CALCULUS OF CALAMITY

This article is an extract from James Shepherd-Barron’s book ‘Calculus of Calamity’, adapted to reflect the emerging realities of nature’s most lethal natural hazard, a viral pandemic.

The Calculus of Calamity is a new stochastic optimisation model which provides a blueprint for re-framing the fundamentals of how disaster risks are measured, mapped and managed. It’s the algorithm that determines who dies, who doesn’t, and why in a crisis. It can be applied before or after a disaster and looks like this:

• 

It is to disasters what Isaac Newton’s famous second law F = ma is to physics … albeit in slightly less elegant form. As with Newton’s equation it demonstrates the relationship between a series of different elements. In the case of so-called ‘natural’ disasters, over two thousand variables interact to determine what happens when a naturally occurring hazard event collides with human nature. In the Calculus of Calamity above, sixty of the most influential have been grouped into six areas of risk (more detail of which can be found in Annex A):

• The type of hazard, including its probability and magnitude
• Society’s tolerance for risk
• The cost per avoidable net excess death or injury
• The vulnerability of people and assets potentially exposed
• Society’s resilience and ability to adapt
• How well preparedness, response and recovery efforts are managed

In plain English, it simply describes what happens when vulnerable people with limited resilience are exposed to potentially lethal naturally occurring phenomena and how the potential consequences are managed. These phenomena range from being extremely violent sudden-onset events such as a hurricane, tsunami or earthquake to something less dramatic but no less lethal such as a drought or epidemic.

The word ‘calculus’ is used to describe the rational and irrational heuristic processes we humans use to deduce and manage risk in our lives, while ‘stochastic optimisation’ is just fancy management-speak for describing the ‘wicked’ process by which each variable (risk factor) interacts and changes over time.

FROM NATURAL HAZARD TO UN-NATURAL DISASTER

As an algorithm, the Calculus of Calamity helps disaster managers^[2] make better, more evidence-based, decisions. This then allows them to prepare better, respond faster, and manage resources more effectively. With uncertainty, chaos and risk at its core, it uses the interdisciplinary approach of ‘Decision Theory’ to provide a deterministic framework for this. Such a framework is needed not just because of the chaos and complexity involved but because humans are erratic and do not always behave rationally, especially under stress. While econometric and epidemiological modelling informs what optimal decision-making should look like, the three P’s of psychology, philosophy and politics introduce elements of human behaviour which suggest ‘most likely’ outcomes, not necessarily the most optimal.

In building on proven concepts of Disaster Risk Reduction, the Calculus of Calamity integrates crucial elements of disaster epidemiology, health economics and business management into a comprehensive and practical decision-making framework. It does so on the basis that disasters are, by definition, essentially public health crises with indirect secondary and tertiary effects that can last for generations. They are hazardous events with the potential to do great physical and psychological harm to large numbers of people over long periods of time. In doing this, it acknowledges for the first time that ‘coordination management’ is second only to poverty, urbanization and population growth as being a major determinant of disaster in its own right.

Disasters are the world’s most unpredictable and chaotic events.

Chaos Theory suggests that complex systems tend to behave in unpredictable ways and even simple events can be inherently chaotic. One of the most famous examples of this theory refers to a hurricane in the Caribbean, the intensity, timing and path of which is influenced by minor perturbations in the atmosphere caused several weeks earlier by the flapping of a butterfly’s wings of the coast of West Africa. This metaphorical story illustrates how one small change in the state of a deterministic non-linear system can result in large differences in a later state. While the equation itself might be correct, tiny errors fed in at the beginning of the estimation process end up in wildly different conclusions by the end. Disasters, in other words, are the world’s most unpredictable and chaotic events.

In recognition of this, disaster modelling has until now tended to focus on the selection of only a limited number of variables, arbitrarily chosen by ‘experts’ from existing statistical datasets. These are then applied with scant consideration given to their relative importance (weighting) or accuracy (sensitivity). Most existing methodologies use only twenty or so, the most frequently used being statistics on things such as population density, educational attainment, and disease susceptibility^[3]. This is not enough for an empirical model – the ‘calculus’ as it is called here – which uses maths rather than myths to inform the decision-making process. That this appliance of science involves the three P’s of psychology, philosophy, and politics is what makes it less about the numbers and more about the process.

It’s also more about Public Health than any other scientific area of enquiry as the evidence which informs the calculus comes from the world of disaster epidemiology where the numbers are derived from complicated health information systems and interpreted through the lens of health economics. Both disciplines prefer relative measures such as ratio’s rather than absolute numbers. When reporting a disaster, for example, the media tend to talk in terms of the numbers killed in order to impart some idea of the scale of the disaster to the general public. This is not of much use to the disaster manager who needs to know how many were killed per 100,000 head of population in order to make appropriate and timely decisions proportionate to need and relative to what has happened in similar situations before. When it comes to the allocation of scarce resources the twin concept of prioritisation and proportionality become critical to making better decisions … which is why disaster managers need to know their epidemiology, especially concepts such as Relative Risk, Attack Rate, R-nought, Incidence, Excess Mortality and Quality Adjusted Life Years.

Let’s take an example: For a given population size, it doesn’t matter how many children have been vaccinated against, say, Measles, as this, while nice for the children and their mothers, doesn’t in itself tell the public health practitioner very much. What he or she needs to know is what percentage of all eligible children have been vaccinated as this coverage data tells them something about the work yet to do – the gap – and whether herd immunity has been reached. As with all types of disaster, it’s what hasn’t happened that matters just as much as what has.

Disasters are social constructs not natural inevitabilities.

Disasters are not ‘natural’ events; they are the consequences of long-term processes of accumulated risk. In their differential impacts across exposed societies, disasters need to be seen as social and political manifestations of inequality and injustice rather than random and unpredictable acts of violence. Across the world, poverty, ideology, politics, class and power relations lead to a build-up of unequal risk that leaves some people considerably more vulnerable than others. Vulnerability is related to social processes in disaster-prone areas and is intimately connected to the fragility, susceptibility and lack of resilience within the population. In other words, disasters are socio-political and environmental by nature, and their manifestation is the result of the socialisation of risk. For every inadequately constructed building and every uninformed person, there is an underlying social context which determines impact. Disasters, in other words, are social constructs.

MANAGEMENT & COORDINATION

Disaster management is essentially a political process but one which must be informed by the evidence. This evidence is captured and analysed in a process we refer to as ‘science’ but it is interpreted and applied by politicians. So we have to be very clear about what so-called ‘scientific evidence’ can and cannot do. Richard Feynman, one of the world’s best-known physicists back in the 1960’s, was as much philosopher as he was theoretical physicist. He used to say that “science is believing the ignorance of experts.” He was also vocal about the way society increasingly expects science to do something it cannot. That is, to characterise highly complex and dynamic systems in terms of linear, unambiguous causes and effects that can be understood, replicated, and applied in the real world.

In practice, the management of humanitarian support is a messy and complex affair that suffers the chaos of uncertainty; insufficient and asymmetric information; and a host of actors with different mandates, different political affiliations and different ways of doing things, most of whom are competing rather than collaborating. It is the task of ‘coordination’ to take such complications into account when managing this complexity.

The Calculus of Calamity challenges and expands the traditional definition of disaster epidemiology, a discipline which, until now, has been mostly concerned with the impact of communicable and chronic disease on disaster-affected populations. Disaster Epidemiology is about much more than population health. Because of this, and because disasters are largely adverse public health crises, disaster managers need to know some basic principles of epidemiology if they are to allocate resources properly.

Coordination costs. But poor coordination costs lives.

