The "Irrational" Failure of Crisis Leadership

David De Cremer

It’s hard to see a crisis coming. Crisis situations usually arrive at moments we do not expect it and as a result, we are ill-prepared. The corona virus, which has now resulted in the COVID-19 pandemic, took China by surprise and forced the country to take drastic measures whereby entire cities became confined and isolated. Shortly after China felt ready to communicate the spread of the virus publicly, their Asian neighbours were immediately confronted with the virus threat. Some of them responded in very effective ways, which means that in the meantime, Hong Kong, Singapore and Taiwan have become the world’s most favorite examples on what to do in the first instance when the corona virus enters your shores. However, the virus did not harbour just in Asia, but moved on to impact the rest of the world. An important observation is that at the moment the virus moved on to Europe, its status also changed. Put simply, the corona virus moved from being an unexpected event to an expected one. So, Europe in a way was confronted with a crisis that was more or less announced. Nevertheless, what we observed in the meantime is a continent that initially acted as if the coronavirus crisis came out of the blue. As if it was something that one did not see coming. How is this possible? And, what does it say about how the response to an announced crisis was led. Indeed, today, many articles are being written about how important leaders are in managing crisis situations and all these insights are being translated into the context of the present coronavirus outbreak. The main conclusion of all these pieces is that crisis requires effective leadership! So, what happened to leadership when the coronavirus hit the European continent?

Today we know that Europe has become the epicentre of the coronavirus – soon to be followed up by the US – with Italy and Spain being hit the hardest. With a sense of disbelief, the world witnessed Italy being hit by a coronavirus tsunami that started at the end of February and resulted in a complete lockdown where no movement of people was permitted at all. Being aware that the virus was coming and that borders with Italy were not being closed, the world also saw many European countries falling prey to an initial kind of inertia where the risk of a quick and escalating virus spread was underestimated. It seemed like no sense of urgency existed to take hard measures like Singapore, Hong Kong and Taiwan did. This was well reflected in the fact that no need for immediate and broad screening of people coming from infected areas – notably Italy and its skiing resorts – seemed to exist. Indeed, in early March, newspapers in different countries reported that there were some identified corona virus cases, but that the circulation of the virus was limited. Limited indeed, until their fellow countrymen came back from the skiing resorts in Northern Italy. Disbelief was uttered when people could freely fly back into their own country without being adequately screened or asked to go into quarantine, even if one did not show any symptoms yet (something a country like Singapore did almost right away). So, although many European countries knew that soon many of their people would return from the skiing resorts in Northern Italy, no precautions were taken to avoid this influx of people to mingle with the rest of the population. Why did this happen and what were the consequences of this decision?

Financial inertia in times of crisis

In my view, this first response, or better, lack of response, created the perfect conditions for what I call the phenomenon of “financial inertia” in times of crisis. This phenomenon refers to the tendency of leaders in times of crisis to avoid spending money in the early stages of the crisis with the result that in the latter stage of the crisis they will be forced to overspend and create economic turmoil that could have been avoided if spending happened earlier. It is thus an irrational kind of behaviour that on the longer term will turn out to be more costly than needed. Let’s return to the example of people coming back from North-Italian skiing resorts. Although there was some awareness that people coming back from an infected area could accelerate the rate of infections within the country, it was only treated as a nuisance, but not as a real emergency. The primary argument was in fact a financial one. Indeed, the virus was not very visible yet (remember only a limited number of cases were known at that time), and so it was reasoned that the costs that would have to be made to screen and test all these families returning were too high in light of the idea that the virus was not much more than an unknown flu. In other words, it was considered unreasonable to hit the entrance of a virus with a financial sledgehammer while we did not even know how severe the virus itself was. This is where leadership in Europe took a different approach than in countries like Singapore. In the latter country, fast tracking of infected people and their social environment was key in containing the first outbreak and assured that hospitals would not be overpopulated and thus able to provide high-quality care.

Fast forward, and now we see that only in the span of a few weeks, the European continent is in chaos and consensus exists that COVID-19 is recognized as a “once in a lifetime” health crisis. Most governments have in the meantime decided to implement harsh measures and decide on (semi)-lockdowns. Even more, and particularly interesting from our perspective, the minds of governments are ready to make significant costs to financially support the unemployed, companies suspending work, and the healthcare system. No more careful considerations whether the financial costs are too high, rather, the opposite mindset now prevails, which is that limitless spending is needed and supported. From nothing to everything. The ECB President Christine Lagarde even said in a tweet, "Extraordinary times require extraordinary action. There are no limits.” This change in investment behavior falls together with the realization that (a) the economic impact of the coronavirus may be three times that of the financial crisis in 2007-2008, and (b) that an uncontrolled escalation of the virus would overwhelm the health care system, which, in turn, would result in a higher number of deaths and financially disrupt society. So, now the consequences of the crisis are visible, suddenly leaders do not know any financial constraints anymore and costs are allowed to raise to unknown levels.

