Electronic voting (eVoting) schemes have been getting a lot of attention lately. This is true in the press see Changes in e-Voting Likely Coming, Experts Say for one example and also in the academic world (at CRYPTO 2006, one of the leading international cryptographic conferences, there were two published papers and one keynote talk on electronic voting).

The cryptographic community has been publishing papers on secure eVoting protocols for years now. These solutions have focused almost solely on making the protocols more efficient and more secure. This is of prime importance, because the correctness of the election result (and the privacy of voters) must be guaranteed even if voting machines are corrupted. Nevertheless, recently our understanding of what is needed to make an eVoting system successful has changed significantly. Specifically, in addition to the system being cryptographically secure, voters (without expertise in cryptography or computer security) must be convinced that the system is secure. For example, the system must be transparent in that voters can verify that their votes have been counted correctly. This is not a new requirement. However, what is new is that this verification must be possible and convincing for people with no mathematical background whatsoever. Furthermore, voters must be convinced that the overall system is one that works (needless to say, a long mathematical proof of the security of the system is not going to help most voters). In short, it has become clear that voter perception is an integral factor that must be taken into account when designing any eVoting system. One very interesting system that takes significant steps in this direction has been suggested by David Chaum; the paper can be found at http://www.voterverifiable.com/article.pdf.

My greatest concern when it comes to eVoting systems is that it is quite likely that the solutions that will eventually be deployed are those that belong to large companies with strong marketing departments, and not those that provide strong guarantees. It is also possible that a solution will be chosen that is very convincing for voters (and thus excels when it comes to voter perception) but can be easily rigged.

We need to make sure that eVoting solutions capture both requirements (i.e., are both secure and convincing). Systems with high security and low voter perception will never succeed, and systems with low security and high voter perception are a threat that no democracy can afford.