As we’ve seen over the course of this journey, the most essential and fundamental challenge of life is to strike a balance between taking risks and staying safe. We’ve learned why animals are appropriately cautious and why we should be too. But this ancient existential balance requires equal consideration of another stark truth: those who are too cautious will either starve or be outcompeted by those less cautious. Those who are too cautious will not survive.
We need to know both what to fear and what not to fear. We have a suite of specialized cognitive abilities at our disposal that themselves were the product of natural selection. We must integrate this specialized knowledge with the intellectual toolkit and framework we’ve developed by studying our successful ancestors. Formal cost-benefit analyses are often used by humans to make decisions about whether or not a risk is worth it, and these analyses are routinely used to inform policy decisions. But our evaluations of risk are often biased: they are swayed by context, history, likelihood, and expected payoffs. Understanding how we make decisions is essential for us to live wisely with fear.
We are a remarkably illogical species, and our decisions are not solely based on quantitative estimates. Why do we fear nuclear power when more people are killed by the process of transforming coal into energy, and many more are at risk from the climate disruption that results from our profligate use of fossil fuels? Why don’t we worry as much about driving fast when we wear seatbelts or have airbags installed in our cars? We overestimate our ability to get ourselves out of problems and under-plan how to avoid getting into them in the first place. So what should we fear? And how should we make informed risk assessments about these fearful things?
A simple way of viewing risk is to consider two factors: 1) the probability of something happening; and 2) the consequence if it happens. A very rare event with a modest consequence may not be that risky, while a very rare event with a huge consequence is something to thoughtfully consider. For example, contrast the risk of slipping in the shower with the risk of a nuclear war. As we age, the consequences of slipping in the shower increase, and in older people, a broken hip may lead to hospitalization, infirmity, and even, on occasion, death. Thus, it’s reasonable that our perceptions of the risk of taking a shower change as we age. We are more careful entering and leaving the shower. We may install a nonslip mat. By contrast, the risk of a nuclear war may be small, but the consequences are profound and horrific, likely affecting everyone on Earth. But except in the most straightforward cases, the exact probabilities of many events are unknown, and the consequences are a function of many variables.
To properly calculate risk, we must collect data. Then we must assess the certainty of our estimates. For instance, if many deer are hit by cars and trucks in a particular section of a busy four-lane highway, there are a number of things that can be done to reduce the risk of injury and death to deer and to humans. Signage can inform people to be aware of deer crossing the road. Speed limits can be reduced. Deer-proof fences can be installed to keep deer off the highway. If we’re really motivated, we can create overpasses or underpasses for the deer and other wildlife. Such wildlife crossings have been very effective strategies for reducing mortality in migratory populations of pronghorn antelope and other species. With time and with data, we become more certain of where risk mitigation should be deployed. Our certainty on a rural road will be diminished because we have less plentiful information; fewer people travel on these roads.
To help calibrate risks and better understand the underlying factors that drive our decision making, I offer fifteen principles of risk assessment. Some are generated from insights we’ve learned throughout the book, and others have emerged from the more specific study of human decision making.
1. Our perceptions of our own mortality risk change as we age. For example, as discussed above, the consequences of slipping in the shower increase as we grow older, and we may compensate for this increased vulnerability. By contrast, we know that teenagers do a variety of risky things that increase their risk of injury and death. This explains why your insurance goes up substantially when you add a teenager to your list of drivers. Common wisdom infers this result from teenagers’ sense of immortality. Yet, data suggest the opposite. Rather than thinking that they will live forever, teenagers overestimate the likelihood of dying in the next year. Given this perspective, it may seem acceptable for them to take larger risks if there are commensurate benefits to those who take risks—nothing ventured, nothing gained. Perceptions of longevity influence our decisions about accepting risks, and these vary by individual and by age. Growing up in a dangerous neighborhood with high infant and childhood mortality may encourage a live-fast-and-die-young mindset. By recognizing these changes in vulnerabilities and perceptions, we can make better decisions that are based on the true costs and benefits of a potentially risky action or activity.
