The Morality Behind Supporting Crowdfunding Campaigns for Eco-Hacktivists

ABSTRACT

Hacktivism, the use of cyber-attacks for a social or political agenda, is becoming increasingly more common. We wanted to investigate if specific aspects of morality and subjective support for nonviolent eco-hacktivism would predict a willingness to donate to a crowdfunding campaign. A total of 350 participants responded to our cross-sectional study. To ensure the dimensionality of our morality components, we ran three principal component analyses (one for each actor in the hacktivism vignette) prior to running our regression model. Our regression model explained 41% of the variance in the willingness to donate to a crowdfunding campaign for the hacktivists. In addition to being younger, there was one significant predictor for each actor in our model. Perceiving the hackers to have higher moral-altruistic beliefs, feeling more moral-social connectivity to the spokesperson, and having more moral behavioral intention with the social media commentator were significant predictors. Preexisting beliefs toward clean water mediated the relationships between moral-altruistic beliefs of the hackers and moral-social connectivity to the spokesperson and a willingness to donate to the crowdfunding campaign. These findings suggest that different perceived morality toward varying actors may better predict support for hacktivism than others, and preexisting beliefs may assist in suppressing objective morality.

The financial cost of cyber-crime is staggering, as there are direct (e.g., ransomware) and indirect costs (e.g., cyber-security counter measures; Anderson et al. 2013; Armin, Thompson, and Kijewski 2016). In 2021, cyber-crime was projected to cost US$6 trillion globally with ransomware costing approximately US$20 billion of that (Morgan 2022). In terms of cyber-security, the costs were expected to exceed US$1 trillion in just a five-year period between 2017 and 2021. This highlights one burden on societies and the need to understand cyber-extremism and support for cyber-extremism in all its forms to identify mitigation strategies.

A recent example of cyber-extremism occurred with the Russian invasion into Ukraine, Anonymous, a decentralized collective of cyber-experts who are known for hacking for political purposes (i.e., hacktivists/hacktivism) launched a concerted cyber-attack on Russian news stations and governmental websites (Tidy 2022; Milmo 2022). These cyber-hacks and interferences effectively shut down numerous pro-Russian sites. Nearly 300,000 accounts liked and supported the Anonymous tweet that announced that they had conducted the cyber-attack (AnonymousTV 2022). The actions were widely viewed as acceptable in the West but were condemned by the Russian authorities. Thus, it is imperative to recognize that the same act could be both perceived as hacktivism (promoting a political cause with no [threat of] violence or self-serving goals to draw attention to a cause [Denning 2001; Eagan 1996; PytlikZillig et al. 2015]) or cyber-terrorism (intending to invoke fear through violence or the threat of violence often along with self-serving goals, such as financial gain [Beck 2008; Embar-Seddon 2002]). Whether an act is viewed as hacktivism or cyber-terrorism would then depend on the individual’s perspective or preexisting beliefs, as is the case with Anonymous’ attack on Russian sites. In turn then, one’s subjective morality would dictate whether this same act was perceived as moral or amoral.

This paper is concerned with hacktivism, a portmanteau of hacking and activism. Hacktivism emerged in the 1990s as hacking but for political means or with a political agenda (Taylor 2005). Prior to this time, hacking was done for experimentation, innovation, gaming, intruding, commercial advantage, or freedom of information. The full history on how hacktivism emerged is beyond the scope of this paper (see Karagiannopoulos 2021 for a more in-depth history of how hacktivism came to be). While we argue that hacktivism and cyber-terrorism are forms of extremism, hacktivism and cyber-terrorism have unique characteristics and are not equivalent (PytlikZillig et al. 2015). Hacktivism and cyber-terrorism are both politically motivated illegal/unauthorized access to computer or electronic systems (Eagan 1996; Embar-Seddon 2002; PytlikZillig et al. 2015). However, the main difference between hacktivism and cyber-terrorism stems from the motivation, or morality, behind the illegal/unauthorized access.

Three common forms of hacktivism are: (1) Distributed Denial of Service (DDoS), whereby hacktivists ultimately increase the number of requests to a website, overloading it, and causing it to crash or malfunction (e.g., Anonymous’ attack on pro-Russian websites); (2) website defacements, where illegal access allows the hacktivists to alter content; and (3) leaking/digital whistleblowing, where illegal access is gained to confidential information that is then leaked on publicly accessible platforms, such as WikiLeaks (Loh 2023). A good example of both website defacement and leaking information occurred when the Ashley Madison website, a dating site for extramarital affairs, was hacked where users found the website with a message of shut down immediately plus approximately 32 million members had their names publicly released to “teach them a lesson” (Doffman 2020). These examples of hacktivism demonstrate how the perceived morality (or amorality) of the hacktivists’ behavior may be blurred depending on an individual’s preexisting beliefs in the “cause.”

For the current study, we chose to examine eco-hacktivism, the use of nonviolent, cyber actions to promote environmental causes or protections (inspired by Eagan 1996), with a combination of website defacement and digital whistleblowing with extortion. There are groups and individuals that have used nonviolent action to protect the environment or raise awareness of environmental crises (e.g., Greenpeace) in person. In Canada, for instance, there are many Indigenous Peoples (e.g., Wet’suwet’en land defenders) who have been fighting to protect their lands from being exploited for oil extraction (Canadian Broadcasting Corporation 2021). Anonymous has also engaged in online eco-hacktivism, like its recent attack on Japan Atomic Energy Agency, Japan Atomic Power, and the Atomic Energy Society of Japan to prevent the release of radioactive water (Japan Times 2023). Yet, historically, there are fewer eco-hacktivism acts compared to what one may expect given the attention the climate crisis receives (Noone and Glover 2023). Hacktivism currently tends to be dominated by political motivation, such as that previously mentioned with Anonymous attacking Russian owned websites after the Ukraine invasion and attacks on the Republican Party of Texas in support of Planned Parenthood. However, as the climate crisis continues, experts have warned that eco-hacktivism could see a drastic increase (Noone and Glover 2023). Thus, understanding the moral support for such acts is crucial for a comprehensive examination of this phenomenon and others similar to it. This then would allow valuable insights into the motivations, justifications, and perceptions of morality that could inform how prosocial unethical behavior and social change are supported by society at large.

In recent years, environmental reports on the amount of plastics and micro-plastics in the world’s waters has raised alarms with environmentalists and individuals (Guzzetti et al. 2018; Walker and Xanthos 2018). We wanted to investigate support for a type of prosocial unethical behavior (i.e., eco-hacktivism) on a salient topic, which is why we chose to focus on a corporation dumping plastics into shared waters. The term prosocial unethical behavior seems paradoxical; however, the intent and motivations behind supporting or participating in eco-hacktivism are meant to protect the environment (i.e., the prosocial) while at the same time the actions to achieve their goals are still unethical or illegal if viewed objectively. Thus, this is one of the overall aims of the current paper: to what extent would participants be more willing to support hacktivism if it aligned with their preexisting beliefs on environmental issues?

