1. A closer look at risk perception and risk governance
Piet Sellke
Dialogik and University of Stuttgart
ICPS
5th Annual Regulatory Affairs International Symposium
March 11th, 2013

2. Outline Part I: Introduction
Part II: Risk Perception and Risk Assessment
Part III: Risk Governance: An Integrated Concept
Part IV: Conclusions

3. Part I: Introduction

4. Deficits in Risk Governance Accountability
Lack of Trust
Framing
Paralysis by
Analysis
Scope
Deficits in Risk Governance
Accountability
Scarcity of data
Inequity
Framing – different stakeholders have conflicting views of the issue
Scope – a risk perceived as only local may have global consequences (and vice versa)
There is a scarcity of data about the risk or people’s perceptions of it or, if data does exist, there is a failure to accept it
Transparency – trade-offs are not made explicit and hidden agendas seem to determine the outcome
Inequity – decisions allot the risk and benefits unfairly
Accountability – decision makers are isolated from the impact of their decision
Alienation – people or organisations are ignored (can lead to social mobilisation) (also “Authority knows best”)
Lack of trust in the process or the communication channel
“Paralysis by analysis” – overly inclusive process leads to inertia
Transparency

5. Policy Dilemma in Risk Management
Follow perceptions of the public?
Follow advice of professional experts?

6. Challenge of Risk Governance
 Integration of factual, technical knowledge and socio-cultural knowledge into one framework
Factual dimension
Physically measurable outcomes
Risk discussed as a combination of consequences and their probability of occurence
Socio-cultural dimension
How risks are viewed if values and emotions come into play

7. Part II: Risk perception and risk assessment

8. Risk perception (I/II)
Human behavior depends on perception, not on facts!
Qualitative Risk Characteristics:
Risk-related patterns
Perceived dread
Familiarity with the risk
Sensual perceptibility
Situation-related patterns
Personal controllability
Voluntariness
Trust in risk management
Fair distribution of gains and losses

9. Risk perception (II/II)
Semantic risk patterns
Risks posing an immediate threat (large dams, nuclear energy)
Risks dealt with as a blow of fate (natural disasters)
Risk as challenge to one‘s own strenght (sports)
Risk as a gamble (lotteries, stock exchange)
Risks as an early indication of insidious danger (food additives, viruses)
Stigmatisation of Risk
Social amplification of Risk
These patterns have functions similar to drawers in a filing cabinet. When faced with a new risk or when obtaining new information about a risk, most people try to file this new information into one of the existing drawers[1]. In addition to the cognitive processing of risk characteristics and risk situations, studies have shown that people tend to stigmatise risk sources that are associated with specific dreadful associations (Kunreuther and Heal 2003). A salient example of stigma is the reaction to products that are deemed to be carcinogenic, although there is often limited, if any, scientific evidence to support this position. The mere suspicion that a substance could cause cancer is often sufficient for generating fear and asking for strict regulatory actions. Stigmatisation leads to a cycle of public outrage and regulatory responses feeding into the process that has been described as social amplification of risk (Kasperson et al. 1988; Kasperson et al. 2003). Stimulated by media reporting, the public’s perception of the risk is often amplified in ways that are difficult to explain if one were focusing on the standard elements of any technical risk assessment – probability and direct losses.
The problems associated with risk perception are compounded because of the difficulty individuals have in interpreting low probabilities when making their decisions (Kunreuther et al. 2001). In fact, there is evidence that people may not even want data on the likelihood of an event occurring. If people do not think probabilistically, how do they make their choices? Psychological research has revealed the following patterns of drawing inferences about probabilities and risks
Social amplification of risk:
Social interactions can heighten or attenuate perceptions of risks
Risk behavior is in turn shaped by heightened or attenuated perceptions
Behavior patterns generate secondary consequences (e.g. loss of trust, drop in business sales, repercussions on other technologies

