Download D.4.3.2 - Report on approaches to the characterization

Front page for deliverables
Project no.
003956
Project acronym
NOMIRACLE
Project title
Novel Methods for Integrated Risk Assessment of
Cumulative Stressors in Europe
Instrument
IP
Thematic Priority
1.1.6.3, ‘Global Change and Ecosystems’
Topic VII.1.1.a, ‘Development of risk assessment
methodologies’
Deliverable reference number and title:
D.4.3.2., part DIA
Report on approaches to the characterization of knowledge of risks,
uncertainties and ambiguity and their use and quality assurance in the IP
domain
Due date of deliverable: 1 March 2006
Actual submission date: October , 2006
Start date of project: 1 November 2004
Duration: 5 years
Organisation name of lead contractor for this deliverable: DIA
Revision: 1
Project co-funded by the European Commission within the Sixth Framework Programme (2002-2006)
Dissemination Level
PU
PP
RE
CO
Public
Restricted to other programme participants (including the Commission Services)
Restricted to a group specified by the consortium (including the Commission Services)
Confidential, only for members of the consortium (including the Commission Services)
X
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
Authors and their organisation:
Ortwin Renn, Christina Benighaus DIA
Deliverable
Nature:
no:
R
D. 4.3.2, part
DIA
Status: submitted
Disseminatio
n level: PU
Date of delivery:
October, 2006
Date of publishing:
October, 2006
Reviewed by (period and name):
April-2006, Ad Ragas, RU
Framing the Perception of Cumulative Stressors
especially Chemical Risks
Ortwin Renn & Christina Benighaus, DIA, Stuttgart, Germany
ABSTRACT
Chemical substances have the potential to effect or harm human health as well as all parts of
the biosphere. At the same time, the chemical industry contributes to the well-being of
society, e.g. fighting against serious diseases, tackling agricultural problems or rising hygiene
levels.
Risks represent mental constructions about what people expect to happen as a result of human
activities or natural events and what they may experience based on observations of the past or
simulations of future events. The link between risk as a mental concept and reality is forged
through the experience of actual harm (the consequence of risk).
Risk perception, in general, denotes the processing of physical signals and/or information
about potentially harmful events or activities and the formation of a judgement about
seriousness, likelihood and acceptability of the respective event or activity. Signals refer to
direct observation by human senses; information refers to verbal and non-verbal exchange of
messages (communication with others).
This list of individual and social factors that shape risk perception demonstrates that the
intuitive understanding of risk is a multidimensional concept and cannot be reduced to the
product of probabilities and consequences. Risk perception studies have revealed the various
elements that shape the individual and social experience of risk.
Context and supporting circumstances are significant characteristics of risk perception. These
perception patterns are not just individual perceptions cobbled together: they originate from
cultural evolution, have been shaped by trial and error and have become and trusted concepts
in everyday life. Perceptions control our actions when facing decisions with uncertain
outcomes or being exposed to a dangerous situation.
2
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
Little is known about how society perceives combined effects of chemical stressors.
Preliminary empirical study indicates that the majority of respondents perceive the combined
effects of noxious substances in the environment as being additive or even super-additive.
There is no convincing concept up to now of how to use the results for risk perception studies
for risk management and regulation. But most analysts agree that data on perception contain
not only interesting insights about how people process complex and uncertain information but
also imply value for normative decision makers. Science may help to determine the
magnitude of the risk but the results of professional risk assessments alone are not sufficient
to make decisions about the acceptability of risks.
Risk communication is a necessary step towards bridging the gap between risk analysis and
perception. The goal of risk communication should not be to persuade people to accept
whatever the communicator thinks is best for them. The ideal communication program
envisions a receiver who processes all the available information to form a well-balanced
judgment in accordance with the factual evidence, the arguments of all sides, and his/her own
interests and preferences.
3
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
Table of Contents
1. Current Risk Debate
1.1 Chemical products in every day live
1.2 Risk in social context
1.3. The NoMiracle Context
2. Definition of risk and risk perception
3. Review on Risk Perception Research
3.1. Cognitive heuristics
3.2. Semantic Images
3.3. Psychometric Factors (Qualitative Characteristics)
3.4. Cultural approaches to risk perception
3.5 Reflexive modernization approach to risk perception
3.6 Summary: Approaches for perceptions of chemical risks
4. Central findings: What shapes risk perceptions?
4.1. Comparative empirical studies about risk perception
4.2 Institutional Trust and Confidence
5. Perception of combined cumulative risks, especially chemical risks
5.1 Survey in Baden-Württemberg
5.2 Perception of combined risks: interaction between environment and lifestyle
5.3 Interpretation of results
6. The Influence of the Media
7. An integrative model of risk perception
8. Conclusion
9. Bibliography
4
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
1. Current risk debate
1.1 Chemical products in every day live
Chemical products are in public use everywhere in the world. They are as much part of
ordinary life as automobiles or electronic devices. Many consumers use chemical products
without deeper knowledge about the risks for human health and the environment and impacts
on all aspects that human societies value (ethical implications). They may also not be aware of
how these products are manufactured and which chemical ingredients these contain. Due to
the variety of substances that the chemical industry has at its disposal, individual consumers
have difficulties to oversee the potential consequences to their health and the environment.
Chemical substances have the potential to effect or harm human health as well as all parts of
the biosphere. At the same time, the chemical industry contributes to the well-being of
society, e.g. fighting against serious diseases, tackling agricultural problems or rising hygiene
levels.
Although knowledge about risks and benefits of chemical products is fragmented at best,
people normally have strong beliefs about the positive and negative impacts associated with
these products. Many studies show that most people tend to associate negative images with
chemical substances rather then benefits (RENN & KASTENHOLZ 2000: 17; JAEGER
2002). There are many reasons for the revealed propensity of most people in industrial
countries to be critical of chemical products (cf. cultural differences in risk perception RENN
& ROHRMANN 2000). Chemical risks are difficult to detect by human senses, they are
perceived as artificial rather than natural, the negative side effects are regarded as hideous and
“creeping into the body” and chemistry is often associated with poison and non-organic
substances. These risks may also induce negative associations in combination with other
negative images pertaining to “big industry”, lack of personal control and missing
governmental oversight. Furthermore, most people are convinced that the long-term risks are
not known to society and that combined effects are hardly ever studied.
Talking about chemical risks and specifically about combined chemical effects faces the
danger that everybody talks about something different. In fact, the understanding of the
concept of risk differs widely across sciences and scientists (see DROTTZ-SJÖBERG 1991,
FISCHHOFF, WATSON & HOPE 1984, RENN 1992a, SHORT 1984, RENN &
ROHRMANN 2000). There is no commonly accepted definition for the term risk - neither in
public understanding nor in the sciences. Since risk refers to a potential of "real"
consequences, it is both a social construction and, in its consequences, a representation of
reality (RENN 1992a; RENN et al. 1992; ROSA 1998). This implies chemical risk is not a
physical entity but refers to a mental inference about implications of a current activity or
event in the future. Risks do not represent real phenomena but originate in the human mind
(IRGC 2005). Actors, however, creatively arrange and reassemble signals that they get from
the 'real world' providing structure and guidance to an ongoing process of reality enactment.
So risks represent mental constructions about what people expect in reality and what they may
experience based on observations of the past or simulations of future events. The link between
risk as a mental concept and reality is forged through the experience of actual harm (the
consequence of risk) in the sense that human lives are lost, health impacts can be observed,
and the environment is damaged or buildings collapse. The invention of risk as a hybrid
between a real phenomenon and mental construct is contingent on the belief that human
action can prevent harm in advance (assumption of human agency).
5
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
The status of risk as a mental construct to anticipate real consequences has major implications
on how risk is looked at. Unlike trees or houses, one cannot scan the environment for risks,
identify the objects of interest, and count them. Risks are created and selected by human
actors. What counts as a risk to someone may be an act of God to someone else or even an
opportunity for a third party. Although societies have over time gained experience and
collective knowledge of the potential impacts of events and activities, one cannot anticipate all
potential scenarios and be worried about all the many potential consequences of a proposed
activity or an expected event. By the same token, it is impossible to include all possible
options for intervention. Therefore societies have been selective in what they have chosen to
be worth considering and what to ignore (THOMPSON et al. 1990; DOUGLAS 1990; BECK
1994: 9 et seqq.). Specialised organisations have been established to monitor the environment
for hints of future problems and to provide early warning of some potential future harm. This
selection process is not arbitrary. It is guided by cultural values (such as the shared conviction
that each individual life is worth protecting), by institutional and financial resources.
Particularly, technology-induced hazards get substantial publicity and are widely discussed in
society. Different assessment and management approaches exist and are being discussed in the
science community for dealing with these risks in a broader and more inclusive sense (IRGC
2005; IAEA 1995; IEC 1993; IPCS 2004; NRC 1983). However, the understanding of
chemical risk perception and its relation to combined effects and is not yet sufficiently
investigated. More research is needed to advance the theoretical foundation of the field as well
as the empirical knowledge.
1.2 Risk in social context
Risk issues create controversies. In many societies, severe conflicts about the evaluation of
risks have emerged, especially with respect to large-scale technologies such as chemical
industries and chemical products, nuclear energy and genetic engineering (BECHMANN
1993, BECK 1986/1992, JUNGERMANN, ROHRMANN & WIEDEMANN 1991,
SJÖBERG 1987, v. WINTERFELDT & EDWARDS 1984). Two concerns are often raised:
risks to human health and well-being and risks to the state of the environment, emerging from
the impact of human activities. In addition, social issues such as equitable risk burden sharing,
environmental justice, or the disturbance of social cohesion.
Furthermore, there is a considerable gap between how experts estimate risks, and how nonprofessional people judge and evaluate risks. Depending on the underlying definitions and
criteria, very heterogeneous risk evaluations are given by different groups (EDWARDS & v.
WINTERFELDT 1987). Many of the risks most prominent in the view of the public are not
those which - according to statistical data - actually result in high accident figures, mortality
rates, health impacts, and so on. Many surveys and psychological experiments have shown
that the perceived seriousness of risks was almost diametrically opposed to the calculated risk
numbers of the risk professionals (COVELLO 1983, SLOVIC 1987, RENN 1990, PIDGEON
et al. 1992). Juxtaposing professional estimates of risks and public perceptions of risk has
been a popular activity among risk researchers ever since the EPA published its "Unfinished
Business" report (US-Environmental Protection Agency 1987, 1990). Most of these studies
confirm a clear discrepancy between the priority list of experts and the mean values of
concerns among the general public. Apparently many more factors than those data influence
6
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
risk perception, risk behaviour and risk management, including a variety of social,
psychological, ethical aspects (O'RIORDAN 1983). Furthermore, cultural factors are crucial
for people's risk evaluation. In fact, "each society has its own distinctive portfolio of risks that
are believed to be of concern" (JOHNSON & COVELLO 1987).Under these circumstances,
chemical risk analysis in the natural sciences or economics is increasingly complemented by
social-science approaches in order to expand how "risk" is conceptualized, and to understand
the way in which humans experience, rate and evaluate risks they are (or might be) exposed
to.
1.3. The NoMiracle Context
This report is worked out as a deliverable in work package 4.3. undertaken by DIALOGIK as
one of the consortium partners in the integrated project1. It is the objective of this part within
the overall NoMiracle project is to integrate a social science perspective in the mainly
technical-economic assessment of risks. The current outcome refers to two of in total seven
purposes of the whole NoMiracle project:
 To quantify, characterise and reduce uncertainty in current risk assessment
methodologies, e.g. by improvement of the scientific basis for setting safety factors
 To improve the provisions for the application of the precautionary principle and to
promote its operational integration with evidence-based assessment methodologies.
This article reviews approaches and models from the social sciences related to risk perception.
It investigates which of the core findings are transferable to cumulative risks especially
chemical mixtures. The paper is guided by the following questions:




Which are the most important approaches of risk perception that fit best the
phenomenon of cumulative stressors?
How can one classify risks arising from the production or distribution of chemicals?
Does risk perception differ from country to country?
What factors influence risk perception (of chemicals risks)?
This report reviews the scientific literature on risk perception with a special focus on chemical
risks. The paper starts with a general introduction about chemical products and risk in social
context (section 1), and it subsequently outlines the concepts of risk and risk perception
(section 2). The third section provides an overview of different approaches to risk perception.
The fourth section summarizes the findings of selected empirical studies and it discusses the
impact of some factors (e.g. institutional trust, risk acceptance and value commitments) on
risk perception. Section 5 discusses the perception of combined chemical risks based on an
empirical study performed in Germany. The influence of the media on risk perception is
discussed in Section 6, and Section 7 outlines an integrative model of risk perception. Finally,
section 8 lists the conclusions.
1
Parts of this report will be published in scientific journals later after some more analysis is completed. This
Report might help to foster the theoretical discussion about risk perception of cumulative stressors among the
various disciplines and areas of expertise represented in NoMiracle, an EU-financed integrated project (European
Commission, FP6 Contract No. 003956).
7
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
2. Definition of risk and risk perception
Within the social sciences, people's evaluations of risks are usually labelled risk perception.
Principally, chemical risks cannot be "perceived" in the sense that they are taken up by the
human senses and processed. However, this terminology has become the standard convention
in the social scientific communities. Risk perception, in general, denotes the processing of
physical signals and/or information about potentially harmful events or activities and the
formation of a judgement about seriousness, likelihood and acceptability of the respective
event or activity. Signals refer to direct observation by human senses, information refers to
verbal and non-verbal exchange of messages (communication with others).Judgements about
seriousness, likelihood and acceptability are governed by internal and external factors which
will be described later in this section. For the purpose of this paper, we define “chemical risk
perception” as individual and social processing of signals and information as well as the
formation of resulting judgments about chemical hazards. Both experiences and beliefs are
part of people’s perceptions. It appears appropriate to juxtapose results from Quantitative Risk
Assessments as, for example "statistical", "probabilistic" or "predicted" risk estimates- to
those estimates of risks that non-professional laypersons associate with the risk in question.
The first estimates are called risk assessment results, the second perceived risk.
Psychological research has revealed different meanings of risk depending on the context in
which the term is used. Whereas in the technical sciences the term risk denotes the probability
of adverse effects, the everyday use of risk has different connotations. In most social contexts,
"risk" refers to the likelihood of an adverse effect resulting from an event or an activity, rather
than an opportunity for desired outcomes. Hence, risks are defined in this paper as the
possibility that human actions or events lead to consequences that affect aspects of what
human’s value. Then "chemical risk" can be defined as the possibility of physical, social
and/or financial harm/detriment/loss due to an exposure to a chemical hazard within a
particular time frame. "Chemical Hazard" refers to a situation, event or chemical substance
that has the potential to harm people, nature, capital or human-made facilities. Hazards differ
in terms of sources and impacts, principal disparities are to be expected for both perceived risk
level and acceptance of risks.
After defining the term risk, it is important to shed some more light on activities that are
associated with the analysis of chemical risk from the social science perspective. How people
think about the seriousness and acceptability of risks and how they make their respective
judgments is influenced by knowledge, values and feelings. The mental models and other
psychological mechanisms which people use (e.g., cognitive heuristics; risk images) are
internalized through social and cultural learning and constantly moderated (reinforced,
modified, amplified or attenuated) by media reports, peer influences and other communication
processes. Technical and solely quantitative approaches for characterizing risks are obviously
inadequate to reflect the complex pattern of individual risk perception. This has been stressed
by many authors, see, e.g., COVELLO 1983, FISCHHOFF et al. 1984, FREUDENBURG
1988, HEIMER 1988, JUNGERMANN & SLOVIC 1993, O'RIORDAN 1983, RAYNOR
1987, RENN 1993, ROHRMANN 1995, SJÖBERG 1987, SLOVIC 1991, SLOVIC et al.
1982, v. WINTERFELDT & EDWARDS 1984.
8
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
3. Review on Risk Perception Research
In total, five approaches of risk perception discussed in literature are summarized and
interpreted here, which gives the reader a quick overview of different “schools of thought”
and how these could be interpreted for chemical risks.
3.1. Cognitive heuristics
The term perception as used in cognitive psychology applies to the mental processes through
which a person takes in, deals with and assesses information from the environment via the
senses (JUNGERMANN & SLOVIC 1993a). Perception as an intuitive risk process is based
on how information of risk is mentally processed. This mental process results in perceived
risk – a collection of notions that people form on risk sources relative to the information
available to them and their basic common sense (RENN 1989). The two major aspects are
information processing and formation of judgements or attitudes.
Most risks that modern society faces are not experienced by human senses but learned
through communication. Rarely do we face disasters personally; however, the media provide
us ample information about hazardous events wherever they take place. The dangers of
technologies or nature, the risks of food additives or chemicals in drinking water, the threat of
nuclear disaster or a chemical explosion would probably never reach public attention unless
society communicates about these adverse possibilities. Risk perception is less a product of
experience or personal evidence than a result of social communication (LUHMANN 1986).
This observation has major consequences: Today’s society provides an abundance of
information, much more than any individual can digest. It is assumed that the average person
is exposed to 7,000 bits of information each day of which s/he perceives around 700,
acknowledges 70, stores seven in the short term memory and may remember less than one in
the longer term (CONARD 2005, p. 4; cf. also COVELLO 1983). Most information to which
the average person is exposed will be ignored. This is not a malicious act but a sheer necessity
in order to reduce the amount of information a person can process in a given time. Human
evolution has provided us with an almost automated and often subconscious tool of selecting
the important information from the abundance of information supplies.
The attention and selection process is not random although random elements may play a role.
People have developed special strategies to select information that they feel is relevant to
them. This is also true for risk information. The major criteria for selection are ability and
motivation (CHAIKEN and STANGOR1987). Ability refers to the physical possibility that
the receiver can follow the message without distraction; motivation to the readiness and
interest of the receiver to process messages. The conditions for both, ability and motivation
are listed in Table 1.
9
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
TABLE 1:
CONDITIONS AND REQUIREMENTS FOR INFORMATION SELECTION
Conditions
Elements of Conditions
Ability
-
Physical access to information
Time to process information
Absence of sources of distraction
Motivation
-
Reference to personal interests, salient values,
or self-esteem
Inducement of personal involvement with issue, the
content, or the source
-
Three conditions have to be met to satisfy the criterion of ability: the information has to be
accessible, the receiver must have the time to process the information, and other sources of
distraction should be absent. Several factors influence the motivation of a receiver to process
actively the information. The information content has to be relevant (referring to personal
interests, salient values, or self-esteem) and it should trigger personal involvement (with the
issue, the content, or the source). Both motivational factors are reinforced if the receiver has
some prior knowledge or interest in the subject or is in need for new arguments to back up
his/her point of view.
If both criteria are met, the individual is ready to absorb the information. Having gained the
receiver's attention does not determine the process, however, by which the information is
acknowledged and further evaluated. A complex procedure of information selection and
processing takes place after the initial attention drawing stimulus. This procedure can be
described in several steps of information processing. Distinguishing concrete steps of
information processing is an analytical tool since in reality individuals perform these steps
simultaneously. After information has passed the initial selections filters, people draw
inferences from this information, compare the content of the information with previously held
beliefs or memories, evaluate the significance, truthfulness and personal relevance of the
information, construct new beliefs (or more often reassure old beliefs) and form an opinion
(initial mental response) or an attitude (enduring mental response) to the object or activity to
which the information provided new insights. These steps are summarized in the following
table (Table 2):
10
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
TABLE 2: STEPS OF INDIVIDUAL INFORMATION PROCESSING
Steps
Description
Passing of attention filters
to select and further process signals coming from the
environment or from other social actors
Decoding of signals
to decipher the meaning of the signals (investigating
factual content, sources of information, value
statements, overt and hidden intentions of information
sources and transmitters)
Drawing one's own
inferences
to come to conclusions about the allegedly revealed
intentions of the source and the transmitter, to employ
intuitive heuristics (common sense reasoning) for
generalizing the information received and to use
symbolic cues for judging the seriousness of the
information
Comparing the decoded
message with encoded
experiences
to analyze the meaning of the message in the light of related
messages from other sources or previous attitudes and beliefs
stored in memory
Evaluating messages
to rate the importance, persuasiveness and potential for
personal involvement on the basis of the perceived
accuracy of the message, the potential effect on one's
personal life, the perceived consistency with existing
beliefs (to avoid cognitive dissonance), reference group
judgments (to avoid social alienation), and personal
value commitments.
Forming specific beliefs
to generate or change beliefs about the subject of the
message or to reassure previously held beliefs
Propensity to take
corresponding actions
to generate intentions for future actions that are in
accordance with the newly formed beliefs
In order to economize information processing, individuals are likely to evaluate whether it is
necessary to study the content of the information in detail or to make a fast judgment
according to some salient cues in the message received. The first strategy refers to the central
route of information processing, the second to the peripheral route (PETTY & CACIOPPO
1986; RENN & LEVINE 1991). The central route is taken when the receiver is so highly
motivated by the message that s/he studies each argument carefully. The peripheral route is
taken when the receiver is less inclined to deal with each argument, but forms an opinion or
even an attitude on the basis of simple cues and heuristics.
In the central mode, the receiver performs two types of evaluations: first, an assessment of the
probability that each argument is true; and second, and an assignment of weight to each
argument according to the personal salience of the argument's content. The credibility of each
11
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
argument can be tested by referring to personal experience, knowledge, plausibility, and
perceived motives of the communicator. The major incentives for changing an attitude in the
central mode are the proximity with and the affinity to one' own interests, values, and world
views. In the peripheral mode, receivers do not bother to deal with each argument separately,
but look for easily accessible clues to make their judgment on the whole package. Examples
of such cues are the length of a message, the number of arguments, the package (colour,
paper, graphic appeal, and others), and the presence of symbolic signals that trigger
immediate emotional responses (cf. KASPERSON et al. 1988).
Once information has been received, common sense mechanisms process the information and
help the receiver to draw inferences. These processes are called intuitive heuristics. They are
particularly important for risk perception since they relate to the mechanisms of processing
probabilistic information. Early psychological studies focused on personal preferences for
probabilities and attempted to explain why individuals do not base their risk judgments on
expected values, i.e. the product of probability and magnitude of an adverse effect
(POLLATSEK & TVERSKY 1970; LOPES 1983). One of the interesting results of these
investigations was the discovery of systematic patterns of probabilistic reasoning that are well
suited for most everyday situations. People are risk averse if the stakes of losses are high and
risk prone if the stakes for gains are high (KAHNEMANN & TVERSKY 1979). Many people
balance their risk taking behaviour by pursuing an optimal risk strategy which does not
maximize their benefits but assures a satisfactory payoff and the avoidance of major disasters
(LUCE & WEBER 1986).
Another important rule of thumb is the overrating of exposure and hazard rather then
probability of harm (RENN et al. 1991). This intuitive heuristic is probably the most powerful
factor for rejecting or downplaying information on chemical risks. If any exposure above zero
or above a defined threshold (including a safety factor for taking variability into account) is
regarded as negative, the simple and intuitively reasonable rule to minimize exposure makes
perfect sense. Most regulatory regimes are based on this simple rule ranging from the
ALARA principle to the application of the best available control technology (BACT). Such
principles imply that any exposure might be negative so that avoidance is the most prudent
reaction.
Another application of such a strategy is to use combinations of risks that are able to
compensate for each other’s losses. This specific deviation from maximizing expected utilities
has been adopted as a normative guideline by portfolio theory used by investors in the stock
market. According to this theory, investors should select a portfolio of stocks in which the
risks of losing money on one share is correlated with the probability of gaining money for
another share (mathematically through co-variance analysis). This example and many other
show that deviations from the rule of maximizing one’s utility are not so often a product of
ignorance or irrationality but an indication of one or several intervening context variables that
often make perfect sense if seen in the light of the context and the individual decision maker’s
values (LEE 1981; BREHMER 1987).
Second, more specific studies on the perception of probabilities in decision making identified
several biases in people’s ability to draw inferences from probabilistic information
(FESTINGER 1957; KAHNEMANN & TVERSKY 1979; RENN 1990; ROSS 1977). These
biases are summarized in the following table (Table 3):
12
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
TABLE 3: INTUITIVE BIASES OF RISK PERCEPTION
BIASES
DESCRIPTION
Availability:
Events that come to people's mind immediately are rated as more
probable than events that are less mentally available.
Anchoring effect:
Probabilities are estimated according to the plausibility of contextual links
between the cause and the effect but not on knowledge about distributions
(People will anchor the information which is significant and available for
them).
Representativeness:
Singular events experienced in person or associated with properties of
an event are regarded as more typical than information based on
frequencies.
Avoidance of
cognitive dissonance
Information that challenges perceived probabilities that are already part
of a belief system will either be ignored or downplayed.
Although these biases constitute clear violations of logical rules, they might have been
overrated in the literature (FISCHHOFF et al. 1981). Many laboratory situations provide
insufficient contextual information to provide enough cues for people on which they can base
their judgments (LOPES 1983). Relying on predominantly numerical information and being
unfamiliar with the subject, many subjects in these experiments retrieve to “rules of thumb” in
drawing inferences. In many real life situations, experience of and familiarity with the context
provide additional information to calibrate individual judgments, particularly for nontrivial
decisions (cf. HEIMER 1988). Nevertheless, risk managers should be aware of these biases
because they are found in public perception and may be one of the underlying causes for the
discrepancy between layperson judgement and professional assessments.
3.2 Semantic Images
Psychological research has revealed different meanings of risk depending on the context in
which the term is used. Whereas in the technical sciences the term risk denotes the probability
of adverse effects, the everyday use of risk has different connotations. With respect to humaninduced risks Table 4 illustrates the main semantic images.
13
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
TABLE 4: THE FOUR SEMANTIC IMAGES OF RISK IN PUBLIC PERCEPTION
(STREFFER et al. 2003)
1.
Pending Danger
• artificial risk source
• large catastrophic potential
• inequitable risk-benefit distribution
• perception of randomness as a threat
2.
Slow Agents
• (artificial) ingredients in food, water, or air
• delayed effects; non-catastrophic
• contingent on information rather than experience
• quest for deterministic risk management
• strong incentive for blame
3.
Cost-benefit Ratio
• confined to monetary gains and losses
• orientation towards variance of distribution rather
than expected value
• asymmetry between risks and gains
• dominance of probabilistic thinking
4.
Avocational Thrill
• personal control over degree of risk
• personal skills necessary to master danger
• voluntary activity
• non-catastrophic consequences
The semantic images allow individuals to order risks and risk sources on the basis of a few
salient characteristics. Reducing complexity by creating classes of similar phenomena is
certainly a major strategy for coping with information overload and uncertainty. The four
semantic images are powerful guides to help individuals to navigate through an abundance of
often contradicting information and provide a rather efficient method to balance the time for
collecting and processing information with the personal need for orientation and attitude
formation. The following list describes the conditions for placing risks in one of the four
image boxes and the consequences with respect to perception and evaluation.
 Risk as a pending danger (Damocles sword): Risk is seen as a random threat that can
trigger a disaster without prior notice and without sufficient time to cope with the hazard
involved. This image is linked to artificial risk sources with large catastrophic potential.
The magnitude of the probability is not considered. It is rather the randomness itself that
evokes fear and avoidance responses. Natural disasters, in contrast are perceived as
regularly occurring and thus predictable or related to a special pattern of occurrence
(causal, temporal or magic). The image of pending danger is therefore particularly
prevalent in the perception of large-scale technologies. People respond to these risks by
looking at the catastrophic potential and the possibilities to cope with the consequences of
a disaster once the unlikely event of an accident occurs. If the catastrophic potential is
high and people perceive hardly any possibility to avoid the consequences of the disaster
14
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
by taking protective actions (such as shelter or evacuation), a negative evaluation of the
risk is almost certain. The probability of an accident is irrelevant for this evaluation.
 Slow agents (Pandora’s Box): Risk is seen as an invisible threat to one’s health or wellbeing. Effects are usually delayed and affect only few people at the same time. Knowledge
about these risks is based on information by others rather than on personal experience.
These risks pose a major demand for trustworthiness in those institutions that provide
information and manage the hazard. If trust is lost, people demand immediate actions and
assign blame to these institutions even if risks are very small. Typical examples of this
risk class are food additives, pesticides, and chemicals in drinking water. Most individuals
develop a "zero-risk-mentality" in responding to slow agents. The threat of being exposed
to risks without sensing the danger or being aware of the creeping hazard overshadows
and even impedes the alternative response of balancing risks and benefits. Most people are
only willing to accept risks if they face the danger and willingly admit to exposing
themselves to the concomitant risk.
 Cost-benefit ratio (Athena’s Scale): Risks are perceived as a balancing of gains and losses.
However, this image is only used in peoples’ perceptions of monetary gains and losses.
Typical examples are betting and gambling both of which require sophisticated
probabilistic reasoning. People are normally able to perform such probabilistic reasoning
but only in the context of gambling, lotteries, financial investment, and insurance. Horse
betting is an excellent example about the degree of sophistication in probabilistic
reasoning that average people master in many countries. The common prejudice of many
technical experts that laypeople are unable to handle risk information and to understand
probabilities is certainly not true in this context. However, some of the biases of
processing uncertainty apply to the perception of financial risks, too. Laboratory
experiments show that people orient their judgment about lotteries more towards the
variance of losses and gains than towards the expected value (POLLATSEK &
TVERSKY 1970). Prospect theory is the mathematical expression of individual
preferences for risk aversion and gain proneness (KAHNEMANN & TVERSKY 1979).
Such deviations from the expected value model (probability times magnitude of damage)
can be a very rational response as explained earlier. If people perceive a risk as falling in
this category, they normally apply a risk-benefit balancing model based on their subjective
values and accept or even seek the risk if this balance turns out to be positive. Potential
conflicts arise when benefits and risks are distributed among different social groups. The
experience of inequities is one of the major driving forces of social responses to risk
situations and a lead to a public debate about fairness and social justice (Mac Lean 1986;
Young 1993).
 Avocational thrill (Hercules Image): Often risks are actively explored and desired
(MACHLIS & ROSA 1990). These risks include all activities for which personal skills are
necessary to master the dangerous situation. The thrill is derived from the enjoyment of
having control over one’s environment or oneself. Such risks are always voluntary and
allow personal control over the degree of risking. People often resist attempts by
governments to regulate these risks even if they pose major threats to those who are
exposed. As long as these risks can be individualized and do not pose major threats to
third parties (such as passive smoking), most people feel that regulatory actions are not
necessary and pose a potential threat to personal freedom. The degree to which external
effects are perceived (and sometimes ignored) and the relative importance of freedom vis15
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
à-vis regulatory actions depend, of course, on personal values and cultural traditions
