The Editor Zambia
There is something deeply troubling about the manner in which some foreign journalists and political analysts descend on African countries to explain their elections to Africans, often with an air of intellectual superiority that borders on condescension.
They arrive with credentials, theories, and preconceived frameworks, spend limited time on the ground, consult familiar circles, and then depart to pronounce sweeping judgments about political realities they have barely experienced. 
This is the problem of helicopter journalism and, in the case of election analysis, helicopter scholarship is an approach in which outsiders fly in, observe from above, identify what they regard as contradictions, interpret them through theories developed elsewhere and then present their conclusions as authoritative explanations of a country’s political experience.
Zambia deserves better.
Professor Nic Cheeseman and Dr. Nicole Beardsworth are perfectly entitled to analyse Zambia’s 2026 general election.
They are also entitled to question figures released by the Electoral Commission of Zambia (ECZ). Academic freedom permits them to advance hypotheses and challenge official narratives.
But Zambians who have swallowed their analysis of the latest general elections should know that academic credentials do not transform a hypothesis into established fact, nor does expertise in African elections confer a licence to pronounce a democratic verdict on Zambia without sufficiently compelling evidence.
Cheeseman is a professor of democracy at the University of Birmingham and co-author of How to Rig an Election, while Beardsworth is a political studies lecturer at the University of the Witwatersrand.
Their credentials are not in dispute, but what deserves scrutiny is the leap from statistical anomalies to the suggestion that Zambia’s presidential election may have been stolen and therefore requires a rerun.
Their analysis, published under the dramatic suggestion that Zambia’s election “may need to be rerun,” rests substantially on differences between presidential and parliamentary turnout in selected constituencies.
The two pedestrian ‘election analysts’ argue that some discrepancies are unusual and therefore require explanation, which is problematic because an anomaly is allowed to become an accusation.
The authors acknowledge an important limitation in their own analysis. Their calculations do not establish who allegedly manipulated the figures, when the manipulation supposedly occurred, how it was carried out, or whether it changed the final presidential outcome.
That admission should be at the centre of the debate. If presidential votes exceed parliamentary votes at a polling station, there can be several explanations. A voter may vote for a presidential candidate but decline to vote for a parliamentary candidate.
A parliamentary ballot may be rejected or spoiled. Candidates can have different levels of popularity and mobilisation. Local political dynamics can produce different turnout patterns. Administrative errors can also occur without constituting an organised conspiracy to steal a presidential election.
The proper response is, therefore, to examine the original polling-station results, ballot reconciliation records, rejected ballots, statements of results, transmission records, and other relevant documentation.
It is not to declare the election stolen before that examination has taken place.
The authors themselves describe their calculations as “unverified” and suggest that the suspicious differences could potentially account for hundreds of thousands of votes. But an estimate of potentially suspicious votes is not equivalent to a verified number of fraudulent votes.
That distinction is fundamental.
A particularly serious weakness is the tendency to interpret figures through a predetermined narrative.
Areas in which President Hakainde Hichilema’s UPND performed strongly are examined through the lens of possible manipulation, while insufficient attention is given to the possibility that political mobilisation, voter preferences, candidate strength and local electoral dynamics may legitimately produce different voting patterns.
If the claim is electoral fraud, the analysis must demonstrate the mechanism.
Who allegedly added the votes? At which polling stations? On which forms? Were the additional ballots physically available? Were ballot-accounting procedures violated? Were polling agents present? Did they sign the results? Were results altered between polling stations and constituency centres? Can the alleged excess votes actually be traced through the electoral chain?
Those are the questions that convert suspicion into evidence.
There is also evidence in the analysis itself that complicates the narrative of a systematically manufactured election. The authors acknowledge that election day was largely peaceful. They cite observations indicating that voting was generally orderly and that independent monitors found substantial correspondence between presidential results announced by the ECZ and results independently recorded by observers.
These observations do not establish that every aspect of the election was flawless. No serious observer should make such a claim. Elections are administered by human beings, and mistakes, disputes, and isolated irregularities inevitably occur.
But isolated irregularities are not automatically evidence of a nationwide conspiracy.
The size of President Hichilema’s victory also matters. The figures cited in the analysis place him at approximately 61 percent against Brian Mundubile’s approximately 38 percent, producing a margin of roughly 1.1 million votes.
Even accepting the authors’ unverified estimate of potentially suspicious votes, their analysis does not demonstrate that Mundubile would have won a clean election.
That is a crucial weakness.
One cannot simultaneously concede that the available evidence does not establish that the opposition would have won and then build a public narrative suggesting that the incumbent’s victory was illegitimate.
The intellectually honest conclusion would be that certain discrepancies deserve further investigation.
That is very different from saying that the president stole the election.
Zambia has institutions designed to deal with electoral disputes. Political parties have agents. Domestic observers have collected information. The ECZ has records. Courts exist to hear petitions and determine whether alleged irregularities were sufficiently serious to affect an electoral outcome.
Those institutions should be allowed to do their work.
Foreign academics can contribute to that process, but they should not become substitutes for it.
There is a broader problem here. Western and international coverage of African politics has too often treated African societies as laboratories for theories developed elsewhere. Local voters are reduced to statistics.
Political choices are interpreted through familiar narratives of ethnicity, authoritarianism, clientelism and manipulation. Local agency becomes secondary to the analyst’s theoretical framework.
That approach is not only patronising. It can also be intellectually lazy.
Zambians are capable of making political choices for reasons that do not fit neatly into an academic model because they can support a president because they approve of his policies. They can reject an opposition candidate because they distrust his political history. They can vote differently in presidential and parliamentary contests. They can change their minds. They can make complicated political calculations.
Not every electoral pattern that looks unusual from a spreadsheet is evidence of manipulation.
The irony is that analysts who rightly demand transparency from African governments should apply the same rigorous standards to their own conclusions.
If there was electoral fraud, expose it.
If officials falsified results, identify them.
If ballots were unlawfully introduced, show where and how.
If polling-station results were altered, produce the forms.
If the alleged irregularities changed the outcome, demonstrate the mathematical chain from the irregularity to the final result.
That is the standard Zambia should demand.
Cheeseman and Beardsworth may be correct that certain questions require answers. But their authority cannot be a substitute for evidence. Their academic reputations cannot turn statistical suspicion into established electoral fraud.
The danger of presenting preliminary analysis as though it were a definitive verdict is that it can undermine public confidence in democratic institutions without establishing that those institutions actually manipulated the election.
Zambia’s democracy certainly needs scrutiny. It needs independent journalism, strong domestic observers, credible electoral institutions, and fearless academics. But it also needs intellectual humility.
The country does not need foreign academics flying in, hovering above its political landscape, and then telling Zambians what their own votes supposedly mean without sufficiently engaging the complexity of the society beneath them.
The central question is, therefore, simple: where is the evidence?
Until the calculations are independently replicated, the underlying polling-station documents examined, and the alleged irregularities established through credible evidence, the Cheeseman-Beardsworth thesis remains what it is: a preliminary and unverified hypothesis.
That means it does not justify overturning an election.
The difference between investigating an election and declaring it stolen is the difference between scholarship and advocacy.
