Rapidly changing conditions and increasing uncertainty of future climate projections make extreme and hard-to-predict conditions challenging to manage the Delta and its watersheds effectively. Similarly, changing social, policy, and economic conditions can alter resource use and desirable management approaches, sometimes substantially.
The Delta Independent Science Board (Delta ISB) has been holding a webinar series on Decision-making Under Deep Uncertainty to explore the scientific tools and concepts that can increase the capacity to anticipate and adapt to the growing uncertainty of future conditions in the Bay-Delta system. The DISB hopes to aid managers, stakeholders, and interested parties in evaluating and planning for a wide range of plausible futures in the Sacramento-San Joaquin Delta.
Decision-making under deep uncertainty is the exploration of tools that can be used to anticipate risks and inform decisions, even when uncertain how likely those future risks may be. Scenario planning has emerged as a key tool for such investigations, so this and the next webinar will focus on understanding the approach and how to get the most out of scenario planning efforts.
This seminar featured Andrew M. Parker, Ph.D., Senior Behavioral and Social Scientist and Professor at Pardee RAND Graduate School, and Jody CS Wong, Ph.D., Associate Policy Researcher at the RAND Corporation. Their presentation covered the fundamentals of scenario-based planning, the benefits of using scenarios, and how scenarios can serve as communication tools for developing a shared understanding of key uncertainties and supporting decisions.
Cognition, cognitive biases, and scenario-based decision-making
Dr. Andrew Parker began with a presentation on cognitive biases.
Over the years, there’s been intense interest in decision support approaches to help planners improve their ability to make advanced types of difficult decisions, such as when considering uncertain long-term environmental changes. To understand the differences between scenario-based and traditional forecast-based approaches, it’s useful to consider the broad set of elements key to effective decision-making processes. The figure on the right is a very simple characterization of some of those processes.
We often spend most of our time talking about choice processes, which include selecting the best decision among available options, but implicit in that are a variety of processes where we evaluate the potential consequences of those options. And even more implicit at times are the decision structure and tasks. These include defining the problem in a way that opens it up through meaningful consideration, objectives that we want to achieve, assembling a menu of options that might achieve those objectives, and identifying any key sources of uncertainty.
Dr. Parker noted that the term ‘decision structuring’ is often used in two different ways. One is equivalent to decision architecture, meaning design choices are made by decision analysts in helping decision-makers structure the decision that they’re facing. The other way is that the decisions are made by the decision-makers themselves on how they want to think about the problem that they’re facing. This presentation will mostly focus on how people frame a decision problem themselves.
“I argue that traditional approaches focus primarily on choice and evaluation,” he said. “But many real-life decisions focus as much on discovering information and designing new alternatives as making the final choice. This is particularly true under deep uncertainty where decision-makers have uncertainty regarding the very nature of the problem that they’re facing.”
Scenarios and scenario-based planning
Scenarios are a set of future states of the world presented to decision-makers as plausible and worthy of consideration, but without initially assigning any likelihood. Scenarios often come with a clear and often narrative policy orientation. Scenarios arguably may facilitate decision structuring, at least in the early stages of deliberation.
However, Dr. Parker cautioned that there are concerns that scenarios may prove insufficient for effective choice, particularly when decision-makers ultimately require likelihood information about alternative futures to choose among decision options. In some more quantitatively oriented DMDU approaches, the likelihood does enter the process once choices need to be made, but initially, the exploration and narratives surrounding scenarios could potentially benefit decision structuring. He noted that scenario literature and scenario proponents argue this is especially helpful for contentious group decision-making.
A favorite quote of Dr. Parker’s that captures the proposed promises of scenarios is, “Scenarios can change decision-makers assumptions about how the world works, compelling them to reorganize their mental models of reality.” This references a concept from psychology of mental models, which is a network of beliefs about something that could be a risk, process, or decision that they have to make. Mental models are a useful way of organizing the world and can act as a shortcut for decision-making. So, the idea that scenarios have the potential to help decision-makers form one or more mental models of the situation they’re facing is a powerful concept.
Regarding scenarios, there are several recommendations:
- Consideration of multiple rather than single futures: When faced with deep uncertainty, a single future or distribution of futures may be more easily rejected by stakeholders with opposing views, while scenarios that focus on multiple distinct features may help to get stakeholders onboard and engaged.
