Revisiting the Sandman Outrage Model
How to Consider Differences in Risk Perception
Working from Home but Missing Your Synergist? Update Your Address
If you’ve been working from home, please consider updating your address with AIHA. You can change your address by editing your profile through AIHA.org. To ensure uninterrupted delivery of The Synergist, designate your home address as “preferred” on your profile. Update your address now.
Have you ever had a conversation with a worker who didn’t want to wear their personal protective equipment because they thought it wasn’t necessary? How about someone who was panicking due to an odor in the workplace, even after you conducted extensive air sampling to demonstrate extremely low levels of exposure?
The disconnect between your perspective and the worker’s can be viewed through the framework of risk perception. Professional analysis of risk and personal perception of risk are typically different. The prototypical example is the risk in transportation. Based on miles travelled, flying in a commercial airline has less chance of injury than driving in a private automobile. Many people, however, view airplanes as riskier than cars.
THE SANDMAN OUTRAGE MODEL About 30 years ago, the risk communication professional Peter Sandman published Responding to Community Outrage: Strategies for Effective Risk Communication. This book presented Sandman’s “outrage model” of risk perception, which attempted to explain why experts and the public often disagreed about the nature and severity of specific risks.
Sandman’s ideas were influenced by environmental controversies in the early 1970s. Following the Love Canal disaster, businesses confronted community outrage over environmental issues. The outrage had many sources, including intolerance of risks that appeared to be well below other risks people accepted in daily life. For example, some people dismissed the risk of harm from not wearing seat belts but would go into the streets to protest much lower risks presented by industrial pollution.
Responding to Community Outrage summarized this discrepancy through the now famous (to OEHS professionals and other risk communicators) formulation:
This simple formula caused a fair amount of confusion. First, as Sandman explains in a footnote, the formula is more accurately presented as:
He used addition because he thought it would be more understandable to lay readers.
Second, the term “hazard” in Sandman’s formula means something different than it does to risk managers and OEHS professionals. In the Sandman model, hazard is what risk assessors call “risk”—that is, a function of exposure and the magnitude of potential harm. To reduce confusion, I will call Sandman’s version of risk “perceived risk” and designate the mathematically calculated risk “objective risk.” I will also change the addition of the original formula to multiplication, which suggests a synergistic effect. With these changes, the model becomes:
CATEGORIES OF RISK Objective risk reflects the idea, well-recognized among professionals, that risk is the combination of a hazard (something with potential to cause harm) and the resulting exposure. It relates to the mathematical likelihood of harm.
When laypeople talk about risk, however, the “absolute” value of the risk is not the key element. There are no commonly agreed-upon estimates that distinguish a “high” risk from a “low” risk. For our purposes, I’m going to place risk into three categories: acceptable, tolerable, and unacceptable.
Acceptable risk is what the general public considers safe. For example, everyone accepts and barely acknowledges the risk of getting hit by a meteorite. The risk is so low that most people will not think of it as a potential hazard. But as risk professionals, we know there is no such thing as zero risk. (We also know that one person in recorded history has been hit.)
Unacceptable risk is a level that will not be tolerated. For most people, driving at a high speed at night without headlights in a storm would be an unacceptable risk; they would slow down or avoid driving in those conditions.
Tolerable risk is a level of risk that is accepted in exchange for some benefit. In industrial hygiene, we encounter tolerable risk on a regular basis. Exposures to hazardous chemicals are accepted in workplace settings at higher levels than would be accepted in community settings. Many factors influence this distinction, such as lifetime exposure averaging and the relative health of the workforce compared to the general population. One of these factors is the benefit of receiving an income.
The purpose of our risk perception model (and Sandman’s outrage model) is to help explain why different people evaluate the same objective risk differently. Skydivers will consider jumping out of an airplane as a tolerable or acceptable risk, while most airline passengers would consider parachutes to be an unacceptable risk management strategy.