How morality can shape our perceptions about unethical behavior

How an act is perceived as legal (or illegal), ethical (or unethical), and moral (or amoral) are inextricably intertwined. For instance, what one perceives as ethical would also be perceived as moral, which would also be perceived as what should be legal. A legal act simply put means that it is not against the laws of a nation. However, laws are constantly in flux, and what today is illegal may be legal in the future or vice versa. Ethics on the other hand can be seen as one’s framework for evaluating and guiding human behavior based on principles of morality (“a code of conduct that, given specified conditions, would be put forward by all rational people” [Gert 2015, p. 1], fairness, and societal norms. Ultimately, someone’s morals are their individual sense of right and wrong, which guide their perception of what is or should be legal or ethical. For the current study, our focus is on moral perceptions of the actors in the hacktivism vignette rather than the legal or ethical components to hacktivism itself, although one’s morality would influence what is perceived as legal and/or ethical.

According to social cognitive theory, morality is created through an adopted form of a learned sense of right and wrong rather than objective moral logic (Bandura 1991). This sense of right and wrong then becomes an individual’s moral standard that influences their behavior through anticipatory, proactive self-regulation to avoid behaving inconsistently to these standards. For example, the act of hacking is objectivelyamoral; however, motivations to help people or the environment through hacking could subjectively be perceived as moral. Some hackers themselves attest that hacking itself is not amoral, but selling the information on how the hacking was done to “criminals” for financial gain or information being posted for free to the public is when the act of hacking becomes amoral/unethical (Jordan and Taylor 1998). Moreover, other researchers have highlighted that hackers perceive their work as legitimate political protest or civil disobedience while others, including those in law enforcement, question the legality and/or morality of the need for hacking (Manion and Goodrum 2000; Karagiannopoulos 2013). These opposing perceptions underscore the subjective nature of these perceptions of morality. These facets support the evidence that suggests that morality or moral reasoning comes from separate cognitive processes where there are objective and subjective differences in morality (Lim, Stocker, and Larkin 2008; Prehn et al. 2008). In the previous example, objective morality can be all hacking is amoral, as it is illegal access to another system; whereas hackers subjectively delineate the act of hacking itself from the personal gain, and hacking only becomes amoral once personal gain is the objective or result. These highlight that we handle objective and subjective morality differently. For this study, we are exclusively focusing on subjective moral facets, particularly: moral subjectivism, cultural relativism, ethical trust, and virtue morality.

Moral subjectivism, also referred to as non-cognitivism, is an individual’s personal moral standards (Monroe et al. 2018). “Do you think their behavior was right?” would be an example of moral subjectivism, where individuals are asked to give their perceptions of right and wrong. Cultural relativism, as we are defining it, is morally relating to a culture and its people who have been harmed. An example of this type of morality would be “Do you think Irish people would support this?” We are operationalizing ethical trust as how individuals trust other’s ethical standards or morality. Thus, “Do you trust their motives?” would exemplify this facet of morality. Lastly, we wanted to investigate virtue morality, which focuses on the traits or characteristics, rather than behavior, of another individual (Annas 2006), with questions similar to “Are they good people?” This is by no means an exclusive list of the types of morality; however, we believe these morality aspects may be more likely to predict support of different actors who were involved in separate components of prosocial unethical behavior in hacktivism, as we discuss more below.

Justifying unethical behavior

What enables individuals to assess behavior, objectively viewed as immoral and unethical, as moral and ethical? Some researchers have suggested that time pressure (Taylor 2009) or taxed cognitive resources (Baumeister et al. 1998) could lead individuals to behave in ways that are inconsistent with their usual morality through the decreased capacity of deeper cognitive processing. However, moral disengagement, where an individual suppresses or disassociates from their usual moral “compass” to engage in unethical behavior without guilt, allows individuals to believe they have retained their moral standards in light of their unethical behavior or support for such behavior (Bandura et al. 1996). For instance, research has found that moral disengagement is associated with unethical behavior from organizational misconduct (Newman et al. 2020) to support for punitive war measures (Aquino et al. 2007) through a variety of cognitive mechanisms.

According to Bandura (1991, 2002), moral disengagement occurs through several cognitive mechanisms, such as displacement of responsibility or euphemistic labeling. These mechanisms then allow individuals to retain a positive self-image in light of unethical behavior (Bandura 2002; Moore 2015). For instance, Bandura (2017) argued that people who retaliate after being victims of a terrorist attack can morally disengage by diminishing their own violence in comparison to the acts of which they were victims. Ultimately, individuals could validate their retaliatory violence by stating their violence was nothing compared to the violence in the terrorist attack, thus morally disengaging from their own unethical behavior. While all the moral disengagement mechanisms vary in the rationale behind the disengagement, they all provide a way to disassociate from or suppress responsibility for the unethical behavior the individual engaged in (see Moore 2015 for a detailed overview).¹ It would then be logical that if participants felt justified in their retaliation (i.e., vengeance against a bigger, more dangerous threat) to support the eco-hacktivists, the conditions are met for moral disengagement, thus participants may demonstrate more support for eco-hacktivists.

While moral disengagement is usually examined in studies that do not contain a prosocial aspect, our study focuses on potential perceived prosociality in support of hacktivists’ actions to protect the environment. Ultimately, individuals could show a more willingness to support eco-hacktivism through a lens of perceived justified moral and/or social superiority. Little research has been done on this area. We argue that it could then follow that specific facets of morality for each actor could be better predictors than others and if the preexisting beliefs toward taking action to clean water from plastics would predict unethical behavior in terms supporting eco-hacktivists.

Support for hacktivism

Hacktivism does not happen in a vacuum. Without some level of public support, hacktivists would likely fail in their objectives to affect change, and, as such, hacktivism would come to a natural end in time. As previously mentioned, hacktivism is utilized for causes or political agendas, yet it is under-researched from a psychological and morality perspective.

Hacktivism has been steadly increasing in frequency, and there are a few reasons that explain this increase (Thackray and McAlaney 2018). Hacktivism can be carried out anywhere in the world, thus proximity to a cause is no longer necessary. It is lower risk compared to in-person public protests, and there is an increase in the audience. Two main goals of hacktivism can be to leak information to draw public or governmental attention to their cause (e.g., WikiLeaks; Caldwell 2015) with the hope of bringing about societal change (e.g., change environmental laws; Sorell 2015). Thus, gaining public backing (macro level) and support from individuals (micro level) are crucial for either of those goals to be effective. One difficulty with researching this area is the gray area or fine line where hacktivism or other prosocial unethical behavior is deemed legitimate or morally correct (Galli 2018). This causes challenges with assessing morality of those who engage in and support these actions.

