When COVID-19 Lockdown Ends, Will ‘Safe Enough’ Be Safe Enough for the Public?

Damon P. Coppola, MEM, ARM, Shoreline Risk

Faced with an inability to contain the COVID-19 pandemic as quickly as hoped, governments are beginning to accept that a protective posture of indefinite lockdown is unsustainable. A time will come that is well in advance of eradication when leaders, elected or otherwise, find they must declare conditions ‘safe enough’ to reopen society. But safety is as subjective a notion as fear is an enduring emotion. Simple reassurance that conditions are safe to venture out will fall far short of what is required to counter the psychological effects that months of lockdown and a constant barrage of frightening news have inflicted on the public psyche. Individuals must re-calibrate their thinking to willingly re-emerge into a world with COVID-19, but many will lack the capacity to do so effectively. It is critical that those presently managing this crisis begin far in advance of any lockdown reversals in providing the required knowledge and tools. Lessons drawn from communication science can support these efforts, but novel approaches will still be required given the unique nature of this incident.

What Do We Mean By ‘Safe’?

Safety is a fluid concept for which there exists little agreement in its meaning even among those with similar knowledge or shared experience. Considered abstractly (i.e., in answering the question ‘what does it mean to be safe?’), we might infer that ‘safe’ implies something is free of risk. In the real world, however, such an absolute level of risk avoidance is virtually unattainable. Individuals and societies are nonetheless innately capable of establishing non-zero risk thresholds for different activities, conditions, experiences, and objects for which they are able to accepted without worry. For instance, despite that countless children are injured riding bicycles each year, most parents generally accept cycling to be a safe activity as long as certain conditions exist (e.g., the bicycle is in operational condition, the rider is trained, a helmet is worn, and cars are avoided).

Derby and Keeney [1]explain that people make a mental trade-off between the existence of a risk and the associated benefits gained by accepting its presence in defining safety to be a condition “associated with the best of the available alternatives, not with the best of the alternatives which we would hope to have available.” We know, for instance, we can reduce million-plus deaths associated with automobile accidents each year by simply mandating more stringent safety standards and drastically reducing speed limits. However, the financial and convenience costs of such actions are of such significant consequence that we’ve instead made a collective social compact to deem cars ‘safe enough’ as they are. The fact that we are able to step into our cars day after day without considering the relatively high likelihood that the trip might end in tragedy is testament to our ability to effectively calibrate our fear response to a non-zero risk.

Interestingly, there is great disparity in what level of risk we consider safe for different things, even when the relative risk between them is similar or even inverted (where we are comparatively less afraid of a greater risk). Societies have banned many life-saving medications after a few hundred or even a few dozen patient deaths were associated with their use, but continue to permit the sale of tobacco products that contribute to the death of almost 500,000 people each year and chronic disease in an additional 16 million. If we calculate the relative risk of a number of different activities across the general population, subjective differences in our views on the acceptability of each is revealed despite the statistical comparability they have on our lives. The following list includes a number of equivalent population-normalized risks that most people would consider to otherwise differ with regards to their safety:

• Smoking 1.4 cigarettes
• Drinking 0.5 liters of wine
• Spending 1 hour in a coal mine (from black lung disease)
• Spending 3 hours in a coal mine (from accidents)
• Traveling 6 minutes by canoe
• Traveling 10 miles by bicycle
• Traveling 300 miles by automobile
• Flying 1000 miles by jet
• Receiving 1 chest x-ray
• Eating 40 tablespoons of peanut butter (from aflatoxin B)
• Drinking 30 cans of diet soda
• Living 20 years near a PVC factory
• Living 150 years within 20 miles of a nuclear power plant

Despite our ability to understand that these risks are statistically equivalent in terms of their impact on our long-term life safety, our fear response is variable (ranging from irrational to nonexistent). And while fear is not the only factor we use to determine safety, it is generally the most influential. We must therefore understand whether it is possible to have a situation where we are fearful of something yet are able to consider it safe.

John Locke described fear to be “an uneasiness of the mind, upon the thought of future danger likely to befall us.” While accurate, his definition is perhaps limited in that there are situations for which uneasiness is too mild a description. The fear of suffocating from COVID-19 induced hypoxia in an intensive care unit, hooked up to a ventilator, is for many people perhaps better described using emotions more powerful than unease. The following is an alternate definition that better captures the sentiment many are experiencing in the current pandemic, especially those with known vulnerabilities:

“[Fear is] a painful emotion or passion excited by the expectation of evil, or the apprehension of impending danger. The degrees of this passion, beginning with the most moderate, may be thus expressed, — apprehension, fear, dread, fright, terror.” [2]

This definition accurately depicts how individuals may experience a spectrum of emotions in response to the danger posed by COVID-19 — not only to themselves but also their friends, family, and community. Almost everyone feels at least some level of apprehension to the crisis, and for many the extent of their fear vacillates significantly from day to day. Fear in this case is therefore not a simple uneasiness of the mind, as Locke suggested, nor is it always a paralyzing condition. It includes a wide range of emotions that are determined by several factors including the individual’s state of mind, the complexity of their living situation, the mood in the press, and even the weather outside that day. It is, with few parallels, a reaction to a perception of a profound lack of safety.

