You are in ePLANETe >
Uncertainty and Risk
Definitions and taxonomies
The notion of uncertainty is used in many scientific fields, often encompassing a multiplicity of related concepts. In broad terms, uncertainty may be defined as being any deviation from the unachievable ideal of completely deterministic knowledge of a relevant system (Walker et al., 2003).
Uncertainty characterises most assessment, policy and management processes that have unpredictable consequences. In a risk assessment context, the United States Environmental Protection Agency refers to uncertainty as our inability to know for sure - it is often due to incomplete data (http://www.epa.gov/riskassessment/). In the Millennium Ecosystem Assessment, uncertainty is defined as an expression of the degree to which a future condition (e.g., of an ecosystem) is unknown. Uncertainty can result from lack of information or from disagreement about what is known or even knowable (MEA, 2003).
Uncertainty may have different types of sources, from quantifiable errors in the data to ambiguously defined terminology or uncertain projections of human behavior. Uncertainty measurements can therefore be represented by quantitative metrics (e.g., a range of values calculated by various models) or by qualitative statements (e.g., reflecting the judgment of a team of experts) (MEA, 2003).
Several nomenclature systems have been developed for describing the different types of uncertainties. For example, Funtowicz and Ravetz (1990) explored the differences between three sorts of uncertainty:
- Inexactness, i.e. a technical level of uncertainty involving the random and systematic errors in empirical quantities;
- Unreliability, which is related to methodological uncertainties arising, for example, from an incomplete understanding and from the approximations made when describing the structural and functional characteristics of a system under study;
- Border with ignorance, which refers to an epistemological level of uncertainty (e.g. omissions of processes and parameters due to ignorance – ignorance of ignorance).
In a context of environmental contingencies and crisis, the checklist developed by De Marchi (1995) supports the identification and ranking of different types of uncertainty (Table 1).
There is an ethical dimension to decision-making and the handling of uncertainty when the lives of others are at stake (e.g. decision to approve new drugs or chemicals that have uncertain human health and environmental consequences). Within this context, Tannert et al. (2007) developed the Igloo of Uncertainty (Figure 1) wherein dangers and risks are discriminated in the field of uncertainty – a danger is present regardless of choice, whereas a risk is either optionally accepted or imposed.
Definition
Institutional
Refers to the role and actions of institutions and their members and stems from the diversity of cultures and traditions, divergent missions and values, different structures and work styles among personnel of different organizations. High institutional uncertainty can hinder collaboration or understanding among agencies, and can make the actions of institutions difficult to predict.
Legal
It is relevant when agents need to consider future contingencies of personal liability for their actions (or inactions). High legal uncertainty may result in defensive responses in regard to both decision-making and release of information. Legal uncertainty may also play a role where actions are conditioned on the clarity or otherwise of a legal framework in allowing one to predict the consequences of particular actions.
Moral
Arises from the underlying moral issues related to action and inaction in a given issue. De Marchi notes that "moral uncertainty is linked to the ethical tradition of a given country be it or not enacted in legislation (juridical and societal norms, shared moral values, mores), as well as the psychological characteristics of persons in charge, their social status and professional roles". Moral uncertainty would typically be high when moral and ethical dimensions of an issue are central and participants have a range of understandings of the moral imperatives at stake.
Proprietary
Arises from asymmetries between potential users of information and knowledge about an issue. Some people or groups have information that others don't and may assert ownership or control over it. Proprietary uncertainty is typically high when knowledge plays a key role in assessment, but is not widely shared among participants.
Scientific
Arises from the scientific and technical dimensions of a problem and is intrinsic to the processes of risk assessment and forecasting.
Situational
Relates to "the predicament of the person responsible for a crisis, either in the phase of preparation and planning, or of actual emergency. It refers to individual behaviors or personal interventions in crisis situations" (De Marchi, 1994) and as such represents a form of integration over the other six types of uncertainty. That is, it tends to combine the uncertainties one has to face in a given situation or on a particular issue. High situational uncertainty would be characterized by situations where individual decisions play a substantial role and there is uncertainty about the nature of those decisions.
Societal
Arises when different communities (with different sets of norms, values, and manner of relating characteristic of their societies) have different approaches to decision-making and assessment. Societal uncertainty would typically be high when decisions involve substantial collaboration among groups characterized by divergent decision-making styles.
Finally, it is also important to clarify the differences between uncertainty, risk and ignorance in relation to different states of knowledge and associated examples of public action (Table 2), since what is sometimes loosely referred to as uncertainty often mixes up these concepts (EEA, 2001).
Dealing with uncertainty
According to the Post-Normal Science framework, the management of uncertainties should rely on explicit guidelines and credible set of procedures such as those provided in the NUSAP notational system. The NUSAP categories stand for Numeral, Unit, Spread, Assessment and Pedigree, enabling the different sorts of uncertainty in quantitative information to be expressed in a standardized way and presented transparently to all the actors involved in a policy process. For extensive guidance on tools for the assessment and communication of uncertainty, see the NUSAP website at http://www.nusap.net.
Adopting a precautionary approach in a context of uncertainty is often recommended as a strategy for public policy action. The precautionary principle is an overarching framework that governs the use of foresight in situations characterized by uncertainty and ignorance, where there are potentially large costs to both regulatory action and inaction (EEA, 2001). The sound application of the precautionary principle to issues of complexity, uncertainty and controversy requires the support of key elements of good governance, such as fairness, transparency and accountability (EEA, 2001).
Scenarios and forward-looking studies are practical tools that can help to explore key uncertainties and their implications across a wider range of contrasting futures. In the face of future uncertainties, scenarios and forward-looking assessments and also visions, can help to structure and explore choices by revealing their possible long-term consequences, thus supporting strategic planning and decision-making.