There is little research on non-state conducted violent eco-extremism and even less on non-state conducted nonviolent eco-extremism that was carried out to defend the environment compared to Islamic extremism, for example (Fletcher 2018; Pickard, Bowman, and Arya 2020). As such, we wanted to investigate support for nonviolent eco-hacktivism that was perpetrated by a non-state organization. Much of the literature and research in regards to hacktivism focuses on the motivations, cognitive reasoning, and morality for the hackers who engage in the unethical behavior. For instance, Chng et al. (2022) found and described 13 types of hackers with hacktivists’ actions led by ideology, notoriety, or revenge, raising the question would those who support hacktivism share similar traits and motivations to those who actually do the hacking? For research that does look at bystander support of hacktivism, it generally examined attitudes and perceptions toward only the hacktivists (Caldwell 2015; Chng et al. 2022; Hadlington 2018; Heering et al. 2020) and rarely their targets (PytlikZillig et al. 2015). New aspects in our research involve investigating participants’ perceived morality toward three unique actors who are involved in the prosoical unethical behavior and the use of crowdfunding as a support mechanism. The actors we utilized in our vignette were: the hacktivists themselves, the hacktivists’ spokesperson, and a social media commentator who demonstrated support for the hacktivism (see Appendix A). Each actor was included to examine different facets of morality and social influences. The hacktivists were to examine direct perceptions of the hacking and hackers. The spokesperson was utilized as an authority figure who justified the hacking and the most public facing figure of the hacktivists. The social media commentator was included to investigate a peer-like social influence from someone who was not directly involved in the hacking but could illicit support through online channels.

For the current study, we wanted to investigate how, at a micro level, individuals could indicate a willingness to support such causes based on their perceived morality of the actors involved in the hacktivism and society’s response to the hacktivism. A progressively more common form of prosocial behavior is crowdfunding. Crowdfunding is a fundraising method, which aims to raise money from a large amount of people to support a cause (i.e., bottom-up, micro-level activism), usually elicited online through social media sites. Our indirect, micro level support for eco-hacktivism was measured by the willingness to donate to an online crowdfunding campaign in support of the hacktivists’ cause. Thus, if individuals have higher moral regard on different facets of morality for the actors and their actions, could these different morality aspects predict more support for the hacktivism?

Anonymity’s influence on hacktivism

One aspect that could influence the results is anonymity. Anonymity is an integral component in hacktivism, where individuals and groups harness the power of online anonymity to engage in activism. Understanding these phenomena is crucial, as they underscore the complex interplay between technology and prosocial unethical behavior. Anonymity can be an influence on hacktivism in several ways. Firstly, anonymity provides protection to hacktivists by concealing their identity (e.g., name, location, IP address), perhaps leading to a decrease in fear of repercussions (legal or social), which in turn could allow hacktivists to partake in prosocial unethical behavior online that they may not engage in in person. The advantages of this anonymity for hacking itself is twofold: reducing the risk of retaliation or being caught while still allowing the individuals to participate in their political movements to affect change. In the case of the hacking collective Anonymous, they additionally use anonymity in order to prevent notoriety (Loh 2023). However, research has found that anonymity can alter individuals’ behavior, such as partaking in more antisocial behavior, such as cyberbullying (Barlett 2015), online aggression (Zimmerman and Ybarra 2016), and general disinhibition (Joinson 2007). More antisocial behavior (or support of antisocial behavior) online can be due to the online disinhibition effect, which ascertains that people engage in more antisocial behavior in part due to the anonymity they perceive to have on the Internet (Suler 2004). In terms of the current study, anonymity of all three actors in the vignette as well as the participants’ anonymity could influence their more willingness to support the hacktivists’ cause compared to if the prosocial unethical behavior had been done in person, such as an in-person protest.

Aims and Hypotheses

The aim of the current study was to investigate if different facets of morality and favorable attitudinal views of three actors in the vignette would predict a more willingness to donate to nonviolent eco-hacktivists. Our hypotheses were (H1) higher levels of concern for the environment (preexisting beliefs) would be associated with and predict more support for the eco-hacktivism, and higher moral regard and connection to the (H2) eco-hacktivists, (H3) spokesperson, and (H4) social media commentator would predict more willingness to donate to the hacktivists’ cause.

Methods

Participants and procedure

Participants were recruited through departmental mail outs at the National University of Ireland, Galway (NUIG), social media posts, participants at talks from the first author, snowballing, and human subject pool. Questionnaires were anonymously completed online. Only those who completed the questionnaire through the human subject pool were compensated with course credit for their participation (n = 90). Full ethical approval was obtained from NUIG Research Ethics’ Committee (19-Aug-21; Amend 1912) prior to data collection. Data were collected from January 18 – March 24, 2020. There were a total of 350 participants (267 female, 79 male, 3 identifying as other, and 1 preferred not to say) retained for analysis.² The average age was 28.25, ranging from 18 to 72 (SD = 11.70).

After the consent and demographic information, participants read a news article that described a company being hacked and blackmailed for dumping plastics into the ocean (Appendix A). Participants were then asked to answer questions on three actors in the news article in this order: (1) the environmental activists (i.e., eco-hacktivists), (2) the spokesperson for the hacktivists, and (3) a person on social media who tweeted support for the hacktivists’ actions. The order remained constant for each participant; however, to avoid any order effects, questions within each section were randomized. The final question asked participants’ their willingness to donate through a crowdfunding campaign to support the activists.

Measures

For the questions on the actors in the vignette, the instructions were identical. For instance, “Thinking about the article you just read, please answer the following questions about your opinion on _____” with a small reminder of who that person/group was (e.g., Enviro-mous are the environmental activists who did the hacking). Followed by, “There are no right or wrong answers, so try to be as honest and accurate as possible.”

The questions for the eco-hacktivist group (Enviro-mous; EM), referred to in the vignette as activists to minimize bias, were created by the authors based on four moral and ethical concepts – moral subjectivism, cultural relativism, ethical trust, and virtue ethics (see Appendix B). Each category had four items with one reversed scored item per category, thus totaling 16 questions. However, upon review, we decided that the item, “Do you think people need to take action to clean up our water?” was included as a control question on the topic, however, as it did not directly ask about the activists, but a more general opinion on environmental issues, we did not include it in the morality scale. Thus, participants were asked to indicate their level of agreement on a 10-point Likert scale from 1 not at all to 10 completely on the 15 remaining items for the hacktivist morality scale.

