Biometric voter registration: Lessons from Ugandan polls
By Rindai Chipfunde Vava
On February 18, I went to Uganda to observe the presidential and parliamentary elections there.
This was the first election to be held using the biometric voter registration (BVR) system in
Uganda.
The first lesson I learnt was that BVR and the use of ICTs is not a silver bullet to all electoral
challenges.
While the BVR machines worked well, they were a number of shortcomings, in particular the
electoral environment whereby the opposition leaders were intimidated and arrested, mistrust of
the Election Management Body (EMB) by stakeholders, late delivery of election materials to
some polling centres on election day, among others.
Biometric identification is where a voter offers a fingerprint, for example, and the device will
look up all the biometric data it contains to see if there is a match (a 1:N match, where N is the
number of records in the database). The question being asked is “who is this person”. Biometric
verification, on the other hand, asks the question “is this person who they say they are?” So, for
example, the voter’s ID number is entered, then a fingerprint and the system matches the offered
fingerprint against just one, a 1:1 match.
Eight candidates were contesting for the presidency, including the incumbent Yoweri Museveni.
One of them was a female. The parliamentary elections were for 413 seats in the National
Parliament; out of this 238 members are elected by universal suffrage, the remainder represent
special groups like women, youth, persons living with disabilities, army, among others. They
were 29 political parties contesting in the 2016 elections and several independent candidates.
The newly introduced system uses human body characteristics to confirm a person’s identity. It
was aimed to improve inclusiveness and accuracy of the voter register. The Ugandan electoral
body compiled a new voter register for the 2016 general elections by extracting data on voting
age citizens from the Ministry of Internal Affairs under the National Identification Register
(NIR).This was a National Security Information System (national ID) project.
There were complaints that these new developments were not communicated on time to key
electoral stakeholders such as political parties, civil society organisations (CSOs) and media. As
a result there was no consensus if it was the best way to go. However the final voters’ roll
contained 15,277 million voters, extracted from the NIR database of about 16,46 million citizens
aged 16 years and above at the time of registration in 2014.
Some of the observers hailed the extraction of the voter register from the NIR as a cost-effective
practical model.