First, the world has become more susceptible to pandemics due to globalisation. Globalisation has made it easier to move goods, services, capital and, notably, people across borders. Equally important, globalisation has increased the variety and intensity of international interconnectedness. Geographic interconnectedness and interconnectedness between various subsystems (e.g. transport, economy, health) creates significant vulnerabilities. It also creates complexity (Page 2007). Interconnectedness, of course, also creates benefits and opportunities. What is incontrovertible is that de-globalization would cause significant economic costs. A more sensible approach to avoiding or managing future global public health crises would be to put in place improved national risk management systems and intensify international cooperation, in addition to creating national-level buffers capable of softening future shocks (Goldin & Mariathasan 2012).
Second, human intuition finds it difficult to deal with non-linearities. Epidemics exhibit exponential growth. Exponential growth remains a phenomenon that is difficult for humans to grasp intuitively. (Mathematics can help.) Therefore, dealing with phenomena involving exponential growth and non-linear behaviour often leads to major mistakes in decision making (Perrow 1984). This is why policymakers are so often behind the proverbial curve in the face of phenomena such as epidemics. It is also worth recalling that epidemics, like many other phenomena, can be described by a logistic function, which means that growth accelerate and then decelerates before ultimately coming to a halt, whether or not measures were taken to slow it down (Kuchnarski 2020, Smil 2019). Obviously it will be worth bearing this in mind when evaluating the relative success/ failure of policies aimed at fighting the epidemic – and even more so when it comes to debating their relative success/ failure in the (non-scientific) political realm.
Third, there is risk and there is the way risk is perceived. Misperceiving risk often leads people and individuals to make avoidable mistakes. Dread risk, for example, can lead people to make decisions that lead to even greater harm than the original risk/ event that triggered the decision. Second-round effects of catastrophic events can be more devastating than the original events themselves (Gigerenzer 2004). Traffic-related deaths due to increased driving increased sharply following 9/11. Suicides by US veterans allegedly exceed total US combat deaths in Afghanistan and Iraq. Cognitively, the recency effect and availability heuristics drive irrational decisions. It is therefore worth estimating the second-round consequences of the wholesale shutdown of the economy in response to the epidemic relative to more nuanced approach, including risk stratification (Katz 2020). Good decision-making needs to take account of both first- and second-round effects (even if the focus is on health outcomes only).
Fourth, the Knightian distinction between risk and uncertainty is crucial – epistemologically and psychologically. Uncertainty can be ontological and/ or psychological. Risk is a way to quantify the unknown. Traditional and Bayesian statistics are key to understanding what is going on and to informing policy decisions. The seasonal flu in the US is estimated to affect up to ¼ of the population and lead to up to 50,000 deaths very year. The current policy response seeks to (1) limit the absolute number of infections, (2) limit the rapid increase of total infections in order to prevent the healthcare system from being overwhelmed (and thus failing to save people who could have been saved with adequate healthcare) and (3) delay their spread in the hope that a drug or vaccination becomes available (before a second and third wave come around). The goal is to limit the absolute number of fatalities. The policy is informed by estimated numbers of (1) infections, (2) required hospitalisations and (3) fatalities. While data (though not always very comparable) is available for (2) and (3), this is of limited use as long as policymakers do not know what the number of total infections has been. It is of limited use because the effective fatality (fatalities/ total infections) rate might be higher (and thus justify more aggressive measures) or the effective fatality rate might be lower (and so the present measures may (may) in retrospect look quite severe). (In the face of reliable estimates and in light of the spike of hospitalisations, it is difficult to oppose policies such as shelter-in-place.) Bayesian and classical statistics are essential to help generate more accurate estimates. Initial uncertainty is gradually transformed into quantifiable risk, as both reliability and validity increase and estimates become more accurate. As an aside, uncertainty is experienced as more stressful by individuals (and financial markets) than significant, but quantifiable risk (de Breker et al. 2016). This is one more reason to generate reliable estimates as quickly as possible.