Disaster managers are confronted with ethical considerations and dilemmas at every turn. Dealing with the aftermath of disaster is no morality play; it is brutal. Disaster managers make life and death decisions all day, every day, and as a result end up “playing God” more often than they would like. Even with the best of intentions, poor management has the potential to cause immense harm if actions to assist and protect are mis-applied. Coordination, cooperation and collaboration costs. But poor coordination, cooperation and collaboration costs lives.

As has already been said, disasters are some of the most complex and dynamic systems imaginable. Not only are there over 2,000 variables involved which interact to determine who dies, who doesn’t, and why, but preventive actions aimed at reducing overall disaster risk – some of the consequences of which might not be felt for years after the event itself – require integrated, joined-up responses across multiple sectors and workstreams^[4]. All this has to be measured, mapped and managed if unnecessary death and injury is to be avoided. Calamity is a complex and complicated business.

Meanwhile, as the management jargon puts it, “You cannot manage what you cannot measure … and you cannot change what you cannot manage”. Disaster management authorities around the world make complex decisions all the time, so the data required must be clear and concise and consistently and coherently communicated.

All disasters set off a cascade of consequences, many of which are indirect and have long-lasting implications. Most direct effects are measured in terms of total loss of life, numbers of people injured or otherwise affected, and the replacement cost of infrastructural damage incurred. Indirect effects can include increases in communicable diseases such as cholera, measles and malaria; increased incidence of heart attack and suicide; and disruption to school attendance.

Disaster deaths are ambiguous in another sense: What is meant by this is that mortality studies focus on people who clearly died or were injured as a direct result of a specific hazardous event. But many deaths occur months or even years after that same specific event from indirect causes, like emphysema from inhaling asbestos fibres in the dust after an earthquake, or suicide from mental stress having been displaced far from known communities. Because establishing cause and effect in such cases is highly subjective, these deaths remain unnoticed and become part of the disaster’s hidden cost.

In almost all cases, a complex interplay between these physical factors and the way people behave is what actually determines who lives and who dies.

Overall then, the term ‘natural disaster’ is not entirely accurate, since the conditions that lead to the calamitous impacts of a natural hazard are inextricably linked to prevailing socio-economic conditions that are determined by human actions and decisions. They are the consequences of long-term processes of accumulated risk. Yes, nature displays its unyielding and immutable force through physical and biological processes but in their differential impacts across exposed societies, disasters become something else; as social and political manifestations of inequality and injustice rather than random and unpredictable acts of violence. Across the world, poverty, ideology, politics, class and power relations lead to a build-up of unequal risk that leaves some people considerably more vulnerable than others. For every inadequately constructed building and every uninformed person, there is an underlying social context which determines impact. Disasters, in other words, are social constructs. This forces us to conclude that all disasters are theoretically avoidable.

Information Management

Governments generally underperform as purveyors of reliable information on disasters. In Sierra Leone, three months after the Ebola outbreak was declared and one week after a national state of emergency was announced, the national ‘Ebola Operations Centre’ established to manage the extensive control measures needed consisted of little more than two flip-charts, one without paper, and a pile of pens strewn across a dusty table. There was nothing to show that this small room, crowded with empty chairs, was an ‘operations centre’ for anything, let alone an unfolding national calamity. There were no maps and no organigrams on the wall to show who was responsible for doing what, where, and when. There were no graphics charting the progress of the disease or of the measures underway to control it.

The World Health Organization has come a long way since then and by the time of the 2018 outbreaks in DRC had designed an operational template for enhanced coordination and control, including on how to report on what they call KPI’s or Key Performance Indicators. As with the COBR C-19 Secretariat in London, the Emergency Operations Centre in Kinshasa would gather data from the affected area and from all over the world, compile it into 13 response areas (each with specialist ministry committees), interrogate the data, make their collective analysis, and report their findings to the Minister of Health at 2 pm every day. Rarely did anybody agree on anything. International health epidemiologists in particular are not good at dealing with uncertainty and find it difficult to share or analyse information that they “cannot defend in front of (their) peers.” This reticence is born of a deep suspicion of ‘noisy’ (conflicting) data and the flawed statistical analysis that can sometimes result. The trouble is, a disaster unfolding in real time cannot afford to wait the weeks required for proper peer review. By then it is too late. Heated discussions between doctors, epidemiologists, statisticians, anthropologists, logisticians, communications officers and managers ensues yet somehow ends up in one coherent, easily readable dashboard. The art lies in understanding the assumptions behind each decision and being comfortable with ‘least worst’ options within a range of possible outcomes.

Disaster managers can’t afford to wait for certainty.

The design for such a dashboard exists. What it should look like is outlined in the ‘World Health Organization Framework for a Public Health Operations Centre’. Produced on the back of lessons learned from countless outbreaks around the world over decades, it was put together by information managers who know how to present qualitative and quantitative data in easily understandable form. It does not need to be re-invented. It includes data on what UN disaster managers call “4W (who, what, where, and when) mapping” combined with health information management data on aspects such as hospital capacity, bed occupancy, laboratory diagnostics, case management, contact tracing, disease surveillance, supply chain management and human resourcing. Lots of numbers are then transformed into easy-to-understand graphics and supplemented with time-lapse maps to create a single gateway into what is going on in real time. The World Health Organization’s dashboard covers one entire wall of its Emergency Operations Centre in Geneva.

Tracking operational outcomes against KPIs, together with an explanation of what is being done to address shortcomings, is useful for both downstream operational planning and upstream accountability and is a standard operating procedure in international disaster responses, as is regular gap analysis. Reflecting operational performance against pre-determined criteria and benchmarks is critical, not just for strategic oversight and programme adjustment but for better integrating risk communication and community engagement into the overall response … a frequent shortcoming in outbreak responses up until now.

TYPE OF HAZARD

Each type of disaster has its own profile and pathology, and these similarities and differences are crucial when it comes to planning, designing, and implementing prevention activities and in executing response programmes. For example, knowing that attack rates^[5] for those diseases found amongst disaster victims in each different type of disaster can help determine what kind of supplies, equipment, and personnel are most urgently needed is useful, but knowing that ingestion of cholera-laden seawater into the lungs after a tsunami gives rise to a particular kind of pneumonia; that inhalation of concrete dust after an earthquake gives rise to another; that dialysis is required to treat crush injuries after a landslide; or that snake bite is potentially one of the biggest killers of children in a flood is just as important.

To lend even more perspective, we must remember that disease epidemics have killed by far the largest number of people of any disaster type during the past two millennia. Smallpox was responsible for an estimated 300 million deaths during the 20^th Century alone, while the ‘Black Death’ is considered to have been the deadliest pandemic in history. Starting in 1347 and lasting five years, this severe and incurable bacterial bubonic plague killed 30% – 50% of the population of Europe. The Spanish flu pandemic of 1918-1919 killed an estimated 50-100 million people worldwide in only two years … around 4%-7% of the world’s population at the time and many times more than were killed in the First World War. Coming as it did at a time of war and food shortages, it has always been assumed that this extraordinary death rate was caused by the population being immuno-compromised, poorly nourished and never having been exposed to this particular pathogen before.

Influenza (the ‘Flu’) or some type of Coronavirus – the virus which causes the common cold – has always been thought of as the most likely cause of any future pandemic^[6]. And epidemiologists agree that it is not a question of ‘whether’ but ‘when’ such a killer will strike. We’re not talking about the kind of Flu that knocks us out for a few days every other winter, but a mutated strain against which we have no built-in immunity and for which there is no cure, only preventive vaccination.

That’s why Avian and Swine Flu caused such a commotion a couple of years ago, with the UN scaring the wits out of the world by suggesting that up to 150 million people could die^[i]. But statistical regression models clearly demonstrate that they were right then and they are right now. The threat has not receded.

