The ongoing novel coronavirus crisis is an opportunity to reflect on how we respond to risk and uncertainty, as well as how governments should communicate risks, in an environment of uncertainty and incomplete and imperfect information. Behavioural scientists have long contended that people often find it very difficult to think in statistical or probabilistic terms, so have highlighted a number of ways in which people’s responses and behaviours depart from what rational choice models predict. Coronavirus: what Xi fears most is people turning on the party Consider this thought experiment, which is sometimes known as the “Linda problem”. Linda is single, outspoken, and deeply engaged with social issues. Which is more likely: A) Linda is a bank manager; or B) Linda is a bank manager who is active in the feminist movement? Studies have found that most people instinctively choose option B. But a moment’s introspection should reveal to most of us that option B is a subset of option A, and therefore option A is the correct answer. The question – conceived by the Economics Nobel Prize laureate, Daniel Kahneman – and others like it show that we do not usually think in statistical or probabilistic terms. Instead, we rely on stereotypes and vivid stories drawn from our memories, our gut instincts, and our emotional reactions – known as the “affect heuristic”. But these are often unreliable, especially in novel and unfamiliar situations. Also, if a scientific or technical analysis does not align with our emotive reactions, instincts or subjective values, we tend to reject the technical analysis, and even find ways to discredit it. Kahneman used the Linda problem to illustrate the availability heuristic, a mental short cut that we use to judge the likelihood of something based on the ease with which it is brought to mind. This heuristic predicts that we form fuller, consistent stories based on the limited available information at hand. In this instance, the idea of Linda being just a bank manager is hardly a good or consistent story based on what we know about her. She being active in the feminist movement makes for a better, more consistent story – this is why we are instinctively drawn to option B. To illustrate the point, consider how you would have responded to the question if you were not told that “Linda is single, outspoken, and deeply engaged with social issues”. You would have had no problems choosing option A. In this instance, giving people more information – especially irrelevant information – led to poorer decision-making. So how does this relate to coronavirus, and the way people think about the risks of a new infectious disease? First, people tend to underreact before the outbreak of a new infectious disease and overreact when the outbreak occurs: they are complacent before an outbreak and too fearful during an outbreak. This is not because people are irrational. It has to do with the fact that before the crisis, the risks of an infectious disease are underestimated because they are not salient. In this regard, we tend to view the risks of an epidemic before one has occurred in the same way we view the risks of climate change: both types of risk are systematically underestimated because their impacts are distant and aren’t immediately felt. This saliency bias leads to probability neglect, disaster myopia, and a naive optimism that if an outbreak of an infectious disease does occur, we would have the governmental, medical or technological solutions to deal with it. When an outbreak occurs, the risks of an epidemic suddenly become very salient in people’s minds and they may react to it with a degree of fear that is not proportionate to the actual risks or harms of the disease. Whereas people are complacent about, and even ignore, the warnings of health authorities at the start of every flu season, their mental energies are now excessively focused on the most salient issue of the day. Government pronouncements of war against a “deadly” virus, media reports of the rising number of infected people and deaths, and the 21st century amplification engine of emotions that is social media all combine to create more fear and socioeconomic disruptions than what the actual risks of the disease suggest. Uncertainty in the early stages of the crisis, over how contagious and fatal the disease is, also plays a big part in creating a self-reinforcing cycle of fear. Second, we rely on our selective and error-prone memory system when thinking about how the future might evolve. There is a biological basis for this: the parts of the brain that are engaged in memory recall are the same parts that are used to imagine the future. What does this have to do with the current crisis? Everything. Because in East Asia, people draw heavily on their traumatic memories of the severe acute respiratory syndrome epidemic in 2003 to imagine how bad this current outbreak might be. Sars has been mythologised to be a terrible pandemic. It was a crisis that was seared into the consciousness of most East Asian societies such that anything which recalls Sars evokes the same reactions of fear and panic. As Dr David Baltimore, a Nobel Prize winner in medicine, commented during the Sars crisis: “People clearly have reacted to it with a level of fear that is incommensurate with the size of the problem.” By the standards of pandemics, Sars was not that contagious or deadly. Worldwide, more than 8,000 people caught it and it did have a moderately high fatality rate: among those who were infected, around 9 per cent (or nearly 800 people) died. Compared to the Spanish flu of 1919, which scientists believed killed 50-100 million people worldwide, the Ebola virus, HIV, and the H1N1 “swine flu” epidemic of 2009-2010 that killed 245,000 people worldwide, the morbidity and mortality rates of Sars were modest. Every year, many more people in Hong Kong or Singapore die of pneumonia than the number of people that Sars killed in either city. Third, Kahmenan’s Prospect Theory predicts that people react to the low probability loss events, such as the low risk of dying from coronavirus, with a great deal of dread and risk aversion – more than what rationality would prescribe. This means that people are willing to pay a very high price to avoid or eliminate that