# Statistics Can Be Misleading, Especially During a Pandemic

This article was written by Ronnie Klein for the International Insurance Society, a sister organization of Insurance Thought Leadership, under the umbrella of The Institutes. To see more IIS articles by Ronnie and other IIS experts, visit internationalinsurance.org.

There is a saying in German: “Traue keiner Statistik, die du nicht selbst gefälst hast.” This translates as, “Do not believe any statistic that you have not forged yourself.”

Many credit this saying to Winston Churchill, but Nazi propagandists actually made up the quote and attributed it to Churchill as a way of impugning him as a liar. It appears that “fake news” is not a new phenomenon.

An example of misleading statistics is when determining whether to take a medical test for a rare but serious disease like spina bifida. This rare disease causes the spine of a baby to form improperly and can lead to serious mobility impairments and possible organ malfunctions. Many doctors will recommend that the patient undergo a blood test to detect this disease. The test has improved over time and is now 95% accurate.

This sounds like an easy choice. The test is 95% accurate and can detect a horrible disease. But let’s explore.

The probability of contracting the disease, according to the U.S. Centers for Disease Control and Prevention (CDC), is 1 out of 2,758. Therefore, out of 1 million pregnancies, there should be approximately 363 babies, or 0.03%, born with spina bifida.

Assuming that all 1 million women opt for the test and that no false negatives occur, there will be 363 actual positives and about 49,982 false positives ((1,000,000 – 363) x .05) for a total of 50,345 positive tests. Receiving a positive result now means that the baby has a 363 in 50,345 chance of having spina bifida — or 0.7%.

Does this sound like a test with 95% accuracy?

Further, once a woman receives a positive test for fetal spina bifida, she must undergo follow-up tests that are a bit more invasive to more accurately determine the status of the baby. However, those additional tests take time to schedule and to generate results.

How much stress is the woman under during this time? What effect could this have on the  unborn child? None of this is usually discussed with the mother.

How can a test that is 95% accurate change the probability of contracting the disease from 0.04% (363 out of 1 million) to 0.7% (363 out of 50,345)? Should an expectant mother take a test for a disease that will affect 363 babies out of 1 million? Armed with the  correct statistics, an expectant mother will be much better prepared to make an informed decision.

COVID-19 statistics can also be misleading. The most common misstatement is that the disease only kills the elderly. According to the most recent data from the CDC at the time of this writing, 80.5% of COVID-19 deaths have occurred in people ages 65 and over in the U.S. On the surface, this seems like a daunting fact.

Of the nearly 540,000 U.S. deaths attributed to COVID-19 as of this writing, almost 435,000 are from people age 65 and over. Breaking it down further, 58% of all COVID-19 deaths occur in  people age 75 and over. It is no wonder that most of the attention has been given to the most vulnerable people in these age groups. Why worry about those under age 65 when only 20% of COVID deaths can be attributed to this cohort?

Examining mortality by age for all causes shows that the statistics for COVID deaths do not vary greatly from all-cause mortality. Said another way, COVID deaths by age are highly correlated to deaths by all causes (see Figure 1).

Figure 1: COVID-19 Mortality vs. All-Cause Mortality by Age Group

Limiting the analysis to age groups over 25, which basically eliminates infant mortality and teen auto accidents, shows an even stronger correlation (see Figure 2).

Figure 2: COVID-19 Mortality vs. All-Cause Mortality by Age Group (25+)

While the media is quick to broadcast that approximately 80% of COVID-19 deaths occur in people over age 65, it fails to state that, in a given year, almost 75% of all-cause mortality occurs in the same age group.

Exploring the data for those ages 25 and up shows that people ages 75 and over account for 59% of all COVID-19 deaths and 56% of all-cause mortality – not far off. This virus is not only a worry for older people, it affects younger adults in a similar proportion to other causes of death.