It is the ease and speed with which these viruses are able to mutate that has meant there is no generic vaccine that is effective against all strains. Some are very simple and fragile scraps of protein. Covid-19, for example, has a DNA sequence of only 30,000 letters, the equivalent of 20 pages of a paperback novel. By comparison, the human genome has over 30 billion letters of DNA code, equivalent to 10,000 copies of the same novel. But, with viruses able to replicate – and mutate – every 20 minutes or so and humans replicating only every 25 years, we should not underestimate the potentially lethal advantage even the simplest of viruses have to cause pandemic calamity.

Failure to resolve ethical disputes over vaccination priorities could kill more people than the original disease.

When a new strain emerges, it can take between five and ten years to develop a viable vaccine and several more before it becomes commercially available at scale. Ethical considerations over who gets vaccinated first and who gets to keep the profits will be difficult to resolve and could lead to social and economic spillover effects that end up killing more people than the original disease.

It is easy to imagine the mass panic that would result if something similar to ‘Spanish Flu’ were to emerge again. Power-station workers, doctors, delivery drivers, water treatment engineers, and petrol station managers would either be sick or dead. Hospitals would close their doors. The lights would go out, the shops would be empty, and cities would grind to a halt. With people fleeing to remote rural areas, law and order would break down. This is the scenario so vividly – and, according to CDC in Atlanta, so accurately – portrayed in the film Contagion.

Anyone who saw Dustin Hoffman play a manic and obnoxious doctor from the US Centres for Disease Control (CDC) in the film Outbreak, or Kate Winslett in the more recent and realistic film Contagion will understand why there is reason to be alarmed about the threat posed by emerging infectious diseases. Emerging and re-emerging infectious diseases are those that are resistant to all known antibiotic therapies and/or for which the population has little or no immunity.

According to the US Institute of Medicine, if the next major infectious disease is not from a previously unknown bug, the biggest threat comes from HIV-AIDs, Hepatitis-C, Tuberculosis (TB), and new, more lethal, variants of Coronavirus and Influenza. They also think that hospital acquired infections will also pose a growing threat as drug resistance increases and new strains of Streptococcus or Staphylococcus emerge. Already, there is only one antibiotic left that controls spread of the “super-bug” Staphylococcus aureus, and there are signs that even this is losing its effectiveness. TB, cholera, and malaria are not only beginning to make a comeback, but are doing so with more virulent and drug-resistant forms.

In reality, there are bugs out there which can kill up to 80% or more of all people they come into contact with, and for which there is no cure. Except for the most exotic – by which we normally mean ‘deadly’ – you will have heard of most of them: Influenza, AIDS, Ebola, and Bubonic Plague, for example. They may well ‘self-limit’ – i.e die out on their own accord – but not before millions are dead or dying.

But you are unlikely to have heard of Henipah, a particularly nasty form of virus found originally in fruit bats. Fruit bats have evolved with this virus over millions of years, and because of this co-evolution, they experience little more from it than the fruit bat equivalent of a cold. But once the virus breaks out of the bats and into a species that hasn’t evolved with it, a horror show can take place, as one did in rural Malaysia in 1999. It is probable that a bat dropped a piece of saliva-covered fruit into a forest piggery. The pigs became infected with the virus, and then amplified it. And then it jumped to humans. It was startling in its lethality. Out of 276 people infected, 106 died and many others suffered permanent and crippling neurological disorders. There is no cure or vaccine. Since then there have been twelve similar, though thankfully smaller, outbreaks in South Asia.

Nearly two-thirds of emerging infectious diseases that affect humans originate in animals, with more than two-thirds of those originating in wild animals. The scope of the challenge this presents is huge and complex, not least because it is estimated that only one percent of viruses that exist in wildlife are known. And, with modern air travel and a robust market in wildlife trafficking, the potential for a serious outbreak in a large population centre is growing all the time. Increased ease of travel – one million humans are in the air at any one time – has radically altered the speed at which microbes can meet and re-combine, and rendered us hideously susceptible to what results. Today, an aggressive transmissible influenza or coronavirus with an incubation period of a few days could be on every continent within 36 hours.

In other words, outbreaks of potentially deadly diseases reflect what we are doing, either deliberately or unwittingly, rather than just being things that happen. In this – and as the Ebola epidemic so vividly demonstrated – epidemics are no different to any other form of so-called ‘natural’ disaster.

We live in a world that, at least from the point of view of a virus or a bacterium, has changed very little. Our world remains fraught with the risk of new pandemics as microbes that have never encountered each other before combine to form mutant stains which will cause diseases capable of spreading in ways neither of their ‘parents’ could ever do.

The appearance of a virus capable of infecting 40% of the world’s population, and killing unimaginable numbers of them is not as far-fetched scenario as you might think. This is what Laurie Garrett said about Bird Flu (Avian Influenza) when it was making headline news in the years after 2005:

“The havoc such a disease could wreak is commonly compared to the devastation of the 1918-19 Spanish Flu, which killed over 50 million people in 18 months. But avian flu is much more dangerous. Doom may loom. But note the ‘may’. If the relentlessly evolving (H5N1) virus becomes capable of human-to-human transmission, develops powers of contagion typical of human influenzas, and maintains its extraordinary virulence, humanity could well face a pandemic unlike any ever witnessed. Or nothing could happen at all.

Scientists cannot predict with any certainty what this virus will do. Evolution does not function on a knowable timetable, and influenza is one of the sloppiest, most mutation-prone pathogens in nature’s storehouse”.

Meanwhile, a Pangolin dies mysteriously in China, a chimpanzee in Central Africa, a few pigs in Australia, and whole flocks of chickens in Indonesia. People in regular contact with these animals fall sick and die. These real-life cases, and others involving bats and unknown numbers of even more exotic species, represent not just isolated events, but a trend in the transmission of new diseases from animals to humans. Covid-19 is just the latest example.

International health experts call such diseases ‘zoonotic’, meaning animal infections that somehow cross over to infect people. About one third of the 15,000 or so diseases known to man – including the modern-day scourges of HIV, Ebola and now Covid-19 – are in this category. For the most part, these diseases are the result of infection by one of three types of pathogen or bug: viruses, bacteria, and fungi. The most troublesome are viruses, mostly because of their abundance, their ability to adapt quickly, and the fact that they don’t respond to antibiotics. In the 1995 film Outbreak, a sweet little Capuchin monkey carrying a “deadly virus” that was going to cause “the greatest medical crisis in the world” caused anxiety in millions of cinema-goers. The film gave zoonotic infections the Hollywood treatment but stripped of the hyperbole, it contained elements of reality. Zoonoses are a major threat to human health, and it is considered “highly likely” that the next pandemic will originate from an animal, as Ebola did.

Within the viral camp, there are two main sub-groups, the DNA and RNA^[7] viruses, with the RNA viruses being particularly worrisome. HIV-AIDS is caused by a zoonotic RNA virus. So was the Spanish Flu Laurie Garrett referred to above. And so are Ebola, Marburg, Lassa, West Nile, Dengue, Rabies, Yellow Fever, SARS, and all those other spooky names which strike the fear of God into anyone who has seen blood oozing from Kevin Spacey’s eyes after being infected by some unidentified bug in the movie Outbreak.

There are an awful lot of RNA viruses. They exist in the oceans, in rivers, in the soil, in forests, and in urban jungles. According to Professor Eddie Holmes of Penn State University, one of the world’s leading virologists, it’s possible that every species on the planet, bacterium, fungus, plant, and animal, supports at least one RNA virus, though, as he puts it, “we don’t know for sure because we’ve only just started looking.”

We do know, however, that influenza viruses – which are RNA viruses – can be lethal and that there are three types, rather unimaginatively called A, B, and C. The A-type viruses cause the most severe epidemics in humans, and only this type is further classified into sub-types on the basis of the two main surface proteins, one called Hemagglutinin (H), the other, Neuraminidase (N). There are 16 known H sub-types, and 9 known N sub-types, which means that at least 144 combinations, or strains, are possible. So far, only three (H1N1, H1N2, and H3N2) are in general circulation among people.

