CCAE

A

I

Tiawhietrr

FMSr

MME

4

ommunication and Marketing As Climatehange–Intervention AssetsPublic Health Perspective

dward W. Maibach, PhD, Connie Roser-Renouf, PhD, Anthony Leiserowitz, PhD

bstract: The understanding that global climate change represents a profound threat to the healthand well-being of human and nonhuman species worldwide is growing. This articleexamines the potential of communication and marketing interventions to influencepopulation behavior in ways consistent with climate change prevention and adaptationobjectives. Specifically, using a framework based on an ecologic model of public health, thepaper examines: (1) the potential of communication and marketing interventions toinfluence population behaviors of concern, including support for appropriate publicpolicies; (2) potential target audiences for such programs; and (3) the attributes ofeffective climate change messages. Communication and marketing interventions appear tohave considerable potential to promote important population behavior change objectives,but there is an urgent need for additional translational research to effectively harvest thispotential to combat climate change.(Am J Prev Med 2008;35(5):488–500) © 2008 American Journal of Preventive Medicine

tbatltwof

atcpsiacca

httPdrta

ntroduction

Either we will achieve an awareness of our place inthe living and life-giving organism of our planet,or we will face the threat that our evolutionaryjourney may be set back thousands or even mil-lions of years. That is why we must see this issue asa challenge to behave responsibly and not as aharbinger of the end of the world.

– Vaclav Havelformer president of the Czech Republic1

he potential health effects of climate changehave been reviewed extensively.2–7 Earth systemchanges, including rising temperatures, increas-

ng climate variability, increased rainfall in some areasnd drought in others, and more frequent severeeather events, have considerable potential to affectuman health. Severe weather events may result in

njuries and fatalities, and heatwaves can cause directffects such as dehydration, heat asthenia, heat exhaus-ion, and heat stroke; excess deaths during heatwavesesult primarily from underlying cardiovascular andespiratory diseases. Ecosystem changes can increase

rom the Center for Climate Change Communication, Georgeason University (Maibach, Roser-Renouf), Fairfax, Virginia; and the

chool of Forestry and Environmental Studies, Yale University (Leise-owitz), New Haven, Connecticut

Address correspondence and reprint requests to: Edward W.aibach, PhD, Center for Climate Change Communication, George

cason University, 4400 University Drive, MS 6A8, Fairfax VA 22030.

-mail: emaibach@gmu.edu.

88 Am J Prev Med 2008;35(5)© 2008 American Journal of Preventive Medicine • Publish

he range, seasonality, and infectivity of some vector-orne diseases. Heavy rainfalls and related factors aressociated with waterborne disease outbreaks, andhese may increase the risk of foodborne illness. Higherevels of carbon dioxide and heat may promote produc-ion of allergens (e.g., pollen) by such plants as rag-eed, and warmer weather may promote the formationf ground-level ozone. Humidity combined with heatacilitates fungal growth and transmission.

Potential indirect effects—for which data are lessvailable and uncertainties are greater—include men-al health consequences, population dislocation, andivil conflict. In addition, changes in the patterns ofests, parasites, and pathogens affecting wildlife, live-tock, agriculture, forests, and coastal marine organ-sms can alter ecosystem composition and functions,nd changes in these life-support systems carry impli-ations for human health.8 The burden of theseonditions is expected to increase as climate changedvances.Successfully addressing climate change as a public

ealth threat will require both mitigation and adapta-ion strategies, or, in more common public healtherms, primary and secondary prevention strategies.rimary prevention requires aggressive efforts to re-uce atmospheric levels of greenhouse gases, both byeducing emissions and by sequestering gases already inhe atmosphere (e.g., through reforestation). Second-ry prevention requires efforts to adapt to a changing

limate in ways that protect population health and

0749-3797/08/$–see front mattered by Elsevier Inc. doi:10.1016/j.amepre.2008.08.016

wlp

iphstamCcvspsc

twembiitwmMvi

It

TstAWana

tbic

apaaka

tbta

dbpepe

N

ell-being. Examples of adaptation include reinforcingevees in coastal areas, implementing heatwave pre-aredness plans in urban areas, and reforestation.There is an urgent need to influence people’s behav-

or—on a large scale or population basis—to helprevent and reduce the burden of climate change onuman and other populations. Social and behavioralcience intervention methods offer important tools tohis end. Here, two broad sets of tools—communicationnd social marketing—are examined as assets forounting a public health response to climate change.ommunication is defined as the production and ex-hange of information to inform, influence, or moti-ate individual, institutional, and public audiences, andocial marketing as the development and distribution ofroducts or services to influence behavior on a largecale for the purpose of societal benefit rather thanommercial profit.

First, a summary is presented of what is known abouthe drivers of population behavior, along with a frame-ork that organizes these influences. Current knowl-dge is then reviewed about (1) the potential of com-unication and social marketing to influence population

ehavior, (2) general public audiences for climate changenterventions, and (3) effective message strategies tonfluence these audiences. It is important to note thathe majority of the research cited here was conductedith people in the U.S., and that recommendations areade primarily for the current American context.any of the general points, however, likely have rele-

ance to other developed nations and even to develop-ng nations.

nfluencing Population Behavior: Understandinghe Challenge

he public health community, like many others, waslow to recognize the threat of climate change. Thehemes of World Health Day 2008 (Protecting Healthgainst Climate Change) and National Public Healtheek 2008 (Climate Change: Our Health in the Bal-

nce) make clear that the threat is now recognized. Theeed for the public health community to mobilize itsssets against that threat is now obvious.

Fortunately, the public health community has mucho offer. Perhaps most notably, this community hasoth breadth and depth of experience in understand-

ng and responding to population behavior-changehallenges (see Text Box for an example).

Historically, when people fail to behave in ways thatre in their own or society’s best interest—as judged byublic health professionals, environmental scientists,nd other similar experts—the tendency has been tossume that the cause must be either a lack of relevantnowledge on their part (i.e., an information deficit)

nd/or misguided attitudes. The prescription that has t

ovember 2008

ended to follow this diagnosis is: to change people’sehavior, we must provide them with the knowledgehey lack and/or persuade them to change theirttitudes.15

In the public health community over the past severalecades, this “information deficit” view of populationehavior—although appealing in its simplicity and ap-arent face validity—has been largely supplanted bycologic views of population behavior. The “people andlaces” framework (Figure 1) is one example of ancologic model.16,17 The framework describes popula-

Text Box

Chronic disease prevention provides an example of thetype of challenges that climate change poses. Over thepast several decades, the public health community hasfocused on understanding and preventing chronic dis-eases through population-based intervention strate-gies.9,10 Many of the underlying behavioral and eco-nomic factors that make chronic diseases sochallenging to control in the modern era appear tohave direct relevance to climate change control. Forexample:

● People have a strong innate tendency to value imme-diate benefits more than future benefits.11 Althoughmany of the costs—including monetary (e.g., retrofit-ting buildings with energy-efficient devices), time andeffort (e.g., using active or mass transit optionsrather than driving in one’s car), and social (e.g.,challenging people’s preferences about whetherdinner should include meat or not)—associatedwith prevention and adaptation are necessarilyborne in the present, many of the associated benefitsdon’t accrue for months (e.g., reduced utility bills,reduced BMI), years, or even decades into the future(e.g., reduced global warming, heart diseaseaverted).

● People have a tendency to consume resources inproportion to how available and affordable theresources are.12 In societies where resources arereadily available and affordable (such as the U.S.),this tendency can lead to excess consumption, thatis, consumption levels that exceed the individu-al’s—or society’s—best long-term interests. Well-documented examples of such excess consumptioninclude calories and electronic media content.13

Production, distribution, and consumption of suchresources are typically dependent on fossil fuels atpresent.

