<<

Scientific Advice Mechanism (SAM)

Scientific Advice to European Policy in a Complex World

Independent Group of Chief Scientific Advisors Expert Scientific Opinion No.7, Sep. 2019 Report

Research and Scientific Advice to European Policy in a Complex World

Group of Chief Scientific Advisors

European Commission Directorate-General for Research and Innovation Unit 03 — Chief Scientific Advisors – SAM, EGE

Email EC-SAM@ec..eu [email protected]

European Commission B-1049

Manuscript completed in September 2019.

The contents of this report are the sole responsibility of the European Commission’s Group of Chief Scientific Advisors. Although staff of the Commission services participated in the preparation of the report and provided information and assistance in assembling it, the views expressed in this report reflect the collective opinion of the Members of the Group of Chief Scientific Advisors, and may not in any circumstances be regarded as stating an official position of the European Commission.

More information on the is available on the internet (http://europa.eu).

Luxembourg: Publications Office of the European Union, 2019

Print ISBN 978-92-76-12558-7 doi:10.2777/68120 KI-02-19-872-EN-C PDF ISBN 978-92-76-12557-0 doi:10.2777/80320 KI-02-19-872-EN-N

© European Union, 2019 Reuse is authorised provided the source is acknowledged. The reuse policy of European Commission documents is regulated by Decision 2011/833/EU (OJ L 330, 14.12.2011, p. 39).

For any use or reproduction of photos or other material that is not under the copyright of the European Union, permission must be sought directly from the copyright holders. Cover image: © leavector, #248160537, 2019. Source: stock.adobe.com EUROPEAN COMMISSION

Chief Scientific Advisors – SAM, EGE INDEPENDENT SCIENTIFIC ADVICE FOR POLICY MAKING

Scientific Advice to European Policy in a Complex World

Group of Chief Scientific Advisors Scientific Opinion No.7, Sep. 2019 (Supported by SAPEA Evidence Review Report No. 6)

Brussels, 26 September 2019

2019 Directorate-General for Research and Innovation

Scientific Opinion Scientific Advice to European Policy in a Complex World

Table of Contents

ACKNOWLEDGEMENTS ...... 5

Executive Summary ...... 6

1. Introduction ...... 13 1.1. Why this opinion and for whom ...... 13 1.2. The approach ...... 14

2. Recommendations ...... 17 2.1. Engage early and regularly ...... 17 2.1.1. Clarify boundaries between science, scientific advice, and politics ...... 17 2.1.2. Define together the questions to be addressed ...... 19 2.2. Ensure the quality of the scientific evidence ...... 28 2.2.1. Use the full scope of good science ...... 28 2.2.2. Ensure rigorous evidence synthesis ...... 30 2.2.3. Ensure rigour in expert consultation ...... 34 2.2.4. Refine the approach to conflicts of interest...... 36 2.2.5. Codify good scientific advice and consider oversight of its implementation .... 41 2.3. Analyse, assess and communicate uncertainties ...... 42 2.3.1. Use the most suitable uncertainty analysis approaches ...... 42 2.3.2. Communicate uncertainties and diverging scientific views ...... 49 2.3.3. Explain the path from evidence to the advice ...... 52

3. References ...... 56

List of Annexes

ANNEX 1 – EC RULES ON INTEREST DECLARATION AND CONFLICTS OF INTEREST ...... 63 ANNEX 2 – CODIFIED PRINCIPLES AND GUIDELINES FOR SCIENCE ADVICE ...... 66 ANNEX 3 – EC PRINCIPLES AND GUIDELINES ON THE USE OF EXPERTISE ...... 68 ANNEX 4 – SCIENCE FOR POLICY LANDSCAPE IN THE EC ...... 69 ANNEX 5 – EXAMPLES OF EVIDENCE SYNTHESIS METHODS ...... 70 ANNEX 6 – MAIN ELEMENTS OF EFSA UNCERTAINTY ASSESSMENT GUIDANCE ...... 71 ANNEX 7 – AN UNCERTAINTY MATRIX ...... 72 ANNEX 8 – NUSAP METHOD OF UNCERTAINTY ASSESSMENT ...... 73 ANNEX 9 – EXPERTS CONSULTED ...... 75 ANNEX 10 – LIST OF ABBREVIATIONS ...... 76 ANNEX 11 – GLOSSARY ...... 77

List of Boxes

BOX 1. CASE STUDY: THE BADGER CULLING CONTROVERSY (ADAPTED FROM MONTUSCHI, 2017)...... 21 BOX 2. STACEY DIAGRAM: A TOOL TO FACILITATE THE DEFINITION OF QUESTIONS FOR SCIENTIFIC ADVICE FOR, WITH AN APPLICATION EXAMPLE ...... 23 BOX 3. A MODEL FOR INTEGRATING PUBLIC PARTICIPATION IN SCIENTIFIC ADVICE ...... 27 BOX 4. CASE STUDY: A RESPONSE TO A PLANT DISEASE EPIDEMIC, AND PUBLIC CONTESTATION ...... 45 BOX 5. AN EXAMPLE OF A KNOWLEDGE-ASSESSMENT TOOL ...... 47 BOX 6. MISCOMMUNICATION DUE TO AMBIGUOUS EXPRESSION OF UNCERTAINTY ...... 49 BOX 7. APPROXIMATE PROBABILITY SCALE RECOMMENDED FOR HARMONISED USE IN EFSA ...... 50 BOX 8. EXAMPLES OF COGNITIVE HEURISTICS AND BIASES IN UNCERTAINTY COMMUNICATION ...... 51

EC Group of Chief Scientific Advisors September 2019 3 Scientific Opinion Scientific Advice to European Policy in a Complex World

Group of Chief Scientific Advisors

Janusz Bujnicki

Pearl Dykstra Deputy Chair

Elvira Fortunato

Nicole Grobert

Rolf-Dieter Heuer Chair

Carina Keskitalo

Paul Nurse

4 EC Group of Chief Scientific Advisors Scientific Opinion Scientific Advice to European Policy in a Complex World

ACKNOWLEDGEMENTS

‘ ’ ‘ ’

 (Making Sense of Science for Policy under Conditions of Complexity and Uncertainty

 ‘ – ’ Kwieciń

 –

 Vladimir Šucha, the Director David Mair (Head of Unit ‘Knowledge for Policy: Concepts and Methods’, JRC) Understanding Our Political Nature

1

EC Group of Chief Scientific Advisors 5 Scientific Opinion Scientific Advice to European Policy in a Complex World

Executive Summary

Why should I care about the topic?

his opinion should e of interest to those who want to now (more) aout:

 what the architects of the C scientific advice system (the College) can do concretely to set in motion further improvements of the system  how scientific advisors and policy maers can est wor to define a clear mandate and uestions for scientific advice  when and how to include staeholders and the pulic in science advice  how to achieve clarity aout values and interests which may e affecting scientific advice, and how to manage them  how to optimise C rules on conflicts of interest to ensure the impartiality of eperts without needlessly losing valuale epertise  how to oost the reliaility and usefulness of epert consultation y reducing ias and aritrariness  how to ensure that the evidence that is collected and used for policy is of high uality  how to deal with different uncertainties in scientific evidence and advice  how to communicate scientific advice effectively, including uncertainties, gaps and divergent scientific views

his opinion recommends ways to generate high uality scientific advice for uropean policy n setting that goal, we ase our wor on the following fundamental principles, informed y evidence and eperience:  Highuality science is the edroc of good scientific advice scientific advice needs to employ effective methods for analysis of the scientific evidence  cientific advisors are intermediaries etween science and policy hence they need to demonstrate their trustworthiness in the eyes of policymaers, the pulic, the scientific community and all others involved in the process, as a prereuisite for doing their wor well  cientific advice needs to e a transparent and impartial process, and to have a clear mandate to ensure that science is separate from politics

o achieve that goal we have adopted the following recommendations:

6 eptemer EC Group of Chief Scientific Advisors Scientific Opinion Scientific Advice to European Policy in a Complex World

Recommendation 1: Engage early and regularly Clarify oundaries etween science, scientific advice, and politics Values, eliefs, opinions and competing interests are integral to society and politics o e trusted and credile, scientific evidence and its analysis must, as much as possile, e clearly differentiated from such factors when advising on policy decisions

cience (and scientific advice) themselves are not completely valuefree ut scientific advice must not e driven y partisan interests and stealth issue advocacy

he oundaries etween science, scientific advice, and policy – and thus the mandate for scientific advice must e clear However, they are often not clearly predetermined n fact they can e contested – y policymaers, the pulic or scientists: eg it may e suect to some deate whether scientific advice should offer policy options he oundaries need to e agreed and set in dialogue when the uestions for scientific advice are defined together with policymaers, and when the decision is made on how to involve the pulic and staeholders in scientific advice are made

Define together the uestions for scientific advice cientists and policy maers should define the uestions for policy advice together, involving staeholders and the pulic, to ensure roust and highuality science advice s evidence emerges, these uestions often need to e refined in an iterative approach

hen defining together the uestions for science on comple policy issues, using simple conceptual aids can mae a difference

ntegrating foresight and horion scanning into scientific advice is important for comple policy issues hese methods can help capture early warnings: lessons from the past include eg the cases of oonedepleting gases and oresight can help the timeliness of the scientific advice on comple emerging issues

hen policy issues informed y science are socially controversial, scientific advice should involve staeholders nvolving memers of the pulic that are directly affected should also e considered o wor well and avoid ecoming an unclear miture of science and politics, staeholder and pulic involvement in scientific advice must have a clear purpose, follow clear principles, and comine delieration with analytical rigour model developed y the U ational Research Council is an eample of good practice

Recommendation 2: Ensure the quality of the scientific evidence Use the full scope of good science ll good uality science that can contriute to the issue at hand should e considered his includes natural sciences, engineering, medicine, social sciences and humanities

EC Group of Chief Scientific Advisors eptemer 7 Scientific Opinion Scientific Advice to European Policy in a Complex World

They have diverse applications (e.g. a ‘scoping review’ is relatively quick

8 EC Group of Chief Scientific Advisors Scientific Opinion Scientific Advice to European Policy in a Complex World clarity o criteria used to eclude eperts due to a with a argin or casey case assessent ( consistency in the pulic disclosure o interests including coparale scrutiny o oth ipartial eperts and interest representatives ( etter understanding o ias ased on the latest science.

odiy good scientiic advice and consider oversight o its ipleentation eveloping a single set o principles and good practices coon to all scientiic advice odies in the uropean oission is recoended. This could e achieved e.g. through an ode o ractice or cientiic dvice accopanied y a living toolo o good practices and ethods designed or collaorative use with policyakers and the pulic. The toolo could uild on the approaches ethods and tools highlighted throughout this pinion while adapting the to use and continually updating the.

rchitects o the scientiic advisory syste should consider entrusting a ody with a lead in developing the set o principles and guidance and entrusting it with oversight o the ipleentation.

Recommendation 3: Analyse, assess and communicate uncertainties se the ost suitale uncertainty analysis approaches oe uncertainties in scientiic evidence are ‘technical’ (e.g. related to limitations in available data) or ‘methodological’ (e.g. related to the reliability of expert judgement). An example of a question which involves such uncertainties is ‘What are the risks to huan health due to the possile presence o sustance in ood products?’.

Technical and ethodological uncertainties can e easured and epressed as proailities which involves statistical data analysis andor epert udgeent. The choice o a ethod involves tradeos etween rigour the tie and resources availale and should take into consideration the iportance o the question to e assessed. The uropean ood aety uthority ( has produced etensive guidance on the use o these ethods which is suitale or use eyond the ield o ood saety.

Other uncertainties are ‘epistemic’ (e.g. about the sources of knowledge that are needed to address the issue) or ‘societal’ (e.g. whether the question is the ‘right’ one. nalysing the should e part o the early delieration with policyakers (and the pulic when appropriate. everal tools are availale including ini checklists and questionnaires and an uncertainty atri as used e.g. y the etherlands nvironental ssessent gency (. ‘edigree analysis’ is a ethod to reduce the aritrariness o soe such udgeents through the use siple ordinal scales ( coined with veral descriptions.

EC Group of Chief Scientific Advisors epteer 9 Scientific Opinion Scientific Advice to European Policy in a Complex World

ommunicate uncertainties and diverging scientific views Where there are uncertainties in scientific evidence or advice provide clarity about what is known partially known unknown and unknowable. As an ethical imperative uncertainties gaps and limitations in available knowledge should always be clearly communicated in the least ambiguous and most comprehensive way achievable. larify the reasons for diverging scientific views whenever possible.

or uncertainties expressed as probabilities numerical and verbal expressions should be combined (e.g. ‘70% certain i.e. likely ’) to reduce ambiguity and their technicality should match the main audience and be expressed in a clear way. A provides extensive guidance in that respect.

xpert panels and evidence reports should aim at consensus but not at the expense of the rigour of the deliberation. issenting views should be documented and explained. egitimate scientific dissent can be useful to policy – for example it may offer ‘early warnings’ on the importance of the problem or minority scientific views.

xplain the path from evidence to the advice

he reasoning applied by scientific advisors when bridging the gap between evidence and policy options or recommendations should be explained. his should include the assumptions made and normative positions taken as well as the limitations and uncertainties encountered.

cientific advisors should consider expressing (through consensus – where achievable) their confidence that their recommendations contribute to achieving the stated policy objective. or this approaches modelled on the ‘pedigree analysis’ may be considered.

