[EM] Modeling Voters for Discussion and Simulation

[Abd ul-Rahman Lomax]

At 05:31 PM 12/20/2008, Brian Olson wrote:
>
>In a recent post Mr Lomax brought up the topic of poorly informed
>voters not being taken into account in various models, but I did
>introduce such a factor in my early simulations. I added some amount
>of 'error' as a uniformly distributed random variable to each opinion
>a voter held and then voted based on that errant preference set. This
>'error' factor could be interpreted as being misinformed, having
>imperfect introspection, or having a ballot failure issue in the
>polling place. Later summary happiness was measured based on the
>original true preferences of each voter.

Actually, I had something else in mind. Voter error, as described, is 
interesting, but I'm not sure that the model proposed deals with it. 
Voter error of the kind described affects "sincere utilities." So 
what is being assumed here is that there is some "informed utility" 
and that some random variable affects the voter's sense of it. 
Determining overall "happiness" based on the "original true 
preferences" is problematic. Interesting, though. I think I'd need to 
see a better explanation of the model, and possibly a better model.

Real distortion, as a result of misinformation, isn't a random 
variable. It's commonly targeted, deliberately created for an 
expected effect on the electorate, and I don't think that the 
solution is in voting methods, exactly. Asset Voting would, in fact, 
fix this problem as well as is possible. (So would FA/DP, in a 
different way.) Asset Voting takes advantage of the capacity of each 
voter to find a trustworthy representative in further process, a 
capacity which need not be dependent -- at all -- on the media, but 
rather on personal contact and judgement. Delegable Proxy would 
function similarly, creating networks of trust based on personal 
contact, not on broadcast reputation.

What I had in mind was the effect of voter *disinterest*. Ignorance 
causes a voter to be disinterested in candidate pairs where the voter 
has no information about both candidates. What ignorance would do, of 
the kind I had in mind, is to depress what might otherwise be 
perceived utilities, for all but a few candidates. It's a real 
effect, and has a major impact on elections. Using an issue space 
model will generate "fully informed utilities" unless the ignorance 
factor is modelled.