● People have a tendency to conserve physical effortexpenditures.14 This innate human tendency hasbeen greatly enabled in the developed world by theproliferation of labor-saving devices (e.g., cars,home appliances, power tools). Like other re-sources, the production, distribution, and use ofthese devices currently depends on fossil fuels.

ion health—and environmental outcomes—as being

Am J Prev Med 2008;35(5) 489

dtfuasmmntsptcTaftsgaatokhio

tt

amfbim

TIB

Rsblaicfs“ur

F

4

etermined primarily by population behavior, which, inurn, is determined by people-related and place-relatedactors. The people-related factors that influence pop-lation behavior are organized into three levels ofnalysis: individual-level factors (such as beliefs andkills), social network-level factors (such as behavioralodeling and social reinforcement), and group-, com-unity- or population-level factors (such as social

orms and collective efficacy). The place-related fac-ors—as identified by Farley and Cohen18—are de-cribed in broad terms as the availability and cost ofroducts and services, the attributes of physical struc-ures, social structures (i.e., laws and policies), and theultural and media messages in our communities.hese place-related factors manifest at two levels ofnalysis: the local level and the distal level. Local-levelactors describe people’s immediate environment—heir homes, schools, workplaces, and neighborhoodhops—and influence population behavior only in aiven locale. Distal-level factors originate furtherfield—in state, national, and multinational capitals,nd in the headquarters of multinational corpora-ions—and exert influence on population behaviorver wider geographic areas. The framework also ac-nowledges that place-related factors—for example,armful environmental exposures—can exert direct

nfluence on population health and environmental

igure 1. A “people and places” framework

utcomes, but the framework intentionally emphasizes n

90 American Journal of Preventive Medicine, Volume 35, Num

he role of human agency in causing or preventinghose environmental exposures.

The potential of communication and social marketings means to influence population health and environ-ental outcomes becomes clear in the context of this

ramework. Specifically, most of the people- and place-ased drivers of population behavior potentially can be

nfluenced through communication and/or socialarketing.

he Potential of Communication and Marketing tonfluence the Drivers of Climate Change–Relatedehavior

emarkably little progress has been made in under-tanding how best to influence climate change–relatedehaviors on a population basis, especially given the

ikelihood of severe negative consequences that mayrise—including health, environmental, and economicmpacts.19 The research literature on individual climatehange–mitigation behaviors has focused primarily onour broad categories: household energy use, recycling,urface transportation behavior, and purchase ofgreen” products.20 The research literature on individ-al-level adaptation has focused primarily on issueselated to increasing household preparedness against

atural disasters, such as hurricanes.21

ber 5 www.ajpm-online.net

IP

Tairfarp

●

●

●

●

●

●

mavav

●

●

Io

Rptsberrdcc

Io

Cn

N

nfluencing Individual-Level Drivers ofopulation Behavior

he intervention studies conducted to date have primarilyttempted to influence population behavior by targetingndividual-level factors. The large majority of 38 recentlyeviewed household energy conservation interventions,or example, used communication to influence individu-l-level drivers of population behavior.22 In this and othereviews, several modes of communication have shownromising ability to reduce energy use23:

The provision of tailored or customized recommen-dations—based on home energy audits—has beenshown in some (but not all) studies to reduce energyuse in the range of 4% to 21%.24–27

The provision of feedback (i.e., specific informationabout the amount of energy being used)—especiallywhen the feedback is frequent or continuous—hasbeen shown to reduce energy use in the range of 5%to 13%.28–32

Encouraging people to set an energy-reduction goal—especially if they aregiven feedback about their progresstoward the goal— has been shown toappreciably reduce household energyconsumption.33,34

Using mass media to model behaviors ofinterest has long been known to be aneffective population behavior-changestrategy35; regrettably, only a singlestudy that has tested this approach to influenceclimate change–relevant behavior was found. In thatstudy, TV was used to model ways to reduce house-hold electricity use. The programming led to a 10%reduction in household electricity use, although thereduction was not apparent a year later.36

Eco-labeling programs have been shown to influ-ence population behavior—at least some people’sbehavior under certain conditions.37 People whohold pro-environmental attitudes are most likely tobe influenced. Moreover, to be effective, peoplemust understand the label, must believe that the“green-designated” product offers meaningful envi-ronmental benefits, and must trust the organizationthat has given the designation (with governmentdesignations being more trusted than industry des-ignations).23 The effectiveness of eco-labeling pro-grams tends to increase over time as consumersdevelop trust in the labeling system.A number of communication campaigns promotinghousehold disaster preparedness have been evaluated.Their behavioral impact has ranged from no behaviorchange at all to a relatively great deal of public andhousehold change.21 The more successful campaignstypically used what are now commonly accepted as

Lisrelatedat ww

onli

good campaign design practices: simple clear messages b

ovember 2008

(e.g., specifying who is at risk, how severe and howcertain the risk is, and what can be done to reduce therisk or diminish losses), repeated often (e.g., through avariety of interpersonal and media channels, electron-ically and in print), by a variety of trusted sources (e.g.,scientists, community leaders, journalists).

Marketing interventions—in which improvements areade to products or services, their prices and availability,

nd how they are promoted (to enhance their perceivedalue to potential customers)—have also shown promises tools for population behavior management with aariety of climate change–relevant behaviors:

Financial incentives to install energy-efficient appli-ances can substantially increase homeowners’ pur-chase of such appliances, especially if the offers areaggressively promoted (i.e., communicated).38

Green energy programs—in which utility companiesoffer their residential customers energy from renew-able sources at a premium price—have a growing

presence in the marketplace. Programs withmore aggressive marketing features—lowminimum purchase requirements, shortcontract lengths, and aggressive promo-tion—have higher customer participa-tion and total energy purchases.39

● Travel demand management programs—which use a variety of marketing methods—have been shown to substantially in-crease use of public transportation(20%–33%) and active transport modes(including walking [16%] and cycling[6%–91%]), and to reduce the numberof car trips taken (10%) and distancetraveled by car (17%).40

nfluencing Social Network–Level Driversf Population Behavior

elatively few studies have attempted to influenceopulation behavior through social network–level in-erventions. One important exception is a series oftudies in which block leaders were recruited in neigh-orhoods to model household recycling behavior andxhort and assist neighbors to recycle. The approachesulted in significant neighborhood-wide increases inecycling.41–43 Opinion-leader interventions of this typeeserve considerable additional research attention be-ause of their potential to influence a wide range oflimate change–mitigation and adaptation behaviors.44

nfluencing Community-Level Driversf Population Behavior

ommunication interventions that influence people’sormative beliefs—that is, people’s beliefs about the

todcastjpm-et.

tenPo

w.ane.n

ehavior of others—have been shown to promote a

Am J Prev Med 2008;35(5) 491

rdtRihgctbrt

IB

Dier

●

●

●

●

abdpoirbacEgtUGoatpasasoaiti(sci

T

Aivbtuh

mcp

4

ange of relevant behaviors including recycling,45 re-uced household electrical use,46 and reduced hotelowel use (which has direct water-use implications).47

ecent advances in understanding how to use thesenexpensive methods to manage population behaviorave readied them for widespread application.48 And,iven the rapidly growing literature suggesting thatommunity-level variables can be far more powerfulhan individual-level variables in shaping populationehavior,49 there is a pressing need for interventionesearch on other community-level drivers of popula-ion behavior.

nfluencing Place-Based Drivers of Populationehavior

espite the importance of the attributes of place, littlentervention research has been conducted to test strat-gies for influencing these attributes. Some illustrativeeal-world examples, however, include:

The limited availability and higher up-front costs ofcompact fluorescent lighting (CFL) have, until re-cently, depressed demand for the product.50 Prior to2007, retail shelf space and marketing promotion forCFLs was severely limited, and prices were exorbi-tantly high by comparison.51 Each of these factorshas improved recently, in part due to retailing giantWalMart’s successful initiative to sell one millionCFLs in 2007.52

Sidewalks and certain other physical attributes makesome communities more walkable than others. Res-idents in walkable neighborhoods get more physicalactivity and do more of their errands on foot, thando residents in less walkable neighborhoods.53 Asmall but growing number of campaigns are advo-cating neighborhood/city reconfigurations to makethem more supportive of active living.54