Responding to Community Outrage discussed 12 elements of outrage. Sandman presented them as “12 questions to ask in risk communication”: 1. Is it voluntary or coerced? 2. Is it natural or industrial? 3. Is it familiar or exotic? 4. Is it not memorable or memorable? 5. Is it not dreaded or dreaded? 6. Is it chronic or catastrophic? 7. Is it knowable or not knowable? 8. Is it controlled by me or by others? 9. Is it fair or unfair? 10. Is it morally irrelevant or morally relevant? 11. Can I trust you or not? 12. Is the process responsive or unresponsive?
Sandman acknowledged that outrage has many more components; the 12 he focused on were the ones he considered most relevant to risk controversies. In some circumstances, different elements might become more important.
What is the underlying source of these elements? I argue that three categories are relevant to risk perception: bias, values, and trust. These categories are lenses through which we can view each of Sandman’s 12 elements of outrage.
Bias There are many kinds of bias. Wikipedia lists more than 100 of them. When we communicate with people, all these biases potentially affect the conversation, some more than others. To communicate risk, we need to address these biases.
Cognitive bias stems from unconscious mental processes affecting how people think. It is created by the brain simplifying information through a filter of personal experience and preference. Although cognitive bias helps us process information quickly, it can result in both overestimating and underestimating risk.
Recency bias is arguably one of the most common biases. When something has just happened to us, we think that it is much more likely to occur again soon. For example, if we have just experienced an accident, we think that the same accident is more likely to occur in the near future. In industrial hygiene and safety, we call instances of recency bias “teachable moments” and use them to motivate behaviors that promote safety.
In my experience, many people are willing to consider that bias influences their risk perception. Of the three categories of risk perception, I believe bias the easiest to address through risk communication.
Understanding our audience’s biases helps us address the following elements of Sandman’s model:
Natural or industrial? This dichotomy incorporates a couple of biases. The public tends to perceive the risk of chemical products made from “natural” ingredients as relatively low even though they are not necessarily safer than other chemicals. Asbestos is natural; that doesn’t make it safe. Apophenia bias, or the tendency to make associations when they don’t exist, is likely in play. In addition, natural hazards like weather events are seen as uncontrollable while industrial events result to some degree from human agency—that is, industrial events would not occur if the conditions were not created by humans.
Familiar or exotic? This element invokes familiarity bias, or the tendency to be more comfortable with familiar options and less comfortable with unfamiliar ones.
Memorable or not? As Sandman observes, to inspire outrage, an event must be memorable. Outrage can therefore be influenced by recall bias, which affects the accuracy and completeness of our memories.
Dreaded or not? A risk that inspires dread is more likely to produce outrage. This observation relates to dread aversion bias, which is the tendency for people to seek pain avoidance over pleasure.
Chronic or catastrophic? This element helps explain why airplane crashes inspire far more outrage than car accidents. Frequency bias is probably involved: the catastrophic nature of airplane crashes makes them seem more common than they are. This is amplified by media coverage.
Knowable or not? This dichotomy is likely an aspect of loss aversion: we prefer a known loss to an unknown loss. The insurance industry is founded on the truth that most people prefer to pay monthly auto collision insurance payments (a known quantity) than accept the risk of having to pay the costs of repair following an accident (an unknown quantity).
The purpose of our risk perception model (and Sandman’s outrage model) is to help explain why different people evaluate the same objective risk differently.
Values The McCombs School of Business at the University of Texas at Austin defines values as “beliefs about what is good or bad about how people should act.” Although this definition is specific to societal values, it also applies to personal values. Conflict occurs when people have different values, which leads to different priorities. Some values are viewed as duties and not to be compromised. In industrial hygiene, we generally think of protecting the health and safety of people as a value, and it is our duty to address it. Other values, such as honesty and freedom, have intrinsic worth.
As with bias, there are many potential values; James Clear, the best-selling author who has written often about self-improvement, lists 57 values on his website. Societal values intended to determine what to do or not to do, and which contribute to normative ethics, become primary values and must be adhered to. Because they are widely shared, primary values are the ones that risk communicators and OEHS professionals can most effectively appeal to in discussions with workers. We can then consider what might be called secondary values, which may be subject to change.
