The risks we take and the life we make - Effects of positional concerns and bounded rationality on risk taking behaviour

Although we today have more information, knowledge, and more advanced technology to prevent accidents than ever before, accidents still occur with a higher frequency than we would like. At the same time, it appears as if some people do not wish to minimize risk. On the contrary, many voluntarily expose themselves to the risk of a serious injury, or even death. People who recreate in avalanche terrain is a good example. Backcountry skiers expose themselves to the risk of being caught by an avalanche every time they venture into avalanche terrain. The aim of the White Heat Project was to find answers to why people voluntarily expose themselves to excessive avalanche risk. When we decide, we compare the expected costs and benefits of our different alternatives. If we use all information available and choose the alternative that can be expected to maximize our wellbeing in the long term, economists view the choice as rational. If we carry the full burden and reap all the benefits of our decision, economists view the decision-environment as efficient. There are many reasons for why choices related to risk may neither be efficient nor rational. Risky behaviors often have external costs. For example, if a skier triggers an avalanche, other people, infrastructure, cars, and forests in the run-out zone can get hit. In addition, rescue operations and health care require resources that are often funded by taxes. Finally, humans are social animals, and our actions can therefore spur emotional and behavioral reactions in others. Skiers who chose to ride risky slopes do not carry the full burden of their decisions and therefore have incentives to take more risk than is optimal from society?s perspective. There are many reasons for why we sometimes make decisions that are irrational: We rely on rules of thumb to preserve energy and we often focus more on immediate than distant goals even when this harms us in the long run, especially when we experience strong emotions. Our emotions, alongside with the physical and social environment also hold power to affect our perception of the risk at hand. The White Heat Project analyzed how emotional, social and cognitive factors affect the perception of, and willingness to expose oneself to, danger in avalanche terrain. Our results suggest that backcountry skiing makes people happy and excited, but we find very little any evidence that the prospect of skiing fresh snow (powder fever) affects willingness to expose oneself and others to risk. Instead, we find that the most important factors behind terrain choices in the backcountry are attitudes to risk, perception of risk, skill, and training. Participants who say that they are either willing to expose themselves to risk or have strong skills in backcountry travel are more likely to ski potentially risky terrain. Individuals with formal avalanche training are less willing to ride risky terrain. These findings support the notion that people are aware of the risk and that knowledge reduces risk-taking behavior. This is reassuring. However, we also find more troubling results. For instance, we find that irrelevant information about past avalanche danger affects both current and future estimates of avalanche danger. Surprisingly, detailed information about the avalanche problem does not remove the effect and individuals with avalanche training are more likely to be affected by consider historic information than individuals without such training. Some of our more preliminary results further indicate that backcountry skiers perceive that they have more control and depend less on luck than comparable others do. These last results are worrying because overconfidence increases the risk of accidents. We are still investigating if irreversible investments in time, effort or money affect risk-taking. Our research on social factors shows that social signaling matters. Our results suggest that many people care about how their consumption and performance compare to that of others. This is especially the case if the behavior is socially valued. Peer pressure and social aspirations in the backcountry increases willingness to exposure oneself to avalanche risk. We find that risk-willing friends push many individuals to accept more risk than they ideally prefer, that many skiers? satisfaction with their riding depend on what other riders do, and that these skiers are more willing to ride risky terrain. Our findings have implications for public natural hazard warnings, education in risk management, and individual risk mitigation; some information in public natural hazard warnings may distort rather than improve risk perception. Warning services should therefore carefully assess the informational content in the bulletin. Knowledge improves decision-ability, but over-confidence and social factors may distort both judgements and incentives. This points to the importance of a greater focus on cognitive and social factors in risk education.

The project has contributed crucially to the establishment of an internationally renowned interdisciplinary research group on decision-making under uncertainty at UiT. It has further created collaboration with distinguished researchers at e.g., Carnegie-Mellon University, Max-Planck Institute, Montana State University, Umeå University and Helsinki University. These collaborations have provided UiT with world-leading competence in decision-making, snow science, and education. The project has created collaboration between researchers and practitioners at avalanche warning services, resulting in e.g., an automatization of the Avalanche Terrain Exposure Scale. The map layer can be used for research and for risk mitigation in avalanche terrain. The research output has been disseminated to policy makers, practitioners, and recreationalists via e.g., open access publications, avalanche seminars, recorded presentations, and podcasts. Some of the material is now used in avalanche courses.

The questions asked in this research proposal are related to mechanisms behind risk-taking behaviour. More specifically, we aim to analyse how (i) positional preferences (i.e., utility based on social comparisons) linked to social identity, (ii) incidental affect and (iii) the trend in the level of risk affect individuals' perception of, and exposure to, risk. In particular, these questions will be applied to a context of risky behaviour in avalanche terrain. To answer our research questions, we will develop new theory and test our derived hypotheses empirically. The main challenges to achieve our goals are to: 1) identify and rank social identity, 2) create a valid measure of risk-taking behaviour, and 3) to compile a sufficiently large and representative dataset of backcountry recreationists. The cross-disciplinary composition of our research group, with expertise in snow science, economics, psychology and geography, along with our location in Tromsø, Norway, and Bozeman, Montana, USA, and our access to a large global dataset on backcountry ski tours, provides us with a unique capacity to face these challenges and reach our goals. The knowledge produced by the research proposed is valuable for policy makers in a number of settings. If risk-taking behaviour is linked to positional preferences and social identity, this implies that the level of risk-taking behaviour is inefficiently high and that the effectiveness of policy may be improved if targeted at certain groups rather than individuals. If incidental moods or trend in the level of risk affect risky decisions, then policy that target rational decision-makers may not work as planned. Our special application on risk-taking behaviour in avalanche terrain will produce knowledge that can be directly applied to avalanche education and forecasts services.