(VOGEL 1986).
Some risk sources evoke more than one image. Particularly interesting are combinations of
the slow agents and pending danger images. For example, nuclear power is associated with
large-scale accidents and slow deterioration of human health induced by low-dose radiation
from normal operation. The same applies to many of the climatic risks. Such combinations
can act as risk amplifiers in public perceptions leading to strong personal as well as social
reactions (KASPERSON et al. 1988).
Risks from chemical products are mostly to be found in the category of slow agents. This has
far-reaching implications. Most agents belonging to this category are regarded as potentially
harmful substances that defy human senses and "poison" people without their knowledge.
Risks associated with food additives, air pollutants, water impurities, and other chemical
agents are mostly invisible to the person exposed. They require warning by regulators or
scientists. Food additives, chemicals or pharmaceuticals are always associated with negative
side effects. Along with that image people tend to believe that toxicity depends less on the
dose than on the characteristics of the substance. Hence they demand a deterministic
regulatory approach when it comes to controlling chemicals in the environment.
Many surveys show that people demand zero-risk-levels for slow agents in the chemical
sector, at least as the ideal target line. Chemical risks in particular when their hazard profile is
characterized by high ubiquity, high persistency and high irreversibility, trigger responses of
avoidance and desires for strict regulatory prohibitions. Something that is regarded as truly
bad and vicious is almost impossible to link with a positive connotation. The only exception
may be the exposure to "natural" agents. Most people believe that anything which exists in
nature cannot be harmful for people if consumed in modest amounts. That is why allegedly
“natural” drugs are associated with fewer or even none negative side effects compared to
chemical drugs. The perceptions of natural toxins as benign reflect the modern impression or
myth of "Mother Nature" who offers an invaluable set of beneficial resources to humankind in
response for taking good care of her. Chemical compounds, however, are associated with
artificiality and seen as threats to human health.
3.3 Psychometric Factors (Qualitative Characteristics)
In addition to the images that are linked to different risk contexts, the type of risk involved
and its situational characteristics shape individual risk estimations and evaluations.
Psychometric methods have been employed to explore these qualitative characteristics of
risks. This approach to risk research was developed by the "Oregon Group" B. FISCHHOFF,
S. LICHTENSTEIN and P. SLOVIC, (see SLOVIC, FISCHHOFF & LICHTENSTEIN 1980;
cf. also SLOVIC 1992). A number of researchers followed their approach, most of them in the
USA and European countries. For a comprehensive review and documentation of this body of
research see ROHRMANN 1995; overviews are provided by JUNGERMANN & SLOVIC
1993, PIDGEON, HOOD, JONES, TURNER & GIBSON 1992, RENN 1986, 1990;
methodological issues are discussed in ROHRMANN 1995, SLOVIC 1992.
The psychometric approach is based on four characteristics:
 to establish "risk" as a subjective concept, not an objective entity,
16
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
 to include technical/physical and social/psychological aspects in risk criteria,
 to accept opinions of "the public" (i.e., lay people, not experts) as the matter of interest,
 to analyse the cognitive structure of risk judgments, usually employing multivariate
statistical procedures such as factor analysis, multi-dimensional scaling or multiple
regression.
Table 5 lists the major qualitative characteristics and their influence on risk perception.
TABLE 5: LIST OF IMPORTANT QUALITATIVE RISK CHARACTERISTICS
Qualitative Characteristics
1.
Personal control
2.
3.
4.
5.
6.
7.
8.
Direction of Influence
increases risk tolerance
depends on confidence in institutional
Institutional control
performance
Voluntariness
increases risk tolerance
Familiarity
increases risk tolerance
Dread
decreases risk tolerance
Inequitable distribution of risks and depends on individual utility, strong social
benefits
incentive for rejecting risks
Artificiality of risk source
amplifies attention to risk, often decreases
risk tolerance
Blame
increases quest for social and political
responses
Furthermore, the perception of risk is often part of an attitude that a person holds about the
cause of the risk, i.e. industrial activity, consumption of food, production method (such as
genetic engineering) and others. Attitudes encompass a series of beliefs about the nature,
consequences, history, and justifiability of a risk cause. Due to the tendency to avoid
cognitive dissonance, i.e. emotional stress caused by conflicting beliefs, most people are
inclined to perceive risks as more serious and threatening if the other beliefs contain negative
connotations and vice versa. Often risk perception is the result of these underlying beliefs
rather than the cause for these beliefs.
With respect to the qualitative characteristics, one would expect that chemical products are
associated with many of the negative qualitative characteristics. First, most chemicals are
associated with negative risk characteristics such as dread, lack of personal control, and
artificiality. The perception of health risks induced by chemicals is usually linked to an
absence of personal control and the preponderance of dread thus amplifying the impression of
seriousness. These characteristics make people even more concerned about the negative
impacts than warranted by the predicted health effects alone. Second, the beliefs associated
with the risk source, for example industry, center around greed, profit-seeking and alleged
disrespect for public health. Third, the possibility of consumers being exposed to risks without
their consent touches upon serious equity concerns if susceptibility to these risks vary
considerably among individuals or rest on probabilistic balancing. For chemical risks factor 6
“Inequitable distribution of risks and benefits” also plays an important role. The more risks
are seen as unfair for the exposed population, the more they are judged as severe and
unacceptable (KASPERSON & KASPERSON 1983; SHORT 1984).
17
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
The importance of these qualitative factors in risk assessment offers a plausible explanation
for the fact that precisely those risk sources that technical risk assessment classifies as
particularly low-risk are the source of greatest concern among the general public. Risk sources
that are deemed controversial, like phthalates, are very often burdened with negative attributes
while leisure activities are associated with more positive ones (JUNGERMANN/SLOVIC
1993).
Psychological research on risk perception brings us a step closer towards an analysis of how
most people assess risk. The observed discrepancy between the results of technical risk
assessment conducted by experts and intuitive assessment of the same risk by society is not,
in the first instance, due to ignorance about the statistically derived expected values or an
expression of erratic thought processes, but rather an indication of a multidimensional
assessment process in which anticipated harm is only one among many factors.
One of the interesting points with respect to the perception of uncertainty is that the
professional distinction in probability distribution and different degrees of remaining
uncertainties (expressed in confidence intervals or in other forms of uncertainty
characterisation) is not echoed in risk perception studies (SPARKS, P. and SHEPHERD, R.
1994). There is a basic understanding among most people that the preferred dichotomy of
judging a situation either as safe or unsafe needs to be replaced by a mental model that
differentiates degrees of certainty. The open space between safe and unsafe, however, is
perceived as an one-dimensional indication of knowledge gaps and less as an indication for
(genuine) probability distributions. The more people associate uncertainties with a specific
risk the more they believe that society needs more science and research to reduce these
uncertainties. Very often they also require that the risk is not taken unless these knowledge
gaps are being bridged and a high degree of safety is accomplished through knowledge
improvement (FREWER, L.J., MILES, S., BRENNAN, M., KUSENOF, S., NESS, M., and
RITSON, C. 2002).
Studies conducted on an international scale also show that people everywhere, regardless of
their social or cultural background, use practically universal risk perception criteria in
forming their opinions (ROHRMANN & RENN 2000). However, the relative effectiveness of
these criteria in opinion-forming and judgments about risk tolerance varies considerably
between different social groups and cultures. While the above-mentioned qualitative
characteristics are accepted (often subconsciously) as assessment values for perceived risk,
their relative contribution to a person’s actual opinion or motivation to take action depends on
individual lifestyles, threatening environmental factors and ingrained cultural values
(SJÖBERG 1997). In assessing risk, people who favour alternative lifestyles tend more than
others to consider both “reversibility of the consequences of risk” and “congruence between
risk bearers and benefactors”, while those with strong material values assess risk more by way
of personal control opportunities and trust in institutional risk control (BUSS & CRAIK
1983).
As OTWAY and THOMAS impressively showed in their studies on attitudes to nuclear
energy, there is a high correlation between different value models and the relative importance
that people attach to either the benefits or the risks of a specific technology (OTWAY &
THOMAS 1982). The conclusion to be drawn from this is that value expectations and cultural
background are significant determinants of subjective risk that do not add to the semantic and
qualitative factors already described but, in effect, presuppose the existence of those factors in
18
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
that they use them as heuristics to incorporate and process information on complex properties
associated with the risk in question. Internalized value expectations and external
circumstances can control the relative effectiveness of intuitive perception processes, but not
their existence. This is not a matter of academic hair-splitting: it has direct relevance to
communication and conflict management. If we assume that intuitive mechanisms of risk
perception and assessment bear practically universal characteristics that can more or less be
reshaped by socio-cultural influences, then they can provide a fundamental basis for
communication of which one can avail oneself regardless of differences between the various
standpoints. In addition to the pool of common symbols and rituals (shared meaning), whose
importance to social integration is in constant decline in pluralistic societies, a new pool of
common mechanisms of risk perception emerges that along with common sense signals the
existence of supra-individual perception mechanisms.
3.4 Cultural approaches to risk perception
The psychological and qualitative approaches primarily address individual risk perception
patterns based on individual or group responses to qualitative factors of risk. However, the
process of risk perception in society has also been thoroughly analysed from a 'macrosociological' perspective as well - see for example BECK 1986, 1992, DOUGLAS &
WILDAVSKY 1982, HEIMER 1988, LUHMANN 1990, 1993, RENN 1992, THOMPSON
1980, and WILDAVSKY 1988, 1993.
Sociologists have particularly stressed that the evaluative process of risk perception is
determined by the norms, value systems and cultural idiosyncrasies of societies (see, e.g.,
BECK 1992, DOUGLAS & WILDAVSKY 1982, SJÖBERG 1995, THOMPSON, ELLIS &
WILDAVSKY 1990). According to the "cultural approach", risk is a "social and cultural
construction" (JOHNSON & COVELLO 1987) - not an 'objective' entity to be measured
independently of the context in which hazards occur. Consequently each society or social
group may have its own specific set of risks believed to be of concern, based on its prevalent
"worldview". Within this area, research has been more often qualitative than quantitative,
including philosophical treatises on risk perception.
A group of distinguished anthropologists and cultural sociologists such as Aaron Wildavsky,
Mary Douglas or Michael Thompson, have identified four or five patterns of value clusters
that separate different groups in society from each other (DOUGLAS & WILDAVSKY 1982;
RAYNER 1990; THOMPSON et al. 1990; SCHWARZ & THOMPSON 1990;
WILDAVSKY & DAKE 1990). These different groups have formed specific positions on risk
topics and have developed corresponding attitudes and strategies. They differ in the degree of
group cohesiveness (the extent to which someone finds identity in a social group), and the
degree of grid (the extent to which someone accepts and respects a formal system of hierarchy
and procedural rules).
These groups are: the entrepreneurs, the egalitarians, the bureaucrats, the stratified
individuals, and –added in some publications - the group of the hermits. They can be localized
within the group-grid space (see Figure 1). Organizations or social groups belonging to the
entrepreneurial prototype perceive risk taking as an opportunity to succeed in a competitive
market and to pursue their personal goals. They are characterized by a low degree of
hierarchy and a low degree of cohesion. They are less concerned about equity issues and
would like the government to refrain from extensive regulation or risk management efforts.
This group contrasts most with organizations or groups belonging to the egalitarian
19
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
prototype, which emphasizes cooperation and equality rather than competition and freedom.
Egalitarians are also characterized by low hierarchy, but have developed a strong sense of
group cohesiveness and solidarity. When facing risks they tend to focus on long-term effects
of human activities and are more likely to abandon an activity (even if they perceive it as
beneficial to them) than to take chances. They are particularly concerned about equity.
The third prototype, i.e. the bureaucrats, relies on rules and procedures to cope with
uncertainty. Bureaucrats are both, hierarchical and cohesive in their group relations. As long
as risks are managed by a capable institution and coping strategies have been provided for all
eventualities, there is no need to worry about risks.
FIGURE 1: PATTERNS OF VALUE CLUSTERS (DOUGLAS & WILDAVSKY 1982;
RAYNER 1990; THOMPSON ET AL. 1990; SCHWARZ & THOMPSON 1990;
WILDAVSKY & DAKE 1990)
Bureaucrats believe in the effectiveness of organizational skills and practices and regard a
problem as solved when a procedure to deal with its institutional management is in place. The
fourth prototype, the group of atomized or stratified individuals, principally believes in
hierarchy, but they do not identify with the hierarchy to which they belong. These people trust
only themselves, are often confused about risk issues, and are likely to take high risks for
themselves, but oppose any risk that they feel is imposed on them. At the same time, however,
they see life as a lottery and are often unable to link harm to a concrete cause. In addition to
the four prototypes, there may be a hybrid group called the autonomous individuals or the
hermit who can be grouped at in the center of the group-grid coordinates. Thompson
20
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
describes autonomous individuals as self-centered hermits and short-term risk evaluators.
They may be also referred to as potential mediators in risk conflicts, since they build multiple
alliances to the four other groups and believe in hierarchy only if they can relate the authority
to superior performance or knowledge
This theory has been criticized on several grounds (NELKIN 1982; SJÖBERG 1997). This is
not the place to review the critical remarks and the counter-evidence provided by many
scholars. The debate is still proceeding without a clear consensus in sight. Most risk
communicators have assured us, however, that this classification has helped them
tremendously in preparing communication programs for different audiences. There is
sufficient anecdotal evidence that people with an entrepreneurial attitude react very differently
to specific arguments compared to people with an egalitarian or bureaucratic attitude. For
example, a reference to cost-benefit ratios makes perfect sense when presented to an audience
of entrepreneurs but would trigger outrage when being referred to in a group of egalitarians.
How can one apply the insights of cultural theories to the perception of chemical risks?
Although the empirical studies by Dake (DAKE 1991 and GRENDSTAD 2000) do not
specify chemical risks per se, they suggest that entrepreneurs will probably welcome
chemicals as they support their conviction that taking limited risks makes life exciting and
provides strong benefits to the risk taker in the long run. The opposite is true for the
egalitarians: They strongly believe that nothing can be beneficial in something that is
artificially produced for special interests. Specifically they will argue that it is wise to avoid a
risk if there is even a chance that it might be detrimental to one’s health. Egalitarians also tend
to be very suspicious towards risk management agencies because they believe these agencies
are influenced by powerful interest groups (in particular those who pursue the entrepreneurial
ideas). Bureaucrats would see all chemical risks as a major challenge, since it makes
management much more difficult. Yet if they were able to see a consistent and implementable
path to incorporate best practice into the dominant regulatory regime, they might be less
reluctant to avoid the issue (Bureaucrats hate to fight, they rather ignore issues when they
don’t like them). The two remaining groups are of less relevance here: they see risk as a
manifestation of fate and would probably be not attentive to any risk debate.
3.5 Reflexive modernization approach to risk perception
The reflexive modernization approach combines the macro and micro level and goes back to
the seminal works of Ulrich Beck (BECK 1986, 1992) and Anthony Giddens (GIDDENS
1990, 1991). Beck claims that on the macro-level the societal production of wealth goes hand
in hand with the societal production of risks, which are incalculable and irresolvable (BECK
1992). On the micro-level the theory describes how people cope with these reflexive
modernisation trends and how that affects their risk perception (KNIGHT & WARLAND
2005).
Beck is convinced that modern risk has an equalizing effect on social stratus. According to
him it would be difficult to target special groups or to distinguish victims from those who
benefit from the risk taking activity. (BECK 1986, 1992). Risks transgress social, national and
cultural boundaries. The release of radioactivity is a major nuclear disaster. For example,
would affect the rich and the poor alike. Similar to Beck, Giddens (1991) understands risks as
generated by modernity and its institutions. He examines a broader array of risks, including
21
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
mundane as well as low-probability, high- consequence risks (KNIGHT & WARLAND
2005).
The preoccupation of modern society with risks has made public institutions become more
reflexive about their role and function in society. Reflexive means that actors need to find reassurance of their objectives and strategies in continuous communication with the outside
world. The result is that any attempt to monitor, decline or eliminate risks through industrial
processes will only serve to hasten unplanned, unforeseen latent consequences (KNIGHT &
WARLAND 2005). The development and implementation of pesticides resulted into a
contamination of ground water and in accumulation of residues in food. Also the
industrialisation of food production has increased the use of additives posing new and
unfamiliar health risks to consumers and new risks to the environment.
As a consequence of the confusion about the merits and risks of modernization, widespread
scepticism has evolved about the role and function of science and technology in producing
social benefit, and the belief in progress, i.e. notion that secular understandings of the world
would lead to a safer and more rewarding existence for humans (KNIGHT & WARLAND
2005). Science and technology are viewed as double-edged -‘‘creating new parameters of risk
and danger as well as offering beneficent possibilities for human kind’’ (GIDDENS 1991:28).
3.6 Summary: Approaches for perceptions of chemical risks
The perception of chemical risks is without doubt a complex process. The characteristics of a
risk source matter when people judge and assess risks. Criteria such as local/global,
private/occupational, present/future impacts exert a major impact on the degree of perceived
risks. In addition, acceptance of chemical risks will largely depend on benefit considerations
and beliefs about the effectiveness of institutional control and regulation. Five different
approaches for risk perception have been introduced in the chapter above, which facilitate our
understanding of the underlying factors that govern risk perception in general and chemical in
particular. The characteristics of these approaches are as follows:
Approach
Cognitive
heuristics
Level
Character
Common sense reasoning for making inferences
Anthropological about the world
Semantic images Social
psychological
Reduces complexity by four powerful images that
help to , affect information overload effectively and
to cope with uncertainty
Psychometric
Psychological
Factors
(Qualitative
Characteristics)
Cultural
Macro
approaches
to sociological
risk perception
characteristics of risks or the risk-taking situation that
help individuals to estimate the degree of seriousness
and to delineate a judgment about acceptability
Risk is defined as a social and cultural construct,
every culture or subgroup pursues its own risk
perception patterns five subcultures have been
defined that can be distinguished by two variables:
belief in hierarchy (grid) and intensity of group
22
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
cohesion (group) distributing people into five classes
due to their values and hierarchical attitude
Reflexive
modernization
approach to risk
perception
Combination of
micro
(individual) and
macro (societal
structures) level
Production of wealth goes hand in hand with
production of risks, risks have an equalizing, effect
on social structure risk actors need permanent
reassurement of their goals and strategies with the
outside world
It is hardly possible to describe all these influential factors by focusing on a single approach or
one model. We cannot recommend a specific approach due to the complex processes that
suggest a combination of all these forces being effective at the same time. It depends on the
individual circumstances under which a chemical risk is socially embedded in order to
determine which of the approaches might be more or less suitable to understand the risk
perception patterns of the various actors.
The perception of chemical risk is often part of an attitude that a person holds about the cause
of the risk, for example industrial activity, consumption of food, production method such as
genetic engineering. Most people perceive risks as more serious and threatening to avoid
cognitive dissonance, emotional stress caused by conflicting beliefs, if the other beliefs
contain negative connotations and vice versa. Often risk perception is a product of these
underlying beliefs rather than the cause for these beliefs (RENN & KASTENHOLZ 2000).
4. Central findings: What shapes risk perceptions?
The following sections describe empirical studies based on surveys, focus groups or
experiments. The main objective of these studies has been to investigate the causal
connections between the factors influencing the perceived seriousness of chemical risks and
the judgement on acceptability. These factors can be regarded as the independent variables in
the complex perception process. The focus will be on trust, social values, cultural and political
structure, amplification processes and the media.
4.1 Comparative empirical studies about risk perception
There are a lot of empirical studies published over the last 15 years that describe differences
of risk perception in various countries. ROHRMANN (1995) identified and documented 20
cross-cultural risk perception studies based on cross-national data collection; he also included
16 studies which looked at cultural differences within countries by comparing different
societal groups.
A more in-depth analysis was conducted by RENN & ROHRMANN (2000) on cross-cultural
risk perception patterns based on four empirical studies of the topic. These studies refer to
investigations on):
23
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus




