Scientific Uncertainty and the Political Process Author(S): Dale Jamieson Source: Annals of the American Academy of Political and Social Science, Vol

American Academy of Political and Social Science

Scientific Uncertainty and the Political Process Author(s): Dale Jamieson Source: Annals of the American Academy of Political and Social Science, Vol. 545, Challenges in Risk Assessment and Risk Management (May, 1996), pp. 35-43 Published by: Sage Publications, Inc. in association with the American Academy of Political and Social Science Stable URL: http://www.jstor.org/stable/1047890 Accessed: 05/06/2010 19:36

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=sage.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Sage Publications, Inc. and American Academy of Political and Social Science are collaborating with JSTOR to digitize, preserve and extend access to Annals of the American Academy of Political and Social Science.

http://www.jstor.org ANNALS, AAPSS, 545, May 1996

Scientific Uncertainty and the Political Process

By DALEJAMIESON

ABSTRACTIn this article, a notion of scientific uncertainty is sketched that is in many ways different from the prevailing view. Scientific uncertainty is not simply an objective value that can be reduced by science alone. Rather,scientific uncertainty is constructed both by science and by society in order to serve certain purposes. Recognizingthe social role of scientific uncertainty will help us to see how many of our problems about risk are deeply cultural and cannot be overcomesimply by the application of more and better science.

Dale Jamieson is professor of philosophy at the University of Colorado, Boulder, and adjunct scientist in the Environmental and Societal Impacts Group at the National Center for Atmospheric Research. He received his Ph.D. from the University of North Carolina at Chapel Hill and previously taught at North Carolina State University and the State University ofNew York.He has also held visiting positions at Cornell, Oxford, and Monash University in Australia. He is the editor of five books and has published many articles on ethics, science policy, and environmental philosophy. 35 36 THE ANNALSOF THE AMERICANACADEMY