10 eptember EC Group of Chief Scientific Advisors

Introduction

Scientific Opinion Scientific Advice to European Policy in a Complex World

1. Introduction

1.1. Why this opinion and for whom

This Scientific Opinion, hereafter ‘the Opinion’, has been produced by the uropean Commission’s Group of Chief Scientific Advisors, hereafter ‘the Scientific Advisors’.

nder resident uncer the uropean ommission ( has committed to putting better regulation principles and scientific eidence at the heart of policymaing. t asserted that ‘high uality scientific adice proided at the right time greatly improes the uality of legislation and therefore contributes directly to the better regulation agenda’. o this end in 0 the set up the roup of hief cientific disors as well as the egulatory crutiny oard both of which mae use of independent epertise.

e the cientific disors hae taen up our role because we are coninced that the use of scientific eidence and adice in policy is necessary for the functioning of democracies and for our societies to thrie and face the challenges of the coming decades. mong other legitimate inputs into decisionmaing science has a special role due to the rigorous and comprehensie nature of nowledge that it offers. efending this role of scientific eidence and adice as a core alue is particularly urgent and important now that scientific uncertainty is often eploited to manufacture a general distrust of science or to pursue narrow interests and agendas that ignore eidence.

s the new ollege of ommissioners will tae office in oember 0 we hae looed bac on roughly three and half years of eperience in adising the 0 0 ollege. ur reflections and eperience hae led us to the conclusion that use of scientific eidence and adice in ommission policymaing can and should be further strengthened.

n woring on this pinion we too into account insights from practice by listening to the eperiences of other practitioners of scientific adice – those adising the ommission international organisations or national goernments. e hae also considered the eidence from the scholarship on scientific adice representing arious schools of thought. ll that eidence together has reealed concrete areas where our own practice too can be further improed – and these are also reflected in the pinion.

he societal issues that policies see to address – such as biodiersity and pollution ensuring nutritious healthy and sustainable societal transformations due to the rise of artificial intelligence and other netgeneration digital technologies – are highly comple. he scientific eidence which is called upon to help address such problems is often eually comple and typically contains arious inds and degrees of uncertainty that are inherent to science.

httpsec.europa.euresearchsampdfc0fcommissiondecisionen77.p df

EC Group of Chief Scientific Advisors eptember 0 13 Scientific Opinion Scientific Advice to European Policy in a Complex World

One of the aims of science advice to poicy is to provide a reiabe and trusted uide throuh a these ayers of compeity and uncertainty. Our ambition is to contribute to that aim throuh this Opinion.

e are addressin in particuar

 oicymaers at the hihest eves of the uropean Commission, incudin the Coee of as the architects of the possibe future chanes to the scientific advisory system in the C

 C poicymaers at epert eve in poicy departments ho rey on scientific evidence and advice in their or

 embers of the euatory Scrutiny oard

 embers of C scientific advisory bodies and epert roups

 Scientists ho are – or are interested in orin as – eperts contributin to scientific evidence that underpins scientific advice to uropean poicies.

1.2. The approach

The or on this opinion started once the Commission endorsed our Scopin aper, pubished in ebruary .

A number of evidence sources underpin this Opinion, notaby

. A scopin orshop une . Consutation meetin ith scientific advice practitioners and C poicy maers arch referenced in the Opinion as ‘ractitioner Consultation’). . The revie report deveoped by the netor of uropean academies of science – uy SAA, . A report pubished by the C oint esearch Centre C, Understanding Our Political Nature. How to Put Knowledge and Reason at the Heart of Political Decision-Making – uy air et a., .

httpsec.europa.euresearchsampdfmeetinshsamscopinpaperscience.pdf httpsec.europa.euresearchsampdftopicsmasosscopinorshopoutcome.pdf httpsec.europa.euresearchsampdftopicsmasosconsutationmeetinsummary .pdf

14 September EC Group of Chief Scientific Advisors

Recommendations

Scientific Opinion Scientific Advice to European Policy in a Complex World

2. Recommendations

2.1. Engage early and regularly

2.1.1. Clarify boundaries between science, scientific advice, and politics

Ensure that scientific evidence and its analysis are differentiated from other factors which influence policy decisions such as interests, values, beliefs, and opinions.

ientii aie is a riin an roerin atiit tat ats as an intereiar eteen siene an oli oars et al iele r ruit et al ) it translates ratial soietal) issues ealt it oli into one or ore tenial sientii) roles i an e aresse siene an ten interrets te sientii ansers to eelo aie to oli aet )

oli an olitis are laen it alues oinions ieoloies an oetin interests arurst C reitt et al ) alues an eotions stronl inluene not onl olitial eaiour ut also eretions o ats air et al )

ientii eiene an aie are not te sole asis or te eisionain roess Cairne lier arurst o ) n line it a ore rinile o sientii aie ie te rinile o reresentation te authority to take final decisions about policies rests with ‘democratically reresentatie an ulicly accountable officials’ arurst )

iene an sientii aie) annot e oletel ree o alues eiter esite it ein a nole ieal oulas lo taan asano ) or eale te ruial role attriute to siene as a soure o eiene or oliies is ultiatel a alue osition too air et al ) u ore alue ositions soul e ae eliit in siene aie

ettin an lariin te ounaries eteen siene sientii aie an olitis are unaental to te iartialit o sientii aie an ene to estalisin an aintainin te trust o all tose inole aner ) – eseiall en sientii aisors intensel enae not onl it sientii eerts ut also it oliaers an te uli as e reoen te o see )

oeer artiularl or ole oli issues te ounaries are not usuall learl reeterine n at te a e onteste – oliaers interest rous an sientists asano )

ee air et al or an eale o a noratie stateent to tat eet

EC Group of Chief Scientific Advisors eteer 17 Scientific Opinion Scientific Advice to European Policy in a Complex World

or eample science is unlikely to ie a definitie answer to the uestion such as ‘should certain drus be decriminalised or legalised’ since the policies in that area are stronly drien by alues and ideoloical standpoints his may lead some policymakers as well as some scientists to state that the question is ‘not scientific’.

onersely others may claim that the debate can be ‘depoliticised’ and science can sole it based e on the consideration of relatie health risks alone ecision makers may also have incentives to ‘depoliticise’ a controversial issue by assigning it to scientific committees arkhurst

nother issue that may be contested concerns the uestion of whether scientific adice can formulate policy options or een recommend some without encroachin on policy and politics see o for an etended eample

a a minimalist role for scientific adice as a ‘science arbiter’; ielke is to synthesise assess and present the eidence and make it as releant to the policy issue as possible but stay clear of formulatin any forwardlookin options

b scientific adisors as ‘honest brokers of policy alternatives’ ielke where scientific adice formulates alternatie courses of policy action based on the analysis of eidence ielke b

c cientific adisors formulatin a recommendation – ie statin a preferred policy option amon seeral considered based on their reasonin and interpretation of eidence b while leain the decisions to policymakers his may involve multiple recommendations based on the ‘ifthen’ reasoning b

he last role listed aboe tends to be the most contested ome practitioners reect it outriht as encroachin on the domain of politics oyd see also o thers adise aainst it for socially controersial issues – because of the risk of scientific adice bein perceied as sidin with a aluedrien preference and thus ettin mied up in the controersy yler kerlof tirlin advocates it in the form of ‘plural conditional’ advice for comple policy issues with hih uncertainties

hereer the boundaries for scientific adice are drawn they need to be clearly areed at the start between policymakers scientific adisors and eperts takeholders and the public may also hae a role where appropriate see cientific adisors as credible and reflectie intermediaries hae a prime role in facilitatin the dialoue and analysis that are necessary for areein and settin the boundaries

n the eample of dru policy this would practically mean to define the uestions pertinent to the issue which scientific adice can usefully address in order to inform policy his approach can include assessin relatie health risks of arious leal and illegal drugs (‘science arbiter’), but also a comparison of the likely health, economic or crimerelated costs and benefits of various options (‘an honest broker of policy alternatives’), and even a review of relevant ethical considerations and dilemmas.

n the deelopin policy options for maor new policy initiaties is a core obectie of impact assessments s hese are drawn up by policymakers at

18 eptember EC Group of Chief Scientific Advisors Scientific Opinion Scientific Advice to European Policy in a Complex World

epert departmental level, rather than scientific advisors. owever, they draw heavily on specialised scientific evidence and epert input and are then evaluated by the egulatory crutiny oard.

he roup of hief cientific dvisors, operating at a crossdepartmental level and advising the ollege of ommissioners, has regularly been tasked with developing broad strategic recommendations to policy, next to occasionally acting as a ‘science arbiter’, e.g. through explanatory notes on comple and controversial topics.

There is a spectrum of legitimate roles for scientific advice in the EC. Defining them explicitly, ideally at the start of advisory work, is important for the clarity of its mandate. Concretely, this can be achieved by:

(a) scientific advisors, experts, and policymakers defining together the questions that scientific evidence and advice can address, including a decision on whether the formulation of evidence-supported policy alternatives is desired, and

(b) deciding together on the precise role and principles of involving stakeholders and the public in scientific advice.

2.1.2. contains guidance on both of the above dialogues.

NB. This recommendation does not in any way prevent any scientific or advisory body (e.g. national academies) from offering pro-active and unsolicited advice on EU policies without any need for defining the questions or the mandate with the policy clients. We are aware of national advisory models which emphasise distance rather than close collaboration, as a means of ensuring complete independence. They are a part of the plural advisory landscape in Europe and play a useful role within it. The co-creation model that we recommend for the EC is geared at facilitating the use of science advice developed by bodies enjoying frequent and direct access to EC policymakers, while ensuring its impartiality.

2.1.2. Define together the questions to be addressed

Engage early and regularly with those requesting scientific advice to ensure the questions for advisors are properly defined.

systematic review of the barriers to and facilitators of the use of evidence by policymakers (liver et al., ) reveals that collaboration and relationship building between scientists and policymakers are among the top five facilitators (net to the availability, clarity, relevance and reliability of research findings).

mpact assessments address the likely economic, social and environmental impacts of each of the shortlisted options which can be positive and negative; direct and indirect, intended and unintended httpsec.europa.eusmartregulationguidelinesdocsbrtoolboen.pdf n httpsec.europa.euresearchsaminde.cfmpgglyphosate, and on new techniques in agricultural biotechnology httpsec.europa.euresearchsaminde.cfmpgagribiotechnology

EC Group of Chief Scientific Advisors eptember 19 Scientific Opinion Scientific Advice to European Policy in a Complex World

A crucial way to help ensure that relevant questions are being addressed by scientific advice is to engage with policymakers from the very early stage , :5-6; KNAW, 2013:6; The Royal Society & Academy of Medical Sciences, 2018:14-15; Mair et al. 2019:64) with the purpose of defining the questions through reflective, nuanced and rigorous deliberation.

his eliberatie engageent is liely to itigate an issue hereby scientiic experts ay consier that policyaers as the the ‘rong’ uestions air et al., hereas ro the policymakers’ perspectie it is the scientists ho ay orulate the ‘rong’ uestions ithout being ase, an procee to anser the , .

he case stuy on bager culling as a response to boine tuberculosis ox illustrates a nuber o issues that ay go rong in scientiic aice, incluing a ispute scope o the proble to be aresse by science.

ac o eliberatie engageent can aniest itsel in a nuber o probles that ay lea to uestions ening up ‘wrong’ ro either perspectie. hese inclue in particular

. nexaine assuptions on all sies about the types o uestions that science can best anser or cannot anser hich ay introuce an ‘issue bias’ into the process arhurst, .

. he pressure to liit the aboe uestions to assessing easurable ris tirling, see also ...

. he act that uestions or scientiic aisors ay be eine ro the perspectie o a single epartental policy e.g. oo saety, enironent, or agriculture hereas they ay be a part o a ore coplex an ultiacete issue airney, .

. ‘lind spots’ in ho policyaers anor scientiic experts see a proble, hich ay hae eelope through eliberation ith staeholers an representaties o the public onig, orsen, eche, , .

. nsuicient unerstaning on the part o scientiic experts o ho policyaing an politics or air et al., yler, airney, as ell as insuicient scientiic literacy aong policyaers, incluing e.g. unerstaning the nature o scientiic eience air et al., or interpreting scientiic clais airney, utherlan et al., .

ue to their role as intereiaries beteen science an policy, scientiic aisors hae a central an coplex role in initiating an acilitating the eliberatie ialogue, as ell as itigating all the issues liste aboe.

The dialogue on problem definition typically needs to be iterative (Mair et al., 2019:69): as preliminary scientific evidence is gathered and analysed, the questions being addressed by scientific advice to policy may need to be refined again.

20 epteber EC Group of Chief Scientific Advisors Scientific Opinion Scientific Advice to European Policy in a Complex World

Box 1. Case study: the badger culling controversy (adapted from Montuschi, 2017)

n an ndependent cientiic eiew rop in te reiewed te aailale eidence aot te link etween terclosis in adgers and in cattle and conclded tat adgers are a signiicant sorce o inection in cattle oweer science was not ale to say at te time weter clling adgers can e eectie in comatting te spread o disease in cattle o answer tat estion a andomised adger lling rial took place etween and oerseen y te ISG. The resulting 2007 report concluded: ‘although badgers do contribute to the disease in cattle, clling adgers is not te soltion – and indeed it might even make things worse’.