Low gasoline taxation in the U.S. fosters higherconsumption. It is estimated that increasing thegasoline tax to $2/gallon would reduce short-termconsumption by �15% and long-term consumptionby �60%. Similarly, federal subsidies on a range ofproducts—including oil production and large sportutility vehicles—lower their price to encourage con-sumption, thereby creating taxpayer-supportedgreenhouse gas pollution. A range of individuals andorganizations are advocating for substantial in-creases in the federal gas tax, or more generally fora carbon tax or a cap-and-trade system to limitcarbon emissions, but their efforts thus far havefailed to gain political traction due to lack of broadpublic support.55

Advertisements and other types of media can rein-force, or even help create, consumerist values andbehaviors. A recent study in China, for example,demonstrated that exposure to consumption-related

and Western-originated media content is contribut- t

92 American Journal of Preventive Medicine, Volume 35, Num

ing to the growth of consumerist values.56 Con-versely, a media campaign in England by the Alli-ance Against Urban 4x4s is attempting to reverse therising trend of SUV sales by creating and sustaining apublic debate in the media about the use of SUVs inthe urban environment, highlighting their effect onsociety and the environment, and countering imagesdepicted in industry advertising (www.stopurban4x4s.org.uk).

When communication and marketing interventionsre used to influence place-based drivers of populationehavior, they often target different audiences and useifferent methods than campaigns seeking to influenceeople-based factors. Segments of the general public—r, stated differently, segments of the voting or purchas-

ng public—can be targets of these initiatives whenallying grassroots support for the proposed place-ased modifications is helpful. The ultimate targetudiences, however, are the people whose decisionsontrol the attributes of place (e.g., elected officials).lected, appointed, and career officials at all levels ofovernment (local, state, national, as well as multina-ional government organizations [e.g., the Europeannion]) are one such category of target audience.overnment officials, through acts of commission andmission, have the capacity to influence the physicalnd social structures of communities directly, and havehe capacity to influence the availability and cost ofroducts and services indirectly through regulationnd, to a lesser extent, the prevalence of media mes-ages. Decision makers in a wide range of businessesnd nongovernmental organizations (NGOs) are aecond such category of target audience. Through theirperating decisions, these people directly influence thevailability and cost of products and services, the phys-cal structures, and the media messages in communi-ies. They also play a number of important roles innfluencing community social structures indirectlythrough their support or opposition). All of thesehould be considered important target audiences forlimate change communication and social marketingnitiatives.

he Potential of Multi-Level Interventions

s has been illustrated above, initiatives that seek tonfluence population behavior with single-level inter-entions—that is, attempting to create change in one,ut only one, of the five levels of influence on popula-ion behavior—can have a measurable impact on pop-lation behavior. In most cases documented thus far,owever, the impact has been modest.That single-level interventions typically have only aodest population impact makes perfect sense in the

ontext of ecologic models of behavior. The causes ofopulation behavior are multifactorial, thus interven-

ions targeting only one of those factors are likely to

ber 5 www.ajpm-online.net

hm

ahmwfiiftpvthtp

cmteveaiebstvr

ischvsftnrnaaileWnadti

it

spGduthoe

luptiinttb

Aa

Rbeibfstllmpci

oipchpfi(wb

I

L

N

ave only modest success. The literature suggests thatulti-level interventions hold greater promise.57

An important interplay also exists between peoplend the attributes of place. Using a traditional publicealth example, programs to prevent micronutrientalnutrition have been shown to be most successfulhen two conditions are ensured: (1) fortified staple

oods are made widely available in the community (i.e.,nfluencing place) and (2) promotional efforts aremplemented to heighten consumer demand for thoseortified foods (i.e., influencing people).58 Similarly,he effectiveness of incentive programs to promote theurchase of energy-efficient household appliances hasaried by a factor of 10, depending on how aggressivelyhey were promoted to members of target house-olds.59 In short, active communication plays an impor-

ant role in stimulating the uptake of useful newroducts and services.There exist only a few examples of multi-level climate

hange interventions. A compelling one was a socialarketing initiative implemented in Hood River OR

hat resulted in a 15% decrease in community-widelectrical consumption as a result of a multi-level inter-ention targeting households.60,61 Built on the basis ofxtensive marketing research, the program influencedttributes of place by offering financial incentives andn-home assistance to help residents install variousnergy-conserving devices, and it influenced people—oth individuals and social networks—through aggres-ive use of media and word-of-mouth initiatives. In sum,his program modified the Hood River community in aariety of ways that made it easier and more normal foresidents to adopt energy-saving measures.

In an excellent review of communication as a policynstrument through which to alter environmentallyignificant behaviors, Stern62 concludes that communi-ation can influence certain important drivers of be-avior (i.e., personal capabilities, habits and routines,alues, attitudes, beliefs and personal norms, and theocial context in which behaviors are or are not per-ormed), but has no capacity to influence the poten-ially more important institutional, economic, and tech-ologic drivers of behaviors (including laws andegulations, financial costs and rewards, available tech-ology, and convenience). Stern concludes that largernd more sustainable changes in population behaviorre likely to require use of the full range of policynstruments, including communication, voluntary col-aborative actions by industry, command and control,conomic instruments, and service and infrastructure.e suggest that, when focused appropriately, commu-

ication and marketing can be used to effect changemong the institutional, economic, and technologicrivers of behavior. Specifically, although communica-ion and marketing are typically thought of as means to

nfluence populations, they also provide means to t

ovember 2008

nfluence the people who control the attributes of placehat drive population behavior.

Governments can—and in many instances have—ponsored communication and social marketing cam-aigns targeting climate change–related behaviors.overnment campaigns typically target people-basedrivers of population behavior, but they can also besed to target place-based factors that are controlled byhe private and NGO sector. California’s success atolding its per-capita energy consumption constantver the past several decades provides an excellentxample.When government policies contribute to the prob-

em, NGO- and citizen-sponsored campaigns can besed to advocate changes in government policy. Theublic health literature uses various concepts and termso describe the use of communication and marketing tonfluence the attributes of place in this manner. Thesenclude policy advocacy, media advocacy, and dissemi-ation of evidence-based practices. Organizations in

he private sector have a different set of concepts anderms to describe these activities including business-to-usiness marketing and lobbying.

udiences for Climate Change Communicationnd Marketing Campaigns

ecent polls indicate that about half of U.S. residentselieve that climate change is already having dangerousffects on people or will within the next decade63—anncrease of 20 percentage points since 2004—and 19%elieve it is a very serious threat to them and to theiramilies. Anecdotally, the frightening projections of risingeas, flooding, mass extinctions, and displaced popula-ions arouse concern and motivate action in some, buteave others with feelings of indifference, despair, disbe-ief, powerlessness, or cynicism. This highlights a funda-

ental truth: There is no such thing as “the generalublic.” To reach and influence audiences effectively,ampaigns must be targeted on the basis of audiences’nterests, values, and current behavioral patterns.64,65

Audience segmentation has traditionally been basedn demographic traits, but demographics alone are

neffective predictors of global warming attitudes andractices.66 Segmentation using a variety of psychoso-ial variables, a method with a long history in the publicealth arena, is likely to offer a more promising ap-roach.15,67 For example, recent research has identi-ed several distinct interpretive communities of riski.e., audience segments) based on differences in globalarming risk perceptions, policy preferences, values,eliefs, and media use.68,69

nterpretive Communities of Risk

eiserowitz and Slovic conducted a nationally represen-

ative survey of adults in the U.S. (n�810) in 2005 and

Am J Prev Med 2008;35(5) 493

ioipilcrEgw

tia

●

●

●

rc

●

●

aphcowua

spsce(bmc

EF

Tiefceadaoamm

ftiMlsaassswtarr

p

4

dentified five distinct interpretive communities basedn people’s perceptions of ten varied hazards (terror-

sm, the Iraq War, global warming, nuclear power,esticides, genetically modified food, gun control, mar-

juana, legal abortion, and homosexuality) (AL, unpub-ished observations, 2008). Each segment exhibited aonsistent pattern of perceptions across the variousisks, driven in part by the group’s underlying values.ach segment also exhibited unique sociodemo-raphic, political, and religious characteristics, whichere used to label them.Focusing specifically on the issue of global warming,

hree audience segments, representing 63% of peoplen the U.S., were found to have high perceptions of riskssociated with climate change:

The Liberal Left (14% of the total sample). Thesepeople tended to be high SES, nonreligious, white,Democratic women with egalitarian values and aliberal political orientation. They also were muchmore likely to perceive a high degree of risk associ-ated with environmental and technologic threats,and a low degree of risk associated with moralthreats (homosexuality, abortion, and marijuanause).Alarmists (12% of the total sample). This interpre-tive community tended to be religious, low SES,minority women who were politically disaffected.They perceived a higher than average degree of riskassociated with all of the risks assessed (environmen-tal, technologic, national security, and moral).Mainstream Americans (37% of the total sample).This segment tended to have a high school educa-tion, be politically independent, and hold moderatepolitical views. They tended to perceive all hazards asrelatively moderate risks, with the exception ofglobal warming, the Iraq War, and terrorism, whichthey rated as high to very high risks.

Conversely, two other interpretive communities hadelatively low perceptions of risk associated with climatehange:

Optimists (21% of the total sample). Optimiststended to be high SES, white, nonreligious, conser-vative, Republican urban men. They perceived all ofthe hazards, including global warming, as relativelylow risks to U.S. society. They also tended to holdstrong anti-egalitarian and pro-individualist values.The Religious Right (16% of the total sample). TheReligious Right tended to be white, highly religious,conservative, Republican rural men. They perceivedmoral issues such as legal abortion, homosexuality,and marijuana as very great risks to U.S. society, butsaw nuclear power, global warming, and the IraqWar as relatively low risks. They held strong hierar-chical values, and like Optimists, also held strong

anti-egalitarian and pro-individualist values. n

94 American Journal of Preventive Medicine, Volume 35, Num

This study supports the view that there are diverseudiences within the U.S. population, each predis-osed to interpret global warming, along with otherazards, in different ways, drawing on different lifeircumstances, experience, social networks, and valuerientations. This understanding of the underlyingorldview of the various audience segments could besed to tailor messages that resonate with the valuesnd predispositions of each group.

Although the example provided here focuses onegmenting the general public for purposes related toromoting climate change prevention, the rationale foregmenting audiences and tailoring messages is equallyompelling when targeting more specialized audiences,specially those who influence the attributes of placee.g., elected and appointed government officials, smallusiness owners, corporate officials). Moreover, seg-entation will also be an asset when pursuing climate

hange–adaptation objectives.

ffective Climate Change Messagesear Appeals

he climate change literature contains frequent warn-ngs to avoid fearful messages,70–72 yet the more gen-ral persuasive communication literature indicates thatear appeals are effective in motivating behaviorhange, especially if they are accompanied by efficacy-nhancing information.73–75 Witte and Allen’s meta-nalysis75 of 93 fear appeal experiments, for example,emonstrated that there is a positive, albeit small,verage correlation (0.16) between fear and behavioralutcomes, and that the effects of fear are significantlyugmented with stronger fear messages and when fearessages are accompanied by efficacy-enhancingessages.This contradiction—between the warnings to avoid

ear in climate change communication and experimen-al evidence indicating its effectiveness—may be driven,n whole or part, by an artifact of the research methods.

ost research on fear appeals has been conducted inab settings with students as subjects, whereas in naturalettings, the probability that people will choose not tottend to fearful messages can be quite high.76 Wittend Allen’s meta-analysis, however, found that defen-ive reactions are also prevalent in lab studies, withtronger fear appeals engendering stronger defen-ive responses, particularly when combined with aeak efficacy message. Given the potential impor-

ance of fear in climate change communication,dditional research—ideally a combination of labo-atory and field research—is urgently needed toesolve the contradiction.

Pending further research, we recommend that whenotentially fearful content is presented, it be accompa-

ied by strong efficacy-enhancing messages. Relevant

ber 5 www.ajpm-online.net

ertdepp

tMwtwppemcrmbsmd

S

RpilntpsarRatapthitNgms

C

AUmpb

hsgttmcsscis

Iacrpicpagspwnscabctelafcdo

ascstimpfepp

fC

N

fficacy-enhancing information can include identifyingecommended actions through which to reduce thehreat, persuasive affirmations that collectively the au-ience is capable of implementing these actions (i.e.,nhancing perceptions of collective efficacy), and sup-orting the individual audience members’ sense ofersonal efficacy in taking action.Further, we recommend a more nuanced approach

o presenting fear-based or threatening information.ore accurate perceptions of the threat of globalarming might be raised with tailored descriptions of

he potential impacts. Much of the description of globalarming to date has focused on the threat to people,laces, and species psychologically, spatially, and tem-orally distant from most residents of the U.S. Forxample, descriptions of the potential impacts of cli-ate change on polar bears have become increasingly

ommon. It is likely, however, that not all audiencesespond equally to the potential threat to these charis-atic animals. Other segments may be more motivated

y descriptions of the potential human health, nationalecurity, economic, or theological implications of cli-ate change, particularly given the importance (as

escribed below) of personal risk as a motivator.77

elf-Protection Versus Altruism

isk communication research typically finds that peo-le must feel personally threatened for messages to

nfluence behavior.78 However, the political scienceiterature on sociotropic motivations suggests that it isot perceptions of personal threat but rather percep-

ions of societal threat that influence people’s sup-ort for public policies.79 Our current research isuggesting that both forms of perceived risk—personalnd societal—may be relevant in shaping climate-elevant behaviors (CR, unpublished observations, 2008).espondents to a nationally representative survey weresked to assess the seriousness of global warming as ahreat to (1) themselves, (2) future generations, and (3)ll life on earth. They were also asked which of 14ro-environmental behaviors they perform. Correlatinghese two sets of measures, behavior was found to be moreighly correlated to the perceived threat of global warm-

ng to future generations (r�0.25) than to the perceivedhreat to self (r�0.21) or to all life on earth (r�0.22;�11,269, p�0.001, two-tailed). These correlations sug-est that people choose environmental behaviors forultiple reasons, with concern for human progeny as the

trongest of the three.

ognitive Outcomes

lthough polling data indicate that the vast majority of.S. residents believe climate change is happening,any do not understand the science underlying the

henomenon (e.g., there is a persistent erroneous

elief that the hole in the ozone is letting in too much a

ovember 2008

eat),77 the human causes of climate change, or thecientific consensus on this point.63 Clearly, there is areat deal that could be taught: the science, the poten-ial consequences, the contribution of people’s actionso the problem, the changes people can make both to

itigate and adapt, and the skills needed to make thesehanges. A few of these issues are addressed below. Ithould be noted, however, that in the absence oftructural changes that make the promoted behaviorsonsiderably easier, knowledge changes are likely to beneffective, except among those who are alreadytrongly motivated.

mproving people’s understanding of the science. Thergument can be made that so long as people know thatlimate change is dangerous, and they understand thateducing fossil fuel use is the most viable means forreventing further climate change, a full understand-

ng of the physical causes and mechanisms of climatehange is unnecessary. The limited research on thisosition, however, is unclear. Bord and colleagues80

ssessed the importance of actual knowledge aboutlobal warming in explaining people’s intentions to doomething about it; they found that the most powerfulredictor of stated intentions to take voluntary actionsas knowing what causes climate change and what doesot. In a recent study, Leiserowitz,81 however, found noignificant relationship between accurate knowledge oflimate change causes and solutions on the one handnd risk perceptions, policy preferences, or reportedehaviors on the other. He found that, in the U.S.,limate literacy—knowledge of the causes of and solu-ions to global warming—was extremely poor. Forxample, he found that most people incorrectly be-ieved that nuclear power plants, toxic waste, anderosol spray cans cause global warming. Ultimately, heound that overall climate literacy was so poor that itouldn’t explain any of the variance among U.S. resi-ents in terms of risk perceptions, policy preferences,r behaviors.This does not mean, however, that more substantial

nd accurate knowledge is not vitally important. Itimply suggests that many people in the U.S. areurrently relying on a variety of factors other thancientific knowledge of the causes and solutions to formheir climate change attitudes, preferences, and behav-ors. It also suggests that given the limited attention and

ental storing capacity available to most people forroblems like global warming, there is a critical needor research that identifies exactly what factual knowl-dge is the most important and useful—either to helpeople understand the potential risks or to guide theirreferences and behaviors.82

