Vous êtes sur la page 1sur 11

Science and Public Policy, volume 26, number 3, June 1999, pages 151161, Beech Tree Publishing, 10 Watford

Close, Guildford, Surrey GU1 2EP, England.

Paradoxes
Scientific expertise and political accountability: paradoxes of science in politics
Peter Weingart

Two paradoxes form the nucleus of the problems of scientific expertise and policy-making. The first is the simultaneous scientification of politics and the politicisation of science. This has destructive effects: the increased use of scientific expertise by policy-makers has not increased the degree of certainty, in fact it becomes de-legitimating. This gives rise to the second paradox: despite the loss of authority of scientific expertise, policy-makers do not abandon their reliance on existing advisory arrangements, nor do the scholars adapt their ideas on science and its relation to politics. How can this stability be achieved? How can sciencepolitics be institutionalised?

Professor Peter Weingart is at the Institut fr Wissenschafts- und Technikforschung, Universitat Bielefeld, Postfach 100131, D-33501 Bielefeld, Germany; Tel: +49 521 106 4655; Fax +49 521 106 6033; E-mail weingart@uni-bielefeld.de.

N THE EVENING of 28 March 1979, Walter Cronkite opened the CBS (Canadian Broadcasting Service) evening news by saying that this had been the most confused day in the history of the news media. He then went on to present the contradicting information from the Three Mile Island Nuclear Reactor at Harrisburg, Pennsylvania. where an accident had occurred. This situation continued for several days, and three weeks later the news magazine Newsweek wrote that one of the first victims of this accident had been scientific credibility. Seven years later almost to the day a much more severe accident at a nuclear power plant in the Ukrainian city of Chernobyl in effect dealt the death blow to nuclear power in most European countries. In Germany, which was particularly affected by a radioactive cloud that swept with shifting winds over the entire country, a spectacle unfolded similar to that in the USA after Three Mile Island. Politicians and scientific experts engaged in media debates which culminated in a politicians ostentatiously eating salad in front of the national TV audience in the futile attempt to demonstrate that vegetables were not affected by the caesium fallout. Three Mile Island and Chernobyl, became symbols not only of the vulnerability of nuclear power technology but more of the loss of scientific authority based on reliable knowledge and unanimous expertise, as well as of the loss of credibility of politicians who relied on it (Weingart, 1979; Krohn and Weingart, 1986).1 Although the public controversy over nuclear power had been going on for some time, the profound impact it had on the relationship between politics and science became apparent to the general public only when the debates between experts
151

Science and Public Policy June 1999

0302-3427/99/030151-11 US$08.00 Beech Tree Publishing 1999

Paradoxes of science in politics


Peter Weingart studied economics and sociology at the universities of Freiburg, Berlin and Princeton, and holds a chair of sociology of science at the University of Bielefeld, Germany. He was director of the Centre for Interdisciplinary Re s e a rc h (Z iF ) 1 9 8 9 1 9 9 4 , Fe l l o w of t h e Wissenschafskolleg 19831984, and currently heads the Institute for Science and Technology Studies at Bielefeld. He has published numerous books and articles in the sociology of science and science policy.

and politicians began to be publicised widely by the media. Several things became visible which contradicted the simpler concepts of scientific expertise and policy-making: The involvement of experts in political controversy exemplified by their dissenting views in public hearings or written testimony. The legitimating function of scientific knowledge in the political arena. Conversely, the de-legitimating function of scientific knowledge, publication of which it was feared would mobilise the public against powerful economic and political interests. The vulnerability of politicians and parties which relied on scientific knowledge as a political resource.

The phenomena are interrelated, and they reinforce each other. The reactions to the two accidents on the part of the politicians and scientists involved could make one wonder about the lack of sophistication, but a kinder interpretation is that they reflected strongly-held concepts about the nature of scientific knowledge, of policy-making, and of the combination of the two. These concepts proved to be an illusion which was shattered. By the time of these accidents, scientists in high circles of government had long become such an accustomed sight that few would have stopped to wonder if this had always been the case. Fewer yet would remember that this had been considered a dangerous liaison as recently as 30 years ago. Then, the most pressing issue was the unaccountable and thus illegitimate influence of scientific experts in democratic governments. Within the relatively short time span of roughly 50 years, we have witnessed the establishment of scientific advice to governments on a large scale and, at the same time, a fundamental change in the concept of science on the part of the observers of the new relationship. This is not just an academic issue. The changes that have occurred in the conceptual apparatus and our perceptions of science are themselves indicative of the process under investigation. For that reason, the first part of the analysis is devoted to the shifts in the conceptions of scientific advice in politics.1 A paradox emerges: by the time scientific expertise
152

in government has become a ubiquitous feature, the initial concerns about the illegitimate influence of scientific experts have vanished almost completely. What, then, has happened to the original concerns about the accountability of experts? The thesis is that the crucial change which has occurred is the general accessibility, in principle, and factual use of scientific expertise by all groups in the political spectrum. This democratisation of expert knowledge has calmed the fears that experts would undermine the democratic mandate.2 Thus, the demise of the lite status of scientists (which previously gave some justification to the concerns) goes hand-in-hand with the democratisation of knowledge in the political arena. From this emerges a second paradox: given the legitimating function of (authoritative scientific) knowledge in politics, the general accessibility of that knowledge has led to a competition for expertise which intensifies controversies in policy-making rather than alleviating them. Finally, a third paradox, or rather a surprising observation, is that, in spite of the loss of authority on the part of the experts and, thus, of their legitimating value, the arrangement remains in place. In the following I will, first, recount the shift from earlier to more recent models of scientific expertise in policy-making. Then I will analyse the causes and consequences of the stated paradoxes, and, finally, conclude with a summary and observations on the future of the sciencepolitics coupling.

