Number twisting could be enticing. If you get carried away with the idea, before you know it, you may get caught citing figures from satirical websites.
By Vanja Stokić (Bosnia and Herzegovina), Zoran Strika (Serbia), Teodora Đurnić (Montenegro), Gentiana Paçarizi (Kosovo) and Jona Plumbi (Albania)
As we have seen, politicians in the Western Balkans frequently manipulate numbers to bolster their reputation or undermine their opponents. This practice becomes especially pronounced during election campaigns, as they seek to sway voters, but it is also common during their time in office. Numbers serve not only as a tool for misleading voters but also as a means of projecting competence—an approach described in Sergei Guriev and Daniel Treisman’s book Spin Dictators.
A well-established tactic, if not unique to the region, is exaggerating the size of your own public gatherings while downplaying those of your opponents. Similarly, fabricated public opinion and popularity polls are often presented as credible. Casual hyperbole seems to have acquired a status of a completely legit figure of speech around numbers. As we have seen, one particularly absurd example featured a politician falsely claiming to have planted a million trees.
Across the region, both ruling and opposition parties seem to feel that casually inflating different numbers is beneficial for their electoral fortunes. During a pre-election rally this year, the mayor of the Central Bosnian city of Zenica, Fuad Kasumovic, claimed that he had paid off 40 million BAM (20.5 million EUR) of the city’s debts, incurred by the previous administration. After reviewing the data from the state Ministry of Finance and Treasury, it was determined that only 12 million BAM (6 million EUR) of debt had been repaid. At another pre-election rally this year, in Travnik, a town not too far from Zenica, the leader of the opposition SDA party claimed that during the mandate of their member Fadil Novalić as prime minister of the Federation of Bosnia and Herzegovina, the GDP growth rate was three times higher than under the successor government. The fact-checking website Istinomjer showed this was untrue.
A recent example from Albania saw opposition leader Sali Berisha claim that the prime minister’s salary was equivalent to 50 pensions. Faktoje fact-checked this statement and established that Prime Minister Edi Rama’s salary was actually equal to 20 urban pensions or 37 rural pensions.
Numbers in Crisis
Numbers have also been central to crisis PR, such as after the roof collapse at the Novi Sad Railway Station in Serbia’s second-largest city in November this year. The incident claimed 15 lives, with two individuals in critical condition at the time of writing. Amid public outrage and demands for accountability, government officials attempted to deflect criticism by referencing global tragedies — in cities like Miami, Paris, and Jakarta, or natural disasters in Spain — where the death tolls were higher.
Pro-government media amplified these comparisons, and Parliament Speaker Ana Brnabić remarked that “the opposition’s problem seems to be that more people didn’t die in Novi Sad.” Following arrests related to the station’s reconstruction, which had been ongoing for three years, Brnabić again deflected by citing slow justice after a 2018 building collapse in Marseille.
This tactic of using numbers to downplay domestic tragedies is not new. In 2018, after two workers died at a Belgrade construction site, President Aleksandar Vučić dismissed the incident, saying, “This is life,” and bizarrely claimed that in the United States, a window washer dies every 10 seconds. His statement was not only false but was sourced from The Onion, a satirical website. (Following an outpouring of ridicule, Vučić apologised for the remark.)
Crowd Counts
In 2019, the “One in Five Million” protests spread across Serbia, including Novi Sad. A local pro-government outlet, NS Uživo, claimed that only 350 people attended a protest. However, the fact-checking website FakeNews Tragač debunked this claim, revealing that the outlet manipulated photos to show a smaller crowd, while actual attendance was closer to 3,000.
Similar tactics were deployed during other demonstrations. Following two mass shootings in May 2023, a wave of protests erupted across Serbia as citizens blamed systemic failures for fostering violence. The “Serbia Against Violence” rallies saw tens of thousands of participants, yet pro-government media used drone footage taken before or at the very start of events—when attendees were still gathering—to downplay the turnout.
Conversely, pro-government outlets inflated figures when advantageous. For example, they claimed that 200,000 people attended President Vučić’s “Serbia of Hope” rally, organized in response to the May tragedies. Independent estimates, however, placed attendance at no more than 50,000.
Frustrated by these discrepancies, a group of media and cultural activists launched the Facebook group “Protest Fan Club,” which eventually became Public Gatherings Archive, which seeks to provide a realistic picture of public events. Aleksandar Gubaš, a member of the Archive, explained that the government frequently inflates attendance at its events while downplaying opposition gatherings.

“This trend of inflating attendance isn’t unique to the government,” Gubaš points out. “It’s common among opposition parties, civic movements, religious organizations, sports clubs, and event organizers worldwide. People crave spectacle and enjoy hearing about massive crowds, which is why we often hear claims of hundreds of thousands on the streets of Tel Aviv, Prague, or other cities. But in reality, political protests rarely surpass a tiny fraction of the population, even in cities like Belgrade.”
