JRC guidelines for integrity in its scientific support activities October 2006 CA(06)55 Joint Research Centre: Robust Science for Policy Making A guideline towards integrity and veracity in scientific support and advice
1 JRC guidelines for integrity in its scientific support activities 2
October 2006 JRC guidelines for integrity in its scientific support activities October 2006 Contents 1 Preface 4 2 Context 5 2.1 Science in public policy: roles and actors 5 2.2 JRC’s mission and role 6 2.3 Scientific integrity: a basic element of robust science support 7 3 Principles and guidelines for robust scientific support 8 3.1 The JRC’s work is based on competence 8 3.2 The JRC is impartial in its advice 9 3.3 The JRC’s work is of relevance for the needs of the EU 4 Summary and a way forward 12 14 4.1 Standard Text 14 4.2 Principal guidelines in headlines 14 4.3 A way forward 15 Bibliography 16 Table 1. Differences between Academic and Regulatory Science 18 Table 2. Sources of Uncertainty 19
3 JRC guidelines for integrity in its scientific support activities October 2006 1 Preface Early in 2006 a JRC value statement was adopted following recommendations of the 5 Year Assessment Panel. It refers to respect for scientific integrity which should give the JRC’s support and advice to policy the necessary credibility in the interplay between different levels of governance in the EU and across national, scientific and political cultures. In addition to this, the JRC introduced its new motto “Robust science 1 for policy making” to show its commitment to achieve the best quality in its work for its customers and stakeholders. Both this value statement and the new motto merit further clarification, taking into account the JRC context and the rich literature on the role of science in the formulation of policy and regulatory decisions (cf. bibliography). This role of science is particularly sensitive in areas involving people’s health and safety and the environment, like in decisions concerning genetically modified organisms, nuclear safety, or climate change. The public profile of these issues is high, the underlying science is complex and different scientific experts may defend opposite views. This has implications for policy makers and the JRC must take care to ensure that its policy advice is clear, scientifically robust and objective. Various national administrations have introduced explicit guidelines for policy makers on how to integrate scientific expertise in policy and how to handle policy decisions in the face of uncertainty. The Commission's White paper on Governance and its Communication dedicated to expert advice highlighted the need to respect the principles of accountability, plurality and integrity when collecting and using expert advice 2 . These documents provide guidelines for policy makers on how to seek and use scientific advice. However, they provide little specific guidance for the JRC on how to produce robust scientific support for the benefit of the full policy cycle 3 with activities like risk assessment, foresight, monitoring, compliance checking or crisis response. Against this background, the purpose of this document is to serve as a guideline for the JRC in its efforts to produce robust scientific support and advice and to seek a high standard of integrity in its work. Many of the orientations in this paper follow universal values in science and governance; some are specific for the JRC. 1 Robust (original Latin: robustus oaken, firm, hard, solid) has a specific meaning in the international language of scientists and computer experts, http://wordnet.princeton.edu/: Robust: strong enough to resist or overcome intellectual challenges 2 3 Detailed references to these administrative documents are given in the section Bibliography. Problem definition, agenda setting, policy formulation/decision making, policy implementation and evaluation.
4 JRC guidelines for integrity in its scientific support activities October 2006 2 Context This section describes and elaborates the three elementary issues underlying this guideline towards “Robust science for policy making”: (a) the role of science in public policy, (b) the policy­orientated mission of the JRC, and (c) the coherence between robust science support and scientific integrity. 2.1 Science in public policy: roles and actors It is true that science has become increasingly important in policy and regulatory decisions in recent years; it is also true that in the identification and characterisation of the policy options science considerations have to be taken into account along with many other factors – economics, social policy, strategic objectives. To appreciate how science and scientific expertise enter into the policy decision making process, it helps to distinguish four generic functions in this process with the following actors and roles 4 :
· The decision maker: the person or the body empowered to take the policy decision. Decision­making in the European Union involves the European Commission, the European Parliament and the EU Council. In general the Commission proposes new legislation, but it is the Council and Parliament that take the decisions and pass the laws.
· The policy maker: the person or the body assisting the decision maker in reaching a decision by providing policy analysis and generating policy options, as is for instance done by the Commission staff and/or by preparing the policy proposals for the decision maker like for instance the Commission does when Community policies are concerned.
· The scientific expert: the person or the body providing scientific support and advice to the policy maker and the decision maker, giving the best evidence on which to base the options. This is an active role covering the full policy cycle from problem definition to policy implementation and, where needed, the anticipation of future questions.