Early models
In the immediate post-war years, when in the USA the military and the scientific community had just dissociated themselves from their successful experience in the Manhattan Project, scientists and policy-makers began to think about the future objectives of a science policy and how to give scientific advice to the President. In 1947, President Truman received the Steelman Report which suggested the establishment of an Interdepartmental Committee on Scientific Research and Development (ICSRD) to monitor federal support for research as well as the creation of a science advisor in the White House. Already in this report the duality of functions, which later came to be called policy for science and science for policy was evident. Truman set up the ICSRD, which never gained any importance, but he did not create a science advisor. Only in 1951 did he follow in part a recommendation to establish a Science Advisory Committee (SAC) which was to provide independent advice on scientific matters especially as regards the objectives and interrelations of the several Federal agencies engaged in research of defense significance (White House Press Release, 1951). Scientific advice was considered relevant by policy-makers only in connection with the military, and therefore SAC was established within the Office of Defense Mobilization (ODM). This only illustrates
Science and Public Policy June 1999

Paradoxes of science in politics

The science advisory structure in the US government paved the way for similar arrangements in other countries and opened a new era in the relationship between science and politics ultimately transforming both the political and science systems
that in its early days the institutionalisation of science in government was closely connected to the Cold War. In 1957, as a reaction to the launching of the Russian satellite Sputnik, President Eisenhower named James R Killian as Special Assistant for Science and Technology and took SAC out of the ODM. As the Presidents Science Advisory Committee (PSAC), it was now directly responsible to the President as part of the Executive Office. The establishment of this science advisory structure in the US government which paved the way for similar arrangements in other countries opened a new era in the relationship between science and politics and ultimately transformed both the political system and the system of science. The first signs of this change came in early. Science had not only gained direct access to the centre of power and thereby hitherto unknown political influence, it had also become established as a source of military technology and then, by implication, a source of public wealth to such an extent that its claims to public funds with hardly any accountability seemed practically irrefutable. The immediate reaction to this new role of science, which appeared to observers to be very close to an establishment ... a set of institutions supported by tax funds, but largely on faith, and without direct responsibility to political control, came from Eisenhower himself (Price, 1967, page 12). In his farewell address, he formulated his famous warning that the nations public policy might become the captive of a scientific-technological elite (New York Times, 1961). Concern about influence of science With this alarm cry, he set the agenda for a discourse on the dangers of scientific experts in government which was to be carried on for another decade or more. A plethora of writings appeared featuring concepts such as scientific power lite, new priesthood, scientific estate, and new mandarins.3 The common concern of these writings had also been expressed in C P Snows tale of the Tizard Lindemann rivalry during the war years in the British government. Snow was deeply worried about the advent of closed politics, the main feature of which was that all countries are not unlikely to be at the mercy of scientific salesmen. Snow (1961, pages 57,
Science and Public Policy June 1999

63) characterised as court politics the relationship which Lindemann had with Churchill, giving him more direct power than any scientist in history. Price did not want to go quite so far but nevertheless thought that the scientific revolution has upset our popular ideas about the way in which policies are initiated and adopted, and in which politicians can control them and be held responsible for them. In his view, the influence of the scientific estate had increased, it was exercised as a matter of advice, not legal right, and it could only be described as a part of the unwritten constitution (Price, 1967, pages 17, 170). The debate reflected the novelty of sciences role in politics, best exemplified by Prices description as the scientific revolution. The basis of experience was still small and, thus, the concerns were real. Yet obviously these anxieties rested on a number of assumptions regarding both the nature of science and its role in the policy process. First they were all related to the discrepancy between the actual influence of scientific knowledge on political decisions (perceived to be considerable) and the lack of public accountability of the science advisors. The threats were seen in varying ways: C P Snow feared undue influence of individuals in secret politics, Don Price saw more reason to worry that decisions of a scientific or technological nature would be popular, that scientists would lobby for particular programmes, and promise technological miracles to get funds for basic research (Price, 1967, page 81). It is noteworthy that he foresaw both a conflict of interest in the advisory role and an invasion of science into the policy-making realm of government. Science could not be limited to providing the means for the attainment of previously established goals. What Price feared most was that the scientists would use the inevitable conflation in their own interest (Price, 1967, pages 181, 262, 277).4 One element of these concerns was the belief that the science advisors influence rested not only on their access to the centres of power but also on the superior knowledge they commanded. Given the nature of the problems facing policy-makers, the fear was that they were not competent to deal with them but instead had to rely on scientific knowledge. Reactions to this perception were by no means unanimous. Opposing those who feared an erosion of the principles of representative democracy were those who hailed the end of ideology or even of politics altogether (Bell, 1960; Lane, 1966; Schelsky, 1965). This view was obviously complementary to the former in so far as the basic diagnosis with respect to the role of scientific knowledge was quite similar and only its evaluation with respect to the fate of politics was reversed. The entire discourse became dubbed the technocracy debate and remained a dominant theme throughout the 1960s and into the 1970s. Habermas reminded readers that the debate actually reflected two major traditions in political thought. The growing reliance of public administrations and
153

Paradoxes of science in politics

Habermass model captures best the iterative process of the definition of problems, their translation into research issues, their re-definition in the light of available new knowledge, and the translation of knowledge into decisions
policy-makers on specialised knowledge had already been analysed by Max Weber as the rationalisation of modern states. The influence of scientific knowledge represented a second stage of this process of rationalisation. However, Weber, in the tradition of Hobbes, and after him Carl Schmitt, insisted on the clear distinction between special knowledge (Fachwissen) and political practice, between the functions of the expert (Sachverstndiger) and the politician. Whereas the choice of means could be rationalised, the choice among values and objectives remained irreducibly irrational. Decisionist and technocratic models This decisionist model of the relationship between political power and (scientific) knowledge, which assumes a clear distinction between objective knowledge and subjective values, seemed to be called into question by an increasing scientification of politics, which would not only render problematic the legitimacy of irrational decisions, but reduce the range of options to an objectively determined singular best decision. In this technocratic model, whose tradition can be traced to Bacon, the politician becomes fully dependent on the expert. Politics is replaced by a scientifically rationalised administration. The crucial and problematic assumption of this model is the notion of a quasi-natural, one-dimensional direction of scientific and technical development (Habermas, 1966). It reflects perhaps the different political traditions that the technocratic model attracted much more attention, if not positive reception, in continental Europe, especially in German discussions, while the academic discussions in the USA leaned more to the Weberian decisionist consensus. With this background, it is ironic that Habermas referred to Dewey to advance what he termed a pragmatist model. In this, the strict separation between politicians and experts is replaced by a critical interrelationship. Habermas envisaged a reiterative communication process between the two sides, in which the development of new technologies would be directed by interpreted value systems, and, at the same time, the interests reflected in these value systems would be controlled by examining them in the light of technical possibilities and the strategic means of their satisfaction.
154