What this means is that the target life insurance-buying population (people ages 30-60) should be very interested in purchasing life insurance to protect against this and future pandemics. COVID-19 increases mortality for adults of all ages at similar percentages. However, this very important fact is not widely broadcast in the news. And the life insurance industry has remained relatively silent on this topic, as evidenced by the continued flat sales of life insurance during the pandemic.

In the largest life insurance market in the world, the U.S., premium sales in 2020 actually dropped while number of policies showed a slight increase. Considering that there are no infectious disease exclusions in the vast majority of life insurance policies and that the world is in the midst of the worst pandemic in the past 100 years, one would think that sales of life insurance would be skyrocketing.

Every life insurance sales person will be familiar with the term “share of wallet.” Potential customers only have so much disposable income, and only a portion of that can be allocated to life insurance. While the pandemic should certainly highlight the need for life insurance, the ensuing financial crisis brought on by travel restrictions, hotel closures, restaurant closures and other lockdowns make certain that a person’s share of wallet is more focused on food, housing, medical supplies and other essentials. Life insurance has  been moved further down the list.

A survey performed by the U.S. Census Bureau revealed that more than 60% of low-income families experienced “income shocks” during the pandemic. This includes food insecurity and delinquencies on rent or mortgage payments. The percentage is even higher for families with children (see Figure 3).

When choosing between paying rent or purchasing life insurance, there is no question at all. However, as bad as things are for these families, it will become much worse if the breadwinner dies due to COVID-19.

Figure 3: Share of Families Experiencing an Income Shock by Household Income and Presence of Children

The U.S. Congress recently passed the American Rescue Plan, which provides aid to all citizens and disproportionately helps low-income families. A family of four earning less than \$150,000 per year received \$6,400 in cash and possibly other benefits, including extended unemployment, tax credits and lower health insurance premiums.

If a family of four is earning \$50,000 per year, this is more than a 12% increase in pay — and the money has already arrived. Similar packages have been offered in most developed countries in the world that are experiencing the same adverse mortality and economic downturn as the U.S.

These low- and middle-income people are suffering from a huge protection gap. One estimate places the middle-market protection gap in the U.S. at \$12 trillion (see Figure 4). This is exactly the market that the life insurance industry has been talking about addressing, but failing to reach, for decades.

Wouldn’t this be a great opportunity to approach these people, as they receive a relatively sizable lump sum of cash? A small term policy that costs less than one-per-thousand for most of these ages would help to protect those families most in need.

Figure 4

But the window for action by the insurance industry is short, as these funds have already been distributed. This money will not sit around waiting to be spent on life insurance.

Selling pure protection to the middle markets during a pandemic is a great opportunity for the customer and the insurer. It provides much-needed protection during a time when excess deaths due to COVID-19 in the U.S. are estimated at about 16% (see Figure 5). It benefits insurers as a way to reach a market that has thus far eluded the insurance industry. And, it will help inform the middle markets of the importance of life insurance and may win over many customers for life.

Figure 5

The life insurance industry has long lived by the motto, “Let sleeping dogs lie.” During a 1-in-100-year pandemic, would it be worthwhile to attempt to make a change and tout the industry’s many benefits – especially to middle-income families? For example, would this be a good time for life insurers to contact all of their existing policyholders to remind them that policies are valid for death due to COVID-19? In-house lawyers can put in all of the caveats such as, “assuming all premium payments are current, assuming the policy is not accident-only, etc.”

J.D. Power performed a life insurance survey in late 2020 and concluded that “…a combination of infrequent client communications and a pervasive perception of high cost and transaction complexity have suppressed consumer interest and customer satisfaction with life insurance providers.”

Following the results of the survey, Robert Lajdziak, a senior consultant for J.D. Power, said that policyowners’ satisfaction with their life insurance products declines the moment the sale is completed. Lajdziak showed his surprise that this trend would continue during a pandemic and implores the industry to “rachet up” its client contact, not just its communication with agents.

To some, telling customers that they are covered for death due to disease is not necessary — but is this really the case? Just one Google search with the tagline, “Does life insurance pay for COVID-19 death?” will show how many articles have been written on this subject.

