A different approach to improvements on the Electoral college.

I have never been truly satisfied with ranking as a way to be rational. It makes for self consistent theory, and while the theory adequately models some realities, I am prepared to argue that the model should model, not actively direct. (Go Adam Smith!)

As such, the rational political _basis_ for breaking up the Left-Right continuum into coalition parties attends a jaundiced bias to my judgment. I may be prepared to recommend it in future anyway, given time for observation in Canada and Germany. Last Election cycle, Canada nearly implemented a Coalition Government, and while I was happy to see Conservatives win out, I had anticipated following the coalitions for adherence to platform promises. Israel has demonstrated over it's short history that even a coalition government can be militarily decisive - I simply can't claim to adequately analyze their news.

If my language is grandiose, please put it down to sophistry - I am enamored of a new idea. The question relates to a methodology, and I don't feel very audacious about it yet - it's a question.

Can rational political voting be studied, modeled or improved by a suggested experiment?

Traditional "Rational Voting," behavior is established by ranking (assigning ordinal value to) ideologue listings of Party positions. If this is so, this allows improvements of weighting and multiplying now sorted weight by intensifiers - if ranks are wrong after weighting, what happened? I'll admit something did happen, but it may no longer qualify for the traditional moniker "rational."

This system can theoretically be extended out to Senators, Congressional Representatives, State and Local government participants, until the rational condensate qualifies for representation by the electoral college system.

That's school environment language for saying take the lists below, decide which things you like in each column, and decide how much you like them. One way to quantify "how much," would be to specify how much money you would spend on each one. Then compare the positions of the candidate you are evaluating on each, and add up the money you agree about on the "plus" side ONLY. The compare the same procedure on the other candidate, and the two added up numbers are supposed to tell you which candidate you "really" want - a rational decision.

My suggested experiment is similar but aims to be instructive anyway:

What if, instead of Party Specific Lists (an abbreviated illustration below)

Republican Democrat

Gun rights Gun Control
Anti-abortion/limits Government subsidized abortion
Tax Reductions for economic stimulus Direct Deficit Spending for economic stimulus
.
.
.
Publius Cato

...we made one authoritative questionnaire listing, used language official to all relevant parties for each item, and had EVERY candidate fill it all out. This would allow a quicker rational count by the average citizen accomplished by personally weighting each issue, and reviewing some "representative" report of his available candidates for ballot casting. The time saving might not be easily observed - citizens typically limit the informed _qualification_ of their vote on time, not the vote count itself. What might be better observed is the true net effect of rationally defined ballot casting.

Does this appear to be a good experiment? What political ground zero would be the best starting place for such an effort? The Libertarians? A PAC or Citizens' advocacy group?