Thus they are partly accepted, partly refuted, reformulated and critically examined for their ideological character (Habermas, 1966, page 134). If we ignore, for the moment, the idealistic elements of this model, Habermas came closest to a realistic description of the advisory process, albeit for reasons other than those he stated.5 His model captures best the iterative process of the definition of problems, their translation into research issues, their re-definition in the light of available new knowledge, and the translation of knowledge into decisions. Because the focus was on the new institutional arrangements of scientific advice to governments, nobody thought to look in the other direction. Scientific expertise was seen as a threat to representative democracy or hailed as a solution of its shortcomings. Yet the consequences which a full-blown integration of science into the political process would have on science itself did not appear as problematic (see Weingart, 1982). Science was regarded as immutable and the scientific community perceived as a homogeneous social group. The simplifications that the models entailed were presuppositioned by theories in political science and the philosophy of science. Since then, and as a result of knowledge generated by science studies, a very different picture has emerged. In this, the fate of the previous problems (the experts public accountability as well as the political responsibility of policy-makers) has taken a surprising turn.

Recursive model
Although sociologists of science have questioned them at an early stage, the two unilinear models of science in policy-making to some extent still dominate perceptions among policy-makers and advisors alike.6 This is all the more surprising as both models represent the two sides of the legitimation dilemma with which policy-makers and their experts are confronted in modern mass democracies. In principle, decisionist policy-making suffers from a rationality deficit, whereas technocratic decisions suffer from the lack of legitimating public consent. Current democracies operate on the basis of two circuits of legitimation of public decisions (Roqueplo, 1995, page 176; compare with Ezrahi, 1990). Decisions have to be rational in the light of existing knowledge, and they have to be made by representatives of delegated power. The legitimating authority for the former is science, which brings experts into governments, for the latter it is public consent, which is normally obtained in elections. At least three basic assumptions underlying the technocratic and decisionist models do not stand up to empirical test: the linear sequence of (political) problem definition, (expert) advice and (political) decision; the value freedom of scientific knowledge; and
Science and Public Policy June 1999

Paradoxes of science in politics

the disinterestedness or political neutrality of scientists.

If, as Jasanoff observes, neither the technocratic notion of the separation of truth from power can be upheld, nor the democratic or decisionist model carry sufficient legitimacy without support from science, a new conceptualisation of the relationship between science and technology is called for. The model of scientific advice to politics which exemplifies these interdependent processes is close, though not identical, to that Habermas described as the pragmatist model. In contrast to his, it is based on empirical evidence rather than philosophical reasoning. The following critique of these assumptions suggests that two interdependent processes best describe and explain the constellation in question: the scientification of politics goes hand-in-hand with the politicisation of science (Weingart, 1982; 1983). Sequence of problem definition, advice and decision As to the first assumption, detailed studies of decision-making processes show that scientists take part in, or have varying degrees of influence on, the form-ulation of the problems put to them. The degree of that influence depends among other things, on the nature of the problem. At the beginning of the chain of problem perceptionexpert advicepolitical decision, not only interests and values but also knowledge play a crucial role. One evidence of scientification is that many issues put on the political agenda are a product of perception through science. For example, the issue of environmental protection was put on the political agenda when scientific knowledge indicated that DDT had accumulated in the food chain (van den Daele, 1979, page 21). The problem of ozone depletion became an issue of international negotiation and agreement only after two scientists, Molina and Rowland, had advanced the hypothesis that CFCs (chlorofluorocarbons) played a crucial role.7 The same is true for a host of other issues. In fact, the discourse on risk, which conspicuously since the mid-70s has been tied to key technologies, is in large measure driven by research probing into potential dangers before the background of expectations of

certainty (Wiedemann, 1993, page 57; Weingart, 1991, page 14). In other words, science plays an increasing role in defining the problems, on which it is then called to give advice, once these problems are on the political agenda. Science is one actor among many in the political system and takes part in setting the political agenda, be it as an interested party, or be it because other actors, such as the media, are interested in the pronouncements of science. Value freedom of scientific knowledge The second assumption of the advisory models that scientific fact and political values can be clearly distinguished, is untenable as well. There is an obvious connection between a positivist concept of science and a decisionist model of scientific advice in politics: according to this view certain knowledge is brought to bear on the choice of means to attain previously chosen ends. Thus, earlier discussions about the role of science in politics focused on its instrumental function only, that is, its role to provide reliable answers to problems.8 This very utility of scientific expertise makes it an inherently desirable commodity in political decision-making. At the very moment scientific experts were admitted to the councils of governments for their instrumental role, they inadvertently assumed a potential legitimating function as well. Today this seems almost a truism but, in fact, the legitimating role of science in politics is a fairly recent phenomenon, having appeared systematically only together with the advent of modern mass democracies and a general recognition of the instrumental value of scientific knowledge. Consequently, it is now commonly accepted that different functions of scientific expertise can be distinguished. Boehmer-Christiansen (1995, page 197) lists ten different functions, among them: legitimacy; persuasion, delaying or avoiding action; justification for unpopular policies; arbitrating disputes; and clarification of conflicting interests. These different functions elucidate the complexity of the relationship between experts and policy-makers at the interface between two systems with their diverging logic of operation. However, on a higher level of abstraction, all these functions can be subsumed under two basic ones: instrumental and legitimating.9 The legitimating function is particularly pronounced in debates over issues with a strong scientifictechnical content. The protracted debate over the safety and economic feasibility of nuclear power revealed for the first time to the public that scientific experts were engaged to support partial positions of governments and interested industries. As this happened, it also became evident that experts disagreed along the lines of the adversaries in the controversy. The public became witness to specialised academic debates, to the fact that scientific knowledge could be uncertain and controversial among the experts.
155