Why do customers need to get this information from a third party? While there may be risks to offering it to in-force policyholders, the benefits of this positive communication could dramatically outweigh these risks.

The life insurance industry protects policyholders from the financial hardships of premature death or disability of a breadwinner. This is especially important during a pandemic. However, sales of life insurance have been flat in most mature insurance markets for decades.

If a pandemic that is responsible for about an eighth of all deaths of people ages 25-64 cannot generate interest among the general population to purchase insurance, at a time when lump-sum stimulus payments are being made to lower-income earners, it is difficult to imagine what will cause an increase in sales. But consumers will not run to purchase this insurance. The industry must think of a coordinated, thoughtful and compelling message.

Epsilon Marketing estimated that there are about 50 million middle-market households in the U.S. A survey performed for this report revealed that reaching the middle market was a top priority for 25 of the 35 life insurance companies that responded. This survey was performed in 2014, so these companies and others had approximately seven years to work out a plan to reach this market.

Now is the time to “pull out all stops” and market aggressively. Doing that will generate sales and create an entire class of new life insurance purchasers who will be able to tell positive stories in the future. Starting this process may be as simple as communicating with existing policyholders about the benefits of their policies. Word of mouth among friends may be the best sales channel to reach the underserved middle markets and to help close the protection gap.

Selling more life insurance during a pandemic can bring peace of mind to customers and  help protect their families. Yet, with all of the talk about new technologies to market, underwrite and speed policies to customers, there has been virtually no perceptible increase in life insurance sales.

This can be easily evidenced by QualRisk’s assessment that, in 2020, all-cause mortality increased in the U.S. by 16%, but there was only a 3% increase for individual life insurance. Some in the industry may look at this as a favorable outcome. What it really shows is the vast protection gap that exists in the U.S. and in all mature insurance markets in the world. It is time to do something differently and reach underserved markets. Now is a perfect time to begin.

# Insurance and Financial Protection

If insurance is a matter of predictive analysis, the most accurate prediction requires no analysis. The prediction is a sentence of pith and precision, economizing words while encompassing phenomena throughout the world. The prediction is a law, Stein’s law, whose namesake, Herbert Stein, was an adviser to presidents and the father of a speechwriter, Ben Stein, to a succession of presidents. The prediction says, “If something cannot go on forever, it will stop.”

What will stop is the obvious: a runaway stock market in which distance is infinite, speed limitless, direction ascendant and principal impenetrable. That people believe this trajectory is not only real but sustainable, that they believe this trajectory transcends Stein’s law as well as the laws of economics and common sense is reason for insurers to act.

What insurers must do is save the many, so people may protect their savings. Failure to act will result in a stoppage of incalculable stress and immeasurable loss, where tomorrow’s retirees have no way to retire save expiring en masse, save death in lieu of destitution.

Without life insurance, those who need a safe and tax-free source of retirement income will not have one. Without life insurance, those unable to work will be unable to live. Not without assistance from the government. Not without a concomitant rise in prices and collapse in dignity. Not without a recession in the economy and a sense of fear among the people.

Avoiding this scenario starts with alerting the public to the imminence of this threat.

The alert must be clear, thanks to the simplicity of the message and the frequency by which it repeats itself. The writing and delivery of the message is a charge the insurance industry must assume and a role it must accept. The costs are minimal, in comparison to the consequences of inaction.

To sound the alarm is to admonish the public, telling the many what to do and where to go.

To sound the alarm is not to sound like an alarmist, because the crisis we face is too extreme to exaggerate.

If the crisis is hard to understand, we must make it easy to comprehend. If the crisis is hard to define, we must make it easy to describe. If the crisis is hard to divert, we must make it easier to diminish.

The insurance industry must lead us through this crisis.

Steadfast in its dedication, the industry can secure its place in history; strong in its devotion, the industry can make history.

Saving a generation from financial ruin is a moral duty. The duty insures lives, providing for the defense of liberty and the pursuit of happiness. The duty demands exemplary counsel, by people of excellent character, for people in exigent circumstances.