- Focus on possibility rather than likelihood: If presented as possibilities rather than predictions, scenarios may be less psychologically threatening and thus easier for decision-makers to explore new decision options that might effectively address inconvenient or contentious futures.
- Increase concreteness, impacting processing styles: Parker said that concreteness can be cut both ways. Scenarios may draw attention to small probability, high consequence outcomes, which can be under-attended to by decision-makers, but at the same time, could mislead decision-makers that such low probability scenarios are more likely than they actually are.
- Focus on policy options rather than expectations: How information is presented can affect several things, such as how intuitively or systematically users process information, what they regard as default options, their focus or lack of focus on worst cases, and their willingness to invest in protective measures. If decision-makers are given policy options, scenarios have the potential to explicitly build that into their deliberation.
- Present uncertainty across rather than within scenarios: To the extent that likelihood is brought to decision support at a later stage, this is typically characterized across rather than within scenarios, which likely changes how some uncertainty is mentally represented.
The main argument is that pursuing scenario-based versus forecast-based approaches doesn’t just matter from a decision analysis perspective; it matters psychologically about how people approach a problem. Perhaps most importantly, decision-makers’ view of the decision will be shaped by the task they face.
Cognitive bias
“From a decision architecture perspective, how it’s presented by a decision analyst will affect how they approach the problem,” said Dr. Parker. “Individually, the mental model the decision-maker adopts for the decision will affect this. More generally, there’s a truism in psychology that behavior is a function of tasks in person, specifically the interaction of tasks in person. So, in this case, the decision-making behavior is a function of the decision task we face and the decision-maker.”
From a task perspective, is the future singular or multiple? Is likelihood placed at the forefront, or does it play a secondary role? Is the starting point expectation or policy? This is often mirrored in cognitive processes. So, is the decision-maker focused on likelihood or possibility, is the perspective one of an average future of multiple futures, and is future diversity more concrete or abstract?
Behavioral decision research originated in psychology but has strong threads in economics, business, engineering, and other fields. This research is often focused on the many pitfalls in human decision-making, especially when talking about narrow conceptualizations of complex problems and especially when uncertainty is not well structured. Indeed, a central tenet of this field is the contrast between the descriptive account of human behavior and normative models of decision-making, such as those coming out of economics, statistics, or engineering.
Some examples of the types of cognitive biases that might come into play in scenario-based approaches:
- People typically bias their processing of information towards prior beliefs; this is called my side or confirmation bias. Psychological distance is a mechanism that may also affect evaluation and perceived likelihood. Arguably, traditional approaches typically require users to agree on a probability distribution before evaluating the desirability of alternatives, which may be difficult with contentious groups.
- Overconfidence in “what is known” worsens with task difficulty (and groups): There’s a general tendency for people to overestimate their own knowledge. This overconfidence can worsen as tasks become more challenging under deeper and deeper uncertainty. This is true for individuals but has been shown to get worse in group deliberation. One of the intriguing aspects of scenarios is that they may serve as counterfactuals, which have been demonstrated as one way to combat overconfidence.
- Brainstorming in groups can be subject to an effect called production blocking. This is an effect whereby interacting groups in a brainstorming setting actually brainstorm fewer ideas than the individuals in the group would have produced on their own. Group discussion limits the generation of alternatives because if I’m focused on discussing your idea, I’m not generating my own. As a largely deliberative approach, this could be a concern for scenario-based planning.
- Framing matters, such as characterizing an outcome as a gain versus loss. There’s a wealth of research in psychology and other related fields on framing effects. The most well-known example comes from characterizing an outcome as a gain versus a loss; people tend to be more risk-taking for losses and risk-averse for gains. Scenarios that are framed correctly may help combat this.
- Cognition and affect are intertwined whenever considering risks: When talking about risk perception, something called the affect heuristic is a mental shortcut that people take when they’re trying to judge how risky something is. Perhaps they don’t have the time or the inclination to process that more completely. They query their emotions about the situation: am I worried? Am I nervous? Am I upset? These are taken as proxies for something being risky. The narrative nature of scenarios very likely draw in affect quite effectively. And that’s something to be aware of when using these approaches.
What we know and don’t know
Dr. Parker said there is very little in behavioral literature about cognitive biases. “You’ll note that I used the terms ‘may’ and ‘could’ a lot. That’s because these effects have been largely under-studied within the context of scenarios.”