The second and third sections had 11 questions with one reversed scored item each (see Appendix C and D) on the same 10-point Likert scale as with EM. These sections were identical but were toward the spokesperson or the social media commentator, respectively. The final question asked participants to indicate their willingness to donate to a crowdfunding campaign to support EM and their cause on a 10-point Likert scale with 1 not at all and 10 completely.

Results and Discussion of the Principal Component Analyses (PCAs)

Since the morality scales were created for this study, it was important to verify the dimensionality of the items. To do this, we ran three separate PCAs, one for each actor in the vignette (e.g., the eco-hacktivists, the spokesperson, and a social media commentator). On a univariate level, all variables, except social media use and how much water needed to be clean (preexisting belief measure), had acceptable levels of skewness and kurtosis. Transformations made both social media use and preexisting beliefs variables more leptokurtotic. Thus, we retained the raw data given the large sample size.

All other assumptions were met, and we also assessed the appropriateness of each PCA based on the Kaiser-Meyer-Olkin measure being above .70, a significant Bartlett’s test of sphericity, some correlations of at least .30, and anti-image with some values approaching zero (Tabachnick, Fidell, and Ullman 2007). Components were extracted based on both >1 eigen values, inspection of the scree plot, and nonredundant residual percentages to attempt to avoid under or over extraction (Cota et al. 1993).

PCA for the eco-hacktivists’ morality scale

The Kaiser-Meyer-Olkin measure was .90 with a significant Bartlett’s test of sphericity (p < .001). The total variance explained was 69.36%. Four factors had an eigen value >1, but we restricted the extraction to three factors based on the scree plot. The factors and loadings were obtained using a Promax rotation with Kaiser normalization, and with 56% of the nonredundant residuals with absolute values greater than .05. The percentage for the nonredundant residuals indicates more work needs to be conducted, as it is slightly above the recommended 50% (Field, 2009). If we extracted four components, the percentage of nonredundant variables deemed the solution good. However, only one item then loaded on the fourth component, indicating there was an over extraction and not suitable.

Two items were eliminated for ambiguous loadings.³ The three components were: EM-Moral altruism, EM-Morally perceived subjective norms, and EM-Morally questionable motivations (reversed scored, as a higher score on the questionable motives indicated less support for the hacktivists). EM-Moral altruism consisted of items that related to the justified “rightness” of EM’s actions in the hacking. EM-Morally perceived subjective norms consisted of items with respect to how participants thought others would perceive EM’s actions; whereas EM-Morally questionable motives consisted of items that related to trusting the true intentions of EM’s hacking. The matrix coefficients and component loadings are in Table 1.

Table 1. PCA component loadings for enviro-mous.

Items	Components
	1	2	3
Component 1: Moral Altruism (a = .92) 5.73% sum of squared % of variance*
EM1 Actions were right	.83	.42	.21
EM2 Made the company clean up in the right way	.75	.29	.10
EM4 Actions were right for Ireland	.81	.47	.14
EM5 Actions protect Irish waters and environment	.75	.36	.08
EM8 Trust EM to use money to clean up the oceans	.82	.10	.37
EM9 Trust EM to use excess £10 million for other environmental issues	.81	.12	.39
EM12 Are good people	.81	.32	.29
EM14 Have the qualities of virtuous people	.80	.29	.32
Component 2: Morally Perceived Subjective Norms (a = .71) 2.62% sum of squared % of variance
EM6 Irish people support EM and their actions	.45	.86	.07
EM7 Irish people would reject EM (R)	.17	.85	−.02
EM13 Considered good people by others	.41	.56	.30
Component 3: Morally Questionable Motives (a = .34) 1.73 sum of squared % of variance
EM10 Hack for other purposes (R)	.12	.11	.80
EM11 Ulterior motives (R)	.50	.01	.67
Eliminated Items
EM3 Actions were irresponsible (R)
EM15 EM members are dishonest (R)

Extraction: Principal Component Analysis. Rotation: Promax with Kaiser Normalization.

*When components are correlated, sum of squared loadings cannot be added to obtain a total variance.

PCA for the spokesperson morality scale

The Kaiser-Meyer-Olkin measure was .92 with a significant Bartlett’s test of sphericity (p < .001). The total variance explained was 67.63%. There were two components with eigen values >1. A Varimax rotation with Kaiser normalization was used which resulted in 46% of the nonredundant residuals with absolute values greater than .05, meaning the solution was deemed good. One item was removed for ambiguous loadings. The two spokesperson components were: Spokesperson-Virtue morality and Spokesperson-Moral social connectivity. The Virtue morality component included items that required judgment on different attributes while the Social connectivity component consisted of items that related to how the person would interact with the spokesperson and/or their statement. See Table 2 for the final PCA structure matrix coefficients and component loadings.

Table 2. PCA factor loadings for spokesperson.

Item	Components
	1	2
Component 1: Virtue Morality (a = .87) 35.89% sum of squared % of variance
Spoke6 Intelligent is this person	.81	.27
Spoke7 Well informed is this person	.77	.41
Spoke8 This person has been educated	.82	.21
Spoke9 Respect this person	.71	.48
Spoke10 Criticize this person (R)	.56	.01
Spoke11 Respect this person’s opinion	.62	.47
Component 2: Moral Social Connectivity (a = .86) 31.92% sum of squared % of variance
Spoke1 Would like this person	.52	.64
Spoke2 Follow this person on social media	.32	.80
Spoke4 Do something similar to this person	.04	.84
Spoke5 Like or share this person’s post	.29	.84
Eliminated item
Spoke3 Agree with this person

Extraction: Principal Component Analysis. Rotation: Varimax with Kaiser Normalization. Rotation converged in 3 iterations.

PCA for the commentator morality scale

The Kaiser-Meyer-Olkin measure was .93 with a significant Bartlett’s test of sphericity (p < .001). The total variance explained was 72.27% with two factors having an eigen value >1. As with the PCA for the spokesperson, the components and loadings were obtained using a Varimax rotation with Kaiser normalization. The solution resulted in 40% of the nonredundant residuals with absolute values greater than .05. While one of the components for the commentator was the same as for the spokesperson (Virtue morality), the second component, Commentator-Moral intentional behavior, was different from the spokesperson in that the component consisted of items on how the participants thought they would behave in response to the commentator. See Table 3 for the final PCA structure matrix coefficients and component loadings.

Table 3. PCA factor loadings for Commentator.

