Journal of Business Ethics

, Volume 104, Issue 4, pp 571–587

Astroturfing Global Warming: It Isn’t Always Greener on the Other Side of the Fence

Authors

    • ESSEC Business School
  • Martin L. Martens
    • Faculty of ManagementVancouver Island University
  • Hakkyun Kim
    • Department of MarketingJohn Molson School of Business, Concordia University
  • Michelle Rodrigue
    • École de Comptabilité, Faculté des Sciences de l’AdministrationUniversité Laval
Article

DOI: 10.1007/s10551-011-0950-6

Cite this article as:
Cho, C.H., Martens, M.L., Kim, H. et al. J Bus Ethics (2011) 104: 571. doi:10.1007/s10551-011-0950-6

Abstract

Astroturf organizations are fake grassroots organizations usually sponsored by large corporations to support any arguments or claims in their favor, or to challenge and deny those against them. They constitute the corporate version of grassroots social movements. Serious ethical and societal concerns underline this astroturfing practice, especially if corporations are successful in influencing public opinion by undertaking a social movement approach. This study is motivated by this particular issue and examines the effectiveness of astroturf organizations in the global warming context, wherein large corporate polluters have an incentive to set up astroturf organizations to undermine the importance of human activities in climate change. We conduct an experiment to determine whether astroturf organizations have an impact on the level of user certainty about the causes of global warming. Results show that people who used astroturf websites became more uncertain about the causes of global warming and humans’ role in the phenomenon than people who used grassroots websites. Astroturf organizations are hence successful in promoting business interests over environmental protection. In addition to the multiple business ethics issues it raises, astroturfing poses a significant threat to the legitimacy of the grassroots movement.

Keywords

AstroturfingBusiness ethicsClimate changeGlobal warmingGrassroots organizationsLegitimacyRhetoric

Introduction

United States President Obama’s call for action on the issue of global warming may have turned out in vain if public opinion heads in the direction found by a study published by the Pew Research Centre in 2009. The latter shows that a majority of Americans (1) believe that the burning of fossil fuels and other human activities are not responsible for global warming and (2) do not view global warming as a very serious problem (Goldenberg 2009; Pew Research Centre, 2010). These results not only contrast sharply when scientists are surveyed, but also with the Nobel Peace Prize winning report indicating that global warming is unequivocal, and that there is at least a 90% likelihood it is caused by human activity (Intergovernmental Panel on Climate Change 2007).

Why such a disparity? Perhaps this apparent state of great confusion and uncertainty can be somewhat indirectly attributed to the Western Fuels Association delivering, for free, to public and university libraries across the United States hundreds of copies of their Greening of Planet Earth video, which shows that plants on earth are lacking carbon dioxide, and that an increase in atmospheric carbon will provide a more fertile world—a potentially serious avenue for confusion for a university first-year student (Hoggan and Littlemore 2009); or, the Heartland Institute sending thousands of brochures and DVDs to Canadian schools, pushing them to teach their students that scientists have been exaggerating the effects of human activity on global warming (De Souza 2008).

Besides targeting educational institutions, the common thread in both of these examples is the rejection of scientific consensus and “convincing many of the public, and often the media too, that the consensus is not based on ‘sound science’ or denying that there is a consensus by exhibiting individual dissenting voices […]” (Diethelm and Mckee 2009, p. 2). Such efforts to reject the case for taking action to fight threats to society and the environment are part of a larger phenomenon known as denialism, which consists of using rhetorical arguments to appear engaging in a legitimate debate where there is in fact nothing to be discussed (Hoofnagle and Hoofnagle 2010). The same tactics used by the tobacco industry to attack and deny the science of second-hand smoke are being used by many of the same scientists to generate controversy about global warming (Oreskes and Conway 2010).

In the socio-political and organizational context, some of these global warming denialists are known as astroturf organizations. The term comes from “AstroTurf,” which is a brand of synthetic carpeting designed to look like natural grass, but is in fact fake grass (generally used for sports fields). Hence, astroturf organizations are simply fake grassroots organizations usually created and/or sponsored by large corporations to support any arguments or claims in their favor, or to challenge and deny those against them (McNutt 2010). Not surprisingly, the use of astroturf organizations appears to be common for political activities (see, e.g., Krashinsky 2009; Mackenzie and Pickard 2009). This may be a major societal concern if large organizations are successful in using astroturf organizations to influence beliefs and perceptions against established science (McNutt and Boland 2007). For example, public relations associations discourage the use of astroturfing in their codes of ethics because of the deceptive nature of this activity (Fitzpatrick and Palenchar 2006). In this article, we argue that such activity purposefully designed to fulfill corporate agendas, manipulate public opinion and harm scientific research represents a serious lapse in ethical conduct.

This investigation is motivated largely by the denialism, and more specifically, the astroturfing phenomenon described above. Hence, our study specifically focuses more on the rhetoric about the causes of global warming rather than the issue itself. In particular, we seek to determine whether people actually trust astroturf organizations in terms of their existence and their messages regarding this specific issue and more importantly, whether these messages affect or change their beliefs about the causes of global warming (human vs. nature). Accordingly, in this study, we draw upon theory about rhetoric (Green et al. 2008; Hartelius and Browning 2008) as it applies to the language of legitimacy (Suddaby and Greenwood 2005) to conduct an experiment to determine whether information provided by astroturf organizations appears to impact the level of certainty and beliefs concerning the causes of global warming. We argue that such an impact is possible if astroturf organizations themselves are effective, which could constitute grounds for ethical and societal concerns.