The ISG took on a broad definition of the problem. Their report reads: ‘…while confirming its commitment to the scientific approach, the [committee] identified its core aim as being ‘to present inisters wit a range o scientiically ased policy options wic will e tecnically environmentally, socially and economically acceptable’. Accordingly, the ISG report included in its eidence ase – esides wide natral science gronding in genetic epidemiological ecological and enironmental stdies economic social practical and animal welare isses tat were deemed critical to assessing te eectieness o clling as a measre o control n response te goernment asked te oernment ie cientiic disor to assess te eidence proided y te report. The brief to GCSA was to ‘eclde economic and oter practical issues’ e report reaced a diametrically dierent conclsion tan stating that culling badgers is ‘te est option aailale at te moment to redce te reseroir o inection in wildlife’ – proided tat it is properly carried ot ollowing an estalised protocol e goernment decided to atorise clling in t stated tat making predictions on te eectieness and sustainability of culling badgers by the farming industry is ‘a matter of judgement not of science’ – sggesting tat te earlier report ad entred inappropriately into wat sold e te presere o policymaking only e in trn criticised te narrow scope o te rie e ie cientiic disor in te epartment or nironment ood and ral airs was o te iew tat ‘nortnately some scientists ae een drawn in to te plic deate aot wic policy option is correct scientists start to say tat ten tey are eginning to take te position o politicians and tey deale te scientiic eidence tey claim to present’ oyd nlike te earlier oweer te pointed ot in tat, because ‘the epidemiology of bovine TB is fiendishly complex’ seeral policy options are easile ased pon te eidence rater tan a single one n act seeral dierent policies were eing prsed in dierent risdictions at te time oyd net to proactie adger clling ngland tey inclded adger accination ales testing and accinating ortern reland and reactie adger clling eplic o reland oyd arged tat it wold e inaccrate to sggest tat ‘ased pon te eidence, any of these policy options is more or less correct’ and that ‘all are possile een i te eidence sggests tat some migt e more sccessl tan oters’ oyd et to eing a sorce o societal controersy and a dispte aot te ondaries o scientiic adice te case was also sect to scientiic controersy and dissent e accsed te report o ndamental scientiic errors in te se and interpretation o te data rom report and making a case or adger clling witot proper asis in eidence ince te second report was atored y a goernment adisor and entsiastically welcomed y te arming indstry accsations o a lack o impartiality also appeared wo adger clls took place in and et to etensie social protests on animal welare gronds stakeolder organisations opposed to clling as well as a nmer o scientists ae contested te goernment claim tat clling as een eectie in redcing te prealence o te disease oday wile te policy is eing contined te goernment as moed towards a mi o interentions rater tan insisting tat clling is te single est option aailale n stated tat: ‘there is no single measure that will provide an easy answer to beating the disease at is wy we are prsing a range o interentions to eradicate te disease y including tighter cattle movement controls, regular testing and vaccinations’

ttpsderamedialoggokoinetandadgerclling

EC Group of Chief Scientific Advisors eptemer 21 Scientific Opinion Scientific Advice to European Policy in a Complex World

When analysing a complex policy issue to define questions for scientific advice, take a systems perspective and consider the use of the best available aids.

efining the uestion for scientific advice should involve a rigorous analysis of the complex policy issue at hand. This is likely to help define aspects that can be most usefully addressed by scientific advice Ansell Geyer, 207 Geyer ihani, 200.

Complexity is a characteristic of a natural or social system where there are strong interactions among its elements, and where the causeeffect links between a multitude of interdependent variables are not fully understood or predictable SAA, 20:270 Geyer ihani 200:2.

xamples of complex systems include a flock of birds, the human brain C, 200:, traffic C, 207:, climate patterns, large interconnected infrastructures SAA, 20:2, healthcare systems Cairney, 20: or educational systems Snyder, 20 C, 207. olicymaking itself is a highly complex system air et al., 20: Cairney, 20.

eatures of complex systems of particular relevance to policymaking and scientific advice include:

 emergence: behaviour resulting from interactions at local level which restricts the behaviour of the whole system, thus making predictions of policy impact difficult. or complex policy issues, this implies that it can be useful to employ pilot schemes before big central policy rollouts Cairney, 20: 7 C, 200: Ansell Geyer, 207 Sanderson, 200.

 path dependence: dependence on the initial decisions and conditions e.g. resources historically committed to a policy ibid: . In policy and scientific advice this often means that, as Tyler 20 put it, ‘starting policies from scratch is very rarely an option’;

 the likelihood of some small actions having large effects, and large actions having small effects. In policymaking that may mean that certain issues are ignored or, conversely, disproportionate attention is paid to them in the policy agenda Cairney, 20:.

22 September 20 EC Group of Chief Scientific Advisors Scientific Opinion Scientific Advice to European Policy in a Complex World

Box 2. Stacey diagram: a tool to facilitate the definition of questions for scientific advice for, with an application example

he tacey iagram is ase on to aes the egree of certainty, an the level of politicalsocietal agreement on the issue in uestion nsell eyer, ; eyer ihani,

t parses a comple policy issue into ifferent ones, ith iffering scope for scientific epertise an avice nsell an eyer provie a case stuy of rug policy in the

one eg efficacy of aiction treatments – reuires specialist scientific epertise an stanar methos to assess efficacy an riss one eg rug legality policies – a valuelaen political one; scientific evience an avice may inform it eg ith evience on comparative health riss of rugs; liely health, economic an crimerelate impacts of various options; hoever, scientific evience is unliely to e the primary factor influencing ecisions; one eg est strategies for reucing rugrelate criminal activity – epert issent is common ue to scientific uncertainties eg comple an multiple causeaneffect relations; evience is largely ase on eperts eighing proailities of ifferent pathays; one – eg helping rug aicts ith a range of aitional prolems, eg health an poverty issues – a chaos zone where ‘adhoc coping and intuition’ predominate; the role of scientific evience is limite; some locally vali ualitative research may still e applicale; one – an area here comple science intersects ith societal compleity ith a ey role for scien tific avisors as intermeiaries, an for continual mutual learning ith policymaers; recommene policy options may involve a preference for smaller pilot proects as learning eercises

ar from Zone 4: agreement hoc c oping a n Zone 2: Zone 5 intuition olitical ecision maing

Zone 5: ra gmat ist c ompleity ecision maing

Zone 1: Zone 3: rofessional e chno ra tio nal Zone 5 lose to e c isi on mai ng e c isi on mai ng agreement

lose to ertainty ar from ertainty ertainty

Stacey diagram (Adapted from Ansell & Geyer, 2017; Geyer & Rihani, 2010)

he one escriptions, incluing their applicaility to science avice, have een aapte an epane

EC Group of Chief Scientific Advisors eptemer 23 Scientific Opinion Scientific Advice to European Policy in a Complex World

Analysing a complex policy issue to define a question for scientific advice requires reflective deliberation – since typically a number of very different questions can be formulated, depending on the level of analysis (see e.g. OECD, 2015:9-10, Kovacic, 2017)12.

The deliberation needs to take a ‘systems perspective’, which helps policymakers and advisors ‘think broadly about the whole picture rather than merely studying component parts in isolation’ (Prewkitt et al., 2012:60).

The deliberation can be facilitated through the use of straightforward conceptual aids. An example (the ‘Stacey diagram’) is presented in Box 2.

A number of scientific methods can and are being used to analyse complex systems (e.g. in epidemiology, traffic management or climate change models; OECD, 2009:11- 12). These methods include e.g. dynamic systems modelling, network analyses, sensitivity analysis or scenario modelling (ibid; OECD, 2017:22).

ntegrating oresight and horizon scanning into scientiic adice is particuar important or ‘wiced proems’ itte eer that cross departments cross disciplines and have timescales longer than the lifetimes of governments’ eddington see aso

oresight and horizon scanning may also mitigate the risk of missing ‘early warnings’. Stirling (2010:1029) provides historical examples: the belated recognition o the eect o seeming enign and inert haogenated hdrocarons on the ozone aer and the sow recognition o the possiiit o noe transmission mechanisms or spongiorm encephaopathies see aso for more ‘late essons from early warnings’.

n the the oint esearch entre has recent estaished the ompetence entre on oresight – egatrends u and has egun wor on understanding potentia impications o megatrends to speciic poic issues in a sstemic contet

Attention to ‘foresight and other forwardlooking tools’ is given in the EC Better eguation ‘Toolbox’ which stresses that those tools ‘complement quantitative modeing with a sstem thining and ongterm approach’. Among the roes o oresight in poic that are highighted is ‘inform[ing] policy generating insights regarding the dnamics o change uture chaenges and options that can e used as an input to poic conceptuaisation and design’

oacic proides an eampe o the issue o water scarcit and water management eicienc in srae which eads to dierent uestions and dierent uncertainties depending on the ‘scale of analysis’ (e.g. the level of the agricultural sector, society as a whole, or the ecosstem as a whoe httpseceuropaeunowedgepoicoresighten oo ‘Evidencebased Better Regulation’ httpeceuropaeusmartreguationguideinesdocsrtoooenpd

24 eptemer EC Group of Chief Scientific Advisors Scientiic pinion Scientiic dvice to uropean Policy in a omple orld

The ’s team supporting the Panel for the uture of Science and Technology (STA)1 – the main scientific advisory body of the parliament includes a foresight team.

reater integration o oresight and horion scanning should be considered in the work o scientiic advisory bodies in the particularly those which oer proactive advice on comple policy issues involving signiicant uture challenges

When policy issues are contested, involve stakeholders - and other members of the public - as appropriate. Consider the use of the best available analytic-deliberative approaches.

Complex policy issues addressed by scientific advice may be socially contested1 – which means that there are widely different opinions in society about the desired goals and courses of action. Contestation is typically due to different values, beliefs and attitudes (SAEA, 2019:, arkhurst, 201:9).

n E policy, examples of currently contested issues include migration and asylum, biotechnology (including e.g. gene editing), the use of pesticides in agriculture, climate change mitigation policies or the conservation of natural resources such as primeval forests1. Box 1 presents a case study of a contested national () policy of badger culling, and Box discusses an Ewide case of measures taken to control the spread of a dangerous plant disease which have met with protests and contestation through courts.

An acute form of contestation is issue polarisation – where there are few middle ground positions, and the issue is typically debated in binary terms (arkhurst, 201:). nternational examples of polarising policy issues include abortion, samesex marriage, the teaching of creationism at schools, or vaccination.

The scientific evidence which informs contested or polarising policy issues may be complex and uncertain – but in varying degrees, depending on the question or sub question.

or example, in the area of climate change, there is overwhelming scientific consensus that climate change is real and caused by humans (cf. Benestad et al., 201), and that taking no or little action will lead to catastrophic global

1 http:www.europarl.europa.eustoaenhomehighlights 1 This condition is also referred to as sociopolitical ambiguity (SAEA 2019:). 1 See (Stokstad, 201) and (onarewski et al, 201) for the societal and scientific controversy around the logging of the Białowieża primeval forest in , which led to a ruling by the European Court of ustice: http:curia.europa.euurisdocumentdocument.sfdocid20110

roup o hie Scientiic dvisors September 2019 Scientiic pinion Scientiic dvice to uropean Policy in a omple orld

onseuenes ithin a e eaes oeer the eat eets o limate hange on loal eather patterns are suet to man unertainties ue to their ompleit as is the ase also hen eiing hih speii mi o limatehange mitigation poliies is liel to e the most eetie enironmentall an soiall

heneer sientii aie is alle upon to inorm onteste or polarising issues memers or representaties o the puli shoul e inole rom earl stages PE see also eg E onig et al )

n suh ases inoling representatie staeholer organisations ill oten e suiient an the most pratial ational esearh ounil see also oeer or highl onteste issues inoling iniiual memers o the puli ho are iretl aete possile poli outomes or ho hae strong opinions aout them shoul e onsiere to help ensure that ‘the proess is not or oes not appear to e apture organise interests that ma not raise the full range of public concerns’ ational esearh ounil

To be done well public participation in scientiic advice must have a clear purpose) it must ollow clear principles and a rigorous method and have clear commitment in terms o time and resources (ational esearch ouncil )

ll o the above success actors should be a part o the early oint relection and analysis by scientiic advisors and policymakers

oles o sientii aisors partiipating in the analtielieratie proess inoling staeholers an the puli ma inlue

 together ith poli maers ientiing an oerlooe aspets o the prolem hih sientii aie might useull aress

 learl ommuniating the aailale eiene inluing its unertainties the sientii riteria that hae een applie to assess the eiene ualit the as in hih the eiene has een ientiie an use an the measures taen to ensure impartialit in snthesising the eiene see also an this is a part o transparen ientiie as one o the eatures o goo goernane o eiene Parhurst

 gathering ne eiene releant to the issue at stae eg the eperiential nolege o pratitioners aete potential poli eisions

 seeing onsensus on a possile aritration approah hen sientii eiene is onteste eg ue to the eistene o issenting eiene or allege ias in the eiene unerpinning the aisor proess

 gauging the impliations o the ierent poli options that sientii aisors are onsiering reommening ase on their analsis o the eiene

 learl ommuniating the reasoning that has le sientii aisors to ormulate poli reommenations or options ase on their analsis o the eiene

eptemer roup o hie Scientiic dvisors Scientiic pinion Scientiic dvice to uropean Policy in a omple orld

oel that is hel up as the ost suitable for science aice to polic is an analticeliberatie one hich cobines a rigorous analsis of aailable scientific eience ith broaer societal ialogue about its iplications such as the analticeliberatie approach eelope b the ational esearch ouncil o suarises its ost releant eleents

Box 3. A model for integrating public participation in scientific advice

dapted rom S ational esearch ouncil ( )

Possible purposes of public participation integrating scientific advice

proing policaing ualit an releance of scientific input eg b clarifing the nature of the proble

proing legitiac eg b seeing consensus on the proble to be aresse the creibilit of releant eience the process for conucting an assessent or inforing the polic ecision

Key principles

nsuring transparenc of ecisionreleant inforation an analsis aing eplicit attention to both facts an alues rooting eplicitness about assuptions an uncertainties ncluing inepenent reie of official analsis anor engaging in a process of collaboratie inuir ith intereste an affecte parties lloing for iteration to reconsier past conclusions on the basis of ne inforation