The reasons Dr. Parker cited include the small number of people engaged in this type of decision-making, the price of their time is high, and the elements of the tasks they are involved in are complex and difficult to capture in an experimental setting.
One exception was an experiment with a realistic fishery management task that asked participants to balance economic and environmental objectives. This was accomplished through a computer-based dashboard that allowed participants to explore a wide range of futures and construct policy options. The options were intentionally large; there were actually 79 different possible combinations of policies, but they could explore 18 that were identified as having relatively high expected value. Participants were randomly assigned to one of two conditions: they either received a display focused on scenarios (shown on the top) or a display focused on forecasts (shown on the bottom). They were then allowed to explore and asked to choose a policy.
“We found the participants in both conditions made reasonable decisions; this was actually quite a comforting result,” said Dr. Parker. “They explored the space similarly and in equal numbers found high expected value solutions. Where they differed, though, was that those in the scenario condition were more likely to identify five of those 18 strategies that we had identified as and designed as more robust, i.e., they performed well over a wide range of futures. This is largely because they paid increased attention to worst-case futures.”
“As a secondary result, those in the forecast condition compared to those in the scenario condition were more likely to endorse a statement that they tended to envision a single future and plan with that future in mind. So using scenarios rather than forecasts encouraged users to consider a wider range of futures, and the full information of the forecast approach, which was the only one that actually provided concrete likelihoods, didn’t appear to help.”
Cautions regarding scenarios
Dr. Parker then wrapped up his presentation with some cautions.
- The narrative nature of scenarios may make it easier to recall or imagine those scenarios, which, in turn, may increase perceived likelihood. Hence, they should be chosen carefully. In particular, they may legitimize multiple viewpoints. While some rare cases may be worth consideration, creating a false sense of equal likelihood may be one potential harm.
- Narrative scenarios often involve a conjunction of details, each adding realism, and may implicitly take advantage of the conjunction fallacy, which is when the conjunction of events feels more likely than either of the events themselves. For example, a hot urbanized future may feel more real and hence more likely than either a hot future by itself or an urbanized future by itself. Scenarios as complex narrative objects are inherently conjunctions of events, and care should be taken to be cognizant of this effect.
- To the extent that decision processes are deliberative, protection should be taken to avoid production blocking, which occurs when consideration of discussed ideas blocks the production of other ideas. Luckily, some solutions have been identified. One is to first brainstorm individually and then discuss the compiled list. Another remedy is heterogeneous groups, as they introduce more diversity and ideas. Similarly, consideration of countervailing evidence can protect against prior beliefs. Scenarios can represent different viewpoints and have interesting promise for combating these processes.
Risk Communication in DMDU
Dr. Jody Wong began her presentation with the tale of the ‘Dodgy Dossier’ to illustrate the importance of communicating uncertainty.
In 2002, the British Government published a document claiming that Iraq had weapons of mass destruction and nuclear ambitions, which was used to justify going to war. However, after the war, this was found to be untrue. An investigation found that expressions of uncertainty in the intelligence were removed or not made clear enough in the public report. Similarly, in the US, a review found that the intelligence community did not accurately explain the uncertainties behind the judgments in their report.
“So the main idea is this: the removal of uncertainty from the documents had a big impact on public opinion and governments and was a key factor in the decision to go to war,” she said.
In DMDU, the assessment of future risks generally contains a strong element of epistemic uncertainty. Put simply, this is uncertainty that arises from a lack of knowledge or understanding. So, further knowledge would subsequently revise our predictions. This suggests that failure to do so can seriously compromise decisions, which is why it’s important to stress the effects of communication on the audience through the cognitive psychological perspective.
One question we might ask ourselves in the context of DMDU is how we might communicate the uncertainty in scenarios honestly and transparently without losing trust and credibility to the benefit of stakeholders who subsequently use the information to form an opinion or make a decision.
A basic model in the field of communication is known as the Lasswells model of communication. This version has been adapted to communicate uncertainty, and it focuses on who communicates what, in what form, to whom, and with what feedback.
So, starting from the left, the first two factors in our framework are about who is communicating. The owners of the uncertainty or the experts assess the uncertainties, such as scientists or official bodies. Communicators might be the technical experts, communication professionals, and journalists who convey view information on behalf of the institutions.
