Community Resilience to Climate Change: Theory, Research and Practice

43 2. DATA AND PERCEPTIONS ABOUT RECORD-BREAKING EVENTS Extreme events are challenging because the probabilities are hard to measure and because decisions about rare events with important consequences pose special challenges. Profound uncertainty makes rational responses difficult, and makes it easier for irrational approaches to take hold. Extreme events are not only uncertain, but themeasure of uncertainty (e.g. the tail of the probability distributions) also is itself uncertain. In some cases probabilities may be fundamentally unknowable, due to the complex interactions between human and environmental systems. The probability is obviously unknown for completely new events, such as the emergence of a particular new disease. However, probabilitiesmay be poorly known even for events that have occurred only occasionally in the past. In the tails of probability distributions, observations are rare and therefore data are sparse for fitting models. As a consequence, trends in extreme events are hard to discern. For example, large data sets and extensive analysis are needed to establish trends in extreme rainfall events, flood damages or sizes of forest fires [8,9,10,11,12]. Where limited data make it hard to measure the probability of a certain kind of extreme event, assessing a trend in that probability is even more difficult. Some classes of extreme events, such as flood damages, earthquake magnitudes, and wildfire sizes, have ‘fat-tailed’ probability distributions. In fat-tailed distributions the probability densities of extreme events are much larger than in more familiar distributions such as the normal distribution [13,14]. In these cases it is highly misleading to estimate the magnitude of the next record-breaking event from the record-breaking events that have been observed so far [13,15]. Sometimes two or more kinds of extreme events co- occur, for example if flooding causes landslides in a watershed previously denuded by fire. In ecology such multiple impacts are called compound disturbances [16]. Fat-tailed distributions tend to magnify the correlations of extreme events and thereby increase the probability of compound disturbances [17]. In addition to these statistical challenges of anticipating extreme events, there are cognitive biases that can lead to irrational decisions when the stakes are high and probabilities are near zero or one [18,19]. For example a ‘tyranny of recent events’ (availability heuristic, [20]) causes people to misjudge risk, as when fears of terrorism are exaggerated for a time after a terrorist attack [21]. In contrast, when risks are familiar there is a tendency to underestimate the danger and overestimate one’s ability to control the situation [21,22]. Sometimes inconsistent decisions are made about gains versus losses [18]. Different preferences for gains or losses of similar magnitude are the subject of prospect theory [23]. Perceiving gains and losses differently is a ‘bias’ in the context of rational choice theory, but not if prospect theory is instead used as the benchmark for expected behavior. Nonetheless, with respect to rational choice theory people can make poor decisions about risk of extreme events. People in isolation have their limits; collectively, however, people can create an institution to improve on these cognitive limits to rational behavior. The challenge is to design institutions so that the aggregate decision creates more good outcomes for the group. These (and other) behavioral phenomena inevitably affect societal decision making about extreme environmental events [24]. These interactions between complex aspects of natural systems and human cognition add to the challenges of understanding and managing extreme events, and make general resilience important. General resilience should protect social-ecological systems against vagaries of human volition aswell as uncertainty about the relevant probabilities. In the remainder of the paper we discuss characteristics of institutions that contribute to general resilience of social-ecological systems. 3. GENERAL RESILIENCE AS A STRATEGY Extreme events, including record-breaking extremes and new kinds of shocks, have been with us forever and may intensify in the future. General resilience—the capacity of social-ecological systems to adapt or transform in response to unfamiliar or unknown shocks—is essential for sustainability in the face of extreme events. However, the wide-ranging nature of general resilience makes it difficult to define specific steps for creating it. Instead it is possible to identify conditions that can enable or support the development of general resilience (Table 1). Diversity provides for different kinds of processeswithin a social-ecological system (functional diversity). It also provides for components that have similar functions but different responses to disturbance (response diversity), so the function is maintained even if one component is damaged. When teams of people are solving complex problems, diversity of perspectives and experience matters as much as individual ability [25]. The cumulative adverse effects of factors that diminish human wellbeing may also reduce the capacity of a society to respond effectively to disasters. There is a strong relationship between various aspects of human well-being and income inequality [26], suggesting that high income inequality reduces the general resilience of a society. Modularity helps contain disturbances by compartmentalizing social-ecological systems [27]. For example, land management with prescribed fire uses firebreaks to limit the spread of the fire. This makes it possible to manage burn units independently, and thereby build landscape diversity. Similarly, quarantine mechanisms may restrict the spread of epidemics or invasive species. Modularity relates