National Election Watch (NEW), an election monitoring group, has released its findings regarding the credibility of Sierra Leone’s presidential, parliamentary, and local council elections held on June 24th, 2023.

Deploying 6,000 observers across the country, including 750 specially trained individuals, NEW employed the Process and Results Verification for Transparency (PRVT) methodology to assess the quality of the election process and verify the accuracy of the official results announced by the Electoral Commission of Sierra Leone (ECSL).

According to NEW, the candidate of the Sierra Leone People’s Party (SLPP) should receive between 47.7% and 53.1% of the vote while ECSL’s official result is 56.1%; The candidate of the All People’s Congress (APC) should receive between 43.8% and 49.2% of the vote while ECSL’s official result is 41.2%.

Read full report below:

To monitor the credibility of the 24th June 2023 presidential, parliamentary and local council elections, National Election Watch (NEW) deployed 6,000 observers across the country covering every polling centre. Out of these observers, 750 were specially trained and deployed to a statistically representative sample of polling stations across all the 5 regions and 16 districts in Sierra Leone using the Process and Results Verification for Transparency (PRVT) methodology. The PRVT allows NEW to assess the quality of election day processes and verify the accuracy of the official election results announced by the Electoral Commission of Sierra Leone (ECSL). NEW conducted its observation in a strictly independent and non-partisan manner on behalf of all Sierra Leoneans to ensure that the results announced are accurate and verifiable.

PRVT is an advanced and proven election monitoring methodology used by citizen election observation groups around the world that allows for the independent and impartial assessment of the quality of election day processes and the official results. The deployment of observers and their reports are strictly impartial. On election day, NEW deployed specialised observers to a statistical sample of polling stations, where they examined crucial aspects of voting and counting at assigned polling stations, reported on strengths and irregularities that impacted on the quality of the election and noted the number of ballots cast and the votes recorded by officials for each candidate at the polling station. NEW’s PRVT analysis is based on data from polling stations, before it proceeds to other tabulation processes, where transparency was limited.

NEW’s PRVT exercise in the 2023 election draws on its experience in successfully executing PRVTs to observe the 2018, 2012, and 2007 elections in Sierra Leone. In all three of these elections, NEW’s projections matched the official results of the Electoral Commission. Unfortunately, in 2023, we find that the published results of ECSL are inconsistent with our PRVT data. In line with the statement released earlier on June 27, titled “PRVT Update: Partial Election Results Released by ECSL”, this statement seeks to share NEW’s PRVT data to the public as follows:

PRVT results assessment

Based on reports from 99% (747 of 750) of sampled polling stations, NEW’s statistical analysis shows that:

● Voter turnout is between 75.4% and 79% (a point estimate of 77.3% with a margin of error of +/-1.7%), while ECSL’s official result is 83.0%.

● Rejected ballots are between 4.5% and 5.7% (a point estimate of 5.1% with a margin of error of ± 0.6%) while ECSL’s official result is 0.4%. The percentage of invalid votes is inconsistent not only with the PRVT data, but also with past elections. In 2018, rejected ballots for the first round were 5.2%, 4.7% in 2012, and 7.3% in 2007.

● The candidate of the Sierra Leone People’s Party (SLPP) should receive between 47.7% and 53.1% (a point estimate of 50.4% with a margin of error of +/- 2.7%) of the vote while ECSL’s official result is 56.1%; The candidate of the All People’s Congress (APC) should receive between 43.8% and 49.2% (a point estimate of 46.5% with a margin of error of +/- 2.7%) of the vote while ECSL’s official result is 41.2%;

● Therefore, no candidate should have met the constitutional threshold of 55% to avoid a runoff.

Analysis of Official Results

Based on the ECSL official release of results by district, NEW’s analysis shows there are inconsistencies in the results.

● The total valid ballots cast as announced by the ECSL for the first batch of results (representing 60% of polling stations) shows an average of 269 valid voters per polling station, while the second batch (representing 40% of polling stations) has an average of only 188 valid votes per polling station.

● Where the average number of voters for each polling station appeared to be set at 300, some polling stations had an excessive number of voters. For example: For eight of the districts in the first batch and for one district in the second batch, the average number of valid votes per polling station is more than 300, and the practice is that the last polling station in a polling centre does not have more than 400 voters. The vast majority of polling stations have 300 voters or less, including more than 84% of the polling stations in each of these districts.

For example, according to official ECSL results, polling stations (PS) in the first batch (60%) of results in Bombali (433 PS) have an average of 340 valid votes, while the second batch (40%) of polling stations (288 PS) have an average of 39 valid votes per polling station. In total, Bombali has only 57 polling stations that have 340 or more registered voters, which would make this average a statistical impossibility. In Kailahun polling stations in the first batch of results (407 PS) have an average of 378 valid votes, while the remaining 40% of polling stations (272 PS) have an average of 113 valid votes per polling station. In total, Kailahun has only 22 polling stations that have 378 or more registered voters, which would make this average a statistical impossibility.

In total, for 9 of the 16 districts, the official ECSL results show that more than 300 valid votes were cast per polling station despite the vast majority of polling stations having a maximum of 300 voters.