[EM] Idea Proposal: Listening Democracy

[Abd ul-Rahman Lomax]

At 07:55 AM 4/20/2010, Raph Frank wrote:
>Btw, you should look into the delegable proxy system.  This is also
>designed to allow effective communication without overloading the
>voters.

Asset Voting was invented when Dodgson realized that most voters were 
people busy with their lives. They were not experts on politics. We 
see constant efforts to get "ordinary people" involved, to get them 
to vote, as if more people voting was necessarily better. It's not, 
unless the system collects good information from them, and when most 
people don't have adequate information, but vote based on media 
impressions (on what else would they vote?) STV -- Dodgson was an 
early analyst of STV -- depends on voters having more information 
than just their favorite candidate, which effectively disempowers 
those with only that much information. Unless party affiliation is on 
the ballot, voters can make assumptions based on being a party 
nominee, but this, then, raises the power of parties and defers the 
exercise of power to them and their nominating process. (Party 
affiliation is, in fact, a solution to the problem of voter 
ignorance, but ... begs the question. How, then, do parties choose 
nominees? If their own process is no better ....)

With Asset, proposed by Dodgson as an STV tweak to deal with 
exhausted ballots, the voter who only knows who their favorite is can 
empower that favorite with one vote. I have no evidence that he 
realized the probable effect of this: a multiplication of candidates, 
even a vast multiplication if the rules permit it. The number of 
candidates could grow so large, and votes so broadly distributed, 
that I'd expect the practice of printed candidate names on a ballot 
would disappear. All votes would be "write-in," though it might be a 
code from a booklet. (For practical reasons, the name has to be 
unique and assigned to the candidate through a registration process. 
San Francisco, I'll note, allows "write-in" votes, supposedly, but to 
be eligible for election, write-in candidates must be registered. 
What San Francisco did that was actually quite offensive was to, 
then, pass a law that write-ins could not be registered for a runoff 
election, basing this on the "promise" of the reform that implemented 
runoff voting that the winner would have a majority. In other words, 
we will guarantee a majority by excluding candidates. Great. No 
wonder they were suckers for RCV. The law was actually tested, once, 
in the last election before RCV was implemented. It's quite possible 
that the write-in would have won. Consider this: if a write-in 
candidate came in third place in the primary, this was actually a 
very strong candidate, if the margin was not very large. Voting 
systems reformers missed a huge opportunity when this was being 
litigated before the California Supreme Court, probably because 
everyone "knew" that runoff voting was worse than IRV. But IRV with a 
true majority requirement (as Robert's Rules actually recommends, not 
what FairVote has claimed over and over), and a runoff with write-ins 
allowed, is not a bad system at all, though Bucklin is better 
(avoiding far more unnecessary runoffs).

In a place like Burlington, a dropped candidate, being the Condorcet 
winner, could win in a runoff. It gets even better if an advanced 
method is used in the runoff, allowing voters to cast ranked votes, 
or other forms of alternative vote, just in case. (Range methods, 
including Approval, are variations on alternative vote; it is just 
that the "alternatives" are simultaneously considered, with full vote 
values in Approval -- full votes sequentially added in with ordinary 
Bucklin -- and fractional alternative votes in Range, or the 
variations on Bucklin that assigned fractional vote values to 
lower-ranked votes.)

Asset, though, can make for a very simple voting system, it works 
fine with vote-for-one or Approval or Bucklin or IRV, even.

Consider what Asset would have done for single-winner in Burlington, 
assuming that the Progressive didn't get a true majority. (I forget 
if that actually happened there). The canvassing would have 
completed, with the Progressive and Democrat being the leaders. The 
Republican would hold a large number of exhausted ballots, and could 
decided the election. Almost certainly the Democrat would win, and 
would be, more or less, forced to govern from the center (as would be 
more or less expected anyway), representing, to some degree, a 
Democrat/Republican coalition. The Republican, if not sufficiently 
"appeased," could say, "Screw it! I'm leaving these votes uncast, to 
make a point." Another option with Asset could be that the 
Progressive and the Democrat, say, come to an agreement and with 
Asset, they could even agree on another winner than themselves. We 
have this idea that candidates are purely selfish. Some are, some 
aren't. Asset would make purely selfish candidates far more visible!

Asset would make for really great news stories. Maybe we could 
convince the media to support it as being butter for their bread.