Digital voting is one of the interesting things suggested as a way forward in the digital democracy. In the Netherlands, all citizens can already apply for a digital id, which can be used for the taxes. This id is suggested as the carrier for modern voting solutions using the Internet. Although digital voting seems to be the future, it is most certainly not ready yet.

Of course, everyone will shout “security”, and they would probably be right. Nevertheless, coming from the scientific field of information security, I can state that we are able to make digital voting protocols at least so secure as their physical counter parts. However, there are still issues, that are mostly on a social level.

It is secure! Or do you not believe me?
If a group of cryptographic protocol experts were to build and test a digital voting protocol, they will probably come up with a beautiful protocol and a lovely proof of its inner workings. Actually, they already have done so. If you fire up Google Scholar and search for “voting protocol”, you will probably find loads of interesting reading materials.

Nevertheless, these papers and protocol description account for huge piles of very mathematical information. For that reason, it is not so strange that one has to be a cryptography expert to really understand what is going on. This makes that the majority of the citizens have to believe it works. Yes, I understand it works that way with a lot of things in society, but in most cases there are no real other solutions. However, in this case, the analogue system is very easy. Everyone can go to the voting office and watch how the votes are being cast and counted, which constitutes a very transparent system.

Additionally, when one proves something, one does this based on a model using certain assumptions. Therefore, one also has to show that the system actually running the voting protocol is equal to the one proven secure and correct. In other words, who is going to make sure that all those nice protocols are really used?

Anonymous: yes, strongly anonymous: no
Voting protocols protect anonymity. In fact, it is one of the core requirements for this kind of research. However, this anonymity starts at the borders of the system. Therefore, it does not protect against shoulder-surfing or coercion.

With the traditional voting system, one goes into a voting booth where he and only he can fill out the ballot. This does not only guarantee anonymity, in the sense of privacy, but also strong anonymity. The latter means that, if one were to sell his vote, he could not prove he voted as he told. Guaranteeing this so-called strong anonymity is very difficult in digital voting protocols.

Digital Voting: Not in the near future
For these reasons, digital voting still is viable, but there are some issues that have to be taken care of. First of all, strong anonymity is very important and needs to be guaranteed. Secondly, the system needs to be proven to work in an intelligible way for all citizens. Thirdly and finally, the theory needs to be shown to be applied in practice, not only on paper. If we can come around these problems, digital voting will be the future, just not tomorrow.

2 Responses to Digital Voting: Possibly the Future, but not Tomorrow

  1. Floor Terra says:

    Although it’s not entirely independant of the points you mentioned I want to mention “single point of failure” as an issue. Lots of electronic voting solutions rely on a single vendor to provide hardware, software, backend services, etc. Given the value (monetary and otherwise) of an election any significant concentration of dependance on any party should be met with the assumption that this party is compromised and is in some way trying to influence the outcome.

    I have never seen a voting protocol that is simple to verify, doesn’t rely on trust in some large party and still is significantly cheaper/faster/reliable that good old-fashioned paper voting.

  2. […] different countries, the Netherlands still vote using a paper ballot and a red pencil. Previously, I made a case why digital voting schemes are insecure. However, in our fear of digital solutions, we tend to forget that traditional methods are not […]

Leave a Reply

Your email address will not be published. Required fields are marked *