But, wait a minute, could these expected financial consequences not be less severe and impactful, if some serious financial investments were made earlier on? Isn’t it the case that medical experts are now saying that if significant investments would have been made earlier, travellers could have been checked and quarantined and thus slowed down the spread of the virus? If that would have happened, the curve would have flattened more quickly and more resources would have been available to manage society. The ugly truth, however, is that human nature is complicated and only when a crisis is experienced in reality, or, in other words, when the consequences of the announced crisis reveal themselves, then people seem willing to take action. But, as we have seen at that moment in time, the actions that were eventually taken will be much more expensive than when those same actions were shown earlier. The fact that we engage in such inertia indicates how irrational we are as human beings when confronted with a crisis. In fact, it suggests that humans by default are very bad at managing the prevention of a fully-fledged crisis, which would save more lives and be less expensive, and only move into actively responding to a crisis when the house is already on fire.

This sobering conclusion reminds me of the following story. A person walks past a restaurant where he notices that the electric wiring in the kitchen is in poor shape and may pose a serious threat to the safety of the people sitting inside the restaurant. It is clear to him that it would not take much for a fire to break out. Convinced by his assessment, this person runs into the restaurant and tries to persuade people to leave the restaurant. He tells them that leaving the restaurant now would be better for their health and survival. What is likely to happen in that situation? People will most likely think he has lost his mind and wonder what he’s on about. No one will listen to him. And, even worse, they will not consider this person to be a leader. No, he will be regarded more as a fool. But, what would happen if this story would take place in parallel universe, where the crisis situation and its consequences are more visible? Imagine that in this parallel universe, the same person is walking past the same restaurant again. But this time, due to poor wiring, a fire has broken out in the kitchen and people in the restaurant are under imminent threat. The same person now runs into the restaurant and saves half of the people sitting there from the fire. What will be their response now? Most likely they will look at this person as a leader, a hero, but definitely not a fool.

What this story teaches us is that it is difficult to move people into action when the threat is not imminent and thus not experienced in the flesh, to say so. And, this tendency displays itself very much so in leaders expected to deal with an announced crisis. Rationally speaking, one would expect that if one knows a crisis is about to happen, we would not wait to experience the consequences, but try to do as much as possible to minimize the experience of those consequences. But, this is not how human leaders are wired. Rather than investing upfront to slow down a crisis and make it manageable, they seem destined to wait for the consequences to reveal themselves and then spend in unlimited ways. The lack of preventive abilities is further demonstrated in failing to ask for the advice of other parties having experienced the crisis one is awaiting. Why is crisis leadership so irrational in the early stages of a crisis? Below, I outline a number of irrational tendencies that prevent us from learning and working with knowledge that is rationally speaking useful to make effective decisions in crisis situations.

On the irrational nature of crisis leadership

Our brain works in the most efficient way possible and this implies that it will use many short-cuts. These short-cuts are useful because they allow us to deal with a massive amount of information that we receive each day. Given the abundance of information it would be impossible to pay attention to and process adequately each piece of information. So, hooray for our mental short-cuts. At the same time, however, these mental short-cuts also may get us into trouble, and especially so under stressful and complex situations. A crisis situation fits that description and it is then also no surprise to see that our mental short-cuts lead us to making decisions that seem badly informed (e.g. letting people come back from the Italian skiing resorts without any screening) or even blind to information that we should be using (e.g. ask Singapore to share their experiences). Below, I list a number of those biases that let us make such irrational decisions.

1.   Normalcy bias: Even though the coronavirus crisis was to some extent an announced one, still it was initially perceived as not being able to disrupt our society and its systems too much. Partly this was due to the fact that the virus itself was treated as a known unknown. We knew it was coming, but what exactly it was, we were less clear about. Under such ambiguous and uncertain situations, our default thinking is to rely on a belief that not too much will change. As there is no concrete information available we resort to the tendency to belief that our society will keep functioning in ways as it has done before. As a result, leaders underestimated the true impact of the virus and refrained from taking more drastic measures in the initial stages of the virus circulation.

2.   Confirmation bias: As the virus crisis was announced, why didn’t we look more carefully at what was happening in Asia? Why did leaders not put more effort into understanding what China was communicating and how much sense their communication made, both in terms of being reliable and accurate? And, why did leaders not pay more attention to important advice of other Asian countries that seemed to be effective in slowing down the impact of the virus? One reason is that stereotypical expectations and ideas about the Asian countries may have played a role where one may have thought that China is a different country, with different rules and authoritarian systems. As such, they are not comparable to the average European country, hence, they are not a relevant comparison point to validate or disprove our beliefs and ways of working. In a similar vein, advice from Singapore could have been ignored as the island state works in ways that no European country does. So, much information coming from these countries was not perceived as confirming our own beliefs about how we live and work. Hence, it is ignored, or, at minimum not used to the extent it should have.

