Zambia: Why five is the magic number when it comes to opposition election victories

Just days before, our research team had feared the worst when President Edgar Lungu claimed the elections had been manipulated  in favour of the Hichilema’s United Party for National Development. This allegation might seem bizarre, as Lungu controlled all of the security forces and his own party cadres committed most of the electoral violence, but it was not unexpected.

In the weeks leading up to the election, the Patriotic Front government had begun to adopt the tactics of former US President Donald Trump, seeking to undermine the credibility of an election that it knew it would struggle to win.

Working with a group of researchers including O’Brien Kaaba, a brilliant legal mind, Nicole Beardsworth, an incisive observer of opposition parties, and Marja Hinfelaar, a central figure in Zambian research for many years and the Director of Research and Programmes at the Southern African Institute for Policy and Research, we had been tracking the different elements of this strategy as they emerged.

While senior PF leaders attempted to intimidate election observers and Western donors by alleging they favoured Hichilema, state media carried front page stories accusing the opposition of using violence and intimidating voters. This propaganda effort was designed to make subsequent claims regarding the unfair treatment of the government more believable.

It therefore felt somewhat inevitable when, after it had become clear that other strategies to manipulate the election in favour of Lungu had failed, the PF issued statements claiming that their supporters and party agents had been intimidated – and that as a result the results should be nullified.

This was not an idle threat – as the final results were being announced, we received news that a team of lawyers was preparing a petition to appeal the process to the Constitutional Court. Because the Court was both created and appointed by President Lungu, this raised serious concerns that Hichilema would be blocked from taking power, plunging the country into political crisis.

This backdrop raises a question with important implications not only for Zambia but also for other countries in which opposition parties are hoping to unseat entrenched autocrats: Why did the government ultimately give up?

While a number of elements played a role, including more effective efforts by the UPND to protect the vote, strong and coordinated international engagement, and the willingness of some civil society groups and the electoral commission to stand up and be counted, perhaps the most important factor that facilitated the transfer of power was the size of the opposition victory.

Opposition parties rarely win by tiny margins in authoritarian countries. There is a simple reason for this – small margins are particularly easy to erode with electoral fraud, especially if opposition parties and domestic monitors cannot observe the vote count in all polling stations.

In other words, to ensure that an opposition party converts their popularity into power, they need to win by more than one to two percent. This is not true of every election – there have been some close opposition victories in Cape Verde, Ghana and Sierra Leone – but it tends to be the case in less democratic regimes.

Think about the most notable opposition wins of the past twenty years: in most cases, they were overwhelming victories.

• In 2002, Mwai Kibaki beat the ruling party candidate Uhuru Kenyatta by a remarkable 31% to end decades of KANU rule in Kenya.
• In 2015 Muhammadu Buhari beat President Goodluck Jonathan by about 10% to overturn the PDP’s dominance of Nigerian politics.
• More recently, Lazarus Chakwera thrashed President Peter Mutharika by around 20% in Malawi in 2020.
• The Zambian election of 2021 saw an almost identical result, with Hichilema capturing 59% of the vote.

In all of these cases, the opposition didn’t just win – it won big.

Where preventing electoral manipulation is concerned, the magic number seems to be a winning margin of 5% or more. While it obviously gets harder to rig an election the fewer votes the government receives, there are two reasons why this threshold is particularly significant.

First, when the opposition wins by much more than 5%, it becomes much harder to fabricate a government win without undermining the credibility of the process.

It also means that even if the results of certain polling stations or constituencies are discounted for spurious reasons the government is unlikely to be able to make up the difference.

In the 2016 Gambian elections, for example, when opposition leader Adama Barrow shocked long-time dictator Yahya Jammeh, a big margin was important to ensure the president was still behind after the electoral commission announced that it had initially “miscounted” and released new figures that removed thousands of votes from Barrow’s tally.

Second, the parallel vote tabulations run by domestic monitors are based on estimating the true outcome of the election using results recorded from a sample of polling stations. With typical sample sizes of 1,100 to 1,500, the margin of error of this projection is usually in the region of +/- 1.2% to +/- 2.5% for the main candidates.

What this means in practice is that in a two-horse race, the Parallel Vote Tabulation (PVT) will “confirm” a range of results up to a 5% spread, and so the outcome can be manipulated within this range and still appear to be “approved”. Once this has happened, it is unlikely that international observers will condemn the poll, and becomes increasingly difficult for international partners to question the government’s victory.

By contrast, once the opposition’s lead goes much above 5%, electoral fraud is likely to be exposed by the PVT – at this point, the election can still be rigged, but doing so becomes more costly.

If governments are determined to retain power come, they can still find a way. In the last election in the Democratic Republic of Congo the results were simply fabricated to ensure that the regime’s most feared opposition candidate lost. But that was facilitated by the fact that domestic observers were afraid to publicly release their findings, and opposition parties and civil society groups were weak.

Where the government is less repressive, or civil society is stronger and backed by international partners, the regime can quickly lose control of the political agenda. In Zambia, for example, the size of Hichilema’s win emboldened a number of the more influential minor presidential candidates to concede defeat and either congratulate Hichilema or endorse a transfer of power.

As the pressure on Lungu to step down started to snowball, gaining support from a broader range of organisations and individuals over time, the electoral commission was empowered to announce that Hichilema had won before the president had conceded and while the government was still protesting the result.

It was only at this point that the calmer voices within the ruling party, which appear to have included the president himself, were able to persuade its hardliners that things had gone beyond the point of no return – and that by improving the PF’s reputation and repairing relations with the opposition, a dignified concession now represented the best way to try and secure the future interests of the party and its leaders.