Did Tinubu Win Fair and Square? Data-Driven Study Says Yes, but Indicts Labour Party Strongholds for Significant Fraud Patterns
A comprehensive data-science analysis of Nigeria’s 2023 presidential election has concluded that President Bola Ahmed Tinubu won the contest legitimately, while simultaneously indicting multiple parties—most notably the Labour Party (LP)—for statistically significant electoral irregularities concentrated in their strongholds.
The findings form the core of a Master’s dissertation in Data Science at Pan-Atlantic University, which examined polling-unit level results from the election. The study analyzed data from 123,918 polling units across all 36 states and the Federal Capital Territory, using statistical forensics and machine-learning techniques commonly applied in international election audits.
Technology Promise Fell Short
Nigeria spent more than ₦300 billion on the 2023 general elections, promoting biometric voter accreditation and real-time electronic transmission of results as safeguards against manipulation. However, on election day, the Independent National Electoral Commission’s (INEC) Result Viewing Portal (IReV) suffered prolonged failures, undermining public confidence and fueling allegations of fraud.
While opposition parties challenged the outcome, the dissertation argues that claims of a centrally coordinated rigging effort favoring the ruling All Progressives Congress (APC) are not supported by the data.
Statistical Red Flags in the Numbers
Applying last-digit distribution tests and Benford’s Law—tools widely used to detect fabricated data—the research uncovered extreme deviations inconsistent with naturally occurring vote counts.
A chi-square statistic of 24,871.37 was recorded for last-digit analysis, far above accepted significance thresholds.
Digits “0” and “5” appeared far less frequently than expected, a known indicator of human number fabrication.
Deviations from Benford’s Law produced a probability of 3.87 × 10⁻¹⁹⁵, effectively ruling out chance.
“These patterns are classic signatures of human interference,” the study notes, adding that modern electoral manipulation tends to favor plausibility over crude inflation.
Machine Learning Identifies Anomalies
Using Random Forest and Isolation Forest algorithms, the analysis flagged 4,351 polling units (3.5%) as anomalous. While seemingly small, the author argues that geographic clustering magnifies the impact.
The highest anomaly rates were found in the South-East:
Anambra: 24.9%
Enugu: 16.7%
Imo: 10.9%
By contrast, Lagos recorded an anomaly rate of 2.3%, while Oyo stood at 0.3%.
Labour Party Strongholds Under Scrutiny
The most controversial finding centers on the Labour Party. Despite winning 29.1% of votes nationally, LP accounted for a disproportionate share of statistical red flags, including 2,328 instances of “perfect scores”—vote shares clustering at psychologically attractive round numbers such as 50% and 75%.
The study argues that LP’s overwhelming popularity in parts of the South-East may have provided cover for localized manipulation. “You can only rig where you are popular,” the dissertation notes, suggesting opportunity rather than ideology drove the irregularities.
Observer Reports Align With Algorithms
Independent election observers corroborated several findings. YIAGA Africa’s Parallel Vote Tabulation (PVT) identified major discrepancies in Rivers and Imo states, both of which ranked among the most anomalous in the computational analysis.
In Rivers State, INEC declared the APC winner with 44.2% of votes, while YIAGA’s PVT estimated APC support between 16.7% and 26.7%. The dissertation’s analysis placed APC’s share at 26.3%, reinforcing concerns of result inflation.
Tinubu’s Victory Stands, Study Says
Despite the irregularities, the research concludes that President Tinubu’s victory remains valid. It finds no evidence of overwhelming fraud favoring the APC nationwide and emphasizes that opposition vote-splitting—63% of votes went to parties other than APC—was the decisive factor.
“The APC’s capacity to manipulate outcomes is limited,” the study states. “Opposition fragmentation, not centralized rigging, explains the result.”
Recommendations for Reform
The dissertation calls for urgent reforms to restore electoral legitimacy, including:
An independent audit of the 2023 election data.
Legal mandates for real-time result transmission.
Unbundling INEC’s regulatory, operational, and prosecutorial roles.
Building in-house data-science capacity within INEC.
Visible prosecution of electoral offenders.
A Troubling Democratic Outlook
With voter turnout at just 26% and anomalies concentrated in politically vocal regions, the study warns of a deepening legitimacy crisis. It describes Nigeria’s 2023 election as “not catastrophically fraudulent, but not credibly free and fair.”
The author concludes that while data cannot decide Nigeria’s democratic future, it can illuminate uncomfortable truths. “Electoral fraud in 2023 was not the monopoly of the ruling party,” the study asserts. “It was democratized.”
As Nigeria looks toward 2027 amid opposition disarray and growing APC dominance, the findings raise urgent questions about trust, participation, and the future of competitive democracy in Africa’s largest nation.