Our study extends the business ethics literature in several ways. First, significant ethical concerns stem from corporations taking advantage of their influence to solely maximize shareholder interests rather than benefit other segments of society (see, e.g., Roberts and Bobek 2004). Therefore, similar to Cho et al. (2006), we connect our findings to wider business ethics issues related to the extent to which it is appropriate for corporations to be implicated in astroturfing activities. Second, language as rhetoric creates and represents social reality by shaping perceptions (Preuss and Dawson 2009); more specifically, the intent to provide environmental rhetoric and disclosure remains strategic (Cormier et al. 2004). Hence, we contribute to the stream of literature showing how corporations employ rhetoric in their social and environmental communications to foster their own interests and shape certain developments (Laine 2005, 2010; Milne et al. 2009; Tregidga and Milne 2006). We take a step further in this sense by demonstrating that corporations may employ rhetoric through the use of an intermediary (the astroturf organization) to fulfill corporate environmental agendas. Third, we provide a small contribution to the growing body of literature on rhetoric and legitimacy (Suddaby and Greenwood 2005) by showing how rhetoric could be used to destabilize a prevailing logic by creating confusion by sowing seeds of uncertainty in place of creating clarity and certainty about an issue. Finally, the existing research in astroturfing primarily focuses on the relationship between astroturf groups or organizations, their sponsor companies, and other relevant stakeholders, such as legislators (Hoffman 2011). To the best of our knowledge, no studies to date have examined the direct impact of the rhetoric used by astroturfing.

Background and Hypotheses

Corporate Political Strategies

Suarez (1998) suggests that the influential position enjoyed by business in American politics (Eismeier and Pollock 1988; Smith 2000) constitutes a driver for corporations to develop political strategies that maximize their political power and complement their overall business agenda. This is particularly true when the outcomes of public policy are potentially going to affect their business interests, leading them to become more politically active (Humphries 1991). Hillman and Hitt (1999) integrated the literature on corporate political strategies into a three-stage sequential decision comprehensive model (see Roberts et al. 2003 for a summary of that model). We briefly focus on the third decision phase of the overall model—strategies and tactics—and more specifically on the “constituency-building” strategy.

A constituency-building strategy makes an effort to affect public policy indirectly. Hence, a firm’s constituency and/or the general public is usually involved; such a strategy is most likely to be used during the public opinion formation stage of a particular policy, and if an organization has a large employment or membership base. Typical constituency-building tactics involve grassroots efforts (i.e., generating the involvement of a legislator’s local voting constituency), advocacy advertising, political education programs, public relations campaigns, and press conferences.

Grassroots or Astroturf?

In more general terms, grassroots organizations are defined as “local political organizations which seek to influence conditions not related to the working situation of the participants and which have the activity of the participants as their primary resource” (Gundelach 1979, p. 187). As such, in contrast to traditional political power structures, grassroots organizations or movements often operate at the local level and are generally fueled by community volunteers who give their time and resources to support a specific cause. Their primary objective is therefore to work their way upward from collective efforts to support a local—and often national and global—cause of social and/or political nature that they deem good for society; their procedures generally include hosting house meetings, putting up posters, setting up websites, talking with people on the street, gathering signatures for petitions, raising money from small donors to support political campaigns, and other activities.

However, while grassroots movements are typically known to connect people with pro-social and pro-environmental issues (e.g., human rights, against child labor, against pollution), corporations also appear to engage in grassroots efforts as part of their own constituency-building strategies.1 Grassroots movements can be used by corporations themselves, in the sense that they can make an attempt to influence and convince their local constituents by advocating facts and arguments aligned with their business interests (McNutt and Boland 2007). As discussed above, this case of simulating a grassroots movement is called astroturfing (McNutt 2010). Hoggan and Littlemore (2009) define an astroturf group as a “fake grassroots organization animated by a clever public relations campaign and a huge budget” (p. 36). McNutt and Boland (2007) emphasize the ethical implications when they define it as “synthetic grassroots organizing created for manipulative political purposes” (p. 165). They go on to note that astroturf organizations are “not legitimate political engagement” (p. 172); instead, they constitute a violation of trust and consist of deceptive and phony non-profits engaging in fraudulent fundraising. A commonly noted example of astroturfing activities often mentioned in the general media is the alleged large-scale campaign and funding support from ExxonMobil Corporation toward creating and funding “think tanks” that spread false information about global warming and climate change science (Greenpeace USA 2007).

A few prior studies have examined the issue of astroturfing in organizational and political contexts from various different perspectives and definitions. McNutt and Boland (2007) provide an overview of the astroturfing phenomenon and discuss its implications for nonprofit organizations. Apollonio and Bero (2007) and Tsoukalas and Glantz (2003) describe the astroturfing activities of the tobacco industry in terms of processes, success, and implications for the public. Lyon and Maxwell (2004) take a slightly different definition of astroturfing (i.e., a strategy “in which a firm that knows the state of the world subsidizes the lobbying activities of a group with similar views” (p. 594)) and, using analytical modeling, show that a law requiring disclosure of astroturfing expenditures would reduce the effectiveness of astroturfing, which would be desirable by public decision-makers. Mattingly (2006) investigates the overall process of corporate political actions by interviewing public relations representatives from industry associations and corporations and a representative from the state legislature. He reports that corporations will be more successful in influencing public policy when an organization accumulates sufficient credibility through political action committees, grassroots and the development of relationships with legislators. Astroturfing is acknowledged by a participant as something to avoid because legislators see through the attempt to manipulate. Fitzpatrick and Palenchar (2006) examine whether governments could force corporations to reveal their participation in front groups without violating constitutional rights. Their findings indicate that legislators are expected to approve legislation on the disclosure of front group financing in the coming years. “Front groups are controversial public relations techniques used by organizations to influence public opinion and public policy on behalf of undisclosed special interests. The groups are created to pursue public policy objectives for organizations that disguise their connection (e.g., financial support) with the effort while attempting to appear independent. The typical objective of front groups is to convince public policymakers that citizen support skews in a particular direction or to influence outcomes in local, state, and national elections” (Fitzpatrick and Palenchar 2006, p. 203). Hence, it appears that a “front group” is similar to an astroturf organization, except that it focuses more specifically on influencing policymakers and election outcomes, whereas an astroturf organization focuses primarily on influencing public opinion.