In my experience, risk communication is most effective when we acknowledge the values involved in the situational risk. Often, this requires that we reduce a risk to a level different than the level we might personally find acceptable. For example, a building occupant’s perceived risk may prompt us to maintain exposure levels for a particular contaminant even lower than strictly necessary to protect health.
Most of the time, values have to be understood and accepted. When discussing the acceptability of risk, we will seldom find any success in trying to change another person’s value system.
Understanding our audience’s values helps us address these elements of Sandman’s model:
Voluntary or coerced? This element concerns freedom, which is a primary value in U.S. society and many others.
Controlled by me or others? Outrage tends to be higher in situations where we don’t have any control over the outcome. Having control is an aspect of freedom.
Fair or unfair? Outcomes perceived as unfair are more likely to generate outrage, and fairness is a widely held value.
Morally relevant or not? In Sandman’s model, risks that are morally relevant contribute to outrage. A company that pollutes a community generates a moral backlash even if the company’s individual contribution to the total pollution is small. What we consider to be morally relevant depends on our values.
Trust The other two components of risk perception, bias and values, cannot effectively be addressed without a solid foundation of trust. Trust can be described as the ability to rely on the accuracy of a source. It is closely aligned with credibility. When Responding to Community Outrage was published, trust was already identified as an aspect of risk communication. For example, the 1991 book Communicating Risks to the Public included a chapter on credibility and trust. The COVID-19 pandemic has brought trust and credibility to the forefront as critical determinants of successful health and safety efforts, as demonstrated by the recent World Health Organization webinar series titled “Trust and Pandemic Preparedness.”
Of the three components of risk perception, trust is the most challenging. A 2014 paper published in Public Understanding of Science observes that trust, like values, has a significant emotional component, which has received limited study. Gaining and maintaining trust—both interpersonal trust and organizational trust—is a challenging prospect. When lost, it is even more difficult to reestablish.
Building trust with our audience helps us address these elements of Sandman’s model:
Can I trust you or not? This element’s relevance to trust is obvious.
Is the process responsive or unresponsive? The consulting firm Stewart Leadership describes trust as “character, competence, and consistency.” Audiences are more likely to trust us if we display competence and good character. Being responsive is a form of consistency.
FINAL MODEL Starting with the original Sandman model, we have identified components that make the model more generally applicable. Incorporating our understanding that objective risk is the combination of a hazard and the resulting exposure, we end up with the following risk perception model:
Communication is the most important aspect of our job. We can have perfect exposure assessment skills, translate them into perfect risk assessments, and make perfect selection of control measures but fail completely by not understanding our audience and not communicating with them effectively. We need to be highly skilled in understanding how people assess risk on a personal level, and how they determine if a risk is acceptable, tolerable, or unacceptable. With this knowledge, we can reduce illness and injury, and improve the overall health of those we serve.
GEORGE GRUETZMACHER, PhD, CIH, CSP, FAIHA, is an industrial hygienist at Wisconsin State Laboratory of Hygiene in Madison, Wisconsin.
Send feedback to The Synergist.
The Sandman Collection
Resources and case studies from Peter Sandman are available on the AIHA website.
AIHA: Strategy for Assessing and Managing Occupational Exposures, 4th ed. (2015).
Jamesclear.com: “Core Values List.”
Psandman.com: Responding to Community Outrage: Strategies for Effective Risk Communication (PDF, 2012).
Public Understanding of Science: “Risk, Communication and Trust: Towards an Emotional Understanding of Trust” (August 2014).
Smithsonian Magazine: “For the Only Person Ever Hit by a Meteorite, the Real Trouble Began Later” (November 2016).
Springer Netherlands: “Credibility and Trust in Risk Communication” in Communicating Risks to the Public: International Perspectives (1991).
Stewart Leadership: “The 3 Most Important Factors in Building Trust with Others.”
TechTarget: “Cognitive Bias.”
University of Texas McCombs School of Business: “Values.”
Wikipedia: “List of Cognitive Biases.”