There is a recurring tendency in African election analysis to treat an opposition victory as evidence that an election was credible, while an incumbent’s victory is often met with immediate suspicion. When the opposition wins, some analysts and commentators point to the result itself as confirmation that the system worked. When the incumbent wins, however, the same system can suddenly be viewed with doubt.
That is the wrong standard by which to judge an election.
On 13 August, Zambia held its eighth general election since the country’s return to multiparty politics in 1991. Incumbent Hakainde Hichilema won a second term with 60% of the vote, while main opposition candidate Brian Mundubile received 38%, with smaller parties sharing the remainder.
Post-election discussions questioned the outcome, with some outrightly accusing the winner of stealing the election. This is precisely where election analysis needs to perform better. The Zambian debate raises a broader question that extends beyond this particular election: what happens when an election result contradicts the expectations of some analysts and commentators?
Among some commentators, the only way an incumbent can win an election is if they steal it. That is a self-sealing argument. It cannot be faulted because such analysis front-loads its conclusions in a way that makes its authors “right” regardless of the election outcome.
There may be factors that analysts failed to account for in assessing an election. These could include turnout, regional voting patterns, late shifts in voter preferences, campaign organisation, incumbency advantages, the state of the opposition or simply a misreading of public sentiment.
The fact that some voices confidently predict a particular outcome does not, in itself, make a different result evidence of fraud. Predictions are not election results. The failure of a prediction is not proof that voters, the Electoral Commission or the entire electoral process were manipulated.
There are three separate issues that should not be conflated - what analysts predicted, what voters decided and whether the electoral process was convincing. A prediction can be wrong without an election being fraudulent. Conversely, analysts can correctly predict a winner even when an election is flawed.
An unexpected result is not the same thing as an illegitimate one. Elections are precisely where political assumptions are tested against the actual choices of voters. Sometimes voters surprise the experts. Sometimes the candidate everyone expected to win loses or vice versa. That may tell us that the analysis was wrong. It does not automatically tell us that the election was stolen.
On the other hand, this does not at all mean that an incumbent’s victory should be accepted without question. Credible allegations of irregularities, intimidation, miscounting of ballots, unequal treatment of parties or other abuses must be taken seriously and investigated. But the standard must be evidence.
The same principle applies in the opposite direction. An opposition victory does not inevitably make every aspect of that election credible. An election can be won by the opposition and still be marred by irregularities. That is why the identity of the winner cannot be the test of credibility.
The real questions are: Were people able to vote freely? Was the electoral process administratively and legally adhered to? Were the votes counted accurately? Were the results transparently reported? Were complaints properly investigated? And, more importantly, is there evidence that irregularities actually affected the result?
Those questions should be asked regardless of who wins. But in many African elections, these questions appear to become more prominent when an incumbent wins and less so, if at all, when the opposition sails through.
Are we judging elections by the evidence of what happened or by whether the result matches what we expected to happen?
That is a pertinent question because political analysis can sometimes become trapped in its own predictions. Once a particular outcome becomes the accepted narrative, an election result that contradicts it can be treated not as evidence that the prediction was wrong, but as evidence that something must have gone wrong with the election.
A democratic election is not a referendum on the predictions of analysts. It is a contest in which voters are entitled to surprise everyone. If we genuinely believe in democratic choice, we have to be prepared to respect the voter even when the voter surprises us.
The question, therefore, should not be whether the winner was the candidate analysts expected or even whether the winner was the incumbent or the opposition. The question is whether the election was conducted with sufficient integrity for the result to reflect the will of the voters.
Neither side has a divine right to win. Judge the election, not the winner.
Dennis Mulilo writes on politics and governance.