Recounting 185 locations found no serious irregularities, CEC does not expand the process

In the December 28th parliamentary elections last year, the verification of counting at several locations raised serious doubts about manipulation of the candidates' votes for MPs. These allegations prompted the Central Election Commission to establish the recounting of all seats, while, at the Prosecution's order, hundreds of people engaged in the voting counting process and representing different political subjects were arrested.

However, just a few months later, in this year's June 7th elections, the recount process has drawn a completely different mirror.

CEC spokesman Valmir Elez has indicated to Kosova Prees that the recounted samples include deployments from all municipalities in Kosovo and, according to findings, are mainly related to recategorisation of ballots from invalid to valid and with human error during the initial count.

“The Central Election Commission has recounted 185 out of a total of 2,498 regular locations, whose ballots were previously counted in the Municipal Counting Centres. The selected sample has included locations from all municipalities in order to verify the accuracy of the counting process. The recount of 7.41% of regular deployments has not included findings that would reason on the need for further recounts. Consistent differences have been minimal and are mainly related to recategorisation of ballots from invalid to valid, as well as human errors during the initial count. As a result, no basis has been found that would require the expansion of the recount process at other locations”, Elesi said, the KP broadcasts Periscopi.

According to Elez, even among the candidates for MPs, the changes have been almost negligible. In 92 percent of cases, the difference has been only one vote for the candidate at the current location, while cases with greater changes have been rare.

“After the recount, from comparison of the candidates' votes for MPs, findings show that most differences are minimal. In 92% of the cases the gap has been only 1 vote per candidate in that vote. This means that if a candidate had 50 votes in a vote by the first count, his score could have come up with 49 votes or 51 votes after the recount. The biggest changes than one vote have been rare: 118 cases with increases of 2 votes; 16 cases by 3 votes; and only 2 cases by four votes more. Meanwhile, 77 cases of lowering 2 have been counted; 4 cases of discounting 3 votes and 2 cases of 4 votes” showed Elez.

Minimum changes have also been recorded in the results of political subjects. In 27 seats of 16 municipalities, corrections have been registered, with a total of 36 added and 20 votes down, resulting in a net change of only 16 votes.

Also, after the recounting of 185 polling, changes in the votes of political subjects have been counted in the 27 seats of 16 municipalities. In general, political subjects have received a total of 36 votes, while 20 votes have been dropped. For example, one subject (A) may have had 100 votes in the initial count and after the recount of 102 votes (+2). Another subject (B) may have had 80 votes and dropped to 78 (-2), while another subject (C) may have had a slight change, for example. +1 or not. This shows that the results of the political subjects before and after the recount are almost the same, with minimal corrections in certain locations” said Elesi.

Elez has indicated that during the recounting process, in addition to verifying the results, several other issues related to the assessment of balloting, verification of signatures on voters' lists, and similarities in the way the paperwork is met. These findings are recorded according to the locations, and by the time they are counted in the QNR, competent organs have been announced for further treatment.

In past elections held on December 28th 2025, the Central Election Commission had decided to recount all the polls, as during initial verifications and recounting a portion of the polling seats, differences were recorded in the number of candidates' votes. According to the CEC, viewing videos from the Municipal Counting Centres and verifying ballots on a polling sample showed that, in some cases, ballots recorded in the results of the count were not compatible with ballots on ballot papers. To ensure the accuracy of the final outcome and the integrity of the electoral process, the CEC had decided to recount all regular locations, 2,557 such ones, and the process had lasted nearly three weeks. /Periscopi