In the mid 1900’s, scientists from the Rockefeller Foundation and other institutions conceived the ambitious goal of eradicating some infectious diseases entirely. They tried hard with Yellow Fever, spending millions of dollars over many years, and failed. They tried hard with Malaria, and failed. They tried again with Smallpox, and succeeded. Why? The differences between these three diseases are many and complex, but probably the most crucial one is that Smallpox resided neither in a reservoir host, nor in a vector such as a mosquito or tick. Its ecology was simple. It existed in humans and humans only, and was therefore much easier to eradicate. The campaign to eradicate Polio, which is still ongoing, begun in 1998 by WHO, is a realistic effort for the same reason: Polio isn’t zoonotic. Eradicating a zoonotic disease, whether a directly transmitted one like Ebola, or an insect-vectored one such as Yellow Fever is much more complicated, because to exterminate the pathogen you either have to exterminate the species in which it resides or interrupt transmission in some other way.

RISK [ r ]

New Zealand lies on the so-called Ring of Fire, a 25,000-mile chain of 452 volcanoes around the edge of the Pacific Ocean. Since records began in 1850, about 90% of the world’s most powerful eruptions have happened along this boundary. White Island – or Whakaari as it is called in its native Maori – is situated just off the North-East coast and is New Zealand’s most active cone volcano. The island has been in a nearly continuous stage of releasing volcanic gas at least since it was sighted by James Cook in 1769 and erupted continually from December 1975 until September 2000, making it the world’s longest historic eruption.

Brad Scott, a volcanologist for the science agency GNS who had been visiting Whakaari for 40 years, was responsible for overseeing the monitoring and surveillance of the volcano, setting the daily alert level, and issuing a regular bulletin informing everyone in the area how it was behaving and, crucially, how it was predicted to behave in the near future. As with every one of the 1,500 active volcanoes in the world, this monitoring applies some pretty sophisticated science, including hyperspectral thermal imagers mounted on orbiting satellites and in aircraft, surveillance cameras and lasers fixed to the crater’s rim for calibrating rock deformation and water levels in the crater, magnetometers for detecting minute magnetic variations, gas analysers for monitoring sulphur dioxide content in released steam, temperature gauges for detecting sudden changes in thermal activity, and seismic accelerometers for detecting minute horizontal and vertical accelerations in the ground. Any one of these sensors helps predict when a volcano will blow. When taken together, however, prediction becomes more accurate still. Nevertheless, it is still possible for eruptions to occur with no more than a few seconds’ warning.

And just after lunch on a sunny day in November 2019, when 47 tourists clad in masks and gloves were clambering about the island, that’s exactly what happened.

With only a few seconds’ warning, gigantic burps of super-heated volcanic gas with temperatures of 1,000 °C blasted over the crater’s rim at over 1,000 kph. Any living thing in its path would not have stood a chance. Eleven people were instantly incinerated, with ten more dying of their injuries in the weeks that followed.

On 18^th November 2019, just ten days before she blew, Brad raised the volcano’s alert level from 1 to 2 on a scale where 5 represents a major eruption. He made this decision having been notified via an automated computer algorithm that the volcano was becoming “noisier” i.e was showing signs of imminent activity. With each volcano having its own ‘tell’ or ‘signature’, the precursors in this case included tell-tale patterns in the magnitude and frequency of seismic tremors and an increase in the temperature and amount of sulphur dioxide being released … both being key indicators of magma rising deep in its bowels. Volcanoes generate low-frequency, infrasonic sounds inaudible to human ears during their eruption processes which can be easily detected by arrays of sensors placed some distance from the site. For certain types of volcano, these systems are extremely reliable. From 2008 to 2016, sensors near Mount Etna in Italy detected 57 of 59 eruptions and sent automated SMS alerts to the disaster management authorities about an hour before each eruption.

Not surprisingly, questions have been asked about the wisdom of allowing tourists onto an island which is an active volcano. The answer, of course, has a great deal to do with money. In recent years, tourism to Whakaari has been on the rise with over 10,000 people visiting the volcano every year, and it had become a significant part of the local economy.

Individual visitors are left to make their own ‘trade-offs’ between the risk and the reward. When making such decisions, it is assumed their consent has been properly informed. But even when fully aware of the risks and possible consequences, are members of the public capable of making entirely rational decisions in such circumstances? For sure, they will have assumed that ‘all reasonable measures’ would have been taken by those in charge to protect them.

But had they? And where does the level of ‘acceptable’ risk lie? While there have been fraught debates about tourist access to other volcanoes overseas, including Japan’s Mt Ontake, where 63 people died in 2014, Mt Yasur, in Vanuatu, where there have been several deaths in recent years, and Kilauea in Hawaii, tourism analysts in New Zealand said the risks of Whakaari bear greater examination.

Since Whakaari was “a disaster waiting to happen,” according to Ray Cas from the School of Earth, Atmosphere and Environment at New Zealand’s Monash University and “was too dangerous to allow daily tour groups to visit … not least because it experiences significant explosive eruptions every three to five years.”

It has only recently been recognised that risk – or, rather, a community’s ability to perceive risk correctly, its tolerance for any residual risk, and its attitude towards risk aversion – is itself a function of risk. And the way a disaster is managed (or mis-managed) is one of the biggest risk factors of them all.

Not all risks can be eliminated. Nor should they be. Whatever the risk reduction efforts made, there will always be some residual risk. But an accountable government should ensure that every effort has been made to protect those they serve from all ‘reasonable’ risks. This, of course, raises the question over what constitutes ‘reasonable’?

The way a disaster is managed (or mis-managed) is one of the biggest risk factors of them all.

We understand risk as a measure of uncertainty. It is the probability of an event occurring multiplied by its possible consequences (impact). Even now, people who think of themselves as modern still find this a difficult concept to grasp. This is for two main reasons: The first, being simply about the mathematics, is explained in Steven Levitt and Stephen Dubner’s thought-provoking book Freakonomics, like this:

“Consider the parents of an eight-year-old girl named, say, Molly. Her two best friends, Amy and Imani, each live nearby. Molly’s parents know that Amy’s parents keep a gun in their house, so they have forbidden her to play there. Instead, Molly spends a lot of time at Imani’s house, which has a swimming pool in the backyard. Molly’s parents feel good about having made such a smart choice to protect their daughter.

But according to the data, their choice isn’t smart at all. In a given year, there is one drowning of a child for every 11,000 residential pools in the United States. In a country with 6 million pools, this means that roughly 550 children under the age of 10 drown each year. Meanwhile, there is one child killed by a gun for every 1 million-plus guns.

In a country with an estimated 200 million guns, this means that roughly 175 children under 10 die each year from guns. The likelihood of death by pool (being 1 in 11,000) versus death by gun (being 1 in 1 million-plus) isn’t even close: Molly is far more likely to die in a swimming accident at Imani’s house than in gunplay at Amy’s”.

The point they are really making is that there are risks that scare people and there are risks that kill people, and the two are very different.

The second challenge with risk is in how it is perceived. An equally thought-provoking book called Nudge by Richard Thaler and Cass Sunstein explains the behavioural psychology involved. In it, they acknowledge that biased assessment of risk can, as in Molly’s case, adversely influence how we prepare for and respond to crises. But it also suggests that we don’t always see what we’re looking at, despite being absolutely convinced that we can.

They demonstrate this ‘blunder of bias’ by showing a drawing of two tabletops, one long and thin, the other short and wide. The funny thing is that both tabletops are identical in size, yet they look different because of the different perspective and visual cues used when they were drawn. The point here is that our judgement is often biased by such subtle shifts in perspective and the way we as individuals interpret contextual cues.