This limited number of prior studies on astroturfing primarily focuses on the more macro-based, organizational (whether it is the corporation per se or the astroturf organization) side of the phenomenon. However, little research has been conducted experimentally at the individual level; for example, on the extent to which astroturfing strategies are effective. We identified a study by Pfau et al. (2007), which assessed the influence of corporate front-group stealth campaigns and examined the effects of post-hoc exposure of their deceptive practices. Using an experimental scenario of a front-group stealth campaign, they find that such campaigns succeed in influencing attitudes toward government restrictions on several business-related issues (e.g., federal legislation to control prescription prices under Medicare).

One specific, but implied objective of astroturf organizations is to increase or instill confusion and uncertainty to the general public about a sensitive issue such as global warming (McNutt and Boland 2007). An example is provided by the American Petroleum Institute’s (API) “Global Climate Science Communication Action Plan.” The stated purpose of this plan is to convince the public, through the media, that climate science is in deep uncertainty. The goal of the plan is to get the average citizen and media to believe that there are critical uncertainties in climate science (Hoggan and Littlemore 2009).

Rhetoric and Legitimacy

A key foundation for the use of corporate political strategies is the concept of legitimacy—“a generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions” (Suchman 1995, p. 574). Corporate political strategies are typically aimed at maintaining the status quo (Tyler 2006), while grassroots activities are typically aimed at creating new norms (Greenwood et al. 2002). This clash between an established norm, value, or belief, termed as an institutional logic, and an emerging norm, value, or belief triggers social conflict (Eisenstadt 1980), as competing interests challenge a contested terrain (Rao and Singh 1999). This contest is a political game, as the competing interests present and contest arguments between a dominant institutional logic and an emergent logic consistent with their self-interests (Hensmans 2003). Hence, rhetorical strategies are a key tool used by actors as they defend established logics against change (Suddaby and Greenwood 2005; Vaara and Monin 2010).

Institutional logics consist of the deeply held and generally unexamined assumptions used by broad social institutions to shape how we interpret and act upon the world at large (Friedland and Alford 1991; Nigam and Ocasio 2010). By unconsciously following established logics, we are able to act upon the world by reducing uncertainty about the world because we place trust in these broader social institutions (Sitkin and George 2005). Multiple social institutions create multiple institutional logics, oftentimes competing and contradictory (Greenwood et al. 2010). In discussing the nature of competing institutional logics, Seo and Creed (2002) argue that actors aware of these contradictions are more likely to act so as to create or inhibit change.

Suchman (1995) argues that legitimacy provides the underlying assumptions that support institutional logics. In evaluating the legitimacy of an issue or argument, actors calculate their own interests and the interests of society, assess whether the argument is normatively correct, or accept it as an inevitable taken-for-granted assumption. Actors presenting competing logics may draw upon these assumptions as a way of contesting the established and emergent logics. We argue that actors presenting competing logics do so by manipulating the levels of uncertainty and trust in the established logics. This can be accomplished by presenting arguments that have an influence on these assumptions.

Legitimacy allows actors to act more confidently in the world by reducing uncertainty about their actions (Ashforth and Gibbs 1990). Actors or entities believe that their actions are appropriate when they engage in behavior consistent with the expectations of the socially constructed system in which they operate. This reduces uncertainty about the appropriateness of actions by conforming to the legitimized expectations of their relevant audience (Aldrich and Fiol 1994). Zimmerman and Zeitz (2002) argue that legitimacy is a basis for decision-making. They contend that social systems create “prescribed scripts, rules, norms, values, and models that are socially reinforced throughout the system” (p. 416) as a way of dealing with the chronic condition of uncertainty found in society. They argue: “When faced with uncertain decisions…, social actors refer back to this stock of scripts, rules, norms, values, and models in order to proceed” (Zimmerman and Zeitz 2002, p. 416). Trust is also an important element of legitimacy. Actors that act in ways that are consistent with the dominant institutional logic are deemed to be more credible, and therefore more trustworthy (Suchman 1995). Conversely, actors that act in illegitimate ways are considered irrational, less credible, and therefore less trustworthy.

At the field level where institutional logics operate, uncertainty and trust are critical to the rhetorics that control how we interpret and act upon the world, especially under conditions of contradictory or competing logics (Lounsbury 2001). Suddaby and Greenwood (2005) argue that contradictions in institutional logics create the conditions by which shifts in logics may occur and that rhetorical strategies determine whether these shifts occur or are resisted (see also Thornton 2001). These shifts happen (or not) “when actors manipulate the degree of uncertainty implied by an innovation” (Suddaby and Greenwood 2005, p. 59). They argue that rhetorical strategies are used to manipulate institutional meaning systems using vocabularies to dampen or “amplify contradictions of meaning inherent in institutional logics in efforts to displace or affirm the dominant logic” (p. 60). Hartelius and Browning (2008) argue that rhetoric is used to sustain or challenge institutional order and the broader discourse supporting this order. Rhetoric supports and shapes the institutional logics in these institutional meaning systems. They contend that we create social order through “dynamic negotiation of socially constructed truths” (p. 23). Rhetoric is used to persuade and manipulate audiences. Rhetorical language can be used to help reduce uncertainty and increase trust by helping people act in an uncertain world, leading to greater legitimacy for the underlying institutional logics. Similarly, rhetorical language can be used to increase uncertainty and decrease trust by making people question the legitimacy of the underlying logics.

Hartelius and Browning (2008) also describe how skilled and persuasive rhetors use a type of indirect communication called an enthymeme as an effective method of persuasion. An enthymeme is a syllogism that contains a nonexplicit assumption (Walton 2001) supplied by the audience. When the audience identifies with the person or organization using this type of indirect communication, the non-explicit assumption is found in the pre-existing beliefs contained in the institutional order. Hartelius and Browning (2008) contend that we can understand the effects of enthymemes by examining the effects of the rhetoric used.