· The stakeholders: people or organisations that are affected by a decision and have an interest in its outcome. They exert influence on the policy maker and decision maker and lobby for a preferred outcome of the decision. They can be consulted in the formulation of the questions on which scientific advice is sought and they may also dispose of scientific knowledge relevant for the advice. In practice people or organisations may find that they fulfil more than one of these four different roles in the decision­making process, for example:
· A stakeholder manufacturer of a chemical substance may also play a role as scientific expert in the preparation of scientific advice by contributing valuable specialised scientific knowledge on the health risks of the substance. These roles 4 OXERA 2000, see Bibliography at the end of this document
5 JRC guidelines for integrity in its scientific support activities October 2006 can be played with the necessary discipline: the manufacturer contributes the best science in an expert group and lobbies separately without mixing up the lobby in the scientific advice.
· A scientific expert may chair a committee with delegated authority to determine whether a proposed process is the Best Available Technology required by environmental legislation. Formally the scientific expert does not take the final decision and can decline all responsibility for it, nevertheless because of this delegated authority the expert will be seen as the de facto decision maker. In this situation the mix of two roles is less welcome and to avoid confusion the expert should make sure that the authority is delegated elsewhere, or if this is not possible, decline the chair of the committee. Proper governance demands that every person or body in the decision­making process understands which role it is playing at all times and avoids mixing up different roles. The JRC’s role is primarily the one of scientific expert, but at times it may find itself in any of the four roles in the decision­making process. Hence it is important that the JRC knows which role it is playing in the decision­making process and acts in the appropriate way as a function of this role. 2.2 JRC’s mission and role Policy makers and decision makers need to be able to rely on the scientific advice that they receive and they seek that from a trusted expert, where possible within their own organisation. For the EU Institutions, the JRC has the role of the trusted in­house research­based policy support organisation that can provide advice for the EU policy­ maker, laid down in its mission: … to provide customer­driven scientific and technical support for the conception, development, implementation and monitoring of EU policies. As a service of the European Commission, the JRC functions as a reference centre of science and technology for the Union. Close to the policy­making process, it serves the common interest of the Member States, while being independent of special interests, whether private or national. Under the catchphrase “Robust science for policy making”, the JRC’s mission statement holds the keys to some general rules for its scientific support, which are elaborated in Chapter 3. The mission statement also draws attention to the JRC’s position as a service of the Commission. With its function to provide scientific support and advice to policy making, the JRC positions itself close to what science policy literature 5 describes as mandated or “regulatory science”. Nevertheless, the JRC also needs traditional or “academic science” in its Work Programme for its credibility as a research organisation and partner. Table 1 in the annex spells out the characteristic differences between regulatory and academic science, since they are also of interest for further positioning of the JRC. 5 See Bibliography at the end of this document, e.g. Jasanoff (1990) and Ruggles (2004).
6 JRC guidelines for integrity in its scientific support activities October 2006 A focus on regulatory science means a focus on typical regulatory science activities like: monitoring, evaluation, screening and meta­analysis. Regulatory science products also differ from those in academic science: the latter places greater value on peer­ reviewed published papers, whereas results from “regulatory” science and research may never be submitted to the discipline of peer review and publication. Within a multidisciplinary research programme with characteristics of regulatory science, the JRC combines elements of academic research organisations, corporate and contract research organisations and even of specialised regulatory agencies, but remains different from all of them. Therefore the JRC does not have the option to model its values and its attitude to customers on that of typical national or international laboratories; it has to establish its own values. 2.3 Scientific integrity: a basic element of robust science support The marriage between science and policy presents society with the challenge to reconcile the scientific uncertainty and the pressure of decision making. This challenge is addressed in a social and institutional context, in which the decision maker, the policy maker, the scientific expert and the stakeholder are expected to operate with integrity. For the scientific experts this translates into an expectation that they respect scientific integrity, which is probably best described as the quality to produce science worthy of trust 6 . Such science demands an attitude of veracity, observant of truth. Hence scientific integrity is an important value for any research organisation and a necessary (but not sufficient) quality for the production of robust science. Trust is hard to win and easy to lose. Numerous examples have shown how easily a single scientist or research group can discredit the work of an entire institute or even a whole research community 7 . Science relies on free and open exchange of theory and experiment on which successive workers build. If they cannot trust the results that they receive, science itself is suffering and loses fragile public trust. As regards scientific integrity for the JRC, the concept must be applied to the full process of giving science and technology support to policies. The integrity and robustness of its scientific products must be assured, so that they can be proved to have these values. Reference guidelines to achieve such robust scientific support, and governance that assures that these guidelines are followed, are therefore essential for the JRC as a whole, meaning both for its customers and for its own existence. 6 The New Oxford Dictionary defines “integrity” as “the quality of being honest and having strong moral principles”. Scientific integrity is therefore the quality that makes the science worthy of trust. 7 Recent examples include http://news.bbc.co.uk/1/hi/world/asia­pacific/4554422.stm the cloning research by Dr Hwang Woo­suk, and Professor Chen Jin of Shanghai Jiaotong University who was dismissed as dean of the university's School of Microelectronics for faking the Hanxin DSP chip, http://washingtontimes.com/business/20060512­101756­8035r.htm