Items	Components
	1	2
Component 1: Virtue Morality (a = .91) 43.34% sum of squared % of variance
Commentator1 Would like this person	.70	.52
Commentator3 Agree with this person	.73	.46
Commentator6 Intelligent is this person	.82	.31
Commentator7 Well informed is this person	.81	.37
Commentator8 This person has been educated	.80	.30
Commentator9 Respect this person	.76	.41
Commentator10 Criticize this person (R)	.54	−.38
Commentator11 Respect this person’s opinion	.79	.29
Component 2: Moral Behavioral Intentions (a = .88) 28.93% sum of squared % of variance
Commentator2 Follow this person on social media	.34	.83
Commentator4 Do something similar to this person	.23	.79
Commentator5 Like or share this person’s post	.34	.83

Extraction: Principal Component Analysis. Rotation: Varimax with Kaiser Normalization. Rotation converged in 3 iterations.

We had anticipated that the spokesperson and commentator components’ dimensionality would be identical. However, one item (How much do you think you would like this person?) loaded onto different components between the two actors. In both PCAs, that item had fairly high and near equal loadings on both components though not statistically close enough to be an ambiguous item. This could in part explain why with slight variation it loaded onto different components for the spokesperson and commentator.

After the PCAs, we created the components and checked the Cronbach alphas. All Cronbach alphas except one were excellent (EM-Moral altruism a = .91; EM-Morally perceived subjective norms a = .71; Spokesperson – Virtue morality a = .87; Spokesperson-Moral social connectivity a = .86; Commentator-Virtue morality a = .91; and Commentator-Moral behavioral intentions a = .88). EM-Morally questionable motives alpha was unacceptable (a = .34), resulting in its removal from further analysis.

Results

The vast majority (n = 330) indicated they used social media daily or several times per week with only nine indicating they did not use social media. In terms of participants’ environmental stance, 243 agreed to the maximum that people need to take action to clean water (M = 9.20, SD = 1.54). In the bivariate correlations, the six components that were retained for the regression analysis were all significantly positively correlated with each other (ranging from r = .23 to .76; see Table 4). Preexisting beliefs on environmental action to clean water was weakly, yet significantly, positively associated with the five of the six components from the PCAs (Commentator-Behavioral intentions was the only component not significantly correlated). However, one’s preexisting beliefs on clean water was not associated with the willingness to donate to the eco-hacktivists’ cause.

Table 4. Zero-order correlations.

	1	2	3	4	5	6	7	8	9	10
1 Gender ID
2 Age	−.13*
3 Social Media Use	.08	−.21**
4 Clean water	.16**	−.05	.09
5 EM-Altruism	.01	−.12*	.08	.25**
6 EM-Perceived subjective norms	−.03	.01	.03	.12*	.43**
7 Spoke-Virtue	.02	−.03	.05	.26**	.71**	.39**
8 Spoke-Social	−.05	−.04	.06	.14**	.67**	.34**	.67**
9 Comm-Virtue	.08	−.09	.05	.23**	.67**	.32**	.76**	.62**
10 Comm-Behavioral intention	−.06	−.07	.11*	.04	.50**	.23**	.44**	.76**	.64**
11 Willing to donate	.02	−.22**	.13*	.09	.50**	.16**	.42**	.59**	.47**	.53**

*Significant at .05 level (2-tailed) **Significant at .01 (2-tailed).

In the hierarchical regression analysis to predict willingness to donate, all assumptions were met. There were some strong correlations, but none were > .80, tolerance ranged from .26 to .99, and VIF ranged from 1.01 to 3.93, indicating there were no issues with multicollinearity (Kutner, Nachtsheim, and Neter 2004; Tabachnick, Fidell, and Ullman 2007). In model 1, we included age, social media use, and how much participants indicated action to clean water was needed (preexisting beliefs on the environment).⁴ The model was statistically significant predicting 6% of the variation in the outcome (F[3, 346] = 7.63, p < .001). In model 2, the six PCA components for the actors in the vignette were added. The additional variables significantly improved the predictive value of the model, R² change = 36.1%, F(6, 340) = 35.45, p < .001. The model as a whole explained 42.3% of the variation in the outcome (Adjusted R² = 41%; F[9, 340] = 27.70, p < .001). Being younger and higher on EM-Moral altruism, Spokesperson-Moral social connectivity, and Commentator-Moral behavioral intention were significant predictors of a willingness to donate to a crowdfunding campaign for the hacktivists’ cause (see Table 5).

Table 5. Willingness to donate to environmental hacktivists through crowdfunding coefficients.

Model	1*				2*
Predictor	β	t	sig	CI(95%)	β	t	sig	CI(95%)
Age	−.20	−3.77	<.001	−.06, −.02	−.16	−3.69	<.001	−.05, −.01
Social Media Use	.09	1.61	.11	−.06, .56	.05	1.22	.22	−.09, .40
Clean Water	.07	1.33	.18	−.05, .27	−.03	−.57	.57	−.17, .10
EM-Altruism					.19	2.91	.004	.01, .05
EM-PSN					−.08	−1.72	.09	−.08, .01
Spoke-Virtue					−.03	−.38	.70	−.04, .03
Spoke-Social Con.					.34	4.18	<.001	.05, .14
Comm-Virtue					.07	.85	.40	−.02, .04
Comm-Beh intention					.15	2.02	.05	.00, .10
R^2	.06				.42
Adjusted R^2	.05				.41

*p < .001.

EM=Enviro-mous; PSN=Perceived subjective norms; Spoke=Spokesperson; Con=Connectivity; Comm=Commentator; Beh=Behavioral.

To analyze the effects of preexisting beliefs, we ran mediation analysis. The mediation path was preexisting environmental beliefs (predictor variable) and the three morality components that were significant in the regression model (mediator variables) to willingness to donate to the hacktivists’ cause (outcome variable). The mediation analysis was conducted in Process v.4.0 (Hayes 2012). Preexisting beliefs significantly predicted EM-Moral altruism and Spokesperson-Moral social connectivity but not Commentator-Moral behavioral intention (see paths a¹, a², and a³ in Figure 1, respectively). We found that the indirect effects of preexisting beliefs through EM-Moral altruism and Spokesperson-Moral social connectivity were statistically significant, as there was no zeroing effect on the confidence intervals (CIs; see Figure 1 paths a*b¹ and a*b²). Zeroing means both CIs were either below or above zero, and this indicates significant mediation is occurring. Thus, preexisting environmental beliefs was completely mediated in those two relationships to the outcome variable. There were no significant indirect effects (mediation) with preexisting beliefs through Commentator-Moral behavioral intention with willingness to donate to the hacktivists’ cause, as can be seen by the zeroing of the CIs (path a*b³ in Figure 1).

Figure 1. Pre-existing beliefs and willingness to donate without mediators in the Model.

