Advanced Review

What shapes perceptions

of climate change?
Elke U. Weber∗

Climate change, as a slow and gradual modification of average climate conditions, is

a difficult phenomenon to detect and track accurately based on personal experience.
Insufficient concern and trust also complicate the transfer of scientific descriptions
of climate change and climate variability from scientists to the public, politicians,
and policy makers, which is not a simple transmission of facts. Instead, worldview
and political ideology, two elements of the cultural context of decisions, guide
attention toward events that threaten the desired or existing social order, and
shape expectations of change, which in turn guide the detection and interpretation
of climate events. Action that follows from climate change perceptions can be
informed by different processes. Affect-based decisions about climate change
are unlikely to motivate significant action, as politicians and the general public
are not particularly worried about climate risks, and because attempts to scare
people into greater action may have unintended negative consequences. Analysis-
based decisions are also unlikely to result in significant action, because of large
discounting of uncertain future costs of climate risks compared to the certain and
immediate costs of climate change mitigation. Rule-based decisions that determine
behavior based on moral or social responsibility may hold out the best prospects
for sustainable action.  2010 John Wiley & Sons, Ltd. WIREs Clim Change 2010 1 332–342

D espite four increasingly unanimous and urgent

reports issued by the Intergovernmental Panel
on Climate Change (IPCC) since 1990,1 national
What shapes the perception of climate change,
both of its extent or even existence and of its causes?
What accounts for the variability of such beliefs? What
and international public debate about climate change imbues perceptions of a changing climate with varying
(its likely magnitude and causes) does not seem degrees of urgency that motivate people to take
converge to a common opinion and reach a consensus protective or mitigative action? This article will review
at the rate that a rational analysis of accumulating social science evidence that speaks to these questions.
scientific evidence would suggest.2 There are multiple Anthropology,7,8 sociology,9,10 cultural psychology,11
reasons for why members of the general public, and behavioral decision research12 have all converged
politicians, and policy makers, members of the media, on a constructivist account of human perception and
and scientists may disagree about climate change,3 action that acknowledges the power of individual,
many of which are strategic and political.4 However, social, and cultural forces, without denying the reality
some of those reasons are psychological and cultural, and power of external, physical, and environmental
and related to the perception of climate change. forces. Understanding the psychological, sociological,
US National Research Council committees, e.g., the and cultural reasons for variations in the perceptions
Committee on the Human Dimensions of Global of climate change has implications for the design of
Change, have also identified people’s perception of educational and policy interventions that can lead to
global phenomena like climate change as a crucial greater convergence in beliefs and willingness to act.
contributor to both environmental problems and
possible solutions.5,6

VARIATIONS IN THE PERCEPTION OF

∗ Correspondence to: [email protected] CLIMATE CHANGE
Department of Psychology and Graduate School of Business,
Columbia University The existence and extent of climate change is a topic
DOI: 10.1002/wcc.41 of great importance to climate scientists as well as

332  2010 John Wiley & Sons, L td. Volume 1, May/June 2010
 WIREs Climate Change What shapes perceptions of climate change?

individuals, groups, and organizations with long- explanations for these deviations from expectations.
term strategic decisions that are crucially affected In a more recent example, farmers in Illinois were
by climate, projected climate change, and potential asked to recall salient temperature or precipitation
climate change mitigation measures. This includes statistics during the growing season of seven preceding
utilities and other companies in the energy and years.14 Those farmers who believed that their
transportation sector, reinsurance companies, and region was undergoing climate change recalled
governments at different levels. Average citizens are temperature and precipitation trends consistent with
typically more concerned about the weather, rather this expectation, whereas those farmers who believed
than the climate in their region.13 However, climate in a constant climate, recalled temperatures and
forecasts and beliefs about changes in existing climate precipitations consistent with that belief. Both groups
variables can also drive individual or household showed about equal amounts of error in their memory
decisions, e.g., the decision to buy a beachfront for salient weather events, but the direction of the
property or whether to support collective action at errors was biased to be in line with farmers’ beliefs
local or national levels to mitigate greenhouse gas and expectations.
emissions.14 Recently, behavioral researchers have shown
differences in the way people learn about uncertain
phenomena or environments from personal experience
Learning about Climate Change from versus from being provided with a statistical (typically
Personal Experience numeric or graphic) description of possible out-
Despite its environmental, social, and economic comes and their likelihood.16 This distinction between
importance, climate change is a phenomenon that learning from experience versus learning from
is not easily and accurately identified by the lay description has received much attention because
public, using their normal tools of observation and ostensibly the same information about events and their
inference. Climate is a statistical phenomenon, a likelihoods can lead to very different perceptions and
term that describes average weather conditions or actions.17 Learning from repeated personal experience
their typical range for a region. Climate change in with outcomes involves associative and often affective
the meteorological sense refers to systematic (yet processes, which are fast and automatic. Learning
usually gradual) changes in average conditions, i.e., to from statistical descriptions, on the other hand,
reliable trends embedded in the random fluctuations requires analytic processing that needs to be acquired
of conditions that can be expected for both stable and requires cognitive effort. When given the choice
and changing climates. Observations are spaced in between attending to information provided in the form
time, and memory of past events can be faulty.14 of statistical summaries or to information provided by
As a result, climate change is not easily detected personal experience, personal experience is far more
by personal experience, even though it appears to likely to capture a person’s attention, and its impact
be open to personal observation and evaluation, as dominates the often far more reliable and diagnostic
most people consider themselves to be experts on the statistical information.18
weather and do not differentiate very strictly between People’s choices under risk and uncertainty
climate (the statistical expectation) and weather can differ in important ways under the two
(what we get). The climate of a region (and changes information conditions, mostly and especially when
in its climate) obviously determines weather. People small-probability events are involved. For better
often falsely attribute unique events to climate change or worse, most extreme weather or climate events
and also fail to detect changes in climate. (frosts, flooding, or droughts) that are the result
Expectations of change (or stability) play a large and indicative of climate change still occur with
role in people’s ability to detect trends in probabilistic only small probability at this point in time in
environments, as illustrated by the following historic most regions of the world. Insights into how these
climate example (Kupperman,15 as reported in rare events will be processed and incorporated
Refs 13, 14). English settlers who arrived in North into perceptions of climate change is thus highly
America in the early colonial period assumed that relevant. The evaluation of risky options under
climate was a function of latitude. Newfoundland, the repeated sampling in decisions from experience
which is south of London, was thus expected to follows classical reinforcement learning that gives
have a moderate climate. Despite repeated experiences recent events more weight than distant events.16
of far colder temperatures and resulting deaths and Such updating is adaptive in dynamic environments,
crop failures, colonists clung to their expectations where circumstances might change with the seasons
based on latitude, and generated ever more complex or according to other cycles or trends. Because rare