One evidence of scientification is that many issues put on the political agenda are a product of perception through science: environmental protection was put on the political agenda when science indicated that DDT had accumulated in the food chain

Science and Public Policy June 1999

Paradoxes of science in politics

Although the USA was to lead in this development, it soon spread to Europe where, foremost in Germany, it triggered the emergence of a green movement and ultimately of a Green Party that became a political force to stay.10 Since then, a multitude of scientifictechnical issues has captured public attention: the safety of recombinant DNA, the ethics of reproductive technologies, the application of biotechnology in agriculture, the ethical, political and economic implications of sequencing the human genome, the depletion of the ozone layer, the implications of CO2 emissions for anthropogenic climate change, the transferability of mad cow disease (BSE) to humans, and other less visible ones. In all these controversies, it has become commonplace that the adversarial parties, be they governmental or non-governmental groups, engage scientific experts to present evidence which supports their respective views. However, if scientific knowledge can be used to legitimate different political positions and decisions, it is obviously not the one-dimensional, hard and objective truth which only relates to one particular solution. Scientific knowledge cannot be separated as neatly from value judgements as both the decisionist and the technocratic model of advice suggest. The relationship between knowledge and decision-making is obviously much more complex: it involves issues such as: how problems are framed; which knowl- edge can be mustered; the stage of certainty, that is, the degree of consensus over knowledge; the ways in which it can be interpreted; the number of answers to the problem it provides; and how that knowledge relates to social values and political interests. If scientific knowledge is linked in any way to interests (in policy-making), it is evaluated as supportive, contradictory, or even dangerous. Knowledge inevitably comes under these evaluative verdicts once it enters the public arena and is considered politically relevant. This is, again, an aspect of the politicisation of science inseparable from the scientification of politics. Political neutrality of scientists The third assumption implied in the decisionist and technocratic models, that scientists remain disinterested in the advisory process and only convey objective knowledge in pure form, is a myth. This is well documented by the many instances in which scientific experts have taken part on opposing sides in controversies in which they had a role as advisors. Sociologists of science have known this for a long time, and, though often by implication, government agencies acknowledge it as well.11 Interest of scientists can take different forms, either as advocacy for one position in a controversy, or as self-interest in favour of the research community to which the scientist belongs. To cite the ozone debate as just one example of many: immediately after publication of the Molina/ Rowland (1974) hypothesis, two coalitions formed,
156

one favouring rapid regulation of CFCs (scientists, environmentalists, consumers, politicians), another opposing such regulation (CFC producers and consumers, lobbies of the chemical industry, scientists and politicians). The core of the proponents of regulation were the few scientists who had published the crucial reports. They became spokespeople and activists on behalf of their convictions even before they could prove their hypotheses. At least after 1986, German scientists, too, openly took position in the media in favour of regulation. Their opponents among the scientists were not a homogeneous group, their interests in opposing rapid political action on CFCs were diverse. Some downplayed the danger in order to receive more research grants to study the problem, others were paid by the CFC producers. When the controversy started, their opposition was equally unfounded by scientific data. The ozone debate, among others, demonstrated the role of scientists as political actors, regardless of the differences in political systems or motives or interests that observers may attribute. Another aspect of this politicisation is the way in which scientific experts are being recruited into their advisory roles. The public controversies point to the most visible incidents of the legitimating function of science in policy-making. It is these that have led Edward Shils (page 201) to comment: Advisors are too frequently chosen not so much because the legislators and officials want advice as because they want apparently authoritative support for the policies they propose to follow. It is obvious that in complying with these desires, the legislators and officials are in collusion with the scientists to exploit the prestige that scientists have acquired for objectivity and disinterestedness. This is obvious in the German system, which is more corporatist than the US system. Scientists who are recruited into the so-called Enqute commissions (set up by parliament) are selected in proportion to the political parties share in parliament and, of course, by their choice. In the commissions, they sit together with politicians which binds them into whatever compromises are struck. The danger is that their advice may be diluted. The advantage usually is that once recommendations are formulated by the commissions, they get translated more directly into political decisions (Grundmann, 1999, chapter 6).12 This arrangement in itself is a sign of the politicisation of science. The crucial point about the politicisation of scientists is that from the outside their positions in a controversy are identifiable and determined by their politics rather than by available knowledge. Intern-ally, this is inescapable. Because of the involvement of the scientific expert in the decision-making process, he/she inevitably assumes a political role. Or as Roqueplo (1995, page 176) put it:
Science and Public Policy June 1999