Development of Hypotheses and Research Question

We argue that the global warming issue represents a clash of competing institutional logics and rhetoric used to support these logics. One prevailing logic is that the dominant method for generating energy in today’s world is various carbon-based means (primarily oil and coal). Hence, we suggest that human-caused global warming presents a contradictory logic because it relies on rhetoric that is counter to the carbon-based energy rhetoric. The basic rhetorical logic used by climate scientists argues that global warming is being caused by human activity through the use of carbon-based energy and that global warming is causing harm to the planet (Intergovernmental Panel on Climate Change 2007). This rhetoric rests on normative legitimacy and that the right thing to do is reduce our use of carbon-based energy. Thus, the global warming rhetoric directly challenges the institutional logics supporting an economy based on carbon-based energy consumption. Actors on both sides of this contested rhetorical field of logics are likely aware of the contradictions and challenges posed by the underlying and often unstated assumptions of both sides of the field. In this context, we contend that the carbon-based energy logic is an established and legitimized dominant institutional order, whereas the global warming rhetoric is the emergent challenger. Actors defending the carbon-based energy logic are creating astroturf organizations in order to challenge and destabilize the emergent global warming logic. These actors use the powerful rhetorical methods available through the use of enthymemes. By hiding their true identity and creating an identity that the audience believes to be shared, astroturf organizations are likely to be persuasive because they draw on the pre-existing beliefs of their audience. That is, they can draw on non-explicit premises of the carbon-based energy logic and use them powerfully as implicit tools of persuasion.

These organizations use enthymemes to persuade audiences, not in support of their position, but instead use it against the rhetoric of the emergent challenger. By doing so, astroturf organizations raise levels of uncertainty and beliefs about global warming, thereby decreasing the legitimacy of the global warming rhetoric. If astroturf rhetoric is effective, we hypothesize that:

H1

Viewing information provided by astroturf organizations will increase uncertainty levels about the causes of global warming compared to viewing information from grassroots organizations.

H2

Viewing information provided by astroturf organizations will decrease the belief that humans are responsible for the underlying causes of global warming compared to viewing information from grassroots organizations.

Broad-based disclosures can be used by the actors of the present contested field of logic in order to gain or maintain their legitimacy (Cho and Patten 2007; Deegan 2002; Gray et al. 1995). In addition to the overall message they send through the disclosure of the astroturf organization, proponents of the carbon-based energy logic can resort to specific pieces of information to reinforce the legitimacy of their established institutional order (Patten 2005). Information users are influenced by the disclosure of information and respond by strengthening the legitimacy they confer to the discloser. For instance, disclosing detailed information about the operational activities on the corporate website renders a corporation more credible in the eyes of certain users (Bansal and Kistruck 2006). Similarly, under certain circumstances, positive environmental disclosures are successful in offsetting negative effects in people’s minds (Milne and Patten 2002). In the context of climate change, astroturf organizations could use specific funding information disclosure strategies in order to reinforce their legitimacy. Such strategies could vary from no disclosure about any funding source to conspicuously disclosing that the organization is funded by seemingly environmental foundations or the like. Such strategies could potentially provide more credibility to astroturf organizations while creating an appearance of increased transparency. Hence, we expect that:

H3

Disclosure of funding sources will moderate the effect of astroturf organizations on global warming uncertainty levels and beliefs.

Finally, because we expect different levels of involvement with global warming issues across our sample participants, we were interested in determining whether some individuals would be more or less likely to be influenced by astroturf information. Specifically, we ask the following as a research question:

RQ

Are individuals with low or high involvement in the global warming issue more likely to be influenced by astroturf messages?

Figure 1 depicts the experimental model of the relationships between variables.
https://static-content.springer.com/image/art%3A10.1007%2Fs10551-011-0950-6/MediaObjects/10551_2011_950_Fig1_HTML.gif
Fig. 1

Experimental model

Research Method

In order to determine the extent to which people’s trust and certainty levels about the causes of global warming are affected by the actual astroturf organizations and their messages exhibited on their websites, we conducted a lab experiment using a four-by-two between-subjects design. The section below details the participants, the experimental task, the experimental procedures and the measurement of variables.

Participants and Experimental Task

Two hundred and eighty-four undergraduate students enrolled in accounting classes at a large Canadian university participated in the study in exchange for course credit. Six participants were dropped from the analysis because they did not complete the survey. Seventy-three participants were part of a control group in which they did not view any website and merely responded to three items regarding the global warming issue, such as uncertainty about the cause, perceived importance of global warming issues, and beliefs as to whether human beings are responsible or not. Table 1 reports demographic information concerning the 278 final sample participants.
Table 1

Descriptive data for 278 participants providing usable responses

 

Sample (%)

Age

 20 or younger

40.40

 21–25

53.10

 26–30

3.30

 31 or older

3.20

Gender

 Male

45.80

 Female

54.20

Annual household income

 Less than $30,000

25.20

 $30,000–$49,999

12.90

 $50,000–$69,999

13.80

 $70,000–$99,999

14.20

 $100,000–$149,999

16.00

 $150,000–$199,999

7.50

 Greater than $200,000

9.00

The experimental task and questionnaire were completed in a lab setting, creating a realistic environment for viewing website disclosures and allowing individuals to complete the experiment on their own time in a natural context (Bryant et al. 2004). The experimental task first consisted of answering a series of questions about opinions, knowledge and concern levels on various social issues (homelessness, racism, fair trade, and global warming). To disguise the purpose of the experiment, participants were told that the purpose of the research was a marketing experiment about effective website design for social issues. Participants were told they would be randomly assigned to view a website related to one of these social issues. The next step was to visit a given website and read some information related to global warming issues contained within the various links of the assigned website. These websites were designed expressly for the experiment and were based on an extensive review of real-world grassroots and astroturf websites relative to the types of global warming-related information commonly provided by these two types of websites. This provided a high level of internal validity, while keeping the task externally valid, as well. There were eight versions of the websites, reflecting the eight treatment conditions, in which both the type of organization (astroturf or grassroots) and the funding source (no funding information, “Funded from donations by people like you,” “Funded by ExxonMobil,” or “Funded by grants from the Conservation Heritage Fund”) were manipulated. Other than the content of the messages and the funding source information, the substance of the websites was identical. In other words, website design, structure and length were the same across all eight conditions, and only the information content was manipulated. This provides the optimal setting in which we were able to isolate and measure the effects of the treatment variables.

