Tom Record writes that media and political framing of Covid-19 has often been at odds with the science and considers the cognitive biases that make focusing on the data so difficult in a crisis.
Reading time: 15 minutes
This article featured in our Summer Journal.
When the news broke on 11 March that Tom Hanks had contracted Covid-19, many people in the US had been expecting the country would be spared from an outbreak of the scale seen elsewhere in the world(1). Hearing he had succumbed awoke many people in America to the reality of the pandemic. That Hanks was on the other side of the world in Australia, and was just one of tens of thousands of people already infected, was irrelevant. He was a well-liked celebrity that many people felt close to. If he could get ill, what chance do the rest of us have? When the dust settles, this might not be viewed as a key turning point in America’s response to the coronavirus pandemic. However, as an anecdote, it offers an insight into the role behavioural biases, emotions and storytelling have played in shaping our understanding and reaction to the virus – often at odds with the science.
Framing Covid-19 has often been more art than science
Emotions strengthen behavioural biases and today’s media is tuned to strengthen our emotional response. Stories about individuals that we can relate to and that resonate with us have a much more powerful influence than crisp scientific facts, or even worse – caveated scientific hypotheses. This is something that the media and effective politicians exploit ruthlessly. Journalists are judged by how often their stories are shared, how many views they receive online and how many comments they garner, ultimately encouraging extreme, controversial reporting. In the recent Covid-19 crisis, we have seen this manifested in the way the media has emphasised rare cases that are easier for a large part of the population to connect with emotionally, such as young and previously healthy people dying. The only scientific models reported on are those that predict huge numbers of deaths (but only in their worst case scenarios – a fact that is conveniently forgotten in the headlines). These capture clicks and, unfortunately, have been the daily staple of journalists looking for ‘engagement’.
Linguistic analysis by a team at Reading University(2) highlights how emotive some of the mainstream reporting has been, especially in the US and the UK where the emotional ratchet has been tightened by the use of military analogies. At around the same time that Tom Hanks’s plight was giving Americans pause for thought, the Daily Mail was attempting to invoke a wartime frame of mind with its 11 March headline: “Can Boris Johnson conjure up the spirit of the Blitz?”(3) Examples of a somewhat less sensationalist ilk can also be found in The Guardian and The New York Times. The German press, in contrast, has tended to use relatively neutral language. Words such as “inform”, “investigate” or “test” are more prevalent, befitting scientific discourse. As the authors of the Reading University study note, “language reinforces reasoning and influences how people act and behave.”(4) In this case, the authors believe that a sensationalist war-like framing in the US and UK media has come at the cost of more thorough scrutiny of the nature of Covid-19 and associated data. The differences in reporting style have been mirrored by starkly different levels of government effectiveness in the face of the pandemic. Many commentators have noted that Angela Merkel’s scientific background has likely been a key reason why her policy and communication response has been superior to that of her counterparts in the US and UK. “Keep Calm and Carry On” might have been a great motto when bombs were falling in London. However, when the intention is for everyone to drop everything and stay at home, it is clear that the messaging should be closer to the opposite of what was invoked during the Blitz.
Distortions in the coverage exemplify a phenomenon described by behavioural scientist Thomas Gilovich as “sharpening and leveling [sic]”(5) which is particularly evident in second-hand retellings of a story. The motivation for sharpening and levelling is inherently linked to the goals of the speaker (e.g. to sell the news or a political agenda), and involves suppressing ambiguous or un-newsworthy details and amplifying those aspects of a story most likely to gain attention.(6) Studies have also shown that second-hand accounts tend to be simpler, with sharpening and levelling occurring with each subsequent iteration, thus creating further distortions. Notably, human-interest stories typically focus on the people involved rather than the context in which an action might have taken place(7). News that Tom Hanks had contracted coronavirus was a prime example of this, as has been the nightly focus on individual virus victims on the BBC news (which has obviously been distressing for the families of those directly affected). Overall, the process of elevating certain elements while de-emphasising ambiguous or inconvenient facts tends to “distort the accuracy of information we receive second-hand, and thus bias some of the most important information upon which we base our beliefs.”(8) Ambiguous outcomes are par for the course in scientific endeavour. The purpose of science is not to find neat solutions that fit assumptions but to test hypotheses and advance understanding. Properly reported and caveated science does not necessarily make for exciting news articles. The science around Covid-19 is no exception.
Indeed, some governments have openly looked to exploit emotional responses to further their agenda. The UK government clearly demonstrated this when considering its options to increase social distancing: a recently published internal document stated that, “The perceived level of personal threat needs to be increased…using hard-hitting emotional messaging.”(9) Meanwhile, SAGE, the UK health advisory panel, directly advocated the use of the media to “increase [the] sense of personal threat.” (10) In short, our emotions have been engaged by the impact of the coronavirus (through physical isolation, change to routine, fear of infection for oneself, friends and family), and those fears have been amplified by government media messaging in an effort to increase the chances of lockdowns being adhered to. We certainly do not wish to diminish the scale of the risk to human health or the urgency of the policy response to the Covid-19 pandemic, but how statistics are represented matters.