Vo lu me 1, May /Ju n e 2010  2010 Jo h n Wiley & So n s, L td. 333

 Advanced Review wires.wiley.com/climatechange

events have a smaller probability of having occurred cases where the outputs from the two processing
recently, they tend (on average) to have a smaller systems disagree, the affective, association-based
impact on the decision than their objective likelihood system usually prevails, because it comes in faster
of occurrence would warrant. When they do occur, and is more vivid. For most people in the United
however, they have a much larger impact on related States, judging the likelihood of climate change based
decisions than warranted by their probability. This on personal experience will lead to judgments that
makes learning and decisions from experience more are low. Even individuals whose economic livelihood
volatile across respondents and past outcome histories depends on weather and climate events (e.g., farmers
than learning and decisions from description.19 The or fishers) and who are thus motivated to attend
public’s perceptions of climate change over time to them, might not receive sufficient feedback from
(as reflected by Gallup or Pew opinion polls, e.g., their daily or annual personal experience to diagnose
Refs 20, 21) seem to reflect a general under-concern systematic changes sufficient to become alarmed.
and greater volatility than warranted by scientific Surveys conducted in Alaska and Florida, where
evidence. Existing laboratory studies of these learning- residents in some regions have increasingly been
from-experience processes (e.g., Ref. 16) have used exposed to physical evidence of climate change show
repeated choices between cards from card decks that that such personal exposure greatly increases their
differed in the likelihood and magnitude of wins concern and willingness to take action.27,28
or losses. The relevance of these observed learning- Personal experience alone may not be a panacea.
from-experience processes with their strong recency Experienced adverse consequences need to be seen as
bias to the perception of climate change still awaits causally connected to the phenomenon or mechanism
systematic investigation. However, anecdotal evidence whose perceived likelihood and severity is being
and general observation suggest that the general studied, to lead to increases in concern or worry.
pattern of generally underweighting small-probability The absence of such perceived connection might
climate events, coupled with a noticeable volatility in be responsible for instances where greater personal
risk perception as the result of recent events (e.g., a experience of adverse events (i.e., flooding) did not
hotter or colder than usual season) is consistent with result in increased concern about climate change.29
learning from personal experience.
The distinction between decisions from personal
experience versus from statistical description maps Learning about Climate Change from
onto another, more general distinction in cognitive,22 Statistical Description
social,23 and clinical psychology,24 namely between Because climate change is so hard to detect and judge
processing of information in more associative and accurately based on personal experience, one might
affect-driven ways versus more analytic ways. Asso- argue that its detection should be left to experts,
ciative processing is a very basic human ability that namely climate scientists, and to their social amplifiers,
does not need to be learned and cannot be turned the media and educators. Such delegation makes
off. Associations are made very quickly and automat- climate change a phenomenon for whose existence
ically. They turn experienced adverse aspects of the and likely magnitude and time course people have
environment into feelings of fear, dread, or anxiety, to rely on their beliefs in scientific observation and
which then influence decisions.25 Analytic processing, modeling, in expert judgments, and/or on reports
on the other hand, works by algorithms and rules about all of these in the mass media. Indeed, most
(e.g., the probability calculus, Bayesian updating, for- people’s knowledge and exposure to climate change
mal logic, and utility maximization) that must be has been almost entirely indirect and virtual, mediated
learned explicitly. Analytic processes are slower and by news coverage and film documentaries of events in
require conscious effort and control. The perceptions distant regions (such as melting ice sheets in Greenland
of climate change and its risks by climate scientists or Antarctica) that ascribe these events to climate
are based in large part on analytic processing, as change, events and arguments for which people’s
these experts have been trained as scientists in the personal experience does not provide concurring
necessary analytic tools and have the necessary input evidence. Reliance on external sources of evidence
required for these tools. Nonscientists, on the other and expertise to form beliefs about climate change
hand, typically rely more on the more readily available raises two important issues, attention and trust.
associative and affective processing of climate-related Attention is a very scarce cognitive resource.12
information that comes their way. Unlike money or other material resources, which
The two types of processes tend to operate can be saved or borrowed, the amount of attention
in parallel and also interact with one another.26 In available to anyone to process the vast amount

334  2010 John Wiley & Sons, L td. Volume 1, May/June 2010
 WIREs Climate Change What shapes perceptions of climate change?