The factors related to what is being communicated might be the object that is uncertain, such as facts, numbers, or scientific models and hypotheses. The source for the uncertainty suggests the reasons for the lack of knowledge range from direct uncertainty about effect to indirect uncertainty about the underlying science.
The magnitude of the uncertainty ranges from a small lack of precision to a substantial degree of ignorance. Factors related to the form of communication might be how the uncertainty is expressed, such as a full probability distribution or a brief mention that uncertainty exists. The format of the communication can be numbers, visualizations, or verbal statements, and the medium of communication would be print, online broadcast, or verbal conversation.
The factors related to whom the communication includes might be the characteristics of the audience, such as their level of numeracy and expertise in the field; the relationship of the audience to the topic, such as whether it is emotionally charged or contested; and the relationship of the audience to the communicators, including their perceived credibility and level of trust or distrust.
The last set of factors is about the effect of the communication on the audience, including how it affects their thinking, emotions, trust, behavior, and decision-making. This ties in with cognitive biases because people’s understanding of uncertainty can vary greatly and have a big impact on decision-making.
There is relatively little research on communicating uncertainty effectively in DMDU; however, Dr. Wong said there’s some hope. Recent research has looked at using scenarios as an effective communication tool. First, scenarios describe a range of possible futures, and narrative theory argues that the stories we tell about ourselves and our experiences shape our identities and worldviews. The authors propose that scenarios or narratives about how the world works and our role in it are important in shaping public opinion, policy, and politics related to the environment. These stories can be intentionally created and spread or emerge and be reinforced through social relationships.
In DMDU, decisions about the future are motivated less by accurate anticipation and risk assessments than collectively held narratives. So, in many ways, narratives can extend cognition, create shared knowledge across space and time, and shape our beliefs, values, and actions in the face of uncertainty. Because environmental decision-making grapples with uncertainty, it gives psychological agency to narratives, which highlight certain contingencies, responsibilities, priorities, and pathways over others. Due to their persuasive abilities, there is a growing interest in using narratives in social sciences and climate research.
Some examples where narratives in the environmental context have conventionally been observed:
- Climate narratives: Scientists, journalists, activists, and other actors use stories with words, images, and statistics to make sense of observations or events. The framing – or how these stories are told can have a big impact, influencing whether people feel hope or fear, take action or do nothing, and feel uncertain or clear about the situation. “For example, climate narratives can place responsibility on individuals or policymakers who have seen them in narratives where they’ve argued that it’s too late to act or, and they’re in narratives that emphasize the need to act now,” said Dr. Wong.
- Embedded in cultural artifacts, such as folklore, songs, customs, and oral traditions that passed down norms from generation to generation: “For example, research shows that folklores that highlight features of the physical environment, such as crop and water-related motifs in fertile areas, or accounts of earthquakes in risk-prone areas can shape shared knowledge, values, and ecological dynamics,” said Dr. Wong.
- Data visualizations that are often accompanied by other texts or images: “For example, the graph are warming charts created by Ed Hawkins in 2018. In blue, it shows the cooler years, and in red, it shows the warmer years. So what the graphic depicts is easily understood by a broad range of audiences without having much scientific information.”
Two potential types of strategic narratives can be leveraged when communicating uncertainty:
- Narratives of human agency and action: Typically, these narratives focus on the exceptional agency of humans, following themes such as how can we play a role in reshaping the Earth’s surface? What is our dependence on the earth system for our well-being? These narratives are often in popular culture in the growing genre of climate fiction, which creates psychologically immersive and visceral experiences of possible futures, including the tragic ones resulting from inaction and utopic ones resulting from societal change.
- Narratives of delay, uncertainty, and inaction: These narratives highlight the difficulties related to knowledge and ethics because they build on real concerns and fears. They also prey on tendencies in decisions under uncertainty. Linguistic analysis shows that these stories are influenced by the norms surrounding other social issues and a hierarchy of more abstract stories about freedom, responsibility, and social division. So, in many ways, they favor the status quo and are more palatable among a broad range of audiences.
Best practices
Dr. Wong then discussed a project she was involved with that concluded this January. They were tasked with developing an issue paper on best practices for environmental risk communication and engagement with communities neighboring military installations. From this, she proposed best practices that could be applied to other contexts, such as DMDU.
