Working on the Metagovernment project, I made this technical analysis of online voting:
We emphasise making decisions by coming to consensus and synthesising proposals rather than deciding through a majority vote. This may be impractical and idealistic if put to the extreme. It has been shown practical in small scale open source projects. Even if it isn't possible to reach consensus for every decision, the fact you are forced to decide on something with a vote raises a flag that something is wrong and that there is the possibility of discrimination against a minority.
In any voting scheme there are some criteria to be considered. I've taken the voting principles used in Germany as a basis (Grundgesetz Artikel 38 Â 1).
Voting must be open, direct, free, equal and secret.
Open means anybody must be allowed to vote.
Direct means that the decision isn't made through intermediaries.
Free means that no pressure may be exerted on the voter to vote
against there will.
Equal means that every vote has equal weight.
Secret means that nobody should be able to determine how an individual voted.
For each of these there are some considerations for online voting.
Open: Using computers to vote may be discriminatory against those who aren't comfortable with them. On the other hand it may increase participation, by making voting easier for those that are comfortable and for people with disabilities.
An exception to this principle is made for people under 18. The intent is for the electorate to have a certain degree of experience. The system would be much more open though, if anybody could participate based solely on merit.
Direct: Representative democracies are by definition indirect, so you could argue, that in spirit Artikel 38 is being violated with the current system. I think it was mainly introduced as a reaction to the electoral collage in the US. Online governance has great potential to do this principle justice.
Free: I can't think of any system that can guarantee this other than a voting booth in a public venue. By voting on the Internet from any computer the risk is introduced of a third party having control over that resource and forcing the voter to vote a certain way.
I think this problem is inherent an haven't been able to come up with a solution in the few years I've been thinking about these systems. I can only find a partial justification that in a civilised society this should be a statistically insignificant problem.
Equal: No difference as far as I can see.
Secret: I could talk for hours about this but the best solution I've seen so far tries to limit the number of parties you have to trust to as few as possible. And even that system relies on Public-key cryptography, which due to a lack of understanding and thereby transparency may not be trusted by every voter, even if they trust the one party they are sending their votes to.
Considering all the problems with online voting, my conclusion is that we should reconsider voting not as a binding for decisions, but as an indicator for the direction the community wants to take, so that gradually consensus can be reached.
The entire thread can be found here: http://thread.gmane.org/gmane.politics.activism.metagovernment/1048/focus=1056