The website for each condition, respectively, consisted of a “Home page” with links to five other pages pertaining to global warming and the organization’s activities. In the grassroots condition, these were labeled as “About us,” “Key issues and solutions,” “Why act now?” “Get involved!” and “Contact us.” Similarly, in the astroturf condition, the pages links were labeled as “About us,” “Myths/facts,” “Climate science,” “Scientific references,” and “Contact us.” All of the content was based on information found on real-world grassroots and astroturf websites. This design allowed us to create a realistic situation across all treatment groups, and it also created a high level of control over the experiment.

A further manipulation consisted of disclosing information regarding the funding source that supported the organization. The organization’s name in all websites, regardless of the condition, was “Climate Clarity.” In each of the funding source conditions, all web pages within the condition specified who funds the organization (donations, Exxon Mobil or the Conservation Heritage Fund). The “no disclosure” condition did not have any information on funding sources anywhere within the web pages.

Experimental Procedures and Measurement of Variables

Participants were randomly assigned to one of the eight conditions (the number of participants in each condition ranged from 25 to 28) and were asked to type their assigned website’s Uniform Resource Locator (URL). Prior to viewing a corresponding website, participants were first asked questions to establish their level of general uncertainty as a control measure (Lind and Van den Bos 2002). (“There is a lot of uncertainty in the world right now,” “Many things seem unsettled in the world currently,” and, “I cannot predict how things will go in the world in the future” on seven-point scales anchored at 1 = “strongly disagree,” 2 = “moderately disagree,” 3 = “slightly disagree,” 4 = “neither,” 5 = “slightly agree,” 6 = “moderately agree,” and 7 = “strongly agree”), as well as their knowledge and concern about four different social issues (noted above). Their personal involvement in the global warming issue, of particular relevance to our study context, was embedded here using these two measures, such as, “To what extent do you know about global warming issues?” and, “To what extent are you concerned about global warming issues?” both anchored at 1 = “not at all,” 4 = “somewhat,” and 7 = “very much.” Next, the participants were introduced to the first screen containing the home page of the participant’s assigned website. Participants were encouraged to spend sufficient time browsing so that they would be able to answer questions about the website’s content. After reviewing the website, participants continued with the questionnaire by responding to a series of multi-item measures of respondents’ uncertainty toward the global warming issue (“There is a lot of uncertainty about whether or not humans are causing global warming,” “The science about global warming seems quite unsettled currently,” and, “Scientists cannot accurately state whether or not humans are causing global warming,” as well as their beliefs regarding the argument that humans are causing global warming (1 = “inappropriate,” “incorrect,” “scientifically unproven,” “inaccurate,” and 7 = “appropriate,” “correct,” “scientifically proven,” “accurate”). Participants also provided opinions on their perceived importance of the global warming issue (“Global warming is not a serious issue that should cause concern,” and, “The impacts of global warming are so minimal that no policy or legislative response is required” on seven-point scales) and the credibility of information from the website (“I think the information on the website is credible,” “I think the information on the website is exaggerated,” “I think the information on this website is not believable,” and, “I believe the claims on the website” on seven-point scales). Next, participants were asked about their opinions concerning the actual website and organization themselves. In particular, the degree of trust toward the organization and the functionality of the website, modified from McKnight et al. (2002) were also measured. Finally, after completing several demographic questions and manipulation check measures, participants were thanked and dismissed.

Results

Confound Assessments

Prior to our main analysis, a 2 (Organization Type) × 4 (Funding Source) analysis of variance (ANOVA) was performed on the participants’ general perceptions regarding uncertainty of the world and the extent to which they were aware of and concerned about several social issues, including global warming issues, to detect whether differences in these variables might confound changes in uncertainty levels about the causes of global warming or beliefs about human responsibility for global warming, after exposure to messages from an astroturf organization. The analysis revealed no significant differences for general uncertainty level, their perceived knowledge, or concern about global warming issues across all experimental conditions (p values > .10).

We also checked the possibility of differential subjective experience that participants had while viewing a respective website. To address this issue, the respondents’ scores from the perceived web functionality were entered into a 2 × 2 ANOVA, which yielded neither main effects nor an interaction effect (p values > .31). Given the absence of any effect on this measure and the fact that the website was simple to navigate and worked well technically (all Ms > 5.00), we could be assured that any effects we observed would not be attributed to any web functionality factor.

Hypotheses Testing

Results from separate 2 (Organization Type) × 4 (Funding Source) ANOVA analyses on key dependent measures (uncertainty and trust of the argument that human beings are responsible for global warming) and on the perceived importance of the global warming issue are presented in Tables 2, 3, and 4, respectively. These analyses yielded a main effect of funding information. Specifically, there were no significant effects involving the funding source factor, neither main effects nor interactions (p values > .10). Disclosing the funding source of an organization did not affect participants’ perceptions of the dependent variables. Thus, we found no support for H3. For the remaining analyses, we pooled the data for different levels of funding information and added participants in the control condition as a standard comparison, yielding a one-way design with three levels (i.e., astroturf vs. grassroots vs. control).
Table 2

Results of ANOVA analyses testing the effect of organization type and funding source on uncertainty about global warming