The duty belongs to the insurance industry, allowing it to earn what no amount of advertising can buy: trust.

Honoring this trust enriches the lives of millions, exceeding the combined wealth of a thousand billionaires, because safety is a treasure of inestimable worth.

# Rational Ignorance and the Protection Gap

A recent study by Tali Sharot, a professor in the Department of Experimental Psychology, University College London (UCL), asked the following: “Would you want to know if your colleagues view you as incompetent?” She found that 55% of respondents would want to know. When the same group was asked the opposite question—“Would you want to know if your colleagues view you as competent?”—the answer was a whopping 80% yes.

Intuitively, that more people prefer to know if they are judged as competent rather than incompetent is not surprising, but it is somewhat illogical. That your colleagues think you are competent might put a temporary spring in your step, but it is not very useful. However, knowing you are viewed as less than competent is potentially much more useful. It could lead to productive long-term changes in attitude and behavior that might save your career.

The results of Dr. Sharot’s survey also provide a classic example of the natural human propensity to turn a blind eye to truly useful information, especially if the information is cloaked in bad news. Dan Ariely, professor of psychology and behavioral economics at Duke University, coined the term “predictably irrational” to describe this tendency.

Rational Ignorance Is All Around

The phrase “rational ignorance” was coined by American economist Anthony Downs in his 1957 book, “An Economic Theory of Democracy,” an early treatise on rational choice theory (and, incidentally, the first to use the left-right axis to political thought). Dr. Downs defined rational ignorance as “refraining from acquiring knowledge when the supposed cost of educating oneself on an issue exceeds the expected potential benefit that the knowledge would provide.”

This concept is particularly associated with politics. One illustration can be seen in the recent municipal elections in Sao Paulo, Brazil. Brazil is one of the few countries in the world where voting in government elections is mandatory. The Brazilian political landscape is also truly byzantine, with representatives coming from 33 different political parties, and the president currently having no party affiliation.

It is not surprising that many Brazilian citizens (myself included) do not have the time to devote to researching every aspect of every candidate’s policies. Indeed, several people wind up using social media posts from political pundits, who are seen as more versed in the issues, for their information, or relying on politically astute friends for suggestions. This relative apathy, however, does not mean that the voters are making poor or biased decisions. Still, such reliance on secondary sources can be seen as “rational ignorance,” as the cost of acquiring direct knowledge about the candidates, the issues and their positions might seem much higher than leveraging someone else’s knowledge and might produce the same decision either way.

Heuristics

Now we come to heuristics. The definition of a heuristic is a rule, method or concept that helps people solve problems faster than they could if they did all the necessary research or performed the necessary calculations to do so. Behavioral economics views heuristics as a driving force behind the human decision-making processes.

Let’s consider a simple example: How does the human brain react when shown pictures of i) a long multiplication problem or ii) an angry man? A person might take a couple of minutes to solve the former, but the latter will immediately trigger a subconscious fight-or-flight reaction. Human brains have, over time, evolved heuristics that instantly recognize an angry face as dangerous.

Have people come to treat social media as a heuristic, believing that posts from pundits and the like can help them navigate complicated political, economic and scientific issues? Two documentaries released in 2020 on Netflix, The Social Dilemma and The Great Hack, go into depth on how social media can manipulate news and politics and stoke conspiracy theories, leaving people less informed, misinformed and irrationally biased.

If people have a tendency to use rational ignorance regarding politics, I have no doubt rational ignorance is also a factor in the superficial understanding most have of personal taxes and investments, diet and exercise, health issues (including reducing the risk of COVID-19) and life insurance.

Insurance Product Offerings and Contract Wording: Mind the gap …

The phrase “coverage gap,” heard often from life insurance company executives, is defined as  “the shortfall in the amount of life insurance cover necessary to maintain the current living standards of dependents.” Life insurance companies devote extraordinary amounts of time, effort and expense trying to educate underinsured individuals about the need to protect themselves and their families from this gap by buying more cover. Could our industry not be addressing one of the key issues leading to the lack of consumer enthusiasm for our products?