of information potentially available on innumerable their decisions and actions, if it comes from
topics is small and very finite. Statistical evidence a trusted source73 . The importance of trust for
provided by scientists and anecdotal accounts of the use of climate information has been well
climate change provided by the media only become documented in the context of climate variability,
information that influences subsequent perceptions where seasonal to interannual climate forecasts are
and behaviors when the general public attends to often provided by multiple (commercial as well as
them. Everyday life provides plenty of competing non-commercial) sources, and are only used when
targets for attention, such as economic survival or provided by trusted intermediaries.13 Different user
family problems. Climate change typically ranks low groups put their trust into different organizations,
to last among the concerns that Americans are from national Meteorological Services to independent
questioned about.20 farm organizations.37,38 Even though the IPCC was
People’s fundamental values and worldviews30 founded as an ostensibly nonpartisan scientific body
influence which phenomena and risks they attend by the United Nations Environment Programme
to and which they ignore or deny. Douglas and (UNEP) and the World Meteorological Organization
Wildavsky7 identify five distinct ‘cultures’ or dis- (WMO) to summarize the scientific evidence about
tinct interpretive communities (labeled hierarchical, climate change and its potential environmental and
individualist, egalitarian, fatalist, and hermitic, respec- socio-economic consequences, there is of course no
tively) that differ in their endorsed patterns of inter- value-free way of framing an issue and of selecting and
personal relationships in ways that affect perceptions presenting information about it.3 Factual information
of risk. Perceptions about the existence of climate interacts with social, institutional, and cultural
change, its causes, and likely consequences are socially processes in ways that may amplify or attenuate
constructed within these communities that are pre- public perceptions of risks like climate change.39
disposed to attend to, fear, and socially amplify This social and cultural amplification of risk by the
some risks while ignoring, discounting, or attenu- news media, cultural groups, interpersonal networks,
ating others. Hierarchical cultures, perceive industrial vested interests, and other groups and institutions
and technological risks as opportunities and thus less occurs in the transfer of information about the
risky, whereas more egalitarian cultures tend to per- phenomenon.40,41 Evidence from the health literature,
ceive them as threats to their social structure and thus the social psychological literature, and the risk
more risky.31 O’Connor et al.32,33 provide evidence communication literature suggests that these social
that such differences in worldview result in differences and cultural processes serve to modify perceptions of
in the perceptions of climate change risks. risk in ways that may generally be socially adaptive,
Group membership of different sorts may also but may also bias reactions in socially undesirable
influence perceptions of climate change. Knowing that ways in specific instances.42
someone is a rancher or an environmentalist is a In summary, both learning from personal
good predictor of their opinions and beliefs about experience and vicarious learning from statistical
environmental risks and interventions.34 Leiserowitz description contribute to people’s perceptions of
et al.35 describe several distinct interpretive commu- climate change. A better understanding of the mech-
nities in the United States (‘six Americas’) who share anisms and challenges in both types of learning helps
values, socio-demographics, and climate change per- explain both the generally low level of concern about
ceptions. Climate change ‘naysayers’ perceive climate the phenomenon, as well as cultural and other group
change as a very low or non-existent danger, and variations in perceptions and concerns. Differences
climate change ‘alarmists’ hold high-risk perceptions in climate change perceptions have been documented
and extreme images of catastrophic climate change. between climate scientists (and other scientists) and
These groups have significantly different values and the general public. A Pew Research Center43 poll,
beliefs about social and political issues and about e.g., found that while 84% of scientists said the
the necessity of individual behavior change and gov- earth is getting warmer because of human activity
ernmental intervention. Political party affiliation has such as burning fossil fuels, just 49% of the general
also shown an increasing relationship to belief in the public agreed. These differences can be attributed
existence and relative threat of climate change in the to both learning mechanism. Climate scientists often
United States.21,36 have personal experience with climate change in
A second important issue related to learning connection with their research activities that take
about climate change from external sources is them to regions of the world where climate change is
that of trust. People pay attention to information apparent. In addition, by virtue of their education and
about climate phenomena and incorporate it into training, climate scientists and other scientists, more

