Science or politics in the global greenhouse? : a study of the development towards scientific consensus on climate change

The major aim of the report is to analyse to what extent, and possibly how, the level of scientific consensus on global warming has changed over the last twenty years, with a particular focus on the IPCC process (Intergovernmental Panel on Climate Change). The report is organized in four main parts. In the first section, I develop the analytical framework that is needed in order to understand the mechanisms at work in politically salient scientific consensus processes. Changes in the level of scientific consensus as brought about by the IPCC is made the dependent variable of the study. Three main dimensions are constructed to help us measure possible changes in consensus level: representativeness, process legitimacy and the substantive edge of consensus. The last category is divided into two parts. If scientific uncertainty strictly speaking is reduced, the pure substantive edge dimension is strengthened, while it is weakened if a consensus process leads scientists to widen the uncertainty ranges. Developments along the soft substantive edge are developments that concern aspects of scientific uncertainty that do not belong to pure or basic science. In brief, the soft substantive edge of a scientific consensus can be strengthened due to e.g. reduced risk-aversion in the handling of uncertainty, 'progressive' changes in the scientists' perception and interpretation of uncertainty and its implications, or improved skills at communicating scientific knows and unknowns across scientific disciplines, and to policymakers and the public at large. The robustness of scientific consensus is an indicator of overall change along all dimensions just listed. Two 'constants' or basic structural variables are important elements of the explanatory framework of scientific consensus processes. The scientific problem structure is probably the most crucial one, and is made up of the following three indicators: the extent of heterogeneity, the extent of maturity and the extent of scientific uncertainty. Scientific problem structures that are heterogeneous, relatively immature and riddled with fundamental uncertainty are considered malign. The actual character of a scientific problem structure significantly conditions the scope for changes in the level of consensus, and not least the possible influences of more 'temporal' independent variables. Turning to the political problem structure, its main feature is viewed to be the affectedness of political actors to policies that are believed to be influenced by for instance a scientific consensus that gradually grows more robust. The stronger one expects to be touched by political action or inaction vis-a-vis a given problem, the stronger incentives one has to try to manipulate the developments of a scientific consensus process. In brief, the scientific and political problem structures of global warming are found to be perhaps unprecedently malign. They are also related, in that scientific malignancy makes a political problem structure more malign, and vice versa. The scope for changes in scientific greenhouse consensus, as well as the possible influences of other independent variables, are strongly conditioned by the state of the problem structures. The 'temporal' independent variables are grouped in three different clusters. Standard epistemic factors is the common denominator for variables that generally are viewed as either elements in or byproducts of 'rational scientific puzzle-solving'. Non-epistemic factors make up the opposite category. Included here are elements like political lobbying or manipulation, ostracism of scientists or scientific disciplines from debates in scientific journals or consensus processes, opportunism (conscious or unconscious) taking place when (or rather if) scientists skew research practices or results in order to please political authorities or funding agencies. Somewhere in the midst between the two other categories we then find non-standard epistemic factors. Elements included here are 'organization' of scientific consensus processes, the workings of 'epistemic communities', the possible influence wielded by political entrepreneurs' and 'leaming' by individual scientists. The second part of the study is devoted to a comparison of the state of scientific greenhouse knowledge at different points in time. Four 'threshold points' are selected: i) Two international scientific assessments in Stockholm, 1971 and 1974, ii) the World Climate Conference in Geneva, 1979, iii) a major scientific assessment finalized in Villach, 1985, and iv) the IPCC process itself. My main aim is to establish a notion of how the level of scientific consensus on global warming has changed over the last twenty years; concerning representativeness, process legitimacy and the overall level of scientific uncertainty. In the third section, the focus is exclusively on the IPCC process. Three politically salient issues under scrutiny by IPCC are selected in order to judge whether, and possibly to what extent, political pressures have influenced the IPCC proceedings on these accounts. The issues are: i) how certain are we that anthropogenic greenhouse gases influence the global climate?, ii) the relative contribution of different energy sources to global warming, and iii) the greenhouse contribution of deforestation and biotic methane releases. The last chapter summarizes and evaluates the empirical findings of the study. First, the changes in the level of scientific consensus since 1970 are assessed, and the overall robustness of the IPCC consensus as compared to earlier assessments is discussed. Then, a part of the chapter is devoted to explanation of the possible driving forces behind the changes observed. Finally, I discuss the relationship between scientific consensus and political consensus; i.e. how scientific consensus influences political consensus, and the possible influences that changes in political greenhouse consensus wield on the state of scientific consensus