Risk perception in the United States and France, P. Slovic, J. Flynn, C.K. Mertz, M.
Poumadere and C. Mays (2000).
Hazard perception and evaluation in Australia, Germany, and New Zealand and a second
analysis on China and Australia, B. Rohrmann (2000).
Risk attitudes in Bulgaria and Romania (with Sweden as reference), L. Sjöberg, D.
Kolarova, A.-A. Rucai, and M.-L. Bernström (2000).
The cognitive architecture of risk perception in the United States and Japan, Rosa,
Matsuda and Kleinhesselink (2000).
Table 6 provides an overview of the four studies. The table lists the sampling methods, the
countries involved in each study, and the primary independent and dependent variables.
TABLE 6: OVERVIEW OF THE DISCUSSED EMPIRICAL STUDIES
Paper
Sampling
Countries
Slovic
et. Al
Random general
population
USA, France
Rohrmann Special groups
(ecological,
technological,
monetarian,
feminist
orientation
Students (three
different
disciplines),
scientists
Sjöberg
Special Groups
et al
(nurses, students
of economics &
business, skilled
workers,
homeless
people)
Rosa et al. Several students
samples
Germany,
Australia,
New Zealand
Independent
variables
Qualitative hazard
characteristics,
cultural prototypes,
levels of emotion,
demographics,
country
Cognitive beliefs,
other attitudes,
group composition,
country
Dependent
variables
Magnitude of risk,
benefits (personal
and social),
acceptability
Demographic
variables such as
gender, respondent
group, anxiety,
credibility of risk
information
sources, country
Culture,
Cognitive
architecture
for
judgement decision
Magnitude
(personal; general);
demand for
personal and social
mitigation, and
others
Australia,
China
Bulgaria,
Romania,
Sweden
(Brazil as
contrast)
US and Japan
Risk appraisals
including
magnitude, health
impacts,
catastrophic
potential, benefits,
risk acceptance
(personal and
societal)
Risk
appraisal
including
familiarity,
knowledge,
voluntary,
controllability dread
and
catastrophic
potential
24
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
a. Differences in cultural affiliations: Comparison of France and the US
The study by SLOVIC et al. (2000) investigates people’s attitudes towards institutions and
policy taken the example of nuclear power. It compares two countries – the US and France –
uncovering many similarities, but also noticeable differences. Clear differences emerge
between the US and France. The French see greater need for nuclear power, and expect
greater economic benefit from it. In general, the French populations appear to embrace
nuclear power, whereas the US population has serious doubts about its safety and
acceptability. French citizens express greater trust in science, industry, and government
officials, and believe that politicians and experts are capable and able to make the necessary
decisions about nuclear power, rather than the people affected by these decisions. The US
respondents show much greater distrust of experts, are more inclined to support direct citizen
participation, and opt for more democratic control over the decision making process.
The differences between the French and US public are not the result of different risk
perceptions but are founded in the evaluation of the citizens with respect to their political
culture. The majority of French respondents have confidence in the centralised political
decision-making bodies that rely on expert consultation and bureaucratic reasoning.
Legitimisation in France involves the proof that the best experts have been consulted, and that
some agency is willing to accept the responsibility and accountability for any decision that
deviates from what the public actually wants or prefers.
This trust-based relationship between citizens and public institutions is alien to the US culture.
People demand that their preferences are taken into account when decisions are being made.
In addition, they do not trust agencies to be honest and independent brokers in managing
public safety, health, and environmental affairs. Thus, US citizens expect public officials to
reflect what citizens perceive as dangerous, and act according to the revealed preferences of
their clients.
Important for decision makers and risk communicators is the question if risk perception can be
influenced over the long run by educational or communication programs. The authors
concluded that attitudes towards nuclear power are conditioned by the interplay between
social, cultural and political factors. These cannot easily be changed by informational or
educational campaigns.
b. Social variables and professional affiliations (Hazard perception and evaluation in
Australia, Germany, New Zealand and China)
The influence of cultural variables to risk perception is emphasized in the second study, which
investigates risk perceptions and attitudes in Germany, Australia, New Zealand and China.
Bernd Rohrmann studied risk perception with respect to more than 40 risks (like parachuting,
living near a airport), classified them into “activities”, “residential conditions”, physical
hazards” and “sources of financial/social risk”. He measured individual risk perceptions and
conducted statistical analyses for different societal and professional groups (ROHRMANN
2000).
Although Rohrmann finds considerable cross-national variations in risk perception among the
countries, he warns against using this data as implicit proof of the validity of the cultural
prototypes that Slovic et al. (2000) explicitly tested in their study. In spite of some similarities
25
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
between the samples, Rohrmann points out that the differences between groups can be better
explained by referring to variations on occupational values (such as health care professionals
or teachers) and worldviews (such as religious convictions) rather than cultural prototypes. It
is indeed surprising that the differences between the countries are less pronounced with
respect to most of the variables used in the first study than between the social and
occupational groups within one country. The finding amplifies the observation that in the
process of globalisation, national identities are partially exchanged for international group
identities. That means that chemical risks are primarily perceived in relation to group
characteristics such as sharing a common occupation rather than national identities or even
countries. A chemical engineer from Spain will perceive similar chemical risks as a chemical
engineer in Norway or France, whereas a Spanish banker will likely perceive a risk differently
than a chemical engineer from Spain.
Rohrmann points out that among the many hazards that people might worry about, several are
always perceived as more serious than others, regardless of country, social group or
occupation. The list of top hazards includes nuclear power, asbestos, and smoking. There are
also some noticeable differences in the ranking of hazards between countries. Most of the
variance can be explained by differences in exposure or significance assigned to these hazards
in each country. However, it would be misleading to assume that risk perception mirrors the
measured level of risk revealed through risk assessments. Rohrmann demonstrates that many
hazards are perceived as either more perilous or less severe than epidemiological data would
suggest. This finding is prevalent in each of the countries included in his analysis.
c. Group attitudes in risk perception (Risk attitudes in Bulgaria and Romania)
The tendency to rank risks similarly was illustrated by the third study. Sjöberg et al. (2000)
compared six different groups in different national contexts. The main data was collected in
Bulgaria and Romania, and compared to the results of studies conducted in Sweden and
Brazil. The six professional groups are nurses, workers, economists, engineers, teachers, and
slum dwellers. The authors chose 28 main risks for his analysis, like smoking, alcohol,
unemployment, bad education, but also pollutants and food additives.
The ranking of risks, both personal and general, are rather similar the population in Bulgaria
and Romania rank most of the risks similarly. Apart from a few exceptions, the same is true
between the six subgroups of the study. Similarities in risk ranking were also detected in the
two comparative samples from Brazil and Sweden. People seem to be concerned about
identical threats all over the world, although the authors observe several notable exceptions.
For example, the Rumanian population was specifically concerned about being attacked by
dogs, a risk that was not even mentioned in any of the other studies. The explanation for this
strange result is quite trivial. The late dictator Ceausescu had hundreds of dogs set free when
he ordered to tear down large parts of the old quarter in Bucharest. In one disregards these
idiosyncrasies, the similarities in risk perception prevail. This is particularly true for
technological risks. Chemicals are usually grouped into the technological risk area, and raised
medium concerns in both countries.
Sjöberg et al also rejected the thesis that the cultural prototypes are powerful predictors of risk
perception. The results of his study suggest that the risk judgments made by the Bulgarian and
Romanian samples are based on the perception of specific threats, attitudes towards nature
and, indirectly, through media coverage. They reject the hypothesis of general, culturally
26
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
contingent values and beliefs. Cultural orientations and biases are important variables that
influence risk perception, but they may differ from those specific biases explicated by the
theory of cultural prototypes. Furthermore, economic status and social aspirations may be
more decisive in shaping people’s attitudes towards technological risks including chemicals
than cultural biases.
d. The hypothesis of universalism (The cognitive architecture of risk perception in the
United States and Japan)
The common observation that risk perception rankings are rather similar among different
nations and cultures, and that some of the factors influencing these rankings are almost
identical, provided the inspiration for the study by Rosa et al (2000). Its intention was to test
the hypothesis of universalism of psychometric variables in determining perceived risk. In
total 70 risks were included in the cross cultural research. The main objective was not to find
the best predictors for explaining variations in risk perception, but to identify hidden clusters
of risk perception based on a variety of psychometric characteristics (by using factor analysis).
The clusters were investigated within and between the different samples.
A particularly interesting result is that the main clusters characterizing risk perception in the
US sample could also be found in the Japanese sample. Based on previous studies (SLOVIC
et al. 1986) about the importance of psychometric variables the two items “perceived dread”
and “perceived catastrophic potential” – were confirmed as valid predictors of overall risk
perception scores in both countries. However, the cognitive routes that let respondents
develop those images are fundamentally different.
For the Japanese, the belief that the risk would lead to catastrophic results was highly
correlated with lack of individual knowledge about risks, and the feeling of having little
control over them. In turn, dread, i.e. the belief that any incident will inadvertently lead to
severe damage s was linked to catastrophic potential, a lack of scientific knowledge, and the
perceived involuntary nature of the phenomena. These relationships were almost reversed
among US respondents. In particular, perceived catastrophic potential was highly correlated
with perceived lack of scientific knowledge about risks and perceived lack of control over
them.
If confirmed, these results may suggest different routes for risk communication. Americans
need to be convinced that science is able to limit the maximum credible accident, whereas the
Japanese need to be assured that people can rely on personal experience when dealing with
such risks, and that personal knowledge is available to cope with the potential outcome of a
hazardous event.
4.2 Institutional Trust and Confidence
The previous empirical studies were based on the assumption that risk judgements and
evaluations are "universal" processes, independent of, e.g., social status, professional
affiliation, ideologies, national heritage and other cultural contexts. The main insights were
that in addition to cultural prototypes, the affiliation to a professional or social group exerted
the strongest effect on risk perception and the evaluation of technological and chemical risks.
27
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
In addition to these cultural and group-related factors in shaping risk responses, social science
research also investigated the role of social networks and reference group judgments which
have been proven particularly influential in other fields of study. These orientations are also
important since most information about risk is not learned through personal experience and
senses but through "second-hand" learning. With the advent of ever more complex
technologies and the progression of scientific methods to detect even smallest quantities of
harmful substances, personal experience of risk has been more and more replaced by
information about risks and individual control over risk by institutional risk management. As
a consequence, people rely more than ever on the credibility and sincerity of those from
whom they receive information about risk. Thus, trust in institutional performance has been a
major key for risk responses. This is particularly important for risks that belong to the
category of “early warning indicators”. Trust in control institutions are able to compensate for
even a negative risk perception and distrust may lead people to oppose risks even when they
are perceived as small. This function of trust as a mediator between negative risk perception
and a positive judgment about acceptability was empirically proven by a study that compared
attitudes towards nuclear power in France and the United States (MARRIS et al. 1999). This
study showed that the French respondents had similar risk perceptions and shared the same
concerns about safety as their American counterparts. However, the French sample
demonstrated a high degree of trust and confidence in the technical elite of their country and
in the effectiveness of regulatory bodies, both of which US respondents were much more
sceptical.
Trust can be substructured in five components (BARBER 1983; LEE 1986; RENN and
LEVINE 1991). These five components are listed and explained in Table 8. Trust relies on all
five components, but a lack of compliance in one attribute can be compensated for by a
surplus of goal attainment in another attribute. If objectivity or disinterestedness is impossible
to accomplish, fairness of the message and faith in the good intention of the source may serve
as substitutes. Competence may also be compensated by faith and vice versa. Consistency is
not always essential in gaining trust, but persistent inconsistencies destroy the common
expectations and role models for behavioural responses. Trust cannot evolve if people
experience inconsistent responses from others in similar or even identical situations.
TABLE 7: COMPONENTS OF TRUST
Components
Description
Perceived competence
degree of technical expertise in meeting institutional mandate
Objectivity
lack of biases in information and performance as perceived
by others
Fairness
acknowledgement and adequate representation of all relevant
points of view
Consistency
predictability of arguments and behaviour based on past
experience and previous communication efforts
Sincerity
honesty and openness
Faith
perception of "good will" in performance and communication
28
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
Trust on a personal level is a subjective exception that a person will refrain from behavioural
options that may harm the trusting person. Trust necessarily entails risk-taking, but, in
contrast to the scientific endeavour of predicting the probability of identified outcomes, trust
implies that the selection of options is left to the entrusted person or institution. Due to the
perceived competency and honesty of the entrusted entity, one does not need to bother with
assessing the outcomes of actions and with controlling the decision making process of that
entity (LUHMANN 1980, 1973). This saves time and effort.
On a more aggregate level, trust denotes a generalized medium of social differentiation and
division of labour (PARSONS 1960). The performance of specialized institutions in economy
and government relies on a prior investment of trust by those who are served by this
institution or finance its functioning. Total control would imply that the control agencies
would need the same expertise and the same time allocation as the performing institution.
Such an arrangement would neutralize the desired effect of social differentiation and
ultimately lead to a society of intimate clans performing all necessary social, economic, and
political functions simultaneously. By shortcutting normal control mechanisms, trust can be a
powerful agent for efficient and economical performance of social tasks. Durkheim's analysis
of organic solidarity as a major structural variable of modern societies focused on trust as one
of the predominant media that helped to shape the division of labour and to differentiate
societal functions (DURKHEIM 1933; LUHMANN 1973).
The relative value of trust varies over time, as empirical surveys clearly indicate (LIPSET &
SCHNEIDER 1983; more references). In some periods, people tend to invest a large amount
of trust in institutions and it takes many disappointments before they withdraw this
investment. In other periods, people tend to be extremely cautious with the investment of
trust, placing more emphasis on functional equivalents, such as more organized control or
increase of direct participation. Trust can partially be substituted by other generalized media,
such as sharing power or control, but not totally replaced.
It is obvious that modern societies face difficulties in providing sufficient trust for reaching
consensus on its complex and differentiated activities. All public institutions have lost trust
and credibility over the last two decades except for the news media (LIPSET & SCHNEIDER
1983; Löftstedt 2005; RENN & LEVINE 1991). Trust and credibility losses are high for
industry, the political system, and many government agencies. Science still has a high degree
of credibility although much less than two decades ago. Most sociologists believe that the
decline of confidence in public institutions is partially a function of better education and the
increase of public aspirations with respect to their share of public resources and welfare
(LIPSET & SCHEIDER 1983; KATZ et al. 1975). In addition, the complexity of social issues
and the pluralization of values and lifestyles may have contributed to a growing
dissatisfaction with the actual performance of institutions (WILDAVSKY & DAKE 1990).
But at the same time, most people are confident in the governmental and economic system
and do not support fundamental changes in the organizational structure of society. Therefore,
the confidence crisis is less a systems than a performance or competence crisis.
Lack of trust does not indicate, however, a declining relevance of trust for governing modern
societies and managing technological risks. The contrary is true. The reliance of the
technological society on trustful relationships between and among its subsystems has never
been stronger than today. However, such a need for trust makes people more and more
sensitive towards situations in which their investment of trust has been factually or allegedly
29
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
misguided or exploited. The more trust is needed for implementing cooperative efforts or for
coping with external effects of social actions, the more cautious are people in assigning
credibility to those whom they are supposed to trust.
In risk debates issues of trust evolve around institutions and their representatives. People's
responses to risk depend, among others, on their confidence that they have in risk initiating
and controlling institutions. Since the notion of risk implies that random events may trigger
accidents or losses, risk management institutions are always forced to legitimate their action
or inaction when faced with an accident. On one hand they can cover up mismanagement by
referring to the alleged randomness of the event (labelling it as unpredictable or an act of
God), on the other hand they may be blamed for events for which they could not possibly
provide protective actions in advance (LUHMANN 1986).
The stochastic nature of risk demands trustful relationships between risk managers and risk
bearers, since single events do not prove nor disprove management failures; at the same time
they provoke suspicion and doubt. The slightest mistake by a risk management agency can be
sufficient to destroy the delicate balance of trust. The handling of risk by private corporations
and governmental agencies has been crucial for explaining the mobilization rate of individuals
for taking actions. The more individuals believe that risks are not properly handled (in
addition to being perceived as serious threats) the higher is the likelihood that people will be
politically active. It has been shown that in the nuclear case the disillusionment of the USpopulation with the nuclear option as well as the number of people becoming political
advocates of antinuclear policies grew simultaneously with the growing distrust in the nuclear
regulatory agency (RENN & LEVINE 1991). Negative attitudes are a necessary but by far not
a sufficient reason for behavioural responses. Public confidence in institutional performance
is another and even more important element in triggering behavioural responses.
It would be beyond the scope of this article to list all the relevant empirical results that
investigated for the significance of trust and credibility, but some of the most interesting (and
even counterintuitive) insights from these studies are worth mentioning (for a more detailed
analysis see KATZ et al. 1975; LIPSET & SCHEIDER 1983; SIEGRIST et al. 2000; RENN
& LEVINE 1991)

