Yehezkel Dror, The Hebrew University of Jerusalem

Issue: The unique knowledge and cognitive perspectives involved in science give it a special role in shaping the future. Thus all policy choices, not only those directly concerned with science policy, need to take science into account. To do so, high quality science advisory systems are therefore essential. Moreover, a balance has to be struck between the need for disclosure and preventing potentially dangerous information from falling into the wrong hands.

Relevance: Mapping out the preferable features of such systems can serve as an aid to both impact assessment and to improving the science advice system. Unfortunately, such an exercise highlights the general inadequacy of current science advisory systems and their often limited impact.

Beneficial measures include the placement of science advisors on all policy staffs, training of science advisors in policy-making, improved interfaces between science advisors and other policy staffs, and providing all main policy advisors with some degree of science literacy. It will also be necessary to provide the public with a balanced evaluation of the science advice that is offered by multiple actors, together with upgrading of science literacy among the main elites and the public at large. At the same time, measures are urgently needed to prevent misuse of information disclosed during the science and technology advisory process.

Science, in the narrower sense of the "natural sciences" but including related parts of "technology", is perhaps the most important force shaping the future and has radical direct and indirect impacts on all domains of human activity. Therefore, all major governmental, social and individual choices need to take science into account. This is especially so when crucial decisions are faced whose impact on shaping the future may be considerable. Also, the modalities of thinking of the various branches of science and their worldviews are very important for comprehending complex policy issues as they can provide different "frames of thinking" and multiple perspectives.1 This fact makes science advice valuable in all policy areas, and not just in those relating to science and technology issues.

Science advice is required on all levels of social and individual choice. To explore this variety, we will look here at two main classes of users namely top-level governmental decision-makers and the public at large. The proposed analysis can relatively easily be adjusted to incorporate additional actors, though in part different science advice systems are needed for different clients and audiences.

One main function of science advice is to supply inputs into the science-related aspects of choices. However, good science advice should fulfil four additional functions regarding high-level decision-making and choice processes: it should (1) provide a minimum of science literacy essential for correctly using or rejecting science advice; (2) improve cognitive maps2; (3) revise decision agendas; and (4) apply scientific frames of thinking to the non-scientific dimensions of main choices. Providing a balanced understanding of the scientific bases of main issues on the political agenda, including a feeling for the ambiguities and uncertainties involved, is the most important service science advice can provide the public at large. The counteracting of "magical thinking" and pseudoscience, and consequently the upgrading of public discourse as a whole, is another important task.

The vast majority of senior governmental decision-makers in nearly all countries lack the minimum of science literacy required to be able to understand and use or reject science advice correctly. Furthermore, they are often unable to utilize scientific modes of thought to better comprehend complex issues. And most parts of the public in all countries are quite unable to evaluate the meanings and judge the validity of the many claims made in the name of science on topical issues, such as environmental policies, uses of biotechnology, hazardous chemicals and so on. Therefore, an important function of science advice is to provide decision-makers and the public at large with at least minimum levels of science literacy and science advice sophistication. This, in turn, requires from science advisors much more than knowledge of science.

The quality of decisions depends a lot on the cognitive maps used by decision makers and the extent to which they fit with a changing reality.3 Therefore, enhancing decision-makers' cognitive maps so as to ensure they reflect reality and its dynamics accurately is an important task for science advice. Related to the above is the role of science advice in helping with agenda setting by alerting decision-makers to issues, opportunities and dangers, especially those stemming from science and technology.

One important way to improve the way complex issues are tackled is by deepening understanding of them using a multiple-perspective approach and processing them with the help of a variety of modalities of thought. The perspectives and modalities of the various academic fields, including the socially and historically oriented ones, with all their diversity, can make significant contributions to meeting this requirement - if applied in sophisticated ways by suitably qualified science advisors.

Science advice contributions to the public at large are no less essential, both as part of the endeavour to move towards more of a "learning and knowledge society"4 and as an essential part of ensuring democratic support for better quality governmental choices.5 Indeed, without the public being more knowledgeable about science and technology, including a correct "reading" of science and technology news and pronouncements with all their contradictions and ambiguities,6 governments are driven towards counter-rational choices on many science and technology related issues.

