The objective of global change vulnerability assessments is to prepare specific communities of stakeholders to respond to the effects of global change (Schröter et al., 2005). There is a growing call to favor "vulnerability" assessments over the more familiar "impacts" approach to research on the human dimensions of global environmental change (e.g., Downing,
2000; Kelly and Adger, 2000; Liverman, 2001; McCarthy et al., 2001; National Research Council, 1999; Parry, 2001; Turner et al., 2003). In this literature, vulnerability is generally defined as a function of exposure to stresses, associated sensitivities, and relevant adaptive capacities. Thus to be vulnerable to the effects of stresses associated with global change, human-environment systems must be not only exposed and sensitive to the changes but also unable to cope. Conversely, systems are (relatively) sustainable if they possess strong adaptive capacity (Finan et al., 2002). In the former case, some form of anticipatory action would be justifiable to mitigate the ecological, social, and economic damages anticipated from global change, whereas in the latter case there would be less reason for concern and action. Vulnerability assessments are therefore a necessary part of sustain-ability science, or basic research intended to protect social and ecological resources for present and future generations (Clark and Dickson, 2003; Kates et al., 2001).
The common distinction between the vulnerability and impacts perspectives is that the former emphasizes the factors that constrain or enable coupled human-environment systems to adapt to stress, whereas the latter focuses more on system sensitivities and stops short of specifying whether a given combination of stress and sensitivity will result in an effective adaptation. In fact, this distinction applies more to the empirical studies of climate change impacts than to the conceptual underpinnings. Adaptation has been at the heart of the debate on reducing vulnerability to environmental stresses for a long time (Turner et al., 2003). Even the early models from the climate change impacts canon (e.g., Kates, 1985) do not exclude the process of adaptation, and the same applies to the broader, related literatures on risk and hazards (e.g., Burton et al., 1978; Cutter, 1996; Kasperson et al., 1988) and food security (e.g., Bohle et al., 1994; Downing, 1991). Thus the recent explosion of interest in "global change vulnerability" is not so much the result of a revolution in ideas—although theories are developing (e.g., Adger and Kelly, 1999)—but a response to a general dissatisfaction with the ways in which adaptive capacity has been captured in empirical research and the associated need to reconnect with this concept if global change models are to improve.
Polsky et al. (2003) suggest that successful empirical research on global change vulnerability should satisfy the following five (overlapping) criteria: (1) exhibit a place-based focus; (2) devote equal energy to exploring future trends and historical events; (3) treat stresses as multiple and interacting instead of unique or multiple and independent; (4) include not only natural and social science but also local ("indigenous" or "user-specific") knowledge; and (5) examine how adaptive capacity varies both within and between populations. This last criterion is especially important for defining vulnerability in the case of drought. In the United States at least, institutions that regulate on the one hand and design and disseminate new technologies on the other hand are the principal pathways for drought response, in anticipatory and reactive modes. To be sure, individual people do actively participate in drought mitigation activities, but the most important current set of options for adaptations to the effects of droughts, we argue, is associated with institutions (detailed in Sections III and IV).
It is difficult to specify quantitative models of how institutions enhance or reduce adaptive capacity. This difficulty is important in the climate change context because quantitative models, for better or worse, have occupied center stage in the debate on possible impacts from climate change and associated policy responses. The majority of these models are grounded in neoclassical economic theory, where the role of institutions in mediating impacts is largely if not entirely discounted. In these cases an individualistic perspective presumes that all people (modeled agents) are "economically rational." These modeled agents will implement any and all necessary adaptations to the effects of climate change "perfectly" (i.e., instantaneously and at greatest individual profit). In this way the role of institutions in influencing social response is implicitly assumed to be trivial, or, if significant, then of equal importance everywhere and always and as such not worthy of specifying in a model.
The canonical example of this approach is Mendelsohn et al.'s (1994) influential Ricardian analysis of climate change impacts in U.S. agriculture. This approach uses a regression model to evaluate the importance of climate in the determination of agricultural land values (in the contiguous United States) relative to other important factors such as population density and soil quality. The possible economic impacts of climate change are projected by multiplying the statistical relationships between historical climate and land values by a hypothetical climate change. Not surprisingly, the projected economic impacts based on the "perfect" adaptive capacity described above, defined in strict profit terms, are lower than in studies that do not allow the modeled agents to respond at all (i.e., where adaptive capacity is assumed a priori to be null).
Of course, if it is unrealistic to assume that agents possess no adaptive capacity, then it is equally unrealistic to assume that they possess perfect adaptive capacity. For example, the decisive factor behind a farmer's choice to prepare for drought through summer fallowing or portfolio diversification may hinge on the advice of an agricultural extension agent—who may or may not have the farmer's profit maximization as the number one priority (Riebsame, 1983). Thus, in principle, greater realism can be achieved by incorporating in the models some of the missing institutional landscape (Hane-mann, 2000). Institutional influences should be particularly important in regions where climate change results in a strengthened drought regime.
Polsky (2004) modified the basic Ricardian framework to explore how institutions modulate agricultural climate sensitivities. In this analysis of agricultural land values in the U.S. Great Plains, statistical relationships are estimated at multiple spatial scales simultaneously: for the region as a whole (n = 446 counties); for the meso-scale (two subregions defined by the boundaries of the Ogallala Aquifer: n1 = 209, n2 = 237); and for the micro-scale (many sets of small numbers of counties; n « 7 on average). For each of the 6 years analyzed, the regression model fit better for the subregion defined by the boundaries of the Ogallala Aquifer than for the rest of the Great Plains. These differences in model fit were modest in
1969, dramatic in 1974, 1978, and 1982, and intermediate in 1987 and 1992, and they suggest that unspecified factors are responsible for buffering fluctuations in land values in the Ogallala relative to the rest of the Great Plains. The Ogallala is characterized by strong S&T and natural resource management institutions developed in response to the challenge of drought and the opportunity of irrigation. Thus an emerging hypothesis is that differences in the form and function of these institutions between the two subregions explain differences in the climate sensitivities of the two subregions (see also Emel and Roberts, 1988). Clearly, testing this hypothesis requires an in-depth study of the ways in which such institutions produce and disseminate knowledge.
Was this article helpful?