Researchers found a low correlation between the perception of institutional competence
and the liking of the tasks and goals that the institutions were performing. The more
sympathy an institution is able to generate the less competent it appears. The institutions
people like most received low ratings on competence and vice versa (LIPSET &
SCHNEIDER 1983).

Perceived competence of institutions was most likely associated with the perception of a
successful task performance and the perceived cost-benefit ratio in meeting these tasks. In
addition, the public image and the social prestige assigned to an institution serve as
preliminary heuristic strategies to assign credibility (MATEJKO 1988).

Perceived fairness and openness is closely linked to the transparency of the decision
making process, the opportunities for public scrutiny and institutional control (check and
balances), and the degree of personal satisfaction with the rationale and procedures for
decision making in the respective institution. Surprisingly, the amount of actual
30
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
opportunities for public involvement and participation was hardly correlated to perceived
openness (LIPSET & SCHNEIDER 1983; cf. theoretical concept LUHMANN 1980).

Institutional case studies demonstrated that the erosion of credibility was often linked to:
incompetence, poor performance, incomplete or dishonest information, withholding of
information, obscure and hidden decision making processes, denial of obvious problems,
and denial of vested interests (BERGESEN & WARR 1979; LIPSET & SCHNEIDER
1983; MATEJKO 1988; MIDDEN 1988).

Credibility can be enforced by good performance, fast responses to public requests,
consonance with highly esteemed social values, availability for communication with
outsiders, unequivocal and highly focussed information transfer, flexibility to respond to
crisis situations or new public demands, and demonstration of public control over
performance and money allocation (LIPSET & SCHNEIDER 1983; ROURKE et al.
1976).
The results of organizational studies on credibility emphasize the close relationship between
perceived performance and credibility. Many risk management institutions face the problem,
however, that their specific task is not well understood and that public expectations do not
match the mandate or the scope of management options of the institution. This is certainly not
unique for risk management agencies: LIPSET & SCHNEIDER (1983) found out that elites
in America complain regularly about the ignorance and misperceptions of the public with
respect to their mandate and performance. Regardless of whether this claim is true, there is a
clear gap between the self-perception of most institutions and the public perception of these
institutions. This is specifically prevalent in the risk arena because the issue at stake, health
and environment, tops the concerns of the public in almost all countries and because the
stochastic nature of risk impedes an unambiguous evaluation of management success or
failure.
5. Perception of combined chemical risks
5.1 Survey in Baden-Württemberg
In what way do individual response patterns change when combined risk is used as a
stimulus? Until recently there was hardly any empirical evidence available, only assumptions
at best. In an attempt to address this deficit, the Center of Technology Assessment in BadenWürttemberg (Akademie für Technikfolgenabschätzung) conducted a representative survey
on this subject among the population of the state of Baden-Württemberg.
The results of the survey are presented and discussed in sections 5.2. and 5.3. The sample
included a representative cross-section of some 1,500 adults in Baden-Württemberg during the
summer of 1998. Its main focus was society’s perception and assessment of technology
(ZWICK & RENN 1998).
31
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
5.2 Perception of combined risks: interaction between environment and lifestyle
Within the framework of the risk survey in Baden-Württemberg, combined environmental
risks were addressed. At first, respondents were asked to assess the importance of
environmental factors in health risks.
According to prevailing expert opinion, more than two-thirds of all cancers can be
apportioned to smoking, alcohol consumption or poor eating habits, while only a fraction of
cancers can be apportioned to the impact of noxious substances in the environment
(HENSCHLER 1993). In the survey the participants were confronted with the following
statement:
“The health risks from environmental pollution are considerably greater than the risks from
personal lifestyles like obesity, smoking, alcohol consumption or leisure activities.”
Some 10.7% strongly agreed and a further 15.3% tended to agree. Practically every second
respondent was undecided (48.1%) and about a quarter (25.9%) thought the risks from
individual lifestyles were greater than the risks from environmental pollution. Adding the
people who agreed and who were undecided, three quarters of the respondents appeared to be
unaware of the statistical data that clearly show a dominance of habitual over environmental
risks. The answers were similar across all groups; there is, for example, no difference as to
whether a respondent preferred green policies or had more conservative values. The only
exception was the level of education: the higher the level of education, the more often
respondents were giving the correct answer. Only 20% of respondents with high school
diplomas but almost 50% of university trained persons expressed the opinion that
environmental risks are less of a threat to human health than the risks involved in personal
lifestyles.
A significantly bigger gap is evident between the opinions of the majority of experts and the
laypeople surveyed regarding their assessment of combined environmental risks. The
respondents were again confronted with a statement:
“Just as combined consumption of tablets and alcohol can cause serious health problems,
relatively harmless substances in the environment can also cause serious damage to one’s
health when they interact.”
Although the super-additive relationship postulated in the statement cannot be ruled out from
a toxicological standpoint, the majority of toxicological studies (cf. STREFFER et al. 2003)
show that super-additive effects are more the exception than the rule. Although the statement
used the soft version “can cause serious harm” for the environmental risks, the doubling of
this phrase in the first part of the sentence (consumption of tablet and alcohol can cause
serious harm) implies a symmetry between the two conditions mentioned. People who fully
agree to the statement perceive a symmetrical situation between the combined effects of
alcohol and drugs and the combined effects of environmental pollutants. Two-thirds of
respondents were convinced that such super-additive effects could be expected from
interactions between multiple pollutants. Just under a quarter (23.8%) were indifferent and
only 9.6% rejected the statement. There were hardly any differences between the various
social groups. Ages, gender, party preference, even formal education, were all variables that
had no significant impact on how the respondents reacted to the statement.
32
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
Because two-thirds of the population thinks, super-additive effects could occur through the
combination of multiple pollutants, the acceptability of further contamination in soil, air and
water is low. On the one hand, this clear response can stem from everyday experience with
medium or high doses of drugs and stimulants. On the other hand, typical aversion to the risk
of “early warning”, especially intolerance of continuous uncertainty, may play a significant
role. The survey included a second item that related to combined effects:
“Even if we don’t know for sure what effect environmental pollution can have on human
health, we should for economic reasons allow a certain amount of pollutants to be released
into the environment.”2
Only 8.9% of respondents agreed with this statement, a third were undecided and a large
majority of 58.7% rejected the statement altogether. There was no evidence of major
differences according to gender, age, class or income in response to this particular statement,
only supporters of the German Green Party (Die Grünen) appeared less willing to approve to
the statement than did the average member of the population. In general a majority of almost
60% rejection indicates that most people associate super-additive effects from exposures to
chemical substances in the low-dose range. In addition, it appears that most people will not
make trade-offs between health impacts and economic benefits if these consequences are
perceived as uncertain. Given the fact that chemicals are most likely to be found in the
category of “slow agent”, the reliance on third party information combined with perceived
uncertainty about risks and benefits reinforces the demand for zero risk.
5.3 Interpretation of results
Apart from the survey about combined environmental risks (ZWICK & RENN 1998), we
know of no sociological or psychological study that explicitly attempts to explain in empirical
form how synergetic risks are perceived. However, there are a variety of related studies that
throw additional light on the subject. One is the empirical works of Kraus, Malmfors and
Slovic (KRAUS, MALMFORS & SLOVIC 1992) on intuitive assessment of toxicological
risk and the works of Jungermann, Schütz and Thüring on the question of risk perception
involving the data sheets enclosed with drugs (JUNGERMANN, SCHÜTZ & THÜRING
1988).
Firstly, the information on large, unimaginable, invisible quantities of chemical substances in
the environment leads us to conclude that not all effects are known and that the possibilities
for combination have not been fully tested. There is thus an element of uncertainty in
individual perceptions, and this cannot be erased by any amount of care or knowledge about
individual noxious substances in the environment. If we look at the studies on qualitative risk
characteristics, perceived uncertainty pertaining to possible impacts plays a significant part in
engendering a more negative assessment of the risks (SIEGRIST & CVETKOVICH 2001).
Most importantly, perceived uncertainty leads to a desire for more stringent regulation
(SJÖRBERG 1994). It is hardly surprising that uncertainty surrounding combined effects
tends to promote caution and demands for strict regulation and enforcement (ZWICK &
RENN 1998).
The phrase “even if” is a translation of the German “auch wenn” which is much less rigid than the word “even”
implies. Literally the phrase “auch wenn” translates into “Also when”
2
33
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
Secondly, combined effects as early warning risks gives raise to the hypothesis that causal
evidence, which plays an important role in intuitive risk assessment, is more readily brought
to mind when assessing combined effects than when assessing individual substances. We all
see ourselves affected by omnipresent chemical substances in some way. At the same time we
can recall at least one example of an illness that is (apparently) caused by environmental
impacts but are unable to name the actual cause-effect chain. Combined effects thus can be
easily recalled and they are practically irrefutable forming an everyday hypothesis on which
we base health risks from unspecified environmental impacts.
The fact that, on the one hand, we accept risk in one context, perhaps even seek it, while on
the other we reject the same or an even lesser risk is not necessarily proof of irrational or
inconsistent behaviour. Chemical risks are more acceptable and predictable for us, when the
relationship between an event or substance and his cause is seen as obvious. We know the
cause and find a way to handle or avoid it. For example we know the carcinogenic effects of
the chemical substances of formaldehyde and benzopyrenes and try to avoid getting in contact
with them. Combined exposure in many situations is less obvious for many scientists but
refers to everyday experience of most people. They may find it strange that scientists find
such cocktails difficult to test and to assess their synergistic effects.
These two heuristics – rapid recall of causal experience and concern based on the lack of
understanding for toxicological or epidemiological testing methods – are reinforced by
perceptions of disagreement among experts (NENNEN & GARBE 1996). As it is currently
not possible to identify all combined effects and many theoretical possibilities for
extrapolation of values are limited to the low-dose range due to a lack of significant causeeffect relationships, experts must also rely on plausible models, theories and assumptions
(PETERS 1996: 63). In reality, the need for interpretation, especially where combined
exposures are concerned, has contributed to a pluralization of expert opinions and
assessments within the science system. There are consequences for the treatment of risk
assessment both inside and outside the scientific community. Firstly, the case of scientific
treatment often referred to as an expert’s dilemma of the first degree: experts conduct risk
assessments in a variety of ways. Four categories can be identified in the plural sphere of
expert opinion (Streffer et al. 2003: 263pp.):




Expert assessments that focus on the experimental results of risk studies and
conventions drawn up by experts, and which deal with remaining uncertainties by
simply ignoring rather than assessing them.
Expert assessments that focus on the empirical results of risk studies but which hover on
the border of the range of conventions drawn up by experts and, in doing so, interpret
uncertainties within the limits of those agreed conventions.
Expert assessments that focus on the empirical results of risk studies, but which reject
conventions agreed to by experts or replace them with their own interpretation models.
Expert assessments that question both conventions and empirical results, and see their
purpose in fundamental critique of the methods and interpretations of the relevant
scientific community.
Without a doubt, the vast majority of experts are to be found in the first two categories, so that
the conflicts that arise in reality are less strongly manifested than one would expect from
looking at all four. At the same time, the various scientific communities use methods such as
consensus conferences, meta-analyses or Delphi surveys in their attempts to resolve conflicts
between the four types of expert opinions (NENNEN& GARBE 1996). The multi-layered
34
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
nature of risk analysis for combined pollutants makes it difficult to find a clear-cut solution
for conflicting expert assessments. In addition, we have a situation where statements on risk
from combined effects are difficult or impossible to falsify. The lack of opportunity to falsify
statements using empirical evidence (at least in the short-term) limits the effectiveness of
knowledge as a tool to evaluate risk assessment studies. Different bodies of knowledge
compete with each other and the competing demands for truth cannot be met to the exclusion
of all possible doubt.
In addition, we need to consider the consequences of expert disagreements beyond the
scientific communities: the existence of discretionary freedom in the assessment of combined
effects, the need to implement conventions that cannot be justified by science alone, the
constant uncertainty in effects analyses and the fact that numerous scientific controversies are
debated in public all have a sustained effect on people’s perceptions and experience. In many
ways, this involves the importance that outsiders ascribe to conflicts between experts: firstly,
most people believe that, in principle, science can come up with clear and precise definitions
of environmental pollution. Confronted with a large number of conflicting assessments by
experts, people are thus forced to conclude that in the course of such conflict at least one or
other of the parties involved is not revealing the truth, be it intentionally or unintentionally. It
is not without reason that in the public eye, what experts put out is seen as a reflection of what
their financial backers, their ideological preconceptions or their blinkered specialized world
put in (COVELLO 1992; RENN 1995).
As most people are unable either to confirm or to reject the statements put out by experts by
applying their own knowledge or experience, they must depend on external criteria if they
want to assess the trustworthiness of those experts. Such external criteria range from assumed
vested interest in the subject (an area in which industrial experts have particular difficulty) to
the perceived likeableness or expertise of those involved (how they handle themselves on TV,
how they dress, their debating style) (PETERS 1991). Laypeople often follow an intuitive
better safe than sorry principle. The assumption that the most pessimistic expert embodies the
epitomy of trustworthiness is reinforced by the fact that in the past environmental pollution
has often been reported as less harmful than it actually turned out to be. Conversely, some
experts often feel pressured by public expectations to place greater emphasis on negative
outcomes in order to improve their public standing, while others financed by industrial
interest groups may tend to underestimate risks. Most consumers are not taking studies
seriously, which are financed by the industry.
6. The Influence of the Media
Most chemical risks that modern society faces are not experienced by human senses but
learned through communication. The dangers of the risks of food additives or chemicals in
drinking water, the threat a chemical explosion would probably never reach public attention
unless society communicates about these adverse possibilities. Risk perception is less a
product of experience or personal evidence than a result of social communication
(LUHMANN 1986).
A vast amount of information about risks stems from intermediary sources. People develop
attitudes and positions with respect to risky technologies and or activities on the basis of
second-hand information. This information is transmitted by the media. Many beliefs about
35
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
risks and risk sources are hence shaped or at least influenced by the information and
evaluations that the media transmit to their consumers. The media perform a dual role in the
communication process: first, they collect information from primary sources and process this
information by applying professional and institutional rules that govern the selection of
received messages and their interpretation. Second, they send information to the final
receiver. The re-coding of messages involves conscious or unconscious changes of the
original information material. Messages from several sources may be integrated into one new
message or comments may be added.
The transformation process of messages during transmission has been a popular topic of
communication research. From a theoretical point of view, many different concepts about the
nature of this transformation have been suggested in the literature (PETERS 1990; PELTU
1985, pp. 129-130; 1989; SOOD et al. 1987, p. 30; LEE 1986, p. 175). The basic differences
between these approaches may be confined to two major questions: First, are the media
creating new messages or are they reflecting existing messages; second, how biased are
journalists in their coverage vis-à-vis their own social convictions and external pressures?
Both questions have not found a final answer yet (PELTU 1985, pp.140-141; MAZUR 1987,
p. 86; LICHTENBERG & MACLEAN 1988, pp. 33-48).
With respect to the first question, the literature suggests a strong influence of the media on
public opinion in the early years of communication research. Through extensive testing,
however, this hypothesis was later substituted by the hypothesis that the media set the agenda,
but do not change the attitudes or the values of the audience with respect to the issues on the
agenda (PELTU 1985, p. 140; LICHTENBERG & MACLEAN 1988). Only in the long term
have media a lasting effect on the attitude and value structure of their consumers.
With respect to the second question evidence has been gathered to support almost all possible
viewpoints. Political and commercial pressures have been detected in media coverage as well
as courageous news reports in conflict with all vested interests. Cultural biases within the
journalistic community have been found, but also a variety of different political and social
attitudes among journalists. Some journalists perceive their job as a mere translation of events
into verbal or visual expressions while others believe they should play a more active role in
shaping and explaining the issue (cf. the controversy about the studies of Kepplinger in the
review by KÖCHER 1986; LICHTENBERG & MAC LEAN 1988, pp.37-45; PETERS 1990;
1991).
In short: the extremes that media are mere reflectors of reality or that they are docile
instruments of social pressure groups may occasionally be true, but they are not the rule. In
reality, the situation is more complex: Media coverage is neither dependent on external
pressures nor an autonomous subsystem within society (LOWERY & DEFELEUR 1983). It
reflects internalized individual values, organizational rules and external expectations. It
depends on the issue itself, the institutional context and the political salience of the issue
which of these three factors is likely to dominate the transformation process. A universal
theory of how this transformation takes place is therefore not likely to evolve. Some of the
common characteristics of media coverage deserve some attention, however:

Media construct reality as well as readers construct their understanding of the media
report (DÚNWOODY 1992). These constructions are results of mental and professional
frames that journalists use in selecting and recoding information. Construction does not
36
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
imply that the coverage is independent of the real events. But there is ample evidence
that the media amplify some elements and downplay others when processing
information (WILKINS & PATTERSON 1987; SOOD et al. 1987). For example, the
number of fatalities is a rather weak indicator for amount of coverage in risk issues,
while the degree of social conflict arising from a risk debate correlates high with media
coverage (ADAMS 1986).

