Voting in elections in the past used to be done with pen and paper. But over the past decades, countries have switched to electronic voting machines (EVMs). Already from the outset, the use of these machines has been heavily criticised by information security experts, concerned about the possibility of election fraud and the undermining of voting secrecy. But experts also raised a more fundamental concern: it’s next to impossible to scrutinise the internal workings of EVMs, and this, combined with the potential for tampering, could undermine trust in the outcome of elections.
Now, decades later, these concerns turn out to be warranted. The inscrutability of EVMs has provided politicians with an easy target for objecting to the outcome of elections – including in some of the largest democracies in the world, such as India, Brazil and the US. And the truth is: these objections could be well-founded – the nature of EVMs makes it impossible to verify these claims. This, in turn, is already sufficient to affect trust of the public in the outcome of the elections, a deeply problematic development, not only because it affects the legitimacy of ongoing elections, but also because trust takes time to earn while it can be lost rapidly, so trust in future elections is also at stake.
Still, EVMs are not going to go away any time soon. Many countries have adopted EVMs as the primary voting method and others show increasing interest in adopting them. This means that these concerns remain relevant.
One of the perspectives which has not been explored in detail is how these concerns relate to human rights, in particular the right to vote. This is relevant, because it could provide an objective framework for assessing the legality of the use of EVMs in the voting process. This study aims to fill part of that gap, by answering the question whether the use of EVMs is compatible with the right to vote, as recognized by international human rights instruments.
The scope of this report is limited. It is not about voting via the internet or other kinds of remote voting, not about other human rights aspects in relation to elections, and not about the relationship between EVMs and personal data. We will also not discuss how EVMs provide new ways to present voting options which could favour certain candidates, potentially nudging voters via “dark patterns”. Furthermore, we will not touch on the broad range of information security-related concerns existing with regard to EVMs, instead focusing only on the more fundamental concern that the inscrutability of EVMs affects trust.
One aspect which we will only touch on in passing is the principle that votes must remain secret to everyone except the voter. We discuss the requirement of voting secrecy only in relation to the requirement of transparency, and do not discuss the requirement itself in-depth. Suffice it to say, however, that the requirement of voting secrecy is also recognised in international human rights instruments, as we will see. So our conclusion that EVMs cannot fulfill the requirement of transparency while safeguarding voting secrecy should not be read to mean that voting secrecy must be watered down to allow the introduction of EVMs.
First, we discuss the human rights framework underlying the right to vote, focusing on international human rights instruments and relevant standards. Here, our conclusion is that transparency in elections is a prerequisite to trust in the outcome of elections, and as such it is a core requirement of the right to vote. Then we focus on how EVMs compare to a paper-based election process when it comes to transparency; we conclude there that EVMs have a fundamental problem satisfying this requirement. We then move on to discuss three case studies of the use of EVMs, namely in India, Brazil and the US. These cases demonstrate that the associated lack of transparency has undermined trust in the outcome of elections in those countries. We then close with observations on the role of EVMs in elections, and how the right to vote stands in the way of the use of EVMs.