Here’s the issue: Insurance products and contracts are not consumer-friendly. To the average person, life and living benefits products are at least as byzantine as Brazil’s political system, and the language of insurance contracts could almost be considered an actual dialect. Insurance is thus fertile ground for the manifestation of rational ignorance among potential customers, who are already known to be more likely to pay attention to information about insurance if it comes from friends and social media posts. (I pity the buyer researching concepts and options such as pure protection, accumulation, critical illness, disability income or long-term care.)

Financial education could be one key to lowering this barrier. The good news is that there are plenty of options for obtaining information, from traditional media such as TV, radio and newspapers, to new media such as podcasts, YouTube videos and blogs. The bad news, however, is that there are plenty of options. Too much information frequently leads to even more confusion.

Many insurtech startups address obstacles by offering quick transactions, clear policy language and highly expedient claims processing. A possible downside, however, is that these startups generally offer only limited coverages. However, the rethinking by insurtechs can be considered tiny steps in the right direction.

Toilet Paper vs. Life Insurance?

One might think COVID-19 would have placed life insurance ahead of toilet paper or disinfectant wipes on the “most desired items to buy during a pandemic” list. I am sure life insurers are still scratching their heads, wondering why the sales tsunami they expected turned out to be only a ripple. Rational ignorance may have squelched any uptick in propensity to buy life insurance.

I would argue that the insurance industry needs to acknowledge rational ignorance as a major sales obstacle. Doing so could be a first step in recovering from the industry’s addiction to complex and layman-unfriendly policy language. Focus will be the key to a successful recovery. The first mover advantage, however, is still up for grabs.

This article first appeared in the March 2021 issue of Society of Actuaries` Reinsurance News.

# Closing the Protection Gap

Drought spells disaster for farmers across the developing world. Most lack insurance because conventional crop insurance is too expensive (where it is available at all). No rain means no income, no food and not enough resources to replant next year. With many countries from sub-Saharan Africa to Southeast Asia already facing an abnormal recurrence of climate risks, natural disasters around the world caused \$232 billion of economic losses last year. Only a small fraction of this was covered by insurance.

Many developing economies depend on improving the productivity and resilience of sectors, including agriculture and tourism, that are vulnerable to climate hazards such as cyclones, heat waves, droughts and flooding. With economic losses from catastrophes growing faster than insured losses, adapting individuals and economies to climate-related impacts has become a major societal priority, outranking other risks like aging populations, terror attacks and social unrest. New insurance products designed to create disaster-risk-financing mechanisms, where no other risk-transfer tool is available, are increasingly being seen as part of the solution in closing this protection gap.

Many of these products are parametric, as opposed to indemnity, otherwise referred to as traditional, insurance. Increasingly recognized as a valuable form of transfer for climate and other natural disaster risks, parametric insurance contracts are based on objective and transparent indices, such as cyclone wind speed, earthquake shaking intensity or amount of rainfall, and payments start to be made as soon as the index reaches a preset threshold. As no costly visits are required to assess the losses, payouts can be made quickly to hard-to-reach insureds in remote locations. Crucially, protection against unpredictable but potentially devastating risks — previously unthinkable with traditional insurance — is now possible.

Rapid relief

For the insured, ease of use, speed, certainty of amount of payout (without dispute) and the resulting ability to plan ahead ensure more rapid relief when disaster strikes, which in turn increases financial resilience.

For the insurer, parametric insurance allows for a more scientific pricing of products that respond to specific isolated parameters, rather than the physical losses that might result from any number of a wide range of occurrences. Together with lower claims management costs, this scientific approach makes lines of business commercially viable that were not previously.

Contrast this with a complex insurance claim on a traditional policy, which may take a long time to adjust and be paid, due to the need to develop claim details and financial components as well as address issues like valuation and other conditions.

A hard sell

This does not solve the issue of how to persuade people in Africa, who are generally somewhat insurance-averse, to buy insurance. This form of risk management requires a certain level of prosperity, as it means spending money on something you hope you will never need.