Not only that, but Thaler and Sunstein go on to point out that we can be absolutely convinced of our own infallibility. Anyone seeing the drawing for the first time is utterly convinced that one table is shorter and wider than the other. It’s the same when managing disasters: when everyone around a decision-making table is trying to reach consensus, personal convictions can be hard to shift even after the introduction of contrary evidence.

At its most fundamental, risk involves the concept of avoiding harm, and in so doing requires us to make trade-offs between our perceptions of hazard, probability, and likely consequences.

Do we really know enough, though, to allow us to make these trade-off decisions? Did the people of Christchurch, New Zealand, know on the 22^nd February 2011 when an earthquake devastated their town that six-to-eight-storey buildings collapse in earthquakes more often than those that are shorter or taller? Would it have made a difference to their daily lives if they had? Is it more important in a flood to know how to swim, or to know what to do if your brother is bitten by a snake? Is it important to know that many of those pulled smiling from the rubble of the Haiti earthquake in 2010 died later because dialysis was not available in sufficient quantities to treat the crush injuries they had sustained? Does it make a difference to know that, without vaccination, tetanus has the potential to kill just as many people as falling debris in an earthquake? Or that, without immediate use of third generation antibiotics, those who have ingested cholera into their lungs in a tsunami are likely to die later of pneumonia?

Insurance companies make their money by knowing something we don’t: Human beings don’t perceive risk rationally. Martin Hartley, Chief Operating Officer of one such company put it very clearly following the Paris terror attacks of November 2015: “The risk is made to seem far worse than it is by its perceived magnitude, not by its probability.”

Even well-educated and comparatively wealthy people fail to take the simplest of measures to mitigate the risk of disaster despite knowing that they live in hazard-prone areas of the world. Although things are improving now, only 17% of people living along the hurricane-prone Atlantic and Gulf coasts in the US had taken steps to storm-proof their homes in 2006. The same figure applies amongst Californians who fail to take any seismic risk reduction measure despite knowing they are living in an earthquake zone.

The problem is, people systematically misperceive probabilities and risks of natural hazard events: low probability events are consistently overestimated, and high probability events consistently underestimated.

So, too, recent events have a greater impact on our behaviour than earlier ones. These biases don’t appear to be related to the frequency of the event, either; people underestimate risks they have not experienced, and overestimate those they have. The sense of invulnerability that comes from a close encounter with a previous event is frequently cited as the reason for failing to evacuate when asked by the authorities to do so. Survivors with first-hand experience believe they are better equipped to deal with future events, only to find out too late that they are just as vulnerable as everybody else.

We also notice trends which aren’t real. People notice clusters of events but they don’t pay attention to the long gaps in between. Nor do we remember for long the impact of the previous occurrence. This helps to explain much risk-related behaviour, including both public and private decisions to take precautions against hazard risk.

Assessing relative risk means being able to qualify and, where possible, quantify a number of risk factors as each will play a role in determining how much risk a system, a process, a person, or a community can bear. Once we have done this, the question then arises as to which hazard poses the greatest risk, and how we weigh up that risk relative to the next so that preventive actions are correctly prioritised.

Given that we can’t eliminate risk altogether, is it morally acceptable to reduce only part of the risk? If we know perfectly well how to reduce this risk, why don’t we? Is total risk aversion even a good thing? Risk might deepen poverty and cause the economy to recede but it can also generate prosperity. With Covid-19, invasive ventilation might save the patient or it might kill him.

Mostly, it’s because the benefits of accepting a certain level of risk far outweigh the costs of eliminating it, especially when the ‘law of diminishing returns’ means that it becomes more and more expensive to reduce one ‘unit’ of risk the nearer we reach total elimination.

And then we have to work out how risk averse we really want to be. Are we willing to pay the emotional and financial cost of total protection? Do we even know what these ‘costs’ are when one is the cost to us as an individual or family, and the other will be a cost borne by the public at large.

Human existence is an inherently risky business. In the end, after all, nobody escapes alive. But how far should we go to control risk and how do we weigh that against unintended consequences? How morally risky is it to attempt to eliminate risk altogether, especially when the result of our failure to do so is felt by society as a whole rather than by us as individuals or families?

The human mind estimates risk through anecdote and imagery amplified and filtered by the psychological algorithm in its internal search engine, the brain, which favours vividness, movement, and recency. People thus think tornadoes kill more victims than asthma attacks, which are in fact 80 times more lethal.

ECONOMETRICS

It doesn’t help that what is valuable is not always valued by society. Although the dead are counted, and injuries logged, damage estimates ignore the value of lives lost and disabilities gained largely because of the difficult conceptual and ethical issues of valuing consequences of risk to life and limb.

It’s a sad truism that not everything that counts can be counted, and not everything of value can be valued. Nevertheless, to arrive at the right choice when it comes to preventing avoidable death and destruction requires assigning ‘value’ to human life.

If it works, the UK Government’s Covid-19 (C-19)^[8] control strategy will cost over £300 billion and save half a million lives, most of them elderly. According to the IMF, this level of expenditure will tip the country into recession and require levels of domestic borrowing not seen since the Second World War. Younger generations fear that it is they that will be left paying the price. Whether they deem this price worth paying depends on what value they are prepared to put on human life.

The UK Government’s C-19 suppression strategy and its demands for extensive country-wide social isolation and ramping up of NHS critical care capacity may result in between 35,000 and 70,000 excess deaths^[9] across the country by the end of December 2020 (UCL). This is in addition to the 30,000 that would die anyway from Acute Respiratory Infections such as seasonal Flu (PHE). Against the number of premature deaths predicted by ‘zero option’ (do nothing) modelling where up to 560,000 UK citizens would die (IC) – loss of life comparable to that suffered by the UK throughout the entire course of the Second World War – this represents the saving of around half a million lives.

However, this will come at great social and economic cost, leaving some to wonder whether such draconian disease control measures are worthwhile? Could the social and economic consequences of the virus be deadlier than the virus itself?

The answer, at least over the short term, appears to be ‘No’. The saving of 500,000 lives is deemed to be worth the billions of pounds committed by the UK Government. But how was that conclusion reached? And in what context does it hold good when the question has yet to be tested against society’s assumption that “every life is priceless” when, from a risk management perspective, they are not?

“Yes, there is a high price to pay. But how do you put a price on life?”

With the first peak of transmission estimated to be in mid-April, estimates of predicted mortality remain uncertain. But one aspect is crystal clear: The length of the epidemic and the numbers that die are, to a major degree, dependent of the public’s psychological resilience and their ability to break the chain of transmission through sustaining physical distancing from one another and upholding strict hand hygiene measures for years to come.

Much clearer is that the strategy now underway will tip the global economy into recession. Most of this money will need to be borrowed now and repaid by future generations.

Are those future generations, some of whom are not yet born, willing to pay such a price, especially when most of those who die are elderly men with pre-existing chronic illnesses (co-morbidities) and who, according to some public sentiments, “only have a few years left anyway”? After all, as the Chief Economist of the UN’s World Food Programme recently put it, “There is only so long an economy can be locked down without inflicting lasting damage.”

The ‘damage’ he is referring to is not measured in short-term mortality alone – a few years from now this disease will be endemic and treated much the same as seasonal Flu or a common cold is – but the long-term consequences of global economic shut-down.

To answer this question, we need to know how much a life saved – or, more specifically, a life-year gained – is worth to society. Only later on, will we need to know what the potential social, economic and political benefits might be in terms of, say, climate change, democratic accountability, and corporate responsibility.

The economic consequences of this disease could end up hurting more people than the disease itself.