Likewise, many researchers have identified the needor an effective metaphor to explain climate change.ommunicators have been searching for bridging met-

phors, and first-rung theories—simple analogies such

Am J Prev Med 2008;35(5) 495

aco“sestc

cgcepLrmcwihfcIiMkaslscm

Tacoc5tjaswrteetmwa

ch

cdarObmtecnadc

Phaojsekws

ipiepripbt

SmUhsffm

F

Teraaorba

4

s “the heart is a pump” to convey the essential pro-esses and significance of climate change.72,83 An anal-gy is the “ozone hole”—commonly thought of as ahole in the ceiling”—which is often presented as auccessful example of a first-rung theory; it was highlyffective in conveying the problem and motivatingupport for policies to address it—so successful, in fact,hat it’s been difficult to separate it from climatehange in the public mind.

The metaphor of the greenhouse effect has beenriticized because many people aren’t familiar with howreenhouses work, and greenhouses are generally per-eived as good things. Likewise the warming metaphormbedded in the term global warming may sound like aositive change to some individuals.84 In response,ovelock85 has suggested the name global heating. Oneesearch paper that has been influential in the environ-ental community argues that describing climate

hange as “a blanket of carbon dioxide around theorld that is trapping heat” is easily understood and

mproves people’s understanding.72 Unfortunately,owever, although the blanket metaphor may be useful

or explaining how these gases trap heat, it carries noonnotative sense of the threats posed by this process.n fact, warm blankets are likely to evoke positivemages and feelings among an American audience.

etaphors engage people’s embodied experientialnowledge, schemas, and mental models (e.g., a hole inceiling is a bad thing and needs to be fixed) and can

trongly determine subsequent inferential processing,eading people to particular conclusions about theignificance or proper response to an issue. Climatehange still lacks a single, powerful, and encapsulatingetaphor.

he “controversy.” Public uncertainty about the realitynd causes of climate change is fed by an emphasis onontroversy in news stories. The belief that there is a lotf disagreement among scientists over the reality oflimate change is held by 40% of the public, and only7% understand that humans are the cause.63 Al-hough dated, research by Wilson86 found that mostournalists didn’t understand climate change, exagger-ted the debate, and underplayed the scientific consen-us, and Wilkins87 identified the tendency for globalarming stories to emphasize a technologic fix frameather than individual contributions and policy solu-ions. The impact of these frames is largest with audi-nce members who hold ideologies that are not pro-nvironmental88; specifically, the people least inclinedo accept climate change as a serious risk find confir-

ation of their beliefs in the balanced reporting styles,hich present both sides of the controversy when nonectually exists.

In combating the misconception that the scientificommunity disagrees about climate change and its

uman causes, one body of research suggests that c

96 American Journal of Preventive Medicine, Volume 35, Num

ommunicators should not repeat the assertions of theoubters. “Myth-busters” research has found that whenfalse statement is repeated in order to refute it, the

epetition merely serves to reinforce the false belief.89

ver time the refutation is forgotten, but the falseelief has been reinforced simply because the audienceember has heard it repeated again. New assertions

hat make no reference to the false claims are moreffective for refuting myths.90 For example, rather thanounter the statement “Climate change is part of aatural weather cycle” in a manner that repeats thessertion, it is preferable to state: “The scientific evi-ence is clear; human activity is contributing to climatehange.”

otential consequences. Studies show that the publicas difficulty understanding the projections and prob-bilities scientists use to estimate the potential impactf climate change. Moreover, debate around the pro-

ected consequences can result in public apathy andtall policy change.91 Communication, then, shouldmphasize what we know, rather than what we don’tnow. Moser and Dilling advise communicators to “leadith the strongest argument—that is, with the greatest

cientific certainty and confidence.”71

Krosnick and colleagues92 conducted research todentify specific cognitions or beliefs that predict peo-le’s perception of climate change as a serious national

ssue that warrants federal public policy response. Inssence, they demonstrated five key beliefs that predis-ose people to support an aggressive public policyesponse: (1) climate change is real, (2) I am certain its real, (3) is it human caused, (4) it is harmful toeople, and (5) the problem can be solved. Theseeliefs, therefore, can be considered important objec-ives for climate change–communication campaigns.

kills. Although “smart meters” are rapidly gainingarket share in some nations, home energy use in the.S. is still essentially invisible. People receive monthlyome energy bills, which is analogous to receiving aingle non-itemized bill at the end of the month for allood purchases.93 This lack of timely and specificeedback discourages involvement and skill develop-

ent in energy-reduction strategies.

ormative Research

he list of options to reduce energy consumption isxtensive, and presenting consumers with a long list ofecommended actions may create an overwhelmingnd confusing disincentive to action. Efforts to encour-ge behavior change should be preceded by an analysisf which behaviors will have the greatest impact ineducing carbon emissions.94 After the most relevantehaviors have been identified, a careful study ofudience knowledge is needed to ensure that the

ampaign is providing new information, not repeating

ber 5 www.ajpm-online.net

wmbHetilWtos

V

Atasoavri

cipewifcmopttt

M

GbeTa

●

●

●

●

●

C

Aaptwihgotnbcg

N

hat the audience already knows or overlooking infor-ational gaps that render audience members incapa-

le of complying with message recommendations.82

owever, identifying the lack of knowledge is notnough. Research must also identify the other barriershat prevent people from changing their behavior,ncluding time and financial resources, social norms,ack of skills, and structural opportunities to change.

ith all this in mind, campaigners may then select andarget behaviors that (1) will have the maximum impactn carbon emissions, (2) are not overly constrained bytructural barriers, and (3) are new to the audience.

alues and Framing

udiences are most receptive to content that is consis-ent with their existing attitudes and beliefs; selectivettention and avoidance make it less likely that incon-istent information will be received.95,96 Some segmentsf the U.S. population may reject or ignore informationbout climate change if they feel it conflicts with theiralues (e.g., libertarian values) or core beliefs (e.g.,eligious beliefs that assert we should exercise domin-on over nature).69

Choosing message frames for climate change that areonsistent with the values of target groups is onemportant way to make the recommended behaviors orolicies easier to accept. Conservation messages, forxample, can use an economic frame (This is an excellentay to save money); an energy independence frame (This

s a means for our country to free itself from dependence onoreign oil); a legacy frame (This is a way to protect ourhildren’s future); a stewardship frame (This is how I honory moral obligation to protect the abiding wonders and mystery

f life); a religious frame (This is a way to serve God byrotecting His creation); or a nationalist frame (Innovativeechnology will keep our nation’s economy strong). Each ofhese frames is likely to resonate more effectively withhe values of different segments of people in the U.S.

essaging Suggestions

iven these considerations, some educated guesses cane made about the most effective communication strat-gies for the five audience segments introduced above.hese suggestions, however, are only educated guesses,nd should be tested empirically.