Is Our Fear of COVID-19 Justifiable?

Life was by no means idyllic or risk-free before Coronavirus mutated into this current strain given there were other health hazards, social hazards like crime and terrorism, natural hazards, accident hazards, and much more that could happen to us. Despite the constant presence of these hazards, most people considered it ‘safe enough’ to venture from their homes, to interact with others, or to spend hours each day in their workplaces. In a matter of a few short weeks, this general sentiment has almost completely reversed in keeping with the associated state and local policy actions mandating people remain at home. A visit to the grocery store or pharmacy is today marked by palpable fear, embodied in the jarring display of masked faces and gloved hands. Danger is suddenly perceived in each other not as a matter of our physical appearance or actions but rather our potential for being infectious. That these profound changes were so quick to have been adopted in nearly every community is indicative of record societal fear. And so new is this fear-driven arrangement that we are only just beginning to ask whether it is justifiable.

In their celebrated article Rating the Risks, social scientists Paul Slovic, Baruch Fischhoff, and Sarah Lichtenstein [3] begin, “People respond to the hazards they perceive.” The way societies have responded to the pandemic at the personal, local, regional, and even national levels would therefore indicate that the factors that shape our perception to the current crisis exist at extreme levels. But fear and perception are imperfect indicators and cannot be used to assess or confirm actual risk. In many cases our fear response is justified and serves us well, such as when we see a tornado approaching and run for shelter in response. But most situations where the danger is not so obvious our fear will be exaggerated or insufficient and the misalignment can stunt our ability to react in an appropriate or proportional way. To assess fear and risk in a given situation, Slovic and his colleagues provide four possible sources of misalignment which they term risk perception fallibility conclusions. They include:

1. “Cognitive limitations, coupled with the anxieties generated by facing life as a gamble, cause uncertainty to be denied, risks to be distorted, and statements of fact to be believed with unwarranted confidence.”

The more uncertainty there is surrounding a hazard, the more it is feared. As people gain knowledge about the associated risks their concern generally wanes. Months into the COVID-19 pandemic, there is still very little understanding about how deadly the virus is, how contagious it is, whether immunity is conferred, how long the crisis will last, and much more. It appears to exist in almost every public place, in every country, every state, and every community. It can strike anyone, anywhere, and at any time. We are thus presented with perhaps the ultimate in uncertainty.

Facing high uncertainty, people typically make personal judgments based on very imperfect information to establish a risk level for themselves. More often than not, they overstate their own risk in the process. The data gathered in response to the COVID-19 crisis shows countries differing considerably in their case fatality rates (CFR) which range from one to ten percent. In the absence of population sampling, and as a result of testing shortages, inaccurate test products, and other testing challenges, the actual fatality rate is not known (though there are indications it may be less than 1 percent given the high number of asymptomatic and mildly symptomatic cases that have since been identified.) Individuals thus have no way to assess how their own COVID-19 risk compares to other health risks they face without or with little concern. Given the precedent-setting levels of resources dedicated and restrictions enacted, they are left to assuming that the threat must likewise be precedent setting in its severity.

2. “Perceived risk is influenced (and sometimes biased) by the imaginability and memorability of the hazard. People may, therefore, not have valid perceptions even for familiar risks.”

People are more afraid of things they easily imagine, remember, or hear about repetitively, and overestimate the likelihood of these so-called available risks. This phenomenon is referred to as the Availability Heuristic. This perception bias can be correct when applied to hazards that are, in fact, frequent, such as automobile accidents. However, when an uncommon but spectacular risk receives constant media attention, such as the 2002 Sniper Attacks in Washington DC or perhaps the ever-so-rare shark attack, people often wrongly assume such events are likely to impact them in a similar way.

In terms of availability, there is no event in history that compares to COVID-19 in terms of the flood of information and anecdotes people receive from all communication channels. Whether on television, the radio, social media, the internet, in journals, in communication from local and state leaders, and through interpersonal communication, people are bombarded with frightening images and stories which include ICUs packed with intubated patients and refrigerated trucks filled with cadavers. Moreover, the regular sight of others donning masks and gloves and reminders to maintain social distance in one’s own community are constant and very personal reminders. People are prone to exaggerate their own risk and easily imagine themselves succumbing to a similar fate when relentlessly assailed with such imagery.

