During a June 2018 review of the Electoral Act Reform Bill, the Nigerian House of Representatives rejected the idea of an electronic voting system but instead made room for the use of electronic card readers in the 2019 general elections. However, with increased calls to digitize the electoral process to ensure more credibility, a working prototype employed in the University of Ilorin student union elections gives an insight into the feasibility of an electronic voting system for Nigeria.
I first met Abayomi Adeola Kareem in 2014 on the floor of the customer service department of Nigeria’s e-commerce giant, Jumia Nigeria. We would banter about service delivery in Nigeria and the ‘Nigerian Customer’ syndrome, a term we coined for the peculiarity of dealing with a client base who were just learning the ropes and adjusting to the art of purchasing items using a mobile device or personal computer.
Abayomi studied Computer Science at the University of Ilorin. Together with six other 20 – 25-year-olds at the time, he had founded a now-defunct techpreneur group in one of the hostels in the sprawling university compound, aptly coined De07 indicating their class set. Mustapha Omotosho, Ahmed Zubair, Jinadu Razak, Ibrahim Akinsanya, Kamaru Onimisi Abubakar, Bilal Kareem Olakunle and Abayomi were at that first informal meeting, incorporating three other members of the group along the way. Their most successful invention would turn out to be an electronic voting system which the school employs in its student union elections to date.
Ibrahim Akinsanya started developing the e-voting software long before a need arose for it.
“I think it started from his good use of Joomla, a CMS (Content Management Software) and then it evolved to building software from scratch and then the e-voting system was created,” Abayomi explained.
A chance finally came to put the voting software to test when the Faulty of Engineering was prepared to go to the polls to elect new executives in 2010.
“The testing of the application idea came up when the Faculty of Engineering was preparing for their election. As a member of the SUG Student Representative Committee (SRC), I sold the idea to a fellow SRC member representing the Faculty of Engineering constituency which he bought into and decided to push it in his faculty,” Akinsanya said. “They accepted it and that was how the e-voting system was successfully executed.”
It would go on to be successfully employed in two other faculty elections, Agriculture and Science, that same year. But they saw a larger need for it in the overall student body and began to speak to the school’s administration to adopt the system in the Student Union Government elections.
Nothing prepares you for the array of activities leading up to a student union elections in any Nigerian university. From full-blown posters slammed on every and any bare block of cement to door-to-door campaigns by a full-fledged campaign team, the makings of a typical Nigerian politician begins in the four walls of a Nigerian university. And the process is not a pleasant one to watch.
The reality at the University of Ilorin during this period was no different. Prior to the adoption of the e-voting system, the SUG elections were marred by issues not alien to elections across many levels in the country. Late collation and determination of election results, multiple votes casting, election results manipulations, and void votes. But even with the group’s alternative system available, the SUG and school administration could not adopt it.
“The SUG electoral law does not recognize electronic voting but ballot papers and boxes,” Omotosho said.
Burdened by bureaucracies, it wasn’t until their graduation that the first electronically run student union elections held in the school campus.
“The law was subsequently amended and e-voting adopted but we had graduated at this point,” he noted.
The e-voting system the group came up with addressed a lot of the irregularities in the student union elections. Students could only cast one vote each with their unique voter identification details, but results could be and were collated in real time resulting in same-day announcements of results.
“The e-voting system is a role-based system developed using a custom built PHP framework,” Mustapha explained.
Voter, Electoral Admin, and Contestant profiles were developed under the framework and students were then assigned unique identification based on which categories they belonged to and access granted accordingly.
When a student logs in as a voter at the polling unit, they are granted access to a list of available positions being contested and candidates for each of these posts. The voter is then able to select his/her candidate of choice after which the votes are then fed into a central system.
“An Electoral Admin Officer who logs unto the platform will be able to perform actions not limited to the following; Create Electoral Post, Create Contestant for each post, Register Voters, View Election’s Results and Reports,” Mustapha said.
The profiles of the contestants were added for optimization. Contestants were also unable to log in or perform any functions on the platform.
“In our first trial, we had five computers,” Abayomi said. “One was the main computer which housed the software and records of the electoral process while the other four were connected wirelessly to the main one, for use by the voters.”
However, the elections didn’t go without a hitch. “Along the line, we found out that the computers were not enough compared to the crowd of voters so we borrowed three extra computers to make a total of seven. And we had one casualty (a voter knocked off one of the computers). We also noticed that at a point due to the influx of votes, the main computer developed a constraint (it became slow) but it was fixed by Al-Mustapha & Ahmed and the end result was superb.”
The group believes this success can be duplicated in the country’s elections with some adjustments. But when compared to a population of about 200 million with more than a quarter eligible to vote in the next elections, the odds are higher.
The fact that it took a full academic year to convince the school to amend the constitution and adopt the software is indicative of the first hurdle a system like theirs will face if in consideration for a larger scale process. Asides from the e-voting software, the group had also developed “a more secure, robust and flexible CBT (computer-based test) application whose outcome was far better than the school’s CBT software.” Akinsanya explained that the administration did not believe the software had come from them and so did not consider swapping it for the already existing software even though it held better results.
In the event that the country seeks an electronic voting alternative in the future, it will be far more cost effective to use a locally developed software as opposed to purchasing from international sources especially at astronomical prices. Nevertheless, finding a credible source for the development of this technology remains a tiny spec in the number of factors that need to come together to create a functional system. What would be more challenging using this prototype will be putting together a system that will not be overwhelmed by the population of voters as well as ensuring that votes cast are transmitted securely to a central processing hub, a very vital element at a time where faith in the electoral process is at an all-time low.
One of the important successes of this system which forms the basis on which other successes are built is its employment of existing student data in the creation of the software database. By collating what can be referred to as a credible voter register, allocating unique voting identification and ensuring accurate capturing of votes, the system ensured a transparent voting process and fraud-free result collation.
For Seyi Ogunseye, a Lagos-based Technology Consultant, the duplication of the UNILORIN e-voting system is not feasible for the singular reason that available voting data in the country is still fraught with inconsistencies. However, he agrees that splitting up the employment of technology in the electoral processes into manageable bits might be the immediate solution that should be enforced ahead of the 2019 elections.
“We need to break the process down and address in pilot stages and see where we can automate and where we can’t. We must not take a one size fits all approach,” which is what the prototype offers.
INEC has said it will be employing technology in the collation of results as well as voter accreditation using data capturing devices that made a debut in the 2015 elections.
With a reported 41 percent failure rate during its testing phase in 2015 and countless malfunctions reported in many polling units on election day, the likelihood that these devices (as well as other technological innovations that may come in during the collation process) will function optimally is still debatable and very crucial for an election that promises to be a memorable one.