3.   Illusory superiority: Closely related to the confirmation bias, many leaders in Europe likely overestimated the positive characteristics of their own society in dealing with the virus and this to the neglect of the more vulnerable aspects of their systems when it came down to dealing with such a health threat. Such illusion of superiority does not motivate people to look for information, let alone to accept advice given by countries already dealing with the virus crisis. If the idea exists that we are ready for what is coming, why ask for advice?

Shouldn’t leadership managing crisis be less human and more automated?

So, yes, we could have slowed down the spread of the virus and made it more manageable, but clearly the human nature of our leadership allowed for failing to spend early to avoid overspending later. If this is the case, wouldn’t it be possible then that in the 21th century we remove human irrationalities as much as possible from the leadership equation and replace it by automated decision-making mechanisms? After all, many tech gurus express hope that the rapid development of AI technologies will facilitate the emergence of a society that works in more accurate, effective and less biased ways.

How intuitively attractive this may sound, I am not convinced that more automated decision-making as an alternative to crisis leadership as we know it today would reveal other and more effective decisions. In fact, letting AI technologies do more of the evaluating side of the equation so humans can follow up the advice of the algorithms calculating the prediction models would likely reveal similar outcomes as we have seen today. How so? The problem that we saw with respect to the failure of screening and following up those people returning from their skiing holidays actually resulted for a large part from humans’ overreliance on data. An important reason why drastic financial investments were not made immediately was due to the fact that the existing models at that point provided inaccurate numbers of how many people were really infected in North-Italy. Since these numbers were underestimating what was really happening, leaders fell back even more to their decision biases of refraining to invest big in a crisis situation that had not revealed itself in its entirety. Put simply, the data underestimated the number of infected cases, and as a result, leaders became blind to the true impact of the crisis and made the wrong initial decisions, which led to more costly decisions later on when the crisis unfolded itself.

A weakness of working with automated systems relying on deep learning principles where algorithms do the math is that the effectiveness of the system depends entirely on the quality of the data. It is the simple application of the garbage in/garbage out principle. By being over-reliant on the data provided and thus relying too much on the trust we have in our algorithmic models, leaders displayed even more biased behaviour, leading to irrational decision-making. I experienced this myself when I was about to fly out from Singapore to give a keynote in Switzerland and asked people over there whether the conference was taking sufficient precautionary measures. Their response was very simple: The Swiss authorities did not close the borders and hardly any infected cases were reported yet in the country. My common sense, however, said that a country being a neighbour of a highly infected country is something like a sitting duck, especially since I realized that the EU would not allow for closing the borders right away. Nevertheless, the organizer decided to follow the data she was told about.

As I explain in my forthcoming book “Leadership by Algorithm”, what makes AI technologies unsuitable to be leaders of humans, and thus to make decisions on our behalf, is that they lack common sense, cannot imagine other realities, and miss the sensitivity to understand the impact and meaning of policies and traditions. By simply looking at the output of the employed model, as calculated by algorithms, this time around, we assumed that the virus was not circulating as much yet in the ski areas, we acknowledged that there were not many cases reported yet in our own country, and became overly optimistic that everything was fine. Being led by machine, to say so, our irrationalities would not be reduced or eliminated and instead they would be aggravated. Our excessive data reliance made us blind and underestimate the true impact of the virus. As a result, our overreliance on data made that we engaged easily in the financial inertia that I identified earlier.

Conclusion

Going back to the restaurant example I provided earlier, we can recognize that it’s relatively easy to be a hero and seen as a leader when you save one life when the house is already on fire. Whereas when you save the same life when the fire is not visible yet, people think of you as a fool. As a result, leaders refrain from making decisions until they cannot avoid it anymore, and by then any decision will be so much more costly than the ones that could have been taken in the earlier stages of a crisis. We see now that leaders are investing without limits and are actually perceived overall as effective leaders. However, it is those leaders who can already act and decide in effective ways when the crisis is about to unfold rather than waiting until the consequences have already surfaced that the world needs most. What this teaches us is that the truly effective leaders in dealing with crisis are those who are not afraid to be labelled as a fool.

 #This piece is also published on the site of The European Business Review

David De Cremer is a Provost’s chair and professor in management and organizations at NUS Business School, National University of Singapore. He is the founder and director of the Center on AI Technology for Humankind at NUS Business school; which is a platform developing research and education promoting a human-centered approach to AI development. Before moving to NUS, he was the KPMG endowed chaired professor in management studies at Judge Business School, University of Cambridge. He is named one of the World’s top 30 management gurus and speakers in 2020 by the organization GlobalGurus and has published over more than 250 articles and book chapters. He is also a best-selling author with his book Huawei: Leadership, culture and connectivity” having sold more than one million copies. His newest book “Leadership by algorithm: Who leads and who follows in the AI era?” will be out in print in May 2020.