The Humanitarian Dilemma

Whatever control strategy is applied to the C-19 pandemic, two things are clear: A lot of people will die prematurely – many unnecessarily – and the social and economic cost will be enormous. Leaving aside the ethical considerations, society is faced with the same ‘humanitarian dilemma’ faced by disaster managers when coordinating responses to international calamities where resources are never enough, the data unreliable, and uncertainty the norm: Is the predicted gain in terms of avoidable deaths averted by any one intervention ‘worth it’ in terms of the financial cost involved? How does one balance the cost of the prevention against reducing the risk of a premature death that may never happen? If there are not enough resources to do both, is it more cost-effective to improve access to safe water in a cholera epidemic or vaccinate the children? Is it better to distribute the full value of cash grant required to keep a family in Northern Syria alive to half the families that need it, or half the amount to all the families?

How does one balance the cost of prevention against reducing the risk of a premature death that may never happen?

The answer to ethical conundrums like these are not as difficult to work out as they appear, and, although away from public view, disaster managers make such life-and-death decisions all the time. So do insurance companies, government transport departments and NHS economists. Each uses a form of cost-benefit analysis (CBA) to help in their decision-making. And each involves the difficult moral question of how to value human life.

The Value of a Life

Although the concept of placing a monetary value on human life is controversial, the ability to do so is essential when making informed and rational decisions about resource allocation. Understandably, for ethical, religious or philosophical reasons, many people oppose valuation of something commonly perceived as priceless and argue that no monetary figure could possibly compensate entirely for the loss of a human life.

Nevertheless, comparing different options requires a common metric against which to measure impact. Usually, this is money. This immediately poses a challenge to planners as it requires a value be put on the direct effects on people vis-à-vis ill-health, injury and death. The UK’s Department of Transport uses a Value of Statistical Life (VSL). Disaster managers in the World Health Organization use Quality Adjusted Life Years (QALYs). Others use ‘Micromorts’ or ‘Judgement Values’ (J-values). Underpinning each approach are the triple concepts of life expectancy, earning potential and quality of life where a monetary value on future years of life is calculated based on discounted income (e.g GDP per head) and work-life balance (the ratio of time spent working to time not spent working). This is part of the Calculus of Calamity.

Although many would argue that the value they place on their life is infinite, reality reveals that this is not the case. There are limits to the amount we are prepared to pay for marginal increases in longevity – for example, to have expensive safety features fitted in our cars – and increases in occupational risk of death are often acceptable if the monetary compensation is substantial enough, as evidenced by the existence of ‘hazard pay’.

Another challenge, and one when we have already discussed, is that most people have great difficulty in understanding the varying levels of risk to which they’re exposed. This is especially true when the impact is intangible and far off, as it is with a deadly pandemic like C-19.

Another is that any attempt to put a figure on the value of human life has to take age and disability into account. Most people agree that it’s reasonable to suggest that the life of a new-born baby should carry a far higher value than that of someone in their eighties. Equally, people usually recognise that a double amputation might save the life of the patient but that this would mean additional years of life gained would be subject to a greatly reduced quality of life.

To accommodate such value judgements, health professionals and disaster managers use something called a ‘statistical life year’ (SLY) when making life and death decisions. Improved life expectancy in terms of ‘life years’ is a much better characterisation of the benefits to be gained by mitigating risk than ‘lives saved’. This is because, in fact, nobody’s life can ever be saved: the best that can be done is to return an individual’s life expectancy back to what it was before the hazard occurred. The SLY calculates the value to each individual of one additional year of healthy life gained by a particular intervention, adjusted for disability.

In the UK, where each citizen’s lifetime value to society is estimated to be worth between £5 – £9 million, the SLY value is estimated to be £248,209 (Thomas, 23 March 2020). This analysis is informed in part by a well-known economic effect called ‘The Preston Curve’ which demonstrates that we get to die earlier as we grow poorer^[10] (Preston, 1975).

This implies that the government would be justified in spending about a quarter of a million pounds on any intervention that would extend a single citizen’s life expectancy by one year. With each victim losing 14.6 life-years on average and, in the best case, 500,000 lives being saved, the cost of control measures must be lower than £1.8 trillion … which, at the moment, they are^[11].

Dividing the cost of the intervention by the aggregated number of life-years expected to be gained by the intervention allows disaster managers to calculate the relative risk of each intervention without having to bring unnecessary value judgements into the equation. Assessing cost-benefits in this way allows railway safety to be put on the same footing as medical treatments, natural hazard risk reduction and public health.

In this sense, our response to the C-19 pandemic is no different to our response to natural hazards such as Earthquakes or Tsunami; various response options, each with different risk parameters and expected outcomes, have different social, political and economic costs against which the short- and long-term implications on life expectancy and quality of life have to be considered.

Meanwhile, the constraint which all but the ‘zero option’ have to consider is that C-19 countermeasures should not decrease GDP per head so much that the UK population as a whole loses more life-years than it gains from such measures. Public health protection schemes should not be put in place if their costs are large enough to cause the nation’s economic output to fall so significantly that it will cause more loss of life and if the scheme had never been implemented in the first place.

It is likely that a recession resulting in a general fall in economic output of 6.4% per person over a prolonged period would cost more life-years than would be restored by current and future C-19 countermeasures. This observation is based on comparisons from the economic recession of 2008-2009 where GDP per head fell by 6% and did not recover until 2015, and its negative impact on life expectancy. Under the scenarios mentioned above, modelling suggests that each victim suffers about 14.6 life-years (years of healthy life) lost.

Current macro-economic analysis suggests that planned countermeasures will severely dampen economic activity. Both the International Monetary Fund and the Centre for Economics and Business Research now predict that the pandemic will cause global GDP to decline twice as much as during the financial crisis of 2008. Furthermore, it raises the prospect of a 1930s-style recession. Such an outcome, if it were to come about, would cause a loss of life-years to the UK population that would far exceed the predicted toll under the ‘zero option’. It would be worse than doing nothing, in other words.

Whatever lies ahead, it is already becoming clear that the impact of C-19 will be deep and lasting; not just in the UK and Europe but in poorer and conflict-affected parts of the world which are woefully unprepared and lack the capacity to respond. The UN estimates that over 100 million people from such areas might die.

VULNERABILITY & EXPOSURE

Vulnerability and Exposure are two sides of the same disaster management coin. But conceptually they are very different. To be exposed to hazard risk is one thing; to be vulnerable to it is quite another. Yet frequently the two terms are conflated, giving rise to false assumptions and flawed decision-making about which mitigation and adaptation measures to apply.

Hazards only result in disasters when they collide with the vulnerability of societies to which they are exposed i.e those people affected by economic, demographic, social, physical, environmental, or political risk factors. We assume, for example, that an expanding population means more people potentially exposed^[12], and that these extra people are just as vulnerable as before. But they are not necessarily more vulnerable, and therefore not necessarily more at risk. They might have taken robust measures to prevent or at least mitigate the threat posed by the hazard they knew was facing them … by building basement shelters, for example, in known tornado-prone areas or seismic-resistant hospitals in areas of known earthquake risk. They might have become more resilient by teaching each other how to treat physical injuries amongst their neighbours or maintain social distancing when faced with an infectious disease.

Exposure

Exposure refers to the people or types of assets located in a hazard zone that are thereby subject to potential losses. Rapid economic and urban development can lead to a growing concentration of people and economic assets in areas that are prone to natural hazards, such as earthquakes, droughts, floods and storms, for example. The risk increases if such exposure grows faster than country’s ability to reduce vulnerability and strengthen their risk-reducing capacity.

Particular risk factors include:

Population Density: The population potentially exposed (PPE) might be living in dispersed rural villages, in urban high-rises, or in peri-urban slums. The population per square kilometre will be very different for each.

Period of Exposure: This does not just refer to those trapped under the rubble in earthquakes, but communities that are cut off from outside assistance, and for whom access to health care, food, and water is not possible without, say, helicopters or boats.

