Journal of Sustainable Marketing

ISSN: 2766-0117

Journal Insights | Publishing Model: Open Access | APC: Waived by the Publisher

Editor-in-Chief View Editorial Board

Dana L. Alden

1141Total Views

Research Article

CSR Fraud and Consumer Opinions: What Simulated “tweets” Can Explain about Consumer Beliefs in CSR

Table Of Contents

Abstract

This research examines what companies can learn from consumer statements made about them over social media, specifically, how the valence (positive or negative statements) and tweet character (emotional or rational statements) of consumer statements coupled with the type of CSR/CSR related fraud impacts consumers’ beliefs about firm CSR use as well as firm confidence. This was conducted via an experimental design manipulating the type of CSR used and related fraud, and then participants responded to the events via simulated tweets and responded to questions about their beliefs in firms engaging in CSR as well as their overall confidence in a firm. Through applying how individuals assess moral judgments via ethical philosophies (idealism vs. relativism) and how those apply to the types of “tweets” made (emotional vs. rational), it was found that the valence and character of a tweet coupled with the CSR type and the related CSR fraud can be used to understand CSR beliefs and confidence in the firm. Additionally, it was found that tweet valence (positive or negative) moderated by the tweet character (emotional vs. rational) impacts the beliefs that firms should engage in CSR. In certain instances, some consumers will not like that a firm engages in CSR, therefore, after a fraud and related negative buzz, it is recommended that firms withhold broadcasting their CSR activities.

The good and the wise lead quiet lives — Euripides

Introduction

Extant research has shown that Corporate Social Responsibility (CSR) has served as a resource for firms (Fauzi & Idris, 2009), which has largely resulted in corporate competitive advantages (Bendisch, Larsen, & Trueman, 2013). However, something that has been intended to be a resource for firms may sometimes not work as intended (Carvalho & Fernandes, 2018). Typical CSR-related fraud involves the misuse, misgoverning, and/or misdirection of funds (Rai, 2020). For example, some companies based out of India, in an attempt to demonstrate that they are engaging in CSR, “donate” to charitable trusts and then have that money discretely directed back into the company through cash transactions (Rai, 2020). Then there are instances of organizations that espouse CSR beliefs while the officials and executives representing these organizations accept and even seek kickbacks and bribes, such as the FIFA corruption scandal (BBC News, 2015).

Firm activities can go viral quickly, especially if they are remotely close to scandalous behaviors (Jansen, Hinz, Deusser, & Strufe, 2021; Zhang & Huang, 2018), and consumers almost instantaneously react via social media based on what organizations are doing, whether it is good or bad (Brown, Broderick, & Lee, 2007; Gomez-Carrasco & Michelon, 2017; Huang, Clarke, Heldsinger, & Tian, 2019). Additionally, Corporate Social Responsibility (CSR) initiatives are often announced via social media (Huang et al., 2019; Kesavan, Bernacchi, & Mascarenhas, 2013); however, CSR activities do not always go according to plan, and positive things a firm does may “backfire” and could result in fraud or wrongdoing if the activities are mismanaged (Huang, 2022).

Additionally, various attitudes toward the use of CSR at firms can impact consumers’ views of that firm, sometimes positively and sometimes negatively (Huang, 2022). However, CSR activities have generally positive implications for firms. While there has been a good deal of recent research surrounding CSR activities from many different angles, this research has primarily looked at broad issues, such as CSR impacting brand reputation, impact on stakeholders, and the role of CSR with corporate governance (Ferri, Pedrini, & Minciullo, 2022; Geng, Cui, Nazir, & An, 2022; Pfajfar, Shoham, Małecka, & Zalaznik, 2022). There is a growing stream of research that examines consumer perceptions of CSR (Harrison & Huang, 2022; Huang, 2022; Lim, 2020; Peasley, Woodroof, & Coleman, 2021) since understanding the impact of CSR on the consumer is integral in the activities leading to satisfaction and long-term advantages (Fatma, Khan, & Rahman, 2018). Thus, there is value in further studying how consumers perceive CSR and their attitudes towards firms that use CSR and subsequently communicate their CSR activities.