The currently has etensive guidelines and tools on stakeholder consultation which have been developed as part o the etter egulation package owever they do not eplicitly consider how scientiic advice and scientiic evidence can be a part o the process e recommend addressing this issue in uture versions o that guidance

httpseceuropaeuinfositesinfofilesbetterregulationguielinesstaeholer consultationpf httpseceuropaeuinfobetterregulationtoolboen tools to

roup o hie Scientiic dvisors epteber Scientiic pinion Scientiic dvice to uropean Policy in a omple orld

2.2. Ensure the quality of the scientific evidence

se the ull scope o good science

Consider all good science – from all scientific disciplines and perspectives - that could contribute to the issue at hand. This includes natural sciences, engineering, medicine, social sciences and humanities.

hile ethoologies such as ranoise controlle trials s an eta analses are highl appropriate to an scientific uestions eg the efficac an safet of eicines the are not suitable for an other uestions of releance to polic that science can aress ocial sciences rigorousl aress an uestions of irect releance to ecisionaing eg about econoic forecasting public acceptabilit an perceptions or eographic trens he eplo a ariet of ethoologies soe of hich eg s are er close to those eploe in natural sciences ualitatie research ethos in social sciences are particularl helpful for unerstaning why an how certain social an huan phenoena eg political attitues or unhealth behaiours such as poor nutrition occur rather than how oten or ith hich probabilit see b

uanities are aluable particularl for placing longter polic an societal issues an science aice itself in a broaer cultural an sociohistorical contet an proiing general conceptual fraeors for thining about the ro oe of the ethos use in the huanities eg in the analsis of ritten recors an aterial artefacts are ain to those in natural sciences soe others eg in soe branches of philosoph rel ainl on logical reasoning an arguentation

esearch in an oain an using an ethoolog fro s through sures to ualitatie ethos can be one ell or poorl hen one ell it has potential to offer aluable – an ierse insights to polic hen one poorl it a easil be misused by interest groups and ‘merchants of doubt’ reses ona to unerine other soli eience see b

cience has its on ualitcontrol an selfcorrection echaniss such as testing to eliinate hpotheses an is recentl taing ualit control further through increase attention to the reproucibilit of results here applicable oo eience snthesis for polic proies further ualit echaniss – eg b ensuring that the

ee also the eaple of the use of ranoise controlle trials in o g econoics sociolog social pscholog political science cognitie an behaioural sciences sustainabilit science public health stuies huan econoic an social geograph g philosoph histor anthropolog ethics his correspons to the uses of science in polic hich are tere conceptual utle an hich coer eliberatie use lucan orientation an enlightenent eiss see also arhurst

epteber roup o hie Scientiic dvisors Scientiic pinion Scientiic dvice to uropean Policy in a omple orld

eidence considered is reeant and comprehensie and reducing the arbitrariness of epert udgements see

eeoping scientific adice on compe poicy issues hich often mean reying on scientific eidence that is euay compe typicay reuires scientific epertise from different scientific perspecties eg the perspecties of different scientific discipines but aso different schoos of thought ithin discipines

ringing in these different scientific perspecties on compe issues is particuary important in the eary stages of probem identification see aboe hen the uestions for scientific adice are not yet edefined ence it may be far from certain hat inds of scientific epertise i be reeant – and a broad pura approach at that stage can hep determine hat focus oud be most appropriate

nother stage hen pura scientific perspecties are particuary important is deeoping poicy options or recommendations et to poicy practitioners and reeant representaties of the pubic see inoement of socia sciences eg economics socioogy or behaioura and cognitie sciences and humanities can be usefu – since formuating scienceinformed poicy options is connected ith assessing the societa scenarios they may ead to incuding any unintended conseuences see airney

eeoping scientific adice on a compe issue can ead to a number of highy specific scientific subuestions n our on practice these hae been in the natura sciences – eg toicoogy epidemioogy or genetics he subuestions are iey adeuatey addressed ithin a singe scientific discipine hich in practice means setting up panes of speciaised eperts from the same fied orma methods of epert noedge eicitation hich are used to uantify uncertainties may go yet further by reuiring eperts to be recruited not ony in a singe discipine but aso to hae a fairy homogenous discipinary profie see

n current practice the highestee permanent scientific adisory bodies are mutidiscipinary he members of the roup of hief cientific disors beyond representing dierse fieds are aso reuired to hae a broad ision hich coectiey aos interdiscipinary or he uropean roup on thics hich adises the oege of ommissioners on ethica uestions posed by scientific and technoogica innoation aso taes a muti and interdiscipinary perspective, with members ‘appointed for their expertise in the fields of law, natural and socia sciences phiosophy and ethics’ ther practitioners of scientific adice outside the stress the important effects of interdiscipinary deiberation ithin such bodies see esp ens

24 ecision on setting up the roup of hief cientific disors httpseceuropaeuresearchsampdfcecdecisionconsoidatedpdf ore on at httpseceuropaeuinforesearchandinnoationstrategysupportpoicy maingscientificsupporteupoicieseuropeangroupethicsscienceandne technoogiesegeen see aso nne

roup o hie Scientiic dvisors eptember Scientiic pinion Scientiic dvice to uropean Policy in a omple orld

urther improvements in multi and interdisciplinarity when advising the can be achieved through greater synergies between the roup o hie Scientiic dvisors and the uropean roup on thics

or adhoc epert panels working on comple topics which support permanent advisory bodies or are set up directly by policymakers in the we recommend that multidisciplinarity is thoroughly considered in the composition o the panels and panel deliberations are structured in a way that avours interdisciplinary work

nsure rigorous evidence synthesis

ood evidence for polic is defined as evidence that is appropriate ie useful to the polic needs b addressin the riht uestions and of hih ualit arhurst, he aspect of appropriateness ie defining the ‘right’ questions) is dealt with extensivel in his chapter deals with the ualit of evidence, and in particular with the ualit of evidence synthesis for polic

vidence snthesis for polic means brinin toether the best available information and nowlede from different disciplines to inform polic main he oal ociet cadem of edical ciences, t can draw both on preexistin documented sources and on new expert input e in the form of expert worin roups or hearins he section focuses on the former, as the latter is addressed in detail in

nthesis can be commissioned directl b policmaers e for the purpose of developin an evidencebased impact assessment of polic options or b scientific advisor bodies which will then use it as a basis for developin their advice to polic

wide variet of approaches to, and methods of, evidence snthesis exist see nnex he choice between them often implies tradeoffs e between speed and riour and is influenced b multiple factors, the most important of which include

 The purpose vidence snthesis ma be reuested e onl to ain a preliminar understandin of the state of nowlede as part of the wor on scopin the issue formulatin uestions as in a scopin review or to sstematicall collate and appraise all the available and relevant evidence as in the full sstematic review

 The time available vidence snthesis is used both in shortterm emerencies , and to inform medium or lonerterm challenes polic

eptember roup o hie Scientiic dvisors Scientiic pinion Scientiic dvice to uropean Policy in a omple orld

idene thus need to e snthesised in tter of hours nd ds during risis s in the se of the uushi disster) or oer onths or ers s in the se of the ntergoernent ne for ite hnge ) hih oersees orehensie snthesis roess ith reorts eer fie ers rig )

 Policy importance ientifi questions of high iortne to oi deisions eg on nning or not er oon used hei gent) re ore ie to require nd e grnted the use of ore tieonsuing ethods hih re ess rone to ises hen eidene snthesis rees signifint noedge g of high iortne for ongter oiies this in ft ed to requests for deeoing ne rir eidene o iustrtes suh se here rir reserh s onduted for ost dede to ddress the question of the effetieness of dger uing s ens of ontroing tueruosis in tte o desries request for ne risssessent stud hih infored the oi resonse to dngerous nt disese eidei ffeting seer ountries

 The breadth o the uestion oe of the ost forised snthesis ethods eg fu ssteti reies) re not usu suite for ddressing rod nd oe questions hih re ie to require eidene fro different disiines nd using riet of sientifi ethodoogies

The resources available – inuding udget nd stff

oe snthesis ethods offer ore ondensed ersions – hih re ess tie onsuing nd resoureintensie ut n e used for siir urose or ee rid eidene ssessent is ondensed ersions of fu ssteti reie ee nne ) sing ondensed ersions ti iies soe oroises in ters of orehensieness nd is riss ut is s to e referred oer een ess rigorous ethods heneer ossie oortion for nironent idene uin iser )

n the eidene snthesis tes not the fooing fors

 snthesis offered the in its role as the EC’s siene nd noedge serie – es through noedge entres in seifi res hih nge noedge eg in the for of sstei reies etnsis nd dt isuistion here the dend for noedge is oordinted ith oi dertents ) n ddition through its oetene entres the ors ith oi ients to deeo rir eidene eg through dned ethods suh s odeing hih n e used eg to

he uushi disster s nuer ident t the uushi uer oer nt in n he ident s strted the tsuni fooing erthque on rh to rioto )

roup o hie Scientiic dvisors eteer

assess the enironental econoic an social iacts o olicies an ata rien icroeconoic ealation

 eience reie reorts coissione scientiic aisor oies ro lic oranisations hich is the crrent oel se the ro o Chie cientiic isors ith a netor o acaeies o science esinate as the eience snthesis o

 ealations an sties coissione EC olic eartents ro eternal contractors

 snthesise otcoes o Ene research roects releant to the olic area in estion roects or olic

 other aailale reie reorts e orl an EC Eroean arliaent

he eneral alit criteria o eience snthesis incle oarsch et al

 corehensieness – ie coerin all releant eience in its arios ors ro eerreiee scientiic literatre throh re literatre to oinion ase nolee hel staeholers hen aroriate an alloin a clear ‐ ictre o the as an ncertainties hat is nnon artiall non or nnoale

 interaction ith clients an iteration e alloin a rerain o the estion an the search aroach ase on the analsis o reliinar reslts

 transarenc o the snthesis ethos e o the rotocols se or searchin araisin an collatin the eience

 aiin to iniise an iases to the etent ossile or the etho in estion onnell et al

or soe o the snthesis ethos articlarl ssteatic reies rai eience assessents scoin reies an eta analses etaile ethooloical an alitassrance iance eists Collaoration or Enironental Eience ‐ Collins et al etticre Roerts oococ llin aiser

he te o eience incle in eience snthesis is larel eterine the estion to e aresse an toloies o eience tes are aailale to ie that choice etticre Roerts or eale estions aot the eectieness o a olic interention are liel to incle riaril eeriental sties e RCs an oserational sties hereas estions relate to the oties o han ehaiors an attites e resistance to soe ne technoloies ill ticall riilee alitatie sties coine ith sres

httseceroaeinoresearchaninnoationstratesortolic ainscientiicsorteoliciesen

eteer

– –

 a ‘groupthink’ bias

 –

ibration 

raining 

ntiing an isosing experts’ intrsts as as assssing an anaging thir onits o intrst is an iportant part o nsuring transparn an intgrit in in snthsis an us h oa oit a o ia ins an sintii ai opoina ukan onig t a

n prt ai to poi pubi srutin an ontrors tn to ous on economic incentives which may adversely affect experts’ impartiality oor t a o t a However, ‘interests’ are not only financial – th a

ptbr

inclde e career advancement, professional reconition, ideoloical convictions or political enaement owe et al, , on et al, he crrently identifies several cateories of interests which memers of its expert rops need to declare financial interests sch as employment, sponsorship, research rants, investments, inflence on decisionmain sch as memership of manain odies, intellectal property rihts, personal interests e interests of immediate family, plic statements and positions, and ‘other’ see nnex for details

‘Interest’ in a field of activity is not synonymos with a conflict of interest, and therefore mst not atomatically disalify an expert xperts have interests in their field practically y definition – e y virte of ein employed y, or receivin research spport from an oranisation active in that field , ractitioner onsltation

Hence, the ‘avoidance’ approach haard, of atomatically disalifyin experts with a certain type of interest may e oth aritrary and detrimental to the ality of the expertise owe et al, nstead, conflicts of interest mst e assessed and comprehensively manaed owe et al, , haard, ive main strateies of manaement have een identified in scholarship haard, hey are smmarised elow, and the implications of each of them for the are discssed he recommendations formlated elow shold apply to all experts conslted y the , whether appointed as memers of formal expert rops, or conslted on an adhoc asis

1. Clarity of the expert selection criteria and a transparent selection process

2. Clarity of the exclusion criteria

Such ‘hard’ exclusion c

eptemer Scieniic inion Scieniic dice o uroean olic in a olex orld

he crrent enera re on I do not incde ch a it o ecion criteria – and thi a in act not e practica, ince thee re app acro the oard to ver divere tpe o epert coittee adviin the he eciion on invite oicia to conider I on a caecae ai

oever, oe odie hich rear et p epert pane ith iiar dicipinar proie and doin a iiar tpe o or have opted or ch it o ecion criteria appicae to a their epert ape incde the ropean ood aet thorit and reearch ndin odie ch a the ropean eearch onci oth oraniation ao e a it o e cearct itation to e aeed on a caecae ai

3. Public disclosure of interests

coon trate o deain ith I i to dicoe interet, o a to ect the to pic crtin, to the deree that repect privac re haard, on, , oe et a, oever, in dicore a the onl trate o anain I i proeatic, ince dicore ha een ond to ead to vario pervere pchooica eect ain et a, haard, on et a, or eape it a acta increae ia in epert ho had dicoed their interest, as they may feel ‘moral licence’ to exaggerate their statements further ain et a, it a ao neative predipoe the recipient o the advice or – convere – ead the to e diproportionate inenced the inoration that i ie to e iaed

