Biometric Voter Registration (BVR) in Zimbabwe: Dispelling the myths

There have been a number of publications in the local media, quoting comments from various ‘experts’, and citing developments elsewhere, using these to cast aspersions on the use of biometrics in the upcoming Zimbabwe elections.  Examples of these publications are “Red Flag over Biometric Registration” (The Herald, 11 March 20017), “France’s Cancellation of e-voting: Eye-opener for Zim” (The Herald 9 March 2017), “BVR, A Luxury We Cannot Afford” (The Herald, 13 March 2017 – Editorial Comment) and most recently “More Thumbs Down for Biometric Voting” (The Herald, 15 March 2017). This effort has been systematic and sustained, culminating into a Newsday publication (16 March 2017) screaming “2018 Polls Hang in Balance”. All this comes after the tender process has commenced and a shortlist of companies compiled – maybe just a coincidence. This, however, is the political side of the process which the author will leave to political analysts.  

What these publications revealed was a clear lack of understanding of the BVR process. This lack of understanding and “misinformation” is being used to discredit the process culminating in the set-up of an agenda giving cues to the abandonment of the biometrics project. This article is intended to correct some of this misinformation and misinterpretation of developments elsewhere. It also aims to clarify the proposed Biometric Voter Registration and Verification process (BVR) which Zimbabwe Electoral Commission (ZEC) is proposing, and has been successfully used in other countries.  

The common theme in these publications has been the misconception that ZEC is going to implement ‘biometric voting or electronic voting’.  This then set the basis for the claim that the system would be susceptible to ‘cyber-attacks’ and ‘hacking’  which would derail the voting process and dis-enfranchise voters, citing France’s abandonment of electronic voting as an example.  ZEC is not proposing to implement ‘biometric or electronic voting’; it is proposing a model of BVR which is very different from electronic voting (even though it can be used as a launch pad for electronic voting). Additionally, the process being proposed is not more vulnerable to cyber-attacks or hacking than any other electronic voter’s register or database. This will be further explained in this article.

The call for the employment of technology in Zimbabwe for both voter registration and facilitation of the electoral process is not new. The issue has been raised in parliament several times (Tongai Matutu, 2010, Pishai Muchauraya and Nelson Chamisa, 2012 for example). The author of this article also advocated for biometrics to be used in a 2012 publication and in a number of follow-up articles thereafter.

The intention to introduce biometrics in Zimbabwe for the 2018 elections has enhanced ZEC’s credibility, and should be applauded as a step in the right direction. Zimbabwe is not re-inventing the wheel, but is following in the footsteps of other countries including Ghana, Benin, Tanzania, Togo, Mauritania, Ivory Coast, DRC and Nigeria among others, which have successfully pioneered this technology.

Before, dealing with the issues that are being raised in the recent publications, a brief explanation of biometrics is given here. Biometrics refers to human physical and behavioural characteristics such as fingerprints, the iris, signature, face etc. These can be used to uniquely identify an individual.  This concept is definitely not new! Zimbabwe has been collecting people’s biometrics for decades; everyone has to have a picture taken and fingerprints captured to obtain a national identity (ID) or passport.  This background and reference is important because BVR is just similar to this process.  In BVR, a voter’s details (name date of birth, address etc) are digitally captured and stored alongside their biometric features (face and fingerprints) on a computer– that’s it. Nothing more nothing less! The advantage of this system is that these biometric features can be used to uniquely identify an individual in a computerised way and additionally , there is inbuilt software to identify and eliminate duplicate voters/registrants; leading to a clean voters roll.

The deployment of personnel for the purpose of collecting BVR information is not different to that done in order to register people in the “old way”.  Personnel will be trained and equipped with mobile voter registration kits. These are portable devices designed to create electoral rolls; equipment that is reusable, extensible and resistant to adverse conditions. These devices are self-contained, autonomous units supported by long-life batteries and can be used in remote areas for registration, even within homesteads. In the end, what is compiled is a normal database or electoral register which includes biometrics information.

The second part of the process is voter verification or authentication which happens on voting day. This is whereby a person appears on voting day, presents an ID or provides a name. The person’s biometrics face and/or fingerprints are then captured and compared to those in the computer database (biometric voters’ register). Again mobile biometric kits/stations are available to achieve this, enabling penetration of remote areas.  If there is a match, the person is verified, gets a ballot paper and continues to vote (manually) in the normal way! The person’s details are then digitally marked as having voted and cannot be used for repeat voting (no need for ink). This is NOT electronic or biometric voting, but manual voting as we are used to!  