Slovic and his colleagues explain that events which are “out of sight [are] effectively out of mind.” It follows then that the opposite has been true with the COVID-19 pandemic; its ever-presence ensures its dominance in our thoughts. An early-April poll conducted by ABC/Ipsos found that 89% of Americans are somewhat or very concerned about catching the virus, and more than 90% reported having changed their routines as a result. Another poll by the Kaiser Family Foundation found that nearly half of Americans reported that the fear of coronavirus infection was so severe it was harming their mental health.[4]

3. “[Risk Management] Experts’ risk perceptions correspond closely to statistical frequencies of death. Laypeople’s risk perceptions [are] based in part on frequencies of death, but there [are] some striking discrepancies. It appears that for laypeople, the concept of risk includes qualitative aspects such as dread and the likelihood of a mishap being fatal. Laypeople’s risk perceptions were also affected by catastrophic potential.”

Statistics are not the sole influence on how the public perceives a risk. In fact, qualitative factors often play a much larger role in how people rate and rank the risks they perceive than quantitative ones. This is because even for risk experts, it is difficult to fully internalize how risk statistics affect one as an individual. In the absence of visual cues, we struggle to conceptualize the difference between a “one-in-a-million” and a “one-in-one-hundred-thousand” chance of occurrence. It is why a person can buy a Powerball ticket and a pack of cigarettes on the same errand, legitimately believing they might win the lottery despite 1 in 292 million odds yet discount their 2 in 3 odds of dying from smoking related illnesses.[5]

Even where COVID-19 statistics do exist, including many updated on a real-time basis by the media and various health organizations (e.g., the John’s Hopkins University Coronavirus Resource Center), we still must closely examine their accuracy and utility. Throughout the ongoing crisis the public has been reminded daily of ‘confirmed’ case numbers in each country and US state, and of the number of reported deaths. These statistics have presented several major problems in terms of personal risk perception. For instance, it has not been explained in any effective way how the differences in testing methods and availability from country to country, and state to state, affect the comparability and significance of these numbers. In countries or states where only those who present with the most classic COVID-19 symptoms are tested, more mild cases will not be captured. Then, when the number of deaths is measured against this numerator, the exhibited case fatality rate is artificially high. For example, there are currently 28,300 deaths from 636,000 confirmed cases, which gives a 4.5% fatality rate. However, if asymptomatic and mild cases represent 90 percent of all COVID-19 infections, as some studies suggest, then the actual number of infected is likely 10 times greater (or more). A potential 6.36 million cases with 28,300 deaths instead gives a 0.45% fatality rate. This represents a very different risk to the average person.

The data also fails to show how other common illnesses, again like influenza, compare in terms of the local and global infection rates, and the absolute number of people who have perished as a result. As of April 15, for instance, influenza has caused as many as 56 million illnesses in the current cycle, 26 million of which required visits to a doctor, up to 710,000 which required hospitalization, and upwards of 60,000 that resulted in death.

But if numbers are not what are primarily shaping fear from COVID-19, what makes the disease so frightening? Slovic and his colleagues identified seventeen unique qualitative fear factors in their article Facts and Fears: Understanding Perceived Risk. These measures help risk communicators to better assess how the fear from COVID-19 might compare to other risks people are facing — many of which they have come to accept as-is. In the following table, risks that exhibit the left form of each pair of characteristics are likely to cause more fear than those that exhibit the right form. For example, hazards that we cannot easily see, like this virus, are more feared than those that we can, such as a cliff. As displayed in the following analysis of COVID-19, wherein the exhibited form of each pair of fear factors is shaded, it should not be a surprise that the pandemic has caused such a strong fear response. In fact, in this regard it ranks among the most fear-inducing threats such as terrorism or crime.

4. “Disagreements about risk should not be expected to evaporate in the presence of ‘evidence’. Definitive evidence, particularly about rare hazards, is difficult to obtain. Weaker information is likely to be interpreted in a way that reinforces existing beliefs.”

The COVID-19 pandemic burst into our daily consciousness over a matter of days. Within a month of the first case in the United States, several cities were facing local ‘hotspot’ conditions, none worse than New York City. By this point, graphic imagery of the crises occurring in China and Italy had already caught the public’s attention. Expectations of how this threat would affect people’s lives were established in their minds well before the virus ever took hold on a national scale. And when it did, with a meteoric increase in infections exceeding all other countries, any lasting hope that the emergency might be avoided was dashed. The collective public fear appeared to have been confirmed, and people quickly came to appreciate the potential for catastrophe.