Source of variation

Type III sum of squares

Mean square

F statistic

Statistical significance

Corrected model

59.822

8.546

4.723

.000

Organization type

55.778

55.778

30.825

.000

Funding source

.852

.284

.157

.925

Organization type × funding source

2.555

.852

.471

.703

Table 3

Results of ANOVA analyses testing the effect of organization type and funding source on beliefs about whether humans are causing global warming

Source of variation

Type III sum of squares

Mean square

F statistic

Statistical significance

Corrected model

44.560

6.366

4.171

.000

Organization type

36.984

36.984

24.234

.000

Funding source

5.207

1.736

1.137

.335

Organization type × funding source

1.756

.585

.384

.765

Table 4

Results of ANOVA analyses testing the effect of organization type and funding source on perceived importance of the global warming issue

Source of variation

Type III sum of squares

Mean square

F statistic

Statistical significance

Corrected model

14.163

2.023

2.476

.019

Organization type

8.248

8.248

10.092

.002

Funding source

.424

.141

.173

.914

Organization type × funding source

5.159

1.720

2.104

.101

Next, we assessed whether viewing the messages from astroturf organization websites had an impact on participants’ uncertainty levels about the causes of global warming (α = .81). The one-way ANOVA run on uncertainty yielded a significant effect (F(2, 275) = 20.54, p < .001). Analyses involving contrasts showed that participants who viewed information from astroturf organizations were more likely to be uncertain about the causes of global warming relative to those who had viewed information from grassroots organizations (Mastroturf = 4.38 > Mgrassroots = 3.33, p < .001) or those in the control condition (Mastroturf = 4.38 > Mcontrol = 3.21, p < .001). Thus, H1 is supported in that astroturf information significantly weakened people’s certainty about the cause of global warming.

Exposure to information from astroturf organizations also influenced beliefs as to whether human beings are causing global warming, which provides support for H2. Specifically, as indicated by the ANOVA performed on this belief measure (α = .92; F(2, 275) = 16.47, p < .001), individuals who viewed information from astroturf organizations were significantly less likely to believe that global warming was caused by human beings than individuals who viewed information from grassroots organizations (Mastroturf = 4.48 < Mgrassroots = 5.33, p < .001). Significance was also observed in the comparison to people in the control group (Mastroturf = 4.48 < Mcontrol = 5.366, p < .001). Taken together, these results indicate that, after viewing astroturf information, people’s attitudes toward the human responsibility argument tend to decrease in terms of strength (i.e., certainty), as well as in magnitude (i.e., beliefs) and support both H1 and H2.

In addition, exposure to information from astroturfing versus grassroots organizations appeared to have an impact on the perceived importance of the issue itself (F(2, 275) = 4.36, p < .05), in that people who viewed information from a grassroots organization perceived the global warming issue as more important (Mgrassroots = 6.42 > Mastroturf = 6.01, p < .01). However, the comparison between people who viewed information from an astroturf organization and people who did not view any website was only directional (Mastroturf = 6.01 vs. Mcontrol = 6.12, p > .49).

Table 5 presents descriptive statistics of the above-mentioned variables, as well as other measured items, such as credibility of the information and trust toward the organization. Interestingly, participants who browsed a website from an astroturf organization found the information as less credible and the organization as less trustworthy, compared to those who browsed a website from a grassroots organization. Despite such a correct assessment of the message and the target, however, participants’ uncertainty and beliefs about global warming were still significantly affected, as astroturf organizations had intended to instill confusion and uncertainty in the general public regarding the global warming issue. This stark contrast (i.e., not trusting, but still being persuaded) may indicate the power of minimal exposure to astroturfing messages.
Table 5

Descriptive statistics for dependent measures

Dependent variable

Grassroots organization

Astroturf organization

Control

Uncertainty about global warming

3.33a

(1.38)

4.38b

(1.27)

3.21a

(1.51)

Beliefs about whether humans are causing global warming

5.33a

(1.21)

4.48b

(1.25)

5.36a

(1.17)

Perceived importance of the global warming issue

6.42a

(.83)

6.01b

(.98)

6.12a,b

(1.29)

Perceived credibility of website Information

5.08a

(1.06)

4.60b

(1.2)

Trusting beliefs toward the organization

4.74a

(.97)

4.45b

(.98)

Certainty of their trusting beliefs toward the organization

4.41a

(1.31)

4.22a

1.33

Web functionality

5.55a

(.92)

5.42a

(1.04)

Interest in further information from the organization

3.12a

(1.54)

3.45a

(1.78)

Number of participants

104

101

73

Standard deviations are given in parentheses. For each row, means with different superscripts differ significantly (p < .05, t tests)

To further shed light on these effects after viewing information from astroturf versus grassroots organizations, we investigated whether a certain group of people would be more or less likely to be influenced by such information. Specifically, in order to examine the possibility that astroturfing messages might be particularly potent among people who are less involved in the global warming issue, we performed a test on the differential reliance on information in forming judgments. We compared the beta coefficients (i.e., weight) of organization type across groups with either high or low expertise on the global warming issue. To create two groups in expertise, we employed a median split (median = 5.00) on the average of two involvement items (i.e., “To what extent do you know about global warming issues?” and, “To what extent are you concerned about global warming issues?” on seven-point scales) and identified low- and high-involvement groups.

