[HacktionLab] Fraud proof voting - will it work?

[Alan Dawson]

On Mon, Dec 09, 2013 at 07:09:07PM +0000, penguin wrote:
> Hi
> 
> Through work, I've come across this thing called fraud proof voting -
> it's been developed by some guy with a (seemingly) specific focus on
> Africa. I wondered if any crypto-experts on this list had any views on
> how fraud proof it really is.
> 
> Details here
> http://cd3wd.com/SEEV/fpv.htm

"The SMS and the envelop data files are destroyed immediately or within a fixed time period after the final result is announced (since they can be used to track the voting choice of every voter).. The data may be used to produce aggregated results down to a specified level (e.g. down to population centers of 5,000 or 50,000) but no further – to avoid retribution by disappointed parties."


This seems like a flaw to me.  3 powerful actors - the voting organisation and the auditors, have access to non anonymised copies of all the votes, also it hopes that those actors do not conspire together to fix the result, and there is no way to verify the results again later, as they are destroyed.  So mathematically I don't think it  provides a secure, anonymous, verifiable, system.  It may be practical though

There are solutions to anonymous voting that are cryptographically secure - like mix nets, but I think they fail ekes' test of understandability.  http://crypto.stanford.edu/pbc/notes/crypto/voting.xhtml

Regards

Alan Dawson
-- 
"The introduction of a coordinate system to geometry is an act of violence"
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <https://lists.aktivix.org/pipermail/hacktionlab/attachments/20131210/9854f91f/attachment.pgp>