Electronic voting could help us but ..
The recent election and the attendant logistic and operational failures once again has generated a loud call for the manual voting to be replaced with an electronic or digital voting system. Many lay people as well as knowledgeable tech personalities have come out to state that this is doable and even to announce a project to deliver same.
To be fair, the frustration with the logistics issue is fully understandable and justified. The scale of the logistics challenge should also be appreciated. Voting materials have to be delivered to almost 120,000 polling units across Nigeria. This distribution is being sabotaged every step of the way by various players intent on either sabotaging the election or rigging the election.
If the elections were digital, these logistics challenges can be overcome or so the expectation goes.
Everyone points to our experience with voting for BBN or withdrawing money from our ATMs or even using USSD to transfer funds and make payments. These systems work and work seamlessly from various locations and their results and effect happen almost instantaneously. Why therefore can we not do our voting same way? Really why not? If we can trust these systems to handle our money, why can’t we trust them to handle our votes?
To answer this question, it is important to differentiate our elections from voting for BBN or transferring money by USSD. The first thing to keep in mind is that elections are far more contentious than either of these. It is therefore extremely important that contenders and the public are able to see the elections as being free and fair. A few conditions usually contribute to elections being considered free and fair.
First and foremost is that as much as possible, voters are not able to prove to anyone who they voted for. This makes it impossible for any one to reliably buy votes or to coerce anyone to vote in their favour. That’s why we can campaign for people to collect their money and vote their conscience. If I can prove to you that I voted one way or the other, then it is easy for you to pay me or to punish me. Any digital voting platform will have to preserve this ie no one can prove that any one voted for any particular contender. This immediately rules out using your mobile phones as the mobile phone network identifies the phone and therefore it is possible to prove that a particular phone voted for a particular candidate. Considering that we have used option A4 before to elect a president, this requirement can be considered as something that can be waived. Given the experience with vote buying and coercion witnessed in the last election, if waived, it becomes an open season for subverting the elections with massive voter coercion and vote buying.
Another difference between voting for BBN and voting for any elective post is that the contenders and the voting public must have confidence that the result and conduct of the election can be challenged in court. Note that ensuring actions can be challenged in a law court is a major plank on which the democractic society is based. No one is likely to go to court for losing a BBN contest. Challengiability imposes a unique constraint — it must be possible to prove in court using the artefacts of the election that the elections were free and fair and also that the ensuring results reflect the proper collation of the individual votes. In a paper vote, you know that at the worst case, the ballots can be retrieved and recounted without compromising the identity of the voters. Of course there is a possibility that the ballots could have been stuffed as is done regularly in Nigerian elections. In an electronic election, you will have to prove that the electronic votes are genuine in themselves. This looks trivial until you realise that proving that an algorithm is correct and that the system in which it is stored is free from intentional and unintended bugs or back doors in a law court is not a trivial thing. How do you prove that the log of the telco was not tampered with? How do you prove that the vote sent from the phone was not tampered with in transit? How do you do that while also preserving anonymity at the same time? In cases where these features are to be implemented in software, how do you prove in a court of law that the software does not have bugs? Can you even prove these technically not to talk of proving it in a court of law?
And of course, people also raise the hype of using blockchain technology. Blockchain relies on every one in the network having a copy of every transaction. Anonymity and challengeability are not one of its promises. And it can definitely be hacked or stolen. Its applicability to democratic elections as we know it is immediately of little use.
These two conditions of anonymity and challengeability alone overshadow the challenges of cybersecurity or infrastructure that already exist.
These issues are not specific to Nigeria. Electronic elections are being researched continuously worldwide and have been researched for decades. The only country with a nationwide digital voting is Estonia, a county of less than 2 million people with high grade digital infrastructure.
It will be more useful to start looking at the voting system and providing technology solutions to particular problems while preserving the anonymity and challengeability. For instance, is it possible to make the ballot paper itself electronic for instance to make it to be e-ink such that they can be reused? Can everyone collect the e-ballot same time they collect the PVCs? Can we make the card reader to be the one that activates the e-ballot so that it cannot be used elsewhere? Or can we have a dispenser that dispenses ballot paper at the voting booth? I am sure there will be a lot of wild ideas but we should constrain the ideas into what can fit into our democratic election protocol especially preserving anonymity and ensuring challengeability.