Direct and Indirect Effects of Pre-existing Environmental Beliefs with Meditators in the Model

All coefficients are standardized. Analyzed by Process v4.0 (Hayes 2012). Bootstrapping was 5,000, and the CIs were set at 95%.

Discussion

Hacktivism, a nonviolent form of prosocial unethical behavior in terms of physical damage, has become an increasingly popular tactic for groups to push their social justice and political agendas. Additionally, there is an increasing amount of online crowdfunding campaigns from asking for assistance with medical bills, research, to support for victims of environmental disasters (GoFundMe 2022). We wanted to investigate if crowdfunding could be a mode that individuals would be willing to donate to in order to support groups who engage in online prosocial unethical behavior, such as hacktivism.

Younger people more willing to donate

In the regression analysis, we had partial support for our hypotheses. Age was one of the strongest, and only background, variables to be significant in line with previous research (Holt and Kilger 2012; PytlikZillig et al. 2015). This in part could be due to the fact that younger people use social media more, as was found in the bivariate correlations. Crowdfunding campaigns are now nearly advertised exclusively through online avenues, such as social media sites like Twitter, Facebook, and Instagram. According to their annual report, GoFundMe, one of the biggest online crowdfunding campaign sites, raised nearly $5 billion in 2021, indicating the prevalence of online crowdfunding (GoFundMe 2022). Thus, if younger participants had more familiarity and trust of online crowdfunding donations, this could explain their willingness to donate. Alternatively, older participants may have been less trusting of crowdfunding and less willing to donate. Younger people have long been involved in activism. However, since 2018, this involvement has intensified especially surrounding environmental causes. Younger people, like Greta Thunberg, and organizations, like RisingUp!, have been heavily involved in environmental activism (Pickard, Bowman, and Arya 2020). Thus, our results are in line with previous research where younger individuals are more likely to contribute to environmental causes even if the causes are using less conventional means (Bowman 2020), such as hacktivism in the current study.

Pre-existing beliefs

One of the main aims and hypotheses of the current study was to examine if participants were more willing to support prosocial unethical behavior if it aligned with their preexisting beliefs. In the bivariate correlations, the preexisting beliefs on the importance of taking action to clean waters was weakly associated with the vast majority of morality components, indicating that those who believed action to clean water needed to be taken regarded the eco-hacktivists and their supporters more morally favorably. However, the preexisting belief was not significantly correlated with the willingness to donate. In the regression analysis, like the bivariate correlation, our hypothesis (H1) was partially supported that an alignment with preexisting beliefs indirectly predicted a willingness to donate to support the hacktivists’ cause. The more participants believed the environment was an important issue, the more moral support that was predicted for EM and the spokesperson, which then predicted a more willingness to donate (preexisting beliefs was completely mediated by those two morality facets). Thus, preexisting beliefs were only significant through indirect effects. Other researchers found that participants’ preexisting attitudes toward hacktivism did predict some support for hacktivism, but not for all forms (PytlikZillig et al. 2015). However, the researchers did not ask about general ideology on the topic but rather if participants supported the concept of hacking (versus cyber-terrorism), which may explain their direct effects and our indirect effects. It is interesting that there was no mediation effect with the social media commentator. This is a novel aspect in our study, and further research is needed to explicate this finding.

Moral perceptions and willingness to donate

A few studies had previously investigated those who support, but do not directly do, the hacking (Heering et al. 2020; PytlikZillig et al. 2015). PytlikZillig and colleagues had a small study (N = 78) where they found that overall perceived shared values and integrity of the hackers were the most reliable predictors. In the open-ended questions, participants indicated that the morality of the hackers was an important factor in whether support was shown. Heering et al.’s (2020) research did not focus on the perceived morality of the hackers but on support for what they called social banditry (i.e., support for others taking from the powerful to help those less powerful). We argue even though morality was not measured in their study, it could have had moral facets to it that led to more support of the hacktivists. In the current study, our hacktivism contained a prosocial component where the hacktivism was to extort money to clean up the dumping of plastics the company had been doing. Thus, it could be the prosocial element that increased support for the hacktivism, which was missing from the climate hacktivism scenario in PytlikZillig et al.’s (2015) studies. In Heering et al. (2020) research, there was an aspect of prosociality in their first study, but their second study had a more self-serving aspect, where participants would directly gain from the hacking, potentially explaining why participants indicated more support.

In the regression model, two facets of morality were included for the hacktivists: Moral altruism and Morally perceived subjective norms. Only perceived moral altruism of the hacktivists was significant, partially supporting our hypothesis (H2). The Moral altruism facet was a perception of the justified actions of the hacktivists. This is in line with previous research where hacktivists being perceived as benevolent predicted more support for them (PytlikZillig et al. 2015), and altruistic beliefs would align with a benevolence factor. Moral altruism’s significance could also be explained by the hacktivists’ actions being seen as a utilitarian (fixing a current issue and preventing similar actions in the future) or retributive punishment (i.e., victim blaming in such that the target brought it on themselves; Weiner, Graham, and Reyna 1997). Some research has found that utilitarian punishment garners more support while others have found retributive does (Carlsmith and Darley 2008; PytlikZillig et al. 2015). However, Carlsmith (2008) found that participants will explicitly claim their reasoning for punishments were more in line with utilitarian concepts, yet the behavior of awarding punishments was more in line with retributive punishments. This indicates there is a disconnect between people’s perceptions of what they support and their actions. Since the hacktivism in the current study could be perceived as either (or both) utilitarian and retributive, the Moral altruism may be influenced by that perception regardless of how the participants viewed the hacking. Retributive or utilitarian punishment perspectives toward supporting unethical behavior are highly individual and rely on personal approaches. This then may explain why morally perceived subjective norms was non-significant in our model because participants do not consider how others would support (or not) the crowdfunding campaign, thus a morality facet of how others attribute support could be insignificant to one’s behavior.

Only one of the morality facets for the spokesperson for the hacktivists (Moral social connectivity) was a significant predictor in participants’ willingness to support a hacktivist crowdfunding campaign (H3). Thus, our findings partially support our hypothesis. The spokesperson was included in the study as often hacktivists, like Anonymous, share their hacking widely through social media platforms, and we aimed to use the spokesperson to represent this aspect while putting a face to the hacktivism that was one step removed from the hacktivists. Moral social connectivity encompassed how participants would engage with the spokesperson online and feel connected, and this could have given participants a feeling of the spokesperson being less anonymous than the hacktivists. Thus, if the spokesperson were perceived as less anonymous compared to the hacktivists, it could explain why the moral social connection facet was the strongest predictor of the willingness to donate. This aspect of anonymity or perceived anonymity was not explicitly examined in the current study, but it could be beneficial in future research to investigate how much anonymity is perceived of each actor and if that influences participants’ willingness to support the cause. While to the best of our knowledge, there is no extant literature with a spokesperson being included in the morality judgments of predicting unethical behavior, one study did find that online political activism did predict more disruptive acts participants were willing to engage in (Holt and Kilger 2012), demonstrating that connection to action may be a better predictor than the dispositional virtue of an actor. Thus, this could also explain why Moral social connectivity with the spokesperson was the strongest predictor in our model. In the PytlikZillig et al. (2015) study, they had found that attitudes toward the hacktivists were more important compared to the targets. In their study, the hacktivists were the public representation of themselves; whereas, in the current study, the spokesperson was the most public face of the hacktivism. Therefore, it could be that it is not necessarily the hacktivists, but the perceived morality of the actor who most engages with society, as the most crucial in predicting support through online crowdfunding.