Time: Much depends on the time-of-day the disaster took place, and whether children were in school, or workers in factories. Equally important is the time-of-year, as seasonal variations may affect tourist inflows or local population outflows.

Climate: Whether the climate is hot or cold, and whether or not this will change over the next few months will have a major bearing on shelter and protection policies. Precipitation patterns of rain and snow and likely windspeeds of seasonal storms will have a bearing on what type of thermal insulation to provide in the forms of clothes, blankets, stoves, and winterised accommodation.

Distance (from epicentre): Energy dissipates over distance. Turbulence will decrease as tsunami waters approach peak run-up elevation, and earthquake intensity at the surface in general reduces the further the building is from the epicentre (though with earthquakes, attenuation and geo-morphology can result in areas of greatest risk some distance away). 

Vulnerability

Vulnerability refers to the characteristics and circumstances of a community, system or asset that make it susceptible to the damaging effects of a hazard.

There are many physical, social, economic, and environmental risk factors, including:

Displacement: Those who have had their homes destroyed might be supported by neighbours within their communities, or may have been obliged to move in with family members far away, or into over-crowded and unsanitary temporary shelters where everyone around them is a stranger and privacy is at a premium.

Demography: With more women than men typically affected by disaster, the proportion of men, women, boys and girls among the survivors will be different to what it was before. Those under five years of age, the elderly, and those with physical or mental disabilities are particularly vulnerable. Minority groups, who might have been facing persecution before the event might find themselves being blamed, too. Levels of poverty may not have changed much, but disasters are great levelers as everybody is equally affected. Some will now have become single parents, female heads of household, or orphaned. Others will have lost contact with surviving family members.

Health Status: Diseases which were endemic before the disaster and vaccination coverage rates will both have a direct and immediate bearing on whether mass immunisation is called for, and what response measures are to be prioritised.

Nutritional status and the proportion of those suffering physical disability are also useful to know as these vulnerabilities become more acute in the weeks and months after a disaster.

Infrastructure: Dwellings and key structures such as hospitals, mobile communications towers, shelters, airport runways, bridges, and power plants (and the electricity distribution grids that go with them) need to be able to withstand the shock of foreseen hazards if relief and recovery operations are to have a chance of getting communities back on their feet quickly.

In an event like Hurricane Dorian, which ravaged the Caribbean islands of Grand Bahama in September 2019, everyone who has not got themselves out of harm’s way by evacuating is ‘exposed’ to the storm. Those left behind are probably least likely to be able to afford the air fare and are therefore more ‘sensitive’ to the hazard bearing down on them. Being poorer, their houses are probably more ‘sensitive’ too; more likely to be located in areas susceptible to flooding or landslide risk and of flimsier construction. These same people are likely to have fewer physical and financial resources at their disposal and, having lower levels of educational attainment than their richer neighbours, lower ‘adaptive capacity’. They are less able to cope and less ‘resilient’, in other words; and, being less ‘resilient’, they are more vulnerable.

By the same logic, increasing their resilience – their capacity to cope and adapt – will lower their vulnerability. This can take ‘soft’ and ‘hard’ forms and take place over short or long time-frames, with first-aid training for schoolchildren at one end and construction of hardened cyclone shelters at the other.

RESILIENCE [ R + Ma]

Resilience is the ability to prepare, plan for, absorb, recover from, and successfully adapt to adverse events.

Being able to anticipate disasters; making people aware of approaching calamity and what they have to do to survive it; and investing in shock-resistant infrastructure and social protection mechanisms, are all part of being resilient, as is the ability to adapt behaviour to changing circumstances. Everyone in the path of a tropical cyclone is exposed to the same hazard, but the young and healthy family who spent that little bit extra tying down their corrugated iron roof and protecting the water well is less susceptible to the hazard than the elderly widow next door who could not. Being poor, female, uneducated, disabled, or elderly also changes the risk profile, as these sub-groups are always disproportionately affected during natural disasters.

In practice, the picture is made more complex through politics, lack of information, dynamic movement within and between populations, and secondary shocks. Often, it is these secondary shocks that prove more debilitating than the original hazard event itself. Salt-water inundation, for example, may have rendered low-lying agricultural land infertile after a storm surge has drowned the cattle and washed away what was left of the family home after high winds had already blown the roof off. Later stresses such as polluted groundwater or rising food prices then begin to pile on top of the original shocks, accumulating slowly to become a shock in their own right.

Disasters are not only tragedies but also opportunities to do things differently and introduce new practices. Done properly, recovery efforts avoid creating new risks or exacerbating existing ones and provide an important opportunity to build back better and enhance resilience. Frameworks have to include at least the following components if they are to be successful:

• raise risk awareness
• carry out hazard mapping and cross-referencing the analysis with vulnerabilities and capacities
• develop knowledge through education, training, research and providing hazard risk information
• commit government and local authorities to institutional frameworks, including organisational, policy, legislation, and community action

So, once again, if we know what to do, why are there still so many avoidable tragedies? It’s not as if the world lacks the knowledge to reduce the death toll. A combination of better land use, better construction practices, and improved civic education would dramatically reduce risk across the full range of disasters. Why, then, are these mitigation measures not adopted systematically and wholesale?

The main reason is that the protection afforded is perceived to provide lower or more uncertain returns on investment than alternative expenditures in physical wellbeing. In other words, since the vast majority of people affected by disasters do not die, it may be easier, quicker, cheaper, and more politically expedient to prevent other causes of avoidable death when budgets are not sufficient to cover everything.

There are also few votes in spending money on preparing for something that may not happen. And, unless remote villagers and urban slum dwellers understand what might befall them, getting them to change their behaviour is extremely difficult. But it can, and, in the most ‘at risk’ places, must be done.

A FIGHT TO THE DEATH

Climbing mountains could never be considered a sport for the risk averse, but, given the range of mitigation measures that climbers take – from equipment, to training, to planning, to evacuation procedures – could it be that the whole balance of risk – the hazard, the exposure, and the climber’s vulnerability – is not what we think it is and prompts us to analyse risk in a new, more evidence-based, light

Disaster movies don’t confront us with our fallibility in quite such brutal fashion, and, in showing us how nature intends to kill us, comfort us in the knowledge that the extinction of our species is somehow not our fault. Disasters are ‘natural’ accidents; unexpected and unpredictable happenings removed from our everyday interactions with ourselves and the environment.

Furthermore, while the heroes survive, these freak events remain positioned outside the moral compass of our culture, and, as a result, no one can be held accountable for them. We forge on, eyes wide shut, aware but oblivious of our responsibilities. This is probably what John Holmes, a former UN bigwig, was implying when he said in 2010, “We are, to a certain extent, sleepwalking our way into disaster”.

In the modern world, it should not be possible, or acceptable, for large numbers of people to die in natural disasters such as earthquakes, volcanic eruptions, or landslides. These are well understood phenomena, and the science has existed for some time for us to understand how they happen, and when and where they might strike. And yet, sudden manifestations of these forces of nature continue to kill thousands, or even hundreds of thousands, of people at a time.

So, if we know what to do, why are there so many avoidable tragedies on our TV screens? Why do some schools become death traps during earthquakes? Why are so many trees chopped down on already unstable slopes, making it almost inevitable that a mudslide will result the next time it rains hard? Why were 20 million people washed out of their mud-brick homes in Pakistan in 2010 when extensive flooding was forecast days in advance? And why is it that the less than 2% of the $24 billion spent on humanitarian assistance every year is devoted to reducing disaster risk?

Some of the answer lies in the shifting demographics; the simple fact that the planet has more people on it now, and even though the number of people vulnerable to hazard events has declined proportionately, this smaller proportion of the larger total means more people are exposed overall.

