The Coalition for Free and Fair Elections “Pod Lupom” (Eng Under the Lens) assessed that Election Day at the 2026 General Elections passed successfully, with significant improvement in the integrity of the election process thanks to the implementation of new election technologies – biometric voter identification and authentication, including optical ballot scanning.
At the same time, the Coalition warned that successfully implementing technology on Election Day does not mean the process of enhancing election integrity is over. According to their findings, many problems and risks remain throughout the election cycle, from the pre-election period to voting and vote counting.
Technical difficulties did not lead to abandoning the new technologies, which was noted as positive.
The issues with electronic voter identification and authentication devices were recorded at 14 percent of polling places, while technical problems and delays in the operation of optical scanners were recorded at 21.3 percent of polling places.
Despite this, the technology, according to findings from Pod Lupom, continued to be used at polling places.
The Coalition stated that it was particularly important that the introduction of new technologies did not result in challenges to the election results because of their implementation. Acceptance of results is an important indicator of public trust in the new method of conducting elections.
“An Important Step, but not the End of Electoral Reform”
“Pod Lupom” warns that technological progress on Election Day cannot solve the issues of the election process. Elections, as they stated, are not only Election Day, but part of a broader question regarding the integrity of the democratic system.
Findings from long-term monitoring indicate the continuation of risks such as vote pressure, vote-buying, misuse of public resources, irregularities during the formation of polling station committees, and other forms of possible damage to electoral integrity.
Thus, the Coalition believes that the next stage of the reform should be directed towards the period preceding the election day. This entails improving the legislative framework, defining the responsibilities of all election participants more clearly, detecting and processing irregularities more efficiently, and ensuring greater certainty, speed and stricter sanctioning for violations of election rules.
The Coalition believes that the 2026 General Elections represent an important turning point, but not the end of electoral reform.
Voter Turnout at 42.9 Percent, Significantly Lower than in 2022
The Coalition, “Pod Lupom”, also tracked voter turnout. According to their data, a total of 42.9 percent of voters voted by 7:00 P.M., with a statistical margin of error of plus or minus 0.34 percent.
The Central Election Commission of Bosnia and Herzegovina announced the turnout of 42.02 percent. For comparison, at the 2022 General Elections, the turnout was around 50 percent – a total of 1,652,501 voters turned out.
According to data from “Pod Lupom”, a decline in voter turnout was recorded at 89 percent of polling places, while an increase was recorded at the remaining 11 percent.
At the polling stations where turnout dropped, an average of 9.82 percent fewer voters voted.
At certain polling places in Prijedor, Zvornik, Posušje, Novi Travnik, Jablanica, Široki Brijeg, Zenica, Ravno, Derventa, Laktaši, and Ljubuški, the drop in turnout ranged between 30 and 53 percent.
“Pod Lupom” states that lower turnout can be a realistic image of voter turnout due to reduced opportunities for certain past abuses. They cited the following examples: voting more than once, voting on behalf of individuals who did not show up or deceased voters, voting at multiple polling places, and manipulating ballots after the fact.
Nearly 900 Non-Partisan Observers Monitored the Elections
The Coalition “Pod Lupom” deployed 893 civic, non-partisan observers who monitored the election process at the polling stations. The opening of election stations was also monitored, as well as voting, closing, the implementation of new election technologies, vote counting, and the packing of election materials.
The observers had unhindered access at most polling stations.
The issues were recorded at 1.6 percent of polling stations, mostly because the list of accredited observers was not delivered to the election committees, or observers were not allowed to enter before 7:00 AM.
The Coalition stated that it was about sporadic procedural lapses which did not compromise their observer mission and were subsequently resolved.
Most Polling Stations Opened on Time
At 7:00 AM, 52.7 percent of observed polling stations were open. Between 7:01 and 7:15 AM, 36.4 percent were opened, between 7:16 AM and 8:00 AM, 10 percent opened, while 0.9 percent of polling stations opened after 8:00 AM.