Media direct their attention to events, not continuous developments. An accident-free
performance of a technology over many years is not newsworthy, unless it is framed as
an event (such as a public celebration). Likewise slow changes of the climate become
hot news issues only if they can be linked to a conference, an exceptional hot summer
(such as 1988 in the Unites States) or political statements (PELTU 1989).

Media have no internal mechanism to resolve conflicts among experts. Journalists have
neither the time nor the qualification to find out who is right in a scientific debate. The
most frequently used method to handle competing scientific evidence in the media is to
give each side room to state or justify claims (PETERS 1991). Most journalists have
lists of people who will provide counter-statements to any statement that they encounter
when working on a story. Neither quality of evidence nor proportionality (with respect
to number of dissidents or professional qualification) determines the amount of coverage
that each side will receive. The amount is either equally distributed among camps or
biased towards the preferences of the journalist or -more frequently- towards the
editorial style of the respective medium. Media in a pluralistic society tend to reinforce
diversity, dissent, and relativity of values (RUBIN 1987, p. 53).
Is there any evidence about specific media treatment of risk-related information? The media
collect information from direct eyewitnesses of hazard events (anecdotal evidence) as well as
systematic information from risk management institutions (RENN 1991). Displaying
anecdotal experience (such as losing property or being injured) contrasts with the statistical
evidence provided by risk experts. This contrast reinforces the constructive nature of media
coverage and its reception (same event through two very different lenses), and often
contributes to the erosion of trust in experts (see below). In addition, the nature and the
magnitude of the original hazard are only of minor interest to most journalists. They prefer to
focus on the way institutions handle risks and communicate about their activities. As Singer
and Endremy have pointed out, the media emphasize hazards that are relatively serious and
relatively rare; it is the combination that gives them their punch (SINGER & ENDREMY
1987, p.13). For example, the Chernobyl accident with 31 acute deaths cases received 129
minutes of CBS News coverage while the 1976 Tandshan earthquake leaving 800,000 people
dead received less than 9 minutes on the average TV evening news (SOOD et al. 1987, p.37).
The literature contains endless lists of factors that are assumed to determine the attractiveness
of risk-related messages for transmitters. Such factors include: technologically induced hazard
(versus natural hazard), cultural distance from the place of occurrence (ADAMS 1986),
human interest component, drama and conflict, exclusiveness of coverage (PELTU 1985, pp.
137-138), proximity to politically hot issues, prestige of information source, and degree of
conflict among stakeholders (PETERS 1990).
Interaction among transmitters, plural input from different sources, the co-existence of
personal, professional, and institutional selection and amplification criteria, and interaction
among different target audiences create enough complexity and uncertainty that the final
37
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
effect of the communication process can hardly be measured at all, let alone be effectively
controlled. Reception studies of media coverage are therefore rare and often very restricted in
the experimental design. It is clear, however, that people tend to form opinions and attitudes
by a selection process in which parts of news-stories are taken out and rearranged in
accordance with personal preferences, existing attitudes, and values (DUNWOODY 1992).
Media consumers create puzzles constructed by many elements (cognitive and evaluative)
from a variety of media reports. It is not so much the intention of the message that consumers
take for granted, but their pre-existing viewpoint that make them select and interpret the
messages.
Up to now, no research has been conducted on media reporting of combined effects. It is to be
expected, however, that the same selection criteria are applied as with other environmental
topics. Events receive a lot of attention especially when they can be presented in a way to stir
public curiosity or outrage. At the same time, the media strive to obtain counter-arguments for
every expert opinion and to publish them in parallel regardless of how representative they may
be of the relevant scientific community. Thus there is a forced emphasis on plurality and on
disagreements among experts.
38
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
7. An integrative model of risk perception
Based on the literature about risk perception, a structured framework can be constructed to
help risk assessors and managers understand and interpret their empirical findings related to
risk perception. Figure 3 illustrates the influences on risk perception of four context levels:

cultural background,

socio-political institutions,

cognitive-affective factors, and

heuristics of information processing.
Each represents substructures of collective and individual influences. Each level is embedded
in the next higher level to highlight the mutual contingencies among individual, social, and
cultural variables.
Figure 3: Four Context Levels of Risk Perception
Four Context Levels of Risk Perception
Personal
Manifestations
Collective
Influences
Cultural Background
Cultural
institutions
Personal
identity and sense of
meaning
Political, societal and
economic culture
Worldviews
Social-Political Institutions
Personal values and
interests
Social values
and trust
Cognitive-Affective Factors
Economic &
politial
structures
Organizational
constraints
Referenceknowledge
Personal beliefs
Stigmata
Emotional affections
Socioeconomic
status
Heuristics of Information Processing
Collective
Heuristics
Risk
Perception
Individual
Common
Sense
Media
influence
1. Level: Heuristics of Information Processing
The first level includes the collective and individual heuristics used when forming judgments
about risks. Many researchers have been able to demonstrate that several common
characteristics such as dread, catastrophic potential, perceived controllability and familiarity
have a major influence on the perceived magnitude of risks, based on empirical data from the
39
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
United States, Germany, France, Canada, Austria, Australia, Japan and elsewhere (Compare
Overview Chapter 3 and in RENN 1990, ROHRMANN 1995; 1999).
2. Level: Cognitive and affective factors
The second level refers to the cognitive and affective factors that influence the perception of
specific properties of the risk in question and also indirectly through the assignment of special
weights to the universal heuristics described above. Cognitions about a risk source – what
people believe is true about a risk – governs the attribution of qualitative characteristics
(psychometric variables) to specific risks (for example dread or personal control options) and
determines the effectiveness of these qualitative risk characteristics on the perceived
seriousness of risk and acceptability It is interesting to note that different cognitive processes
can lead to the same result (in the case studied, the arousal of catastrophic images, in ROSA et
al. 2000). While the Japanese respondents based their image of catastrophe on individual
knowledge and familiarity, US respondents associated collective scientific experience and
knowledge with catastrophic potential. This difference in cognitive routes also justifies the
distinction between cognitive factors and heuristics (the first and second levels in Figure 3).
While cognitive factors have been explored extensively, emotions have been widely neglected
in risk perception research. More recently, however, psychologists have discovered that affect
and emotions play an important role in people’s decision processes (SLOVIC et al. 2002;
LOEWENSTEIN et al. 2001). These factors are particularly relevant when individuals face a
decision that involves a difficult trade-off between attributes or where there is interpretative
ambiguity as to what constitutes a ‘right’ answer. In these cases, people often appear to
resolve problems by focusing on those cues that send the strongest affective signals (HSEE &
KUNREUTHER 2000). Both Rohrmann and Sjöberg demonstrated in their studies, that
emotional affections about risk sources influence risk perception and co-determine the
salience of psychometric variables such as dread and catastrophic potential. On the collective
level, stigmata referring to risk sources or activities play a similar role in stimulating
emotional responses (SLOVIC et al. 1992). Recent work on risk perceptions regarding
technological hazards shows that emotional and cognitive factors are mutually related
(ZWICK & RENN 1998). It is not yet clear whether cognitive beliefs trigger the respective
emotional responses, or whether emotional impulses first ‘search’ for arguments supporting
one’s emotional stance.
3. Level: Social and political institutions
The third level refers to the social and political institutions that individuals and groups
associate with handling risks. Most studies on this level focus on trust in institutions, personal
and social value commitments, organisational constraints, social and political structures, and
the socio-economic status. Another important variable in evaluating risk is the perception of
fairness and justice in allocating benefits and risks to different individuals and social groups
(LINNEROOTH-BAYER & FITZGERALD 1996). Other studies have shown the relevance
of political and social organisations and their strategies to communicate with other
organisations and society at large (CLARKE 1989).
The media, social groups, institutions, and organisational bonds also shape individual and
societal risk experience. Several of the studies in this article were discussing the role of the
media. Press coverage appears to contribute substantially to a person’s perception of risk,
40
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
even when the extent of direct harm to people and property is taken out of the equation. In
contrast to popular belief, however, there is no evidence that media create opinions about risks
or even determine risk perceptions. Studies on media reception suggest instead that people
select elements from media reports and use their own frame of reference to create
understanding and meaning (cf. e.g. LICHTENBERG & MAC LEAN 1988). This creation is
certainly influenced by the cognitive content and the evaluative statements of the reports, but
the influence is more indirect and ambiguous than expected. Most people reconfirm existing
attitudes when reading or viewing media reports (PETERS 1994).
4. Level: Cultural background
The last level refers to cultural factors that govern or co-determine many of the lower levels of
influence, including the selection of universal heuristics. Opinions on the validity of the
cultural theory of risk differ widely. SLOVIC et al. (2000) regard this approach as useful in
explaining some of the differences in risk perception; SJÖBERG et al. (2000) found the
variance explained by cultural prototypes to be so low that they rejected the whole concept.
ROHRMANN (2000) expressed a sceptical view, mainly because of measurement
deficiencies. All the authors agree, however, that specific, culture-based preferences and
biases are indeed important factors in risk perception. The disagreement is about the relevance
of the postulated four prototypes within the realm of cultural factors.
All four levels of influencing factors are relevant to a better understanding of chemical risk
perception. The comparative studies show that universal yardsticks to evaluate chemical risks
do exist, at least to a certain degree, but that their relevance and their selective use in
assigning these characteristics to specific risk situations are contingent on cognitive and
affective patterns as well as on the social, political and cultural environments.
In spite of many open questions and ambiguities in risk perception research, one conclusion is
beyond any doubt: abstracting the risk concept to a rigid formula and reducing it to probability
and consequences does not match people’s intuitive feeling of what is important when making
judgments about the acceptability of risks. Rather than evaluating risk with a single formula,
most people use a set of multiple attributes.
8. Conclusion
This list of individual and social factors that shape risk perception demonstrates that the
intuitive understanding of risk is a multidimensional concept and cannot be reduced to the
product of probabilities and consequences alone (ALLEN 1987). Although risk perceptions
differ considerably among social and cultural groups, the multi-dimensionality of risk and the
integration of beliefs related to risk, the cause of risk, and its circumstances into a consistent
belief system appear to be common characteristics of public risk perception in almost all
countries in which such studies have been performed. This is not to say that quantitative
assessment results do not matter for people’s perception but they are only one element among
many others that shape the formation of attitudes towards risks and judgements about their
acceptability (BOHOLM 1998).
41
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
Risk perception studies have revealed the various elements that shape the individual and
social experience of risk. First, individual and social risk experience appears to be stronger
related to exposure than to actual casualties on which most risk assessment is based (RENN,
BURNS, KASPERSON, KASPERSON & SLOVIC 1992). An exposure of few people
resulting in several casualties is likely to be less influential for risk perception and public
response than an exposure of many people resulting in minor injuries or only few casualties.
Second, the processing of risk by the media, social groups, institutions, and individuals shapes
the societal experience with risk and plays a crucial role in determining the overall intensity
and scope of societal impacts. Press media coverage, perception of dread, and the individual
intention to take action contribute substantially to an event’s socio-economic impacts even
when controlling for the extent of direct harm to people and property. This finding supports
the theoretical claim in the beginning of this article that risk reflects both, real harm and social
constructions.
Third, individual perception is highly influenced by qualitative risk characteristics and
semantic images. These images constitute tools of reducing complexity by providing easily
identifiable cues for ordering new risks into one of four images: pending danger, slow agents,
cost-benefit ratio, and avocational thrill. These images are internalized through cultural and
social learning. They cluster around qualitative variables that specify the context and the
situation in which the risk manifests itself within each image. Although these variables allow
a certain degree of abstraction with respect to perceiving and evaluating risks across different
risk sources, they provide sufficient contextual specification for making the necessary
semantic distinctions between negligible, serious, and unacceptable risks. Rather than
evaluating risk with a single formula, most people use a set of multiple attributes, many of
which make normative sense. The application of qualitative characteristics seems to be a
universal mechanism of risk perception which has been discovered in many countries, their
specific content and relative importance depend, however, on the cultural context in which the
individual is raised.
Fourth, among the perception variables, dread, personal control, familiarity, and blame, are
good predictors for risk perception in most countries. This has been confirmed by empirical
investigations in the United States, Germany, France, Canada, Austria, Australia, Japan and
many others (ROHRMANN & RENN 2000). The degrees to which these qualitative
characteristics are assigned to specific risk sources depend on cultural context, but also on the
media coverage. In contrast to popular belief, there is no evidence that media create opinions
about risks or even determine risk perceptions. The studies on media perception suggest
instead that people select elements from media reports and use their own frame of reference to
create understanding and meaning. This creation is certainly influenced by the cognitive
content and the evaluative statements of the reports, but the influence is more indirect and
ambiguous than many people think. Most people reconfirm existing attitudes when reading or
viewing media reports.
Fifth, in a social environment in which personal experience is largely constructed by secondhand information, trust is an essential prerequisite for communication and social coordination.
In risk arenas, trust can easily be destroyed by non-predictable or non-avoidable disasters; at
the same time it can easily be exploited by referring to random events as explanations or
excuses for risk management errors or hubris (LÖFTSEDT 2005). As both processes occur
simultaneously, trust is constantly at stake in institutional responses to risk. One way to cope
with this problem is to develop an organizational style that either leads to high-reliability
42
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
performance or to immunization against public scrutiny. In addition, the ambiguity in
assigning causality or blame to various actors (including nature or God) makes risk an ideal
issue for political arenas. Struggling for social resources, actors in risk arenas can mobilize
public support by placing blame and responsibility to other actors. They know that the
probabilistic nature of risk prevents any of the actors to come up with compelling evidence
that they are correct in their assessments or can justify their claims. This ambiguity has two
consequences: resources other than evidence become major chips in the game for more social
influence in risk debates and the position of rule enforcement agency tends to be weak as they
struggle for the best evidence available.
What benefits can scientists and policymakers gain from the study of risk perception? What
guidance can be derived from studies on intuitive risk perception for risk and technology
policymaking? Even if there are no recipes to be derived from the current situation, analysis
of risk perception still provides a number of lessons for policymakers, especially if one
accepts pragmatic technology and risk policy as normative objectives.