The Liberal Left, with relatively high SES and highlevels of education, has a greater propensity to seekand process information deeply, and can likely bereached through a variety of print channels. Becausethis group already perceives climate change as a highrisk, the information of greatest value to them maybe the relative efficacy of various actions they cantake (and how to take those actions) and policiesthey can support, to reduce their personal and their

community’s carbon emissions.97 n

ovember 2008

Alarmists also already perceive climate change as ahigh risk, but some may lack basic knowledge, skills,financial resources, or structural opportunities tochange their climate-related behaviors. Efficacy-enhancing messages, as well as messages that high-light the monetary benefits of some behavioralchanges, are likely to be of value. As their primarysource of information, television is probably themost effective medium.Mainstream Americans are also highly concernedabout climate change. Like the Liberal Left andAlarmists, however, they probably lack a clear under-standing of the changes in behavior they mightmake, and the possible costs and benefits to them-selves, their communities, and the world at largeassociated with those behaviors. Given the size, po-sition, and importance of this segment, we encour-age immediate in-depth investigation of this group’sclimate change perceptions and behaviors with ap-propriate research methods.Optimists, with strongly individualist worldviews andlow perceptions of climate change as a threat, areunlikely to be receptive to most environmentalistmessages about climate change. Messages emphasiz-ing energy independence and the economic benefitsof conservation are more likely to resonate withthem. They may best be reached through newspa-pers and the Internet—their primary sources ofinformation.The Religious Right, who also generally do notperceive climate change as a significant threat, maybe most receptive to messages framed in moralterms, including the stewardship ethic found inGenesis and the moral duty of Christians to help thepoor and needy (i.e., those millions likely to be mostaffected by climate change). Television and talkradio may be the most effective channels for reach-ing members of this audience.

onclusion

recent meta-analysis of the health campaign liter-ture found that, on average, persuasive media cam-aigns evoke personal behavior change among 9% of

heir target audience.98 Somewhat larger effect sizesere found among campaigns that (1) promoted behav-

ors enforceable by law (e.g., seatbelt use); (2) achieved aigher than average exposure to the campaign (i.e.,reater message reach and frequency) among membersf the target audience; and (3) presented new informa-ion (versus information that had already been commu-icated previously in other ways). Regrettably, to theest of our knowledge, no similar analysis has beenonducted to assess the impact of campaigns seeking toenerate public support for policy solutions.For a variety of reasons—including the conservative

ature of meta-analysis, the modest levels of funding

Am J Prev Med 2008;35(5) 497

taibbwa(pipmbPh�nTfs

cvfieuictfa(sumodndiosrae

ismcsrc

syt

phpseait(

eirretfdp

asrattitvsiv

Np

R

1

4

ypically devoted to health communication campaigns,nd the fact that all campaigns reviewed sought tonfluence only individual-level drivers of populationehavior—we see this 9% level as establishing the lowerounds for behavior change that can be accomplishedith public health communication campaigns. Moreggressive communication and marketing campaignsor multiple overlapping campaigns) that target botheople- and place-based drivers of population behavior,

ncluding public policy, when sustained over longereriods of time, have the potential to multiply theinimum effect size into a broad-based shift in societal

eliefs, norms, and practices. The National High Bloodressure Education Program—which is credited foraving helped lower U.S. stroke mortality rates by60%—is one example of such a campaign (www.nhlbi.ih.gov/about/nhbpep/),99 and the Campaign forobacco-Free Kids is a second illustrative example

ocused exclusively on influencing public policies toupport the public’s health (www.tobaccofreekids.org).

At the individual level efforts should be made to craftommunication and marketing campaigns targetingarious strategically important audiences. Among theve audiences identified by Leiserowitz and Slovic, forxample, Mainstream Americans appear to be a partic-larly important target audience given their proportion

n the overall population and their apparent interest inhanging their climate-related behaviors. Communica-ion campaigns targeting such audiences should beocused tightly on providing information that will helpudience members pursue both personal and societali.e., policy) action. Marketing campaigns targetinguch audiences should improve the availability of prod-cts and services that make it easier for audienceembers to reduce their use of fossil fuels and act in

ther ways that reduce greenhouse gas emissions. Ad-itional audience segmentation research is urgentlyeeded to validate and refine, or replace, the segmentsescribed here, and to improve understanding of the

nformation and other factors that will move membersf each segment to action at a personal, family, andocietal level. This will entail conducting formativeesearch to identify cognitive and skills deficits, socialnd environmental barriers that can be modified, andffective framing strategies.At the social-network level there is an urgent need to

dentify and activate popular opinion leaders within alltrata of society, including the government and com-ercial sectors. Personal influence, especially that of

ommunity opinion leaders, is a powerful source ofocial change that will be needed to engage U.S.esidents in responding rapidly to the issue of climatehange.

At the community level, there are fewer models ofuccess on which to base climate change interventions,et the emerging literature in public health indicates

he importance of community-level attributes in driving

98 American Journal of Preventive Medicine, Volume 35, Num

opulation behavior. Social norms campaigns, whichave been shown to be an effective way to influenceopulation rates of a range of conservation behaviors,hould be made a high priority both for their potentialffectiveness and low cost. Campaigns that specificallyddress people’s collective efficacy—the belief that thiss a problem we can solve—may help overcome theendency to continue to overuse common resourcese.g., as happens in the “tragedy of the commons”).

At the place level—local and distal—aggressive strat-gies need to be implemented to improve the availabil-ty and price of products and services that help peopleeduce their carbon emissions, remove structural bar-iers to behavior change, and implement policies thatncourage energy conservation. Among other things,his will require building public support (or demand)or local, state, federal, and multinational policies thatramatically reduce greenhouse gas emissions and helpeople adapt to unavoidable climate changes.All of these actions will require investment of public

nd philanthropic resources. A portion of those re-ources should be directed to conducting translationalesearch that will help to ensure that communicationnd marketing programs created with public, philan-hropic and private sector resources are, in fact, effec-ive at motivating and supporting the necessary changesn population behavior. If the experience of the Na-ional High Blood Pressure Education Program is aalid indicator, public and philanthropic investments inuch research will, in turn, stimulate large investmentsn program development by organizations in the pri-ate sector.

o financial disclosures were reported by the authors of thisaper.

eferences1. Havel V. Our moral footprint. The New York Times 2007 Sep 27.2. Haines A, Patz JA. Health effects of climate change. JAMA 2004;291:

99–103.3. Patz JA, Campbell-Lendrum D, Holloway T, Foley JA. Impact of regional

climate change on human health. Nature 2005;438:310–17.4. Patz JA, Olson SH. Climate change and health: global to local influences on

disease risk. Ann Trop Med Parasitol 2006;100:535–49.5. McMichael AJ, Woodruff RE, Hales S. Climate change and human health:

present and future risks. Lancet 2006;367:859–69.6. Haines A, Kovats RS, Campbell-Lendrum D, Corvalan C. Climate change

and human health: impacts, vulnerability and public health. Public Health2006;120:585–96.

7. Intergovernmental Panel on Climate Change. Climate change 2007: im-pacts, adaptation and vulnerability. Contribution of Working Group II tothe Fourth Assessment Report of the Intergovernmental Panel on ClimateChange. Parry ML, Canziani OF, Palutikof JP, van der Linden PJ, HansonCE, eds. Cambridge UK: Cambridge University Press, 2007. www.ipcc.ch/ipccreports/ar4-wg2.htm.

8. Epstein PR. Climate change and human health. N Engl J Med 2005;353:1433–6.

9. Brownson RC, Remington PL, Davis JR. Chronic disease epidemiology andcontrol. Washington DC: American Public Health Association, 1998.

0. Yach D, Hawkes C, Gould CL, Hofman KJ. The global burden of chronic

diseases: overcoming impediments to prevention and control. JAMA 2004;291:2616–22.

ber 5 www.ajpm-online.net

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2

3

3

3

3

3

3

3

3

3

3

4

4

4

4

4

4

4

4

4

4

5

5

5

5

5

5

5

5

5

5

6

6

6

N

1. Simpson CA, Vuchinich RE. Temporal changes in the value of objects ofchoice: discounting, behavior patterns, and health behavior. In: Bickel W,Vuchinich R, eds. Reframing health behavior change with behavioraleconomics. Mahwah NJ: Lawrence Erlbaum Associates, 2000:193–215.

2. Bickel WK, Marsch LA. The tyranny of small decisions: origins, outcomes,and proposed solutions. In: Bickel W, Vuchinich R, eds. Reframing healthbehavior change with behavioral economics. Mahwah NJ: Lawrence Erl-baum Associates 2000;341–91.