In the toda, eer o epert rop actin in perona capacit ie epected to act ‘independent and in the public interest’ are reired to decare interet heir decaration o interet I are eent dicoed pic throh the reiter o epert rop to the etent that i copatie ith privac re pert invited the on an ad hoc ai are not reired to it ritten I t t inor the eore the eetin aot an I the have

oiion eciion o etaihin horionta re on the creation and operation of Commission expert groups (henceforth, ‘the 2016 Decision’). httpeceropaetranparencreepertIIII pd eciion o the ective irector o the ropean ood aet thorit on opetin Interet anaeent httpeaeropaeitedeatiecorporatepicationiecopetininte retanaeentpd httpeceropaereearchparticipantdatareherchercere aendedenpd pert rop a deined oiion eciion o , ie either ora etaihed a oiion deciion or inora, ie et p an individa departent that ha otained the areeent o the oiioner and icereident reponie and o the ecretariatenera httpeceropaetranparencreepertindec he eciion

epteer rou o hie Scieniic disors Scieniic inion Scieniic dice o uroean olic in a olex orld

n he ineres o urher enhanceens in ransarenc and consisenc he should consider inroducin he sae reuireens ih resec o declarin and disclosin ineress or all he exers consuled

ot all participants in the scientific adisory process are reuired to be impartial. taeholders representing specific interests (such as ciil society, industry, business) can hae a legitimate role in the scientific adisory process (see 2.1.2.), as long as their interests are clearly stated and disclosed. hey openly participate in the process as issue advocates (or ‘interest representatives’ in C terms). heir contributions may involve ‘science‘ or ‘evidencebased advocacy’ hich refers to the use and promotion of scientific eidence in order to influence policy in line ith a particular agenda (utchings tenseth, 2016 arhurst, 2012), and for hich the interpretation or selection of scientific eidence is therefore expected to be subect to bias.

neres reresenaies ineracin ih he are currenl reuired o e reisered in he ransarenc eiser he eiser reuires disclosin eneral oals and rei o an oranisaion as ell lisin he ain iniiaies olicies and leislaie iles olloed ha oranisaion

his aroach could e urher reined reuirin clear saeens on he osiions aen or ulic saeens ade ineres reresenaies on hose seciic iniiaies olicies or iles – analo o iarial exers ho are reuired o disclose such osiions and saeens

4. Oversight ersight as a C management strategy refers to the actions of a superisors or peers of an expert, ho are tased ith identifying the Cs to hich the expert in uestion may be obliious (hagard, 200). possible eaness in the oersight system that needs to be managed is that those tased ith oersight may be also be biased or act in arbitrary ay (ibid).

n the C, the oersight role is entrusted to C officials responsible for setting up expert groups. t inoles screening declarations of interest (Ds)6 prior to the expert’s participation in an expert groups, and taing one of the measures defined in the 2016 Decision if a C is found to exist. he main to measures are to

httpsec.europa.eutransparencyregexpertindex.cfm 6 n the case of experts inited on an adhoc basis, ho currently do not need to submit Ds but are reuired to inform the C of any C, the C also needs to assess such information and decide on the measures to tae.

rou o hie Scieniic disors eptember 201

exclude the expert or to restrict their contribution (eg by exclusion rom discussing speciic items on the agenda

5. Understanding bias he inal strategy is improving the understanding the nature o bias based on the best available scientiic evidence (air et al and considering that bias and conlict o interest are related but dierent concepts (hagard oore et al

n the a conlict o interest (I) is ‘any situation where an individual has an interest that may compromise or be reasonably perceived to compromise the individual’s capacity to act independently and in the public interest […]’ hus a mere or a o bias o a certain ind is enough or a to exist there is no need to prove that it occurred or that it did not occur

iases present in the creation selection and interpretation o scientiic evidence are not limited to those consciously driven by personal interest such as inancial or ideological motives (arhurst ognitive biases and heuristics are integral to human thining ( arhurst air et al see also or example o cognitive bias in understanding uncertainty hey are not a matter o deliberate choice to act in personal interest at the expense o public interest (oung

areness o cognitive biases does not usually mean that humans can avoid them (ahneman though spotting them in others is liely to be more successul (ibid One example of cognitive bias is motivated reasoning, i.e. the ‘tendency to arrive at conclusions about evidence that match people’s preexisting beliefs’ (Mair et al otivated reasoning is not related to reasoning ability (ibid and in act more analytical relection has been ound to increase ideologically motivated reasoning (ibid

he third option is to appoint the expert as an interest representative rather than an impartial expert i that possibility exists or the group in uestion (see nnex or details he ecision

eptember

Many valuable guidelines for and codified principles of scientific advice already exist at national and international level (see nnex ). he odes of practice for scientific advice are different in nature and purpose from other types of scientists’ codes of practice, (e.g. those regarding research integrity (, ), health research or research and industry), though they may share with them a common core, for example in matters of integrity (ato rimoto, ).

here is currently no single code of practice for scientific advice. number of principles of scientific advice and scientific expertise, as well some guidance for implementing them, can be gleaned from various existing documents. owever, they are fragmented and sometimes dated. his is a systemic weaness observed already by the EC’s first Chief cientific dvisor (lover, ).

he main currently existing documents include

 the principles and guidelines on the collection and use of expertise by the ommission (see nnex ). hey date bac years at the time of writing and the extent to which they are followed in the is no longer clear (lover, )

 ommission ecision on the creation and operation of ommission expert groups (which concerns all experts consulted by the ) – see nnex and ...

 he ommission ecision on the setting up of the roup of hief cientific dvisors, which lists a number of ualities which must be met by the roup members and its activities

 tatement on cientific Integrity

39 httpscodeofconductforhealthresearch.eu httpseuraxess.ec.europa.eusitesdefaultfilesamceeene.pdf httpsec.europa.euresearchsampdfcecdecisionconsolidated.pdfviewfi tpagemodenone

eptember

 etter eulation toolbo hich includes uidance to policymaers at epert leel on usin scientific eidence and epertise

he full tet of the statement nne httpseceuropaeurcsitesrcshfilesrc strateyenpdf etter reulation toolbo httpseceuropaeuinfobetterreulationtoolboen esp oolbo and

eptember

ncertainty at its most eneral refers to all types of limitations in aailable nolede E E a oeer there are idely different types of uncertainty E

 – hich results from limitations in aailable data t coers e missin data uncertainty about a sinle nonariable alue of a parameter uncertainty about the ariability of data multiple true alues ithin a rane or randomness uncertainty about measurement precision or about etrapolation alidity

 – hich can be about ho to assess a uestion it can also be produced e by limitations in the analytical strateies employed e uncertainty about ecluded factors about the reliability of epert udement or the appropriateness of search criteria to identify releant literature

 – ie uncertainty about hich inds of nolede are releant to the uestion at hand and about hat is at stae it may be produced by the hih compleity and openendedness of the system under study

 – hich can be uncertainty about hether all releant aspects of the policy problem at hand hae been considered about rial problem definitions or about the alueladen assumptions or biases that may be present in the problem definition he term may be etended also to ‘scenario uncertainties’ i.e. those concerning the possible outcomes of the different course of action includin unintended conseuences cf etersen et al

ssessin technical and methodoloical uncertainties belons to the core actiities of specialised ris assessment bodies such as the European ood afety uthority E or – at the national leel – e the erman ederal nstitute for is ssessment f hese to types of uncertainty are addressed in etensie and recent E uidance deeloped for the field of food safety E ab but are suitable to other fields due to the eneral nature of the approach

E uidance does not prescribe specific methods for uncertainty analysis but instead provides ‘a harmonised and flexible framework within which different methods may be selected, according to the needs of each assessment’ (EFSA b see nne for the main elements of the E approach he uidance

he four distinct cateories are an idealisation since in reality the boundaries beteen e epistemic and societal uncertainty or beteen methodoloical and epistemic uncertainty may be fuy E b proides an illustration of uncertainty about a nonvariable quantity (‘the total number of animals infected with a specified disease entering the EU in a given year’) and uncertainty about a variable quantity (‘body weights in a population’) httpefsaeuropaeu httpsbfrbundde

eptember

(EFSA b) provides a detailed evaluation of the strengths and weaknesses of all the existing methods.

Under EFSA’s model, uncertainties can be assessed through (a) statistical analysis of data or (b) expert udgement (see ... and ... for a discussion of how different biases in expert udgement are managed). Uncertainties for different parts of the assessment – relying on either of the approaches can then be combined, using recognised calculation methods.

he most elementary udgements of this kind typically take the form of (a) probability statements (on a scale of 100%), e.g. ‘less than 5% likely’; (b) probability bounds, combining (a) with a specified range of a parameter – e.g. ‘it is less than 10% likely that mean exposure exceeds 10 mg/kg bw per day’ (EFSA 2018a:23).

September

Xylella fastidiosa is a bacterium endemic in parts o the Americas. t causes plant diseases. any plants are asymptomatic carriers o the bacteria, which can contribute to its spread. he bacterium lies in the plant tissue and is normally spread by insects eeding on the aected plants. n 2013, the bacterium was ound inecting olie trees in the o Apulia in . he disease was causing a rapid decline in olie plantations and, by 2015, it was aecting the whole proince o ecce and other ones o Apulia. he bacterium had neer preiously been conirmed in Europe. n the EU territory, seeral cultiated plants o high economic alue other than olies (e.g. plums and cherries) can be aected, as well as many widespread ornamental plant species. Science adice in the orm o a new risk assessment study reuested rom and proided by EFSA50 conirmed that there is no biological or chemical control aailable that can eliminate the bacteria rom a diseased plant in open ield conditions. ther aailable scientiic eidence indicated that certain insects which act as ectors or the disease can coer distances o ca. 100 meters in 12 days. ence, the response o the EU and national authorities has been ocusing on the insect ector. n 2015, the E adopted a decision51 which reuired remoal not only o the inected plants, but also o all other plants susceptible to the disease– een in the absence o any symptoms o inection – situated within 100 metres o the inection. espite a maority scientiic iew that the response was appropriate or eradicating the disease, the measures met with public contestation in taly. Farmers hae protested against them as a threat to their lielihoods, and were supported by enironmentalists who deplored the uprooting o ancient trees (including healthy ones). ndiidual national court rulings hae ound in the farmers’ aour, stopping tree elling and the spraying o insecticide. A minority o dissenting scientists claimed that the trees could be healed through natural methods52. he European ourt o ustice (E), in a case brought by a group o indiiduals, conirmed the alidity o the eradication measures under EU law. n September 201, the E – in a case brought against taly by the E, ruled that taly was in breach o two o its obligations: to immediately remoe at least all the inected plants in the containment area, and to conducting annual sureys53. eanwhile, in Apulia, the disease has been spreading to the neighbouring ones. he demarcated area has been repeatedly changed and extended to address that. ue to the number o outbreaks, eradication o the disease in the buer one was no longer possible and in une 2018, the E adopted a decision5, extending the demarcated area by 20 km. he disease has also aected other EU countries. n , there was the irst outbreak in orsica in 2015. Some 25 outbreaks occurred in mainland , and some 350 in orsica; the entire island was placed under containment in 201. n , an isolated inding in Saxony was reported in 201. he plants were destroyed. Sureying in the buer one conirmed the absence o any other positie cases. n 2018, the disease the authorities declared the disease eradicated. n , the irst outbreak, in cherry trees, was reported in allorca in 201. n 201, the entire territory o the aleares was declared an area under containment. n 201, the bacterium was also reported in the region o alencia, and in 2018 in one tree the adrid region. he areas were demarcated and eradication measures taken. he EU territory with the exception o the oicially demarcated areas is now considered ree rom . he

he case study is based on E and E oicial data (except i indicated otherwise): https://ec.europa.eu/ood/plant/planthealthbiosecurity/legislation/emergencymeasures /xylellaastidiosaen 50 https://esa.onlinelibrary.wiley.com/doi/10.203/.esa.201.5 51 https://eurlex.europa.eu/legalcontent/E//F/uriEE:0201508 201802romE 52 https://www.theguardian.com/world/2018/ul/22/italyarmersolietreesxylellablight cowdungpuglia 53 https://curia.europa.eu/cms/upload/docs/application/pd/2010/cp1010en.pd 5 https://eurlex.europa.eu/legalcontent/E//uricelex:3201802

September 201

as se sre resear res a total of ca. € 9m, with the aim to improve e ee f e aerm a s ers a f sse ree measres

era ae s ese fr ea eefe e mama ams ess a fr eress essea a era assessmes aae s ra a see as as eress ees frm s se a eraes e fram f e es fr sef assessme a se rese e es ma as e as se ma e e ara f saeers a e sera f aes

ras esem a sea eraes are a es eresse esre a as raes eerse e a e are ese rese e fram f e es a are ere sea a ae reae ases s eaes em f se e se f era aass frameworks such as EFSA’s.

eereess ar e ae eraes frm e eares saes f rem ef s ra e esre a e ess e aresse sef ae are e ms reea es a a ese ess are remare ree measrae rs r see

erse e se rem ef rea affes aer ess es f eraes are sere e e ms mra ea a er s e ase ses aer e a a a sease eem a srae a ma r f esem a sea eraes are sere frm e se e ea a mea es

ara a e are erse s f era frm e eares saes f rem ef s ee a ssessme as ee ae seera ss a es e eeras rmea ssessme e ea sef assessmes as ar f a a rem aass a er s e a

er ms eera ee s are sae fr a eera eee sef asrs sef eers a maers as e as saeers a memers f e as arrae – see se e are ese fr seass a seass ae a a e frm f srafrar ess r esares

eame f a sae rre se a aa ee s re

eame f a eefe es fr rs assessme s e e aresse fr e ase f a a sease esre ‘a s e effeeess f r measres aas ?’’ ras a reae eame f a ra a ae se es fr a eera s Also referred to together as ‘deep uncertainties’: Stirling 2010; EFSA 2018b:34.

eemer

– he uidance has been developed b the Agenc in collaboration with trecht niversit. t has been in use since 2002, and is credited with ‘stimulating colearning processes among scientific advisors and policmakers for a deeper understanding and awareness of uncertaint and its polic implications’ SAEA 2019:8. he uidance is built as a laered set of tools, with increasing level of detail and sophistication, whereb the recommended depth of analsis depends on the resources available and the importance of uncertainties for the issue at hand.

an der Slus et al. 2008 illustrate the application of the uidance to a polic issue of regulating the health risks of particulate matter . he inihecklist and the uickscan uestionnaire are generall suitable as an aid in the deliberation with policmakers. he are built around the main uestions listed below. . As needed, the responses to uestion can be developed in further detail with the aid of the ncertaint atri, which is presented in Anne . . etailed uidance and the ool atalogue are the most suitable for further epert analsis. he include an ncertaint atri see Anne , and the SA methodolog see Anne 8.

