News and events Tilburg University

Can blockchain contribute to combating electoral fraud?

Published: 26th February 2018 Last updated: 12th April 2019

Gijs van de Water’s Master’s Thesis
Gijs van de Water, 24, recently got his Master’s degree in Information Management. The Information Management program prepares students for a career at the interface of information technology, business and management. Before Gijs started on the Master’s program in Information Management, he studied computer science at a university of applied sciences.

In determining the topic of his Master’s thesis, he was looking for a subject that would not only allow him to apply his knowledge and skills in the field of ICT but that also linked up with a currently relevant societal issue. That is how he came up with the question if blockchain technology can be employed to make elections less vulnerable to fraud.

There are many examples of elections abroad that prompt suspicions of fair amounts of vote rigging taking place. Blockchain technology is known to be safe, so Van de Water’s hypothesis was: if electoral votes are put on a blockchain, nobody can tinker with them, and election results will be fairer and more transparent. 

Blockchain

Blockchain can be seen as a ‘write-only ledger’: you can add records, but records cannot be changed. Theoretically, this means that if you put a vote on it, this vote cannot be manipulated in any way.

Devil’s Triangle

Gijs first established what criteria countries need to satisfy in order for them to be able to organize free and fair elections. These criteria, determined by the UN, include the following: People are allowed to vote only once, the results may not be fixed in advance (uncertainty), and the elections must be open to all and transparent. He subsequently analyzed the steps in the voting process (from the voting booth to the counting of the votes) and checked if various models of blockchain technology could replace steps in this voting process. In doing so, he continuously checked if the criteria that voting processes need to meet were indeed being met. He found out that whichever model he applied, there were always one or two criteria that were not being met. In each case he would end up with a ‘devil’s triangle - a criterion at each corner, two of which would be met, but never all three. For an election process to be fair, however, all criteria need to be met.

In a public blockchain, the criterion at issue is uncertainty, because people can witness live how ballots have been cast. This may influence the voting behavior of people who still have to vote. If you opt for a private blockchain, in which only one or a few organizations (such as the EU) can see what is going on, the criterion of uncertainty is met, but the criterion of transparency is not, because voters do not know if this organization is politically motivated to change the results. There are methods available that make it possible to vote in a public blockchain without people being able to witness the results ‘live’. However, this method requires voters to come to the voting booth twice, which diminishes accessibility. An additional problem is that computers are not transparent: a voter voting on a computer for Party A can never be sure if the computer does indeed save the vote at Party A. The computer may have been hacked, or the manufacturer may have manipulated the software. While blockchain is fairly safe, things can still go wrong at the surface.

So unfortunately the conclusion is that blockchain is not going to help us make voting less vulnerable to fraud. We will have to look for other methods to make electronic voting safer.

Gijs van de Water, Blockchain ballot. Electoral Enhancement or Danger to Democracy? December, 2017. Supervisor: dr. H. Weigand.

Masterthesis Gijs van de Water