- Acknowledge the values, feelings, and worries expressed by the target stakeholders. The first is to acknowledge the values, feelings, and worries expressed by the target stakeholders. This helps build a solid partnership based on trust and credibility.
- Recognize that target stakeholders are predisposed to distinct stressors and biases. This helps bolster a communication approach that meets the needs of target stakeholders. Internal stresses may, for example, include a lack of literacy skills, while external stressors may include challenging economic factors or conditions. And biases might also color perceptions of risk.
- Communication should avoid jargon and explain risk information in clear and consistent ways to minimize excessive fears and anxiety: Jargon and highly technical language can hinder narrative comprehension by audiences, and as a result, critical risk information and recommended actions may be misunderstood or ignored. Best practice suggests using terms and language familiar to the larger audience and focusing on local and observable consequences described in non-probabilistic terms.
- Incorporate tools to learn about the target concern: This increases the likelihood that the target audience will receive narratives more positively. Explaining technical details through stories that resonate with stakeholders can increase the likelihood that the information will be understood, personalized, and acted upon.
- Evaluate successes and failures to help ensure communications evolve as situations change and ideally improve over time. Regular monitoring and evaluation are necessary and should include the involvement of stakeholders in assessing narrative effectiveness, noting any reasons for changes in a narrative that could impact an organization’s trust and credibility.
“So, in our opinion, there’s a need for continued exploration of cognitive psychological impacts of scenarios,” said Dr. Wong. “I’ll end here with a big-picture question: How can we construct scenarios and narratives with DMDU to understand how shared knowledge and values are shaped?”
She noted that she is part of a team of researchers and journalists who have been exploring that question. In an upcoming edition of Frontiers in Climate, they will publish a commentary on how media narratives intersect with DMDU to inform and leverage the complexities of modern contemporary public challenges.
“One of the key ideas is to explore how uncertainty might be actionable as opposed to fearful. We’re proposing that narratives can be potent drivers that propel people to act according to or despite uncertainty. So stay tuned if you’re interested.”
Q&A
CHAIR LISA WAINGER: Can you give us a definition of cognitive bias? And what are some of the major concerns?
DR. ANDREW PARKER: The field of behavioral decision research typically defines cognitive bias as a result of a natural cognitive process that results in a systematic deviation from what we refer to as normative decision-making or decision-making that might be considered by rational models. The field tends to focus on a contrast between descriptive accounts of decision-making and normative theories of decision-making, and cognitive biases are the result of natural cognitive processes that cause a deviation from those normative models.
CHAIR WAINGER: You said people tend to see losses as bigger than gains, even if the magnitudes are the same … ?
DR. ANDREW PARKER: The classic framing effect studies contrasted two different presentations of risk information that were informationally equivalent to each other, whereas one focused on the losses and the other focused on the gains within the situation. They found that people tend to react very differently to them in two different ways. One, they tended to be risk-seeking in the loss frame and risk-averse in the gain frame, but also that the losses weighed heavier than the gains. So, if you want to think of it this way, the slope of the utility function is greater in the loss than for a limited gain frame.
DR. DIANE McKNIGHT: We are the Delta Independent Science Board, so we’re focused on the Bay Delta system and California’s water issues. To what extent that scenarios may be place-based or not place-based influence the bias in terms of whether individuals may consider the scenario impactful or not? Does place matter?
DR. JULIE WONG: “It does make me think about the idea of psychological distance, which is the degree to which people feel removed from a phenomenon. “The idea of the proximity, or the distance to which people are closer to the issue at hand versus whether they are far removed from it, has an effect on how they take action or do not take action at the end of the day.”
DR. ANDREW PARKER: My initial reaction is that closeness to the problem or something in your backyard reduces the psychological distance, increases salience for the individual, and may increase our connection with a variety of fairly strongly held beliefs and values. So biases, such as my side or confirmation bias, play out more strongly when people have very strongly held beliefs and values. Similarly, recall will be greater for something that has greater salience to an individual. We know that recall is related to perceived likelihood. So, something that is more valued to you is scarier because of that, and perhaps that has a greater salience to you, is more memorable, and hence may feel more likely. This, of course, speaks to the power of scenarios because it gets engaged stakeholders quite readily. But it also speaks to one of the things to keep your eye out for: our perceived likelihoods getting increasingly divorced from reality.”