3. Maibach E. The influence of the media environment on physical activity:looking for the big picture. Am J Health Promot 2007;21:S353–62.

4. Epstein LH, Saelens BE. Behavioral economics of obesity: food intake andenergy expenditure. In: Bickel W, Vuchinich R, eds. Reframing healthbehavior change with behavioral economics. Mahwah NJ: Lawrence Erl-baum Associates 2000;293–312.

5. Rogers E, Storey D. Communication campaigns. In: Berger C, Chaffee S,eds. Handbook of communication science. Newbury Park CA: Sage,1986;817–46.

6. Maibach E, Abroms L, Marosits M. Communication and marketing as toolsto cultivate the public’s health: a proposed “people and places” framework.BMC Public Health 2007;7:88.

7. Abroms L, Maibach E. The effectiveness of mass communication to changepublic behavior. Ann Rev Public Health 2008;29:219–34.

8. Farley T, Cohen D. Prescription for a healthy nation: a new approach toimproving lives by fixing our everyday world. Boston: Beacon Press, 2005.

9. Lutzenhiser L. Marketing household energy conservation: the message andthe reality. In: Deitz T, Stern P, eds. New tools for environmentalprotection: education, information, and voluntary measures. WashingtonDC: National Academy Press, 2002:49–65.

0. Deitz T, Stern P, eds. New tools for environmental protection: education,information, and voluntary measures. Washington DC: National AcademyPress, 2002.

1. Mileti DS, Peek LA. Understanding individual and social characteristics inthe promotion of household disaster preparedness. In: Deitz T, Stern P,eds. New tools for environmental protection: education, information, andvoluntary measures. Washington DC: National Academy Press, 2002;125–39.

2. Abrahamse W, Steg L, Vlek C, Rothengatter T. A review of interventionstudies aimed at household energy conservation. J Environmental Psych2005;25:273–91.

3. Th�gersen J. Promoting “green” consumer behavior with eco-labels. In:Deitz T, Stern P, eds. New tools for environmental protection: education,information, and voluntary measures. Washington DC: National AcademyPress, 2002:83–104.

4. Winett RA, Love SQ, Kidd C. The effectiveness of an energy specialist andextension agents in promoting summer energy conservation by home visits.J Environmental Systems 1982;12:61–70.

5. Hirst E, Grady S. Evaluation of a Wisconsin utility home energy auditprogram. J Environ Systems 1982;12:303–20.

6. McMakin AH, Malone EL, Lundgren RE. Motivating residents to conserveenergy without financial incentives. Environ Behav 2002;34:848–63.

7. McClelland L, Cook SW. Promoting energy conservation in master-metered apartments through group financial incentives. J Appl Soc Psychol1980;10:20–31.

8. Hutton RB, Mauser GA, Filatrault P, Ahtola OT. Effects of cost-relatedfeedback on consumer knowledge and consumption behavior: a fieldexperimental approach. J Consumer Research 1986;13:327–36.

9. Van Houwelingen JH, Van Raaij FW. The effect of goal-setting and dailyelectronic feedback on in-home energy use. J Consumer Research1989;16:98–105.

0. Winnet RA, Neale MS, Grier HC. Effects of self-monitoring and feedbackon residential electricity consumption. J Appl Behav Anal 1979;12:173–84.

1. Seligman C, Darley JM. Feedback as a means of decreasing residentialenergy consumption. J Appl Psychol 1977;62:363–8.

2. Hayes SC, Cone JD. Reduction of residential consumption of electricitythrough simple monthly feedback. J Appl Behav Anal 1981;14:81–8.

3. Becker LJ. Joint effect of feedback and goal-setting on performance: a fieldstudy of residential energy conservation. J Appl Psychol 1978;63:428–33.

4. McCalley LT, Midden CJH. Energy conservation through product-integrated feedback: the roles of goal-setting and social orientation. J EconPsychol 2002;23:589–603.

5. Bandura A. Social foundations of thought and action: a social cognitiveapproach. Englewood Cliffs NJ: Prentice Hall, 1986.

6. Winett RA, Leckliter IN, Chinn DE, Stahl B, Love SQ. Effects of television

modeling on residential energy conservation. J Appl Behav Anal1985;18:33–44.

ovember 2008

7. Banerjee A, Solomon BD. Eco-labeling for energy efficiency and sustain-ability: a meta-evaluation of U.S. programs. Energy Policy 2003;31:109–23.

8. Stern P. Information, incentives and proenvironmental consumer behav-ior. J Consumer Policy 1999;22:461–78.

9. Wiser R, Olson S, Bird L, Swezey B. Utility green pricing programs: astatistical analysis of program effectiveness. Berkeley CA: Lawrence Berke-ley National Laboratory, 2004. eetd.lbl.gov/EA/EMP/.

0. Taylor MAP, Ampt ES. Traveling smarter down under: policies for volun-teer travel behavior change in Australia. Transportation Policy2003;10:165–77.

1. Burn S. Social psychology and the stimulation of recycling behaviors: theblock leader approach. J Appl Soc Psychol 1991;21:611–29.

2. Hopper J, Nielsen J. Recycling as an altruistic behavior: normative andbehavioral strategies to expand participation in a community recyclingprogram. Environ Behav 1991;23:195–220.

3. Shrum LJ, Lowrey TM, McCarty JA. Recycling as a marketing problem: aframework for strategy development. Psych Marketing 1994;11:393–416.

4. Dearing JW, Maibach EW, Buller DB. A convergent diffusion and socialmarketing approach for disseminating proven approaches to physicalactivity promotion. Am J Prev Med 2006;31:S11–23.

5. Schultz PW. Changing behavior with normative feedback interventions: afield experiment on curbside recycling. Basic Appl Soc Psychol 1999;21:25–36.

6. Schultz PW, Nolan JM, Cialdini RB, Goldstein NJ, Griskevicius V. Theconstructive, destructive and reconstructive power of social norms. PsychScience 2007;18:429–34.

7. Goldstein NJ, Griskevicius V, Cialdini RB. Invoking social norms: a socialpsychology perspective on improving hotels’ linen-reuse programs. CornellHotel Restaurant Admin Q 2007;48:145–50.

8. Cialdini RB. Descriptive social norms as underappreciated sources of socialcontrol. Psychometrika 2007;72:263–8.

9. Kawachi I, Berkman L. Social cohesion, social capital and health. In:Berkman L, Kawachi I, eds. Social epidemiology. New York: OxfordUniversity, 2000:74–190.

0. Tiedemann K. Price, product life and market share for CFLs. Paperpresented at Advanced Technology in the Environmental Field, Feburary 7,2006, Lanzarote, Canary Islands, Spain.

1. Sandahl LJ, Gilbride TL, Ledbetter MR, et al. Compact fluorescent lightingin America: lessons learned on the way to market. Richland WA: PacificNorthwest National Laboratory, 2006.

2. Barbara M. Wal-Mart puts some muscle behind power-sipping bulbs. NewYork Times 2007 Jan 2. www.nytimes.com/2007/01/02/business/02bulb.html.

3. Saelens BE, Sallis JF, Black JB, Chen D. Neighborhood-based differences inphysical activity: an environment scale evaluation. Am J Public Health2003;93:1552–8.

4. Yancey A, Fielding J, Flores G, Sallis J, McCarthy W, Breslow L. Creating arobust public health infrastructure for physical activity promotion. Am JPrev Med 2007;32:68–78.

5. Frank R. A way to cut fuel consumption that everyone likes, except thepoliticians. New York Times 2006 Feb 16. www.nytimes.com/2006/02/16/business/16scene.html.

6. Paek HJ, Pan Z. Spreading global consumerism: effects of mass media andadvertising on consumerist values in China. Mass communication. Society2004;7:491–515.