2. Our decisions may be influenced by our pre-existing beliefs. We, like other species, learn in a Bayesian way—past experience matters! And once we get something in our mind, it’s difficult to shift it out. For instance, if we are debating whether or not guns make people safer, data can be interpreted in ways that support or reject the need for and morality of personal gun ownership. Consider that between 2010 and 2015, the United States was ranked fifty-ninth in global homicide; there were 2.70 murders per 100,000 people annually. Honduras was ranked first, with 67.19 homicides per 100,000 people per year. However, Honduras only had an average of 5,218 gun-related murders annually while the United States had 8,592. Brazil, ranked twelfth, had the most murders annually—38,494. Given these numbers, there’s both room for interpretation and room for rationalization based on your pre-existing beliefs about whether citizens should or should not possess guns.
3. We, like other species, hate to lose. We are more concerned about losing something small than excited about potentially gaining something large; we made more upset by a pay cut of $100 per week than happier by a raise of $100 per week. Loss-aversion bias, as this is called, is associated with the amygdala—the part of the brain responsible for managing our fears. Remember that the amygdala is directly activated by fearful stimuli, including a predator’s scent. By studying two women with an exceptionally rare genetic disease that damaged their amygdalae and comparing them to healthy subjects, researchers found that all had the same ability to estimate rewards. However, those patients with damaged amygdalae showed no evidence of the loss aversion that was present in the healthy subjects. These results match those found in studies of nonhuman primates, in which the amygdala has been suggested to be associated with loss aversion. We fear loss.
Marketers capitalize on this loss-aversion bias. Marketers manipulate their messages to focus on how we can avoid potential losses by buying a certain item without necessarily discussing the potential risk versus benefit. Insurance sales capitalize on this fear. Political lobbyists and those trying to influence policy also carefully word their messages because they know that how a risk is presented may change our perception of whether something with some risk is considered acceptable or not. To inoculate yourselves against this, reframe statements about loss to reflect the potential gains associated with a policy or a purchase. You may very well decide to try to avoid losses anyway, but at least you will have had the opportunity to consider the magnitude of the associated benefit.
4. Our assessments of risk are influenced by whether we accept the risk willingly or whether a risk is being imposed on us. We accept the risk of injury or death while skiing but often shun risks that our employers may make us accept. When we accept a risk, we also are sensitive to the benefits we are willing to accept. For instance, it’s fun to ski, and fun is its own reward. Similarly, we see that animals accept the risks associated with play (which takes time and energy and may increase the risk of an injury) because the benefits (improved motor skills, neurogenesis, and better condition, among others) presumably outweigh the costs. But play, in a proximate sense, is fun, and fighting is risky and scary, even though play may employ the same actions and movements as fighting. Because getting hurt at work is not high on most people’s list of fun things to do, risky jobs often require greater financial compensation to offset the risks.
Commercial fishing and crabbing off Alaska is highly profitable—if you survive. The oceans are rough, and ships occasionally disappear without a Mayday call. Many people are swept overboard or maimed by heavy equipment in a dynamic, slippery, and cold environment. Given the relatively small size of the commercial fishing fleet, it’s one of the most dangerous jobs on the planet. The US Centers for Disease Control and Prevention concur. Between 1992 and 2008 there were 128 deaths per 100,000 commercial fishermen, compared with four per 100,000 workers for all other jobs. Commercial fishing off coastal Alaska was slightly more dangerous than commercial fishing in the northeastern United States—the second most hazardous place to fish in the United States. On Alaskan fishing boats, the hazards must be compensated for. A captain on a commercial crabbing boat can take home $200,000 a year, and crew may earn as much as $100,000 per year for only a few months of very hard and dangerous labor. This increased compensation further illustrates that we’re sensitive to both costs and benefits as well as the amount of control we have over risky situations.
5. The type of outcome influences what we fear. We are more likely to fear dreadful, horrific outcomes. It’s likely difficult to induce too much concern about the risk of having yellowed toenails, which might at most influence your status at the fitness club, unless you’re a supermodel, and in that case you probably have an insurance policy. The risk of a traumatic, bloody amputation, however, will capture most people’s attention. All of us wish to avoid major, physically traumatic events.