Message received, loud and (un)clear
Thus far I have mainly focused on how messages have been constructed and distorted. How we receive these messages is of equal importance and cognitive biases have a particularly strong impact on our ability to interpret the complex data behind the coronavirus pandemic. Covid-19 has brought statistics into people’s living rooms in a highly unusual and emotive way. Since the start of the pandemic, many governments around the world have held daily press briefings run by a mix of politicians and medical advisers who have explicitly couched policy decisions in statistical terms. The outbreak has also spawned many an amateur statistician, virologist and pathologist – subjects the human brain is not hardwired to understand intuitively in the same way it is, say, to grasp the rules of grammar, the building blocks of language and storytelling. As someone who studied these subjects as part of a degree in Natural Sciences, I can attest to their complexity (with the caveat that I am a sample of one). In contrast, the strength of our innate linguistic skills is clearly demonstrated in studies of second-generation speakers of Pidgin English, who have been shown to intuitively develop a grammatical framework and create a more complex and nuanced language than that spoken by their parents.
Indeed, a new research paper by behavioural scientists Traci H. Freling et al. suggests that at times of crisis, people tend to suffer from the “anecdotal” bias, trusting stories rather than statistics when determining what actions we should take.(11) Ordinarily people tend to take a more rational view of statistics and they have greater persuasive power than anecdotes, so times of heightened threat, such as today, present a particular behavioural anomaly. It is not just the general public that suffers from this bias. The paper concludes “factors which heighten emotional engagement render decision makers more susceptible to the anecdotal bias.”(12)
Somewhat ironically, given our earlier points about the media framing the virus in war rhetoric, the Blitz provides an interesting historical example of how anecdotes can lead to a spurious and potentially dangerous understanding of risk in a highly emotive context. During WWII, Londoners believed German V-1 buzz bombs and V-2 rockets were landing in clusters in certain areas of the city, making these locations more dangerous to be in, when in fact these bombs fell in random locations and many failed to reach London at all. Nevertheless, these anecdotes gained prominence and influenced perceptions about which parts of the city were safe. This is an example of the cluster illusion, borne out of a human tendency to detect order where none exists, and also demonstrates the power of anecdotes in times of distress(13).
Thinking more broadly, all of us have ingrained biases that have evolved over many centuries to protect us. Perhaps the best known of these is the flight or fight response. When threatened or under stress, you naturally get a jolt of adrenaline, the more primitive limbic system of the brain takes over, and you prepare to fight or run. When running, the instinct is to join the pack, to follow the rest and not be left straggling as easy pickings for predators. The amygdala is the key part of the limbic system that is responsible for emotions – experiments on monkeys in the 1930s showed that removing their amygdalae left them fearless, hypersexual and either very aggressive or immensely passive. The aspects of the brain that effectively shut off at times of fight or flight are those that act best when we are in a positive frame of mind – they are responsible for us seeking new experiences, being creative, planning ahead and being able to adapt more easily to change.
Keeping calm, collectively
Clearly for long-term investors, avoiding a heightened emotional response that is activated by stress is a positive. One of the advantages of working in a team of independently minded individuals, is that different things will stress each of us differently. We saw this when the crisis hit – a range of individual responses from excitement at being able to add to existing holdings cheaply, to caution stemming from thoughts about second and third order effects to everything in between. What has helped us to navigate through the crisis has been a strong sense of teamwork, open dialogue informed by differing opinions, and acceptance that there is a lot of uncertainty that we can only estimate. We have tried to keep an open mind, while benefitting from preparation (such as a November 2019 team discussion of central bankers’ responses to the next crisis), and a consistent, rigorous process based on individual companies’ fundamentals and the potential for returns to be skewed in our favour. Above all, we draw strength from the sense that we are in this together. The culture at Majedie is something that we are proud of and is the ultimate enabler of our success. We aren’t distracted by office politics or trying to appear right in front of our colleagues – we just want to get the best ideas we can find into the portfolio, ensure that the portfolio remains diverse, and ultimately generate the best risk-adjusted returns for our clients, in a responsible manner.
When clouds of stress are billowing from the market and the media, and government is fanning these into tornadoes of emotional turmoil, the ability for a team to be calm, thoughtful, open and willing to back their judgement sets a good active manager apart from the rest. We hope that our newsletters in recent months have given a flavour of the kinds of things we think about, and more importantly, how we think during a crisis. Looking out to the horizon, we believe that the fruits of the decisions we have made in recent months should benefit our clients’ returns over the long term.
(5) Thomas Gilovich’s How We Know What Isn’t So: The Fallibility of Human Reason in Everyday Life page. 90
(6) Thomas Gilovich’s How We Know What Isn’t So: The Fallibility of Human Reason in Everyday Life page. 90
(7) Thomas Gilovich’s How We Know What Isn’t So: The Fallibility of Human Reason in Everyday Life page. 92
(8) Thomas Gilovich’s How We Know What Isn’t So: The Fallibility of Human Reason in Everyday Life page. 90
(11) https://www.sciencedirect.com/science/article/pii/S0749597819301633″>When poignant stories outweigh cold hard facts: A meta-analysis of the anecdotal bias”, Traci H. Freling, Zhiyong Yang, Ritesh Saini, Omar S. Itani, Ryan Rashad Abualsamh, Pages 51-67; https://doi.org/10.1016/j.obhdp.2020.01.006
(12) https://www.sciencedirect.com/science/article/pii/S0749597819301633″>When poignant stories outweigh cold hard facts: A meta-analysis of the anecdotal bias”, Traci H. Freling, Zhiyong Yang, Ritesh Saini, Omar S. Itani, Ryan Rashad Abualsamh, Pages 51-67; https://doi.org/10.1016/j.obhdp.2020.01.006
(13) Thomas Gilovich’s How We Know What Isn’t So: The Fallibility of Human Reason in Everyday Life page. 20