of the most controversial The conventional wisdom about SOMEpublic policy decisions in Ameri- why science is often so ineffective in can society involve risks that are pri- providing solutions to problems with marily understood through scientific important scientific dimensions fo- processes and institutions. The evi- cuses on the role of uncertainty. In dence for climate change, for exam- this view, problems such as climate ple, comes mainly from experiments change are characterizedby high lev- run on highly complex climate models of scientific uncertainty about the els rather than from our everyday likelihood and effects of key events, experience. Other issues with impor- and so partisans of various policies tant scientific dimensions include can use-or misuse-scientific infor- ozone depletion, biodiversity loss, mation and authority for their own acid rain, and exposure to radon and purposes. For example, although the various toxic chemicals. Without sci- weight of scientific evidence suggests ence and scientists, there would be that large-scale emissions of green- little public concern about a wide house gases are likely to change cli- of range important issues. mate, there are so many uncertain- science has been ef- Although very ties about the roles of clouds, carbon fective in these issues into bringing sinks, and various possible feedbacks the it has been public arena, quite that both greenhouse "hawks" and ineffective at solutions. providing "doves"can reasonably enlist science There are a number of views about as an ally while accusing their oppo- this is the case. Over lunch and why nents of misusing The at scientists science.3 only professional meetings, way out of this situation, some argue, often complain about the lack of un- is for uncertainties to be reduced to derstanding or downright perversity the at which science can deter- the of leaders who point on part political mine a rational What is scientific information. On the policy. ignore needed is a new of other generation super- hand, many policy analysts remote fault scientists for to each computers, greater sensing talking and a and more ac- other rather than capability, larger producing "policy- tive research relevant" own community. science.1 My view, In the conventional uncer- which cannot be view, fully developedhere, is seen as an is that the characteristics of sci- tainty objectivequantity very whose value can be reduced in- ence that enable it to have its unique by vesting in more science. While this cultural authority as a knowledge may usefully be thought of as one of producer disable it from bringing several understandings of uncer- public decisions to closure.2 tainty, it is at best simplistic and mis- 1. See, for example, E. S. Rubin, L. B. Lave, and M. G. Morgan, "Keeping Climate 3. The typology of greenhouse "hawks," Research Relevant," Issues in Science and "doves," and "owls" is developed in Michael H. Technology, 8(2):47-55 (1991-92). Glantz, "Politics and the Air Around Us: Inter- 2. I have developed this view more fully in national Policy Action on Atmospheric Pollu- a number of papers. See, for example, "Ethics, tion by Trace Gases," in Societal Responses to Public Policy and Global Warming," Science, Regional Climate Change: Forecasting by Anal- Technology and Human Values, 17(2):139-53 ogy, ed. M. Glantz (Boulder, CO:Westview Press, (1992). 1988), pp. 41-42. SCIENTIFICUNCERTAINTY 37 leading to think of it as the only or the recognitionof the chronictoxicity most important one. Rather than be- of DDT.6 ing a cause of controversy,scientific Fallibility looms large with re- uncertainty is often a consequence of spect to many health and environ- controversy.4This suggests that the mental risks. In some cases, we may social world is active in the construc- know that various exposures are as- tion and characterization of uncer- sociated with harms, but we may tainty, and if we want to understand have little idea ofwhat causal mecha- uncertainty, we need to understand nisms are at work. Although the sta- the socialfactors that help to produceit. tistical evidence may be strong enough for some to attribute causal- FALLIBILITY,UNCERTAINTY, ity, even in these cases we may worry ANDINDETERMINISM about the fallibility of such claims. Ourview of the matter may simply be The first in step understanding wrong-not in details, but thor- uncertainty involves distinguishing so. We not even be in a it from some related notions with oughly may position to assess the probability of which it is often confused.5 our The fact of our fal- is often conflated being wrong. Uncertainty is often with relates to libility usually-indeed, fallibility. Fallibility must but it the fact that we could be about be-ignored, constantly wrong the of to which we presents possibility bringing virtually any proposition down an entire edifice of our from the most knowledge. give assent, arises from (for "Iknow how old Uncertainty ignoring homely example, We take various features of I am")to the most exotic (for fallibility. example, a as and focus on other "I know how old the universe is"). problem given dimensions. For it is Fallibility lurks in the backgroundof example, widely scientific claims and moves to the agreed that the case for climate is weakened the fact that foreground when new evidence change by we are uncertain about the effects of comes flooding in that suggests that clouds on the climate The our previous views about some mat- system. solution is more intensive of ter were not just wrong, but deeply study cloud formation and effects. But to and profoundlywrong. The discovery clouds as an area of of the ozone hole, which was not pre- identify uncer- is to dicted by any of the atmospheric tainty presuppose that our gen-- models, is one example of this, as is eral knowledge of the climate system is not uncertain, that the climate 4. This point is argued forcefully in Brian models are and so L. Campbell, "Uncertainty as Symbolic Action basically correct, in Disputes Among Experts," Social Studies of on. This background knowledge is Science, 15:429-53 (1985). "blackboxed"--it is taken as a set of 5. Although I draw the distinctions in a assumptions from which we proceed somewhat different discussion in this way, my to try to reduce uncertainty.This ap- section is indebted to Brian Wynne, "Uncer- tainty and Environmental Learning: Recon- 6. For discussion of these cases, see D. ceiving Science and Policy in the Preventive Budansky, "Scientific Uncertainty and the Pre- Paradigm," Global Environmental Change, cautionary Principle," Environment, 33(7):4-5, 2:111-27 (1992). 