7 JRC guidelines for integrity in its scientific support activities October 2006 3 Principles and guidelines for robust scientific support The mission statement of the JRC holds three keys for the principles of JRC’s robust scientific support: ­ … reference centre of science and technology for the Union => The JRC’s work is based on competence ­ … independent of special interests, whether private or national => The JRC is impartial in its advice ­ … customer­driven … close to the policy­making process => The JRC’s work is of relevance for the needs of the EU The following subsections expand these principles into some more general and some more concrete guidelines in a non­exhaustive way. In most cases these guidelines serve more than just the principle to which they are linked; they all serve the general goal of providing robust support that respects scientific integrity. 3.1 The JRC’s work is based on competence The JRC strives for excellence in all of its scientific work by seeking to employ the best scientists and support staff. Its competence is also demonstrated in the fact that it conducts work for third parties, provided that this does not, and is not perceived to, create a conflict of interest. The JRC applies rigorous quality assurance on its work. It meets the demands of its policy customers either from its own resources or, where JRC’s in­house know­how is not sufficient, by hiring external experts or subcontracting certain elements of the work. 3.1.1 Pursuing scientific excellence The JRC strives for excellence in fields in which it might be called to offer scientific support and advice. This excellence is needed to be able to attract the best workers, motivate the staff and achieve respect and credibility in the eyes of peers. It requires both competent staff and appropriate facilities, together with adequate reliable funding. The need to deliver sustainable high­quality results has induced the JRC to pursue a quality management approach, where necessary backed up by external certification and accreditation (ISO 9001, ISO/IEC 17025, ISO Guide 34). This rigorous quality assurance should further guarantee an excellent performance of the JRC in knowledge production, knowledge synthesis and prediction, which are the three basic components of regulatory science. The knowledge­synthesis activities, in particular in JRC bureaus and legally required data base management, rely on external knowledge and the collating of work of others. These activities enable the JRC to operate with excellence and authority in areas where it is not necessarily at the forefront of research.
8 JRC guidelines for integrity in its scientific support activities October 2006 The JRC can also mobilise external excellence to help generating its scientific support and advice in the relevant areas through extensive networks: it has more than 1000 partners in numerous networks and cooperates with more than 1500 partners in Indirect Research Actions of the Framework Programme. 3.1.2 Ensuring that all scientific advice is traceable Robust scientific policy support and advice have a clear audit trail of all evidence, reasoning, assumptions and procedures so that, if the decision based on the advice is subsequently challenged, the policy maker and decision maker may show how they reached their conclusions. This is not unique for the JRC or to regulatory science; the basic requirement of any scientific process is that it can be repeated and validated at any time on the basis of rigorous information on its production history. A useful reference tool for elaborating an audit trail in regulatory science issues has been published 8 by the Dutch National Institute for Public Health and the Environment, RIVM. 3.1.3 Focusing on tasks within scope and competence of the JRC The JRC’s institutional Work Programme is driven by its customers’ needs and funded from the Framework Programme for Research. The JRC also provides on­demand support and solutions at short notice via specific contracts. The JRC has a responsibility to ensure robust science and may not agree with the customers’ pre­determined wishes regarding the outcome of the analysis. The JRC always has the choice to decline work that it considers unacceptable, either because the work does not lie within its scope and competence or because there it sees an ethical or commercial conflict. When the JRC works for third parties, care is taken that the work does not conflict with its primary function to offer authoritative and manifestly unbiased scientific support and advice for the development of EU policies. As a general rule, military issues are outside the scope of the European Research Policy, its Framework Programme and the Work Programme of the JRC and consequently JRC scientists are not supposed to engage in tasks with a military connotation. 3.2 The JRC is impartial in its advice The JRC strives to be impartial in its advice. Its quality assurance procedures promote objectivity and are designed to prevent unconscious assumptions or well­intentioned bias distorting its advice, while documenting uncertainties in a transparent way and communicating the basis of choices and assumptions made in the development of the advice and their influence on the results. It avoids potential conflicts of interest with its policy customers. 8 Van der Sluijs et al. 2003. RIVM introduced a more conscious dealing with uncertainty following extensive public debate triggered by an employee who publicly criticised unjustified levels of certainty reported in RIVM studies for the Dutch government.
