This piece is the second of a series exploring the current state of electoral technology in Nigeria. It begins with a brief summary of how technology has been deployed in several strands of Nigeria’s election management, drawing cross-country comparisons. It then proceeds to discuss the technologies that will be used in the 2023 elections, highlighting specific challenges associated with each and proffering possible solutions for INEC.
Nigeria’s general elections have included the use of technology since 2015. The Independent National Electoral Commission (INEC) introduced the earliest innovations in electoral technology when it initiated the use of the Permanent Voters’ Card (PVC), which encloses the biometric information of voters using an embedded microchip on the card. Smart Card Readers (SCR) were then used to accredit voters by scanning the microchip on the PVC.
Combined with the PVC, the SCR was a reliable deterrent to voter fraud by identifying voters’ identity (fingerprint and display picture) before voting. A replacement for the temporary voters’ card used in the 2011 elections, PVCs were first issued in 2011 but deployed fully in 2015.
That same year, INEC, for the first time, invested in making election results publicly available on its website, although this was only after all votes had been tallied and final results announced.
2015 was indeed the watershed in the history of electoral technology in Nigeria.
How Has Nigeria fared so Far?
While the number of technological innovations introduced since 2015 may seem few, each technology has brought several changes to election management.
Table 1 below summarises the use of technology across 5 stages of election administration in Nigeria so far: Voter Registration and Identification, E-voting, Results Processing, Use of Open-Source Software in Election Administration, and Online Publication of Election Data. These results are drawn from International IDEA’s ICT in Elections Database, which provides data on the use of electoral technology in more than 170 countries of the world. The table shows that Nigeria has substantially implemented the use of technology across all stages of its election management – except for E-voting which has not been implemented at all, and Online Publication of Data which is still in its formative stage,
Table 1: How does INEC use ICT in election administration?
Note: The ICT in Elections Database provides up-to-date data for each country. Thus, responses on certain aspects of election tech in Nigeria may reflect their adoption for sub-national and off-cycle elections, even when they are yet to be used for national elections.
Nigeria is one of many African democracies that show impressive levels of progress in the use of electoral technology when compared with other regions of the world where democratic elections are held. For instance, out of the 177 countries listed on the International IDEA database, only 58 have digital voter registers, half of which are African countries.
A quick visual representation of selected use of ICT in African elections is presented below. 4 metrics are reviewed:
Figure 1: Which African countries use biometric data for voter identification?
Of the 52 countries where biometric data is used for voter identification at polling units, 20 are from Africa, and Nigeria is one of them.
Figure 2: Is technology used for the voter identification process at polling units (electronic polling book)?
Nigeria is one of the 8 African countries and 47 globally where technology is used for voter identification in elections, whether this is via an offline method or connected to a central voter database.
Figure 3: Were results processed using an electronic tabulation system?
Nigeria is one of the 27 African countries and 105 globally where election results are processed using an electronic tabulation method. This began with the 2015 elections, although votes seem to have been electronically tabulated at the state and national level collation centres only.
Figure 4: Which African countries publish election results online?
The majority of democratic countries globally (114) publish their election results online, but these are mostly in non-machine readable formats – only 57 publish results in formats that are machine-readable. In Africa, this number is even smaller, as only 6 of the 42 countries that publish results do so in machine-readable formats. Nigeria has a lot more work to do in this regard, as its online election results are currently published in the original hand-filled formats that are not machine-readable.
Note: Results are for the last national elections in each country.
Source (Figures 1-4): International IDEA ICT in Elections Database
How will technology be used in 2023?
The core of electoral technologies that have been introduced is fostering a more efficient voter identification process, making voting and results collation faster and more reliable, preventing electoral fraud, and overall, improving the credibility of elections. Yet, in this regard, INEC’s technology adoption so far has had a few drawbacks. For example, even though results were made public on the INEC website during the 2015 and 2019 elections, these were not live updates and the public could not access them online until final results were collated— which in most cases, never happened on time. Thus there is no difference between following the results update on INEC’s website and simply waiting for the final results announcement. Also, the process of using SCR to identify voters was done offline and unconnected to a central voters’ register, making fraud deterrence less effective.
What will be different in 2023?