At first ZEC had indicated that biometric verification would not be done (thus just creating Biometric Voter Register – or simply an electoral register which contains a person’s face and fingerprints which would not improve the voting process itself but provide a clean and credible voters’ roll). However Vice President Mnangagwa in response to a question by MP Chamisa in parliament regarding use of biometrics on voting day had this to say; “Hon. Chamisa has forgotten that we agreed that we need BVR.  We never said it would not go full throttle.  We agreed that the biometric system would be used in coming up with a Voters Roll up until the actual voting.” So, it is expected that biometric verification will be used on voting day.

The other dominant theme of the publications attacking the BVR process was the ‘susceptibility to hacking and cyber-attacks’. A biometric voter register, as mentioned before, is no different from any electoral register (as prescribed by the Electoral Act) or any other database. Therefore it’s susceptibility to hacking and cyber-attacks should just be at the same level; but this is not even the case as these biometric databases are more robust and designed to protect the sensitive personal information they contain. 

The issue of data privacy features dominantly in the development of biometric processes. Consequently, the BVR process has inbuilt protection included in the software packages (for example, template protection) which makes it more robust than the current electronic register which has been used in the previous elections. It is difficult to hack, and even if the data is somehow stolen it would be in an unusable format for the perpetrator. It is accepted that the outcry might have been based on the misconception that “electronic voting” and automatic tallying of votes would be carried out; an assumption which is very wrong.

Another debate and negative concept being cast about the BVR process is its perceived cost, but before delving into the intricacies of financial cost, it is important to look at why Zimbabwe has embarked on this path. It is not by accident that ZEC has embarked on the Biometrics project. The history of disputed elections and unclean/suspicious voter registers is a known political burden to Zimbabwe. This has damaged the credibility of Zimbabwe elections leading to violence, leading to loss of lives, people being displaced and some fleeing the country. The cost in terms of human lives and the country’s economy has been monumental and cannot be quantified. It is clear that the current scenario cannot be sustained, and an improvement/change in the electoral process is crucial. Reverting to the use of national IDs or licences will create the same cycle of rigging accusations and discrediting of the electoral process – a vicious circle which needs to be avoided.

In 2012, ZEC said they would need about US$20 million to spruce up the widely-condemned roll after which constituency boundaries would be drawn up for general elections(The Herald 21/12/12). It is on record that a proposal for biometrics registration was made at that time, detailing that the exercise could be carried out within 3 months, costing USD20 Million; the same figure that ZEC had said it needed to clean up the voters’ roll!

The current proposal for BVR is based on a budget of US$29 million; to produce a NEW clean and credible voters’ roll – surely not an expensive exercise especially if put into context of what it will achieve. The cost of acquiring the equipment needed is no more than US$15 million. Therefore the “unaffordability” claim is unfounded. Furthermore UNDP had offered to fund the BVR procurement process through their structures to ensure transparency, a proposal which has now been rejected for ‘sovereignty’ reasons. However the government has now made US$17 million available to fund the process. In addition, this process is sustainable, and will be much cheaper in the next elections (no/low procurement cost) in addition to the bonus of sustainable dispute free elections.

Having said all that, BVR in itself does not guarantee successful, fair or credible elections. The author does not propose the use of biometrics as a “silver bullet” capable overcoming all obstacles Zimbabwe faces in ensuring a level playing field in which all eligible voices have their say in the political future of the country.  Its effectiveness can only be recognised if applied in tandem with the political-will and sincerity of authorities in charge, who are tasked with guaranteeing fairness and ensuring inclusion of all citizens.  Biometric technology cannot solve problems rooted in issues such as mistrust among stakeholders or lack of political freedoms. Elections, at the end of the day, are a political process.

In spite of all the challenges, the introduction of biometrics in the compilation of voter registers should improve the accuracy of the voter registers and provide the foundation for clean and violence free elections. Ghana has used biometric registration and verification in three consecutive elections (the latest occasion being in 2016) proving that the process can be reliable and sustainable.  It is therefore urged that ZEC and all stakeholders embrace biometrics technology to ensure integrity, inclusiveness, accuracy, transparency and accessibility in the coming elections. The media should act responsibly and report facts accurately, and ZEC should take a pro-active role in explaining the BVR process and educating the public.

Dr. Samuel Chindaro is an Electronics Engineer, biometrics expert and researcher, trained at NUST in Zimbabwe, the University of Birmingham and the University of Kent in the UK. At Kent, he was the leader of a specialist research group on biometrics technology. He can be contacted at S.Chindaro@gmail.com