CCMG PVT findings for the presidential elections falls within the official results as announced by ECZ

10 Min Read

Note: ECZ official presidential results for 155 of 156 constituencies (excluding Mandevu constituency) while CCMG PVT estimates for all 156 constituencies.

CCMG Verification Statement on the Accuracy of the Official 2021 Presidential Results


Today, the Christian Churches Monitoring Group (CCMG) is releasing its verification statement on the accuracy of the presidential results. CCMG conducted a parallel vote tabulation (PVT) for the presidential election as part of our comprehensive monitoring of the 2021 general elections. As we stated during preliminary statement, CCMG completed its PVT by 12 noon on Friday 13 August 2021 with over 99% of our monitors having sent in their observation reports. Had it proved necessary, CCMG would have released its PVT estimates if we had any indication there was manipulation of the official results. We commend the Electoral Commission of Zambia (ECZ) for fulfilling its constitutional duty with honour and are grateful to them for not having to do so. CCMG now confirms that its PVT verifies that ECZ’s official presidential results reflect the ballots cast at polling stations.


CCMG’s PVT deployed carefully recruited, highly trained accredited independent non-partisan monitors to a nationally representative random sample of 1,500 polling stations with 866,689 registered voters of whom 463,625 are female (53.5%)[1]. The PVT sample was carefully constructed according to well-established statistical principles to ensure that it was nationally representative and included polling stations in every province, district and constituency of the country. This is done by stratifying by province, district and constituency to ensure the percentage of polling stations, registered voters and female registered voters in the sample closely matches the percentage for the entire country (see Annex 6). For example, Eastern province has 11.8% of all polling stations, 12.8% of all registered voters, and 13.1% of all female registered voters in the country, and so the CCMG’s PVT sample matches this distribution with 11.8% of polling stations,12.8% of registered voters, and 13.3% female registered voters coming for Eastern province in the sample. This ensure that CCMG’s PVT sample is truly nationally representative and therefore, can determine the quality of the voting and counting processes and can estimate accurately estimate how Zambians cast their ballots in all 12,152 polling stations.

On election day, CCMG’s PVT monitors witnessed and reported on the entire voting and counting process, arriving at 5:30 in the morning and staying at their assigned polling station until counting was completed and results have been posted. Because CCMG’s PVT monitors are deployed to a nationally representative random sample, the PVT methodology provides the most accurate information on the conduct of voting and counting and it is the only methodology that can provide independent verification of the accuracy of official election results. CCMG’s PVT is not an exit poll. Our monitors do not ask any voter for whom he or she voted. CCMG’s PVT monitors also do not count the presidential ballots as that is the responsibility of the polling officials. However, our monitors are present throughout the counting process and report on its conduct. CCMG’s PVT estimates, therefore, are based on the official and publicly available official results from polling stations.

PVTs are used around the world and across Africa to independently determine if official election results reflect the ballots cast. The PVT methodology has been successfully employed by citizen observers for multiple elections in numerous African countries, including: Burkina Faso, Côte d’Ivoire, Kenya, Malawi, Nigeria, Ghana and Zimbabwe. PVT’s have routinely been conducted for presidential elections in Zambia with PVTs previously conducted in 1991, 2008, 2011, 2015 and 2016.

The PVT methodology is impartial and non-partisan and reflects the ballots cast regardless of the outcome of an election. Thus, PVTs in Zambia have verified results in past elections both when the ruling party candidate won the election as well as when an opposition party candidate prevailed. For example, the PVT verified the results of the 2008 presidential election when the late Michael Sata lost as well as confirmed his victory in 2011 (see Annex 5).

PVTs serve to independently verify official election results by comparing the official result with the PVT’s estimated range for each candidate (as well as for rejected candidates and turnout). If the official result falls within the PVT estimated range, then the official result reflects the ballots cast at polling stations. However, if the official result falls outside of the PVT estimated range, then the official result has been manipulated and the PVT suggests the true result.

CCMG is an independent, non-partisan citizen monitoring network that conducted a comprehensive observation of the 2021 general elections, including a long-term observation of the pre-election period and the election day PVT. Our observation effort was undertaken to ensure a credible electoral process for all Zambians, and we are beholden to no candidate nor the ECZ.

CCMG PVT Findings

As Table 1 shows, for the two leading candidates as well as for all other candidates combined, the official results announced by the ECZ[2] fall within the PVT estimated range and therefore, we can declare with confidence that the official results reflect the ballots cast at polling stations (see Annex 1 for CCMG PVT estimates for all presidential candidates).

The PVT estimated range for Hichilema’s vote share is 57.7% to 60.5% and his ECZ official result is 59.4%, which falls within this range. Similarly, for President Lungu, the ECZ official result is 38.4%, which falls within the PVT estimated range for his vote share of 37.1% to 39.7%.

Note: ECZ official presidential results for 155 of 156 constituencies (excluding Mandevu constituency) while CCMG PVT estimates for all 156 constituencies.

Because CCMG’s PVT estimates show that the lowest possible result for Hichilema (57.7%) is greater than the highest possible result for President Lungu (39.8%), the only possible outcome of the election is for Hichilema to receive the most votes of any presidential candidate.

Further, according to CCMG’s PVT estimates, the lowest possible result for Hichilema is 57.7%, which is greater than the 50%+1 vote share required to prevent a presidential runoff election.

CCMG’s PVT estimate for turnout is 70.4% +/-0.7%, with a range of 69.7% to 71.1% and for rejected ballots the PVT estimate is 2.5% +/-0.1%, with a range 2.4% to 2.6%. With 155 of 156 constituencies announced, ECZ’s turnout figure is 70.8%[3] with 2.3% rejected ballots.

CCMG’s PVT monitors reported the election day process was inclusive and transparent, as both PF and UPND party agents participated throughout voting and counting across the entire country (see Annex 2 and 3). As previously reported, PF and UNDP had party agents at 98% and 99% of polling stations during the counting of ballot papers. The PF and UNDP party agents also agreed with the official vote count for the presidential election at 98% and 99% of polling stations, respectively. At 97% of polling stations, a PF party agent signed the official presidential results form (GEN 20), while at 99% a UPND party agent signed the official presidential results form (GEN 20). In addition, polling officials posted the official results for the presidential election at 95% of polling stations (see Annex 4).

Overall, while there were serious challenges during the pre-election period that raised questions about the fairness and credibility of the electoral process, the ECZ conducted a well-organized and credible election day process that largely adhered to official procedures, and both PF and UPND party agents were involved through the voting and counting across all provinces. Given that the CCMG’s PVT estimates verify the ECZ’s official results for the presidential election, all Zambians should have confidence the ECZ’s official presidential results are accurate and reflect the ballots they cast at polling stations. We commend the ECZ for meticulously tabulating the presidential results and announcing results in line with its mandate.




Share this Article
Leave a comment