September 2019

– ‘scenario uncertainties’, and societal uncertainties.

 –

 – –

ressin uncertainties in ters o roailit can e aiuous. eral expressions such as ‘very likely’ or ‘unlikely’ mean different things to different eole , . o sos an etree eale o te conseuences o suc aiuit.

An extreme example is provided by the L’Alquila case a aor and traic eartuae in tal in . cientiic eerts ere irisoned ater ein cared, aon oter tins, it ain contriuted to te sread o reassurin essaes tat a lare eartuae in te sort ter as ‘unliel’, desite teir consideration tat eartuaes are not redictale in te deterinistic sense. e use o te ter ‘unliel’ as aiuous. , .

o itiate tat aiuit, te nternational anel on liate ane and recoend usin aronised aroiate roailit scales – ic coine eral and nuerical eressions o roailit a o

eteer

contains the A scale he ambiguity can be further reduced if the numerical expression is presented before the verbal expression – e.g. ‘70% certain (i.e. likely)’ rather than vice versa AA

ommunicating uncertainty is not ithout the limitations related to many universal cognitive effects in perceiving such information As a result of hich complete unambiguity and neutrality of such communication is not achievable air et al xamples of cognitive biases and heuristics affecting risk perception are presented in ox see also AA A

or quantitative expressions of uncertainty eg for approximate probability or probability distributions A has developed very comprehensive guidance A hich is tailored to three broad audience categories entry informed and technical level he guidance also covers communicating inconclusive assessments among others

ther relevant uncertainties including those hich often result from the complexity of the issue – methodological epistemic and societal – also need to be clearly outlined As a general rule though this can be done only descriptively

eptember

 rein ncertainty are alay irectinal (eigen 0). r eale tey can e framed positively or negatively (‘40% chance of success’ v. ‘60% risk of failure’) which is knn t inlence e interret an react t te inratin (anean 0 0 enn 00 0).

 ele ten t e c re reccie y te erine te aar (e.g. raiatin) rater tan te rik an aere eent (e.g. a raiatin e tat i anger) ( 0).

 ele ten t e aere t aigity –i.e. tey ten t reer a riky t naig option over an ambiguous one (e.g. ‘30% chance of win’ v. ’2040% chance of win’; EFSA 0) an ten t ee te latter tye inratin a a ign inclete cience rater tan a genine itritin ( 0).

 ele ten t iate n ler i r iger ale in a raility itritin eening n tle cange in a etin at te i rae. r eale en reente it a rectin teeratre rie ‘y t egree’ ele ten t ick te i ale a a cenari en ake at tey tink aen t ick te rtcae cenari en ake at aen (eigen 0).

nicating ncertaintie (a ell a ienting cientiic inin – ee el) a te raack tting te in te tligt te lic tat ay e cetical at te eience. i can ce at te eene cnicating c i alreay knn an ra cnen it ay eny (an er li 0). ncertaintie tat are cnicate ay al e trategically erlaye y ie acate e interet are aerely aecte y te cientiic aice eing ere (ee reke an nay 0 arkrt 07 ee al ...).

ere i n knn eirical eience tat cnicating ncertaintie increae trt ang te ra lic. ntea tere i liite eeriental eience tat gget tat it e nt increae it r een ten t ler it (iegrit 0). i reactin te lic ccr ecae ncertainty ten t e een a nell in aking decisions, and a sign of ‘sloppy’ or ‘weak’ science (ii ee al 07).

eite tee limitations, drawbacks and ‘uncertainties about uncertainties’ (igrit 0) en cnicatin i an etical ieratie r cientiic aice ince t terie y kningly eaggerating certainty an cnen l e atal t te rincile te cientiic air a an net intereiary.

00

eteer 0

Epert consensus is an ideal outcome but should not be pursued at all costs. First, ecessive focus on consensus in epert panel deliberations may lead to various cognitive biases, such as overconfidence bias. anel deliberation techniues are available and recommended in this pinion to mitigate such adverse effects (see 2.2.3.). Seeking full consensus at the cost of overemphasising certainty and clarity can ultimately undermine the rigour of the evidence and the advice as well as trust in them (van der Sluis, 202).

Second, legitimate difference of views occurs even within single scientific disciplines. egitimate dissent is defined as the occurrence of minority scientific positions which have been based on rigorous and recognised methods and meeting the relevant uality criteria, but arriving at different conclusions than most other scientists (van der Sluis et al., 200). t must be clearly distinguished from dissent driven by poor (or even fraudulent) science. he latter can be used by interest groups to undermine other solid evidence (see e.g. enestad et al. for an eample of climate change denialism); it needs to be weeded out within science itself and in the evidence synthesis (see 2.2. and 2.2.2.).

For a comple uestion, which is typically approached from different disciplinary perspectives, the probability of legitimate minority views is typically higher.

Scientific dissent can be of great value to policy – for eample it may offer for eample ‘early warnings’ that are relevant to policy both in terms of minority scientific views and warnings that the problem is more severe than it might appear at first (Stirling, 200; van der Sluis, 202; see also 2..2.).

2.3.3. Explain the path from evidence to the advice

For any comple issue that is informed by scientific advice there tends to be a gap to be bridged by scientific advisors between (a) what is known (or partially known) about the different aspects of the problem and (b) the conclusions, informed by that knowledge, about the best courses of action for tackling the problem (SA, 204b; airney, 2064) – if the scientific advice mandate includes the latter (see 2...).

September 20

or eaple clear evience o the health riss o ecessive sgar consption an o the ain rivers o that nhealthy ehavior oes not atoatically lea to clarity aot the optial policy i to aress that prole in a particlar local national or international contet which col involve eg health warnings conser ecation behavioural ‘nges’ or financial isincentives see airney tley et al

riging the gap etween evience analysis an orlation o the avice is liely to reire spport ro soewhat ierent grops o eperts or the two types o activity airney see also

t also reires holistic an interisciplinary reasoning y scientiic avisors who oten analyse iverse sorces o scientiic evience hey also nee to e aware o the sociopolitical contet relevant to the prole at han heir reasoning st e rigoros an incle enhance awareness aong the avisors o the possile cognitive heristics an iases that ay aect the in orer to rece their averse ipact on the otcoe see also

cientiic avice in the operates at ltiple levels an taes varios ors see nne

t the top political level it typically aresses highly cople isses or which roa estions are orlate n eaple ro or own practice is the estion we aresse in the opinion on oo ro the oceans

n sch instances scientiic avisors typically orlate roa evienceinore strategic options or recoenations or policy orlating the reires holistic an interisciplinary elieration see aove nerpinne y evience spanning

ne o reent reasoning errors is that o overgeneralisation – wherey internal valiity a policy has een proven to wor soewhere is taen as evience o eternal valiity ie preicting that the policy will wor everywhere or in a very ierent contet his ay lea to costly policy errors artwright arie oer relevant case sties incling one o a orl an ntrition proect see also arhrst httpseceropaeresearchsapeetingshlgsascopingpaperoceano opviewitpageoenone

epteer

variou cienific iciline an ehooloie he avice i uch cae concern olic acion in hihl cole e uch a a foo roucion an conuion e ee ihin hich b efiniion i a uliue of caueaneffec lin oe of hich are uncerain

orulain oion or recoenaion a hi level of colei oe no norall len ielf o uaniaive ereion a robabiliie irlin eeron e al

nea he oion are uuall ore uefull aee an eree ualiaivel in ‘plural conditional’ terms (ibid), i.e. as a range of possible scenarios (which can be construed in ‘ifwhen’ terms; cf. also WRR 2017b) ih conieraion of oible uninene oucoe

vice on olic acion in hihl cole e ih an uncerainie a alo involve a reference for recoenin ilo olic acion hich allo for raaic learnin an a oible chane of coure before lare olic rollou ee

everhele he liel iac of oe e of olic oion ienifie hrouh hi reaonin roce a be aenable o bein aee hrouh avance oellin eho ee hich houl in uch cae be corie in he evience bae unerlin he avice

cienific avior a alo be ae o aner ver ellefine ueion here he anae for he avice erel eclue forulain an olic oion – hu ain he aviors to act as ‘science arbiters’ (see 2.1.1.) hi i he core acivi of ri aeen boie n eale of a ueion coe fro eene cae u on elaine b

nerin ueion of he above in involve uanificaion hich i icall achieve hrouh a cobinaion of aiical aa anali an eer ueen ee uch cienific aeen a ell be a ar of he evience ue for cienific avice arein iue a hiher level of colei

hile uanificaion of he avice in er of robabiliie i no norall ueful for he broa raeic cienific avice on cole iue ore rucure an riorou a of erein i ualiaivel raher han onl hrouh a narraive coul be coniere

eeber

 – –

 – ethod modelled on the ‘pedigree ’

eptember 201 ientii pinion ientii die to ropen oli in omple orld

. (2017). he uropean ode of onduct for Research ntegrit. Reised edition, 1–20. httpsdoi.org10.107e2011002 Ansell, C., & Geyer, R. (2017). ‘Pragmatic complexity’ a new foundation for moving beyond ‘evidencebased policy making’? oli tdie, (2), 1–17. httpsdoi.org10.10001272.201.1210 eddington, . (201). he science and art of effectie adice. n tre iretion or ientii die in hitehll (p. 1). Retrieed from httpwww.csap.cam.ac.umediauploadsfiles1fdsaw.pdf enestad, R. ., uccitelli, ., ewandows, ., ahoe, ., gen, . ., an orland, R., oo, . (201). earning from mistaes in climate research. heoretil nd pplied limtolog, (–), –70. httpsdoi.org10.1007s00700117 od, . (201). Research standard for policreleant science. tre, (7), 1– 10. httpsdoi.org10.10011a rom, . W. . (201). nstitutionaliing applied humanities enabling a stronger role for the humanities in interdisciplinar research for public polic. lgre ommnition, (1), 1–. httpsdoi.org10.107s1010212 urgman, . . (201). rting dgement o to et the et ot o pert. ambridge niersit ress. ain, ., oewenstein, ., oore, . (200). he irt on oming lean ererse ffects of isclosing onflicts of nterest. he ornl o egl tdie, (1), 1–2. httpsdoi.org10.102 airne, . (201). he oliti o ideneed oli ing. he oliti o idene ed oli ing. ondon algrae acmillan . httpsdoi.org10.10771 17171 airne, ., lier, . (2017). idencebased policmaing is not lie eidencebased medicine, so how far should ou go to bridge the diide between eidence and polic elth eerh oli nd tem, (1). httpsdoi.org10.11s121017012 artwright, ., ardie, . (2012). ideneed poli prtil gide to doing it etter. ford niersit ress. httpsdoi.org10.10177110712.00 ollaboration for nironmental idence. (201). ideline nd tndrd or idene nthei in nironmentl ngement erion . (. . . . ( ullin, rampton, d.). Retrieed from httpswww.enironmentaleidence.orginformation forauthors ollins, ., oughlin, ., iller, ., ir, . (201). he rodtion o i oping eie nd pid idene ement o to ide. Retrieed from httpnora.nerc.ac.uideprint12112R.pdf raig, . (201). o doe goernment liten to ientit o oe oernment iten to ientit ondon algrae acmillan. httpsdoi.org10.1007710 Donnelly, C. A., Boyd, I., Campbell, P., Craig, C., Vallance, P., Walport, M., … Wormald, C. (201). our principles to mae eidence snthesis more useful for polic comment. tre, (7710), 1–. httpsdoi.org10.10d10101 ouglas, . . (200). iene poli nd the leree idel. ittsburgh, a niersit of ittsburgh ress. Retrieed from httpwww.stor.orgstable.cttwrc7 . (201). te leon rom erl rning iene pretion innotion eport o . uropean nironment genc. httpsdoi.org10.200700 172 177 7221

eptember 201 rop o hie ientii dior ientii pinion ientii die to ropen oli in omple orld