Boeing should have known this when their new 737 Max airplanes began falling from the sky. Airline crashes have a special place in the news: we’re captivated by them. The real causes often take time to sort out, but each crash immediately attracts rampant speculation and discussion. Why? A commercial airplane crash taps into our pre-existing biases. Crammed like sardines in a pressurized tin hurtling along at nearly the speed of sound, we have no control when we fly. The outcome of hitting the ground (or the sea) at over 500 miles per hour is simply horrific; planes and bodies are pulverized into small bits. Despite these extreme and extremely rare outcomes, modern commercial air travel is exceptionally safe, and the widely cited statistic that it’s safer to fly on a commercial jet than to drive to the airport is true. Nevertheless, maintaining trust in the regulatory system that certifies planes, plane parts, and pilots is essential, and any rumor hinting of corruption or poor quality control will immediately be seized upon.
It is possible that Boeing could have mitigated some of the fear surrounding two crashes involving the Boeing 737 Max in late 2018 and early 2019 by being completely transparent about their investigation with the public and with the US Federal Aviation Administration and other regulatory agencies around the world. Had they voluntarily called for the immediate grounding of the fleet until the problem was solved, they might have avoided losing trust. The last thing any company wants is to have their product generate fear and anxiety in its customers.
6. We are more likely to fear the unknown than the known. This likely explains why nuclear power plants are feared even though there have not been any deaths associated with nuclear energy in the United States since the Shippingport Atomic Power Station began operation in 1957. By contrast, between 1999 and 2015 alone, there were over half a million gun deaths in the United States. Guns are the number twelve cause of death overall and the number one cause of death by homicide and suicide in the United States. An average of 33,400 people die by guns annually. Yet, at least as viewed through the lens of policy, we fear nuclear power plants more than guns.
As we learn more, we fear less; education can inoculate us against false fears. When HIV was first recognized to be a fatal and transmissible disease, HIV-positive patients were stigmatized and isolated because of people’s fears of contracting it by touch. As people learned that HIV was transmitted only through bodily fluids, many people’s initial fears were reduced, and patients were no longer stigmatized. Beware, however, of excessive familiarity that may breed complacency. If something is genuinely risky, it pays to keep your guard up.
7. Evolution works at the level of the individual. We should recognize that we are the product of natural selection when we pay attention to our personal welfare, the welfare of our kin, and perhaps that of close associates who aren’t kin. As we learned in Chapter 8, ground squirrels are quite sensitive to their audience when emitting risky alarm calls and are more likely to call when close relatives are within earshot. Squirrels who adopt this strategy leave more descendants than those who emit calls with less discrimination.
We too are primed to care about relatives and individuals more than population-level statistics. Most gun violence incidents have a single victim. For many of us, these victims are strangers. Media reporting that personalizes these murders and describes the suffering of their survivors may be an effective way to communicate the tragedy of gun violence and make the risks and costs more tangible. Following a terrorist attack or other mass casualty event, reports that personalize the victims are those that are most likely to help us understand the magnitude of the event. Politicians know this and often try to manipulate our perceptions of security by telling stories about individual victims, which they link to a larger problem. Their policy (border walls, health care reform, and so on) is then proposed as a solution. I saw this firsthand when Los Angeles mayor Eric Garcetti shuttered the city and instructed Angelenos to stay at home during the COVID-19 pandemic. Social distancing was presented as a measure to save our neighbors and our relatives and to reduce the burden on our health-care providers, who were desperately working to save lives. If we understand why we respond to messages that tell stories about individuals, we can then arm ourselves with data and do a better job of objectively evaluating the true risk.
8. The scope of damage may also influence our evaluation of risks. Is damage concentrated on an individual or spread across the environment? Most organisms occupy relatively small patches of the Earth and have evolved mechanisms to make assessments about things that happen in areas they occupy. Marmots are more vulnerable when they are far from protective burrows, and they perceive greater risks in areas with limited peripheral visibility. Bushbuck, however, seek dense wooded areas for safety from large carnivores. If these species’ local landscapes are modified, their risk assessments will also change.
