Risk Management Under Uncertainty : Strategies for protecting health and the environment

Sammanfattning: This thesis deals with problems related to the management of environmental and health risks and problems. The general hypothesis of this work was that there are possibilities of making considerable improvements in the decision analysis and management of complicated environmental and health risks and problems. Specific attention has been paid to decision-situations involving great uncertainty and to how different decision-makers handle these extraordinarily difficult risk management situations. The studies focus on the risk management of the threat of anthropogenic climate change and chemical hazards, and on the legal and administrative risk management system in Sweden.A general conclusion is that there is room for considerable improvement in the risk management of environmental and health risks in the investigated areas. Increased use of decision theory and development of priority-setting strategies are two important tools for such improvement. Further, increased attention to the following issues will also contribute to improved risk managment: 1) the demarcation of the analysis, 2) surprises and worst cases, 3) separation and identification of scientific and non-scientific issues, 4) the distribution of the burden of proof, and 5) the scientists' and regulators' different roles.It can be concluded from the experience reported in the four papers in this thesis that it is fruitful to combine different perspectives when studying risks and the management of risk in a comprehensive approach. Three perspectives have been identified and used: (I) the actor perspective, (II) the cause perspective and (III) the risk/problem perspective. The three perspectives complement each other by giving different kinds of information eventhough they are closely linked together. A general model applicable to all kinds of risks has been developed. It serves to conceptualize the three perspectives and the three levels where it is possible to take action against risks/problems. Papers II and IV are detailed studies of more narrow issues compared to the other two more comprehensive studies.In paper (I) we compare three legal and administrative systems of risk management in Sweden, namely those responsible for work environment, environmental protection, and chemicals control. Large differences were found in terms of the organizational structures and the general modi operandi of government activities in the three areas. Many of these differences seem to be the result, not of deliberate choice but rather of the lack of coordination between policy areas. It is concluded that systematic comparisons of experiences from different areas can be helpful in improving the efficiency of risk management.Paper (II) is a case study of how the Government, governmental agencies and private companies acted when a dye stuff and a chemical marking agent were introduced to the Swedish market. Their activities are compared to the Swedish act and ordinance on chemical products. The introduction of these chemicals was given big attention in the Swedish newspapers and other media during the winter of 1993/94 when several health problems were reported and related to these substances. It is concluded that the introduction of the chemicals did not comply with the Swedish chemicals legislation. The responsibility to make risk assessments and the reversed burden of proof principle are two fundamental bases of the legislation which were set aside. It is also concluded that the unusual actions and role of the Government in this decision process may explain the actions and positions of the authorities and the companies.Paper (III) is a decision-theoretical study of an extraordinary complex risk management and decision problem - the threat of an anthropogenic climate change. The problem involves a multitude of both natural and social causal factors and a large amount of scientific uncertainty. We found that the method used in all economic decision-studies of global warming that we are aware of is more risk-taking than the standard method used in risk analysis - the expected utility method. We argue that more attention should be paid in the scientific community to less probable but more serious effect scenarios and that a less risk-taking method of analysis should be used. We also argue that the standard method of discounting future costs is inadequate, misleading, and probably not consistent with the concept of sustainable development. We conclude that the principle of sustainable development, the scientific uncertainty, the risk of irreversible damages and the long time lag from negotiations to the point in time when most of the reversible damages will be gone, all together imply that we should reduce anthropogenic emissions of greenhouse gases and that we should act as soon as possible.Paper (IV) examines reasons for and the importance of analysing worst case scenarios and surprises in the area of anthropogenic climate change. Three main reasons are given for assuming that a worst case scenario may be realised: 1) The unpredictibility of the climate system, 2) historical experiences, and 3) conceivable severe scenarios and mechanisms. Although the uncertainty of surprises is huge by definition, and the probability of worst case scenarios cannot be assessed, there are reasons for and ways of incorporating them into the risk analysis. Not including them in the analysis could very well mean that the largest risk is excluded and that decision makers make a more risk-taking decision than they would otherwise have made. However, the IPCC (Intergovernmental panel on climate change) have avoided analysis of the possibility of worst case scenarios and major factors that could give rise to such scenarios. Possible reasons why the IPCC have not included them in their analysis are discussed.