It is extremely difficult for an individual to alter the initial frame of reference they establish for a risk, especially when it is reinforced over time. In the current crisis, the initial assessments people made about their own risk have been unceasingly reinforced with a flood of both indirect (communicated) and direct (experiential) stimuli. There do exist articles that have been published with the intent to better calibrate public risk perceptions, including those that attempt to explain risk disparities between different age groups (e.g., an almost complete lack of risk for those under the age of 10), that report how as many as half of those infected show no symptoms, or that explain how increased population sampling is likely to indicate a much lower virulence of the disease. The majority of articles, however, express a generalize sense of danger which overshadows all other viewpoints.

There is actually very little we know with certainty about the virus, even as governments begin to explore reopening. There has been no successful attempt by a country to conduct population sampling, which would help to determine whether many people might have already contracted the virus but had never known thereby driving up observed fatality rates. Without population testing, there is no means to assess whether a community is likely to be virus free, or to understand how prevalent it is if cases do exist. And there is no means for a person to determine with any accuracy their likelihood of surviving infection in light of their age, gender, physical characteristics, medical history, and other factors.

The fear response people are experiencing would thus appear to be fully justified. Moreover, the persistence of fearful conditions is something few people are likely to have previously experienced. It is oftentimes only time that is able to change a person’s opinion about risk, and in fact the primary reason people are more afraid of new risks than old ones is that they have not had time to gather enough information to calibrate their initial impressions. We can expect that through time, if people’s initial expectations are not realized (especially if they were extreme), they will begin to question their views. As an increasing number of people either become sick with the virus itself or know friends or relatives who have, and in keeping with the statistics a large majority of these cases resolve without hospitalization or death, the fear response will adjust. But where fatalities do occur, and where individuals know or hear of those who have perished (including through the media), this will serve as a tangible confirmation of one’s internalized fears and only serve to strengthen them.

What level of COVID-19 risk is safe?

As with other hazards, society should not expect to achieve safety in the face of COVID-19 characterized by a zero level of risk, but rather at a level of risk that is associated with a tolerable balance between intervention costs and benefits. In other words, safety will be found at the point where the majority of society is unwilling to sustain or increase the financial, social, and other forms of sacrifice required to further reduce any residual risk. We have not yet determined what that level is, but there exist guidelines to help us in doing so.

It is first and foremost our goal to protect all citizens’ most basic physiological needs, which includes their health and safety. Without a viable COVID-19 vaccine or treatment the next best option is to reduce transmission. Unlike bloodborne or vector-borne viruses, the airborne nature of the virus that causes COVID-19 requires much greater restriction on even the most casual human contact as is achieved with lockdowns and stay-at-home orders. As a mitigation measure, lockdowns can be effective at breaking the transmission cycle, but they carry monumental costs that impact our ability to address our wider scope of human needs.

The most obvious need that has been impacted, and the most widely reported impetus behind efforts to quickly reverse lockdown orders, is the need for financial security. While many people in office and tech jobs can easily continue working from home, many people in the service industries, hospitality, transportation, manufacturing, agriculture, and many other sectors are not. Just a few weeks of lockdown has led to the loss of tens of millions of jobs — the most rapid decline in the nation’s history. Extending these losses for several more weeks or months will assuredly cause an economic depression and perhaps even a mass humanitarian crisis.

Maslow’s Hierarchy of Needs

Beyond finances, there are many other needs that are sacrificed to support a lockdown order. Psychologist Abraham Maslow mapped out the full extent of human needs in 1943, and examining these provides a more comprehensive understanding of the full costs of lockdown as a COVID-19 intervention measure. Maslow depicted the extent of human needs as a pyramid, with the most vital physiological needs at the base (food, water, health, shelter, clothing). For the time, these life-sustaining resources are what response efforts have targeted. Above physiological needs are safety needs, which include personal, emotional, and financial security. Calls to reopen society to enable people to return to work are reflective of the criticality of these second-tier needs. The third level of human needs includes social belonging such as friendships, intimacy, and family. Social distancing has required that people betray the human desire for physical contact and togetherness with friends, family, classmates, and neighbors, and this is causing an increase in loneliness, stress, and depression. The next level of need is self-esteem, which is primarily the result of our feeling we are useful and recognized in society. Social distancing is keeping people from most of the hobbies, sports, recreational activities, educational opportunities, and opportunities for professional achievement that we need to promote self-esteem and self-worth. And at the top of the pyramid is self-actualization, which is what we perceive to be the realization of our greatest potential, including the achievement of our goals.