Yet, the scope of damage that we’ve created is global. It’s extremely difficult for us to link our personal carbon footprint to the consequences of the sum total of our global energy consumption and release of carbon dioxide. We can’t comprehend the spatial scale of a nuclear war, which has a global impact. And, we can’t understand the temporal scale of a nuclear meltdown. A really bad meltdown in a populated area could easily kill or shorten the lives of numerous people while making the area uninhabitable for many generations. By personalizing the damage and by putting a human face on it, we may be able to better understand the risks. Sadly, climate-driven catastrophes are becoming much more common, and we can see all too well the suffering that follows a superstorm. We evolved from a long line of storytellers, and good stories can effectively spur us to action. To ensure that it’s the right action, consider verifying the sources of information.
9. Fear can depend on location. Elk in the northern part of Yellowstone National Park maintain a complex and temporally variable assessment of where they are likely to encounter wolves. Much like the elk, it’s to be expected that we may become anxious at certain locations where we’ve had bad experiences. Stepping back a bit, this insight can explain some odd human reactions. Janice, my wife, was driving her visiting parents to the beach one day when a red-light camera flashed. She later recounted to me that she was crossing Beethoven Street in Culver City, California, and thought that the yellow light turned red while she was still in the intersection. She dreaded the day when the automatic traffic ticket would appear in the mailbox. Remarkably, it never did. But for at least ten years after this experience, whenever we approached Beethoven Street (and only Beethoven Street), she pointed out the red-light camera. To get to Beethoven Street, of course, we’d probably driven through ten other red-light cameras that didn’t bother her. By understanding our inner marmot, we can explain some of our own peculiar behaviors and perhaps those of our loved ones.
10. Risk is context dependent. We have learned that our physiological state can influence our perceptions of risk, as we saw with the decisions marmots make about whether to emit alarm calls. This means that we should expect our state or condition and a variety of other external stimuli to influence both our actual well-being and safety, and our perceptions of well-being and safety. In the aftermath of a failed 2016 coup in Turkey, there was a widespread purge of university faculty. For this reason tenured university professors in Turkey had much more to fear from speaking their mind than tenured professors at the University of California. The fear of punishment has major impacts on personal well-being and optimal responses in a given situation.
11. We systematically overestimate small risks and underestimate large risks. For instance, we overestimate the number of deaths due to botulism and underestimate the number of deaths due to cancer. Error management theory (EMT) explains the underlying biological basis of our biases. This theory predicts that we should behave in ways consistent with the folk advice of nothing-ventured-nothing-gained when the benefits are huge, even if the chance of success is small. EMT also reveals that if the cost of a mistake is particularly large, we should behave in conservative ways. In many situations, particularly when the cost of a mistake is great, overestimating risk is generally adaptive. We saw this with Randy Nesse’s smoke detector principle (it’s better to have your smoke alarm go off when you burn your toast to ensure that it will go off when there’s a real fire), and error management theory gives us the tools to understand when and why. Our anxiety is adaptive, and having cautious responses makes sense when the cost of errors is especially high. A large empirical decision-making literature reveals specific biases. Despite all of the idiosyncrasies that make us human, formal decision theory, which mathematically calculates optimal decisions, provides a valuable framework with which we can evaluate risk.
Human decision theorists have come up with statistical methods to integrate different attributes of risk. One powerful statistical technique combines them in a way that allows us to visualize risks on two different dimensions. One dimension includes the following risk attributes: how involuntary or delayed an outcome is, how unknown it is, how well science understands it, how uncontrollable it is, how catastrophic it is, and how dreadful it is. Along this axis, risks like nuclear power, pesticides, and food coloring score particularly high while skiing, alcoholic beverages, swimming, and mountain climbing score particularly low. People make conscious decisions to ski, drink, swim, and climb mountains, but they are subjected to nuclear power, pesticides, and food coloring. A second dimension further differentiates these risks and is characterized by the degree of certainty about whether an exposure is fatal, how dreadful it is, and how catastrophic it is. Along this axis, risks like general aviation, handguns, and nuclear power score high, while home appliances, power mowers, and food coloring score low.
