Internet voting systems have both strong support (from vendors and committed users of e-voting systems) and what appears to be equally strong opposition (from the technical community).
There appears to be two aspects to internet voting:
- Ensuring that only those authorised to vote are the only people who do vote.
- Ensuring that there is no tampering with the votes between point of voting and point of counting.
The first of these will always carry some level of risk and that risk is no different from that with traditional paper based voting systems where the identification of the voter is required. How are we to be absolutely sure that the voter is in fact the voter we have on our list of authorised voters? What is proposed here is that solutions to the problem be allowed to evolve with time and advancing technology. Initially we could start with distribution of a "vote authorisation token" to authorised voters via the postal services. Though this is not perfect it appears to carry fewer risks than what might be achieved currently with electronic distribution. Also, the basis for most voter eligibility is residence. Eligible non-resident voters also have some physical connection that authorises them to vote such as being stationed somewhere or having equity in a property located within the electoral area.
Assuming that the distribution of a "vote authorisation token" is viable the next question then is how might we ensure that a cast vote progresses to the counting process without intervention?
The proposal is that the same principles used by bitcoin can be used to both transmit votes (transactions) and count them (account balances).
We propose that the distribution of "vote authorisation tokens" be nothing other than a bitcoin-like transaction of a vote from an electoral commission to the voter. This exactly parallels the payment of an amount of bitcoin money to a voter. The voter's balance then represents the number of votes he or she is authorised to cast. The voter in casting a vote is in effect making a payment-like transaction from his/her vote "wallet" to the "wallet" designated to be the wallet of the vote counting point for a candidate - in other words the candidate's "account" held on a publicly visible "wallet".
The handling of the vote "transaction" is then exactly the same as that for monetary transactions in bitcoin. The "miners" would then be those organisations and individuals who have the most interest in the outcome of the election or contest. Major political parties would invest the most to ensure their rivals do not have dominance over transaction validation.
Every voter would be able to trace his or her vote through the transaction history giving full transparency.
The results would be available in real time - even where complex distribution of preferences are utilised. In such cases the false "random" nature of distributions would also be eradicated - leaving one and only one path through the distribution.
see also "what is a bitcoin"
i discover i was not the first person to think of this