We cannot expect that control of a risk like COVID-19 will come without costs, but given that the threat from the virus is not existential we cannot sacrifice all other needs to prevent it. The costs of lockdown as just described are extreme. Lockdown also creates new risks that didn’t exist previously or that are magnified as a result. For instance, the economic pain of income and investment losses and a decrease in government capacity will together lead increases in crime. Impacts on social programs that support mental health, domestic abuse, drug abuse, and other social challenges will result in further pain and suffering (including assault, abuse, psychological injury, and suicide). Those with chronic and non-COVID related diseases and injuries are foregoing medical treatment due to a reduction in medical system capacity, and prevention programs (e.g., routine physicals and dental cleaning) are being put off. Reductions in recreational activities and exercise are causing physical and psychological effects that negatively impact health.

W. Kip Viscusi [6], in the article “Economic Foundations of the Current Regulatory Reform Efforts”, describes how risk acceptability is often driven by more than simple economics. He found that there are cases where societies are willing to control a risk even though the financial costs are excessively high, especially when there are strong political and social influences on the decision. Using a cost-benefit threshold of $2.8 million per life saved as a socially-acceptable measure, he produced a list of regulations were passed despite having failed such a “test” (where the cost should be perceived as greater than the benefit). For comparison, he produced a similar list of regulations that had passed the same test. Examples are shown in the following tables:

This research showed that there are situations where societies are willing to expend huge financial costs to save lives. But that all depends on who incurs those costs. As previously described, societies are generally unwilling to increase the cost of automobiles any more to increase the number of lives that would be saved, despite that more than one million die and many more are injured as a result of the residual risk. In the end the decision is not one that is quantified in absolute dollar terms. Risk management decision making processes are very imperfect, and in fact biases and “injustices” are an unfortunate outcome in the consensus process through which acceptability levels are established and treatment options are chosen. There will always be winners and losers in this bargaining process. There are three things that decision-makers can keep in mind while driving this process to ensure they achieve the greatest level of acceptability in the measures ultimately implemented:

1. It is vital that risk acceptability processes are not overly influenced by those with financial and other vested interests.
   The risk acceptability process, including the selection of interventions, is influenced by politics, corporate interests, and social norms. This is painfully obvious in the struggle to identify a common level of acceptability for climate change, firearms, and even construction in high-risk coastal zones. Political influence in the case of COVID-19 lockdowns is complex given there are divisions in ideology according to how restrictions have impacted individuals’ income and investment potential. It has also become an indicator of how leaders are being judged for their ability to successful managing the crisis. The degree to which citizen participation can be increased in this process will determine the extent to which the outcome is considered acceptable.
2. The public generally considers dollar-valuation of human life to be unethical.
   To those whose lives are at greatest risk, any dollar figure will seem low or inappropriate as a trade-off for the acceptance of the risk. Most people feel there is no cost to great to save their own life, and can only consider this type of analysis in the abstract. This works well for voluntary risks such as smoking in public places as there exists a choice. However we cannot simply choose to end the existence of COVID-19. Despite that it will be possible to calculate a probable cost per life saved, any effort to frame the argument according to such figures will be (and has already been in the case of COVID-19) fiercely contested by the public.
3. Risk management is an undemocratic process if those who may be harmed by the outcome are not asked if the danger is acceptable to them.
   It is important to understand who benefits the most and who is harmed the most when risks are mitigated. It’s easy to recall past cases where vulnerable or disadvantaged groups were exposed to risks whose benefits were enjoyed by others, including the placement of toxic dumps or oil pipelines. With COVID-19, vulnerability is not just a matter of economics or job type, but also one that cuts across all socioeconomic groups (e.g., age, presence of underlying health conditions, capacity of local medical system). Some people of very different political ideologies may be in close agreement when it comes to the selected measure, while others who generally agree on policy decisions find they see the lockdown issue differently. Those who are vulnerable will consider there to be more at stake in opening society than others who have a lower perceived or actual risk of dying, especially if alternative measures are not implemented to address the continuing prevalence of the virus in the community.