On these two dimensions, therefore, both guns and nuclear power are seen as certainly fatal and are perceived with some dread, but nuclear power is viewed as involuntary, and its effects are delayed. Thus, we expect different assessments about these risks, and this is why nuclear power is perceived as being more risky than guns. To properly use our fear to thrive, we must hone our risk assessment abilities. We must become comfortable making decisions under uncertainty, and we must wisely let context dictate the correct decisions. We should embrace decision theory and use it more when making consequential decisions—including about whom to vote for.
12. Sociality can buffer fear. Whether it’s by grouping to avoid predators or just listening to others, animals receive antipredator benefits from being around each other. Recent work has even suggested that social buffering can block fear conditioning in rats. I suggest that this applies to us too. Fears and other emotions can be contagious, and by being sensitive to others’ emotions we can quickly respond to threats. These responses are likely rooted in our capacity for empathy.
But these days empathy is in short supply. We are quick to point the finger at those who have different beliefs without trying to understand where they are coming from. We harden our positions, associate with those who share similar ones, and lose respect for others’ positions. Worse yet, as Sherry Turkle argues in Reclaiming Conversation, many of us avoid conversations, preferring to communicate by text, which eliminates all the social cues so important for meaningful, empathic communication. Text-based communication may be exacerbated by open-plan offices. Such office layouts are marketed as a structured way to encourage interactions, so it’s a shame that the data suggest otherwise. When Ethan Bernstein and Stephen Turban quantified how office workers communicated in open-plan and more traditional offices, they found that people retreated to their own desks or cubicles in open-plan offices and actually communicated more by text than in person.
One way to reclaim empathy is to have more dinner parties with our neighbors to discuss and debate controversial topics and the fears that underlie them. By talking with and listening to others, we can understand why people have views divergent from our own. By having meaningful conversations about our fears with others we let our capacity for empathy reduce them. Of course, social transmission is a force multiplier that could go the other way as well. So we must guard against having our fears socially enhanced, and we must guard against only listening to people with similar views. To make the best decisions it is essential to seek out contradictory perspectives to challenge our assumptions. A growing literature shows that when groups composed of diverse people with different perspectives work together, they make better decisions than less diverse groups.
13. Learning, whether it is through fear conditioning or habituation, has a profound impact on our assessments of risk. We know that animals have some innate predispositions to respond fearfully to certain things, but as we learned from tammar wallabies and a variety of fishes, experience with predators is often required to hone those abilities. For complex, long-lived species like us, we should expect that learning is an important mechanism in how we respond fearfully to various events and objects. We have biases to learn to fear snakes and spiders even if we’ll never really be threatened by them.
Our culture, which is driven by social learning, teaches us both reasonable and unreasonable fears. We should be aware of the force-multiplying effect of social learning and realize that we may sometimes learn the wrong things to fear. I believe this risk of learning the wrong things compels us to properly estimate the risks of events and base our decisions, whether personal or political, on the best evidence available.
14. We must be aware of our surroundings. Like the calling peacock or the go-away birds that communicate to their entire community that something scary is around, when in threatening situations we should be open to new sources of information to help us better estimate risk, even if it’s untraditional. Yet we should constantly evaluate these sources of potential information. From an ecological perspective, the ubiquitous connectedness between different species means that maintaining diverse ecological communities may be essential for their survival, as well as for ours.
15. We are especially vulnerable to those who seek to manipulate us with fear. Manipulating fear in others can be an effective way to motivate change, in certain circumstances. Fear can be used to nudge us into healthier practices, get us to evacuate before a hurricane, and prepare us for earthquakes. Our loss-aversion bias can be used profitably to encourage people to make properly informed decisions about the true risks of a medicine, a surgical procedure, or the costs of things like pollution.