10. Challenges of Risk Assessment
Quality of explanatory power depends on the accuracy and validity of (real) predictions
Problems in validating risk assessment results
Prove of correctly assigned probabilities to a specific outcome difficult
Risks with difficult to discern cause-effect relationships
Risks with rare effects
Risks with effects difficult to interpret
Variations in both causes and effects obscuring the results
Risk assessment is confronted with three major challenges that can be best described using the terms ‘complexity’, ‘uncertainty’ and ‘ambiguity’. These three challenges are not related to the intrinsic characteristics of hazards or risks themselves but to the state and quality of knowledge available about both hazards and risks. Since risks are mental constructs, the quality of their explanatory power depends on the accuracy and validity of their (real) predictions. Unlike some other scientific constructs, validating the results of risk assessments is particularly difficult because, in theory, one would need to wait indefinitely to prove that the probabilities assigned to a specific outcome were correctly assessed. If the number of predicted events is frequent and the causal chain obvious (as is the case with car accidents), validation is relatively simple and straightforward. If, however, the assessment focuses on risks where cause-effect relationships are difficult to discern, effects are rare and difficult to interpret and variations in both causes and effects are obscuring the results, the validation of the assessment results becomes a major problem. In such instances, assessment procedures are needed to characterise the existing knowledge with respect to complexity, remaining uncertainties and ambiguities (WBGU 2000, 195ff.; Klinke and Renn 2002).

11. Integration of Risk Assessment and Risk Perception
Option 1 (only science):
Only scientific concepts of risk can claim inter-subjective validity and applicability
(Erroneous) risk perceptions have to be corrected via communication and education
Option 2 (only people):
No overarching universally applicable quality criterion available in order to evaluate the appropriatness and validty of risk concepts
Scientific concepts should compete with concepts of stakeholders and public groups
Loss of Public Support
Loss of Scientific Expertise

12. Integration of Risk Assessment and Risk Percpetion
Option 3 (science + people):
In identifying aspects of concern and worry, all groups of society have the same right to raise them and to bring them to the negotiation table
Question of the degree to which these concerns / worries are violated by the risk-bearing activity / events should be primarily answered by experts
Integrated Approach

13. Part II: Risk governance: An integrated concept
Taking Control in Pensions Planning 1999
3 May 1999
Part II: Risk governance: An integrated concept
Draft 1

14. CONVENTIONAL RISK MANAGEMENT
Deciding
Understanding
Who needs to know what, when?
The knowledge needed for judgements and decisions
Management
Communication
Appraisal
Who needs to do what, when?
Most risk management processes do not go beyond these steps

15. IRGC’s RISK GOVERNANCE FRAMEWORK
Risk assessment PLUS Concern assessment
Getting a broad picture of the risk
Deciding
Understanding
Pre-Assessment
Characterisation
and Evaluation
Appraisal
Management
Categorising the knowledge
about the risk
Who needs to do what, when?
Communication
Communication
Who needs to know what, when?
Is the risk tolerable, acceptable or unacceptable?
Is the risk simple, complex, uncertain or ambiguous?

16. Taking Control in Pensions Planning 1999
3 May 1999
Phase 1
Pre-Assessment
Characterisation
and Evaluation
Appraisal
Management
Communication
Getting a broad picture of the risk
Draft 1

17. Importance of Framing
Looks like a high risk from the outside

18. Importance of Framing
But consider this…

19. Importance of Framing
Or this…

20. Different regulatory approach
NOVELTY AND PRECAUTION: THE IMPACT OF FRAMING ON THE RISK-HANDLING OF GMOs
Comparing USA and Europe:
Different framing
Different regulatory approach
In the EU, GM crops were framed as a radical departure from any previous products and were seen as requiring path-breaking regulatory approaches.
The US, in line with the OECD approach, framed them as inherently similar to existing products developed through conventional plant breeding programmes and therefore not requiring any additional scrutiny beyond existing regulatory systems, for example for pesticides, food for human consumption or animal feeds (i.e. they were seen as requiring path-dependent and evolutionary regulation).