9 JRC guidelines for integrity in its scientific support activities October 2006 As part of the Commission the JRC knows that 9 “the precautionary principle is essentially there to be used by decision­makers in the management of risk and should not be confused with the element of caution that scientists apply in their assessment of scientific data”. 3.2.1 Delivering science free from bias Biases may slip into science because of the human factor: from cognitive bias (e.g. from mental errors or shortcomings usually caused by simplified and subconscious human information processing strategies), via disciplinary bias (disciplines tend to develop their own characteristic way of viewing problems and selecting objects of inquiries), to motivational bias (when people have incentives to reach a certain conclusion or see things a certain way) 10 . Fraud is an extreme and most negative case of motivational bias and obviously a serious scientist would never consider introducing an intentional dishonesty in his or her results. Usually there is also no difference of opinion about the rule that conflicts of interest must be avoided, because directly or indirectly they can trigger motivational bias. However, the question whether a scientist may deliberately introduce bias by erring on the side of caution needs a more sophisticated answer than a simple “no”. Whereas it is relatively clear that filtering measurements and selectively picking data for reasons of well­intended caution is wrong practice in the exact sciences, in socio­ economic modelling or prospective analyses, for example, a scientist may prefer to select certain data or measurements to build in caution. Scientific results thus obtained are valid, but must be reported identifying the assumptions as an integral part of the outcome of the research. Robust science keeps a continuous guard against all hidden biases; the scientist has to ensure that the best science is presented and that all of the caveats, uncertainties and possible built­in caution are properly reported. 3.2.2 Promoting objectivity Objectivity is a powerful resource in governance, since it allows policy makers and decision makers to claim that they act for the benefit of all without serving any particular interest. Objective experts avoid subjective preferences in their advice, and show that they have considered all current scientific views in a balanced way, i.e. it does not give equal weight to views that are held by 5% or by respectively 95% of the scientists. 9 Quote from the Commission’s Communication on the Precautionary Principle, COM(2000)1. The precautionary principle is most authentically described in the 1992 Rio Declaration on Environment and Development as follows: “Where there are threats of serious or irreversible environmental damage, lack of full scientific certainty shall not be used as a reason for postponing cost­effective measures to prevent environmental degradation”. 10 Bias is a constant or systematic deviation as opposed to random error. An extensive description of the various kinds of bias is incorporated in the glossary of Van der Sluijs et al. (2003).
10 JRC guidelines for integrity in its scientific support activities October 2006 The objectivity requirement and the associated approach seem obvious, but it may become difficult when the JRC has scientific evidence that goes against the policy being pursued by its customer. The JRC’s duty is clear – it is to present this evidence to the policy maker. It may also propose publication as long as it is not confidential material and presuming that no policy harm is done by such publication. Policy makers may exercise pressure on the JRC to offer greater certainty than the science can justify; or to interpret uncertain results in such a way as to support a preferred policy option. In respect of veracity the JRC is alert to this possibility and can react in line with ethical and integrity principles, discussed in part 3.3.3. The JRC has a special challenge when it acts on behalf of the policy maker, for example when it provides the secretariat for committees of independent expert advisers. For such tasks the JRC must be faithful to its role as an objective actor and extremely careful not to influence (or to give the impression of influencing) the outcome of committee meetings by the way in which it ­ possibly unconsciously ­ convenes, conducts and, most important, records the outcome of the meetings. 3.2.3 Documenting uncertainty Science contributions to policy making are usually full of uncertainties. Whereas academic scientists may have the luxury of suspending judgement because of soft components in the results, politics cannot wait until all uncertainties in scientific input have been resolved 11 . Therefore, regulatory scientists need to pay particular attention to the different ways in which uncertainty is present in their supporting scientific assessment. The Dutch Institute for Public Health and the Environment RIVM produced several useful reference documents with guidelines for uncertainty assessment 12 and with a classification of sources of uncertainty, applicable to all contexts 13 . At the highest level there are two major sources of uncertainty: “variability” of the system/process under consideration and “limited knowledge” of the analyst or of mankind in general. Instructive details of these sources of uncertainty are given in Table 2 in annex. Scientific experts should insist, and policy makers should demand, that the uncertainty is considered, documented and stated explicitly and they must resist pressure to stating greater certainty than they can defend. Policy makers need to understand the nature, location and level of uncertainty in the scientific advice that they receive. Transparency about uncertainty is to be considered an integral part of scientific integrity (“intellectual rigour”), also because uncertainty is often exploited in a politicized way, as part of strategic interest­defending moves, stirring ever more controversy (e.g. tendencies towards “manufactured uncertainty” and “consciously created confusion”). 11 See also the table with differences between regulatory and academic science at the end of this document. 12 J.P. van der Sluijs et al. (2003) 13 M. van Asselt et al. (2001)