Paradoxes of science in politics

The crucial point about the politicisation of scientists is that from the outside their positions in a controversy are identifiable and determined by their politics rather than by available knowledge
What transforms scientific knowledge into an expert appraisal is its inscription within the dynamics of decision-making. Yet this inscription, at least in the case of scientifically and politically complex questions, immediately leads the scientist to express opinions or convictions which (however scientifically founded) cannot in any way be identified with knowledge in the strict sense which science generally affords this term. It is a systematic result of this coupling rather than the idiosyncratic behaviour of individuals that experts often fall prey to the temptation to oversell their expertise to provide recommendations far beyond their realm of knowledge (Renn, 1995, page 149). Simple models flawed From the above it should be evident that the simple models of scientific advice, the decisionist or democratic and the technocratic models, were flawed from the beginning because of naive views of the nature of scientific knowledge, of scientists behaviour and, in view of the actual operation of existing institutional arrangements, of scientific advice to politics. On account of the agenda-setting role of scientists alone, the linear conception of the advisory process has to be replaced by a recursive model. Instead of the linear chain of problem perceptionexpert advicepolitical decision, the process has to be modelled as a recursive loop, which entails the following stages: Perception of the problem may come either from the scientific community or from policy-makers. In the political process it is transformed according to political criteria of relevance. As a political programme funding research for further clarification of the initial problem, it is handed back to the scientific community. The scientific community, in turn, executes the pertinent research whose results become the basis for continuous adaptation of the initial problem perception.13

implementation of funding programmes for policy-oriented research) the findings are in remarkable accordance with Jasanoffs studies of the advisory processes in the American regulatory agencies (compare with Jasanoff, 1990, especially page 234). The recursive nature of the relation between science and politics as exemplified in the case studies on which the model was based underscore the scientification of politics. The first paradox is thus resolved: the initial concerns about the accountability of experts disappear because access to scientific advice has become demo-cratised and the resulting relationship between science and policy-making appears as recursive and reciprocal rather than linear. Yet now another paradox emerges.

Inflationary use of scientific expertise


The legitimating function of scientific knowledge in politics as well as the political agenda-setting role of scientists are the main instances which indicate the reciprocity of the scientification of politics and the politicisation of science. These terms denote a dynamic link, which is the key to understanding the role of science in present-day politics and some features not normally associated with it. It has become fashionable to speak of a blurring of the boundaries or of hybridisation. However, it is misleading as it incorrectly suggests that the functional differentiation between science and politics disappears. The novel aspect of the use of scientific expertise is the close coupling of science and politics. This coupling is dynamic in the sense that it is driven from both sides to become ever closer. With this, another paradoxical development emerges. The legitimating function of scientific knowledge, once discovered, has initiated a competition among political actors for the most recent research results. In regulatory practice, this mechanism was institutionalised some time ago. This is highlighted in Germany where regulatory practice is proceduralised on the basis of the formula according to the state of current knowledge in science and technology (nach Stand von Wissenschaft und Technik). This means that, in all cases in which regulatory agencies and ultimately courts, that is, judges, have to decide on the release of new pharmaceuticals or the operation of new technologies such as nuclear power plants, the decision is based, not on normative criteria derived from legal provisions and interpreted by judges, but on the pronouncements of scientific experts. The administrative courts in whose jurisdiction these decisions fall have thus, in principle, deferred a large part of the responsibility for these decisions to the respective communities of specialists, even though the actual decision rests with the judges. The crucial issue obviously is who is justified to pass judgement on the state of science and technology,
157

This is an idealised reconstruction which means that the different stages are not necessarily neatly separated either institutionally or temporally. However, even though this model is drawn from studies of a particular class of advisory processes (those which entail policy for science issues, that is, the formulation and
Science and Public Policy June 1999

Paradoxes of science in politics

The quasi-experimental implementation of new technologies with the implication of unknown risks has led to a marked increase in the number of experts admitted to, and called upon by, administrations and eventually the courts
and what counts as such when the judgment is conflicting (Hagenah, 1986; 1995). With the development of nuclear power technology, the arrangement between the legal system and science has changed significantly in several respects. The safety issues have become too complex for courts to judge on their own by simply referring to the textbook literature. Complexity refers, in this case, to the fact that technologies are implemented at a stage when research into their safety, health and environmental impacts still goes on and no fixed and consensual standards have been formulated. In fact, this quasi-experimental implementation of new technologies with the implication of unknown risks has become a more common pattern. This has led to a marked increase in the number of experts admitted to, and called upon by, administrations and eventually the courts, an observation first made with respect to the regulation of nuclear power but probably generalisable for other areas of regulatory practice.14 The same fact is reflected in Germany in the emergence of a sizeable infrastructure of expert commissions (for instance, the reactor safety commission (RSK), the radiation protection commission (SSK)) which are supposed to advise the government and monitor research for that purpose.15 The manifest function of these commissions is to deal with the risks of the technologies for which they are set up. Expertise is used instrumentally, but the latent function is to absorb uncertainty, in other words, to use scientific expertise for legitimating purposes (Krcken, 1996). Non-scientists decide between conflicting advice The seeming paradox is that the intensified use of scientific expertise has not increased the degree of certainty on the part of judges, administrators and policy-makers; on the contrary, it has left them witnessing the ongoing debates among scientific experts and forces them to decide between conflicting advice. Put more concisely, the competition for, and inherently inflationary use of, scientific advice for legitimating (and even instrumental) purposes is selfdestructive and de-legitimating. The paradox arises because, in principle, the competition for the latest, and therefore supposedly
158

