Much like U.S. President-elect Donald Trump, the German right-wing populist party has a rocky relationship with the media.
The party, ‘Alternative for Germany’ (AfD), recently decided to bar a range of individual journalists as well as the entire German public broadcasting sector from a congress of European right-wing populist parties. Its functionaries and supporters regularly dismiss the news media as “Lügenpresse,” a term used to denounce the media in Nazi Germany.
Party leader Frauke Petry prefers the term "Pinocchio press," which carries less historical baggage while conveying the same sentiment.
In that light, many saw the headlines in mid-June of 2016 with some degree of Schadenfreude: "Study exposes Petry as the least truthful politician in talk shows," wrote Die Welt. "Petry is the Pinocchio among talk show guests," said Zeit Online. "Petry lies the most," wrote Hannoverische Allgemeine.
But does she?
It was a good idea, in principle. Students from the Cologne School of Journalism analyzed every statement made by seven politicians on four major political talk shows over a four-month period. On that basis, they calculated the percentage of incorrect statements for each politician and published their results as Faktenzoom, an online ranking that identified Petry as the least truthful politician.
The party was quick to dismiss the ranking as yet another example of the dishonesty of the media. "If that is the future of German journalism, then we have to worry even more about this sector than we already do," Petry wrote on her Facebook page on June 15.
Her complaints were not entirely unjustified. Fast-forward seven months, and Faktenzoom is offline. In late December, a German court prohibited the school from disseminating the ranking. A defamation trial looms unless it will be changed or retracted by Jan. 23. What went wrong?
"I think our students did a thorough job," said Ulric Papendick, the director of the Cologne School of Journalism. "There were some errors, but these errors were corrected immediately and in a transparent manner."
In spite of these corrections, Petry remained at the top of the ranking. It turns out there's another thing she has in common with Trump: a love of litigation.
After Petry’s case, the higher regional court of Cologne ordered Faktenzoom to refrain from disseminating two specific verdicts, as well as the overall ranking. Key to the court's decision was Faktenzoom’s practice of rephrasing statements for the sake of clarity.
For example, when Petry claimed that an "upper limit" for refugees coming to Germany was called for within the Social Democratic Party (SPD), the formulation Faktenzoom put on its site read "The SPD calls for an upper limit." Faktenzoom rated this statement as false. But while such a limit is not an official party line, individual party politicians, such as the mayor of Regensburg, did indeed call for it.
Carsten Brennecke, Petry’s lawyer, has more qualms with the project.
"It can’t be proven that the school intentionally damaged my client," Brennecke said. "But if you look at the way the study was conducted, you get the impression that everything was done to ensure Petry would end up topping the ranking."
The Faktenzoom team rejects this allegation, insisting that they treated all politicians the same way. But while there is no evidence for intentional discrimination against Petry, their methodology has been criticized for featuring problematic elements.
The team categorized all statements as "checkable," "common knowledge," "platitude" or "uncheckable." Only statements deemed “checkable” were verified. But according to media expert Stefan Niggemeier, the criteria for this categorization were quite vague, making the selection of statements excluded from the analysis look a bit arbitrary.
For example, Faktenzoom considered the claim that "there were 2.5 million refugees in Turkey in January" by Conservative politician Armin Laschet as common knowledge. The Left’s Katja Kipping’s statement that "you can’t force a refugee to apply for asylum" was considered a platitude. Petry’s claim that "Turkey spends five per cent of the costs per refugee that we spend in Europe" was deemed as uncheckable. Moreover, a large number of statements were excluded because the Faktenzoom team didn’t find sources to either prove or disprove them.
"In our view, this artificial exclusion of various correct statements is what led to our client’s negative result in the first place," said Brennecke.
Faktenzoom’s raw data doesn’t seem to indicate that methodological choices affected Petry more than other politicians. However, the project does highlight the challenge of quantifying the truth in a comparable way across different politicians.
Faktenzoom did try to circumscribe the sample it would choose from for its analysis clearly: All claims made in talk shows over a certain period of time.
However, while not all factually incorrect statements are equivalent, they all counted the same when it came to calculating the percentages for the ranking. For example, Faktenzoom deemed false a claim by conservative politician Markus Söder that only 400 out of 16,000 refugees had been relocated in Europe by Feb. 18 — when the actual number was 583. While inaccurate, the correct number still essentially supported Söder’s overall argument that progress on the EU’s relocation scheme was painstakingly slow.
"Looking back, calculating these percentages and using a ranking might be seen as problematic," says Papendick. "But it certainly wasn’t our objective to identify or expose the biggest liar in German politics. We stated quite clearly that we don’t accuse any politician of intentional lying. The idea was just to examine how often they deviate from the facts."
Prominent media organizations ran the Faktenzoom results without any caveats about the methodology, however.
“Some media are obviously willing to spread the most outlandish nonsense in order to discredit the AfD," Petry wrote in her June Facebook post.
An ironic complaint coming from a party whose functionaries have also struggled with the facts. But not an entirely unjustified one, at least in this case.