But this knowledge can also be used maliciously, and we should guard against becoming a target and protect others from becoming targets. One way to do so is to pause and reframe the question. For instance, rather than focusing on how many people are killed by terrorist attacks annually, focus on how many people are not killed by terrorist attacks. The probability of being killed by a terrorist attack, in most countries, is fortunately quite small. Another way to inoculate people against misinformation is to show how sensitive we are to misinformation. Given this sensitivity, we can teach people to be skeptical about simple messages designed to instill fear, and to demand evidence of a simple causal link.
But we have a new challenge. We are in the midst of a great mismatch that we have created by culturally modifying our environment, and the decision rules we’ve evolved may no longer be relevant. Like the birds that can’t estimate the velocity of an approaching object above a certain threshold because they have not evolved mechanisms to avoid hitting an approaching airplane, we also have not evolved to comprehend many novel features of our modern environment. We used to fight with sticks and rocks and can’t really estimate the impact of nuclear weapons. We have evolved to modify our behavior in response to immediate changes in temperature; we put on sweaters if it’s cold and remove our hands quickly from heat. But we are unable to properly respond by modifying our behavior now to prevent even worse anthropogenic climate change at some point in the future. Yet, if our behavioral changes caused the problems, our behavioral changes must also be the solution.
We do have the power to make certain thoughtful behavioral changes in response to stimuli we’re evolutionarily unprepared to deal with. One example of such a stimulus is the twenty-four-hour updates we are constantly receiving of threatening news, available at all times on our television sets, cell phones, and computers. We may habituate to this overstimulation, but if so, we risk becoming less sensitive to important messages. Losing our fear removes something vital to who we are—we become less human. We’ve evolved to respond immediately to threats by engaging a sophisticated series of neurochemical responses that have served us and our ancestors well in the past. Yet for many of us, the constant activation of our HPA axis just causes stress and a feeling of indecision. We should be aware of these responses and, if necessary, reduce the rate at which we receive information so that we can respond appropriately when necessary.
As I age, I become more cautious. And, at a very proximate level, the structure of our brain changes as we age. A recent study found that the right parietal cortex shrinks with age and the relative size of this part of the brain is associated with risk taking. I don’t think that I’ve become more fearful per se, but the consequences of decisions have changed since I was younger and stronger and more resilient. This is natural.
While I grew up in the ocean—swimming, body surfing, and body boarding—it wasn’t until I was forty that I really began to surf. I’ve had precious moments with my family in the ocean—like the day when David and I sat in awe while two pairs of dolphins herded a huge mixed species school of fish into a bait ball beneath us. Brown pelicans and double-crested cormorants dove next to us into the feeding frenzy, emerging triumphantly with fish, only to be attacked by western gulls who, directly out of a scene in Finding Nemo, sought to claim the fish as theirs. Or the days when seals or sea lions poke their heads up next to us, fuzzy whiskers reflecting in the morning sun, curiously checking us out. Or the days when it is quiet, and the waves come slowly, and we just sit and enjoy the moment.
I know that I will never surf a big wave. I also will no longer ski down forty-five- to fifty-degree chutes after ascending them with ice axes and ropes. I will never climb that 7,000-meter peak or climb long, exposed rocky arêtes. Those days are all in my past. These days I’m quite content to hike to a pretty meadow or overlook, take a nap, and, upon waking, see the life and beauty in nature.
Now, with the responsibilities of family and home ownership, I have more to lose, and I crave more stability. Because I have something to lose, I have fear of losing whatever stability I perceive I have.
But I find it comforting to know that my fear comes from a long line of my ancestors, both human and nonhuman. It is an inherited treasure, a powerful ally. Yet, it is also an annoying and sometimes intolerable companion. It is a compass that, when calibrated properly, guides us away from danger and toward opportunity.
At some level, our relationship with fear is a lesson from life. Since it’s impossible to eliminate risk, our fears and anxieties assist us in making the right decisions. Since we cannot eliminate them, we should both embrace our fears and challenge them. As Mary Schmich, a journalist at the Chicago Tribune, wrote in 1997, “Do one thing every day that scares you.”