Aspen Global Change Institute Elements of Change 1996

AGCI Session II: Characterizing and Communicating Scientific Uncertainty

Session Chairs: Dr. Richard H. Moss and Dr. Stephen H. Schneider

July 31 to August 8, 1996

A Social Constructionist View of Scientific Objectivity

Simon Shackley

Lancaster University

Lancaster, United Kingdom

It is assumed by many that greater precision, quantification and clarification of uncertainty will create more persuasive science for policymaking purposes and will help advance the climate change policy agenda. While there is much value in this, Shackley questions whether it is as important a priority as commonly assumed. Empirical studies have illustrated the importance in successful environmental policymaking of policy coalitions between environmental groups, industry, scientists, and others with a stake in the outcome. The parties in these coalitions generally have substantial differences in their outlooks, aims and objectives, but they revolve around a core of consensus that might be thought of (following Hajer) as a "storyline."

A storyline is an account of an issue in terms of its causes and effects, the impacts of remedial actions and assumptions about the motivations and behavior of the key policy actors involved. Storylines simplify arguments and interpretations over a wide range of scientific, economic and policy issues. A specific storyline serves to hold the policy coalition together despite substantially different interpretations, values and objectives. It is simply not possible to encapsulate all the complexity surrounding many environmental issues in any one definitive, agreed-upon, account. Such "closure" is not, fortunately, required for effective political and policy action. And as is discussed later, ambiguity and uncertainty can actually help hold the coalition together.

Studies have illustrated the importance in successful environmental policymaking of policy coalitions between environmental groups, industry, scientists, and others with a stake in the outcome.

Referring to a theory from Ted Porter's book Trust in Numbers , Shackley says that the desire for greater precision through quantification may be symptomatic of "weak" institutions which feel distrusted. It is possible that weak institutions collectively attempt to re-build trust through using storylines which rely on quantification. The latter have an appearance of objectivity and rationality which assists in holding policy coalitions together and is difficult to argue against by those not involved (even though quantification does not automatically imply less ambiguity). However, Shackley says that it is questionable whether this sort of quantification will secure political consensus in the way that it might have done 50 or even 20 years ago. Public alienation and distrust of governments has increased and policy making institutions frequently send out conflicting signals (environmental protection versus increased consumption, for example). In this new political context, it is not clear that quantified storylines will succeed in re-building trust between governments and the public.

Political Influences on the Scientific Process

Shackley then turned to examples of where science is being "constructed" in advisory contexts in order to illuminate how non -scientific factors enter into the formulation of scientific advice. Shackley began by discussing changes in the global warming potential (GWP) for methane in the IPCC Working Group I 1994 assessment. The 1992 IPCC report expressed reduced confidence in the 1990 estimates and declined to estimate indirect effects. Following the 1992 report, two new three-dimensional atmospheric chemistry models were developed which accounted for indirect effects of methane. The two models did not produce the same value but the researchers were able to agree on a new GWP for methane with an uncertainty of plus or minus 35 percent, based on the range of the models' results and the modelers' interpretation.

The anticipated political implications had a direct effect on how the GWP and the uncertainty associated with it was represented in the report.

At a meeting of lead authors in Geneva in July 1994, several authors (notably Robert Watson) expressed concern that the new GWP value for methane would have a great deal of political significance because it meant that on a 10 to 20 year time frame, methane might be a more important contributor to climate change than CO2. Because these authors could foresee a controversy over this new value, they stressed the importance of clarifying how the new GWP was calculated, and in particular, how the uncertainty range was determined. The point Shackley makes here is that the anticipated political implications had a direct effect on how the GWP and the uncertainty associated with it was represented in the report.

Later that year, in the plenary session for Working Group I, methane's GWP was one of the most controversial issues. Some government delegates, particularly New Zealand's government scientists, raised concerns about this and an ad hoc meeting was held to discuss the issue. (New Zealand has very high methane emissions, primarily due to the large numbers of sheep.) The New Zealand government scientists expressed their belief that a 35 percent uncertainty was too tight a range, representing more certainty that we actually have about this value. They argued that policy might be made based on what they saw as provisional figures policies such as trading emission reductions in methane for increases in CO 2. This argument was rejected, and one of the key people in this rejection was Robert Watson, who said it was not up to scientists to tell policymakers how they could use GWPs. Shackley calls Watson a classic example of a "gatekeeper:" a scientist and policymaker who is very influential in deciding how science will be expressed and transferred to the policy arena.

Shackley alluded to the analogy with Ozone Depletion Potentials (ODPs), in which case the negotiators requested a single value for each ozone-depleting chemical. Yet, if advisory scientists accede to this request, is there not the danger that over time policymakers come to have greater confidence in the certainty and precision of the science than the scientists do themselves? Given their confidence, might policymakers not also come to believe that such precision and certainty can be expected and requested of advisory science? By contrast, if advisory scientists were to be more forthright in expressing the uncertainties, then policymakers might come to accept that it was not possible to obtain precise values.