11 JRC guidelines for integrity in its scientific support activities October 2006 The habit of documenting the uncertainty in the various pieces of science is an adequate way to promote transparency and providing an audit trail at the same time. 3.3 The JRC’s work is of relevance for the needs of the EU The JRC through its unique relationship with the Commission is in a position to meet immediate needs for scientific support in the EU institutions and to anticipate future EU policy questions. Its activities extend from “horizon scanning” to identify prospective science that might affect EU policies through to checking compliance with science­ based policies and providing support with putting them into action. It will enforce its own ethical guidelines, in consultation with its customers that will determine the areas of policy within which JRC is prepared to operate. 3.3.1 Serving the EU with robust scientific support It is the JRC’s mission to support EU policies that are conceived and implemented by the EU institutions and its institutional customer base consists of the Commission with its Directorates General, the European Parliament, and the Council of the European Union. In addition to this direct support, the JRC also has strong links with the regulatory and advisory bodies of the EU. In some cases pre­operational regulatory and advisory support of the JRC led or is leading to the creation of new agencies (ECB, EFSA) and in general the JRC is developing relations with a growing number of regulatory agencies of the EU. It provides support to their functioning directly or more often at the request of and in close consultation with the Commission service responsible for the agency. Playing this sometimes invisible but therefore not less important role of in­house science expert, the JRC is aware that the EU could have to suffer in credibility and effectiveness if it were to provide its institutional customers with weak, contestable scientific input. As part of the EU institutions the JRC therefore promises to deliver “Robust science for policy making” from a competent and impartial organisation that keeps an audit trail of its advice which is unbiased, objective and detailed in its uncertainties. 3.3.2 Linking with policy and scanning emerging issues The JRC meets the immediate needs and anticipates future policy questions in the EU, as part of the Commission and in relation with the other EU institutions. Its activities link to the full policy cycle and are subject to the subsidiarity principle, meaning that JRC interventions need the European dimension and are not better taken care of by (a) national institute(s). These activities range from pure research, via foresight, risk assessment, implementation support and monitoring through to compliance checking and crisis response. Special attention is given to cases of crisis response, since the time for reaction is often too short to follow a proper scientific process of review and reflection. When policy makers and decision makers are under time pressure, the science may have an important and even determinative impact on the policy, which makes it even more important that
12 JRC guidelines for integrity in its scientific support activities October 2006 the JRC in these cases does the utmost to maintain the quality and guarantee the integrity of its scientific support and advice. Anticipation and preparation for possible crises is the best way to produce reliable advice in time. Therefore, the JRC is actively scanning science and technology that emerge, and social reactions to them, firstly to ensure that it has the necessary scientific understanding in place when it is called upon for policy support and secondly to warn the policy maker of prospective policy concerns before they turn into crises. 3.3.3 Following EU ethical and integrity principles The JRC carries out its work in respect of its own Ethics at the Workplace 14 , in respect of the Baseline requirements of the Commission’s Internal Control Standards regarding Ethics and Integrity, in respect of the ethical rules and principles of the countries in which it operates and in respect of the personal ethical standards of its own workers. For actions funded under the Framework Programme the JRC is also held to its obligations laid down in the JRC Specific Programmes, which declares the respect for fundamental ethical principles, including amongst others the principles reflected in the charter of Fundamental Rights of the EU and when participating in indirect actions to the same requirements applied to any other participant in such actions. In addition, Council directives and Protocols annexed to the Treaty take care that full regard is paid to the welfare requirements of animals in Community research. In principle all these administrative guidelines also hold the keys to a proper reaction in case that the JRC, through one of its scientists, receives dubious demands from its customers, e.g. a request:
· to abstain from commenting incomplete and/or incorrect scientific claims in policy statements;
· to force the interpretation of uncertain results in favour of a preferred policy option;
· to hold back results that are unwelcome to the policy maker, i.e. focussed on the JRC not to publish results that might undermine a Commission policy. It requires good judgment to recognise when such demands are in conflict with standards for scientific integrity. In any event, the way to make known facts and conclusions that are unwelcome to the customer (also when this is the mother organization) is not different from the usual way, i.e. through the hierarchy. Existing Commission rules 15 foresee in cases when scientists feels that their concern is not dealt with in a satisfactory manner by their direct superior and these rules describe how these rare but not imaginary situations can be reported without damage for the one who reports possible wrongdoing. 14 http://www.cc.cec/home/dgserv/jrc/general/ethics/ethics_en.html 15 Commission Internal Control Standards, Baseline requierement 14 on Reporting Improperties.