Science-based risk assessment is a beneficial and necessary instrument of pragmatic
technology and risk policy. It is the only means by which relative risks can be compared
and options with the lowest statistical expectations selected. However, it cannot and should
not be used as an exclusive guide for public action. The price for its universality is
abstraction from context and the overshadowing of other rational and meaningful
perception characteristics. Without taking context and situation-specific supporting
circumstances into account, decisions will not, in a given situation, meet the requirement of
achieving collective objectives in a rational, purposeful and value-optimising manner.
 Context and supporting circumstances are significant characteristics of risk perception.
These perception patterns are not just individual perceptions cobbled together: they stem
from cultural evolution, are tried and trusted concepts in everyday life and, in many cases,
control our actions in much the same way as a universal reaction to the perception of
danger. Their universal nature across all cultures allows collective focus on risk and
provides a basis for communication (ROHRMANN 1995).
 From a rational standpoint, it would appear useful to systematically identify the various
dimensions of intuitive risk perception (concerns assessment) and to measure those
dimensions against prevailing, empirically derived characteristics. Many psychometric
variables that matter to people are open to scientific study and scrutiny. In principle, the
extent to which different technical options distribute risk across the various groups of
society, the extent to which institutional control options exist and to what extent risk can be
accepted by way of voluntary agreement can all be measured using appropriate research
tools. Risk perception supplies lessons in the need to incorporate these factors into
policymaking. This is based on the view that the dimensions (concerns) of intuitive risk
perception are legitimate elements of rational policy, but assessment of the various risk
sources must follow rational, scientific procedures on every dimension.
 Risk perception is no substitute for rational policy. Just as technical risk assessment
should not be made the sole basis for decision-making, factual assessment of risk should
not be made the political measure of its acceptability. If we know that certain risks, like
passive smoking, can lead to serious illness, then policies to reduce risk can certainly be
normatively justified even if there is a lack of awareness of the problem among the general
public. Many risks are ignored because no-one wants to deal with them. This applies
especially to risks that are triggered by natural forces. To allow oneself to be guided by
ignorance or obviously false perceptions hardly meets the prescription of pragmatic risk
and technology policy. Knowledge of these perception patterns can, however, be used to
43
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
structure and implement informational and educational measures in a beneficial manner.
The inability of many people to understand probabilistic statements or to recognize the
long-term risk from familiar risk sources is surely one problem area in which targeted
education and information can be of benefit (RENN & LEVINE 1988). What is really
needed is mutual enhancement between technical risk assessment and intuitive risk
perception.
 Little is known about how society perceives combined effects. Preliminary empirical
study indicates that the majority of respondents perceive the combined effects of noxious
substances in the environment as being super-additive. The explanations and
interpretations contained in this chapter assume that exposure to combined noxious
substances triggers a particularly high degree of concern and that scientists’ ability to allay
that concern is limited. Such concern leads to a cautious approach which sends out an
important signal for the need to protect ourselves against nasty “surprises”. It also acts as a
vent for expression of the anxiety caused by the automation and chemicalization of
everyday life. But while such anxiety can be alleviated to an extent, no amount of
information or education can dispel it altogether. Once relationships have been recognized,
it is up to policymakers to combine scientific expertise on the possible outcomes and
remaining uncertainties with the assessments and desires for change of those affected by
the risks in question, and to integrate them into a holistic policy that is both knowledge and
values based. Risk policy should neither be purely science-based nor purely value-based.
 Option trade-offs require policy weighting between the various target dimensions. Such
trade-offs are dependent both on context and on the choice of dimension. Perception
research offers important pointers concerning the selection of dimensions for focus. The
aspect of fairness that rates high among people as an evaluation tool for the acceptability of
risks plays a significant role in such trade-offs and in weighting the various dimensions. In
their roles as scientists, experts have no authority to take such things into account. This is
where risk comparison reaches its limits. Even if we remain within the semantic context
accepted by most people – a pool of comparable risks – multidimensionality in the intuitive
risk model and variable targets in risk management prevent risk policy from focusing onesidedly on minimization of expected impacts. A breach of the minimization requirement,
however, implies acceptance of greater damage than is absolutely necessary (although this
can be justified in individual cases depending on the risk situation).
The inclusion of risk perception studies for normative use in policy making faces two major
drawbacks. First, the advice of social scientists will vary considerably depending on the
perspective that the scientists prefer. Second, unlike the technical or economic understanding
of risk, the social science concepts offer no common denominator for measuring cultural or
social acceptability (DOUGLAS 1985). What constitutes a major risk for one group may be
perfectly in line with the perceptions of another group. Who is going to decide which risk
perception has more validity than another perception?
There is also no impartial referee available to judge the appropriateness of risk perceptions.
Science may help to determine the magnitude of the risk but this information alone is not
sufficient to make decisions about the acceptability of risks. The only viable resolution of
these conflicts in democratic societies is by initiating a discourse among the major parties
involved in the decision making process or affected by the decision outcomes (HABERMAS
1971). Such a dialogue can be organized in form of advisory committees, citizen panels,
formal hearings, and others (FIORINO 1989; WEBLER & RENN 1991). Risk
communication and participation are integrative parts of risk pre-assessment, risk appraisal,
44
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
risk judgement and management. Discursive elements need to be inserted in risk assessment
as well as in risk management although the formats and the type of input from stakeholders
and the public at large may differ in both steps (IRGC 2005).
Risk communication can be organized as providing information to the public, as a mutual
learning process, or as an attempt to reconciling conflicts about risks. The empirical study of
risk communication has demonstrated the need for trust and credibility between the
communicators and their audience, but they also revealed the continuous trend towards
distrust and suspicion on the side of those who are supposed to bear the risks. What can be
learned from the evidence collected so far is that credibility of a communication source is
closely linked to the perceived past performance record and its openness for public demands.
The more institutions comply with the expectations of the public, the more confidence people
will have in these institutions and the more trust they will assign to their messages.
Communication efforts alone may be successful to change excessive aspirations or to correct
misperceptions of the actual performance record, but it is more than unlikely that
communication can compensate poor performance.
Risk communication is a necessary step towards bridging the gap between risk assessment,
management and perception. The goal of risk communication should not be to persuade
people to accept whatever the communicator thinks is best for them. The ideal communication
program envisions a receiver who processes all the available information to form a wellbalanced judgment in accordance with the factual evidence, the arguments of all sides, and
his/her own interests and preferences. Risk communication is therefore a crucial element of
any risk assessment and management strategy. It is aimed at informing appropriating, and
supporting public formation of judgments and evaluations. The ultimate goal is to reconcile
expertise, interests, and public preferences.
45
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
9. Bibliography
Adams, W. C. (1986): “Whose Lives Count?: TV Coverage of Natural Disasters”,
Communication, 36, 2, 113-122.
Allen, F. W. (1987): "Towards a holistic appreciation of risk: the challenge for
communicators and policymakers," Science, Technology, and Human Values, 12, Nos.
3 and 4, 138-143
Barber, B. (1983): The logic and limits of trust. New Brunswick: Rutgers University Press
Bechmann G (1993): Risiko als Schlüsselkategorie der Gesellschaftstheorie. In: G. Bechmann
(ed.) Risiko und Gesellschaft. Grundlagen und Ergebnisse interdisziplinärer
Risikoforschung. Westdeutscher Verlag, Opladen, pp. 237-276.
Beck, U. (1986): Die Risikogesellschaft: Auf dem Weg in eine andere Moderne. Frankfurt/
Main: Suhrkamp.
Beck, U. (1992): Risk Society: Toward a New Modernity. Translated by Mark A. Ritter
(Sage: London 1992)
Beck, U. (1994): “The Reinvention of Politics: Towards a Theory of Reflexive
Modernization,” in: U. Beck, A. Giddens and S. Lash (eds.): Reflexive Modernization.
Politics, Tradition and Aesthetics in the Modern Social Order (Stanford University
Press: Stanford 1994), 1-55.
Bergesen and Warr (1979). A Crisis in the Moral Order: The Effects of Watergate Upon
Confidence in Social. Academic Press, 277-295.
Boholm, A. (1998): Comparative Studies of Risk Perception: A Review of Twenty Years of
Research, Journal of Risk Research, 1, No. 2 (1998), 135-163
Brehmer, B. (1987): The Psychology of Risk. In: Singleton WT and Howden J (eds.), Risk
and Decisions Wiley, New York, pp. 25-39.
Buss, D., Craik, K. (1983): Contemporary Worldviews: Personal and Policy Implications,
Journal of Applied Psychology, 13, 259-280.
Chaiken, S. and Stangor, C.(1987): "Attitudes and attitude change," Annual Review of
Psychology, 38, 575-630
Clarke, L. (1989): Acceptable risk: making decisions in a toxic environment. Berkeley, CA:
University of California Press.
Conard, Bruce R. (2005): Thermodynamics, Kinetics, Chaos and Communication.
Presentation to MDEC, Oct 12, 2005, Toronto
Covello, V.T. (1983): "The perception of technological risks: a literature review,"
Technological Forecasting and Social Change, 23, 285-297.
46
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
Covello, VT. (1992): Trust and Credibility in Risk Communcation, Health and Environmental
Digest, 6, 1, 1-3.
Dake, K. (1991), "Orienting Dispositions in the Perceptions of Risk: An Analysis of
Contemporary Worldviews and Cultural Biases," Journal of Cross-Cultural Psychology,
22, 61-82
Douglas, M. (1985): Risk acceptability according to the social sciences. New York: Russell
Sage Foundation
Douglas, M. (1990): “Risk as a Forensic Resource,” in: DEADALUS, 119, No, 4 (Fall 1990),
1-16
Douglas, M. and Wildavsky, A. (1982): Risk and Culture (University of California Press:
Berkeley 1982)
Drottz-Sjöberg, B.-M. (1991): Perception of risk. Studies of risk attitudes, perceptions, and
definitions. Stockholm: Center for Risk Research
Dunwoody, S. (1992): "The media and public perception of risk: How journalists frame risk
stories." In: D.W. Bromley and K. Segerson (eds.), The Social Response to
Environmental Risk: Policy Formulation in an Age of Uncertainty. Dordrecht: Kluwer
Durkheim, E. (1933): The division of labor in society. New York: The Free Press (orig.
1893)
Edwards, W. & von Winterfeldt, D. (1987): Public values in risk debates. Risk Analysis, 7,
pp. 141-158.
Festinger, L. (1957): A theory of cognitive dissonance. Stanford: Stanford University Press
Fiorino, D.J. (1989): "Technical and democratic values in risk analysis," Risk Analysis, 9, No.
3, 293-299
Fischhoff, B., Lichtenstein, S., Slovic, P. Derby, S.L. and Keeney, R.L. (1981): Acceptable
risk . Cambridge, MA: Cambridge University Press
Fischhoff, B, Watson, SR. and Hope C. (1984): Defining Risk, Policy Sciences, 17 pp. 123129.
Freudenburg, W.R. (1988): Perceived risk, real risk: social science and the art of probabilistic
risk assessment, Science, 242, 44-49
Frewer, L.J., Miles, S., Brennan, M., Kusenof, S., Ness, M., and Ritson, C. (2002). Public
preferences for informed choice under conditions of risk uncertainty. Public
Understanding of Science, 11(4), 1-10.
Giddens, A. (1990): The Consequences of Modernity. Stanford: Stanford University Press.
Giddens, A. (1991): Modernity and Self-Identity. Stanford University Press.
47
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
Grendstad, G. (2000): Grid-Group Theory and Political Orientations: Effects of Cultural
Biases in Norway in the 1990s. Scandinavian Political Studies 23 (3), 217-244
Habermas, J. (1971): Toward a rational society. London: Heinemann
Heimer, C. (1988): “Social structure, psychology, and the estimation of risk,” Annual Review
of Sociology, 14, 491-519
Henschler D (1993): Krebsrisiken im Vergleich, Folgerungen für Forschung und politisches
Handeln. In: Forschungszentrum für Umwelt und Gesundheit (GSF): Mensch und
Umwelt: Risiko, issue 8, p. 70.
Hsee, C. and Kunreuther, H. (2000): “The Affection Effect in Insurance Decisions,” Journal
of Risk and Uncertainty, 20 (2000), 141-59
IAEA (1995): Guidelines for Integrated Risk Assessment and Management in Large
Industrial Areas, Technical Document: IAEA-TECDOC PGVI-CIJV. (International
Atomic Energy Agency: Vienna 1995)
IEC (1993): Guidelines for Risk Analysis of Technological Systems, Report IEC-CD (Sec)
381 issued by the
IPCS and WHO (2004): Risk Assessment Terminology. (World Health Organization: Geneva
2004)
IRGC (2005): White Paper on Risk Governance: Towards an Integrative Framework. Geneva:
Intenrational Risk Governance Council 2005
Jaeger, W. (2002): Gefahrensituationen meistern. In: WLB Wasser, Luft und Boden –
Zeitschrift für Umwelttechnik 8-9/2002).
Johnson, B.B. and Covello, V.T. (Eds.) (1987): The social and cultural construction of risk.
Boston: Reidel.
Jungermann H, Schütz H. and Thüring M. (1988): Mental Models in Risk Assessment:
Informing People about Drugs. Risk Analysis, 8, 147-155.
Jungermann, H. and Slovic, P. (1993): "Charakteristika individueller Risikowahrnehmung."
In: W. Krohn and G. Krücken (eds.), Riskante Technologien: Reflexion und Regulation.
Franfurt/Main: Suhrkamp
Jungermann H. and Slovic P (1993a): Charakteristika individueller Risikowahrnehmung. In:
Bayerische Rückversicherung (ed.) Risiko ist ein Konstrukt. Wahrnehmungen zur
Risikowahrnehmung Knesebeck, München, pp. 89-107.
Jungermann, H., Rohrmann, B. and P.M. Wiedemann (1991): Risiko-Kontroversen Konzepte, Konflikte, Kommunikation. Springer Verlag Berlin
Kahneman D. and Tversky A. (1979): Prospect Theory: An Analysis of Decision Under Risk,
Econometrica, 47, 263-291.
48
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
Kasperson, R.E. and Kasperson, J.X. (1983): Determining the Acceptability of Risk: Ethical
and Policy Issues. In: J.T. Rogers and D.V. Bates (eds.), Assessment and Perception of
Risk to Human Health, Conference Proceedings. Royal Society of Canada: Ottawa, pp.
135-155
Kasperson, R E., Renn, O., Slovic P., Brown, H.S., Emel, J., Goble, R., Kasperson, J.X. and