43-44 (Sept. 1991). 38 THE ANNALSOF THE AMERICANACADEMY proach of taking some propositionsas or other failings.7 fixed while interrogating others is a Uncertainty should also be distin- fundamental part of scientific prac- guished from indeterminacy. Often tice. Scientific progress would be im- what appears to be uncertainty can- possible if every proposition were not be reduced because there is no problematizedin every investigation. reliable fact of the matter to be The general point can be seen from learned that directly bears on im- an everyday example: I discuss sell- proving our beliefs. At least three ing my bike to a friend. In this con- sources of indeterminism can be text, there is no uncertainty about identified: agency, underdetermina- whether I own the bike. We both take tion, and categorical relativity.8 it as given that this is the case. Of Many of the most serious environ- course, it may be that due to fraud or mental and health problems we face forgetfulness I do not own the bike. involve agency.Part of why we do not But in our discussion, these possibili- know what will happen to global cli- ties are not on the table, and so there mate in the twenty-first century is is no uncertainty about whether I because we do not know how people own the bike even though it could will behave in the future. Will they turn out that I do not. Now imagine continue to increase their use of fossil a situation in which we are highly fuels? Or will other energy sources be suspicious of each other: it is well substituted? Will governments un- known that I was once convicted of dertake policies to geoengineer cli- running a bike theft ring, or that I mate? Will there be other responses suffer from amnesia. When the con- to early signs of global warming? text is changed in one of these ways, These are just a few of the questions the problem of uncertainty may whose answers matter in determin- arise. My friend may demand proof ing what will happento future climate. that I really own the bike before she Similar questions could be raised will continue the discussion with me. about the effects of tobacco smoke, What this homely example shows is the prevalence of HIV,and so on. that while we can always be wrong The indeterminism that results about (most) things, uncertainty re- from agency is made worse by the fact that about human behav- quires particular contexts and social predictions conditions. ior can themselves change the behav- is Indeed, this very example has im- ior that being predicted. Consider a case. At 8 a.m. on a warm plications for uncertainty about risk. simple summer the local radio station Uncertainty disappears or is mini- day, that there will be massive mized when we have complete trust predicts traffic as thousands of in the institution, person, or data set jams people flock to the beach. The traffic that is being interrogated. It is mag- jam nified or accentuated when there is 8. In addition, some have argued that indeterminism is a of mistrust, whether founded on fraud fundamental property nature. See, for example, John Dupr6, The 7. For further discussion, see Paul Slovic, Disorder ofThings: The Metaphysical Founda- "PerceivedRisk, Trust, and Democracy,"Risk tions of the Disunity of Science (Cambridge, Analysis, 13(6):675-82(1993). MA:Harvard University Press, 1993). SCIENTIFICUNCERTAINTY 39 fails to materialize. Many people on, say, average temperatures? And heard the radio broadcast and de- why bring together in the single class cided to stay home. of extreme events such diverse phe- A second source of indeterminism nomena as hailstorms, droughts, flows from the underdetermination hurricanes, heat waves, cold snaps, of theory by data.9Any particular ob- and so on? What are the baselines servation is consistent with an indefi- from which the claims of increasing nite number of logically distinct theo- frequency or increasing temperature ries. For example, the observation are projected?What may appear to be that there are a variety of life forms an increase from a baseline of 50 is consistent both with evolutionary years ago may appear to be a de- theory and creationism. Often we try crease from a baseline of 500 or 5000 to distinguish theories by designing years ago. Of course, stories can be a crucial experiment, one in which told about why one form of categori- distinct theories support different zation is better than another; the predictions. But there are distinct point is that empirical investigation theories that cannot be distinguished presupposes categories, without be- in this way. In such cases, people ing able to justify them empiricallyin often appeal to conceptual concerns advance. in order to justify the choice of one Rather than being epistemological theory over another--one theory is problems, fallibility and indeter- simpler, coheres better with other be- minism are metaphysical conditions. liefs, and so on. While there may be We have no idea how to overcomeour grounds for preferringone of two em- fallibility or how to tame those re- pirically equivalent theories, in such gions of the world that are indetermi- cases there is no empirical fact of the nate. Uncertainty, on the other hand, matter about which theory is true; is an epistemological problem. Un- rather, the matter is indeterminate, certainty arises fromignoring our fal- for there are no empirical discoveries libility and winking at indetermina- that would support one theory at the cies. What allows us to do this is a expense of the other. substratum of conventions, shared The third source of indeterminism purposes, common contexts, and col- is even morebasic than the other two. lective knowledge. Uncertainty is Knowledge claims presuppose cate- producednotjust by narrow scientific gories, but categories are relative. mechanisms but also by broad cul- For example, some people point to tural processes. Assertions of uncer- increases in global mean tempera- tainty are not just expressions of our tures and extreme climatic events as ignorance but part of what brings evidence of global warming. But why order to our world. Uncertainty im- is global mean temperature a signifi- plies both the existence of certainty cant category?Why not instead focus and the existence of a path from one to the other. Claims of uncertainty reflect and establish 9. The classic argumentfor underdetermi- epistemological nation can be foundin W.V.O.Quine, Word and order and imply a research program Object(Cambridge: MIT Press, 1960). and a way of moving toward closure. 40 THE ANNALSOF THE AMERICANACADEMY