Vo lu me 1, May /Ju n e 2010  2010 Jo h n Wiley & So n s, L td. 335

 Advanced Review wires.wiley.com/climatechange

generally, can also be expected to rely more on their of 47% in 2006, which was maintained through
analytical processing system than members of the April of 2008, to only 35% in October of 2009,
general population, and put greater trust and weight consistent with the ‘finite pool of worry’ hypothesis
on extrapolations from statistical evidence and model described below. Greater concern about other risks,
outputs, that will thus make them more likely to like the late 2008 global economic melt down and
consider global climate change to be a more serious subsequent unemployment, replaced concern about
risk than typical nonscientists. For others, including environmental issues.
the general public and politicians, variation in climate Regardless of stated levels of concern, few
change perception seems to be associated with polit- Americans see climate change as an immediate risk and
ical beliefs and other deeply held values in ways that tend to rank it as less important than other social issues
go beyond simple personal and strategic concerns. like the economy and terrorism.46,47 While giving lip
These cultural values and allegiances influence infor- service to its importance, the Obama administration
mation processing and causal attribution processes by also appears to have assigned a low priority rank
guiding attention to messages and shaping trust in the to climate change legislation, putting it close to the
messengers.a bottom of its list of legislative concerns, in a similar
Even for those citizens across different regions location to where the US public has placed it.
of the world who believe that climate change is This section explores reasons for the generally
occurring, an understanding of its causes and likely low level of concern about climate change among
effects is limited.44 As described in a recent report by citizens, as well as factors that account for variations
an American Psychological Association task force,42 in the worry and concern that motivates action.
psychological research has started to examine reasons Following a similar organization as in the previous
for these limitations. Analogies from failures to section, I will review different classes of factors
fully understand and take action in other domains that shape concern about and action on climate
provide hypotheses about cognitive and motivational change. These influences are organized around the
challenges that need to be overcome. Tests of these different ways or modes (affect-based, rule-based, and
hypotheses applied to climate change are starting to analysis-based) in which environmentally relevant
emerge, though much work remains to be done to decisions can be made.
illuminate the special questions and challenges related
to cognitive and emotional shortcomings that the
proper understanding of the anthropogenic causes Affect-Based Processing of Environmental
of climate change and their action implications bring. Decisions
Many of these are summarized in the next section, If the understanding of climate scientists is driven
as the perceived causes of climate change and the to a large extent by analytic processes (e.g., by
extent of worry expended on the risk are often highly the probability estimates of adverse consequences
related. provided by climate models), the discussion in the
previous section about the speed and automaticity
of affective and associative processing of information
VARIATION IN CONCERN AND
suggests strongly that the general public will rely far
ACTION ON CLIMATE CHANGE more on affective processes in their climate change
Citizens’ perceptions of the importance and severity decisions. A series of seminal studies in the 1970s
of climate change do not seem to match those of most identified several psychological risk dimensions that
climate scientists. In 2006, the Pew Global Attitudes influence the public’s judgments of the riskiness of
Project, an international representative opinion poll, physical, environmental, and material risks in ways
showed that 75% of Americans assessing global that go beyond, and often disagree with, their objective
warming as a ‘very’ or ‘somewhat’ serious problem.20 consequences (e.g., Ref. 48). People’s judgments of
This level of concern is similar to the level in Russia the similarities between pairs of different health and
(73%) and lower than that in many other nations safety risks could be placed into a two-dimensional
(87% in Canada, 81% in Mexico, 95% in France, space that has been replicated across numerous studies
88% in China, 97% in Japan, 96% in Brazil, and in multiple countries.49 The first dimension of this
94% in India). Thus concern might seem high, space, ‘dread risk,’ captures emotional reactions to
but it does not appear to be either very stable or hazards like nuclear weapons fallout, nuclear reactor
very committed.45 A recent21 report found that the accidents, or nerve gas accidents, i.e., things that make
proportion of Americans who view climate change people anxious because of a perceived lack of control
as a ‘very serious’ problem dropped from its high over exposure to the risks and consequences that

336  2010 John Wiley & Sons, L td. Volume 1, May/June 2010
 WIREs Climate Change What shapes perceptions of climate change?

are catastrophic. The second dimension, ‘unknown kind graphically depicted in the film ‘The Day after
risk,’ refers to the degree to which a risk (e.g., DNA Tomorrow’) have the ability to raise visceral reactions
technology) is seen as new, with unforeseeable to the risk.58 Climate change that is construed as rapid
consequences not yet tested by time, and with hazards is more likely to be dreaded than climate change that
and exposures not easily detectable. is expected to be gradual.
What then is known about the general public’s
emotional reactions to the prospect of climate change?
Public opinion data in the US document a low level Analysis-Based Processing of Environmental
of concern.50 Leiserowitz51 applied a methodology Decisions
developed by Slovic et al.52 to quantify people’s People consider climate change risks to be consider-
emotional reactions to the risk of climate change, ably uncertain and as being mostly in the future and
asking people to provide the first thought or image geographically distant, all factors lead that them to
that came to mind when they heard the term ‘global discount their possible consequences. Economic anal-
warming,’ which was subsequently rated on a scale ysis allows for the discounting of future and distant
ranging from −5 (very negative) to +5 (very positive). costs and benefits by some amount (e.g., by the rate
Associations to distant events like melting glaciers and of interest offered by financial institutions) as a func-
polar ice were most common, followed by generic tion of the time delay, a mechanism that is described
associations to heat and rising temperatures, and mathematically by an exponential discount function.
impacts on non-human nature. Mean affect scores However, contrary to the assumptions of rational-
indicated that these images of ‘global warming’ had economic discounting, people are inconsistent in their
only moderately negative connotations for almost all discounting, applying different discount rates to out-
respondents. comes in different domains (e.g., financial, health,
Differential personal experience affects risk or environmental outcomes59,60 ), showing a strong
perceptions, as discussed above, often by way present bias (i.e., strongly preferring immediate ben-
of people’s affective reactions.53 Whereas personal efits and disliking present costs relative to delayed
exposure to adverse consequences increases fear options), and discounting future benefits far more
and perceptions of risk, familiarity with a risk than future costs.61,62 Whereas the costs of actions
acquired by exposure without adverse consequences that could lead to a mitigation of CO2 emissions
can also lower perceptions of its riskiness.54 One are incurred immediately, their uncertain and future
non-climate example of reduced risk perceptions due benefits are heavily discounted, making the delibera-
to greater familiarity is the home bias in equity tive consideration of such actions unlikely to arrive
portfolios in financial investing, where investors at socially responsible and long-term sustainable
overinvest in equities from their home country, with behavior.
which they are familiar and about whose risks they To explain people’s large impatience when
therefore worry less55 and therefore fail to diversify immediate rewards are an option,63 Trope and
sufficiently internationally. Thus greater familiarity Liberman suggest that future events are construed
with climate change and its risks, unless accompanied differently from events in the present. In particular,
by alarmingly large negative consequences may events in the distant future (an invitation to give
actually lead to smaller rather than larger perceptions a paper at a conference 2 years from now, or
of its riskiness. the prospect of coastal flooding 30 or 50 years
If people conceive of climate change as a from now) are construed in abstract terms, whereas
simple and gradual change from current to future events closer in time (the upcoming trip on Monday
values on variables such as average temperatures and to attend the long-scheduled conference, or the
precipitation, the risks posed by climate change would prospect of a major hurricane passing through
appear to be well-known and, at least in principle, town tomorrow) are construed in more concrete
controllable (e.g., by moving to a different part terms. Abstract versus concrete representation of
of the country). Of course, in most cases, people the consequences of possible actions differ in their
do not move away from hazards even when they affective strength and impact. Abstract representations
are aware of them. Status quo biases or change of consequences in the distant future usually lack
inertia have been documented in many situations, the concrete associations connected to present or
e.g., for organ donations56 and insurance decisions.57 near-present events and thus may not be feared as
The perceived ability to take corrective action is much.
an important determinant of emotional reactions. Protective or mitigating actions against global
Potential catastrophes from climate change (of the climate change can require the sacrifice of concrete,