However, this challenge can be surmounted. African governments and policymakers understand the cost-benefit analysis, combined with the experience from developed countries showing that insurance can play a cost-effective role in a country’s efforts to increase disaster resilience.

The African Risk Capacity (ARC), launched by the African Union in 2013, demonstrates what is possible. Set up as a mutual (known as ARC Ltd.), and designed to provide rapid payouts to covered nation state members, initially for droughts but planned to include floods, tropical cyclones and epidemics, the insurance pool started with four countries and now has a dozen or so policyholders and 34 member states. ARC has become highly efficient in pooling risks and their transfer at very low marginal cost to the global reinsurance markets and now protects tens of millions of people.

This is a remarkable achievement when you consider that until recently the concept of selling droughts in Africa to the global reinsurance market would have been unthinkable.

Insuring the ‘uninsurable’

Parametric insurance solutions have mostly been used in the reinsurance space around catastrophe risks, but the boundaries defining what is “uninsurable” are being increasingly pushed to new limits. A policy covering hurricane-related damages to coral reefs was purchased in 2018 to cover a part of the vast Mesoamerican Reef along Mexico’s Yucatán Peninsula. Once verified, the agreed policy would be paid within one week. Such a rapid disbursement of funds is crucial as much of the initial reef repair following a severe storm needs to be done very quickly to avoid further damage and set up a successful recovery.

Not just for developing countries

These same innovative parametric applications being adopted in emerging economies also have significant potential in developed markets. In contrast to emerging economies, where the problem is more likely due to cover being unavailable, businesses in developed countries are increasingly seeking protection against losses for which traditional insurance is not best-suited.

Emerging climate risks are a key driver behind this growing demand for more innovative insurance products in both the public and private sectors. The impact from a crop loss following a major weather event or supply chain delays is smaller in developed markets, but still significant. The Bank of England, for example, downgraded its expected first quarter GDP growth from 0.4% to 0.3%, following the impact on businesses from the “Beast from the East” – the cold weather snap that hit the U.K. in the winter of 2018.

Whether it’s reducing the protection gap, financing resilient infrastructure or improving risk management and return optimization across the financial sector, insurance technology and innovation has a decisive role to play in responding to climate risk and smoothing the world’s transition. While protection gaps remain an issue as greater costs are borne by the uninsured, these gaps are closing slowly. Innovative risk transfer structures and new products based on parametric triggers have a key role to play and will continue to help increase resilience of households and companies to growing climate risks.

# COVID-19 Risk and Buyers’ Psychology

Despite all the attention devoted to the continuing devastation from COVID-19, compliance with mitigation measures is not turning out to be universal. Even public officials have been bending the rules. Indeed, a recent study from the U.K.-based behavioral science consultancy Dectech found that nearly a third of those surveyed had spent time with someone outside of their household despite restrictions on such socializing.

Similarly, many families are underinsured, suggesting that perhaps they do not see the value of such financial protection, or do not understand their risk in not having it.

Is the problem that people just don’t understand the risks of this pandemic, or is the issue with risk in general?

A handful of studies published since the summer have examined the psychology of how the public perceives the pandemic’s risks. What can these results tell us about the public’s perception of COVID-19 risk, and how can this information help insurers understand consumer mindsets and needs?

Understanding the threat of COVID-19

When people understand that a reasonable threat exists they will generally be more likely to take actions to avoid it. This was shown to be the case during 2014’s avian flu epidemic, and COVID-19 research has consistently been telling a similar story — the larger the threat people feel, the more likely they are to take protective measures such as washing hands, wearing masks, social distancing and even panic shopping.

But the bigger challenge may well be the public’s sense of the threat in the first place. In one U.S. study, half of the respondents substantially underestimated their risk of dying if infected. Perhaps more worryingly, older respondents and individuals with underlying health conditions also underestimated the virus’s potential threat despite understanding that they were at higher risk than the population average.