THEUSES OFUNCERTAINTY just another playground for compet- Sometimes, uncertainty claims are ing ideologies. While it is true that scientific un- used in attempts to bring pol- directly and the debates that it en- debates to closure. For example, certainty icy can be corrosive to scientific the precautionary principle, which genders those who see scientific un- has been endorsed by various nations authority, and international bodies, states certainty as destructive and delegiti- overlook the fact that virtu- (roughly) that if an action or policy mating all to various conflicts potentially has catastrophic effects, ally parties then we should refrain from under- have an interest in maintaining sci- The interest of sci- taking it even if the probabilities are entific authority. On the other hand, some entists in maintaining scientific uncertain.1o But scientists argue that unless it is certain that an authority is obvious. action or policy will have harmful also benefit from the right amount of consequences, then it should be per- uncertainty. If there is too much un- mitted. Both views figure in the cli- certainty, an area of research looks mate change debate. Greens argue hopeless; if there is too little, re- that since there is a significant search appears not to be needed. The chance that climate change will occur right amount of uncertainty supports and have catastrophic consequences, a call for further research. we should "purchase some insur- Political actors of whatever ideo- ance" by capping greenhouse gas logical outlook have an interest in emissions. "Browns"argue that un- preserving scientific authority be- less it is certain that greenhouse gas cause science can provide a rationali- emissions will cause catastrophic cli- zation for decisions that are made on mate change, we should not impose other grounds. When a policy deci- the costs on the economy that cap- sion can be presented as dictated by ping emissions would entail. science, it is a way for a decision Direct appeals to uncertainty are maker to evade responsibility for his rarely effective in bringing policy de- or her choice. A decision backed by bates to closure. Instead they often science can be viewed as implied by open the door to the spectacle of du- the nature of things, not as a decision eling experts-scientists of equal for which a leader should be held training and stature who have dia- accountable. Although political ac- metrically opposed views about what tors have an interest in preserving is the case and what ought to be done. scientific authority,they also have an If the experts cannot agree about, for interest in keeping it in its place. The example, climate change, what is an optimal role of scientific information ordinary person to think? Rather for decision makers is to enable and than providing a rational means for structure decisions, not to determine resolving epistemological differ- them. ences, uncertainty reduces science to What I have suggested is that sci- 10. For further discussion, see Budansky, entific uncertainty mediates between "ScientificUncertainty." the closed world of scientific knowl- SCIENTIFICUNCERTAINTY 41 edge and the open world of public eral project of predicting future cli- policy formation. If what I have said mate on the basis of computermodels. is correct, the cultural imperative Scheduling reductions in uncer- with respect to scientific uncertainty tainty is a third way in which uncer- is not simply to reduce it but more tainty is managed. The 1990 IPCC generally to manage it. In a recent report speaks confidently of reduc- article, Shackley and Wynne have tions in uncertainty that will occuras identified some of these management a result of better data sets and more strategies."1 powerful computers. In 1988, the Quantifyinguncertainty is one way British Department of the Environ- of managing it. In 1990, the Intergov- ment laid out a 25-year plan for elimi- ernmental Panel on Climate Change nating all of the uncertainties with (IPCC)estimated that a carbondiox- respect to future climate.12 Of course, ide doublingwill producean increase no one knows exactly how these un- of global mean temperatures of 1.5- certainties will be eliminated. None- 4.0 degrees centigrade. This estimate theless, simply attaching a date to summarizes the results of some ex- their elimination appears to make periments run on what are regarded the problems more tractable. to be the best climate models. The IPCC estimate does not a represent IMPROVINGDECISION MAKING probability estimate nor any kind of normal distribution. Yet specifying Despite the fact that scientific un- this range as the likely result of a certainty plays a functional role in carbon dioxide doubling sets limits our public decision-making process- on the uncertainties, thus making es, many people are unhappy about them more manageable. how we make decisions that have im- Locating uncertainty is another portant scientific dimensions. The way of managing it. When a climate core of the unhappiness, I believe, is model fails to successfully retrodict a that the gap between science and pol- past climatic condition, this could be icy seems unnecessarily wide. As a regarded as evidence against the society,we have a large investment in model. Typically,however, the uncer- science, yet science seems to influ- tainties are located not in the model ence policy only indirectly. Science but in the data that the model ma- and policy can be brought into closer nipulates. We are directednot toward contact,but there is a price that must a fundamental rethinking of the be paid. Here are some positive sug- model but toward improving our data gestions for how science can be collection.When the uncertaintiesare brought into closer contact with pol- located in the data rather than in the icy questions. models, they do not threaten the gen- First, greater attention can be paid to problem definition at the be- 11. Simon Shackley and Brian Wynne, ginning of a decision-makingprocess. "RepresentingUncertainty in Global Climate When policy problems are not Scienceand clearly Change Policy:Boundary-Ordering defined and characterized,it is Devices and Authority," Science, Technology quite and Human Values (in press). 12. Ibid., p. 25 (in manuscript). 42 THE ANNALSOF THE AMERICANACADEMY unclear what scientific informationis tact with public decision making.14 relevant to bringing them to closure. Science, as it is practicedin American Better problem definition involves society, is an elite institution, to a being clear not only about what ques- great extent self-governing,with pri- tions are being asked but also about mary allegiance to its own internal the context in which they are asked values. While many people have ac- and the purposesthat answersto these cess to the deliverances of science, questions are supposed to serve. The very few people are involved in the debate over climate change policy is production of science, and scientists an example of how things can go themselves are overwhelmingly wrong when there is little agreement white, male, and upper middle class. about what question is being asked. Finally, there are various reforms Some people claim that it is uncer- in our public decision-making pro- tain whether emitting greenhouse cesses that would also help to bring gases will change climate; others science into greater contact with pol- seem to deny this. In some cases, they icy. As things now stand, science and are not really disagreeing. Both par- science policy are scattered through- ties to the dispute may agree that, for out the federal government. We are the purposes of counting as scientific one of the few industrialized nations knowledge, the proposition is uncer- that does not have a cabinet-level tain. More research needs to be done, department of science and technol- data collected, and so forth. But those ogy.Moreover, the adversarial way in who seem to deny that there is sig- which policy debate is conducted in nificant uncertainty are often claim- this country may also have the effect ing not that there is no scientific un- of marginalizingor needlessly proble- certainty but that there is no matizing scientific information. Sci- uncertainty for the purposes of public entific institutions are in many ways decision making. In their view, the authoritarian and directed toward risk of climate change is known to be the creation of consensus and thus great enough, and the costs of mitiga- are often at odds with the prevailing tion and prevention are low enough, values of policy debate. that some "no regrets" strategies Broad changes in the areas I have identified would scientific in- ought to be pursued. This is an exam- bring formation into closer contact with ple of a case in which it is clear that but as a science the scientific data may rightfully be policy, result, would become less and regarded as uncertain for some pur- autonomous, public decision become more poses but not for making might others."3 technocratic. Even if it were Second,various reforms in the prac- thought that this was not too to tice of science would help in bringing price high pay, scientific information into closer con- the effect that science would have on