Gubaš explained that they count individuals whenever possible, relying on clear footage and patience. When counting is impractical, they use approximations.
“After analyzing hundreds of protests, we’ve learned that crowd density at open-air events never exceeds two people per square meter and is often closer to 1.5. Using this knowledge, along with estimates of the area covered, we can reliably determine attendance at large, stationary gatherings,” Gubaš adds.
In Bosnia too politicians have long exaggerated the size of gatherings they organize while downplaying those of their opponents. For example, in April, the “Republika Srpska is Calling You” rally in Banja Luka was claimed by organizers and government representatives to have drawn 50,000 participants. However, the Archives of Public Gatherings from Belgrade later stated the actual number was no more than 9,000.
“Numbers are always manipulated by politicians to favor their narratives,” explains Banjaluka-based political analyst Tanja Topić. “Governments often inflate their own turnout figures to demonstrate strength while minimizing opposition numbers to portray them as weak and unsupported. For instance, during opposition protests in Banja Luka in October 2021, the police estimated 3,500 participants, but opposition sources claimed 10,000.”
The same tactic is used to discredit social movements perceived as threats. In 2018, the “Justice for David” movement formed in Banja Luka, seeking justice for the murder of 21-year-old David Dragičević. Despite its non-political nature, the government labeled it a state enemy, accusing it of attempting to destabilize Republika Srpska with foreign backing. During its largest gathering on October 5, 2018, organizers estimated attendance at 40,000, while the government-controlled public broadcaster RTRS reported only “several thousand.”
“State-controlled media serve as extensions of political authority,” warns Topić. “They inflate pro-government event numbers and downplay opposition gatherings to reinforce the illusion of societal unity under the ruling elite. This manipulation creates a cycle of propaganda, where dissent is consistently diminished and government strength exaggerated.”

Ghost Pollsters
In Montenegro, the manipulation of numbers for political purposes is most evident in the presentation of public opinion polls ahead of elections.
“Depending on which party commissioned the research, the numbers vary, and the percentages always favor the political party or group that ordered the survey,” explains Biljana Matijašević, a political journalist at ND Vijesti.
Darvin Murić, editor-in-chief of the fact-checking platform Raskrinkavanje.me, highlights that fake public opinion polls are a frequent tool for political manipulation, often designed to persuade voters with messages like, “We are the strongest; our opponents are performing poorly.”
“This type of manipulation is quick and cost-free,” Murić explains. “Arbitrary figures are presented as party or candidate ratings, a fabricated research agency name is attached, and the information is then disseminated through social media and sympathetic media.”
As a notable example, Murić cites a fictitious research agency allegedly based in Belgrade, the so-called Center of Independent Researchers (CSI). This agency regularly releases “polls” about elections in Montenegro, despite there being no records of CSI in the business registers of either Serbia or Montenegro, nor is there evidence of any activities beyond its election-related claims.
Murić alleges that CSI uses specific disinformation tactics, including publishing charts under the name of reputable agencies. In nearly all its polls, CSI predicts strong performances for the Democratic Party of Socialists and the Democratic Front, the two biggest parties, along with their candidates. Unfortunately, several Montenegrin media outlets—such as Borba, Adria, and IN4S—have published results from this fraudulent agency.

Similar fake public opinion polls surfaced ahead of the recent local elections in Podgorica. One such poll, attributed to the Ipsos agency on behalf of the New Serbian Democracy (NSD) party, contained inconsistent data regarding both the sample size and the results. Ipsos’s executive director, Vladimir Raičević, denied that the agency had conducted any such survey.
Another poll, published on Facebook, was allegedly conducted by methodology expert Prof. Dr. Miloš Bešić. Accompanied by charts showing party ratings, it was presented as “authentic polling.” But Bešić told Raskrinkavanje.me he had no connection to the data.
A third poll, published on the CGinfo portal, claimed to have been conducted on behalf of the Democratic Party of Socialists (DPS). This poll suggested that the candidate leading the Europe Now Movement–Democrats list was the frontrunner for mayor of Podgorica. DPS MP Andrija Nikolić dismissed the poll as fake.
Manipulating crowd numbers and polling results sows confusion, leaving audiences unsure of whom to trust. This erodes confidence in public opinion research, particularly pre-election polls.
“Many believe these polls are commissioned, with results tailored to favor those who pay for them,” Topić notes. “Such tactics foster a winning image to sway undecided voters. Post-election, all parties declare victory by selectively interpreting results to highlight their strengths. Electoral engineering, political corruption, and clientelism are skillfully hidden within this system.”
Beyond election-related polls, Darvin Murić also points to manipulations surrounding the recent population census both before and after the census. “Political and nationalist agendas led to the release of false information about the results early on, and later the actual results were presented in a manipulative way. We must also not overlook the ongoing manipulation of numbers regarding the victims of past wars,” Murić explained.