Ratick, S. (1988): "The social amplification of risk: a conceptual framework," Risk
Analysis, 8, 177-187.
Katz, D., Gatek, B.A., Kahn, R.L. and Barton, E. (1975): Bureaucratic encounters. Institute
for Social Research, Ann Harbor: University of Michigan
Knight, A. and Warland, J. (2005): Determinants of Food Safety Risks: A Multi-disciplinary
Approach. Rural Sociology 70(2), pp. 253–275
Köcher R. (1986): Bloodhounds or Missionaries: Role Definitions of German and British
Journalists. European Journal of Communication, 1, 43-64.
Kraus N., Malmfors T. and Slovic P. (1992): Intuitive Toxicology: Expert and Lay Judgments
of Chemical Risks. Risk Analysis, 12, 215-232.
Lee, T.R. (1981), “The public perception of risk and the question of irrationality.” In: Royal
Society of Great Britain (ed.), Risk perception , Vol. 376 (5-16), London: The Royal
Society
Lee, T.R. (1986): "Effective communication of information about chemical hazards," The
Science of the Total Environment, 51, 149-183.
Lichtenberg, J., and D. MacLean (1988): "The Role of the Media in Risk Communications".
In: Risk Communications: Proceedings of the International Workshop on Risk
Communications, H. Jungermann, et al. eds.,Kernforschungsanlange, Julich, France. pp.
33-48.
Linnerooth-Bayer, J. and K. Fitzgerald (1996): Conflicting Views on Fair Siting Processes:
Evidence from Austria and the U.S. Health, Safety & Environment, 7:2:119-134.
Lipset, S.M., and Schneider, W. (1983): The confidence gap, business, labor, and government
in the public mind. New York: The Free Press.
Loewenstein, G., Weber, E., Hsee, C., Welch, E. (2001): “Risk as Feelings,” Psychological
Bulletin, 127 (2001), 267-86
Löftsedt, R. (2005): Risk Management in Post Trust Societies. London: Palgrave Macmillan
Lopes, L.L. (1983): “Some thoughts on the psychological concept of risk,”Journal of
Experimental Psychology: Human Perception and Performance, 9, 137-144
Lowery, Sharon and DeFleur, Melvin (1983): Milestones in mass communication research.
New York: Longman
49
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
Luce, R.D. and Weber, E.U. (1986): “An axiomatic theory of conjoint, expected risk,” Journal
of Mathematical Psychology , 30, 188-205
Luhmann, N. (1973): Vertrauen: Ein Mechanismus der Reduktion sozialer Komplexität,
Second Edition. Stuttgart: Enke
Luhmann, N. (1980): Trust and power. New York: Wiley.
Luhmann, N. (1986): Ökologische Kommunikation. Opladen: Westdeutscher Verlag.
Luhmann, N. (1990): “Technology, environment, and social risk: a systems perspective,”
Industrial Crisis Quarterly, 4, 223-231
Luhmann, N. (1993): Risiko und Gefahr. In: Krohn W, Krücken G (eds.) Riskante
Technologien: Reflexion und Regulation. Suhrkamp Frankfurt am Main, pp. 138-185.
(English version: Luhmann, N. 1990: Technology, Environment, and Social Risk: A
Systems Perspective. Industrial Crisis Quarterly, 4: 223-231
Machlis, G.E. and Rosa, E.A. (1990): “Desired risk: broadening the social amplification of
risk framework, Risk Analysis, 10, 161-168.
MacLean, D. (1986): "Social Values and the Distribution of Risk," in: D. Maclean (ed.),
Values at Risk (Totowa: Rowman and Allanheld), pp. 75-93
Marris, C., and Joly, P.-B. (1999): Between consensus and citizens: public participation in
technological decision-making in France. Science Studies vol. 12, n 2, 1999, pp 3-32
Matejko, A.J. (1988): the sociotechnical perspective on trust at work. Speyrer Arbeitshefte,
86:1-251
Mazur A. (1987): Putting Radon on the Public’s Risk Agenda. Science, Technology, and
Human Values, 12, 3 and 4, 86-93.
Midden, C., (1988): Credibility and Risk Communication. Manuscript for the International
Workshop on Risk Communication, October, 17-21 1988. Research Center KFA: Jülich
Nelkin, D. (1982): "Blunders in the business of risk," Nature, 298, 775-776
Nennen HU. and Garbe D. (eds.) (1996): Das Expertendilemma. Springer Berlin.
NRC (1983): Risk assessment in the federal government: Managing the process. National
Res. Council of the National Academies. Natinal Academic Press, Washington,DC
O'Riordan, T. (1983): "The cognitive and political dimension of risk analysis," Environmental
Psychology, 3, 345-354
Otway, H. and Thomas, K.(1982): “Reflections on risk perception and policy,” Risk Analysis,
2, 69-82
Parsons, T.E. (1960): "Pattern variables revisited," American Sociological Review, 25, 467483.
50
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
Peltu, M. (1985): The Role of Communications Media. in: Otway H, Peltu M (eds.)
Regulating Industrial Risks. Butterworth, London, pp. 128-148.
Peltu, M. (1989) Media Reporting of Risk Information: Uncertainties and the Future. In:
Jungermann H, Kasperson RE, Wiedemann PM (eds.) Risk Communication.
Forschungszentrum KFA, Jülich, pp.11-32.
Peters, HP. (1990): Technik-Kommunikation: Kernenergie. in: Jungermann H, Rohrmann B,
Wiedemann
PM
(eds.)
Technik-Konzepte,
Technik-Konflikte,
TechnikKommunikation. Monographien des Forschungszentrums Jülich, Vol. 3
Forschungszentrum Jülich, Jülich, pp. 59-148.
Peters, HP. (1991) Durch Risikokommunikation zur Technikakzeptanz? Die Konstruktion
von Risiko”wirklichkeiten” durch Experten, Gegenexperten und Öffentlichkeit. In:
Krüger J, Ruß-Mohl St (eds.) Risikokommunikation. Edition Sigma, Berlin, pp. 11-67.
Peters, H.P. (1994): "Wissenschaftliche Experten in der öffentlichen kommunikation über
Technik, Umwelt und Risiken", in: Kölner Zeitschrift für Soziologie und
Sozialpsychologie
Peters, HP. (1996): Kommentar zu Hans Mohrs Studie über das “Expertendilemma”. In:
Nennen HU, Garbe D (eds.) Das Expertendilemma. Zur Rolle wissenschaftlicher
Gutachter in der öffentlichen Meinungsbildung. Springer, Berlin, pp. 61-74.
Petty, R.E. and Cacioppo, E. (1986): "The elaboration likelihood model of persuasion,"
Advances in Experimental Social Psychology, 19, 123-205
Pidgeon, NF., Hood, CC., Jones, DKC., Turner, BA., Gibson, R. (1992): Risk Perception. In:
Royal Society Study Group (ed.) Risk Analysis, Perception and Management. The
Royal Society, London, pp. 89-134.
Pollatsek, A. and Tversky, A. (1970): “A theory of risk,” Journal of Mathematical
Psychology, 7, 540-553
Rayner, S. (1990): Risk in cultural perspective: acting under uncertainty. Norwell: Kluwer
Rayner, S.(1987): "Risk and relativism in science for policy." In: V. T. Covello and B.B.
Johnson (eds.), The social and cultural construction of risk (5-23), Dordrecht: Reidel.
Renn, O. (1986):, "Akzeptanzforschung: Technik in der gesellschaftlichen Auseinandersetzung," Chemie in unserer Zeit , 2, 44-52.
Renn, O. (1989): Risikowahrnehmung. Psychologische Determinanten bei der intuitiven
Erfassung und Bewertung von technischen Risiken. In: Hosemann G (ed.) Risiko in der
Industriegesellschaft. Universitätsbibliotheksverlag Erlangen-Nürnberg, Erlangen, pp.
167-191.
Renn, O. (1990): “Risk perception and risk management: a review,” Risk Abstracts, 7, No.1 ,
1-9 (Part 1) and 7, No.2, 1-9 (Part 2)
51
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
Renn, O. (1991): "Risk communication and the social amplification of risk. In: R. Kasperson
and P.J. Stallen (eds.), Communicating risk to the public (287-324), Dordrecht: Kluwer
Renn, O. (1992): Concepts of Risk: A Classification. in: Krimsky S, Golding D (eds) Social
Theories of Risk. Praeger, Westport, pp. 53-82.
Renn, O. (1992a): "The social arena concept of risk debates." In: S. Krimsky und D. Golding
(eds.), Social theories of risk (170-197), Westport, CT : Praeger
Renn, O. (1993): "Risikoabschätzung” und ”Risikobewertung”. In: P. Wiedemann and H.
Schütz (Hrsg.): Technik kontrovers. Aktuelle Schlüsselbegriffe für die öffentliche
Diskussion. Ein Handbuch. Frankfurt am Main (Institut für Medienentwicklung und
Kommunikation GmbH 1993), S. 181-186 und 193-196
Renn, O. (2005): Style of Using Scientific Expertise: A Comparative Framework. Science and
Public Policy, Vol. 22 (June 1995), 147-156
Renn, O. and Levine D. (1988): Trust and Credibility in Risk Communication. in:
Jungermann H, Kasperson RE, Wiedemann PM (eds.) Risk Communication.
Forschungszentrum, Jülich, p. 51.
Renn, O. and Levine, D. (1991): "Trust and credibility in risk communication.” In: R.
Kasperson and P.J. Stallen (eds.), Communicating risk to the public (175-218),
Dordrecht: Kluwer
Renn, O. and Rohrmann, B. (2000): Cross-Cultural Risk Perception. A Survey of Research
Results. Kluwer, Dordrecht
Renn, O., Burns, W., Kasperson, R.E., Kasperson, J.X. and Slovic, P. (1992): The Social
Amplification of Risk: Theoretical Foundations and Empirical Application. Social
Issues, Vol. 48, No. 4, Special Issue: Public Responses to Environmental Hazards
(1992), 137-160
Renn, O., Kastenholz, H. (2000): Risk Communication for Chemical Product Risk. An OECD
Background Paper, Berlin: BGVV.
Renn, O., Webler, Th. und Johnson, B.: Public Participation in Hazard Management (1991):
The Use of Citizen Panels. Risk - Issues in Health and Safety. Vol. 2, No. 3 (1991), S.
197-226
Rohrmann, B. (1995). Technological risks - Perception, evaluation, communication; in: R.E.
Melchers & M.G. Stewart (Eds.): Integrated Risk Assessment: new directions;
Rotterdam: Balkema.
Rohrmann, B. (1999): Risk perception research - Review and documentation; Research
Center Juelich: RC Studies 68.
Rohrmann, B. (2000): Cross-national studies on the perception and evaluation of hazards. In:
Ortwin, R. and Rohrmann, B. (eds.) (2000): Comparative Risk Perception. Dorbrecht,
The Netherlands. Kluwer.
52
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
Rohrmann, B. and Renn, O. (2000): Risk perception research - An introduction. In: Renn, O.
and Rohrmann, B. (2000): Cross-Cultural Risk Perception. A Survey of Research
Results. Kluwer, Dordrecht
Rosa, E.A. (1998): “Metatheoretical Foundations for Post-Normal Risk,“ Journal of Risk
Research, 1 (1998), 15-44.
Rosa, Eugene A., Noriyuki Matsuda, and Randall R. Kleinhesselink (2000): The Cognitive
Architecture of Risk: Pancultural Unity or Cultural Shaping?", in Ortwin Renn and
Bernd Rohrmann (eds.): Comparative Risk Perception. Dorbrecht, The Netherlands.
Kluwer., pp.185-210
Ross, L.D. (1977): “The Intuitive Psychologist and His Shortcomings: Distortions in the
Attribution Process,” in: L. Berkowitz (eds.): Advances in Experimental Social
Psychology, Vol.10 (Random House: New York 1977), 173-220.
Rourke, F.E., Free, L. and Watts, W. (1976): Trust and Confidence in the American System.
Washington, DC: Potomac Assoc.
Rubin, DM. (1987): How the News Media Reported on Three Mile Island and Chernobyl.
Communication, 37, 3, 42-57.
Schwarz, M. and Thompson, M. (1990): Divided we stand: redefining politics, technology,
and social choice. Philadelphia: University of Pennsylvania Press.
Short, J.F. (1984): The social fabric of risk: toward the social transformation of risk analysis,
American Sociological Review, 9, 711-725.
Siegrist, M., and Cvetkovich, G. (2001): Better negative than positive? Evidence of a bias for
negative information about possible health dangers. Risk Analysis, 21(1), 199-206.
Siegrist, M., Cvetkovich, G. and Roth, C. (2000): Salient value similarity, social trust, and
risk/benefit perception. Risk Analysis, 20 (3), 353-361.
Singer, E. and Endremy, P. (1987): Reporting Hazards: Their Benefits and Costs.
Communication, 37, No.3 (Summer), 10-26
Sjöberg, L. (Ed.) (1987): Risk and society. Studies in risk taking and risk generation. Hemel
Hempstead, England: George Allen and Unwin.
Sjöberg, L. (1994): Perceived Risk vs. Demand for Risk Reduction. Risk Research Report No.
18. Center for Risk Research, Stockholm School of Economics, Stockholm.
Sjöberg, L. (1995): Workshop 2. Personality impression and credit decisions. In B. Green
(Ed.), Risk behaviour and risk management (pp. 93-96). Stockholm: Risk Research
Group at the Stockholm University.
Sjöberg, L. (1997): Risk Sensitivity, Attitude and Fear as Factors in Risk Perception. Paper at
the Annual Meeting of the Society for Risk Analysis-Europe. Manuskript. Stockholm.
53
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
Sjöberg, L., Kolarova, D., Rucai, A.-A., and Bernström, M.-L. (2000). Risk perception in
Bulgaria and Romania. In: Renn, O. and Rohrmann, B. (Eds.), Cross-cultural risk
perception. A survey of empirical studies (pp. 145-184). Dordrecht: Kluwer.
Slovic, P. (1987), "Perception of Risk," Science, 236, No. 4799, 280-285;
Slovic, P. (1991): Perceptions of risk: Paradox and challenge. In Hazmat ‘91: Proceedings
(pp. 4-3 to 4-22). Evanston, IL: Northwestern University, Transportation Center.
Slovic, P. (1992): “Perception of Risk: Reflections on the Psychometric Paradigm,” in: S.
Krimsky and D. Golding (eds.): Social Theories of Risk (Praeger: Westport and London
1992), 153-178.
Slovic, P., Finucane, E., Peters, D. and MacGregor, R. (2002): “The Affect Heuristic,” in: T.
Gilovich, D. Griffin, and D. Kahneman (eds.): Intuitive Judgment: Heuristics and
Biases. (Cambridge University Press: Boston 2002), 397-420.
Slovic, P., Fischhoff, B. and Lichtenstein, S. (1980): Facts and fears: Understanding
perceived risk. In R. Schwing & W.A. Albers, Jr. (Eds.), Societal risk assessment: How
safe is safe enough? (pp. 181-214). New York: Plenum.
Slovic, P., Fischhoff, B. and Lichtenstein, S. (1982): “Why Study Risk Perception?,” Risk
Analysis, 2 (June 1982), 83-94.
Slovic, P., Fischhoff, B. and Lichtenstein, S. (1986): The psychometric study of risk
perception. In V.R. Covello, J. Menkes, & J. Mumpower (Eds.), Risk evaluation and
management (pp. 3-24). New York: Plenum.
Slovic, P., Flynn, J, Mertz, C.K, Poumadere, M. and Mays, C. (2000): Nuclear power and the
public: A comparative studyof of risk perception in the United States and France,: In:
Ortwin, R. and Rohrmann, B. (eds.) (2000): Comparative Risk Perception. Dorbrecht,
The Netherlands. Kluwer, pp. 55-102.
Sood, R., Stockdale, G. and Rogers, EM. (1987): How the News Media Operate in Natural
Disasters. Communication, 37, 3, 27-41.
Sparks, P. and Shepherd, R., “Public Perceptions of the Potential Hazards associated with
Food-Production and Food-Consumption: An Empirical Study,” Risk Analysis, 14
(1994), 799-806).
Streffer, C., Bücker, J., Cansier, A., Cansier, D., Gethmann, C.F., Guderian, R., Hanekamp,
G., Henschler, D., Pöch, G.; Rehbinder, E., Renn, O., Slesina, M. and Wuttke, K.
(2003): Environmental Standards. Combined Exposures and Their Effects on Human
Beings and Their Environment. Berlin (Springer 2003)
Thompson, M. (1980): An outline of the cultural theory of risk, Working Paper of the
International Institute for Applied Systems Analysis (IIASA), WP-80-177, Laxenburg,
Austria: IIASA
Thompson, M., Ellis, W., and Wildavsky, A. (1990): Cultural theory. Boulder: Westview
Press
54
EU-Project NoMiracle, Deliverable no D 4.3.2., part DIA
Ortwin Renn and Christina Benighaus
US-Environmental Protection Agency (1987): Unfinished Business: A Comparative
Assessment of Environmental Problems (Washington D.C. 1987)
US-Environmental Protection Agency (1990): Reducing Risks: Setting Priorities and
Strategies for Environmental Protection. SAB-EC-90-021 (Washington, D.C. 1990)
Vogel, D. (1986): National Styles of Regulation. Ithaca, N.Y.: Cornell University Press
von Winterfeldt, D. and Edwards, W. (1984): ‘Patterns of Conflict about Risk Debates,’ Risk
Analysis, 4, 55-68
Webler, T. and Renn, O. (1991): Conditions for a Rational Discourse on Risk Between
Engineers and Social Scientists. Report to Ernst Basler and Company, Zürich,
Switzerland. (Center for Technology, Environment, and Development: Worcester, MA.
Wildavsky, A. (1988): Searching for Savety, New Brunswick.
Wildavsky, A. (1993). Vergleichende Untersuchung zur Risikowahrnehmung: Ein Anfang. In
Bayerische Rück (Ed.), Risiko ist ein Konstrukt. Wahrnehmungen zur
Risikowahrnehmung (pp. 191-211). München: Knesebeck.
Wildavsky, A. and Dake, K. (1990): “Theories of risk perception: who fears what and why?”
Daedalus, 119, No.4, 41-60
Wilkins L. and Patterson P. (1987): Risk Analysis and the Construction of News.
Communication, 37, 3, 80-92.
Young, Peyton (1993): Equity in Theory and Practice (Princeton: Princeton University Press)
Zwick M. and Renn, O. (1998): Wahrnehmung und Bewertung von Technik in BadenWürttemberg. Presentation. Akademie für Technikfolgenabschätzung in BadenWürttemberg, Stuttgart
55