13. These suggestions are more fully developed in Charles Herrick and Dale Jamieson, 14. I have discussed some of these reforms "The Social Construction of Acid Rain," Global in "What Society Will Expect from the Future Environmental Change, 5(2):105-12 (May Research Community," Science and Engineer- 1995). ing Ethics, 1(1):73-80 (1995). SCIENTIFICUNCERTAINTY 43 policy decisions would still remain certain, is of limited effectiveness in limited. shaping people's view of the world. One reason the role of science would remain limited is that our CONCLUDINGREMARKS most important public policy decisions involve questions of value that In this article, I have sketched a cannot be addressed by science. Asec- notion of scientific uncertainty that is ond reason why science would con- in many ways different from the pre- tinue to have a limited role relates to vailing view. Scientific uncertainty is the prevailing cultural attitudes that not simply an objectivevalue that can frame our decision-makingpractices. be reduced by science alone. Rather, We are living in a time in which citi- scientific uncertainty is constructed zens are deeply insecure about their by both science and society in order own futures and those of their chil- to serve certain purposes. Recogniz- dren and have very little trust in in- ing the social role of scientific uncer- stitutions of any sort. In such an at- tainty will help us to see how many mosphere, the bonds of community of our problems about risk are deeply are strained and the willingness to cultural and cannot be overcomesim- make trade-offs is limited. Against ply by the application of more and such a background,science, however better science.