Vo lu me 1, May /Ju n e 2010  2010 Jo h n Wiley & So n s, L td. 337

 Advanced Review wires.wiley.com/climatechange

immediate benefits for the sake of abstract, distant appeals of American evangelical churches, remind-
goals. The strong negative affect associated with the ing their members of their duties as stewards of
concrete, immediate costs and sacrifices may well drive the earth fall into that category. Rule-based decision
ecologically damaging consumption decisions and making also highlights the importance that people’s
actions. However, this tendency is moderated by the attributions of responsibility for climate change can
way people think about changing their consumption. have. Whereas only 35% of all American respondents
When asked to delay consumption, people first believed that climate change was a serious problem
generate arguments for the status quo, immediate deserving of action in the21 poll previously mentioned,
consumption, and only then latter generate arguments this figure went up to 97% for the subset of respon-
for delaying consumption. Yet, argument generation dents who believed that the observed warming was
for the first action considered (e.g., immediate due to human activity. Attributions of responsibility
consumption) tends to interfere with the subsequent seem to have a high likelihood to trigger corrective
generation of arguments for other action.64 Weber action rules.
et al.65 succeeded in drastically reducing people’s
discounting of future rewards by prompting them to
first generate arguments for deferring consumption, IF THERE IS INSUFFICIENT WORRY
followed by a prompt to generate arguments for OR CONCERN, SHOULD WE
immediate consumption. Social norms and/or positive GENERATE MORE?
or negative affective reactions to a choice options One might conclude that the research reviewed in
typically determine which option is considered first. the previous section suggests that we should find
Therefore, Hardisty et al.66 found that 65% of ways to evoke stronger affective reactions toward
Republicans were willing to pay a CO2 emission the risk of climate change in citizens, managers, or
reduction fee on such purchases as airline tickets public officials, by making expected climate change
when the fee was labeled as a carbon offset (and first effects more vivid or concrete. Such a course of action
generated arguments for purchasing it), but that this may, however, have some unintended consequences.
percentage dropped to 27% when the fee was labeled As worry increases about one type of risk, concern
as a carbon tax, a label that generated negative visceral about other risks has been shown to go down, as
reactions in this group and led them to first generate if people had only so much capacity for worry or
arguments for not purchasing it. a finite pool of worry.70 Increased concern about
global warming may result in decreased concern about
other risks. Hansen et al.71 tested this hypothesis
Rule-Based Processing of Environmental in a climate context, examining whether increases
Decisions in concern about climate variability on the part of
Many decisions, including environmental decisions Argentine farmers resulted in a decrease in concern
are made by recognizing the situation as one for about other risks. In two scenarios of a farm
which a rule exists. These rules can be laws (‘stop at a decision experiment, farmers rated the extent that they
red traffic light’), self-imposed admonishments (‘never worried about (a) the political situation in Argentina,
eat after 5 pm’), or social norms (‘if you broke it, you (b) weather and climate, (c) prices of input variables,
fix it’).67 Many such rules of conduct derive from the and (d) prices of crops at harvest, on a scale from 0
social roles of decision makers,68 including positions (‘not at all worried’) to 10 (‘extremely worried’). The
of responsibility within society (e.g., professor, two scenarios involved the same crop-selection and
parent, or Christian). Implicit rules and role-related cultivation decisions, but differed in the provision of
obligations are often acquired through observational a seasonal climate forecast that indicated unfavorable
learning and imitation. Copying the observed behavior La Niña conditions for the upcoming growing season.
of others is a widespread phenomenon of which the Not surprisingly, stated concern about climate risks
imitator is typically unaware and plays a large role in among the farmers significantly increased from the
human development.69 first to the second scenario, essentially providing
Since rule-based decision habits often develop a manipulation check for the provided climate
to prevent undesirable decisions made as the result information. At the same time, however, concern
of applying analysis- or affect-based processes (e.g., about political risk decreased from the first to the
overriding impulsive eating), promoting a more second scenario, even though the objective political
widespread use of rule-based processes in environmen- risk was identical.
tal decisions may be an effective way of encouraging There was some indication that concern and
more long-term oriented, sustainable behavior. Recent worry was a finite resource even within each scenario.

338  2010 John Wiley & Sons, L td. Volume 1, May/June 2010
 WIREs Climate Change What shapes perceptions of climate change?