There are few votes in spending money on preparing for something which may not happen.

But perhaps the fact that risk reduction programmes are extremely complex in nature and complicated to put into practice has a lot to do with it too. Engineering solutions are expensive, especially when outcomes are so uncertain. Mayors and Ministers are no different to the rest of us in tending to discount low-probability-high-impact hazards and seem reluctant to invest in disaster risk management as a result. Despite the magnitude of disaster costs, and the proven cost-benefits, reducing disaster risk is often perceived as less of a priority than mending the street lights or filling in the potholes. There are few votes in spending money on preparing for something which may not happen.

There is also a perverse political incentive. Political leaders have always understood the symbolic power of responding to disasters, especially natural disasters. Saving lives and assisting disaster victims is a moral, humanitarian, and political imperative that few would contest. As such, disaster relief is a powerful tool for leaders, enhancing their political profile and facilitating patronage. In contrast, the incentives for reducing disaster risk, a public good, are far less visible and far less obvious. The benefits also take longer to accrue than democratic election cycles allow.

In an ideal world, though, national and local authorities would prevent as much at risk as is cost-efficient, reduce or transfer what they can of the remaining risk, and then, recognising that there will always be residual risk, prepare for it. You would think that such an approach by those authorities whose job it is to protect us – our elected officials and public servants – would be obligatory. But it isn’t.

Ultimately, it comes down to economic fundamentals and on deciding how much risk we can tolerate and afford to accept. But the social and economic costs of ignoring disaster risk are substantial, especially if entire societies isolate and turn in on themselves. Not making decisions on reducing disaster risk is to accept high numbers of deaths, extensive damage, disruption to business, prolonged economic decline, and lost opportunity for years, if not generations.

A paradigm shift is needed from post-disaster compensation for reconstruction and recovery to pre-disaster investment in risk management and adaptation. This requires investment in enhanced early warning systems, better land-use planning, the establishment of sustainable insurance systems, and the building of resilient infrastructure such as hazard-resistant hospitals, schools, transportation systems, ports, energy grids, water treatment plants, and communications systems. It also requires the much cheaper, but much more difficult expedient of informing the people, so that we may be made aware of the risks facing us and know what options are available to confront those risks ourselves.

While this shift relies on political and technical innovation, it is the economic dimension that will ultimately be the foundation for change by developing incentives and market mechanisms to align actions among national and local governments, the private sector, and civil society toward greater resilience. Working with communities to enable them to prepare for, cope with, and adapt to, future hazards at the local level is the key.

To achieve this, all of us need to take greater responsibility for protecting ourselves and those around us. However, we also need to demand greater accountability than we do now from our elected representatives for reducing our vulnerability to so-called natural hazards in the first place.

The traditional discourse from those whose livelihoods depend on disasters continuing to occur is that the pace at which we are creating exposure to hazard risk has become greater than the pace at which we are reducing our vulnerability to it.

But when some of the long-held myths are unpacked, something else suggests itself: We know what to do, but we’re just not doing it.

Scepticism might linger over the true nature of climate change, its causes and effects, and more attention is already being paid to slow-onset disasters driven by such things as over-fishing, de-forestation, dependence on certain crops, sea-level rise, and excessive extraction of groundwater. But, as the response to typhoon Haiyan so cruelly demonstrated to the people of the Philippines at the end of 2013, the Ghorka Earthquake to the people of Nepal in 2015, and as the Covid-19 response around the world is doing now, resources are never enough. And it is here, in this vortex of competing priorities and misinformation, that the discipline of disaster epidemiology provides decision-makers, opinion-formers, and disaster planners with the evidence they need to confront hard and potentially unpopular choices.

It may be that the scientific community is petrified by its impotence in the face of these microscopic bugs. Certainly HG Wells was, as, in his prescient 1930’s book War of The Worlds, his Earth-destroying Triffids were eventually overcome, not by the ingenuity of man, but by a common virus. This was then, and is now, a fight to the death between bugs and drugs, and it’s not clear who the winner will be.

© James Shepherd-Barron

16 May 2020

The author is a disaster management consultant. During the period 2012–2018 he acted as a special humanitarian health adviser to the UK government’s Department for International Development (DFID) for Ebola and Cholera epidemics in Sierra Leone and the Democratic Republic of Congo. He has an M.Sc in International Health and an Honorary Doctorate in Public Health. He is currently advising the Cash Management Industry on cash-based approaches to Covid-19 response.

He can be contacted on [email protected]

ANNEX A

D = Coefficient of Disaster Impact (includes Social, Economic and Environmental impacts + secondary & tertiary effects)

H = Hazard Type (event mortality and morbidity per location)

p = Probability (Return Interval)

m = Magnitude (intensity, duration)

r = Risk (acceptable, relative, residual, perceived, transferred)

Ɵ = Econometric of marginal utility cost (Quality Adjusted Life Year) and net excess mortality

λ = Total ‘at risk’ population (potentially) exposed

v = Vulnerability (poverty, dependency, food insecurity, disease susceptibility, demographics, infrastructural tolerance)

e = Exposure (location, population density, air pollution, water contamination, extreme temperatures, early warning, risk mapping)

Δ = Economic Loss (physical, social, political and economic capital)

E = Environmental Loss

R = Resilience (coping capacity, knowledge/attitude, access to basic services, cash, electricity, livelihoods, vector-borne diseases, (neonatal) tetanus, asbestosis, amputation; systems disruption e.g business continuity)

M = Mitigation (protection, preparedness, prevention, retrofitting, training, rehearsal)

a = Behavioural adaptation

C = Disaster Management efficiency and effectiveness (technical qualifications, experience, directive v consensual, information, communications)

---

[1] More detail on each element can be found in Annex A.

[2] Though the humanitarian world uses the term ‘coordination’ rather than ‘disaster management’, this article assumes that both terms mean the same thing. In practice, they’re very different.

[3] On average, 34% of variables relate to the social context, 25% to the disaster environment, 20% to the economic situation, 13% to the built environment, and 6% to the natural environment (with 3% to other indices).

[4] Education, Protection, Water-Sanitation, Nutrition, Food Security, Camp Management, Livelihoods, Early Recovery, Health, Shelter, Logistics, Telecommunications, Cash & Voucher Financial Assistance, Information Management.

[5] ‘Attack Rate’ refers to the number of people infected with a disease divided by the total number of people exposed. If 70 people are taken ill out of 98 in an outbreak, the Attack Rate is 70÷98 = 0.714 or 71%.

[6] A pandemic is an epidemic that spreads through human populations simultaneously in different parts of the world over the long term.

[7] DNA stands for ‘deoxyribonucleic acid’, a self-replicating material which is present in nearly all living organisms as the main constituent of chromosomes. It is the carrier of genetic information. RNA stands for ‘ribonucleic acid’ and its principal role is to act as a messenger carrying instructions from DNA for controlling the synthesis of proteins.

[8] C-19 is the name given by WHO to the disease caused by the virus SARS-CoV-2

[9] Excess Deaths are those that occur over and above those that would occur anyway over the same period, based on previous data.

[10] The Preston curve indicates that individuals born in richer countries, on average, can expect to live longer than those born in poor countries. However, the link between income and life expectancy flattens out. This means that at low levels of per capita income, further increases in income are associated with large gains in life expectancy, but at high levels of income, increased income has little associated change in life expectancy. In other words, if the relationship is interpreted as being causal, then there are diminishing returns to income in terms of life expectancy.

[11] UK Government control measures for the three months March-May 2020 inclusive are estimated at £320 billion.

[12] λ in the Calculus of Calamity equation refers to ‘Population Potentially Exposed’.

---

[i]              www.un.org/news/briefings/docs/2005/050929_nabarro.doc (accessed 26 June 2014)