In order to test for statistically significant differences across the two groups, we combined the two equations into one by using dummy coding (Gujarati 1970, 2003):
$$ {\text{Beliefs }} = \beta_{0} + \beta_{ 1} \times D + \beta_{ 2} \;{\text{Organization}}\;{\text{Type}} + \beta_{ 3} \;{\text{Organization}}\;{\text{Type}} \times D + \varepsilon $$
where D = 0 (if a high level of involvement exists), D = 1 (if a low level of involvement exists)

In terms of interpretation, we can conclude that the impact of organization type (either astroturf or grassroots) for a high-expertise/concern group (β2) is not the same as that for a low-expertise/concern group (β2 + β3) if β3 is significantly different from zero. That is, β3 indicates the differential weight of organization type on beliefs for the high versus low group. In our analysis, we found β2 to be significantly smaller than zero (β2 = −.49; p < .05) and β3 to be significantly smaller than zero (β3 = −.67; p < .05). These results indicate that, in terms of beliefs about the cause of global warming, individuals who were not highly involved in the issue were more influenced by exposure to the astroturfing messages, relative to individuals who were highly involved in the issue. As such, these findings echo the notion that individuals who are less (vs. more) aware of or concerned about the global warming issue appear to utilize the information from astroturf organizations to a greater degree, which results in trivializing the emergent logic that human activity is directly responsible for the issue. However, it is noteworthy that even highly involved participants were significantly influenced by the astroturf message, as indicated in β2, which is significantly smaller than zero. This result of our analysis points to the power of using rhetoric to create uncertainty and lower trust in an emergent logic.

Conclusion, Discussion, and Implications

This section summarizes our paper and its contributions, lists the limitations inherent to our investigation and discusses some ethical issues and implications emerging from our study.

Summary and Contributions

The purpose of our study was to determine whether information from astroturf organizations affects people’s trust and certainty levels in terms of their existence and their messages about global warming, and whether these messages actually affect or change their beliefs regarding its causes (human vs. nature).

Analyzed in light of rhetoric as it applies to the language of legitimacy, our results show that, in the clash of competing institutional logics, astroturf organizations appear to be capable of successfully using enthymemes that hide the assumptions contained in the context of carbon-based energy institutional logics to challenge the emergent institutional logic of the global warming rhetoric used by real grassroots organizations. In addition, we found that while participants with a low level of involvement in the global warming issue were more influenced by astroturf information than high-involvement participants, these high-involvement participants were still significantly influenced. Prior research has used rhetorical theory to show how argument structure is used to justify and institutionalize a practice (Green et al. 2009). Our results show how rhetorical practices can be used to lower trust and create uncertainty in order to defeat a competing or emergent logic. In the case of global warming, instead of using rhetoric designed to reduce uncertainty and increase trust in support of their own assumptions, astroturf organizations appear to use misinformation to increase uncertainty and decrease trust about a competing logic, thereby decreasing its legitimacy. Thus, rhetorical strategies are not only a key tool used by actors in support of a logic (Suddaby and Greenwood 2005), but can also act as a tool to attack competing logics. This is politically useful for advocates of a carbon-based economy. Astroturf organizations may be able to minimize or preclude a political debate about activities they support by keeping the debate focused on the emerging logic presented by global warming rhetors.2

A surprising result was the lack of support for the moderating effect of funding source disclosure. We expected that participants who viewed denialism information on a website that claimed to be funded by ExxonMobil would be less trustful of the information, and thus the information would not affect their beliefs about global warming. This group of participants did have lower levels of trust on the website, yet their beliefs were still significantly affected by the information. Such a lack of support for our expectations may be attributed to the differences in user responsiveness driven by different types of disclosure and impression management tactics (Bansal and Kistruck 2006), or that some website disclosure features affect trusting intentions, but not trusting beliefs (Cho et al. 2009). More importantly, however, our research findings suggest that astroturf organizations are effective in creating the sought uncertainty in the minds of people exposed to their message.

Information from astroturf organizations appears to effectively undermine the certainty concerning the causes of global warming, the beliefs that humans cause global warming and the importance of the phenomenon per se. Our research examines a current social conflict in which competing interests are challenging each other in a contested terrain (Eisenstadt 1980; Rao and Singh 1999). Our results support the idea that rhetorical strategies used within broad narratives (Hartelius and Browning 2008) can be manipulated by actors as they attack or defend established logics (Suddaby and Greenwood 2005) by challenging an emergent contradictory and competing logic.

Our research provides an empirical test of Suddaby and Greenwood’s (2005) argument that rhetoric can be used to emphasize contradictions in emergent logics by manipulating uncertainty. A counter-intuitive result comes from our finding that uncertainty about the cause of global warming increases, even when participants distrusted the source of information. This builds on Sitkin and George’s (2005) argument that we are able to act on our trust in a broader social institutional logic because our uncertainty about the world is reduced. Our research results suggest that we may be less likely to follow an emergent logic when we are more uncertain about the effect of our actions. Increasing uncertainty about an emergent logic may destabilize it because we are unsure of the appropriateness of the actions dictated by the new logic.

Limitations

Like all empirical studies, our investigation is subject to certain inherent limitations. First, while our research design focused on a high level of internal validity with preset organization-type conditions, information/messages and funding sources, it required some tradeoffs with external validity. However, because the structure and content of the websites designed for use in our study were based on information found on real grassroots and astroturf websites, we believe our study’s external validity is relatively strengthened. Second, participants in our experiment were university undergraduate students. While we believe these sample members were appropriate subjects for this investigation, given their substantial web reliance, we concede that they may not be representative of the general public. However, many researchers have found that student samples are appropriate, as long as the task is matched to their familiarity level and abilities (e.g., McKnight et al. 2002). In this aspect, we believe that using a student sample was appropriate for this particular task (e.g., Cho et al. 2009). Finally, we tested only one type of social issue—global warming, which may raise the usual concerns about generalizability. However, we specifically chose an important social and environmental issue for which astroturfing is more common and problematic than for other matters. In addition, this specific topic appears to create a conservative setting for the study, given that people are moderately knowledgeable about global warming issues and must have opportunities to form some attitudes as a result of previous exposure to arguments concerning the cause. In this regard, the fact that simply being exposed to astroturfing messages can alter public perceptions and weaken their beliefs as to the human cause somewhat alleviates some of the concerns regarding generalizability. An important next step would be to investigate whether our results hold for other social issues or different types of media employed by astroturfing organizations.

Social and Ethical Implications and Directions for Future Research

Our study examines the direct impact of the rhetoric used by astroturf organizations. Our findings on this issue point towards important social and ethical implications which represent valuable research opportunities for the development of the scientific body of knowledge on the consequences of the astroturfing practice. We briefly discuss some of these potential implications to conclude.