For both climate scenarios, farmers who worried indicated that they had the capacity to store grain on
more about political risk tended to worry less their farms were significantly less likely to indicate
about climate risk. Judgments of worry or perceived that they used irrigation or that they had signed up
risk were not inconsequential, in that differences for crop insurance.
in farmers’ perceptions of the degree of risk posed
by political, climate, input costs, and crop price
variables were associated with differences in their CONCLUSIONS
production and pricing decisions. Other illustrations
of the finite pool of worry effect are provided by Behavioral research over the past 30 years strongly
the observation that increases in the concern of suggests that attention-catching and emotionally
the US public about terrorism post 9/11 resulted engaging informational interventions may be required
in decreased concern about other issues such as to engender the public concern necessary for individ-
restrictions of civil liberties as well as climate ual or collective action in response to climate change.
change,72 or that the recent financial crisis reduced Such interventions would need to be conducted
concern about climate change and environmental with full awareness about unintended side-effects
degradation.21 (e.g., reductions in concern about other important
As another caution against inducing stronger risks) and in ways designed to help people overcome
concern about climate change,14 Weber coined cognitive and affective capacity limitations (e.g., the
the phrase single-action bias for the following single-action bias). To the extent that time-delayed
phenomenon observed in contexts ranging from consequences of our actions do not attract the atten-
medical diagnosis to farmers’ reactions to climate tion or generate the concern ex-ante that they would
change. Decision makers are very likely to take seem to warrant ex-post, behavioral research provides
one action to reduce a risk that they worry about, some corrective actions. The concretization of future
but are much less likely to take additional steps events and moving them closer in time and space seem
that would provide incremental protection or risk to hold promise as interventions that will raise visceral
reduction. The single action taken is not necessarily concern. Guided protocols by which decision makers
the most effective one, nor is it the same for different consider arguments for energy conservation and
decision makers. However, regardless of which single climate change mitigation before they are allowed to
action is taken first, decision makers have a tendency consider arguments against such actions may help to
not to take any further action, presumably because improve the balance between the desire for immediate
the first action suffices in reducing the feeling of gratification and the goal of sustainable development.
worry or vulnerability. Weber14 found that farmers Finally, for at least a subset of the public, better
who showed concern about global warming in the (environmental) science and statistics education can
early 1990s were likely to change either something in create the familiarity with the scientific presentation
their production practice (e.g., irrigate), their pricing of information and the mental habits that will create
practice (e.g., ensure crop prices through the futures citizens who give greater weight to the output of
market), or lobbied for government interventions (e.g., their analytic processing system, moving the risk
ethanol taxes), but hardly ever engaged in more than perception of the general public and its officials closer
one of those actions, even though a portfolio of to that of climate scientists. A better understanding of
protective actions might have been advisable. The the human contributions to climate change promises
fear of climate change seemed to set a ‘flag’ that to activate norms about social responsibility to rectify
some action was required, but remained in place problems caused by human action. Even though
only until one such action was taken, i.e., any single changes at a social or cultural level are harder to
protective action had the effect of taking down the effect, they may have the most lasting consequences.
‘impending danger flag.’ While such behavior might Changes in the prevalence and salience of different
have served us well in our evolutionary history goals, roles, and obligations related to the human
where single actions generally sufficed to contain interface with its natural environment may result in
important risks, in more complex environments where greater attention to climate change, more accurate
a portfolio of risk management actions is advised, perceptions of its scope and possible consequences
purely affect-driven, single-action biased responses on the part of policy makers and the general
may not be sufficient. Hansen et al.71 found evidence public, and a greater willingness to act on these
for the single-action bias in farm practices that can perceptions.
be interpreted as protective actions against climate Failing these efforts, the problem discussed
change and/or climate variability. Thus farmers who in this paper is ultimately self-correcting. Increasing

Vo lu me 1, May /Ju n e 2010  2010 Jo h n Wiley & So n s, L td. 339

 Advanced Review wires.wiley.com/climatechange

personal evidence of climate change and its potentially NOTES

devastating consequences can be counted on to be an a There is need, however, for empirical evidence about
extremely effective teacher and motivator, if and when differences in concern or risk perceptions related to
they occur and are recognized as the result of human climate change between climate scientists and mem-
activity. Unfortunately, such lessons may arrive too bers of the general public, which have not yet been
late for corrective action. compared directly.

REFERENCES
1. IPCC. In: Pachauri RK, Reisinger A, eds. Climate Environmental and Ethical Issues in Management. San
Change 2007, Synthesis Report. Geneva, IPCC: 2007. Francisco, CA: Jossey-Bass; 1997, 314–341.
2. Oreskes N. The scientific consensus on climate change. 15. Kupperman KO. The puzzle of the American cli-
Science 2004, 306:1686. mate in the early colonial period. Am Hist Rev 1982,
3. Hulme M. Why We Disagree about Climate Change: 87:1262–1289.
Understanding Controversy, Inaction and Opportu- 16. Weber EU, Shafir S, Blais A-R. Predicting risk-
nity. Cambridge: Cambridge University Press; 2009. sensitivity in humans and lower animals: Risk as
4. Hoggan J. Climate Cover-up. Vancouver: Greystone variance or coefficient of variation. Psychol Rev 2004,
Books; 2009. 111:430–445.

5. National Research Council. Global Environmen- 17. Hertwig R, Barron G, Weber EU, Erev I. Decisions
tal Change: Understanding the Human Dimensions. from experience and the effect of rare events. Psychol
Washington, DC: National Academy Press; 1992. Sci 2004, 15:534–539.

6. National Research Council. In: Stern P, Dietz T, Rut- 18. Erev I, Barron G. On adaptation, maximization, and
tan VW, Socolow RH, Sweeney J, eds. Environmentally reinforcement learning among cognitive strategies. Psy-
Significant Consumption: Research Directions. Wash- chol Rev 2005, 112:912–931.
ington, DC: National Academy Press; 1997. 19. Yechiam E, Barron G, Erev I. The role of personal
7. Douglas M, Wildavsky A. Risk and Culture: An Essay experience in contributing to different patterns of
on the Selection of Technological and Environmental response to rare terrorist attacks. J Conflict Resolut
Dangers. Berkeley: University of California Press; 1982. 2005, 49:430–439.