Another explanation for this could be higher social connectivity of the public face of the hacktivism could lead to more moral disengagement. Two ways in which participants could have morally disengaged are through moral justifications, which portray the unethical behavior as an act benefiting society, and advantageous comparison, which allows individuals to downplay the severity of the behavior by indicating the unethical behavior was not as severe as what caused it (Bandura 1999, 2017; Moore 2015; Sykes and Matza 1957). Both of these mechanisms could be found in how the spokesperson justified the hack in the article (“While this hack may seem too extreme for some, it pales in comparison to the damage PRQ was doing to the oceans and environment”). Thus, if participants felt that dumping plastics into the water was worse than any financial cost or loss of jobs due to the hack, the conditions for moral disengagement may have been met. We tried to balance out the support for the hacktivists with mentioning specifically how both the government-run company and employees were harmed by the hacking. However, that may not have sufficed if the hacktivism was still perceived as prosocial unethical behavior and participants were more socially connected to the actor stipulating these facets. Moral disengagement has additionally been found to be associated with moral relativism (Moore 2015). Moral relativism is a very broad generalized concept; however, based on our findings, the perceived morally altruistic intentions of the hacktivists and the perceived moral social connectivity to whomever the public face of the hacktivists is may provide more specific insight into the relationship between moral disengagement and moral relativism. Future research should investigate if specific aspects of morality are more associated with the propensity to morally disengage and how crowdfunding may inherently be a platform that allows individuals to morally disengage and support unethical behavior.

The social media commentator actor in the vignette was to investigate if a non-involved peer-like actor (i.e., social influence) perceived morality would predict a willingness to donate. The moral behavioral intention facet was significant, thus our hypothesis (H4) was partially supported. Again, no previous research has examined this type of actor to which to compare our results. However, there is a significant amount of research that has found participants, especially adolescents or young adults, support more antisocial behavior (Eamon 2001; Thompson, Mehari, and Farrell 2020; Yavuzer et al. 2014) and positive activism (Altbach 2007; Li and Liu 2021; Rosenberg 2013) based on peer influence. Thus, our findings are in line with this previous research. The effect may have been stronger if the social media commentator had been a friend/colleague, known social media contact, or a known social media influencer.

Surprisingly, the virtue morality facets, a specific type of moral relativism, for either the spokesperson or social media commentator were non-significant predictors in our model. This indicates that the willingness to donate to an eco-hacktivist crowdfunding campaign was not predicted by perceived moral dispositional facets when the other types of morality were included in the model. As these actors were novel aspects in our study, further research is needed. One aspect of the study was we gender neutralized all actors. It may prove interesting in future research to ask participants which gender participants perceived the actors to be and analyze if that perception influences both morality facets and the willingness to support prosocial unethical behavior.

Limitations

The study was cross-sectional in nature, and, therefore, causation cannot be determined. There are many forms of hacktivism, and the predictors of one may not predict another. Moreover, predictors for online prosocial unethical behavior may not be predictors for physical in-person forms of prosocial unethical behavior. Thus, generalization of our findings regarding eco-hacktivism must be done with caution to other contexts. However, conducting the PCAs provides a strong base for future research on different morality facets to be investigated in other forms of unethical behavior. We euphemized, or sanitized, how the hacktivists were named in the vignette (i.e., activists) primarily to attempt to avoid bias, but this is also a main aspect in the process of moral disengagement (Bandura 1991, 1999, 2002, 2017). Thus, sanitizing the naming of the hacktivists may have assisted with moral disengagement, creating a more willingness to donate. Had we referred to them as eco-extremists or cyber-terrorists, we could have biased the results in the other direction. It may be important to note that the term hacktivist in and of itself could be perceived as euphemistic language, and the use of language should be examined in future research to investigate if the terms produce different results.

Summary of the discussion

• Younger people were more willing to donate to support the hacktivists’ cause, which in part may be due to younger people being more familiar and comfortable with crowdfunding platforms.

• Participants’ preexisting beliefs that environmental causes are important predicted higher moral regard from the hackers and their spokesperson, which in turn predicted more willingness to donate to a crowdfunding campaign.

• The more participants perceived the hacktivists to be justified in their actions (Moral altruism), the more they were willing to support the hacktivists through crowdfunding.

• The more Moral social connectivity the participants had for most public-facing actor (or potentially least anonymous actor), the spokesperson for the hacktivists, the more willingness there was to donate to support the hacktivist crowdfunding campaign. This connection facet of morality to, rather than the disposition of, the spokesperson was the strongest predictor of participants being willing to donate. No extant research has previously investigated this facet.

• Participants’ reported moral behavioral intention to the social media commentator predicted more willingness to donate to the crowdfunding campaign. This type of actor in hacktivism has also not been investigated before.

Conclusion

With technological advances, there is always a positive and negative side, and hacktivism and cyber-terrorism are the essence of this. Overall, only specific facets of morality may predict support for prosocial unethical behavior. With online activism and increased potential for hacktivism, our results could predict how hacktivism will be perceived by many as well as predict the attributes of hacktivism, which may lead to financial support by society at large. This then could create bottom-up activist/extremist movements. While in research, it is generally the motives or trust of the hacktivists themselves that garners the most support, the moral social connectivity of the spokesperson for the hacktivists was the strongest predictor of willingness to donate to a crowdfunding campaign in our study. It could be then that it is the public face of the hacktivism that requires the higher moral perceptions in order to predict support, not just the perceptions of the hacktivists. Perceived moral “goodness” of the hacktivists, however, was also an important and significant moral facet. As younger people were more willing to donate to the hacktivist cause, it could indicate crowdfunding could become an important method to gain financial support for hacktivism going forward, and as such, counter-extremism measures should investigate this to supplement interventions. Based on our findings, crowdfunding could be a new social movement perspective (bottom-up social movements) on supporting activism.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Additional information

Funding

This research was supported by the Hardiman Scholarship at the National University of Ireland, Galway.