First, the lack of transparency characteristic of astroturfing organizations constitutes serious societal and ethical concerns (Bodensteiner 1997; Oreskes and Conway 2010). From an ethical perspective, all information users “have the right to expect objective, understandable and honest reporting by [an] entity” (Dillard 2007, p. 48). This is far from what is observed in research on astroturf organizations. Indeed, deception is used to manipulate public opinion in favor of business interests through the use of fake grassroots organizations (Fitzpatrick and Palenchar 2006; McNutt and Borland 2007). In our situation, environmental information is employed therein with the purpose of deceiving users (Cormier et al. 2004). More specifically, astroturf organizations can resort to two aspects of deception: confusion and fronting (Laufer 2003). On one hand, they instill confusion in the mind of the public by casting doubt on the importance of a particular issue or by emphasizing the uncertainty surrounding it. On the other hand, astroturf organizations can also act as a “front” for corporations since they constitute a front group dedicated to the opposition of an action in order to protect hidden corporate interests (Laufer 2003). Although they do not constitute the main object of our study, we indeed observe both confusion and fronting in our results. Astroturf organizations are (1) successful in creating uncertainty about the importance of global warming in the mind of the public and (2) utilized by corporations to attack the emerging logic of climate change and defend the carbon-based energy logic which encompasses corporate interests. Future research could investigate more closely how astroturf organizations employ different aspects of deception to influence public opinion in the context of climate change or other social causes. Aside from representing an apparent serious lapse in ethical conduct from corporations, the deception tied to astroturf organizations is likely to pose serious social consequences. The general public is more likely to be deceived by astroturfing than governments primarily due to a lack of awareness about these front groups (Bodensteiner 1997; McNutt and Borland 2007). Future research could examine the impact of astroturf organizations’ message on different audiences. This kind of study would deepen the insights of our initial results implying that astroturf organizations appear to be trying to control the flow of information in an effort to manipulate the public in favor of the corporate agenda.

Another potentially significant implication lies with the impact of astroturf groups on actual grassroots organizations. Our findings not only indicate that corporations are successful in fostering their own interests through astroturfing, but they also imply that this corporate political activity is likely to be detrimental to grassroots organizations (McNutt 2010). Astroturf organizations take the social movement approach to fulfill corporate agendas. Such fraudulent replication of grassroots organizations is likely to raise questions about the legitimacy of those organizations in the mind of the public. Marketing research indicates that a deceptive advertisement generally increases consumers’ suspicion towards other advertisements (Darke and Ritchie 2007). A similar analogy can be drawn between astroturf and grassroots organizations for the public. Once aware that they have been manipulated by a fake grassroots organization, members of the public are likely to be much more cautious and skeptical with all grassroots organizations, whether they are genuine or fake. In other words, the uncertainty created by astroturfing may be likely to reduce people’s desire to participate in legitimate grassroots movements and to support them financially. With people being the core of grassroots organizations (Gundelach 1979), astroturf organizations possibly represent a serious threat to these movements (McNutt and Boland 2007). Hence, the impact of astroturf organizations on the legitimacy of grassroots organizations warrants future research. Such research could shed light on whether corporations, by using astroturf organizations, simultaneously strengthen their business interests and weaken their activist opponents.

Finally, a key potential ethical implication stems from an environmental perspective. Both natural and human arguments shape the ethics of corporate environmental protection. One may argue that the natural environment deserves ethical consideration from the corporation due to its intrinsic value, i.e., plants, animals and all other components of the natural environment are as worthy of being treated by high moral standards as human beings (Hoffman 1991). Others may claim that embedded in the ethics of the collective good is business’ responsibility towards future generations and their environmental rights (Jeurissen and Keijzers 2004), which translate into moral obligations towards the environment today. Astroturfing global warming goes against the ethics of environmental protection. It challenges the ethical arguments advocating the respect of the environment in human activities. Specifically, our results show that astroturf organizations undermine the certainty concerning the causes of global warming, the beliefs that humans cause global warming and the importance of the phenomenon per se. This downplay of humans’ role in the global warming issue is likely to prevent action against climate change from governments, corporations and individuals. This broader, latent, consequence of astroturf organizations could be studied in future research. Indeed, global warming is recognized by many as an important issue (e.g., Bebbington and Larrinaga-González 2008; Romar 2009; Worldwatch 2009) which will have serious social and environmental consequences if not addressed properly and rapidly (Intergovernmental Panel on Climate Change 2007).

Footnotes
1

We acknowledge that grassroots organizations can also promote or support a cause that may or may not be necessarily viewed as positive for society (e.g., the National Rifle Association). In our paper, however, given our climate change context, we will make the implicit assumption that the grassroots organization examples used in this study support pro-social and pro-environmental causes.

 
2

We are grateful to an anonymous reviewer who pointed out this subtle but critical point.

 

Acknowledgments

We would like to express our thanks to Editor Adam Lindgreen, two anonymous reviewers, Sylvie Berthelot, Yves Gendron, Den Patten, and participants of the 12th Annual Alternative Accounts Conference and Workshop in Toronto, the Colloque “Comptabilité, Multivocalité et Diversité” in Rouen, the 2010 Greening of Industry Network Conference in Seoul, the 2010 International Federation of Scholarly Associations of Management Conference in Paris, the 22nd International Congress on Social and Environmental Accounting Research in Saint Andrews, and the 2010 Society for Marketing Advances Conference in Atlanta for their helpful comments and feedback on previous versions of this paper. Charles Cho notes that this project was started while he was at Concordia University and acknowledges financial support received from the Fonds Québécois de la Recherche sur la Société et la Culture (FQRSC) and the Social Sciences and Humanities Research Council (SSHRC) of Canada.

Copyright information

© Springer Science+Business Media B.V. 2011