Long time since last post; but the reasons as always being that I am busy trying to get in rhythm, learning to cook efficiently :P and so on..
Recently Australia had its first experiment with e-voting, which while good doesn't exactly fits the ideal description of what is expected from e-voting schemes. I myself am not a firm believer in the e-voting systems. The main problems that are not answered, which according to me as the bases of any voting scheme, are the properties of Verification, Privacy and Classification
- Verification: How to verify if votes casted had lead to legitimate elected candidate? How is it possible for voters to verify if the person elected was indeed from majority?
Example: In a classical voting system, a voter puts a vote on paper, folds it and drops it inside the designated box. In this case, a voter knows for sure that his/her vote is being considered and would be counted in the final outcome. Furthermore, the voter can also be assured that the vote cased is a real and valid vote. This verifiability property seems to be missing from e-votings. Although it is worth noting that many protocols do offer this property where the voter can check his/her vote against final tally. (Estonian voting protocols are prime example)
- Privacy: In an ideal voting system a voter will be free from social aspects of voting i.e. from the pressure of friends, family and society in general. In classical setting, a voter is free (theoretically*) to vote as there wont be anyone peeking over the shoulder of the voter or a family verifying that it is voted for bias voting. In e-voting, this scenario seems unlikely! How and when would be voter cast a vote such that he/she is not being influence by social environment or is isolated enough to cast a vote that is in a sense anonymous. Given this situation, e-voting makes a very bad candidate for a democratic e-voting system.
Some protocols do exist that allows for the voters to re-cast votes and the last vote casted is considered as a valid one. But that does not exactly provide a optimal solution.
- Classification: How would a vote that been nullifies be verified as being intentionally nullified or due to some technical glitch?
Example: (Valid for India) Whenever a voter wants to decline voting due to no suitable candidate standing for the election, then he has two options:
- Declare that at the polling station which might possible earn him retribution from the winning candidate after the election session.
- (in case of Ballot paper)To invalid the vote by stamping (casting vote) at odd location or stamping more than one candidate.
This option of invalidating the vote (and remaining anonymous), besides not casting the vote, seems to be missing in e-voting. Furthermore, current e-voting systems are also not being 'democratic', so to speak.
* Theoretical in a sense that there are lots of realistic stuffs that do infact influence the voting outcomes (Bribing, fake-voting, vote-for-money...)