most compelling, scientific knowledge drives the recruitment of expertise far beyond the realm of consensual knowledge right up to the research frontier where knowledge claims are uncertain, contested, and open to challenge. The dependency of the regulatory and legal system on the progress of research codified in the formula according to the state of science and technology dramatises the dynamism which is entailed in the basic mechanism of scientification, namely that in principle all statements are revocable (Weingart, 1983, page 228). In many instances there is consensus over the state of knowledge, but the competition between political adversaries for legitimating knowledge pushes the demand for expertise in the direction of yet uncertified knowledge, that is, controversy. The inherent dynamics of this process is exacerbated in policy-making contexts where, because of the nature of the problems at stake (for instance, practical problems such as risk management) facts are uncertain, values in dispute, stakes high and decisions urgent (Funtowicz and Ravetz, 1993). The fact that the process is driven from within is the essence of the close coupling of science and politics. Even where the legal peculiarities of the German system are not to be found, basically the same mechanism is at work. Looked on from up close, the example of regulatory practices by administrative courts illustrates how, paradoxically, scientification of politics literally produces its opposite, the politicisation of science. It should be noted here in passing that it would be fruitful to look on the entire debate on risk and uncertainty in the light of this mechanism. The discourse on risk and uncertainty, which is so intimately connected to the issue of science in government, is not only propelled by continuous research into potential hazards of technologies. This research constitutes expectations of certainty, reliability and safety. The closer the integration of science into politics the higher are these expectations, and so on. It has not gone unnoticed that the increased expectations posed to scientific knowledge are transferred to the experts who allegedly represent that knowledge and, because they systematically exceed their abilities, are the main reason for the growing distrust of scientific expertise. Policy-makers still reliant on scientific advice This paradox gives rise to yet another intriguing observation. Given the general recognition of the loss of authority of scientific expertise and, thus, of its function as a political resource, we would assume a renaissance of political decisionism as well as a demise of any technocratic expectations. However, although the naive technocrats may have lost out, policymakers have not abandoned their reliance on existing advisory arrangements nor have scholars changed their concepts of science and politics. Therefore, the stability of these arrangements and the conceptions of
Science and Public Policy June 1999

Paradoxes of science in politics

The inflationary use of scientific authority in public debates and the ensuing loss of legitimating power has been recognised by governments and scientific communities: as both have an interest in maintaining that authority, they have both reacted

science and politics which support them are surprising and need explanation. It has long been noted that, since in modern societies science is the only source of privileged knowledge, all criticism and counter movements have the existing scientific discourse as their ultimate frame of reference (unless the basis of communication becomes something other than knowledge, for instance, revelation). There is no alternative to using scientific expertise to give political decisions an instrumentally more reliable base and provide them with a higher legitimacy, even if uncertainty is encountered. Thus, the science/politics link is firmly institutionalised and not likely to change fundamentally or disappear. The changes occur along the boundaries between science and politics.16 The relationship, or what I would rather term the novel type of coupling between science and politics could be characterised as an unstable equilibrium. The instability results from the interaction of two strategies of both science and politics in response to the paradoxes. One may be dubbed contraction, the other expansion. Although it could seem that the strategies are contradictory, in fact they are complementary. Creating an ultimate authority The inflationary use of scientific authority in public debates and the ensuing loss of legitimating power has, of course, been recognised by both governments and the scientific communities. As both have an interest in maintaining that authority they have both reacted.17 The debate in German policy-making circles to establish a National Academy of Science is in large measure motivated, though not publicly, by the fact that the country does not have an institution in its science system that could speak as the ultimate authority. By establishing such an institution the Federal government would introduce a hierarchy within the system which would, it is hoped, neutralise the many expert voices in public controversies that are recruited by non-governmental organisations. (The attempt has failed so far for reasons unrelated to the issue.) Another international example of the same strategy is the mechanism of the Intergovernmental Panel
Science and Public Policy June 1999

on Climate Change (IPCC). IPCC is a consortium of climate research institutes and policy-makers from across the world who, before reporting their research findings and recommendations to the media, create a consensus among themselves in confidentiality. In this way, the experience of the nuclear power debate is avoided, when contradicting experts neutralised each other and presented the image of an academic farce (Elzinga, 1996). Here again, the attempt to contract the amount of expert knowledge to augment its authority and utility is obvious, and up to this point in the case of the IPCC this policy has been very successful. In essence, the attempt at contracting the knowledge admitted into the policy process is an effort to control the influx of knowledge and thereby the de-legitimating effect of contradicting pronouncements of scientific experts. The second strategy of expansion is a response of the scientific community only. It, too, is a reaction to the abundance of scientific expertise in policymaking. However, here the issue for certain communities is how to capture attention for their causes. The answer is seen increasingly often in overselling to the media. There is evidence that fields of science try to capture media attention, and thereby indirectly the attention of policy-makers by establishing what could be called the catastrophe-cycle.18 The physicists have thrived on such a scheme for many years by playing on both the military threat of nuclear destruction and the civilian threat of an end to energy sources. When the latter lost its impact to the even greater fear of nuclear accidents, the last attempt was to play on the threat of climate change as a result of the greenhouse effect.19 Here it can be demonstrated that the issue of climate change was captured by the meteorologists who since then have cooperated with the media in monitoring this ultimate global threat. Again, the pattern is obvious: originally, rather crude extrapolations indicated anthropogenic temperature increases with catastrophic water-level rises, desertification, storms and the like. Meanwhile, although the models used are still much too simple, the predictions have become more cautious. However, climate research has become a thriving field, and climate change is a new focus of policy-making affecting many other areas. When serious warnings of man-made threats are issued by science, they usually have an immediate impact because of the general credibility of science, but the paradoxical effect already becomes visible as the discourse of ever-growing risks and man-made threats tends to become inflationary. The credibility of science is challenged by counter claims alleging that science is speaking in its own interest.20 Once these allegations have become established as part of the discourse, it will be very difficult to differentiate between reliable knowledge and biased information. This difficulty will no longer be a choice between different institutions such as science and religion: it will affect science proper.
159

Paradoxes of science in politics

Conclusion
The institutional reality of todays coupling of science and politics may now be summarised and focused. The basis of the relationship is that of an exchange of resources for knowledge which has an instrumental and a legitimating aspect. The inherent dynamics of this relationship is a result of the operation of several mechanisms: The increase of resources which leads to more, possibly an abundance of, knowledge. The inherently limitless (over-)production of knowledge creates for politics technological potentials and/or political expectations, threats to legitimacy and, thus, imperatives to act (= scientification of politics). Here the media play a decisive role in amplifying and thereby structuring the supply of knowledge (expansion). The demand for knowledge to solve scientific/technical (instrumental) and/or obtain support for preconceived decisions (legitimating) ultimately leads to the inflationary use of expertise and, as it pushes to the realm of yet uncertain knowledge, to controversy (politicisation of science).
3. 4. 5.

