Lioness hunt in Berlin
Life in southern Berlin took an unexpected turn on 20 July, 2023, thanks to the hunt for a lioness. Just outside of the city borders, the local police received notice of a smartphone video clip showing a large animal, possibly a lioness, half-hidden behind bushes on the edge of a forest. Two police officers sought out the producer of the video clip – who had initially believed that the animal was a wild boar – and reported back to their headquarters that the video’s origin was valid and that it was indeed a possible sighting of a lioness. The local authorities turned to action. Warnings to the population were issued. The hunt started, with drones, helicopters, armored vehicles and countless police. Most other outdoors activities moved indoors. Special ammunition was procured. In the general flurry, only a handful of public voices expressed doubts about the actual existence of a lioness – doubts that, alas, turned out to be justified. About 36 hours after the animal’s sighting, forensic evidence confirmed that it was indeed a wild boar (Bückmann et al., 2023).
The belief error underlying this frenzy is an example of “base rate neglect” (Kahneman and Tversky, 1973): decision makers are often too credulous of a new signal, relative to how informative this signal is in comparison to the knowledge that existed prior to receiving the signal. All Berliners know that the chance of meeting a lion is tiny. A large animal that one happens to spot in Berlin’s periphery is much, much, much less likely to be a lion than a wild boar: no lion was known to roam around Berlin during the past few hundreds of years, whereas wild boars are ubiquitous in the area. A signal with contrary evidence would have to be extremely precise in its indication that the animal is a lion – far more precise than the video clip – to serve as a justification of costly security measures. Nevertheless, the authorities alarmed hundreds of thousands of people, effectively restricting their lives, and many people believed them. Prior knowledge about the frequency of meeting lions (the base rate) was essentially ignored.
The belief in the existence of the lioness also illustrates an interesting connection between base rate neglect and communication. In this episode, like in much of everyday life, conversations shaped beliefs. The video was widely available, but its interpretation was up for discussion. The local authorities made their assessment after receiving the report from the police officers who visited the video owner’s house. The public, in turn, made its assessment on the basis of receiving the warnings from the authorities. Somewhere in this short chain of Chinese whispers, the base rate was lost.
Research in behavioral economics and in linguistics describes a related phenomenon: conversational listeners, when interpreting a statement, often ignore the relevant context (see the scholarly discussions in Sperber and Wilson, 1995, Yus, 1999, Eyster et al., 2018, Enke and Zimmermann, 2019, or Chapter 5 of Weizsäcker, 2023). Context ignorance may distort the listener’s updating in many different ways and, in particular, it may make a statement appear much stronger than it would be if seen in light of its context. An underlying reason for this effect is that the statement is salient, whereas the context is not: hearing a dramatic utterance like “a possible sighting of a lioness” makes it easy to forget that the chance of such an event is tiny. Another reason is that a conversation, as a mode of information transmission, may induce base rate neglect. The listener cannot know whether or not the talker has already accounted for his prior knowledge. That is, it is unclear whether statement reflects a talker’s posterior belief, or just the new bit of information that he wants to share. If the listener fails to understand that a message is only about the new bit, then her inference is bound to be distorted. Just like a base rate neglector overemphasizes the signal, a context-neglecting listener overemphasizes the message, perhaps even to an extreme degree.
Who is to blame? Mostly, a listener cannot know what is on the talker’s mind, so it is the talker’s responsibility to be clear in his utterances. In the present case of the lion hunt, multiple talkers were active in the communication chain – the police officers who visited the video’s owner, and the authorities – raising the question of who exactly should have been clearer. The answer is that the authorities, not the police officers, have to take the blame. After all, the political administration and the police chiefs are responsible for the most widely distributed communication: the warnings to the public.
We note that the above discussion, about priors and posteriors in communication, makes the assignment of responsibility to the authorities, and away from the police officers, even stronger. This becomes evident when seen through the lens of a very simple example by Gilboa (2015), on the question of whether or not one should account for one’s own prior in a statement: A juror sits on a judicial committee and votes to acquit an accused murderer, on the grounds that the evidence is insufficient to prove the absence of reasonable doubt. Yet, when the juror comes home in the evening, he implores his daughter not to go on a date with the same accused person, believing that this person was, in fact, guilty of murder with high probability. The justification for supporting both of these statements by the juror is that the judicial system, for good reasons, aims for rulings that ignore the prior beliefs of the jurors. In private life, however, it is normatively good for the juror to use his prior beliefs.
In the case of the statements about the lioness, we observe the same dichotomy. Like the juror in his juror role, the two police officers did their job and reported about the validity of the evidence at hand. And, like the juror in his private life, the authorities would have been well advised to factor in their prior knowledge when considering the issue of a warning. The public was right to assume that the authorities accounted for the rarity of lions, in their public statements. The authorities, quite simply, failed to do so.
Author: Georg Weizsäcker
Bückmann, Anna Kristina, Sebastian Leber, Julius Geiler and Christian Müller, 2023. Rekonstruktion der Löwenjagd: Wie kam es zur Sommersensation, die keine war? Der Tagesspiegel, 20 August, 2023.
Enke, Benjamin, and Florian Zimmermann, 2019. Correlation neglect in belief formation. Review of Economic Studies86(1), 313-332.
Eyster, Erik, Matthew Rabin and Georg Weizsäcker, 2018. An experiment on social mislearning, CRC TRR 190 Discussion Paper 73.
Gilboa, Itzhak, 2015. Rationality and the Bayesian paradigm, Journal of Economic Methodology, 22:3, 312-334.
Kahnemann, Daniel, and Amos Tversky, 1973. On the psychology of prediction. Psychological Review 80(4), 237-251.
Weizsäcker, Georg, 2023. Misunderstandings: False beliefs in communication. Open Books Publishers.
Yus, Francisco, 1999. Misunderstandings and explicit/implicit communication. Pragmatics 9:4, 487-517.