low risk, much like how people pay too much for insurance against low-risk catastrophic events - especially after one has occurred. This may also be why societies are willing to bear the enormous costs of trying to contain the novel coronavirus. Fourth, from an evolutionary perspective, we are primed to fear animals that cause death and disease – such as snakes, rats, spiders, and bats – even though the number of deaths they cause today is much lower than modern killers such as motor vehicles and guns. (This is one of many examples of how our brains have not evolved as quickly as technologies have.) We are therefore more likely to overreact emotionally to animal-borne diseases than to the risks posed by guns and cars, even though the latter cause far more deaths today. Fifth, there is always a social element to how we respond to new infectious diseases. Far from being the individual utility maximisers that we find in standard economics textbooks, an epidemic often reveals how our behaviours are more the result of social norms and beliefs, and of our herd mentality, rather than individual assessments of risk and cost-benefit analyses. This has given rise to panic buying of face masks and even toilet paper. Communicating risks What do these behavioural insights imply for how governments should communicate risks in an environment of uncertainty and imperfect information? As a general rule, governments should not view or portray people’s fears and emotions as irrational. People’s psychological responses – even if they are driven by emotion rather than facts and statistical reasoning – are the products of evolution. It is neither realistic nor possible to fight centuries of evolution, especially given the uncertainty in such situations and the real possibility that the facts may change. Instead, governments should acknowledge, and empathise with, people’s fears rather than dismiss them. They should also be upfront about what they know and don’t know about the disease, and they should communicate and share that uncertainty with the public. All this is difficult to do, especially for a Hong Kong government that suffers from a lack of trust from the populace. But the alternative strategy of trying to reassure the public by downplaying the risks of a pandemic is even more likely to reduce trust, especially when the statements of reassurance turn out later to be wrong. Acknowledging and empathising with people’s fears also means that rather than aiming to eliminate fear altogether, governments should tell their people what to expect and what they should do. This is especially important given that this crisis may last many months; it may even be a recurrent one. In the case of Sars, governments in the region got lucky. The disease went away during the warm months and never came back in the following flu seasons. We may not be so lucky with 2019-nCoV. On the other hand, it appears that while this coronavirus is more transmissible than Sars, it is much less fatal. Across China, the mortality rate of the disease is around 2 per cent; outside Wuhan and Hubei, the mortality rate is below 0.2 per cent. Furthermore, the vast majority of those who have died from the virus were over the age of 60 or had pre-existing conditions. Can Asian economies survive the coronavirus? The rising number of infections, combined with the disease’s relatively low mortality rate, points to the real possibility that China’s containment strategy based on the Sars experience may not work. If so, it may eventually shift from a strategy of containment to one of mitigating the impacts of the disease. Mitigation would mean focusing the efforts of the health care system on reducing the case mortality rate while accepting that there would be community spread of the virus. If mainland China were to do this, it is hard to imagine that Hong Kong can persist with a containment strategy. This is not a fatalistic or defeatist attitude; it is simply recognition of the fact that containment does not work, that we overcame Sars partly because we got lucky with its epidemiological characteristics – it was not that contagious and patients were infectious only after they showed symptoms – and an acceptance that the costs of containment and its disruptions to daily life are simply too high relative to what it can achieve. A decision to switch from containment to mitigation may be very unpopular. The less trust a government commands, the longer it has to persist with a containment strategy, and the higher the economic price it has to pay. Many people also want their governments to take, or be seen to take, the toughest, most aggressive containment measures even if those measures do not work. They focus on the losses – infections and deaths – avoided as a result of containment, while ignoring the potentially much larger losses of other more prevalent medical conditions that are not treated as a result of health care resources being diverted to deal with the novel coronavirus, the losses in economic output, and the disruptions to normal life. It is still too early to declare that the containment strategy in China has failed. At this point, any official who dares to say “let’s learn to live with this virus” would be driven out by the fearful masses. But while it is not time yet to declare that containment has failed, it is not too early for governments to start planning for mitigation and to prepare their populations for it. This begins with communicating the possibility that unlike Sars, this novel coronavirus does not just go away and that we have to think about how we lead our lives as normally as possible – while taking the public health and personal hygiene precautions to minimise the spread of this virus – in the knowledge that the virus may never be eliminated. In short, governments may have to start communicating that this novel coronavirus is here to stay, that it is not like Sars but much more like H1N1, and that we have to adapt to it as such. Donald Low is Professor of Practice in Public Policy at the Hong Kong University of Science and Technology, and director of the university’s Leadership and Public Policy Executive Education programme. He is also the editor of Behavioural Economics and Policy Design: Examples from Singapore , published in 2011.