A well-established finding from behavioral economics is that humans are often poor at understanding their own risk objectively. For example, a majority of people implausibly believe they are better than average drivers, and many also believe their chances of avoiding cancer are better than other people’s, despite no objective basis for the belief. This comparative optimism — the belief that negative events are more likely to happen to others — might also explain how people are perceiving the risk of COVID-19. A study early in the current pandemic on perceived risk and self-reported personal behavior conducted by researchers at the California Institute of Technology found that participants perceived a high level of risk to the general population but tended to rate themselves as being at lower risk than the average person. This perception was not influenced by whether the respondents were actually at lower risk than average.

This finding, replicated by a group of researchers from the U.K., U.S., Europe and Australia using the U.K. data, also suggests that, despite the inescapable factual information available about COVID-19, people still believe “it won’t happen to me.”

Perhaps most interesting is the finding that factors beyond objective facts influence COVID-19 risk perception. According to a study from the University of Cambridge of almost 7,000 individuals in 10 different countries, although participants had relatively high threat perception, men, who are statistically far more likely to die from COVID-19, were less likely to report feeling threatened by the pandemic than women. The researchers also found that risk perception can be amplified by how a person’s friends and family are perceiving the risk and by their having existing prosocial values (values that promote behavior benefiting society as a whole). The Cambridge study also showed that the most powerful predictor of COVID-19 risk perception was direct personal experience with the disease.

Clearly, despite all the objective information we have at our disposal, visceral experiences — i.e., concrete evidence obtained and interpreted through our own senses, emotions and social interactions — and existing attitudes are powerful drivers of people’s reactions to external threats.

Is risk knowledge or a feeling?

These studies paint a picture of reactions to the COVID-19 pandemic that are consistent with psychologists’ view of risk perception: Perceiving a threat may push people to take action, but perceiving a threat accurately is not guaranteed. While perception of COVID-19 risk is relatively high on the scales used in the studies, it is still underestimated and subject to biases such as over-optimism and social influence, and context such as pre-existing worldview, direct personal experience and gender identity.

Risk, to humans, is subjective — fundamentally about what one is feeling. Fear, disgust and the prospect of guilt and regret turn people away from danger, and, when not sure how to respond, humans look to other people’s behavior and their own expectations for answers. This is how people experience the world first-hand, and this experience, rather than objective facts, frequently dominates judgments and decisions.

To understand the physical effects of COVID-19, seeing someone in a hospital bed who is either intubated or on mechanical ventilation is likely to have a far more powerful visceral effect on a person than case numbers. People can empathize with the horror of the situation and grasp the risk in a way that knowledge of the numbers just cannot achieve.

Perhaps this was the idea behind one U.K. city’s campaign to promote compliance with COVID-19 restrictions. Its shocking slogan, “Don’t Kill Grandma,” sought to elicit an emotional response to promote a personal understanding of COVID-19’s risks rather than presenting statistics and then expecting a highly rational and analytical response.

Communicating to inspire protective action

Should communicators use shock tactics such as fear appeals to help people gain an appropriate understanding of risk? One of the most widely discussed uses of fear appeals, i.e., tactics designed to elicit emotional discomfort and motivate particular responses, are the campaigns to push smoking cessation. Cigarette packets in the U.S. are required to display warning messages, and in the U.K. must have both warning language and harrowing images of the physical effects of smoking, in the hopes that user exposure to these horrifying images may promote behavior change.

Fear appeals can be effective, but there are serious caveats. Overly shocking communications can make people disengage and ignore or react defensively to the intended message altogether. Finding the right balance of engaging emotions while sustaining people’s attention is vital for risk communicators. In particular, being able to show that a risk is personally relevant and that one’s own actions can minimize the danger of a particular risk are key elements of making fear appeal messages work.

Insurance marketers face a similar challenge in persuading consumers to protect their assets and finances. It can be off-putting to communicate risk by scaring potential customers. Also, the industry is dominated by a highly objective and numbers-based understanding of risk, but, as the COVID-19 research shows, people do not usually think of risk statistically.