A. (201). Guidance on xpert nowledge licitation in ood and eed afety Risk Assessment. ornl, (), 7. ttpsdoi.org10.20.efsa.201.7 A. (201a). Guidance on ncertainty Analysis in cientific Assessments. ornl, (1), . ttpsdoi.org10.20.efsa.201.12 EFSA. (2018b). The principles and methods behind EFSA’s Guidance on Uncertainty Analysis in cientific Assessment. ornl, (1), 2. ttpsdoi.org10.20.efsa.201.122 A. (201). Guidance on Communication of ncertainty in cientific Assessments. ornl, (1), 7. ttpsdoi.org10.20.efsa.201.20 GCA . (2010). The Government Chief Scientific Adviser’s Guidelines on the Use of Scientific nd ngineering die in oli ing. ondon. Retrieved from ttpsassets.publising.service.gov.ukgovernmentuploadssystemuploadsattacme ntdatafile20710gcsaguidelinesscientificengineeringadvicepolicy making.pdf Geyer, R., & Riani, . (2010). ompleit nd li oli ne pproh to tentirt entr politi poli nd oiet. Abingdon, Routledge. Glover, A. (201). A moment of magic realism in te uropean Commission. In . Wilsdon & R. Doubleday (ds.), tre iretion or ientii die in rope (p. 10). Cambridge. Retrieved from ttpwww.csap.cam.ac.ukmediauploadsfiles1fdsaw.pdf Gluckman, P. (201). e art of science advice to government. tre, , 1–1. Gluckman, P. (201). e sciencepolicy interface. iene, (0), . ttpsdoi.org10.112science.aai7 G cience. (2010). The Government Chief Scientific Adviser’s Guidelines on the Use of ientii nd ngineering die in oli ing. oernment ie or iene. ttpsdoi.org10.1002220.2017.1100 addaway, . R., & Bilotta, G. . (201). ystematic reviews eparating fact from fiction. nironment nterntionl, –, 7–. ttpsdoi.org10.101.envint.201.07.011 utcings, . A., & tenset, . C. (201). Communication of cience Advice to Government. rend in olog nd oltion, (1), 7–11. ttpsdoi.org10.101.tree.201.10.00 IGA. (201a). iene die to goernment diere tem ommon hllenge rieing pper or the lnd onerene. Auckland IGA. IGA. (201b). nthei report iene die to goernment onerene gt lnd e elnd httpingorgpontentplodntheieportienedie tooernmentgtpd. asanoff, . (17). Contested boundaries in policyrelevant science. oil tdie o iene, (2), 1–20. ttpsdoi.org10.117700127017002001 RC. (201). trteg . Brussels. aneman, D. (2011). hining t nd lo. (Penguin, d.). ondon. arlsson, . ., & akaasi, R. (2017). reore or deeloping n eidene nthei report or poliming. Copenagen. AW. (201). nl onerning dem dior report i priniple proedre nd lit rne. Retrieved from ttpswww.knaw.nlsaredresourcesadvieenbestandenManualConcerningAcadem yAdvisoryReports.pdf Konarzewski, M., Zabielski, R., Kowalczyk, R., & Duszyński, J. (2018). Białowieża Forest: ogging data lacking. iene, (7), .

rop o hie ientii dior eptember 201 Scientific inion Scientific Advice to uroen olic in Comle orld

httpsdoi.or10.112science.aat02 oni . orsen T. Emmeche . (201). The ethos o postnormal science. utures (S) 12–2. httpsdoi.or10.101.utures.201.12.00 oacic . (201). nestiatin science or oernance throuh the lenses o compleity. utures 80–8. httpsdoi.or10.101.FUTUES.201.01.00 Kowarsch, M., Garard, J., Riousset, P., Lenzi, D., Dorsch, M. J., Knopf, B., … Edenhofer, O. (201). Scientiic assessments to acilitate deliberatie policy learnin. lrve Communictions (ay) 102. httpsdoi.or10.10palcomms.201.2 eopoldina. (201). Guidelines for dvisin olicmers nd societ. erlin. air . Smillie . a laca G. Schendiner F. ayosa . astor . an ael . (201). Understndin our oliticl nture o to ut nolede nd reson t the hert of oliticl decisionmin. (ublications ice o the European Union Ed.) (1111 ed.). uembour. httpsdoi.or10.2010822 ontuschi E. (201). Usin science main policy hat should e orry about uroen ournl for hilosoh of Science (1) –8. httpsdoi.or10.100s1101 01 oore . A. Tanlu . aerman . . (2010). onlict o interest and the intrusion o bias. udment nd ecision in (1) –. ational esearch ouncil. (2008). ulic rticition in environmentl ssessment nd decision min. ulic rticition in nvironmentl Assessment nd ecision in. ashinton The ational Academies ress. httpsdoi.or10.12212 ootny . (201). The Cunnin of Uncertint. (. ress Ed.) uroen ournl of rnic Chemistr (ol. 2012). ambride. httpsdoi.or10.1002eoc.201200111 E. (200). Alictions of comleit science for ulic olic ne tools for findin unnticited conseuences nd unrelised oortunities. aris. httpsdoi.or10.101.1828 E. (201). Scientiic Adice or olicy ain. rnistion for conomic Cooertion nd eveloment (21). httpsdoi.or10.18sl1cpben E. (201). ete the ssues Comleit nd olic min. C nsihts (E nsi). aris E ublishin. httpsdoi.or10.1882211en lier . nnar S. orenc T. oodman . Thomas . (201). A systematic reie o barriers to and acilitators o the use o eidence by policymaers. C elth Services eserch . httpsdoi.or10.1181212 reses . onay E. . (2010). Merchants of doubt : how a handful of scientists oscured the truth on issues from tocco smoe to lol rmin. ondon loomsbury ress. ens S. E. (201). Knowledge, policy, and expertise : the UK Royal Commission on nvironmentl ollution . ord rod Uniersity ress. arhurst . (201). The olitics of vidence rom evidencesed olic to the ood overnnce of evidence (1st ed.). e or outlede. etersen A. . anssen . . . an der Sluis . . isbey . S. aet . . ardeer . A. artinson uhes . (201). Guidance or uncertainty assessment and communication. …, 34. Retrieved from www.pbl.nl/en etticre . oberts . (200). Eidence hierarchies and typoloies horses or courses. idemiol Communit elth 2–2. etrieed rom httpech.bm.comcontentech2.ull.pd etticre . oberts . (200). Sstemtic evies in the Socil Sciences A rcticl Guide. (. etticre . oberts Eds.). ord U lacell ublishin td. httpsdoi.or10.100280088

September 201 C Grou of Chief Scientific Advisors Scientific inion Scientific Advice to uroen olic in Comle orld

Piele, R. . . The onest roer in sense of science in olic nd olitcs. ew or ambride niversit Press. https//doi.or/DO ./BO Prewitt, K., chwandt, . ., traf, M. L., Knowlede, ., Behavioral, D., ciences, . . sin cience as Evidence in Public Polic. https//doi.or/./34 Pullin, A., Frampton, G., Jongman, R., Kohl, C., Livoreil, B., Lux, A., … ittmer, . . electin appropriate methods of nowlede snthesis to inform biodiversit polic. iodiversit nd Conservtion, , –3. https//doi.or/./s3 3 Ravetz, J. R. . Scientific nolede nd its socil rolems. Oford larendon Press. Renn, O. . is overnnce Coin ith uncertint in comle orld. London Earthscan. https//doi.or/.434/444 Rittel, . J., ebber, M. 3. Dilemmas n a General heor of Plannin. olic Sciences, , –. https//doi.or/./B43 Rowe, S., Alexander, N., Clydesdale, F. M., Applebaum, R. S., Atkinson, S., Black, R. M., … edral, E. . undin food science and nutrition research inancial conflicts and scientific interit. utrition Tod, 3, –3. https//doi.or/./.b3e3a4b34 Rowe, S., Alexander, N., Weaver, C. M., Dwyer, J. T., Drew, C., Applebaum, R. S., … Westring, M. E. 3. ow eperts are chosen to inform public polic an the process be improved elth olic, 3, –. https//doi.or/./.healthpol.3.. anderson, . . ntellient Polic Main for a omple orld Pramatism, Evidence and Learnin. oliticl Studies, 4, –. https//doi.or/./.4 4... aner, M. . emporal and spatial dimensions in the manaement of scientific advice to overnments. lrve Communictions, , –. https//doi.or/./palcomms.. PE. . in sense of science under conditions of comleit nd uncertint. , German. https//doi.or/.3/MO ato, ., rimoto, . . ive ears after uushima cientific advice in Japan. lrve Communictions, . https//doi.or/./palcomms.. ierist, M. . ncertainties about the communication of uncertainties. n nterntionl Conference on Uncertint in is Anlsis Chllenes nd Advnces in Assessin nin nd Communictin Uncertint. https//doi.or/./cr343 lob, M., taman, J. . olic nd the evidence est thenu nstituut. he aue Rathenau nstituut. Retrieved from http//www.rathenau.nl/en/publications/policand theevidencebeast.html nder, . 3. he imple, the omplicated, and the omple Educational Reform hrouh the Lens of ompleit heor. https//doi.or/./3tnptlnren pruit, P., Knol, . B., asileiadou, E., Devilee, J., Lebret, E., Petersen, . . 4. Roles of scientists as polic advisers on comple issues literature review. nvironmentl Science nd olic, , –. https//doi.or/./.envsci.4.3. tirlin, . . Keep it comple. ture, –4. tostad, E. . Last stands. Science, 3, 4–43. https//doi.or/./science.3.3.4 utherland, . J., pieelhalter, D., Burman, M. . 3. Policwent tips for interpretin scientific claims. ture, ovember, –. eien, K. . . he tamin of uncertaint how we mae sense of chance b words and numbers. n roceedins of the nterntionl Conference on Uncertint in is

C Grou of Chief Scientific Advisors eptember Scientific inion Scientific Advice to uroen olic in Comle orld

Assessment Belin. httpsdoi.org.cry Thagard, P. . The moral psychology o conlicts o interest nsights rom aective neuroscience. ournl of Alied hilosoh, , –. httpsdoi.org.....x The Royal Society Academy o Medical Sciences. . vidence snthesis for olic A Sttement of rinciles. London. Retrieved rom httpsroyalsociety.orgmediapolicyproectsevidencesynthesisevidence synthesisstatementprinciples.pd Tyler, C. . Top things scientists need to know about policymaking. The Gurdin, p. Monday December. Retrieved rom httpswww.theguardian.comsciencedecscientistspolicygovernments science Tyler, C., Akerlo, K. . Three secrets o survival in science advice. ture , , . httpsdoi.org.dx Tyshenko, M. G., Oraby, T., Darshan, S., Westphal, M., Croteau, M. C., Aspinall, W., … Cashman, N. . xpert elicitation on the uncertainties associated with chronic wasting disease. ournl of Toicolo nd nvironmentl elth rt A, –, –. httpsdoi.org... van der Sluis, J. P. . ncertainty and dissent in climate risk assessment A postnormal perspective. ture nd Culture, , –. httpsdoi.org.nc.. van der Sluis, J. P., Petersen, A. C., Janssen, P. . M., Risbey, J. S., Ravet, J. R. . xploring the uality o evidence or complex and contested policy decisions. nvironmentl eserch etters, , . httpsdoi.org. van Gelder, T., odicka, R., Armstrong, N. . Augmenting xpert licitation with Structured isual Deliberation. Asi nd the cific olic Studies, , –. httpsdoi.org.app. Weiss, C. . . . The Many Meanings o Research tiliation. ulic Administrtion evie, , –. Woodcock, P., Pullin, A. S., Kaiser, M. J. . valuating and improving the reliability o evidence syntheses in conservation and environmental science A methodology. ioloicl Conservtion, , –. httpsdoi.org..biocon... oung, M. S., Mahoud, J. M., Stanton, N. A., Salmon, P. M., Jenkins, D. P., Walker, G. . . Conlicts o interest The implications o roadside advertising or driver attention. Trnsorttion eserch rt Trffic scholo nd ehviour, , –. httpsdoi.org..tr... WRR a ‘Scientiic and administrative integrity o the WRR’. The Scientiic Council or Government Policy WRR. httpswww.wrr.nlpublicatiespublicatieswetenschappelikeen ambtelikeintegriteitvandewrr WRR (2017b) ‘The profile and working methods of the Netherlands Scientific Council for Government Policy’ (in Dutch). The Netherlands Scientiic Council or Government Policy WRR httpswww.wrr.nlbinarieswrrdocumentenpublicatiesproielen werkwiewetenschappelikeraadvoorhetregeringsbeleidWerkwieproielWRR .pd

September C Grou of Chief Scientific Advisors

Annees

Scientific inion Scientific Advice to uroen olic in Comle orld

Although the Commission has considerable inhouse epertise, it needs specialist advice from eternal eperts as a basis for policymaking. This may be provided by groups of eperts or eternal consultants, or take the form of studies.

An epert group is a consultative body set up by the Commission or its departments to provide them with advice and epertise composed of public andor private sector members which meets more than once.

Gathering epertise from various sources may include gathering the views of various stakeholders.

pert groups may be composed of the following types of members A

A. ndividuals appointed in their personal capacity who are to act independently and in the public interest

. ndividuals appointed to represent a common interest shared by stakeholders in a particular policy area, who do not represent an individual stakeholder, but a policy orientation common to different stakeholder organisations. Where appropriate, those individuals may be appointed on the basis of proposals put forward by the stakeholders concerned

C. Organisations in the broad sense of the word, including companies, associations, Non Governmental Organisations, trade unions, universities, research institutes, law firms and consultancies

D. Member States authorities, at national, regional or local level

. Other public entities, such as third countries' authorities, including candidate countries’ authorities, nion bodies, offices or agencies and international organisations

Type and Type C members can only be appointed if they are registered in the Transparency Register0.

Only individuals applying to be appointed as Type A members of epert groups or subgroups are reuired to declare interest, i.e. to disclose ‘any circumstances that could give rise to a conflict of interest. and are thus reuired to submit a declaration of interests ('DOI’) form on the basis of the standard DOI form, together with an updated curriculum vitae ('CV’), as part of their application to become members of an epert group or subgroup.