But science advice given to the public (and also top decision-makers) has an even broader mission, namely to counteract clearly false beliefs such as in astrology. Beliefs in various forms of "magical" thinking or pseudoscience are widespread, including among high-level decision-makers. When the power of humanity and its leaders to shape the future was limited, the damage caused by such beliefs was, from a long-term historic perspective, tolerable. But given the power to shape the future, for better or worse, that modern science and technology gives humanity, the damage that can be caused by choices based on ignorance can be catastrophic. Therefore, the task of science and technology advice to improve people's views of reality and counteract growing reliance on "magical" thinking and pseudoscience is becoming increasingly crucial.

It is the growing burden imposed on science and technology advice that makes assessment (and improvement) of its quality and impact an urgent task. It may not be an overstatement to claim that radical upgrading of the quality of science and technology advice and its impact is critical for the future of humanity in terms of the path it follows, or indeed even its very survival.

A related issue is the need to control access to potentially dangerous scientific and technological knowledge, given the power of science and technology to enable ever smaller groups of people to cause ever greater harm.7 Therefore, when improving scientific advice the potential misuse of the information disclosed also has to be taken into account. This is an increasingly acute issue, and while it is one which should be kept in mind when pondering science advice impact assessment and upgrading, it is beyond the scope of this article.

A variety of methods are available for assessing the impact of science advice. Case studies can show to what extent science advice has influenced a decision and shaped its outcomes, interviews may provide some indications of impacts, writings by decision-makers and science advisors provide some information, and so on. However, all these methods are of limited feasibility and validity in practice. Diffuse and long-term effects are hard to identify and often impossible to conceptualize. The specific contributions of science advice in multi-disciplinary teamwork often cannot be identified. And the multiplicity of intervening variables often casts doubt on cause-and-effect relationships and on correlations between the quality of advice and its impact. Therefore, efforts to evaluate science advice impacts directly must be supplemented by indirect evaluation based on inputs, structure, etc.

However, caution is necessary when basing impact evaluation on indirect criteria. The latter must be based on an in-depth study of the relationships between indirect criteria and outputs in those cases where the latter can be identified. And, when this is not feasible, indirect criteria must be grounded in reliable theories dealing with organizational behaviour, knowledge processing in decision-making, policy planning and so on, together with lessons of experience distilled from knowledgeable advisors, decision-makers and opinion formers.8

Based on the arguments presented above,9 I propose the following map of the preferable10 characteristics of science advice for top-level decision-makers, with some additional features of science advice aimed at the general public. This map is meant as a tool to help assess the impact of science advice indirectly, with a view to improving it.

Location: Science advisors must be located in main decision centres and participate in central choice processes. Thus, science advisors need to be placed near heads of government with access to the principal deliberations involved in decision-making.

Tasks: The formal tasks of science advisors should include contribution of science knowledge to concrete policy spaces and decisions; and application of multiple science perspectives and thinking modalities to all main choice processes. Informal tasks of science advisors include upgrading the scientific literacy of main decision-makers and their staffs, improving their cognitive maps and putting emerging issues that arise out of science and technology on the agenda.

Composition and knowledge resources: Science advisory units need to have a multi-disciplinary composition and easy access to additional scientific knowledge and perspectives as relevant. Individual science advisors should seek help from other disciplines where necessary.

Integration with policy advisory units: To fulfil their expanded tasks in top-level decision-making, science advisors would ideally be included on staffs serving heads of governments.

Interface: A good interface between science advisors, other policy advisors and decision-makers is essential. The main responsibility for achieving such an interface lies with science advisors, who must understand the power environments in which they operate, the policy-making processes in which they participate and the information processing habits of their partners and clients. The language of science advice has to be matched to that of policy-makers (or that of the general public, where appropriate). And the behaviour of science advisors must get across the barriers raised by the "two cultures11". But all this must not reduce the integrity of the advice given, but adjustments in formatting, timing, presentation etc. are inevitable.

Self-scepticism: Science advisors must be very sensitive to the limits, ambiguities and uncertainties of science advice, especially when applied to multifarious issues on which little reliable knowledge is available. A good dose of self-directed scepticism, including a critical approach to the main assumptions beyond (or even including) the domain of knowledge we hold as incontestable,12 is therefore an important feature of high-quality science advice.