13 JRC guidelines for integrity in its scientific support activities October 2006 4 Summary and a way forward The essence of this guideline is to ensure scientific robustness, integrity and veracity that is reflected in a standard text (see 4.1) that might, for example, serve as explanatory statement for JRC’s motto “Robust science for policy making” or as an essential element in JRC’s value statement, as described below, or in headlines relating to the section headings in Chapter 3 (see 4.2). 4.1 Standard Text The JRC seeks to assure the robustness and the scientific integrity of its work by its competence, impartiality and EU policy relevance. For this purpose it employs competent scientists and support staff and draws on external experts, subjecting all of their work to rigorous quality assurance. It seeks to avoid any actual or perceived conflicts of interest and ensures that the scientific support and advice that it offers to EU institutions is free from bias, including unconscious or well­intentioned bias. Respecting fundamental ethical principles, together with those reflected in the Charter of Fundamental Rights of the European Union, it seeks to meet all of the present and prospective needs of its customers, ensuring that it offers advice without usurping the proper process for making policy decisions. 4.2 Principal guidelines in headlines
· JRC’s work is based on competence o Pursuing scientific excellence o Ensuring that all scientific advice is traceable o Focusing on tasks within scope and competence of the JRC
· The JRC is impartial in its advice o Delivering science free from bias o Promoting objectivity o Documenting uncertainty
· The JRC’s work is of relevance for the needs of the EU o Serving the EU with robust scientific support o Linking with policy and scanning emerging issues o Following EU ethical and integrity principles
14 JRC guidelines for integrity in its scientific support activities October 2006 4.3 A way forward The new motto “Robust science for policy making” underlines the JRC’s wish to promote the quality and the trustworthiness of its support. Common sense and good practise will implement the motto. It should reinforce the attitude and activities of scientific staff and management with respect to the integrity in their work and the robustness of their results. The standard text and the headlines may be used to recall the important issues for the JRC’s staff and its management. They also clarify the JRC’s basic principles as regards its scientific integrity for its Board of Governors as well as for its customers and stakeholders. However, the challenge is not to develop principles, practices, standards and values for scientific integrity and robust support to be hidden away somewhere in a JRC manual, but to ensure that the JRC staff and scientists know them and see them as their own. There is no simple way to enforce the ideas or to make sure that people live up to the good intentions. The robustness and the integrity of the whole scientific support process is a shared responsibility of JRC’s management and staff and the many useful contributions in the drafting phase of this paper showed that the issues are of concern to scientists and management. As a first step towards common ownership this guideline has been discussed with the Scientific Committee of each Institute in a series of internal seminars which indicated high interest and strong involvement. Examples of issues worth further exploring are:
· The production of a list of possible examples in which scientists and/or the organisation faced a situation that was not dealt with in accordance with the present guideline on scientific integrity and solutions to such cases.
· Integrity breach through forgery or fraud is certainly an offence under Commission staff rules, but do the Internal Rules and Procedures of the Commission in the JRC cover basic principles of integrity in scientific research? Are there procedures, practices or issues that merit being addressed separately and specifically for the JRC?