All members of the polling station committees were present at 6:00 AM at 92.6 percent of observed stations, while all members were not present at 7.4 percent of locations.
Manual counting of ballots before opening polling stations was conducted at 96.8 percent of polling stations.
Secrecy of the vote was guaranteed at 99.9 percent of observed polling stations. At 98 percent of polling stations, the list of all polling committee members along with their affiliated political parties was visibly displayed.
At 2.3 percent of polling stations, there were doubts, indications, and complaints that the president or deputy president of the polling committee were affiliated with a party.
Suspicions that a certain member of a polling committee was falsely representing another political entity were recorded at 1.3 percent of polling stations.
16.8 Percent of Polling Stations Were Not Accessible to Persons with Disabilities
One recurring issue concerned the accessibility of polling stations. According to observer data, 83.2 percent of polling stations were accessible to persons with disabilities using wheelchairs or other mobility aids.
Such accessibility was lacking at 16.8 percent of polling stations. Badges, political posters, religious symbols, or signs carrying potentially discriminatory meanings within 50 meters of polling stations were recorded at 2 percent of locations.
At 75.7 percent, no such markings were observed, while observers lacked information for 22.3 percent of polling stations.
New Technologies Mostly Ready Before Opening
New technology operators were present at 6:00 AM at 97.4 percent of observed polling stations. Biometric identification/authentication devices and optical scanners were ready for use before opening at 99.5 percent of locations.
Instructions for voters on proper marking and scanning of ballots were posted at 98.5 percent of polling stations.
The initial report of the biometric identification and authentication device, showing a zero baseline, was printed at 99 percent of polling stations.
The same percentage was recorded for the scanner’s initial report. Observers were allowed to inspect the initial reports from these technologies in 99.8 percent of cases.
Scanners Had Issues at Over One-Fifth of Polling Stations
The largest part of technical issues related to optical scanners. Problems and interruptions that disturbed the voting process were recorded at 21.3 percent of polling stations.
Problems with the electronic voter identification and authentication devices, which disturbed the voter identification process, were recorded at 14 percent of polling stations. However, not one polling station recorded abandoning the use of scanners during the voting process. Before or immediately after opening, scanners were abandoned at 0.6 percent of polling stations.
For the devices for identification and authentication, no abandonment was recorded, while at 0.8 percent of polling stations the abandonment occurred before or immediately after the opening. In the cases of technical issues, polling station committees and operators acted in accordance with prescribed procedures in 36.1 percent of documented cases.
Recorded Voting without Scanners and Fingerprints
The Coalition documented the cases in which voters were not identified via scanning their fingerprints. At 39.8 percent of polling stations, more than 10 such cases were recorded.
At 0.2 percent of polling stations, ballots were deposited into the ballot box without prior scanning, even though the scanner was functional.
At 1.3 percent of stations, observers noticed a polling committee member, operator, or other individual inserting ballots into the scanner on behalf of voters.
Voting without required identification documents occurred at 0.1 percent of polling stations.
Family Voting Recorded at 44.7 Percent of Polling Stations
One of the most frequently observed problems was family voting, multiple people entering a voting booth together. Such cases were recorded at 44.7 percent of polling stations, mostly involving up to 25 instances per station.
More than 25 such cases were recorded at 0.2 percent of polling stations.
Misuse of voter assistance was also present. At 21.3 percent of polling stations, it was recorded that the same person helped a larger number of voters, generally up to five cases. At 0.1 percent of polling stations, more than five such cases were recorded.
According to the findings “Pod Lupom”, the technology for identification and authentication prevented offering voter assistance at 13.6 percent of polling stations, up to five cases each. Of voters who received assistance, 26.7 lacked required proof or certification verifying their actual need for help.
Photographing Ballots and Compromising Vote Secrecy
At 2.9 percent of polling stations, observers spotted voters taking photos of their ballots inside the booth.
Voting secrecy inside the booth or during ballot scanning was compromised at 2.8 percent of stations.
Overt campaigning or shouting at voters to support specific political entities was recorded at 0.7 percent of polling stations or in their immediate vicinity.