6.

7. 8.

9.

10. 11.

The (over-)supply of knowledge and its politicisation leads to de-legitimation of politics and loss of authority on the part of science. Thus, it creates imperatives of selecting among variants of expertise. This selection must operate internally to science and may take the form of pooling and monopolising expertise (contraction of knowledge supply) on the part of science and of creating institutional hierarchies of expertise (contraction of knowledge demand). It is evident that these interacting mechanisms, which constitute what I term the coupling of science and politics, do not come to a stable state. Rather, it is foreseeable that there will be no end to the continued production of knowledge which, to capture public attention and support, will be sold on promises and threats, a strategy that could be self-defeating in the long run. Likewise, it is foreseeable that there will be no end to the competition among policy-makers for the most recent scientific knowledge in the search for its legitimating power. Thus, the boundary between science and politics has to be constantly redrawn and reiterated.

12.

13.

14. 15. 16. 17.

18.

Notes
1. 2. For an overview from the vantage point of science studies but also in the persistence of views, see Cozzens and Woodhouse (1995). This is not to say that the problem of accountability has disappeared altogether. There remains a fundamental tension between political power and its legitimation on the one hand and scientific knowledge on the other, which becomes immediately apparent in governments persisting interest to keep expert advice confidential because it is considered a threat to

19.

20.

the autonomy of political decision-making (Roqueplo, 1995, page 177). Apart from Prices book (1967) see Gilpin and Wright (1966), Lapp (1965), Lakoff (1966). This fear proved not to be unwarranted, see Weingart (1980) and Greenberg (1967). To be sure Habermas was not alone and observers like Price not naive to take the simplified models for granted. Price noted that the process of responsible policy-making is a process of interaction among the scientists, professional leaders, administrators, and politicians; ultimately authority is with the politicians but the initiative is quite likely to rest with others, including the scientists in or out of government (Price, 1967, page 67). This point will assume importance later on. See Jasanoff (1990). Jasanoff distinguishes between the democratic and the technocratic model: in the democratic model, legitimacy of decisions accrues from the participation of the populace. The decisionist aspect is only camouflaged, however, since decisions have to be reached by, albeit representative, political leaders. Thus, there is no difference between the decisionist and the democratic model. The article which triggered the debate was Molina and Rowland (1974). Renn distinguishes enlightenment, pragmatic or instrumental, interpretative and catalytic functions all of which fall under what I call instrumental. He does acknowledge that policy-makers overemphasise ..expert advice because it provides legitimacy to their decisions and opportunities to use experts as scapegoats (Renn, 1995, page 147). These functions of expertise refer to policy-making. We could take the position of science, too. In that case, the instrumental function of expertise refers to its uses in obtaining resources for research, the legitimating function refers to obtaining public support and acceptance of research. One of the early studies in the USA, see Nelkin (1971). For a detailed analysis of the Austrian controversy, see Nowotny (1979). For Germany, see Kitschelt (1980). See Jasanoff (1990, page 230) and the literature cited there. Boehmer-Christiansen (1995, page 199) states that the model underlying much of expert advice in Europe is naive, rationalist and far too linear. In this particular case, Grundmann argues, the political style in each of the countries, adversarial in the USA, consensual in Germany, does not explain the outcome of the ozone debate, because they practically inverted into their opposites during the 1980s (compare with Grundmann, 1999, page 335). Compare the reconstruction of four funding programmes in Germany (Environmental Research, Nuclear Fusion, Biotechnology, Electronic Data Processing) and the War on Cancer in the USA in van den Daele et al (1979). For a detailed account of the emergence of the Program for the Environment (Umweltprogramm) see Kppers et al (1978). Hagenah (1986, page 106) for nuclear power in Germany; DiFabio (1990) for regulation of pharmaceuticals in Germany. Compare Willke (1992) on the increased importance of this knowledge-based infrastructure in the regulation of risk. Although alluding to Gieryns (1995) work, here the issue is not the boundary between science and non-science. There are no reliable data on the numbers of expert advisors in the German Federal government. The Ministry for Science and Technology (BMBF) reduced the number of its advisors in 1983/84 from 1033 to 540, and the number of advisory panels by about 30% to 63. In 1994, the number of expert advisors was near 300. This does not necessarily indicate that expert advice is consciously contracted to augment its impact. Rather, because of the introduction of indirect funding mechanisms it is diffused to other organisations (BMFT-Journal 1984). Boehmer-Christiansen (1995, pages 202, 203) lists fusion, many health issues, cancer, limits to growth, star wars as examples aside from the environmental domain. In each case, unsubstantiated promises or threats are made. Global warming is the threat through which Global Change research is funded. In Germany, this connection is explicit: the German Physics Society (DPG) launched a memorandum together with the German Meteorological Society (DMG) (DPG/DMG, 1987) warning of climate change in order to save nuclear power. Compare with Engels and Weingart (1997). The NASA (National Aeronautics and Space Administration) press conference in February 1992 on ozone depletion over