7. Valente TW, Schuster DV. The public health perspective on communicat-ing environmental issues. In: Deitz T, Stern P, eds. New tools for environ-mental protection: education, information, and voluntary measures. Wash-ington DC: National Academy Press, 2002;105–24.

8. Darnton-Hill I, Nalubola R. Fortification strategies to meet micro-nutrientneeds: successes and failures. Proceedings of the Nutrition Society2002;61:231–41.

9. Stern PC, Aronson JM, Darley DH, et al. The effectiveness of incentives forresidential energy conservation. Eval Rev 1986;10:147–76.

0. Gardner G, Stern P. Environmental problems and human behavior. NewYork: Allyn and Bacon, 1996.

1. Keating K, Flynn CB. Researching the human factor in Hood River:buildings don’t use energy, people do. Proceedings of the AmericanCouncil for an Energy Efficient Economy Washingon DC: ACEEE Press,1984.

2. Stern P. Changing behavior in households and communities: what have welearned. In: Deitz T, Stern P, eds. New tools for environmental protection:education, information, and voluntary measures. Washington DC: National

Academy Press, 2002;201–12.

Am J Prev Med 2008;35(5) 499

6

6

6

6

6

6

6

7

7

7

7

7

7

7

7

7

7

8

8

8

8

8

88

8

8

8

9

9

9

9

9

9

9

9

9

9

5

3. Leiserowitz A. American opinions on global warming: a Yale University/Gallup/Clear Vision Institute poll. 2007. environment.yale.edu/documents/downloads/a-g/AmericansGlobalWarmingReport.pdf.

4. Maibach E, Rosthschild M, Novelli W. Social marketing. In: Glanz K, RimerB, Lewis FM, eds. Health behavior and health education: theory, research,and practice, 3rd edition. Indianapolis IN: Jossey-Bass, 2002;437–61.

5. Slater M. Choosing audience segmentation strategies and methods forhealth communication. In: Maibach E, Parrott R, eds. Designing healthmessages: approaches from communication theory and public healthpractice. Thousand Oaks CA: Sage, 1995:186–98.

6. Pew Research Center. Little consensus on global warming: partisanship drivesopinion. 2006 Jul 12. people-press.org/reports/display.php3?ReportID�280.

7. Maibach E, Weber D, Massett H, Hancock G, Price S. Understandingconsumers’ health information preferences: development and validation ofa brief screening instrument. J Health Commun 2006;11:717–36.

8. Leiserowitz A. American risk perceptions: is climate change dangerous?Risk Anal 2005;25:1433–42.

9. EcoAmerica and SRI Consulting, Business Intelligence. The AmericanEnvironmental Values Survey: American views on the environment in anera of polarization and conflicting priorities. 2006. ecoamerica.typepad.com/blog/2006/12/american_enviro.html.

0. Moser SC. More bad news: the risk of neglecting emotional responses toclimate change information. In: Moser SC, Dilling L, eds. Creating aclimate for change: communicating climate change and facilitating socialchange. New York: Cambridge University Press, 2007:64–80.

1. Moser S, Dilling L. Making climate hot: communicating the urgency andchallenge of global climate change. Environ 2004;46:33–46.

2. Frameworks Institute. Talking global warming. 2001. www.ucusa.org/assets/documents/ssi/UCSpretotorev.pdf.

3. Boster FJ, Mongeau P. Fear-arousing persuasive messages. In: Bostrom R,Westley B, eds. Communication yearbook, 8. Newbury Park CA,1984:330–71.

4. Sutton SR. Fear-arousing communications: a critical examination of theoryand research. In: Eiser JR, ed. Social psychology and behavioral medicine.New York: John Wiley & Sons, 1982;303–37.

5. Witte K, Allen M. A meta-analysis of fear appeals: implications for effectivepublic health campaigns. Health Educ Behav 2000;27:591–615.

6. Hastings G, Stead M, Webb J. Fear appeals in social marketing: strategicand ethical reasons for concern. Psychol Market 2004;21:961–86.

7. Leiserowitz A. Communicating the risks of global warming: American riskperceptions, affective images and interpretive communities. In: Moser S,Dilling L, eds. Creating a climate for change: communicating climatechange and facilitating social change. New York: Cambridge UniversityPress, 2007:44–63.

8. Weber E. Experience-based and description-based perceptions of long-term risk: why global warming does not scare us (yet). Clim Change2006;77:103–20.

9. Sears DO, Funk CL. The role of self-interest in social and political attitudes.Adv Exp Soc Psychol 1991;24:1–91.

0. Bord RJ, O’Connor RE, Fisher A. In what sense does the public need tounderstand global climate change? Public Understanding Sci 2000;9:199–212.

00 American Journal of Preventive Medicine, Volume 35, Num

1. Leiserowitz A. Global warming in the American mind: the roles of affect,imagery, and worldviews in risk perception, policy preferences and behav-ior. Dissertation. Eugene OR: University of Oregon, 2003.

2. Fischhoff B. Nonpersuasive communication about matters of greatesturgency: climate change. Env Science Tech Online 2007;41:7204–8.

3. Ungar S. Public scares: changing the issue culture. In: Moser SC, Dilling L,eds. Creating a climate for change: communicating climate change andfacilitating social change. New York: Cambridge University Press, 2007:81–8.

4. Auburn A, Grady J. The missing conceptual link: talkback testing ofsimplifying models for global warming. 2001.

5. Lovelock J, Lovelock JE. The revenge of Gaia. New York: Basic Books, 2006.6. Wilson KM. Drought, debate, and uncertainty: measuring reporters’ knowl-

edge and ignorance about climate change. Public Understanding Sci2000;9:1–13.

7. Wilkins L. Between facts and values: print media coverage of the green-house effect, 1987–1990. Public Understanding Sci 1993;2:72–84.

8. Corbett JB, Durfee JL. Testing public (un)certainty of science: mediarepresentations of global warming. Sci Commun 2004;26:129–51.

9. Schwarz N, Sanna L, Skurnik I, Yoon C. Metacognitive experiences and theintricacies of setting people straight: implications for debiasing and publicinformation campaigns. Advances Exp Soc Psychol 2007;39:127–61.

0. Mayo M, Schul Y, Burnstein E. “I am not guilty” vs “I am innocent”:successful negation may depend on the schema used for its encoding. J ExpSoc Psychol 2004;40:443–9.

1. Dilling L, Moser SC. Introduction. In: Moser SC, Dilling L, eds. Creating aclimate for change: communicating climate change and facilitating socialchange. New York: Cambridge University Press, 2007:1–27.

2. Krosnick JA, Holbrook AL, Lowe L, Visser PS. The origins and conse-quences of democratic citizens’ policy agendas: a study of popular concernsabout global warming. Clim Change 2006;77:7–43.

3. Kempton W, Montgomery L. Folk quantification of energy. Energy Int J1982;7:817–28.

4. Stern PC. Toward a coherent theory of environmentally significant behav-ior. J Soc Iss 2000;56:407–24.

5. Wicks R. Message framing and constructing meaning: an emerging para-digm in mass communication research. In: Kalbfleisch P, ed. Communica-tion Yearbook, 29. Mahwah NJ: Lawrence Erlbaum, 2005;331–61.

6. Lazarsfeld P, Berelson B, Gaudet H. The people’s choice. New York:Columbia University Press, 1948.

7. Costanzo M, Archer D, Aronson E, Pettigrew T. Energy conservation: thedifficult path from information to action. Am Psychol 1986;41:521–8.

8. Snyder L, Hamilton M. A meta-analysis of U.S. health campaign effects onbehavior: emphasize enforcement, exposure, and new information, andbeware the secular trend. In: Hornik R, ed. Public health communication:evidence for behavior change. Mahweh NJ: Lawrence Erlbaum Associates,2002:357–83.

9. Roccella EJ. The contribution of public education toward the reduction ofCVD mortality: experience from the National High Blood Pressure Educa-

tion Program. In: Hornik R, ed. Public health communication: evidence forbehavior change. Mahwah NJ: Erlbaum, 2002;73–84.

ber 5 www.ajpm-online.net