatchers probably has enough votes to win.
>>
>> Summary
>> =======
>> Although I haven't gone into details here, a trust network for voting
>> advice
>> can make money irrelevant to politics. We can build a fairly
>> decentralized
>> version of that system using RSS feeds and smart aggregators. The
>> heart of the
>> system is the standard message format, accompanied by reference
>> implementations
>> for an aggregator and a publisher.
>>
>> The reference implementations will allow for widespread adoption and
>> encourage
>> better, more fully-featured versions of the tools. Some desirable
>> features will
>> have to be left out, at first, because they depend on the ability to
>> preserve
>> anonymity in a way which also guarantees that a living, breathing,
>> vote-eligible
>> human being is at the other end of the RSS feed, and that only one
>> feed is
>> being counted for that person. But in exchange for giving up those
>> potentially
>> valuable features, we gain the ability to implement something that
>> will work
>> with technology that is available today.
>>
> (06)
--
This message is archived at: (07)
http://collab.blueoxen.net/forums/cgi-bin/mesg.cgi?a=tools-yak&i=441E3362.2080206@sun.com (08)
|