Notes on contributors

Christie Tetreault

Christie Tetreault received her PhD from the National University of Ireland, Galway (now University of Galway). She is currently a post-doctoral fellow with the Centre for Forensic Behavioural Science and Justice Studies at the University of Saskatchewan. Her research focusses on how group dynamics, cognitive shifts, and dispositions influence susceptibility and resilience to various forms of extremism and interpersonal violence.

Kiran M. Sarma

Kiran Sarma is an Associate Professor of Psychology at the University of Galway (formerly National University of Ireland, Galway), a chartered forensic psychologist and Chair of the Division of Forensic Psychology of the Psychological Society of Ireland.

Notes

¹ There are other theoretical frameworks that could also explicate support for prosocial unethical behavior, such as Sykes and Matza’s (1957) neutralization theory, whereby there are “neutralizations,” such as euphemisms, individuals use to assuage their guilt of or justify partaking in antisocial behavior.

² Three were removed for being multivariate outliers; six for reading the article too quickly, and two for being under the age of legal consent.

³ Ambiguous loadings were removed when an item had loadings on two or more components with less than .10 difference.

⁴ Gender was not included as it was not significantly correlated to the DV, and analysis revealed it acted as a suppressor variable in the model.

Appendix A

Full Text of News Article

On January 10, 2019, international media reported that PRQ Chemicals, a Welsh State-owned oil production company based in Cardiff, had been dumping its excess plastic and micro-plastics into the Irish Sea. The plastics were washing up on local and international shores along with many marine animal carcasses that had died from chocking on or ingesting these plastics. Environmental experts estimated that it would cost £30 million and a decade to clean up the illegal dumping.

On March 19, 2019, PRQ Chemicals were subject of a ransomware attack by a new group of environmental activists, hackers called Enviro-mous. The hackers were able to gain control over all of PRQ’s computer systems and halt all production and distribution. The hackers promised to give PRQ their computer systems back once £40 million was paid, the cost of the cleanup and extra for other environmental causes in the area. While no one was physically harmed in the ransomware hack, PRQ lost an estimated £200 million in profit and had to lay off approximately 30% of its workforce until they regained control over their computer system, three months later.

Enviro-mous issued a press release through a spokesperson on social media, Earth Defender @95Earth:

While this hack may seem too extreme for some, it pales in comparison to the damage PRQ was doing to the oceans and environment. They have been dumping plastics for decades but finally got caught, and despite warnings of the harm they were causing by their own scientists, they continued to dump their waste into the waters we and marine life depend on. At some point, someone had to do something to get them to pay attention and stop their illegal practices. We felt we were in a good position to shine a light on their actions for the public and force them to take the environment, animals, and human lives seriously. The dumping cannot continue, and we forced them to stop.

The spokesperson added that Enviro-mous would be doing a crowdfunding campaign to “out” other companies that were destroying the environment and other possible ransomware attacks if needed, so any financial support would be welcomed.

There has been strong support for the hackers on social media. For example, one Twitter user, @keepitclean87, in a thread showed their support:

“This is the govt [government] being the govt, again. They knew what was going on & looked the other way. #Greedy. It’s 2019 and still happening. Our oceans are dying from plastic. If the govt still isn’t listening, then we have to force them to. The attack made the govt do that! Good! 1/3

For those saying it wasn’t an appropriate response, wake up! Someone had to do something! Standing up to the govt for the damage they’re doing. Dumping does a lot more damage, and the ransomware hack was just to clean up the govt’s mess anyway. 2/3

What water do you expect to drink from if it’s all polluted? What the govt is doing to OUR water is FAR worse than hacking them to force them to do what’s right. Where’s the @gofundme campaign to donate? 3/3”

Appendix B

Initial Items and Corresponding Ethical/Moral Categorization for the Activists

Table

Item	Category
In general, do you feel Enviro-mous’ actions were right?	Moral subjectivism
(2) Do you think Enviro-mous made the company clean up in the right way?	Moral subjectivism
(3) How much do you feel Enviro-mous’ actions were irresponsible? (R)	Moral subjectivism
(4) Do you think Enviro-mous’ actions were right for Ireland?	Cultural relativism
(5) Do you think Enviro-mous’ actions helped protect Irish waters and environment?	Cultural relativism
(6) Do you feel Irish people would support Enviro-mous and their actions?	Cultural relativism
(7) Do you think Irish people would reject Enviro-mous’ hacking? (R)	Cultural relativism
(8) Do you trust Enviro-mous to use the money to clean up the oceans?	Ethical trust
(9) Do you trust Enviro-mous to use the excess £10 million to clean up other environmental issues?	Ethical trust
(10) Do you think Enviro-mous will hack for other purposes in the future? (R)	Ethical trust
(11) Do you feel Enviro-mous has ulterior motives than making the government pay for the cleanup? (R)	Ethical trust
(12) Do you feel Enviro-mous members are good people?	Virtue morality
(13) How much do you think Enviro-mous members would be considered “good people” by others?	Virtue morality
(14) Do you think Enviro-mous members have the qualities of virtuous people?	Virtue morality
(15) How much do you feel EM members are dishonest? (R)	Virtue morality

Items 1–2, 4–5, 8–9, 12, and 14 were on the Altruism component. Items 6, 7, and 13 were on the Perceived subjective norms component. Items 10–11 were on the Questionable motivations component.

Appendix C

Opinion Items for the Spokesperson

Table

How much do you think you would like this person?

(2) How likely would you be willing to follow this person on social media?

(3) How much do you agree with this person?

(4) How much do you think you would do something similar to this person?

(5) How likely would you be to like or share this person’s social media post?

(6) How intelligent do you think this person is?

(7) How well-informed do you think this person is?

(8) Do you feel this person has been educated?

(9) How much do you respect this person?

(10) How much would you criticize this person’s social media post? (R)

(11) How much do you respect this person’s opinion?

Items 6–11 were on the Virtue component. Items 1–2 and 4–5 were on theSocial connectivity component.

Appendix D

Table

Opinion Items for the Commentator
How much do you think you would like this person?
(2) How likely would you be willing to follow this person on social media?
(3) How much do you agree with this person?
(4) How much do you think you would do something similar to this person?
(5) How likely would you be to like or share this person’s social media post?
(6) How intelligent do you think this person is?
(7) How well-informed do you think this person is?
(8) Do you feel this person has been educated?
(9) How much do you respect this person?
(10) How much would you criticize this person’s social media post? (R)
(11) How much do you respect this person’s opinion?

Items 1, 3, and 6–11 were on the Virtue component. Items 2 and 4–5 were on the Intentional behavioral component.