The DO form consists of a series of standard uestions reuesting individuals who wish to act as eperts appointed in a personal capacity to disclose any interest relevant to the subect of the work to be performed. The uestions are in the following categories of interest

1. mployment, consultancy and legal representation (within the past years)

Summary of key aspects based on Commission Decision of 0..201 establishing horiontal rules on the creation and operation of Commission epert groups httpec.europa.eutransparencyregepertPDC201011COMMSSONDCS ONN.pdf 0 httpec.europa.eutransparencyregisterpublichomePage.do

C Grou of Chief Scientific Advisors September 201 Scientific inion Scientific Advice to uroen olic in Comle orld

embership of managing bodies, scientific advisor bodies or similar (within the past ears)

esearch support (eg grants, fellowships, sponsorship) – within the past ears

inancial interests (eg investments including stocs or shares)

Intellectual propert (eg patents, trademars, coprights)

ublic statements and positions within the past ears (a) epert opinions and (b) representing an interest or defending an opinion in an official capacit

ach individual must assume full responsibilit in relation to the content of the declaration submitted Individuals who answer uestions in the affirmative are ased in the DOI to suppl further details n affirmative answer in the DOI form does not automaticall disualif the individual concerned, but reuires the competent Commission departments to have it screened in accordance with the Decision, in order to determine if a conflict of interest eists

In principle, the conflict of interest assessment are to be performed b officials of the nit responsible for the management of the group or subgroup in uestion Officials operating in other departments ma also be associated, as appropriate or the purposes of the assessment, a number of factors are to be taen into account including the nature, tpe and magnitude of the individual's interest, as well as the degree to which the interest ma be reasonabl epected to influence the individual's advice n interest is considered to be insignificant or minimal where it is unliel to compromise or to be reasonabl perceived as compromising the epert's capacit to act independentl and in the public interest when advising the Commission

here the responsible officials consider partiall or full ecluding an individual from the wor of an epert group or subgroup, the ma contact the individual in order to obtain an additional information that ma be needed for the final assessment of an conflict of interest

here the competent Commission department concludes that no conflict of interest eists, the individuals in uestion ma be appointed as members acting in a personal capacit, provided the possess the epertise reuired and the other conditions in these rules are fulfilled

here the competent Commission department concludes that the individuals' interests ma compromise or be reasonabl perceived as compromising their capacit to act independentl and in the public interest when providing advice to the Commission, one of the following measures must be taen to deal with the conflict, depending on the specific circumstances

(a) the individual's application shall not be retained; […];

(b) the individual's appointment as member of the epert group or subgroup in a personal capacit shall be made subect to specific restrictions, eg eclusion from certain meeting andor activities, (in particular from drafting opinions and recommendations), or reuiring abstention from the discussion of, or vote on, specific items […];

(c) the individual shall be appointed as member of the epert group or subgroup representing a common interest shared b a number of staeholders (pe member), after consultation of the stakeholders concerned. […].’

he DOIs of pe members of eperts groups are made public through the register of epert groups

pe members and representatives of pe C members do not have to declare interest – because the are not reuired to act independentl, but instead represent an interest that is openl declared he ransparenc egister, where the are reuired to register before the can be appointed, ass them to state the general goals of the organisation

eptember C Grou of Chief Scientific Advisors Scientific inion Scientific Advice to uroen olic in Comle orld

epresentatives o pe and ebers do not have to declare interest or be reistered in the ransparenc eister since the are not reuired to act independentl but epress the vies o public authorities hich the represent.

he representative in an epert roup a invite eperts ith speciic epertise ith respect to a subect atter on the aenda to take part in the ork o the roup or subroup on an adhoc basis. uch invited eperts do not have to subit s. he are hoever reuired to inor the copetent departent beore the eetin o an potential conlict o interest.

C Grou of Chief Scientific Advisors epteber Scientific inion Scientific Advice to uroen olic in Comle orld

 set o eerin uidin principles ( b)  dditional ualities or successul science advice ( a)  rucial eleents o a scientiic advisor sstes ( )

 olic recoendations raeorks and echaniss o science advice ( )  uidelines and ood practices or the advisor process ( )

 ood use o researchbased evidence in policakin (echnopolis i ii)

 uidelines or advisin policakers and societ.

 he cornerstones o political consultanc’

 uidelines on ood science advice to polic

 uidelines on science advice to polic and societ

 cientiic and adinistrative interit o the

 he proile and orkin ethods o the (in utch)

 anual concernin acade advisor reports. asic principles procedure and ualit assurance.

https.leopoldina.oruploadstleopublicationeopoldinaeitadenolitikberatun.pd https.akadeienunion.deentheunionconsultancorpoliticsandsociet http.bba.deservicepublikationenbestellenaniesteund leitlinienolitischeeitlinien.pd https.acatech.deublikationleitlinienuerdiepolitikundesellschatsberatun https.rr.nlpublicatiespublicatiesetenschappelikeenabtelikeinteriteitvan derr https.rr.nlbinariesrrdocuentenpublicatiesproielenerkie etenschappelikeraadvoorhetreerinsbeleiderkieproiel.pd

epteber C Grou of Chief Scientific Advisors Scientific inion Scientific Advice to uroen olic in Comle orld

 ’ ‘ssueschllenes nd oortunities in science dvice to olic’

C Grou of Chief Scientific Advisors

‘ ’

‘ ’

– ‘ ’

http://ec.europa.eu/governance/docs/comm_expertise_en.pdf

‐ ‐ ‐ ‐

influence analysis time and resources available, it may be sufficient to characterise overall defined –

– – ’ ’

 ‘assessment’ refers to expert judgement of reliability;  numerical expression of ‘scientific strength’

edian scores obtained through ualitatie pedigree analysis ie expert judgement can then be plotted against uantitatie parameters eg expressed as ariability ranges

n example belo uses case study of melamine ris assessment t shos a three uantitatie parameters healthbased guidance alue for melamine; consumption of melamine containing chocolate and melamine concentration in mil poder ith their technical uncertainty ariability ranges shon as purple boxes and b the median of scores of experts on all four dimensions of the ‘pedigree analysis’ (green diamonds ith the error bars shoing the range of all scores

an der luijs illustrates its application of to expert judgement concerning monitoring the emissions of three air pollutants

he reported strengths of the method include the fact that is that it adds rigour to ualitatie judgements by encouraging systematic and consistent analysis of arious uncertainties is easy to use including for uic informal analyses – and the result are easy to understand for nonexperts

he reported eanesses of the method include the fact that it still results in fairly approximatie judgement and – as for other structured methodologies the need to train experts in the method b

eptember

aria niersity ollege ulin oerto World Health Organization Representation to the EU, member of the EC’s ientifi ommittee on ealth nironmental and merging iss ( en niersity ollege rans he etherlands ientifi ounil for oernment oliy ( and ilurg niersity aul niersity of tirling assimo niersity of aples ederio asha ffie of the hief iene disor of anada lexandra entre for nironmental oliy mperial ollege ondon arah entre for iene and ehnology tudies eiden niersity aniel niersity of onstan irsten niersity of outhern enmar ouis ational ademy of ehnologies of rane ar hief iene disor of reland aruh arnegieellon niersity ilio niersity of ergen oert anaster niersity ario niersitat utnoma de arelona uy uropean olitial trategy entre ( rg inrih erman ational ademy of ienes eopoldina eter hief ientifi diser for ales ndre Non-affiliated-expert athrine niersity of slo arta uropean ood afety uthority ( lan openhagen usiness hool ( atthias niersity of ergen nne reet uropean iene die orum ( ora niersit utonoma de arelona oahim oint esearh entre uropean ommission tefan niersity of ente ustus otsdam nstitute for daned ustainaility tudies adeus Wrocław University of Science and Technology, PL ara ioenter inland i hristina oyal nstitute of ehnology toholm leonora ondon hool of onomis and and niersity of enie a osari lexandru niersity olitehnia of uharest elga oial tudies of iene urih usan niersity of amridge ustin ondon hool of onomis and olitial iene ( rthur niersity ollege ondon arara University of Vienna and King’s College London, AT and UK uia openhagen ussiness hool rtin nternational nstitute for daned ustainaility tudies ( ilsri und niersity ndy ussex niersity eroen niersity of ergen osalia inia ia oen loal oung ademy and ario niersity of agre ames niersity of heffield ilippo ational nstitute of strophysis

eptemer

Conflit of interest

elaration of interest

Eropean ood afet thorit

Epert noledge eliitation

Eropean Union

noledgealit ssessment

ongoernmental organisation

meral Unit pread ssessment edigree an nertaint assessment method

iene die for oli b Eropean ademies

eptember

mbigit n this Opinion, ariabilit of interpretations based on idential obserations or data assessment

naltideliberatie model model of pbli partiipation in sientifi adie hih ombines a rigoros analsis of aailable sientifi eidene ith broader soietal dialoge abot its impliations

pproimate probabilit sale hih ombines erbal and nmerial sale epressions of probabilit

Cognitie bias n inolntar pattern of thining that leads to distorted pereptions and dgements that an reslt in errors in reasoning, logi and ealation

Compleit harateristi of a natral or soial sstem here there are strong interations among its elements, and here the aseeffet lins beteen a mltitde of interdependent ariables are not fll nderstood or preditable

Coneptal se of siene Comple and often indiret as in hih siene in poli an hae an impat on the noledge, nderstanding and attitdes of poli maers and pratitioners

eliberatie se of ientists and polimaers engage in a dialoge siene in poli ith the prpose of defining the estions to be ansered throgh a refletie, naned and rigoros deliberation

Emergene ehaior reslting from interations at loal leel hih restrits the behaior of the hole sstem, ths maing preditions of poli impat diffilt or omple poli isses, this implies that it an be sefl to emplo pilot shemes before big entral poli rollots

Enlightenment as a se of eing informed abot the stateoftheart of fatal siene in poli isses and asalfntional relationships that form reliable noledge

Epistemi nertaint Unertaint abot hih inds of noledge are releant to the estion at hand, and abot hat is at stae

Eidenebased adoa he deliberate proess, based on demonstrated

eptember

eidene, to diretl and indiretl inflene deision maers, staeholders and releant adienes to spport and implement ations that ontribte to isses of ommon interest

Epert noledge sstemati, domented and reieable proess eliitation EE to retriee epert dgements from a grop of eperts, often in the form of a probabilit distribtion

Eternal alidit he etent to hih the findings of a std an be generalised or etrapolated to the assessment estion at hand

oresight nforming poli b generating insights regarding the dnamis of hange, ftre hallenges and options that an be sed as an inpt to poli oneptalisation and design

ll sstemati reie tpe of literatre reie that ses sstemati methods to ollet seondar data, ritiall appraise researh stdies, and snthesize findings alitatiel or antitatiel

Hazard he intrinsi potential of an eent, sbstane et to ase harm

Horizon sanning he sstemati eamination of potential ftre problems, threats, opportnities and liel ftre deelopments

Heristis n approah to problem soling or selfdisoer that emplos a pratial method, not garanteed to be optimal, perfet, or rational, bt instead sffiient for reahing an immediate goal

nterdisiplinarit ilding lins beteen disiplines

nternal alidit nternal alidit refers to the degree of onfidene that the asal relationship being tested is trstorth and not inflened b other fators

sse adoate n siene adie, an agent representing and adoating speifi interests – eg the staeholders onslted hen deeloping the adie

sse bias a of eidene reation, seletion and interpretation hih is biased in faor of partilar soial or politial onerns, or priileges partilar sientifi methods, in nontransparent as

eptember

e ‘assessment’ refers to

e roems roem tat s fft or mosse to soe ease of nomete ontrator an ann rerements tat are often fft to reonse

etemer

etemer

etemer Getting in touch with the EU

IN PERSON All over the European Union there are hundreds of Europe Direct information centres. You can find the address of the centre nearest you at: https://europa.eu/european-union/contact_en

ON THE PHONE OR BY EMAIL Europe Direct is a service that answers your questions about the European Union. You can contact this service: – by freephone: 00 800 6 7 8 9 10 11 (certain operators may charge for these calls), – at the following standard number: +32 22999696 or – by email via: https://europa.eu/european-union/contact_en

Finding information about the EU

ONLINE Information about the European Union in all the official languages of the EU is available on the Europa website at: https://europa.eu/european-union/index_en

EU PUBLICATIONS You can download or order free and priced EU publications at: https://publications.europa.eu/en/publications. Multiple copies of free publications may be obtained by contacting Europe Direct or your local information centre (see https://europa.eu/european-union/contact_en)

EU LAW AND RELATED DOCUMENTS For access to legal information from the EU, including all EU law since 1952 in all the official language versions, go to EUR-Lex at: http://eur-lex.europa.eu

OPEN DATA FROM THE EU The EU Open Data Portal (http://data.europa.eu/euodp/en) provides access to datasets from the EU. Data can be downloaded and reused for free, for both commercial and non-commercial purposes. We live in a complex and ambiguous world, where trust in both science and politics can be put into question – where scientific evidence is dismissed and policy decisions contested on the assumption of partiality and private interests. Nevertheless, at present the role of science is crucial to inform policy-making in order to provide clarity and assess the consequences of policy options in a systematic way. In order to ensure that trust in science is maintained, science advice needs to be provided in an impartial, reliable, relevant and transparent way, following a set of principles and building on existing best practices.

This scientific opinion, informed by a scientific evidence review report by the Scientific Advice for Policy by European Academies (SAPEA) consortium, makes the following recommendations to ensure the quality and relevance of scientific advice: — Engage early and regularly with policy-makers and define together the boundaries of the advice, the question and its scope, as well as the best way to address it. Involvement of stakeholders or the public in the process can also be envisaged — Improve the quality of scientific advice by rigorous synthesis of existing evidence and transparent debate — Analyse and communicate uncertainty and diverging views related to the scientific evidence and the policy options recommended

Studies and reports