The International Society for Fair Elections and Democracy (ISFED) used the Serbian model in the parallel vote tabulation (PVT) during the October 31 parliamentary elections. 

Tamar Sartania, the ISFED board member, told journalists that the NDI experts involved in the process didn’t get a written notice about the model that had a technical fault and counted invalid ballots.

A member of the ISFED board elaborated on the parallel vote tabulation methodology. She explained that the PVT formula changes periodically. This time the tabulation formula used during one of the last elections, namely in Serbia in June 2020, was applied.

“This is not Georgian know-how, this is a proven methodology used in many countries around the world, and this methodology is constantly updated,” Tamar Sartania said.

Invalid ballots do not count in Georgia for a long time. So, the question raised that if ISFED carried out PVT in 2016 correctly, why there is a technical fault in the formula this time, a member of the Board continued.

“The fact is that the 2020 software is different from 2016 software. The methodology has been used globally for four years. During this period, it was updated, added extra functions, revised. Each subsequent PVT relies on the most recently used formula and not on the PVT previously used in the country. 

According to the commission, this is the reason why the Serbian formula was used in Georgia. However, the invalid ballots were counted to calculate the percentage of candidates in Serbia, which held elections in June 2020. How is the PVT methodology process done, who is responsible? The commission found out that ISFED does not do PVT independently, it is the responsibility of ISFED, but it hires NDI experts and works with them on software and precinct selection. NDI helps, supports ISFED in data analysis to some extent, but ISFED is responsible for the final product, as it is the contracting authority of this process.

Communication is crucial during this process. No matter who develops the software, when you do not give the right task to a person or group of people, they would not be capable of writing a relevant soft. The program found that the following was a flaw, ISFED provided NDI experts with the list of logical tasks, although it did not provide written information on how the votes received by candidates were calculated in Georgia,” Sartania said. 

She called this fact a human mistake.

ISFED released a statement on the parallel vote tabulation of Georgian parliamentary elections on December 11. The statement read that the internal audit results confirmed the accuracy of the PVT methodology. However, a technical fault was identified in the vote calculation formula. Namely, the formula counted the number of invalid ballots along with the votes received by the electoral subjects, ISFED informed.

The adjusted ISFED PTV results coincided with the final results of the CEC.