160

Science and Public Policy June 1999

Paradoxes of science in politics


the Arctic and its echo in the press is an incident which demonstrates the increased vulnerability of science under the circumstances described (compare with Grundmann, 1999, page 392). Regulierung technisch-kologischer Gefahren in der Risikogesellschaft (Opladen, Wiesbaden). G Kppers, P Lundgreen and P Weingart (1978), Umweltforschung Gesteuerte Wissenschaft? (Suhrkamp, Frankfurt). S A Lakoff (editor) (1966), Knowledge and Power ( New York). R E Lane (1966), The decline of politics and ideology in a knowledgeable society, American Sociological Review, 31, pages 658. R E Lapp (1965), The New Priesthood (Harper and Row, New York). M J Molina and F S Rowland (1974), Stratospheric sink for chlorofluormethanes: chlorine-atomcatalysed destruction of ozone, Nature, 249, 28 June, pages 810812. D Nelkin (1971), Nuclear Power and its Critics (Cornell University Press, Ithaca and London). New York Times (1961), 22 January, page 4E. H Nowotny (1979), Kernenergie Gefahr oder Notwendigkeit (Suhrkamp, Frankfurt). Don K Price (1967), The Scientific Estate (The Belknap Press of Harvard University Press, Cambridge MA) page 12. O Renn (1995), Style of using scientific expertise: a comparative framework, Science and Public Policy, 22(3), June, pages 147156. P Roqueplo (1995), Scientific expertise among political powers, administrations and public opinion, Science and Public Policy, 22(3), June, pages 175182. H Schelsky (1965), Der Mensch in der wissenschaftlichen Zivilisation, in H Schelsky (editor), Auf der Suche nach Wirklichkeit (Bertelsmann Universittsverlag, Dsseldorf) pages 439480. E Shils, Science and scientists in the public arena, The American Scholar, 35, pages 85202. C P Snow (1961), Science and Government (Harvard University Press, Cambridge, MA) pages 57, 63. W van den Daele, W Krohn and P Weingart (editors) (1979), Geplante Forschung (Suhrkamp, Frankfurt) page 21. P Weingart (1970), Die Amerikanische Wissenschaftslobby (Bertelsmann Universittsverlag, Dsseldorf). P Weingart (1979), Das Harrisburg-Syndrom oder die DeProfessionalisierung der Experten, in Nowotny (1979) pages 917. P Weingart (1982), The scientific power elite - a chimera ; the de-institutionalization and politicization of science, in N Elias, H Martins and R Whitley (editors), Scientific Establishments and Hierarchies. Sociology of the Sciences Yearbook, Vol VI (Reidel, Dordrecht) pages 7187. P Weingart (1983), Verwissenschaftlichung der Gesellschaft Politisierung der Wissenschaft, Zeitschrift fr Soziologie, 12(3), pages 225241. P Weingart (1991), Large technical systems, real-life experiments, and the legitimation trap of technology assessment: the contribution of science and technology to constituting risk perception, in T R LaPorte, Social Responses to Large Technical Systems, Control or Anticipation (Kluwer, Dordrecht) pages 517. White House Press Release (1951), 20 April, cited in J L Penick et al (editors), The Politics of American Science, 1939 to the Present (Rand McNally, Chicago) page 182. P M Wiedemann (1993), Tabu, Snde, Risiko: Vernderungen der gesellschaftlichen Wahrnehmung von Gefhrdungen, in Bayerische Rck (editor), Risiko ist ein Konstrukt Wahrnehmungen zur Risikowahrnehmung (Knesebeck, Munich) pages 4367. H Willke (1992), Ironie des Staates, Grundlinien einer Staatstheorie polyzentrischer Gesellschaft (Suhrkamp, Frankfurt).

References
D Bell (1960), The End of Ideology (Free Press, Glencoe, Ill). BMFT-Journal (1984), 2, page 11. S Boehmer-Christiansen (1995), Reflections on scientific advice and EC transboundary pollution policy, Science and Public Policy, 22(3), June, pages 195203. S E Cozzens and E J Woodhouse (1995), Science, government, and the politics of knowledge, in Jasanoff et al (1995, pages 533553). U DiFabio (1990), Verwaltungsentscheidung durch externen Sachverstand, Verwaltungsarchiv, 81, pages 193227. DPG/DMG (1987), Warnung vor drohenden weltweiten Klimvernderungen durch den Menschen, Physikalische Bltter, 43(8), pages 347349. A Elzinga (1996), Shaping worldwide consensus: the orchestration of global climate change research, in A Elzinga and C Landstrm (editors), Internationalism and Science (Taylor Graham, London) page 228. A Engels and P Weingart (1997), Das politisierte Klima die Entstehung anthropogenen Klimawandels als politisches Hand-lungsfeld, in P Hiller and G Krcken (editors), Risiko und Regulierung (Soziologische Beitrge zu Technikkontrolle und prventiver Umweltpolitik, Frankfurt). Y Ezrahi (1990), The Descent of Icarus (Harvard University Press, Cambridge MA). S Funtowicz and J Ravetz (1993), The emergence of post-normal science, in R von Schomberg (editor), Science, Politics and Morality. Scientific Uncertainty and Decision Making (Kluwer, Dordrecht) pages 85123. T F Gieryn, Boundaries of science, in S Jasanoff et al (1995, pages 393443). R Gilpin and R Wright (1965) (editors), Scientists and National Policy Making (Columbia University Press, New York). D S Greenberg (1967), The Politics of Pure Science (The New American Library, New York). R Grundmann (1999), Transnationale Umweltpolitik: zum Schutz der Ozonschicht. USA und Deutschland im Vergleich (Campus, Frankfurt). J Habermas (1966), Verwissenschaftlichte Politik in demokrati-scher Gesellschaft, in H Krauch, W Kunz and H Rittel (editors), Forschungsplanung (Wien, Mnchen) pages 130144. E Hagenah (1986), Stand der Wissenschaft Stand des Rechts, Report Science Studies 30, Bielefeld. E Hagenah (1995), Prozeduraler Umweltschutz (Nomos, Baden Baden). S Jasanoff (1990), The Fifth Branch (Harvard University Press, Cambridge MA) page 15. S Jasanoff, G E Markle, J C Petersen and T Pinch (editors) (1995), Handbook of Science and Technology Studies (Sage, Thousand Oaks). H Kit s c h e lt (1 9 8 0 ) , K er n e n e r gi e p o l i t i k , A r e n a e i n e s gesellschaftlichen Konflikts (Campus, Frankfurt). W Krohn and P Weingart (1986), Tschernobyl das groesste anzunehmende Experiment, Kursbuch, 85, pages 125. G Krcken (1996), Risikotransformation: Die politische

Science and Public Policy June 1999

161