· The production of knowledge for concrete real­world problems is not the same as academic science. The JRC’s focus on regulatory science in support of EU policies can be used for a clearer positioning of its work and organisation. In addition to any comments from the Board of Governors it will be useful to solicit reactions on scientific integrity issue from JRC’s major stakeholders in an upcoming meeting with the High Level User Group.
15 JRC guidelines for integrity in its scientific support activities October 2006 Bibliography Literature S. Jasanoff (1990): “The Fifth Branch: Science Advisers as policymakers”, Harvard University Press, ISBN 0­674­30062­9 and many references therein A. Stirling, (ed), (1999), On Science and Precaution in the Management of Technological Risk: Volume II case studies, European Commission, JRC­IPTS, Seville, EUR 19056/EN/2; this report draws on studies conducted by O. Renn and A. Klinke, A. Salo and A. Rip, as well as A. Stirling OXERA 2000, Policy, risk and science: Securing and using scientific advice, p ii, http://www.hse.gov.uk/research/crr_pdf/2000/crr00295.pdf,(Oxford Economic Research Associates Ltd), ISBN 0­7176­1852­8 S, Funtowicz et al. (2000), “Science and governance in the European Union: a contribution to the debate”, Science and Public Policy, 23(5), October, pages 327–336 M.B.A. van Asselt et al. (2001) Uncertainty and RIVM’s Environmental Outlooks: Do­ cumenting a learning process, http://www.rivm.nl/bibliotheek/rapporten/550002001.pdf J.P. van der Sluijs et al. (2003); “Guidance for Uncertainty Assessment and Commun­ ication” http://www.nusap.net/downloads/detailedguidance.pdf ISBN 90­393­3536­2 A. Ruggles (2004): “Regulatory vs. Academic Science”, Essay with an illustrative table of differences between the two branches of science at the following website, http://sciencepolicy.colorado.edu/ogmius/archives/issue_9/research_highlight.html. S. Jasanoff (2005): “Designs on Nature, Science and democracy in Europe and the United States”, Princeton University Press, ISBN 0­691­11811­6 L. Levidow et al. (2005), EU Regulation and agri­biotechnology: Precautionary links between science and policy, Science & Public Policy 32(4) pages 261­76, http://technology.open.ac.uk/cts/peg/sppaug2005eu%20fin.pdf Administrative reports and documents “The Politics of GM Food”, Publication of the UK Economic and Social Research Council ESRC Global Environmental Change Programme, Risk, Science & Public Trust, (1999) European Commission: Communication on the Precautionary Principle, COM(2000)1 White paper on European Governance, COM(2001)428
16 JRC guidelines for integrity in its scientific support activities October 2006 “Democratising Expertise and Establishing Scientific Reference Systems”, Report of Working Group 1b, Commission White Paper on Governance, Work Area 1: Broadening and enriching the public debate on European matters, R. Gerold, A. Liberatore, May 2001 “DTI ­ Guidelines 2000: Scientific Advice and Policy Making”, URN No: 00/1026, UK Office of Science & Technology with Foreword by the former Chief Scientific Adviser Sir Robert May, http://www.dti.gov.uk/science/page15432.html “Code of Practice for Scientific Advisory Committees”, Department of Trade and Industry, December 2001 “Scientific Advice and Policy Making”, Department of Trade and Industry, Report by the Chief Scientific Adviser Professor David King, December 2001 On the collection and use of expertise by the Commission: Principles and Guidelines, COM(2002)713 “Risk: Improving government’s capability to handle risk and uncertainty”, UK Government Strategy Unit November 2002 “Typifying Scientific Advisory Structures and Scientific Advice Production Methodologies”, Final Report prepared for DG RTD, PREST, University of Manchester, December 2003 “Science in trade disputes related to potential risks”, case studies, IPTS, JRC, August 2004 OMB Final Information Quality Bulletin for Peer Review, US Office of Management and Budget, December 15, 2004 “Policy through dialogue: informing policies based on science and technology”, UK Council for Science and Technology, March 2005 “Guidelines on Scientific Analysis in Policy Making”, HM Government, UK, October 2005 HSE Statement for the implementation of the Chief Scientific Adviser’s Guidelines 2005, UK Health and Safety Executive
17 JRC guidelines for integrity in its scientific support activities October 2006 Table 1. Differences between Academic and Regulatory Science [A. Ruggles (2004)] INSTITUTIONS GOALS REGULATORY SCIENCE ACADEMIC SCIENCE Government/industry Universities Information needed to meet regulatory requirements and to provide reliable information for decision makers. Original research framed by scientists and driven by rational analysis and expert judgment. Research “questions” are framed by legislators and regulators and have immediate social and economic implications. Ultimate goal is conflict resolution via public debate over competing interests and values. To expand the understanding and knowledge of the natural world through an ongoing process of questioning, hypothesizing, validation, and refutation. Predictive certainty is