 |
| Source: New York Times |
Fifth, governments play a pivotal role in addressing public health crises. Competent political decision-makers, a competent bureaucracy and state capacity are crucial in efficiently addressing crises. Transparency and data-driven decision-making can help engender public trust and facilitate the implementation of important policies, especially where their success depends on voluntary compliance. Similarly, a lack of transparency may not only complicate decision-making (e.g. withholding of information by lower-level officials), but public suspicion over official transparency may undermine the effectiveness of policy decisions. A highly centralized political regime has advantages and disadvantages. If it pursues the right policies, it will be more effective than more fragmented regimes (e.g. China vs US). At the same, centralised and uniform magnify policy mistakes, limit opportunities to oppose mistaken policies and limit learning. While democratic regimes may, on average, be more transparent, they do not necessarily have a more competent and capable of bureaucracy (cf. Singapore, Korea, China, US). Similarly, in case of a severe crisis, autocratic regimes may be thought to have an advantage in terms of their ability to take extreme measures. However, most democratic regimes have an extensive arsenal of powers and measures they can resort in the event of a severe national crisis. Admittedly, much of this is speculation. What is needed is a thorough empirical examination of how various regimes types (democracy, centralization, bureaucratic competence/ state capacity) fare in the face of (public health) crises.
Sixth, the rally-around-flag effect and the yearning for a strong leader are very common features of a national crises. This is related to the psychological need for certainty and an abhorrence for uncertainty (see above). It is well-understood that uncertainty tends to lead to anxiety, a feeling of helplessness and even passivity and in people. It is therefore not surprising that people have a psychological need for leadership and authority, especially in moments of heightened uncertainty, and therefore rally around the flag. Intriguingly, even in the case of non-crisis-related epistemic uncertainty, people will tend to look leaders and authority for guidance or, rather, they look to individuals/ institutions that come across as authoritative. Unfortunately, this is also why so many people are prone to fall for the advice of well-dressed investment bankers, eloquent professors and “hedgehogs” appearing on primetime TV, while they really should be listening to the more thoughtful (if complex) arguments and less deterministic/ more probabilistic predictions of “foxes” (Tetlock 2005, Tetlock 2016). A desire for authoritative leadership in situations of uncertainty and increased risk may be hardwired in humans. If it is, it is arguably more likely to stem from humanity’s more recent agricultural rather than our (more egalitarian) hunter-gatherer past (Jathe & Ryan 2010). Perhaps not surprisingly, people are quite willing to accept government guidance in crisis situations. In part, this may be explained by the fact that such guidance can help overcome coordination and collective action problems (Sheppard et al. 2006). Speaking of which.
Seventh, externalities – costs or benefits related an action affecting others – is a key concept in understanding several aspects of the epidemic. For example, young people, less at risk from the epidemic, generate negative externalities for the rest of society by not complying with restrictions on public gatherings. Governments, national or state, buying up healthcare supplies reduce their availability to and/ or increase the price for others. Externalities are also closely related to the problem of collective action and successful cooperation (Olson 1965). In the present situation, coordination and cooperation among individuals or countries would lead to improved aggregate social welfare outcomes but are often undermined by “free-riding”. Furthermore, people who quarantine themselves do so for legitimate reasons. But as long as the fatality risk remains unchanged and a certain share of the population will get affected no matter what (absent effective drugs or vaccines), they are playing a zero-sum game that generates obvious externalities. Effective coordination could involve government-guided risk stratification policies that would help generate better aggregate welfare and health outcomes. If, for example, the herd immunity threshold can be reached, while helping vulnerable groups avoid infection, lives could be saved and aggregate welfare (a horrible economic term in this context) improved.
Ending on a note of a cautious optimism, research suggests that people overestimate the severity and duration of significant events in psychological terms. Crises related to wars and terrorist attacks as well as personal trauma are often experienced as very severe. Empirical research, however, demonstrates that people overestimate the longer-term psychological impact of such events (Brockman 2013). One is tempted to link this empirical fact to the concept of complex adaptive systems. To the extent (sic!) that the international economy, nation-states, societies and cities are complex adaptive systems, they have a tendency to exhibit occasional instability followed by adaptation and survival. Following major instability, they continue to perform many of the same functions, albeit in potentially different ways, as they did before. This time will be no different (Dower 2000, Schivelbusch 2000).