DR. TANYA HEIKKILA: One of the takeaways I have from your presentations is that if we are using narrative-based scenarios, there are both advantages but also potential risks, especially in terms of how those scenarios could either help us overcome some of the cognitive biases or potentially how those narratives could even trigger were reinforced certain cognitive biases. Suppose you’re a decision-maker trying to buy into this idea of using some DMDU approaches and scenarios more robustly. How much pre-work should be done before devising some scenarios to understand the mental models that the stakeholders might have and to think ahead of time about how the design and the narrative structure of the scenarios could potentially be triggering or overcoming some of those cognitive biases? Then, how much work does it take to get the scenarios right to hit that sweet spot to encourage the potential advantages you could achieve?
DR. ANDREW PARKER: We conducted a study some years before where we were mostly interested in how people react to analytically derived scenarios. We took advantage of simulation and computational capabilities to analytically derive policy-relevant scenarios. We were very interested in how people received this, in this case, the public. One of the really interesting parts of that study was that people generally liked the idea of analytically derived scenarios and that we were showing them different outputs from some scenario discovery methods. They generally liked them, but mostly when the types of dimensions characterized in those scenarios were intuitive.”
DR. ANDREW: Now, in hindsight, this makes total sense. When you’re doing multi-dimensional analysis, it’s very possible that you come up with dimensions that are hybrids of a bunch of natural things that don’t make sense to the average human. Those weren’t very helpful. What was really helpful was when they aligned with concepts that were relatively straightforward. So, things like wetter or drier, hotter or colder, more urbanized versus less urbanized … you can imagine a variety of underlying factors that would all align with that. At the time, we were thinking of this as a rotation in the space of a multi-dimensional result. It was saying, okay, maybe we accept a slightly less than optimal rotation if it better aligns with the original axes that we started with. And that was a pretty clear result from that study.
So, to get to your question in terms of the pre-work, I think analytically derived scenarios are a good idea, but take care of how those are constructed. I think it gets into the idea of narrative as something that tells a natural story. If it does, it’s going to be a more engaging and more effective scenario.
DR. JODY WONG: When I think deeper about the construction of narratives, we can certainly benefit more from working with researchers, scholars, and people who have dedicated their lives to this line of work. For instance, in the field of communication, there are narrative experts who will tell you that constructing a good narrative is much more complex than we think because, done the right way, narratives can hit the sweet spot – they can model or inform the kind of action that people might take. So I think it’s as much of an art as science.
And to clarify, we’re not suggesting that narratives are everything. Among the broader audience, one’s opinion on different environmental issues depends on many other factors. But all these factors act in concert with a story that can capture people’s imagination, so even if we do not know all the factors and processes by which a narrative can gain traction with the public over time, we can certainly observe its presence.”
DR. JAYANTHA OBEYSEKERA: We are trying to come up with a range of plausible scenarios; some could be ‘black swans’ or very extreme scenarios that one cannot even associate any probabilities with, and the stakeholders or community might not be very interested in those scenarios, either they’re not useful, or there’s not much you can do, but yet they might be important because of the irreversible implications that could have on the world. So how do you convince the community to look at that range of scenarios, the plausible and those that are very rare? There might be a reluctance to look at those scenarios, so what would the risk communication strategy be to convince them to be open to complete range?
DR. JODY WONG: “It depends on who you’re trying to communicate to and the issue you’re trying to communicate. First, I think it’s important to really recognize that the people we’re trying to communicate with are predisposed to a lot of stresses and biases. So, trying to evaluate the demographics of who you’re communicating to can really foster an approach that meets their needs. An example of internal stress that people might face is their level of literacy skills; external factors could include challenging economic factors or conditions that they’re personally facing. And then cognitive biases could color their perceptions of risk. So knowing your audience is one way to engage this risk communication framework and approach and try to build into it.”
DR. ANDREW PARKER: The idea of coproduction with communities who are concerned is a very powerful way to go about this. Get them involved in the scenario design. This is where some focus on possibility rather than probability can help get people on board with considering things. I, frankly, would be honest that something is a very low likelihood event or low and unknowable, given the deep uncertainty. But consider that the possibility will help explore policy options that may be robust to even those relatively unthinkable situations. In my mind, the path forward is not to pretend that they will buy into the scenario but to get them to the point where they’re willing to consider it as part of a robustness goal. Having them involved will definitely help with buy-in.