The point of discussing this example is not that the judgment of the IPCC scientists was wrong, but that it was more than simply a scientific judgment. It involved judgments about how policy works or should work, and what policymakers need. A small number of "gatekeepers" now make these kinds of decisions. Shackley argues that more explicit discussions of these non-scientific judgments are needed. The way that scientific uncertainty is represented in advisory reports is influenced by advisory scientists' judgments about the needs of policymaking.

The way that scientific uncertainty is represented in advisory reports is influenced by advisory scientists' judgments about the needs of policymaking.

Another example of the construction of science involves the representation in the IPCC of flux adjustments in coupled ocean atmosphere General Circulation Models. A dominant theme of industry comments during the Madrid IPCC WG I Plenary in November 1995 was to emphasize flux adjustments as a reason climate model results should be seen as conditional. The advisory scientists' response to such criticism is colored by recognition of the politically-inspired nature of such attacks. In such a context, who (scientists or skeptics) introduces the uncertainty into the debate is key in the wider perception of whether the advisory process has integrity.

But Shackley also argues that the actual substance of the science, as well as its portrayal, is influenced by policy considerations. As evidence of this, he describes his analysis of why flux adjustments have been used in some climate modeling centers but not in others. Most of the debate is scientific in character, but since there are good scientific reasons for both using and not using flux adjustments, more than just science is involved, he contends. Shackley and colleagues believe that factors such as the institution's mission and funding, policy roles and relations, and relations with the climate-change-impacts community can enter into the decision as to whether or not the institution uses flux adjustments in its models. Policy implications influence the debate because without flux adjustments, GCMs cannot be used to make long-term projections, which is one of the key requests made by policymakers of climate scientists. Shackley argues that the use (or non-use) of flux adjustments therefore emerges from the complex interplay of scientific debates with institutional and policy commitments and preferences.

Instead of seeing ambiguity and uncertainty as something to avoid or purify from the climate debate, it could actually be a benefit.

The Role of Ambiguity in Coalition Formation

Shackley says that both scientific arguments, and the non-scientific context of advisory science which comes to influence advisory science, contribute to the potential ambiguity of science for policy. He believes that a certain degree of ambiguity in scientific knowledge is useful and perhaps even necessary in producing a "storyline." So instead of seeing ambiguity and uncertainty as something to avoid or purify from the climate debate, it could actually be a benefit. Some of the necessary actors in the policy coalition on climate change might drop out if the uncertainties cleared up too much. Among the ambiguities he believes are important in holding together an emergent policy coalition are those in the physical science realm, such as ambiguities regarding climate sensitivity and global warming potentials.

Ambiguities in the meaning of climate sensitivity include the rules by which GCM results are adopted in the IPCC assessment, the domain of uncertainty accounted for, changes in the "best guess" to accommodate new knowledge, and changing from a narrow definition which includes CO2 only to a wider definition which includes all GHG forcings. Ambiguity in the meaning of GWPs includes the choice of GHGs for which GWPs are calculated, which indirect effects are included and how they are included, the time horizons and discount rates used, in what parameter of climate change the GWP is being measured, in what atmospheric residence time is used for CO 2, and whether GWP is calculated with sustained or pulse emissions.

The ambiguities around climate sensitivity help to hold together the GCM community, as well as helping those using simple models and those assessing impacts to calibrate their models to the GCMs. It further allows policymakers to have easy access into the climate issue, providing an "anchoring devise" and an entry point for many different communities. It also justifies ongoing modeling activities to continue research to hone these values. The ambiguities in GWPs serve similar functions, for example bringing together research communities to address problems which had not previously been tackled such as those relating to the global carbon cycle. These ambiguities also help to accommodate the preferences of policy communities, including the desire of many governments for a comprehensive approach to all GHGs, not just CO2.

Discussing the significance of policy coalitions for the climate issue, Shackley says that there is a need to link climate change to other policy agendas and issues in a meaningful way.

Shackley presents a model of the chain from physical effects to human impacts of climate change. He calls the physical processes "up stream" and the effects on human systems "down stream." As one moves down stream in this model, and includes more knowledge, debates and uncertainties, there is an opportunity for the whole to be destroyed in intellectual, policy and political discussions. Therefore, he says, up stream processes, and ambiguities in these processes, may serve in practice as better anchoring devices. Down stream processes are less able to generate a clear storyline and to hold the coalition together.

Discussing the significance of policy coalitions for the climate issue, Shackley says that there is a need to link climate change to other policy agendas and issues in a meaningful way. Here a dilemma emerges because in order to secure these effective cross-issue linkages, the knowledge needed is of the down stream variety which is less able to secure an effective storyline. It may be that we have to revise downwards our expectation that scientific knowledge is able to hold together policy coalitions in the climate case and look to other shared knowledge and values to provide the rationale instead.

It may be that we have to revise downwards our expectation that scientific knowledge is able to hold together policy coalitions in the climate case and look to other shared knowledge and values to provide the rationale instead.

Forward to Next Section//Back to Table of Contents//AGCI Homepage//Comments: agcimail@agci.org