Policy professional practice: Science advisors, while being in the main scientists, are also a sub-set of policy professionals and participate in policy advice as a whole. Therefore, they need some knowledge of policy professionalism, such as how to cope with uncertainty and inconceivability and how to relate policy advice to political advice.

Special training: One of the conclusions of all the features of good science advice is that special training is needed for science advisors, knowledge of science being necessary but not sufficient for them to perform their main tasks.

It is not difficult to adjust the map given above to the preferable requirements of science advice from the point of view of the public. However, four additional features require emphasis:

• In addition to the many scientists and science writers working on their own or on behalf of various interest groups, independent "science and technology think tanks" which specialize in providing inputs into public opinion are essential.
• Without setting up "Science Courts", independent "science ethics panels" warning the public against clearly biased and sloppy science writings could be useful. The Danish experience with the scientific ethics panel may be worth exploring.13
• Special attention must be given to upgrading the science and technology literacy of main mass media commentators, as distinct from specialized science correspondents.
• The science and technology literacy of publics at large needs to be raised, inter alia by radical changes in science and technology teaching in schools (e.g. including science literacy courses in core curricula of all university students, and integration of science and technology subjects with main television news programmes).

Preventing misuse: An obvious requirement is high quality scientific knowledge and maximum efforts to provide as balanced and "objective" advice as possible. To this must be added various ethical demands on whom to serve. Also, as mentioned, it is essential to place institutional restraints on the availability of science and technology knowledge to those who may potentially misuse it. This raises the difficult issue of striking a balance between preventing potentially dangerous information from falling into the wrong hands and preserving the integrity and openness of science. However, any discourse on assessing and improving science advice that does not take up the dangers of misuse is sorely lacking.

Preliminary comparison of this partial map of the characteristics that might be desirable in science advice systems with those actually found in practice, as examined in an empirical study of mine of more than forty offices of prime ministers and presidents in a variety of countries, reveals major gaps. Most of the features of the preferred system are either met only in a small part or not at all. Furthermore, nearly all writings on the desiderata of science advice ignore many of the proposed requirements. Hence, my conjecture that science advice impact as a whole falls far below levels that are both necessary and possible, with the partial exception of domains where science inputs are of obvious importance, as in some aspects of military affairs.14

If this conjecture should be borne out by systematic studies based, inter alia, on comparing reality with "preferable characteristic" maps, which too need much improvement, then it may be possible to arrive at a more reliable and differentiated profile of science advice impact and derive concrete proposals for improvements. Additionally there is a need to locate science advice impact studies and improvement efforts in the context of upgrading of capacities to govern as a whole and making government more knowledge-intensive and innovative in particular, as examined elsewhere.15

science advice, top-level decision-making, public opinion, impact assessment

1. This is recognized in nearly all the literature dealing with policy analysis as well as cognitive and organizational processes. See, for example, David Dery, Problem Definition in Policy Analysis (Lawrence: University Press of Kansas), 1984 and Harold A. Linstone, Decision Making for Technology Executives: Using Multiple Perspectives to Improve Performance (Boston: Artech House), 1999.

2. A cognitive map is a mental representation of the layout of one's environment. In a broader sense it is a mental image empowering thinking, whether for an individual or for a culture.

3. See, for example, Yaacov Y. Vertzberger, The World in Their Minds: Information Processes, Cognition, and Perception in Foreign Policy Decisionmaking (Stanford: Stanford University Press), 1990.

4. On efforts in this direction, though needing expansion and deepening, see OECD, Reviews of National Policies for Education: Lifelong Learning in Norway (Paris: OECD), 2002.

5. See the discussion of "Empowering the People with Public Affairs Enlightenment", Yehezkel Dror, The Capacity to Govern: A Report to the Club of Rome (London and Portland (OR) USA. Frank Cass), 2001, paperback 2002, ch. 11.

6. Fully applicable is the following statement on coping with ambiguities in texts, as requiring "that the reader understands a great deal already, and that he is able to guess by sympathy (I add "and some relevant literacy") the way the contradiction must be solved". William Empson, Seven Types of Ambiguity, third edition (Norfolk (CO), USA. New Direction books), 1953, p. 191.