CHAIR WAINGER: We see this reticence to look at extreme events and negative scenarios in general in the ongoing activities. You’re talking a lot about coproduction, with the idea being that there’s plenty of opportunity to have a dialogue about the benefits or harms of using such scenarios. But is there another answer besides that to motivate people to look at more extreme events?
DR. ANDREW PARKER: I’m intrigued by the idea of reframing a problem in terms of action, as opposed to risk and or loss. There are going to be scenarios that are, on the one hand, sort of unthinkable for one reason or another, politically or otherwise … there may be reticence to take it on. But reframing that scenario, in terms of action, is one way you might approach that.
DR. JODY WONG: I think it’s really more about the idea of reducing the psychological distance. For example, at an abstract level, climate change is perceived as far away, like the concept of psychological distance. This results in a dissension of the problem and an unwillingness to tackle the issue. On the other hand, if we can draw the distance closer, then it leads to a higher likelihood of acceptance of climate change by the public … I think making the issue of any topic issue more localized, more relevant, and more urgent will help to reduce the struggles and estrangement by people and help to increase pro-environmental behaviors. So, if we can reframe the narrative to reduce the psychological distance that people experience, this could be one way.
DR. THOMAS HOLZER: I just wanted to make an observation. My career was spent dealing with earthquakes. And for a long time, we couldn’t get anybody to do much about it. But then we had the Loma Prieta earthquake, and all of a sudden, the scenarios that were much worse than the Loma Prieta earthquake became relevant. So, you need to pick your timing for the scenarios. We know we’re going to have floods in California, so rather than talking about the Ark storm on just a random basis, connect it with a real flood so people see a lot of the damage that can be done. You need to pick the timing of the scenarios to make them effective.
DR. ANDREW PARKER: If you have a salient event, that will certainly be a motivator. And connecting it to real-world events, even if it’s a more extreme scenario – those events which have affected people’s lives, will certainly increase their willingness to connect with scenarios.
CHAIR WAINGER: You mentioned the problem of presenting a series of scenarios, some of which may be highly improbable, and decision-makers treat them all as equally likely. That could be dangerous if people start making inefficient investments. But if you had a positive case study, a little bit of a manipulation of your audience in that when you gave them probabilities, they only focused on the central tendency type scenario. And then, when you gave them scenarios, they were more likely to look across futures and find a solution that was robust across futures, which I think is kind of the benefit of thinking about the improbable conditions. If you can protect against extreme conditions without spending that much more than you would have for the best guess scenario, then you’re in good shape. Is it really bad to have people misinterpret that? Or is it sometimes beneficial?
DR. PARKER: We were fairly careful to express to them that just because these scenarios were being presented next to each other, they weren’t necessarily equally likely. They were just scenarios to consider. So I don’t think we encouraged them to consider them equally likely. It is true that in the study, they were more likely to explore that area of the policy space and discover some of the things that you’re talking about. Because they explored that policy space, they were able to discover things like how they could cover a lot of other things that were unlikely but scary for not that much extra cost. And that’s where they landed on the more robust solutions in many ways. So I think that there’s a there’s a real element of truth in that. There’s a fine line between not putting that forward as an organizing concept, but at the same time, being open to the idea that these things have varying probabilities.”
CHAIR WAINGER: We want to understand these cognitive biases so we don’t inadvertently manipulate people. We know we can scare them more easily with losses than with gains. So, if we frame things as losses, we are pushing them toward a particular action. So I think it’s almost like we’re trying to do the opposite, trying to avoid triggering people in certain ways.
DR. PARKER: I remember the first time I was in a grocery store, and I saw a package of ground beef labeled 80% lean, 20% fat; it had both labels on it. This was the subject of a study run by Irwin Levin, a psychologist in Iowa, some years ago. I took a photograph and sent it to him because finally, somebody got it. What he had shown was that when people bought ground beef, they saw it as labeled as 20% fat. He ran a condition where he presented both pieces of information. And it seemed to have a lot of promise to it. So, to your point, Lisa, I think you’re right. It’s being aware of these cognitive biases and trying to be cognizant of the community in constructing scenarios, so we don’t inadvertently build those things into it.”
Previous webinars in the series:
