Thousands of monitors, a PVT and behind-the-scenes diplomacy played a key role, yet questions about the future of observation remain.

From left: Commonwealth observer mission lead Jakaya Kikwete; former Zambian president Rupiah Banda; Edgar Lungu; Hakainde Hichilema; AU observer mission lead Ernest Bai Koroma.

Zambian voters went to the polls on 12 August to vote in presidential, parliamentary, and local government elections. By the time the final results were announced in the early hours of 16 August, it was already clear that opposition leader Hakainde Hichilema and his United Party for National Development (UPND) had won convincingly.

Ahead of the election, there had been much speculation, that regardless of the outcome, the incumbent would attempt to cling to power through authoritarian means. In the end, however, Edgar Lungu conceded quickly and congratulated his successor openly. At a time when there are mounting questions about the future of election monitoring, how much are observer missions to thank for this smooth transition of power?

The role of observers in Zambia’s transition

Credit for the election victory primarily belongs to the UPND, who ran a well-organised campaign, and the voters who turned out in huge numbers and queued for as much as 12 hours to cast their ballots. Nonetheless, observers also played a key role in events that unfolded around election day. A limited number of international “observers”, who were permitted only to gather information but not intervene, were present in a small fraction of polling stations. Meanwhile, thousands of domestic “monitors”, who were empowered to raise concerns to polling station staff, covered much more of the country.

The UPND also deployed a large, internally-vetted and well-trained network of party agents to polling stations. The decision of the Electoral Commission of Zambia (ECZ) to announce local results at polling stations as well as constituency tabulation centres – a measure lobbied for by observer groups and others – helped ensure these agents could provide the party with accurate results to compare to those announced at the central collation centre in Lusaka. This allowed the UPND’s chairman for elections, Gary Nkombo, to dramatically challenge the first result the ECZ tried to declare, pointing to a discrepancy that favoured PF. Following some discussion, the electoral commission corrected its figures.

The most prominent domestic group, the Christian Churches Monitoring Group (CCMG), prepared a parallel vote tabulation (PVT) with technical assistance from the US-based National Democratic Institute (NDI). By law, they were unable to make their findings public until all the official results had been confirmed. Nonetheless, the CCMG was prepared to make its tabulation public sooner if early ECZ announcements were not consistent with their numbers or if the commission did not meet its self-imposed deadline of announcing results within 72 hours of the polls closing. In the end, the PVT, which proved to be highly accurate, was not required to contest the official results. However, it was shared with members of the diplomatic community and other observation groups as early as 13 August. This helped inform behind-the-scenes diplomacy aimed at ensuring that Lungu accepted the election outcome.

As results began to flow, it was evident there had been a huge swing towards the UPND and that it would go on to win a majority. Whether Lungu would concede was another matter. Indeed, on 14 August, the incumbent alleged that the election had not been “free and fair”, claiming his party’s agents had been chased away from polling stations in UPND strongholds. Unfortunately for Lungu, at around the same time he was making this argument, international observer groups were issuing their own preliminary statements. The European Union’s mission was the most critical of the ruling party, making it clear that it was PF, not the UPND, who had been primarily responsible for attempting to manipulate the election. Along with CCMG statistics that showed PF agents had been present in almost all polling stations, this served to undermine Lungu’s claims and to assuage voters.

These statements also likely added to the diplomatic pressure on Lungu to concede. These negotiations involved former Zambian president Rupiah Banda as well as former Tanzanian president Jakaya Kikwete and former Sierra Leonean president Ernest Bai Koroma, from the Commonwealth and African Union observer missions respectively. They persuaded Lungu to step down without opening a court case, which would have prolonged the period of uncertainty and could have led to violence. The three former presidents then facilitated an apparently cordial meeting between Lungu and Hichilema, pictures of which were released to the public.

Persistent questions about observer missions

When elections are won by the opposition, there tends to be less criticism and scrutiny of observer missions as the result is assumed to be credible. This has been the case in Zambia so far. However, some of the persistent problems with election observation were still evident.

International groups spent little time in the country before the election and, overall, were still disproportionately concentrated in urban areas, particularly Lusaka. The AU, Commonwealth, and some smaller regional observation groups all released statements that were mild or vague in their criticisms of the significant problems in the pre-election and campaign periods. These missions, as well as the EU’s, released their preliminary statements before the election results were announced despite this having created problems for them on other previous missions. And there were questions about the partiality of local observers, with some domestic monitoring groups being accused of siding with Hichilema.

Observers played a critical role in Zambia’s successful election, but these ongoing shortcomings mean questions about the future of election observation will not go away soon – especially with the court annulments of elections in Malawi and Kenya still relatively fresh in the memory.