When determining treatment options for COVID-19 or any risks faced by the community or country, the starting point is to consider whether the risk might be safe if left as-is. We typically do this in situations where a risk is of such low consequence that any level of resources spent to reduce it would be better dedicated to hazards of greater significance. For the four common coronavirus strains that cause run-of-the-mill colds in humans, and have so for decades, this is precisely the approach that is taken. Because this particular strain is so contagious and has a higher case fatality rate, accepting the risk as-is is not a condition most would consider to be safe. It is why there is so much resistance to simply ‘opening up society’ without an associated plan of incrementally-decreasing protections. We are therefore left to consider which of the more restrictive or responsive options would provide an acceptable and thus safe level of risk. These include:

• The “no go” alternative
  Complete elimination of a hazard’s risk is generally not possible. With technological hazards such as those associated with a particular chemical, we can simply choose to no longer allow production or use of that chemical. We cannot, however, simply choose to no longer allow COVID-19 to exist. Once a vaccine or treatment exists, the potential to eliminate the virus may finally exist. But even then, as we have seen with influenza, vaccines do not guarantee eradication.
• Establish a “de minimis risk” level of risk
  De minimis interventions are those that establish a statistical risk level below which people are not likely to be concerned. A standard often applied is one in one hundred thousand or one in one million chance of occurrence over a selected time period (usually a year or a 70-year lifespan). This concept widely used for technological hazards, which include the risks that are associated with mitigation measures for natural hazards (including biological threats like COVID-19). For instance, de minimis risk will likely come into play if when a vaccine is developed there is a health risk associated with people receiving it. Because there is a theoretically high number of people who would receive the vaccine (perhaps 330 million people in the US alone), some might feel even a one in a million risk would be too high given this would result in at least 330 US deaths and as many as 7,000 worldwide. If or when a vaccine is finally developed COVID-19, society will have to decide whether the death of 7,000 otherwise healthy individuals is an acceptable cost to ensure the safety of billions including millions of COVID-19 related fatalities that might be prevented.
• Establish a “de manifestis risk” level
  Related to de minimis risk is the concept of de manifestis risk, or “obnoxious risk.” With this option, a risk level is chosen above which treatment at all costs becomes mandatory. A common level that is used is one in ten thousand deaths per exposed individuals. With estimates that as much as 70% of the global population may be infected, and early estimates of CFR hovering between 1 and 3%, there could be as many as 1 in 50 killed by the virus. In the absence of other options, a lockdown is an attempt to control the risk at all costs. Continuation of this posture is likely until more complete, timely, and accurate information is available and decision makers can formulate more balanced interventions.
• Perform a cost–benefit analysis
  The most severe impact of COVID-19 infection is death, and so any cost benefit impact will have to assess what cost is acceptable to save additional lives. As previously mentioned, such analyses are rarely seen as acceptable by the public, especially given it is their own lives that would be weighed against costs in this example. Where property damage and economic costs are the primary factor assessed, perhaps this might be done, but that is not the case in the present crisis.
• Determine acceptable risk to be the best choice among alternatives.
  Derby and Keeney [7]write that the answer to determining what level of safety is ‘safe enough’ is a factor of “what alternatives are available, what objectives must be achieved, the possible consequences of the alternatives, and the values to be used.” In the absence of a vaccine to eradicate the virus, an effective treatment that ensures survival, or some other miracle cure, this last option is the most likely to result in an exit from the current lockdowns. What life may look like under an incremental opening is assured to be very different from either the current stay-at-home measures or the extent of freedom that existed prior to it’s being imposed. It will include a more structured, incremental process of easing and tightening that is responsive to actual risk levels and which balances control of the epidemic and a functioning of society.

Given their fear, the public is most likely to consider a path to safety as one that acknowledges and addresses the existence of a risk they know to be present (even if it is unseen). People are aware that every moment of our lives entails risk, and so we can expect that few people will demand a life entirely free of COVID-19 risk (or risk from any viral or bacterial illness at that). People deserve an accurate assessment of how the actions they will be required to take will positively or negatively influence their own infection likelihood, what the associated costs of those actions might be, and what the likely consequences will be if those measures ultimately fail (and they are infected). We can compare this to airline travel. We know that accidents do occur, and many people have died as a result. We also know that more accidents will occur in the future. Rather than ban air travel and eliminate all benefits it provides, we regulate safety through the application of standards known to have a positive effect on risk reduction, and we recommend specific behaviors and actions for different populations to help them assess their own risk and use air travel if they so choose.