required by the political process and the legal system. ROLE OF UNCERTAINTY Knowledge is frequently and necessarily generalized to situations very different from those in which the original data was collected. Uncertainty is expected and “embraced”. Uncertainty is unwelcome by the public, legislators and the courts. INFORMATION Must frequently act before all the necessary information is developed. Publish when a body of information has been developed, tested, and validated. STATISTICAL SIGNIFICANCE/ ACCEPTABLE ERROR/BURDEN Often work with a legal mandate to minimize Type II error with the result that Type I error is increased. Strive to minimize Type I error. Regulatory scientists are required to consider and work with the values of many including the public, politicians, the scientific community, and the regulatory community. Academic scientists work primarily with their own and their collaborators’ values; seldom have to incorporate public or political values. “Gray literature,” baseline data, monitoring data, regulatory documents. Published, peer­reviewed papers and books, presentations at professional meetings. COMPLETENESS OF OF PROOF ROLE OF VALUES PRODUCTS TIME­FRAME Determined and driven by statute, regulation, and the political process; finite and often quite short (90 days to 2­ 4 years). Resolution of problems being reacted to is often crisis­ driven or driven by court­mandated timelines. POLITICAL INFLUENCE ACCOUNTABILITY INCENTIVES Open­ended; usually carried out relatively free of an urgent need for the information generated. Directly influenced by politics – upper­level administrators are appointed by the President; funding is at the will of Congress; ultimate oversight is by the courts. Indirectly influenced by the researcher’s own political philosophy and by their perception of the preferences of grant and tenure review committees. Legislatures, courts, and the public. Professional peers. Compliance with legal requirements, working for the public good. Professional recognition, advancement in tenure system; university administration.
18 JRC guidelines for integrity in its scientific support activities October 2006 Table 2. Sources of uncertainty Uncertainty is a not a clear concept: it has a meaning, but no content. To understand the concept better M. van Asselt et al. (2001) made a classification of sources of uncertainty with two main categories of “variability” and “limited knowledge”. Variability
· Variability is a property of the system/process under consideration that can behave in different ways or is valued differently. Variability is an attribute of reality. Different sources of variability can be distinguished, i.e.: o Inherent randomness of nature: the non­linear, chaotic and unpredictable nature of natural processes o Value diversity: differences in people's mental maps, world views and norms and values due to which problem perceptions and definitions differ o Human behaviour: “non­rational” behaviour, discrepancies between what people say and what they actually do, or deviations of “standard” behavioural patterns (micro­level behaviour) o Social, economic and cultural dynamics (societal randomness): the non­linear, chaotic and unpredictable nature of societal processes (macro­level behaviour) o Technological surprises: unexpected developments or breakthroughs in technology or unexpected consequences of technologies. Variability as defined by these sources goes beyond established seasonal variations. Due to variability in combination with limited resources to measure and obtain empirical information, reality inhibits inherent uncertainty and unpredictability. As such variability contributes to limited knowledge. Limited knowledge
· Limited knowledge is a property of the analysts performing the study and/or of our state of knowledge. Limited knowledge results partly out of variability, but knowledge with regard to deterministic processes can also be incomplete and uncertain. A continuum can be described that ranges from inexactness to irreducible ignorance: o Inexactness also referred to as lack of precision, inaccuracy, metrical uncertainty, measurement errors, or precise uncertainties. “We roughly know”. o Lack of observations/measurements: lacking data that could have been collected, but haven’t been. “We could have known”. o Practically immeasurable: lacking data that in principle can be measured, but not in practice (too expensive, too lengthy, not feasible experiments). “We know what we do not know”. o Conflicting evidence: different data sets/observations are available, but allow room for competing interpretations. “We don’t know what we know”. o Reducible ignorance: processes that we do not observe, or theoretically imagine at this point in time, but probably in the future. “We don’t know what we do not know”. o Indeterminacy: processes of which we understand the principles and laws, but which can never be fully predicted or determined. “We will never know”. o Irreducible ignorance: there may be processes and interactions between processes that cannot, or not unambiguously, be determined by human capacities and capabilities. “We cannot know”.
19