Selective Figures
In Kosovo, the past three years have been marked by violent protests in Serb-majority municipalities, mass resignations of Serbian officials from public institutions, boycotts of local elections in these areas. In 2023, Kosovo faced EU sanctions, with the government of Albin Kurti frequently accused of straining relations with both the EU and the U.S.
The government framed its measures in these municipalities as efforts to assert national sovereignty. These measures included closing parallel Serbian institutions, banning the use of the Serbian dinar, and encouraging the opening of businesses owned by ethnic Albanians in these areas.
To counter growing criticism, the ruling party published a graphic on its official Facebook page in February, claiming that its efforts to combat crime had improved security perceptions among Kosovo Serbs. The data for these claims was sourced from the UNDP’s Public Pulse survey, conducted biannually. The survey addresses socio-economic, political, security, and environmental issues. The ruling party highlighted data from the most recent report, published on February 28, which showed that 37.6% of Serbs were satisfied with security conditions, compared to 33.8% in April 2023 and 15.3% in November 2022. These comparisons were presented as evidence of significant improvement in security perceptions among Serbs in Kosovo.
However, the Hibrid.info fact-checking platform found that the ruling party selectively cited only the three most recent reports (November 2022, April 2023, and November 2023), excluding data from earlier years that contradicted its claims. Notably, the data presented as “2024” was actually from November 2023, as indicated in Public Pulse 25.
UNDP reports tell a different story. In 2018, only 4.8% and 10.4% of Serbs felt secure. By 2019, this figure rose to 16.2% and 27.4%. In 2020, it climbed further to 47% and 46.2%, and by the end of 2021, it reached 50.9%. However, under the current government, security perceptions have declined significantly, dropping from 46.1% in early 2022 to 15.3% by the end of that year.
Fact Checkers to the Rescue!
Across the region, numerous media outlets and individuals are engaged in fact-checking efforts. In Bosnia, platforms like Istinomjer and the website Raskrinkavanje regularly debunk disinformation. However, their reach and influence remain significantly smaller than the actual sources of manipulation.
“Politicians in Bosnia receive significant attention from mainstream media, which often uncritically transmits their statements without additional verification. Furthermore, some media outlets operate directly in favour of specific political groups. For example, Radio Television of Republika Srpska (RTRS), technically a public service broadcaster, serves the interests of a political party and never publishes corrections. By contrast, the state-level public broadcaster BHT has been relatively responsive, though this depends on its management at any given time,” said Dalio Sijah, editor-in-chief of Istinomjer.
In Montenegro, the approach to reporting and verifying information varies significantly, reflecting the country’s highly polarized media landscape, according to Marko Banović, an analyst at the Digital Forensics Center.
“Even the deconstruction of manipulative content is often selective in Montenegrin media,” Banović explains. “Outlets aligned with certain political parties frequently disseminate manipulative information without critical examination. Meanwhile, those attempting to expose such content often overlook manipulation by political actors they support. This exacerbates societal polarization and erodes public trust in the media.”
In Kosovo, the media play a prominent role in scrutinizing government claims. Fact-checking platforms such as Hibrid.info, Kryptometer by Kallxo.com, and Reporteri.net verify government statistics. However, these platforms often lack the reach of the government’s narratives. Mainstream media also address these issues through TV programs and follow-up articles that debunk government statements, particularly on topics like economic growth and employment figures. Still, fact-checkers and investigative journalists face challenges in accessing reliable data, particularly on general population statistics.
In Albania, civil society organizations like Faktoje and BIRN Albania work diligently to debunk false information. However, the fact-checking process takes time, often delaying corrections due to the slow release of public information. Journalists sometimes wait weeks for data that should be readily available.
A lack of mainstream media support worsens the problem. Debunked claims published by civil society organizations often fail to gain traction in the mainstream media, even in outlets that originally spread the false information. Albanian mainstream media, largely controlled by businesspeople with political ties, frequently act as “spokespersons” for political parties—a concern highlighted in European Commission progress reports. As a result, these outlets rarely correct false information, effectively becoming sources of disinformation themselves. This dynamic makes it increasingly difficult for civil society organizations to disseminate accurate information to the public.
In Serbia, outlets such as Fake News Tragač, Istinomer, and Raskrikavanje, along with investigative journalism centres like CINS and KRIK, tirelessly debunk disinformation churned out by the government and other actors.
Although it is often noted that the reach of fact-checking operations in the Western Balkans is limited, it may be more appropriate to consider what these media landscapes would look like without them. The fog of disinformation and propaganda in these societies would undoubtedly be much thicker. It is also worth noting that the influence of fact-checking outlets has grown steadily over the past decade. Their readership often surges—sometimes dramatically—during moments of crisis or when glaring factual uncertainties surround a particularly contentious issue. This growth underscores their increasing relevance in society.
Read the first part of the article The Art of Deception (Part I): Cherry-picking, Spinning and Inventing Numbers