8. Strauss S, Orlove BS. Weather, Climate, Culture. 20. Pew Research Center. No Global Warming Alarm in the
Oxford: Berg Publisher; 2003. U.S., China. Washington, DC: Pew Research Center;
2006.
9. Dunlap RE, Van Liere KD. The new environmental
paradigm: a proposed measuring instrument and pre- 21. Pew Research Center. Modest Support for ‘‘Cap and
liminary results. J Environ Educ 1978, 9:10–19. Trade’’ Policy. Fewer Americans See Solid Evidence
of Global Warming. Washington, DC: Pew Research
10. Dunlap RE, Van Liere KD. Commitment to the Center; 2009a.
dominant social paradigm and environmental con-
cern: an empirical examination. Soc Sci Q 1984, 22. Sloman SA. The empirical case for two systems of
65:1013–1028. reasoning. Psychol Bull 1996, 119:3–22.

11. Weber EU, Morris MW. Cultural differences in judg- 23. Chaiken S, Trope Y. Dual-process Theories in Social
ment and decision making: Insights from constructivist, Psychology. New York: Guilford; 1999.
structuralist approaches. Perspect Psychol Sci 2010. In 24. Epstein S. Integration of the cognitive and the psycho-
press. dynamic unconscious. Am Psychol 1994, 49:709–724.
12. Weber EU, Johnson EJ. Mindful judgment and decision 25. Loewenstein GF, Weber EU, Hsee CK, Welch E. Risk
making. Annu Rev Psychol 2009, 60:53–86. as feelings. Psychol Bull 2001, 127:267–286.
13. National Research Council. In: Stern PC, Easterling 26. Marx SM, Weber EU, Orlove BS, Leiserowitz A, Krantz
WE, eds. Making Climate Forecasts Matter. Washing- DH, et al. Communication and mental processes: expe-
ton, DC: National Academy Press; 1999. riential and analytic processing of uncertain climate
14. Weber EU. Perception and expectation of climate information. Glob Environ Change 2007, 17:47–58.
change: precondition for economic and technological 27. Arctic Climate Impact Assessment. Impacts of a Warm-
adaptation. In: Bazerman M, Messick D, Tenbrunsel ing Arctic. Cambridge, UK: Cambridge University
A, Wade-Benzoni K, eds. Psychological Perspectives to Press; 2004.

340  2010 John Wiley & Sons, L td. Volume 1, May/June 2010
 WIREs Climate Change What shapes perceptions of climate change?

28. Leiserowitz A, Broad K. Florida: Public Opinion on Cli- Challenges. Washington, DC: Author; 2009, Retrieved
mate Change. A Yale University / University of Miami from http://www.apa.org/releases/climate-change.pdf.
/ Columbia University Poll. New Haven, CT: Yale 43. Pew Research Center. Public Praises Science; Scientists
Project on Climate Change; 2008, http://environment. Fault Public, Media. Washington, DC: Pew Research
yale.edu/uploads/FloridaGlobalWarmingOpinion.pdf. Center; 2009b.
29. Whitmarsh L. Are flood victims more concerned about
44. Moser S, Dilling L. Creating a Climate for Change:
climate change than other people? The role of direct
Communicating Climate Change—Facilitating Social
experience in risk perception and behavioural response.
Change. Cambridge, UK: Cambridge University Press;
J Risk Res 2008, 11:351–374.
2007.
30. Schwartz SH. Are there Universal aspects in the struc-
45. Weber EU, Kates RK, Krantz D, Leiserowitz A, Parris
ture and contents of human values? J Soc Issues 1994,
TM, et al. Political and societal goodwill. In: Sachs JS,
50:19–45.
Schlosser P, eds. Is Sustainable Development Feasible?
31. Dake K. Orienting dispositions in the perception of risk: New York City: Columbia University Press; 2010.
an analysis of contemporary worldviews and cultural
biases. J Cross Cult Psychol 1991, 22:61–82. 46. Krosnick JA, Holbrook AL, Lowe L, Visser PS. The
origins and consequences of democratic citizens’ pol-
32. O’Connor RE, Bord RJ, Fisher A. The curious impact icy agendas: a study of popular concern about global
of knowledge about climate change on risk perceptions warming. Clim Change 2006, 77:7–43.
and willingness to sacrifice. Risk Decision and Policy
1998, 3:145–155. 47. Leiserowitz A, Kates R, Parris T. Do global Attitudes
and behaviors support sustainable development? Envi-
33. O’Connor RE, Bord RJ, Fisher A. Risk perceptions, gen-
ronment 2005, 47:22–38.
eral environmental beliefs, and willingness to address
climate change. Risk Anal 1999, 19:461–471. 48. Slovic P, Fischhoff B, Lichtenstein S. The psychometric
study of risk perception. In: Covello VT, Menkes J,
34. Opotow S, Brook A. Identity and exclusion in rangeland
Mumpower J, eds. Risk Evaluation and Management.
conflict. In: Clayton S, Opotow S, eds. Identity and the
New York, NY: Plenum Press; 1986.
Natural Environment: The Psychological Significance
of Nature. Cambridge, MA: MIT Press; 2003. 49. Slovic P. Perception of risk. Science 1987, 236:
35. Leiserowitz A, Maibach E, Roser-Renouf C. Global 280–285.
Warming’s ‘‘Six Americas’’: An Audience Segmen- 50. Dunlap RE, Saad L. Only One in Four Americans are
tation. New Haven, CT: Yale Project on Climate Anxious about the Environment. Princeton, NJ: Gallup
Change; 2008, http://environment.yale.edu/uploads/ News Service; 2001.
SixAmericas.pdf. 51. Leiserowitz A. Climate change risk perception and pol-
36. Dunlap RE, McCright AM. Social movement identity: icy preferences: the role of affect, imagery, and values.
validating a measure of identification with the environ- Clim Change 2006, 77:45–77.
mental movement. Soc Sci Q 2008, 89:1045–1065.
52. Slovic P, Flynn J, Layman M. Perceived risk, trust,
37. Orlove B, Broad K, Petty A. Factors that influence the and the politics of nuclear waste. Science 1991,
use of climate forecasts: evidence from the 1997/98 254:1603–1607.
El Nino event in Peru. Bull Am Meteorol Soc 2004,
53. Keller C, Siegrist M, Gutscher H. The role of the affect
85:1735–1743.
and availability heuristic in risk communication. Risk
38. Roncoli C. Ethnographic and participatory approaches Anal 2006, 26:631–639.
to research on farmers’ responses to climate predictions.
Clim Res 2006, 33:81–99. 54. Fischhoff B, Slovic P, Lichtenstein S, Read S, Combs B.
How safe is safe enough? A psychometric study of atti-
39. Kasperson RE, Renn O, Slovic P, Brown HS, Emel J, tudes towards technological risks and benefits. Policy
et al. The social amplification of risk: a conceptual Sci 1978, 8:127–152.
framework. Risk Anal 1988, 8:177–187.
55. Weber EU, Siebenmorgen N, Weber M. Communicat-
40. Weinstein ND, Lyon JE, Rothman AJ, Cuite CL. Pre-
ing asset risk: how name recognition and the format of
occupation and affect as predictors of protection action
historic volatility information affect risk perception and
following natural disaster. Br J Health Psychol 2000,
investment decisions. Risk Anal 2005, 25:597–609.
5:351–363.
56. Johnson EJ, Goldstein D. Do defaults save lives? Science
41. Taylor SE. Adjustment to threatening events: a the-
2003, 302:1338.
ory of cognitive adaptation. Am Psychol 1983,
38:1161–1173. 57. Johnson E, Hershey J, Meszaros J, Kunreuther H. Fram-
42. American Psychological Association Task Force on ing, probability distortions and insurance decisions.
the Interface Between Psychology and Global Cli- J Risk Uncertain 1993, 7:35.
mate Change. Psychology and Global Climate Change: 58. Leiserowitz A. Surveying the impact of ‘‘The Day After
Addressing a Multi-faceted Phenomenon and Set of Tomorrow’’. Environment 2004, 46:23–44.