Intimidation or pressure on voters was noted at 0.1 percent of polling stations. At 0.8 percent of stations, unauthorized persons lingered inside and disrupted the voting process.
Election Day Mostly Wrapped Up by 8:00 PM
At 7:00 PM, 58.7 percent of polling stations closed. Between 7:01 PM and 7:15 PM, 24.9 percent closed; between 7:16 PM and 8:00 PM, 13.3 percent closed; and 3.1 percent closed after 8:00 PM.
At 30.8 percent of polling stations, voters who were present at or outside the station at closing time were allowed to vote. This was denied at 3.8 percent of stations, while at 65.4 percent of stations there were no voters waiting to vote at closing time.
Electronic and Manual Results Matched at 59 Percent of Polling Stations
After closing the polling stations, the scanner printed a Consolidated Results Report (FER) at 99.2 percent of polling stations.
A copy of the FER was provided to observers at 94.6 percent of polling stations. Electronic transmission of scanner results before the start of manual counting was completed at 98.3 percent of stations.
Problems during the electronic transmission were recorded at 10.7 percent of polling stations. An especially important finding relates to the comparison between manual and electronic vote counting.
According to the reports by observers, the results of electronic and manual vote counts matched across all government levels at 59 percent of polling stations. At 41 percent of stations, the results did not match.
This data relates to the discrepancy rate between electronic and manual vote counting methods, and it doesn’t mean that electoral fraud was recorded at polling stations. Differences between the two count methods can stem from various technical or procedural causes and require further investigation by relevant authorities.
Altering Ballots Recorded at 0.1 Percent of Polling Stations
Declaring ballots as invalid was conducted in accordance with the rules of CEC BiH at 99.7 percent of polling stations. Adding votes directly to ballots during counting was recorded at 0.1 percent of polling stations.
At 37.7 percent of polling stations, ballots without the QR code were declared invalid, while at 0.4 percent of polling stations this was not the case. At 61.9 percent of polling stations, no ballots without the QR code were recorded.
All observers could clearly and smoothly monitor the counting process across all government levels at 97.3 polling stations.
Most Summary Forms Posted, but FER Not Delivered Everywhere
Finalized SIT ZR summary result forms were signed by all polling committee members at 98 percent of polling stations.
Yellow copies of SIT ZR forms were prominently posted at 92.2 percent of polling stations, whereas copies of the FER electronic results report were not posted at 6.8 percent of stations. Observers photographed posted forms at 93 percent of stations and retained their own copy of the FER at 91 percent of stations.
Biometric identification devices and scanners were properly packed and handed over at 98.7 percent of polling stations.
All polling station procedures were completed at 97 percent of stations.
“Pod Lupom” Calls for Earlier Preparations Ahead of Next Elections
Based on Election Day observations, the Coalition “Pod Lupom” recommends that preparations for the next elections start earlier. A special focus, as they stated, should be placed on troubleshooting procedures for technical issues, timely and comprehensive training for election administration, and strengthening technical support at polling stations.
At the same time, technological reform must be paired with stronger mechanisms to prevent, monitor, and penalize irregularities prior to Election Day. The Coalition emphasizes that election integrity cannot be guaranteed solely through the deployment of biometric identification devices and scanners.
Voter pressure, vote-buying, misuse of public resources, polling committee malfunctions, and other forms of irregularities remain part of the broader election process.
For these reasons, “Pod lupom” believes the next phase of electoral reform must target the pre-election period – focusing on regulations, accountability of political entities and election officials, oversight of the electoral process, and more effective detection, prosecution, and punishment of violations.
At the conclusion of the monitoring process, the Coalition noted that its observers experienced some form of pressure at 4.4 percent of polling stations.
The data thus presents a dual picture of the elections: new technologies were successfully introduced and utilized across nearly all polling stations – overcoming technical hurdles without systemic abandonment – yet various procedural irregularities and risks persist, demonstrating that technological advancement represents only one component of comprehensive election reform.
