When Machines Glitch: Electoral Technology and the Dynamics of Vote Trading

By Micheal Eze, Founder and Principal Researcher, Critical Frontiers, in collaboration with Risalat Consultants International

Electoral irregularities remain an abiding challenge in Nigeria's fourth democratic dispensation. During the same period, the integrity of every election has been heavily contested across political divides. These challenges range from pre- and post-voting vices to actual manoeuvres at polling units, with consequential impacts on election outcomes. One particular ill I wish to engage here is the thorny problem of vote buying. This problem has a long history in the Nigerian electoral process. However, the advent of the BVAS machine in 2021 changed the optics and practice of vote trading during elections. Prior to its introduction, vote-buying merchants and sellers only needed to agree on simple grounds, say, the offering of rewards for votes before or after the ballot had been cast, provided the intended seller is eligible to vote. But the BVAS machine altered this arrangement. One of the machine's functions is to maintain the integrity of the voter register by ensuring that only voters whose details are captured on the INEC voters’ register participate. Consequently, it is not enough to appear on the voters’ register pasted on the wall at polling units; the BVAS must authenticate voters’ biometrics and clear them to participate – this process is called voter accreditation and is usually completed before handing out the ballot.

For context, the introduction of BVAS was a necessary intervention aimed at eradicating any undue alteration of the voters' register which may inflate the list and allow unqualified persons to vote – as has been seen in some parts of the country where underage people brazenly voted because their names somehow appeared in the register. Once the machine fails to accredit any voter, in principle, such a person cannot vote. Even when an unaccredited person managed to vote, such a vote may not be valid, given that the number of votes recorded by the BVAS must tally with the counted ballots in the ballot box. Any discrepancy will raise eyebrows among stakeholders. In my experience, BVAS has moderated the dynamics of vote buying, ensuring merchants don't pay until the voter successfully casts her vote; otherwise vote buying merchants might be throwing money to the wind.

But the exchange gets more complicated: a voter can be duly registered, with a valid PVC and her details appearing on INEC register, yet the BVAS machine fails to accredit the voter. This experience puzzled me when I served as a polling officer during the senatorial bye-election in Edo State in 2025. I had no explanation for the technical glitch that disenfranchised eligible voters, and my complaint to the technical officer provided by INEC during the exercise didn't ameliorate the ordeal, as I was simply asked to wait it out. It got even more puzzling in specific instances where the machine repeatedly rejected the same person, while recognising everyone else. This was concerning because the overall functionality of the BVAS machine was not compromised. I kept wondering how a technical glitch of this kind might reshape vote buying practices, as both the person who intended to sell his vote and the buyer who needed to boost his ballots were visibly frustrated. Like vote buying, technical glitches have been a recurring feature in elections since the introduction of card readers and subsequently voting technologies. A case in point is the malfunctioning of the IReV portal, which further undermined the integrity of the 2023 presidential election. Although the IReV glitch did not disenfranchise the electorate, it nonetheless created doubts about the transparency of results uploads. INEC has repeatedly argued that the few cases of technical glitches don't bear any defining impact on the overall outcome of elections. Yet, the minutiae of these glitches, however minor, cannot warrant sweeping the whole technical issue under the carpet. My hunch is that technical glitches will resurface during the voting exercise in 2027. This raises the question of how technical malfunctions can define electoral participation, and the regular vice – vote buying. The BVAS, meant to keep the register clean, had the unintended consequence of disenfranchising eligible voters.

To prevent the risk of buying invalid voters, the bargain between the buyer and seller has shifted towards a post-ballot settlement where the transaction is settled after a successful ballot. The practice involves the voter subtly showing his ballot to the merchant to confirm his party/candidate got the vote, a practice that further compromises the secret ballot system. This has invariably tilted the bargaining power in favour of the merchant who already exerts considerable weight on the scale of the transaction. The intending voter is now compelled to part ways with her only bargaining chip, trusting the merchant will hold his end of the bargain. From experience, the promise of post-ballot settlement is hardly satisfactory for the selling voters. Merchants often exploit their leverage by paying voters less than the agreed bargain or refusing to pay at all (especially when the numbers are not in their favour). The post-ballot settlement is particularly dangerous for vote sellers because payment after voting is uncertain. The logic of vote buying is primarily to induce undecided or uncommitted voters – especially the poor – who are easily swayed with material or monetary inducement. This means party stalwarts are not always the primary targets of vote buying, as they already commit their allegiance. The uncommitted voters who are willing to trade are now exposed to the risk of not getting any rewards, with no reliable basis to enforce settlement after voting. Conversely, the merchant may not be willing to make a pre-ballot purchase. Some instances of post-ballot settlement gone wrong have become public disputes, with disgruntled voters jointly berating merchants for failing to settle them. Vote buying has become riskier because the buyer is not willing to buy votes that may be invalid due to a technical glitch or the voters changing their minds after receiving pay. Neither is the seller willing to risk voting without settlement. Ironically, the last-minute rush may influence merchants to intensify vote buying. Towards the closing periods of the voting exercise, where competing merchants are desperate to boost their ballot, vote buying is driven into a late frenzy. Voters suddenly regain bargaining power, and merchants can easily let their guard down and even offer pre-ballot settlement to outbid their competitors. Vote buying is illegal and criminalised in Nigeria's Electoral Act 2026 under Section 22, so it is a dark market with shady transactions. But electoral technologies like BVAS have reconfigured vote-buying transactions from a simple buying and selling logic to a high-risk commitment for both parties without eliminating the trade.

This raises the concern of how the same piece of technology meant to enhance electoral integrity can, invariably, compromise it.
It should be emphasised that INEC remains the ultimate electoral umpire who decides the degree and fashion in which its technological apparatus is deployed during elections. Yet, the improvements heralded by these technologies are too precious to warrant their reversal. Moreover, the use of BVAS is mandated in the Electoral Act 2026, s.47. Electoral integrity is a dynamic and fragile concept. Regardless of the effort made to enhance electoral exercise, a little speck of irregularity can severely taint the integrity of the election.

The recent controversy surrounding INEC regarding alleged manipulation of voters’ register ahead of the 2027 polls reinforces the need for thorough voter verification before voting. My proposal is that the BVAS machine remains the highest reference for voter accreditation. BVAS accreditation uses both biometric and camera systems to verify voters, which increases the chances of accreditation, including for the elderly and artisans whose faces and fingerprints have been altered due to ageing and the impacts of physical work. INEC should work assiduously to reduce to the barest minimum, if not eliminate, the issues of technological glitches through pre-election testing and system upgrades. Specifically, to curtail the incidence of BVAS failing to accredit eligible voters with valid PVCs, INEC should consider documenting and announcing the rate of accreditation failures and the causes. This rate of accreditation failure among valid voters can help inform preparations for subsequent elections.