With COVID-19, we can likewise choose to address risk through actions that are not as blunt and restrictive as an extended societal lockdown. To establish a sense of safety that may be associated with the outcome of any such interventions, there are the two conditions that must be true:

1. The risk associated with COVID-19 must be assessable with a degree of accuracy that allows individuals to compare it to other risks for which they accept without worry. This is especially true in light of the benefits that will be gained through one’s agreeing to conditions that will potentially expose them to the risk (namely the benefits of emerging from lockdown). To calculate this, a monumental increase in testing capacity and conduct will need to occur. Through increased testing, much more can be learned about how different people have been affected by infection with the virus, including methods of transmission and probabilities of different outcomes (no affect, mild illness, hospitalization, permanent impacts, and death).
2. There must exist opportunities for people and groups of varying vulnerability to act in a way that they consider to be appropriate for their own vulnerabilities. In other words, those who believe that they stand to lose more through exposure (e.g., death or permanent disability) than the average person must be able to take actions they still consider to be safe. For some, this will include maintaining a lockdown stance until a vaccine or effective treatment is developed, and so to receive their support for an incremental opening there must exist measures to support these holdout populations (e.g., support for grocery shopping, income support, secure transportation services). People whose jobs require them to be exposed to high numbers of people will also need to have easy access to personal protection equipment (PPE) including masks and gloves. And people will need to know that there exists a legal and regulatory framework to ensure that all public and private organizations are held accountable to agreed-upon safety measures to control virus transmission including minimum space requirements in common areas, limits on event size, regular sanitation of surfaces and equipment that are shared (e.g., in stores, restaurants, theaters, or schools), and others as determined in the scoping process.

If, and only if, society determines that the residual risk associated with an ongoing presence of COVID-19 in their community or country is acceptable, and individuals have the capacity to establish how that risk applies in their own situation, then life with a non-zero risk of COVID-19 can theoretically be considered to be safe.

COVID-19 and the Difficulty of Communicating Risk in Reverse

The fear response that people have exhibited thus far will not so easily be countered when the time comes to reverse them. Slovic and his colleagues found that, “people’s beliefs change slowly and are extraordinarily persistent in the face of contrary evidence. New evidence appears reliable and informative if it is consistent with one’s initial belief; contrary evidence is dismissed as unreliable, erroneous, or unrepresentative.” They add that, “Demonstrating safety would require a massive amount of evidence” [8].

Most of the time, risk communication seeks to raise awareness of a risk and influence people to take increased action accordingly (good examples include the click-it-or-ticket and September Hurricane Preparedness Month campaigns). In the case of COVID-19 lockdown easing, the goal will instead be to promote an increased level of risk taking. People will need to be coaxed from the safety of their homes despite their fear of a very real hazard. The methods that are required to achieve these two very different responses, however, will ultimately be the same.

The goal of risk communication is to change attitudes on an issue for the purpose of creating, reinforcing, or changing responses. This is often called persuasion. Much of the difficulty of communication science is discovering the right technique to deliver a message in a way that makes it most persuasive. Among the most important factors is the degree to which recipients are involved with the issue being addressed, or stated otherwise, how much people feel that decisions about the issue will have direct implications for their lives and the things that are important to them. There is arguably no issue with a higher level of collective involvement than the current pandemic. And those who are highly involved with an issue are much more likely to carefully scrutinize the message itself, such that the degree to which a message persuades them will be based mostly on the strength of the argument presented. They will also be more likely to grasp upon inconsistencies and question any counter-intuitive facts presented.

There has also never been an issue that has so dominated the news as the current pandemic, just as no issue has so dominated policy agendas in every country and society as the current one has. Campaign planners are thus targeting audiences that are not only highly involved, but also highly informed. Unfortunately, much of the information received thus far has focused on avoiding public places and the company of other people, and this must all be counteracted. When communicators know beforehand that their highly involved audience already holds unfavorable attitudes toward the behavior(s) being presented, there is a risk in presenting any viewpoints that counter the audience members’ beliefs. For example, imagine a community located downstream from a dam that has been reinforced to decrease hazard risk to the community. Also imagine that there has long persisted an awareness among the community’s members that the dam in its previous condition could have burst and inundated the town, destroying property and threatening residents’ lives. Any communicator, whether from outside or within the community, risks low credibility if they fail to address these existing perceptions head-on. People have strong pre-existing perceptions about coronavirus and the disease that it causes, so pursuit of any messaging approach that is counter-attitudinal to those beliefs will backfire when attention becomes focused on counter-arguing (i.e., criticizing the points that are made, finding things wrong with them, and settling on an even more extreme viewpoint against the message).

