Transparency International in Bosnia and Herzegovina (TI BiH) calls on the Central Election Commission to investigate suspicions of electoral fraud in the early elections in Republika Srpska, given that data from a large number of polling stations cast doubt on the integrity of the entire electoral process. TI BiH reminds that numerous court rulings in BiH in recent years have established the mass occurrence of identity theft and vote manipulation for voters who do not participate in elections, and the same phenomena were recorded on election day itself, which some voters have publicly testified to.
Due to the small difference between candidates and a number of polling stations that reported unnaturally high turnout compared to the average, suspicion arises that irregularities could have affected the result, as votes from just 40 disputed polling stations (Table 1) make a difference of over 8,500 votes. These stations are mostly located in Doboj, Laktaši and Zvornik and reported much higher turnout than average, and in some cases over 80 or 90 percent.
At most of these polling stations, the ruling coalition candidate won over 80% of votes, and in some cases over 97%. Particularly noteworthy are the polling stations Doboj – Opsine, where turnout exceeded 90%, and Doboj – Majevac with 80%, where in both cases the opposition candidate received only 5 votes each.
At 9 of the 40 disputed stations, such high turnout was reported that according to this, during the entire 12-hour election day, one voter would have had to vote every minute and a half, which is hard to believe. Therefore, TI BiH requested from CEC data on turnout until 11 AM and 3 PM from these polling stations, where the average time in some intervals would have to be even less, and this was precisely the basis for annulling the elections in Doboj in 2020.
The CEC decision from 2020 states that the reported turnout at certain polling stations in certain intervals was such that one voter voted every 50 seconds, and subsequent graphological examination of signatures established forgery of between 10-20 percent of signatures. Court proceedings that were subsequently conducted against polling board members established the mass occurrence of vote theft from voters living abroad and persons who had died but were not removed from the voter list.
At certain polling stations on this list, the turnout itself is somewhat lower, but there was a drastic increase in the number of voters between two electoral cycles (DOBOJ – OŠ DOSITEJ OBRADOVIĆ 2B 382 more voters, ZVORNIK – KARAKAJ D 137 more, ZVORNIK – SNAGOVO 72 more).
TI BiH has previously warned that outdated voter lists in BiH compiled by competent MUPs leave enormous room for these types of electoral manipulations. The scale of this is best evidenced by the verdict for electoral theft in the village of Klisa near Zvornik in the 2022 elections. It was reported that out of 230 voters in that location, 197 (85.65%) came to vote, of whom 103 voted for Milorad Dodik, 62 for Senad Bratić, and the rest for other candidates.
The BiH Court subsequently sentenced three polling board members to suspended prison sentences “for signing instead of a large number of voters and falsifying the election result”. That disputed result at this polling station was never annulled, and the room left for manipulation of election results is sufficiently illustrated by the fact that on Sunday only 12 voters came to this polling station instead of 190 as shown in the previous elections.
In this, as well as in previous verdicts for Doboj, Novi Grad, Brčko, etc., courts in BiH have established that there is mass abuse of the identity of voters living abroad, and such phenomena, according to TI BiH information, were also reported in these elections. In Laktaši, testimony appeared from a citizen who reported that someone voted instead of her and members of her family living abroad, and a similar case was reported in Doboj.
Suspicions regarding election regularity
Apart from the mentioned 40 polling stations cited as an indicator that potential irregularities could affect the outcome of the election for President of Republika Srpska, it should be emphasized that electoral forensics of the overall results opens additional suspicions.
Data from all polling stations shows that the difference between the two candidates unusually increases with rising turnout. If a so-called descriptive analysis of vote distribution is performed on official data, which is a standard and internationally recognized tool in statistical electoral forensics, there is an evident sharp increase in vote difference at high turnout. The difference moves linearly up to the average turnout level, while the share of votes for the SNSD candidate unusually accelerates with the trend of high turnout at a number of polling stations with extreme values that are statistically improbable.
The difference is not in accordance with expected statistical patterns and indicates the presence of unusual outliers that cannot be explained by natural variation of the electorate. Given the nature of vote distribution, such extreme trends require further investigation to determine whether manipulation or other factors that undermine election regularity exist. It should be noted that this same phenomenon has been established by numerous studies in countries with seriously compromised electoral process integrity.
The opposition candidate, according to this data, has an advantage at all polling stations that reported turnout up to 53 percent, while the SNSD candidate makes a difference of 10,000 votes at a smaller percentage of polling stations that reported much higher turnout.
Also, the data shows an unusual trend of growth in the number of invalid votes in Doboj, Zvornik and Laktaši relative to the total number of votes in those cities.
Note: Overview of total number of votes by municipalities/cities, and number of invalid ballots. Size indicates the advantage achieved by the winning candidate. CEC data from November 24, 2025, retrieved at 4:41 PM and is not official and serves for informational purposes.
In addition to this, CEC should consider numerous allegations of electoral manipulations that have appeared in the media and are supported by certain evidence, as they seriously undermine citizens’ trust in the electoral process.
TI BiH believes that the Central Election Commission should initiate proceedings ex officio and investigate whether the rules of the electoral process have been violated. We remind that a similar procedure was conducted in 2020 in Doboj, when graphological examinations of voter signatures were performed and compared with IDDEEA data. Findings showed that 10-20% of signatures were forged, and after dozens of court proceedings, it was later established through voter testimony that their identity had been abused.
Also, the fact that a similar scenario has been occurring for several electoral cycles indicates the need for electoral technologies for voter identification and vote counting to be implemented in upcoming elections in order to restore citizens’ trust in the electoral process and protect one of the fundamental rights.
View data on turnout and candidate results by polling stations HERE.