For the 2023 elections, INEC has more ambitious goals to solve these challenges. They are introducing 3 key technological innovations: the INEC Voter Enrolment Device (IVED); the Biometric Voters Accreditation System (BVAS), and the INEC Election Result Viewing (IREV) Portal. As well as improving its user friendliness, the commission has also increased the range of information publicly available on its website. For example, information about candidates and political parties, updates on PVC registration and collection, information about polling units, electoral guidelines, and many more are now available on its website.
These technological innovations are unique in that they are clearly backed by the recently passed 2022 Electoral Act, which authorizes INEC to electronically accredit voters, as well as manage and transmit results. Elsewhere, Dataphyte has written about what these technologies are, and how they will work on election day.
Figure 5 below presents a quick summary of how they work in practice.
Figure 5
By shifting the management of specific election procedures away from people (who could be partisan) to technology, INEC was able to improve election procedures in off-cycle elections where both the IREV and BVAS have been trialed. In the recent Osun and Ekiti elections, both the IREV and BVAS were described as the 3rd force, enhancing consistency between INEC’s final results and the aggregate from polling units, as confirmed by civil society observation groups.
However, there are a couple of general concerns with both technologies. Perhaps most concerning is that they have been received with skepticism— especially from the political class, who might consider technology use in elections highly volatile. For example, some key chieftains of the APC have recently expressed concern about IREV, stressing that they do not “see how transmission of results would work,” given specific regional disparity in the infrastructure required for such technology to function well, such as internet connectivity and electricity.
The second challenge is that, while INEC has recorded massive successes in using these technologies in 4 sub-national elections so far, there are concerns about scale and replicability — obviously, using technology across 119, 771 polling units would present greater challenges compared to just 700 polling units.
Thirdly, technologies could potentially worsen election administration by introducing new layers of bureaucracy. New technologies would involve new lines of communication, new management structure, and new manpower that could potentially create bottlenecks in the interaction between human agents and resources within each election management value chain.
More importantly, there are specific challenges with each of the BVAS and IREV, which may have a substantial impact on the elections if not well managed by INEC. Some of these are discussed below
Addressing challenges with the BVAS
In theory, using biometric data to confirm voters’ identity at polling units should work without glitches and ensure voter fraud is minimised. However, this is not always the case, and INEC has had to also use Smart Card Readers in the2021 Anambra Guber polls, where it had mainly planned to use BVAS. As such, the possibility of INEC resorting to the SCR as a backup for BVAS devices in certain polling units cannot be entirely ruled out, despite the Commission’s proclamations to the contrary. However, as Dataphyte has shown, INEC’s planned BVAS supply is insufficient. The commission is deploying 200,000 BVAS devices across 176, 846 polling units; an average of 1 device per polling unit. This figure fails to consider factors such as a device failing to work properly in a polling unit, or some polling units having way more voters than others.
There are also questions on quality assurance. If BVAS devices fail, what back-up system could guarantee the same level of authenticity as BVAS? A probable alternative as mentioned above is INEC staff reverting to SCR (or manual forms) to verify voters’ PVC, but this does not ensure consistency in voters’ data, seeing that the SCR only accredits PVCs, whereas the BVAS performs more sophisticated functions, such as capturing images of results sheet and transmitting it in real time onto the IREV? Ultimately, where contentions on election results arise, INEC must be able to match the results sheet in each polling unit with the register on the BVAS. Using SCR in some polling units and not in others presents significant legal and technical challenges to this outcome.
Furthermore, much of the challenges with the BVAS in past elections is attributed to poor staff training on how to use the device, as noted by INEC’s Chief Press Secretary in a recent interview with Dataphyte. For instance, during the last FCT Area Council elections, many INEC staff reportedly lacked adequate knowledge of how the BVAS works and had to wait for technicians assigned to their polling unit to resolve this.
Ahead of the 2023 elections, INEC has disclosed that it is recruiting and undertaking extensive training for ad-hoc staff members who would handle the BVAS in each polling unit, in addition to recruiting 8,809 technical assistants who would manage the BVAS processes in each registration area (ward).
While comprehensive staff training is commendable, the planned number of technical assistants implies that each official would be supervising an average of 20 polling stations on election day. This, again, requires substantial logistical planning, such as transport and security, to cover allocated polling units within limited voting time. To boost confidence in its staff preparedness, INEC should consider expanding its number of technical assistants. Alternatively, experts have suggested enlisting IT professionals as volunteers to monitor and observe polling units. This might help offset the manpower shortage, although it would require significant ethical training to ensure the neutrality of individuals at the polls.