21. Taking Control in Pensions Planning 1999
3 May 1999
IMPORTANCE OF FRAMING
Frames represent social and cultural perspectives
Challenge or problem
Opportunity or risk
Innovation or intervention
Frames determine boundaries of what is included and excluded
Time and duration (future generations, sustainability)
Location and space (the universe, all nation, Norway, Stavanger)
Social class and stratus (vulnerable groups, poor, immigrants)
Types of adverse effects (physical, mental, social, cultural)
Primary or secondary impacts (ripple effects)
Criteria taken into account (risk reduction, cost, benefit, equity, environmental justice, value violations…)
Draft 1

22. Taking Control in Pensions Planning 1999
3 May 1999
Phase 2
Pre-Assessment
Is the risk simple, complex, uncertain or ambiguous?
Appraisal
Management
Communication
Characterisation
and Evaluation
Categorising the knowledge
about the risk
Draft 1

23. Components of Risk Knowledge
Complexity in assessing causal and temporal relationships
Uncertainty
variation among individual targets
measurement and inferential errors
genuine stochastic relationships
system boundaries and ignorance
Ambiguity
Interpretative ambiguity (What does it mean?)
Normative ambiguity (Is it tolerable?)

24. RISK APPRAISAL Risk Assessment Concern Assessment
Hazard identification and estimation
Exposure assessment
Risk estimation
Concern Assessment
Socio-economic impacts
Economic benefits
Public concerns (stakeholders and individuals)

25. BRENT SPAR – UNDERESTIMATING STAKHOLDER CONCERN
Greenpeace’s campaign included occupation of the platform but did not include calling for a consumer boycott. Nonetheless, Shell is estimated to have lost between £ million, mostly from lost sales across northern Europe; petrol stations were fire-bombed in Germany.

26. Taking Control in Pensions Planning 1999
3 May 1999
Phase 3
Pre-Assessment
Characterisation
and Evaluation
Appraisal
Management
Communication
Is the risk tolerable, acceptable or unacceptable?
Draft 1

27. Characterization and Evaluation
What are the broader, value-based questions to consider?
Characterization:
What are the societal and economic benefits and risks?
Are there impacts on individual or social quality of life?
Are there ethical issues to consider?
Is there a possibility of substitution?
Evaluation:
What are possible options for risk compensation or reduction?
How can we assign trade-offs between different risk categories and between risks and benefits (or opportunities)?
What are the societal values and norms for making judgements about tolerability and acceptability?
Do any stakeholders have commitments or other reasons for desiring a particular outcome of the risk governance process?

28. Evaluation and Characterisation
The scientific evidence from the risk assessment is “simple” – civil aviation emits significant pollutants into the atmosphere. Grounding the fleet would stop those emissions.
Modern values – particularly economic and societal – are such that no decision has been or is likely to be made to ban civil aviation.
In this instance, the policy judgement is therefore to give greater weight to the values than the scientific evidence.

29. Taking Control in Pensions Planning 1999
3 May 1999
Phase 4
Pre-Assessment
Characterisation
and Evaluation
Appraisal
Management
Communication
Who needs to do what, when?
Draft 1

30. RISK CHARACTERISTICS AND THEIR IMPLICATIONS FOR RISK MANAGEMENT (I/II)
Knowledge Characterisation
Management Strategy
Appropriate Instruments
Stakeholder Participation
1 ‘Simple’ risk problems
Routine-based:
(tolerability / acceptability judgement)
 Applying ‘traditional’ decision-making
Risk-benefit analysis
Risk-risk trade-offs
Instrumental discourse
(risk reduction)
Trial and error
Technical standards
Economic incentives
Education, labelling, information
Voluntary agreements
2 Complexity-induced risk problems
Risk-informed:
(risk agent and causal chain)
 Characterising available evidence
Expert consensus seeking tools, such as Delphi or consensus conferencing, meta analysis, scenario construction
Results fed into routine operation
Epistemological discourse
Robustness-focussed:
(risk absorbing system)
 Improving buffer capacity of risk target via:
Additional safety factors
Redundancy and diversity in designing safety devices
Improving coping capacity
Establishing high reliability organisations