7. For broader implications of this rupture in history, see Yehezkel Dror, "From My Perspective: Lucifer Smiles", Technological Forecasting & Social Change, 69 (2002), 431-435.

8. To give just three illustrations out of a large body of theory, empirical studies and lessons of experience that can serve as grounding for indirect criteria for assessing science advice impacts: Daniel Kahnemann, Daniel, Paul Slovic and Amos Tversky, eds., Judgement Under Uncertainty: Heuristics and Biases (Cambridge, UK: Cambridge University Press), 1982, which presents the main studies for which Kahnemann received the 2002 Nobel Prize in Economics, illustrates well-based theories on requirements for coping correctly with uncertainty, which is a crucial issue in science advice. Alexander L. George, Presidential Decisionmaking in Foreign Policy: The Effective Use of Information and Advice (Boulding (CO), USA. Westview), 1990, processes relevant theories and experiences and transforms them into structural features of effective advice. And Aaron Lobel, ed., Presidential Judgment: Foreign Policy Decision Making in the White House (Hollis (NH), USA. Hollis Publishing), 2001, illustrates experiences of senior advisors from which quite a number of indirect effectiveness criteria can be derived. For a detailed theoretical discussion and justification of using preferable models and other indirect criteria for evaluating policy processes, see Yehezkel Dror, Public Policymaking Reexamined (New Brunswick (NJ) USA. Transaction), supplemented edition, 1989, part II.

9. The proposed map is qualitative, but to evaluate reality some quantitative ideas of satisfactory levels for the qualitative dimensions are also needed. This need can be partly met by comparisons with the past and with similar units in other locations and countries. More important in practice, with respect to many of the features of the "preferable characteristics map" reality is so far away from it as to make quantitative standards unnecessary. For improvement purposes quantitative scales are even less essential given the present weaknesses of nearly all high-level science advisory setups, with the need for doing much more to fulfill the requirements of the map serving as a source of practical guidelines for action.

10. "Preferable", in the sense of being significantly better than reality, but without the presumptions of being "optimal", which should be avoided in light of the underdevelopment of reliable and salient theories.

11. The expression was originally used by C. P. Snow in the 1959 Rede lectures (later published as Two Cultures, Cambridge 1964). Snow's basic thesis was that the breakdown of communication between the sciences and the humanities (the "two cultures" of the title) was a major hindrance to solving the world's problems.

12. Very relevant are parts of James A. Dewar, Assumption-Based Planning: A Tool for Reducing Avoidable Surprises (Cambridge, UK: Cambridge University Press), 2002. Put in more literary form, avoiding the regret of "if only we had known", as underlying for instance the writings of Anton Chekhov, is an important science advice contribution which requires much sell-doubt. See Vladimir Kataev, If We Could Know! An Interpretation of Chekov (Chicago: Ian R. Dee), 2002. Translated from Russian.

13. As illustrated by its critical comments on a book dealing with environmental trends. See Andrew C. Revkin, "Book on environment is flawed, Danish ethics panel concludes", International Herald Tribune, January 8, 2002, p. 4.

14. Relevant is Colin S. Gray, Strategy for Chaos: Revolutions in Military Affairs and The Evidence of History (London and Portland (OR), USA. Frank Cass), 2002.

15. See Yehezkel Dror, The Capacity to Govern: A Report to the Club of Rome", op.cit.; and Yehezkel Dror, "Towards Knowledge-Intense and Innovative Government", The IPTS Report, No. 64, May 2002, pp. 28-34.

Professor Yehezkel Dror, The Hebrew University of Jerusalem

E-mail: Contact Form

• Yehezkel Dror is Professor of Political Science at the Hebrew University of Jerusalem and a member of the Club of Rome. He has held senior policy planning positions in the Israeli government, was a senior analyst with the RAND Corporation and worked on European Union problems as a Distinguished Professor at the European Institute of Public Administration in Maastricht. His main research interests include governance, policy planning and statecraft, on which he has been advising many countries and working as a consultant at the OECD, the UN and UNDP. He is the author of many articles and 14 books published in nine languages, the most recent being a Report to the Club of Rome on The Capacity to Govern, published in German, Spanish and Portuguese, with a revised version in English published in 2001 and in paperback in 2002.