Finally, and perhaps most importantly, INEC needs a clear framework to ensure data privacy and security. Interestingly, much of the public debate on the use of BVAS has been centered around the security of the BVAS and IREV joint system, to which INEC has responded that it is well fortified against cyber attacks, evidenced in its resistance to several hacking attempts during the Osun and Ekiti elections. However, a key conversation around individual data protection has been missing in INEC’s electoral preparations.
For instance, the 2022 Electoral Act, which essentially empowers INEC’s use of digital technology in the elections contains no mention of the protection of voters’ data. Yet, thousands of INEC officials and ad-hoc staff would be interfacing with voters’ sensitive data across different stages of the election— from voter registration to accreditation and verification and management of digital registers. As the elections draw near, INEC must comprehensively apply the Nigeria Data Protection Regulation (NDPR) to every aspect of its digital innovation. It must also educate the public on how their data would be used and managed.
Addressing challenges with IREV
The key purpose of IREV is enabling a transparent and timely transmission of election results to the public. Previously, election results were published on INEC’s website only after final election results were established, which took an average of 3-4 days in both the 2015 and 2019 presidential elections (see figure 6 below for comparison with other African countries).
Figure 6: It took INEC 4 days to establish final results and 2 more days to certify/announce the final results during the 2019 presidential elections.
Note: i) Data represents the number of days it took for EMB to establish final results in the last presidential elections only, legislative elections are excluded. ii) 0 (Eswatini)= results were provided the same day as voting day. iii) Sierra Leone held 2 rounds of national elections in 2018; data above is for the first round.
While IREV would not necessarily speed up the process of certifying the final results, it would enable citizens and civil society groups to independently gather and analyse election results in real-time. It would also specifically aid the Parallel Voters Tabulation (PVT) process done by many election observation groups. Barring any unforeseen changes, the public would be able to veritably confirm election winners in their units, constituencies, and states, without waiting for INEC’s final announcement.
There is however a worrisome challenge with the IREV portal as it is now; where it has been trialed, many of the results sheets uploaded are not legible, either due to poor image quality or poor handwriting. Investing in BVAS devices with clear image quality and ensuring that staff clearly document results and take quality pictures are the most basic expectations on IREV for INEC.
Secondly, although the intention to upload results sheets in their original manual form to preserve results integrity is well understood, the current model is not machine readable and requires individuals and civil society organisations to either manually compute results from thousands of results sheets or use additional technology to extract and aggregate this. Even with an extra technology, individuals would have to walk through thousands of clicks to get the final result for each State; something very time consuming when analysing the Presidential election.
If the public cannot directly get aggregated information on election results either at Local Government or State level, the current design of IREV makes it less useful in the bigger picture. Afterall, most voters can already access Form EC60E (copy of election results) in their respective polling units; the real essence of a real-time IREV portal is to enable real-time aggregation. Realistically, individual citizens are less likely to go through the rigour of scrutinising 13, 325 results in a state such as Lagos, for instance.
While INEC might not need to convert its paper sheets into machine-readable formats, it could redesign its portal to have landing pages to harvest results sheets at both local government and state levels.
Also worth mentioning is ensuring a 100 percent upload rate for the results sheet. INEC was able to achieve this in both Osun and Ekiti, but other elections were aggregated at a 99.13% rate. A 0.8 shortfall might seem negligible, but not for a process as simple as uploading a results sheet – especially when such a shortfall could become substantial where legal contentions arise about the accuracy of results in certain polling units.
The final point on IREV is ensuring routine testing and maintenance of all IT systems relevant to the portal— both software and the devices that will be used for the elections. Not unexpectedly, the IREV portal would witness more traffic and cyber threats during the Presidential elections than it did in past elections where it has been used. Notably, on 18th November, INEC’s website was unusually down for approximately an hour; displaying the usual ‘404 error’.
The commission may have been performing a maintenance check, but such downtime during the 2023 elections, where the public is not able to assess its website for an hour during election results collation, would be more grave.
Overall, while the degree of confidence with which INEC is set to implement these technological changes in the next elections are commendable, the commission must minimise risks by having credible alternatives where technology fails. In addition, prioritising staff training and support, investing in the security of its systems, ensuring data privacy and implementing user-friendly interfaces on its website are very crucial preparatory tasks for the commission.