Vo lu me 1, May /Ju n e 2010  2010 Jo h n Wiley & So n s, L td. 341

 Advanced Review wires.wiley.com/climatechange

59. Hardisty DH, Weber EU. Temporal discounting Strategic Reflections on Human Behavior. Cambridge,
of environmental outcomes: effects of valence out- MA: MIT Press; 1991.
weigh domain differences. J Exp Psychol Gen 2009, 68. March JC, Heath C. Primer on Decision Making: How
3:329–340. Decisions Happen. New York: The Free Press; 1994.
60. Hendrickx L, Nicolaij S. Temporal discounting and 69. Meltzoff MA, Moore MK. Persons and representa-
environmental risks: the role of ethical and loss-related tion: why infant imitation is important for theories of
concerns. J Environ Psychol 2004, 24:409–422. human development. In: Nadel J, Butterworth G, eds.
61. Ainslie G. Specious reward: a behavioral theory of Imitation in Infancy. Cambridge: Cambridge University
impulsiveness and impulse control. Psychol Bull 1975, Press; 1999.
82:463–496.
70. Weber EU. Experience-based and description-based
62. Loewenstein G, Elster J, eds. Choice Over Time. New perceptions of long-term risk: why global warming does
York: Russell Sage; 1992. not scare us (yet). Clim Change 2006, 70:103–120.
63. Trope Y, Liberman N. Temporal construal. Psychol 71. Hansen J, Marx S, Weber EU. The role of climate
Rev 2003, 110:403–421. perceptions, expectations, and forecasts in farmer deci-
64. Johnson EJ, Haeuble G, Keinan A. Aspects of endow- sion making: the Argentine Pampas and South Florida.
ment: a query theory of loss aversion. J Exp Psychol Technical Report 04-01. Palisades, NY: International
Learn Mem Cogn 2007, 33:461–474. Research Institute for Climate Prediction (IRI); 2004.
65. Weber EU, Johnson EJ, Milch K, Chang H, Brod- 72. Leone RC, Anrig G. The War on Our Freedoms: Civil
scholl J, et al. Asymmetric discounting in intertemporal Liberties in an Age of Terrorism. New York: Century
choice: a query theory account. Psychol Sci 2007, Foundation; 2003.
18:516–523. 73. Slovic P. Trust, emotion, sex, politics, and science: sur-
66. Hardisty DH, Johnson EJ, Weber EU. A dirty word or a veying the risk-assessment battlefield. In: Bazerman M,
dirty world? Attribute framing, political affiliation, and Messick D, Tenbrunsel A, Wade-Benzoni K, eds. Psy-
Query Theory. Psychol Sci 2010, 21:86–92. In press. chological Perspectives to Environmental and Ethical
67. Prelec D, Herrnstein RJ. Preferences and principles, Issues in Management. San Francisco, CA: Jossey-Bass;
alternative guidelines for choice. In: Zeckhauser R, ed. 1997, 277–313.

342  2010 John Wiley & Sons, L td. Volume 1, May/June 2010