Decisionmaking, risk, and uncertainty: an analysis of climate change policy.

• Author: Dholakia-Lehenbauer, Kruti
• Position: Essay

This article explores four questions. First, what theoretical frameworks help describe policy failure and success? Second, how might the decision that leads to failure or success be understood in terms of differing concepts of rationality and decisionmaking? Third, how does the discussion of risk and uncertainty as originally proposed by Frank Knight (1921) apply to a better understanding of both the first and second questions? Fourth, what is the relationship between serial and parallel processing and how are these administrative systems related to important aspects of the prior questions? Our chief contribution in this article is to show the ways in which these questions and their respective theoretical frameworks are interrelated as applied to one important contemporary policy question--climate change. We think our proposed integration of the various literatures offers important insights into the challenges policymakers face in deciding whether or not to adopt a particular policy.

The first question is based on the notion that decisionmakers can commit two fundamental types of policy errors: the wrong policy can be implemented or there can be a failure to implement the needed policy. Based on work by Heimann (1993), we construct a simple model of hypothesis testing using Type I versus Type II errors that is common in the social and behavioral sciences.

The second question is based on Vernon Smith's (2003) discussion of two fundamental types of decisionmaking: constructivist and ecological. The former approach, which has important similarities to expected unity theory, represents a highly deductive, top-down approach that places a high degree of confidence in model building and the ability to make accurate predictions about phenomena over extended periods. The latter represents a more bottom-up strategy that assumes decisionmakers do not have all encompassing information available to them and thus rely on heuristic models of decisionmaking. Such decisionmaking does not reject theory, but allows for theory-building to be done incrementally with ongoing adjustments of beliefs.

The third question provides an analytic framework for evaluating the probability of different kinds of events. Knight (1921) drew the important distinction between risk and uncertainty, and the consequences for understanding the probability of the occurrence or non-occurrence of particular events (see Raines and Jung 1992, Jarvis 2008). Risk refers to a calculated and predictable outcome, whereas uncertainty refers to an event for which the probability of outcomes is difficult to objectively measure. We believe that policymakers, as well as policy analysts, are all too likely to misinterpret one set of events for another, leading to faulty policymaking.

The final conceptual framework in which we examine the relevance of the forgoing issues to questions of administrative systems offers additional and important insights into the nature of decisionmaking in organizations. Ultimately, we think our analysis offers a roadmap for future research that will be of assistance to scholars and decisionmakers in better understanding decisionmaking and its consequences. In the following section, we discuss the desirability of policy action or inaction (Type I versus Type II errors). We then move to a comparison of constructivist as opposed to ecological decisionmaking and the distinction between Knightian risk and uncertainty in the subsequent two sections. We then apply our arguments to the issue of climate change.

To Act or Not to Act: The Nature of Policy Success or Failure

What is the nacre of policy success or failure? Many scholars have explored that question from the standpoint of what happens after a decision has been made. Our approach is somewhat different. Instead of asking how many resources are devoted to a policy or how principal-agent problems influence policy outcomes, we inquire about the fundamental nature of decisionmaking--that is, should one act or not act? Of course, we recognize that the decision of whether or not to act is fraught with uncertainty and is based on a subjective assessment of the probability of success (or failure). Yet focusing on it helps capture the most fundamental choices that policymakers face.

One important contribution to this research is the work of Heimann (1993), who builds on work by Perrow (1984) and Landau (1969), who pioneered research in the area of organizational failure and the need for organizational redundancy. In his discussion of the decisions leading up to the Challenger Space Shuttle disaster of January 1986 (see Vaughan 1997), Heimann notes two types of policy errors. In the first type (Type I error), the wrong policy is implemented. In the second type (Type II error), the correct policy action is not implemented. The former represents an error of commission, and the latter an error of omission. Figure 1 presents the stylized model developed by Heimann.

Decisionmakers face two possibilities when confronting binary choices: act or not act. We know the correct decision was to take action--that is, stop the launch. The decision taken, however, was to do nothing--that is, allow the launch to proceed when it should have been aborted. This resulted in a Type I error. And, as already noted, it is crucial to recognize the unavoidably subjective nature of the assignment of probabilities to committing a Type I versus a Type II error. Heimann (1993) makes a strong case that, given the knowledge available to NASA, committing a Type II error (wrongly stopping the launch) was preferable to committing a Type I error (wrongly launching), even if the probability of failure had been determined to be remote. In other words, halting the launch of Challenger was a small price to pay to avoid a possible failure, even if the chances were good there would be no failure.

Contrasting Approaches to Decisionmaking

Here we examine and contrast two different rationality assumptions to decisionmaking: the constructivist and the ecological approaches. Both have their place in formulating and designing policy, under different conditions and contexts having to do with the information available and the costs of making the wrong decision. Constructivist rationality is based on the logic of Descartes (Cartesian logic), which advocates that human reason can deductively construct human institutions through the conscious molding of human behavior. Cartesian logic is based on the notion that either society derives its structure from the top down along rational principles or through conforming to the preferences of a social planner who could know objectively what is "best" for the society.

An alternative to constructivist rationality is ecological rationality, which is closely related to the principles and assumptions of bounded rationality. Smith (2003: 470) notes, "Ecological rationality uses reason--rational reconstruction--to examine the behavior of individuals based on their experience and folk knowledge.., to discover the possible intelligence embodied in the rules, norms, and institution of our cultural and biological heritage." Thus, ecological rationality is a decisionmaking process that emerges out of cultural and biological evolutionary processes utilizing the homegrown principles of action, norms, traditions, and morality. Decisionmakers, acting on the assumptions of ecological rationality, develop rules of thumb or heuristics in decisionmaking. This approach implies that determining global optima or maximizing strategies for a particular problem may not be possible.

From the perspective of decisionmakers, the implications for constructivist versus ecological rationality are found in the kinds of policy steps to be taken or the policy actors...