31. RISK CHARACTERISTICS AND THEIR IMPLICATIONS FOR RISK MANAGEMENT (II/II)
Knowledge Characterisation
Management Strategy
Appropriate Instruments
Stakeholder Participation
3 Uncertainty-induced risk problems
Precaution-based:
(risk agent)
 Using hazard characteristics such as persistence, ubiquity etc. as proxies for risk estimates
Tools include: Containment, ALARA, BACT
Reflective discourse
Resilience-focussed:
(risk absorbing system)
 Improving capability to cope with surprises
Diversity of means to accomplish desired benefits
Avoiding high vulnerability
Allowing for flexible responses
Preparedness for adaptation
4 Ambiguity-induced risk problems
Discourse-based:
 Application of conflict resolution methods for reaching consensus or tolerance for risk evaluation results and management option selection
Integration of stakeholder involvement in reaching closure
Emphasis on communication and social discourse
Participative discourse

32. Taking Control in Pensions Planning 1999
3 May 1999
Complementary Phase
Pre-Assessment
Characterisation
and Evaluation
Appraisal
Management
Communication
Who needs to know what, when?
Draft 1

33. Objectives of Risk- Communication
Enlightenment: Making people able to understand risks and benefits (and their interactions)
Behavioral changes: Making people aware of potential risks and benefits help them to make the right choices
Trust building: Assisting risk management agencies to generate and sustain trust
Conflict resolution: Assisting risk managers to involve major stakeholders and affected parties to take part in the risk-benefit evaluation

34. Risk Communication – Essential throughout the process
Pre-assessment
Informing other agencies and assessing who is affected and who is mandated to take responsibility
Inviting views of affected stakeholders
Appraisal
Requesting and receiving appropriate scientific advice on the risk
Requesting and receiving scientific advice on people’s concerns
Evaluation
Communication of appraisal findings (if they are clear)
Involving all affected agencies and stakeholders if risk appraisal findings are uncertain or ambiguous
Deliberations concerning values / perspectives and to evaluate trade-offs
Management
Inclusion of appropriate stakeholders in the decision making process
Communication of the decision / regulation / advice

35. STAKEHOLDER INVOLVEMENT
Actors
« Civil society »
Affected stakeholders
Affected stakeholders
Scientists/ Researchers
Scientists/ Researchers
Scientists/ Researchers
Agency Staff
Agency Staff
Agency Staff
Agency Staff
Type of participation
Instrumental
Find the most cost-effective way to make the risk acceptable or tolerable
Epistemic
Use experts to find valid, reliable and relevant knowledge about the risk
Reflective
Involve all affected stakeholders to collectively decide best way forward
Participative
Include all actors so as to expose, accept, discuss and resolve differences
Dominant risk characteristic
Simple
Complexity
Uncertainty
Ambiguity

36. Taking Control in Pensions Planning 1999
3 May 1999
PART V: Conclusions
Draft 1

37. CONCLUSIONS I Problems in handling risks:
Part V: Conclusions
CONCLUSIONS I
Problems in handling risks:
Plural values and knowledge claims
Expert dissent on risk and benefits
Transboundary nature of risks
Social amplification and attenuation via perception and social mobilization
Emergence of systemic risk that cross national and sectoral boundaries (ripple effects)
Need for integration of risk analysis and perception
Communication must be tailored to the risk class

38. CONCLUSIONS II
Four risk management regimes should be used to deal with these new risk challenges:
simple risk management: standard risk assessments
risk-informed management: expanded risk assessments; seeking expert consensus and epistemological clarification
precaution-/resilience-based management: negotiated safety level under uncertainty; seeking stakeholder consensus and relying on containment and resilience
discourse-based management: value-based orientation; seeking more public input and stakeholder involvement for interpretative variability and normative controversy

39. Thank you!

40. RISK GOVERNANCE GOES MUCH FURTHER
Core Risk Governance Process
pre-assessment
risk appraisal -- risk assessment -- concern assessment
evaluation: tolerability / acceptability judgement
risk management
communication
Organisational Capacity
assets
skills
capabilities
Actor Network
politicians
regulators
industry/business
NGOs
media
public at large
Social Climate
trust in regulatory institutions
perceived authority of science
degree of civil society involvement
Political & Regulatory Culture
different regulatory styles

41. Adequate structure of working groups?
Local structures, leading structure?
Who is initiating a process as this?
What partners does one need?

42. New Challenge: Systemic Risks
Emergence of systemic risks that are...
transboundary
socially amplified via perception and social mobilisation
subject to expert dissent regarding risks and benefits
unmanageable by single organizations
difficult to communicate