To achieve behavior change in the face of pre-existing knowledge and fear about COVID-19, people will need a diverse set of cognitive, social, and emotional skills that enable them to understand the information they are receiving, to interpret its relevance for their own lives and communities, and to articulate their views to others. Target populations, like risk communicators, have their own goals that frame and influence how they interface with the messaging efforts. While there will always be a certain segment of the population that pays absolutely no attention to communicators or their messages, most will recognize how the messages pertain to their own lives and seek more information upon which they will base any future action. How they go about doing this and what kind of information they seek will differ among the different issues they face. For those who do elect to pay attention to the communicator’s message, there are generally three areas that describe their intentions:

1. Advice and Answers
   This group consists of people who want only to be told exactly what to do. They need step-by-step instruction on how to address the problem of which they have just been informed. This group of people would rather receive the product of analysis than be given all of the necessary information to draw such conclusions themselves. This can be compared to the faith individuals have in advice provided by a trusted doctor, lawyer, or other confidant.
2. Numbers
   There are certain people who will not take advice at face value. Rather, this group would prefer the option to draw their own conclusions about what they need to do from a body of statistical evidence that the communicators provide to them. This group would rather be faced with quantitative summaries of expert knowledge than to be fed step-by-step instruction offering little room for individual analysis and perspective.
3. Process and Framing
   This group is composed of those individuals who prefer to fully analyze as many factors relevant to the problem as possible, assuming full ownership of their actions and any likely outcomes that result.

Conclusion

Lockdowns are safe, but they are costly. They represent their own source of risk, and they will have to end eventually given the complex cost of sustaining them. It is difficult to explain to the public that the best path forward is one in which people may die, or for which we have not achieved as significant a reduction in risk as we might hope under more ideal circumstances. Risk decisions are a part of our daily lives, and for most of these we can proceed with very little thought. But for risks wherein there exist no obviously correct or straightforward choices — especially where human lives are at stake regardless of the intervention — the decision making process will be contentious. It is precisely this type of problem that our democratic process, which entrusts difficult decisions in the hands of elected leaders, is meant to address. And while we should respect the difficulty they face in making such decisions, we must also ensure they are made with the required ethics, diligence, and consideration to ensure a viable and acceptable outcome.

Safety is a sentiment that cannot be established through assurance alone. There is a segment of the population that will be willing to accept any prescribed option irrespective of the risk reduction achieved on account of their trust in the communicator. But most will demand the means to confirm for themselves that their fear response may be eased. Communicators must become acutely aware of how the priorities and goals of their target audiences do or do not match those of their own and use this knowledge to ensure that a diverse notion of safety is achievable. By ensuring the intervention shaping process has identified and accommodated the goals of all or most recipients, and that communication strategies are applied which indicate to recipients that communicators are cognizant of their goals, there will be a lower likelihood that target audiences reject messages in their entirety. There will be some recipients who will never feel safe until COVID-19 is eliminated, or whose goals are directly opposed to those of the communicators (and may even run counter to what those tasked with managing the crisis are trying to achieve), but this is expected and okay. By ensuring that the process of opening the country remains conducive to an ongoing dialogue, and there exists flexibility in the intervention measures such that individuals can adapt as they see fit to fluctuating risk conditions, it will be possible to re-establish an enduring national sense of safety.

Damon P. Coppola, MEM, ARM is Principal of Shoreline Risk LLC and author of Communicating Emergency Preparedness, Introduction to International Disaster Management, Introduction to Homeland Security, and several other emergency management professional texts.

• For more information on the risk communication processes described in this paper, see: https://bit.ly/34JEVkd.
• For more information on framework proposals to support incremental easing of lockdown restrictions, see: https://bit.ly/2JYk833.

[1] Derby, S.L., Keeney, R.L., 1981. Risk analysis: Understanding “How safe is safe enough?”. Risk Analysis 1 (3), 217–224.

[2]The Medico-Legal Society of New York. 1905. Medical Jurisprudence. The Medico-Legal Journal. Volume 23. https://bit.ly/2RJ8rBD

[3] Slovic, P., B. Fischhoff, and S. Lichtenstein. 1979. Rating the Risks. Environment Vol.21 №3. Pp. 14–20, 36–39.

[4] Fredericks, Bob. 2020. Nearly 9 in 10 Americans Fear Getting Coronavirus, Poll Finds. The New York Post. April 3. https://bit.ly/3epDdsW; Achenbach, Joel. 2020. Coronavirus is Harming the Mental Health of Tens of Millions of People in the US, New Poll Finds. The Washington Post. April 2. https://wapo.st/2RHtIvp

[5] Paquette, Danielle. 2015. The Terrifying Rate At Which Smokers From Smoking. The Washington Post. February 26. https://wapo.st/2VJNEz9

[6] Viscusi, W., 1996. Economic foundations of the current regulatory reform efforts. Journal of Economic Perspectives 1 (3), 119–134.

[7] Derby, S.L., Keeney, R.L., 1981. Risk analysis: Understanding “How safe is safe enough?”. Risk Analysis 1 (3), 217–224.

[8] Slovic, P., B. Fischhoff, and S. Lichtenstein. 1979. Rating the Risks. Environment Vol.21 №3. Pp. 14–20, 36–39.