 Recent research on CSR has demonstrated that perceptions of CSR impact brand reputation, product innovation, and consumer trust (Geng et al., 2022). Additionally, CSR can be utilized as a marketing and branding tool, positively impacting consumer beliefs and attitudes (Doorn, Onrust, Verhoef, & Bügel, 2017; Huang, 2022). Other benefits from CSR are customer satisfaction, improved product quality, enhanced reputation, and improved company performance (Fatma et al., 2018; González-Rodríguez, Martín-Samper, Köseoglu, & Okumus, 2019; Goyal & Chanda, 2017; Latif et al., 2020). Therefore, it is very well known that CSR can lead to firm competitive advantages (Huang, 2022). However, what happens when there are problems with implementing CSR or there is CSR-related fraud? Little research has addressed this yet. While there has been a great deal of research on greenwashing activities, which is the inauthentic use of CSR for the sole purposes of improving brand image and profit, there has been a small amount of research on CSR and CSR-related fraud with many gaps left unexplained (Huang, 2022; Kurpierz & Smith, 2020). The research in this area has looked at the relationship between CSR and fraud, where generally it has been found that firms engaging in CSR activities experience a decrease in the likelihood and severity of corporate fraud (Harjoto, 2017; Hu, Dou, & Wang, 2019; Liao, Chen, & Zheng, 2019). Additionally,Kurpierz and Smith (2020) theoretically demonstrate that greenwashing, which is essentially a misrepresentation of CSR, is akin to committing fraud.

Additionally, while there is a stream of research that analyzes consumer tweets and other social media engagements (Araujo & Kollat, 2018; Huang et al., 2019; Okazaki, Plangger, West, & Menéndez, 2020); this current research examines statements made about a fictional firm in a controlled environment. Because of the use of experimental design and the elimination of confounds, this research is able to cleanly show how statements made about a firm can mean specific, and not always obvious, things for that firm. It is a reality that firms planned use of CSR will not always go according to plan. This research aims to delve into this gap and understand better what consumers think of CSR and CSR-related fraud and, importantly, what firms can and should do to mitigate damages. Thus, the goal is to understand ways to minimize the harm that can be done from fraudulent events that occur in conjunction with CSR activities. Thus, this research has two main contributions: 1) a deeper understanding of CSR and CSR-related fraud and 2) what firms should really do based on what and how things are being said about them.

Various elements drive consumer attitudes toward CSR, such as the type of CSR used and/or individual orientations toward CSR (Huang, 2022; Kolodinsky, Madden, Zisk, & Henkel, 2010). Information that is available to marketers, such as public statements made about a firm over social media, could be analyzed to help elucidate how consumers feel about a firm's CSR use. Thus, to test this, an experimental design of simulated tweets was created, and later analyzed for emotionality/rationality and positivity/negativity. Then participants were surveyed for beliefs in the appropriateness of CSR used as well as their attitudes towards a firm that uses CSR. The findings offer both theoretical and practical implications. Theoretically, the findings show a link between ethical ideologies and the tone and character of a statement made about a firm. For marketing managers of CSR, when there are rational tweets/statements made that are either positive or negative, they will want to downplay or not communicate CSR activities because consumers are more likely to think that firms should not be engaging in CSR. But, when there are positive tweets/statements and CSR that focuses on external stakeholders, marketing managers will want to advertise and promote CSR activities as these consumers are more likely to have positive attitudes about CSR activities at these firms. The rest of the paper is organized as follows: literature review, theoretical discussion and hypotheses development, methodology, analysis and results, discussion, and limitations and future research.

Literature Background

Corporate Social Responsibility (CSR)

Legitimacy, sustainability, and profitability are three main motives for CSR (Brønn & Vidaver-Cohen, 2009), where legitimacy is related to firm image, sustainability refers to ecological and societal well-being, and profitability deals with a firm generating sustainable financial results (Ferri et al., 2022). There are many definitions of CSR, but generally, it has been defined as “doing good for society” while considering ethical, discretionary/philanthropic, economic, and legal impacts (Schwartz & Carroll,2008, p. 156; Carroll , 1991 ;Weller ,2017). CSR activities focus on social issues, sustainability, consumer protection, corporate governance, legal, and regulatory concerns (Ferrell, Fraedrich, & Ferrell, 2017). CSR is how firms connect with customers, suppliers, retailers, and their stakeholders – it is a firm’s commitment to financial growth while also focusing on the well-being of stakeholders, which means customers, employees, managers, and their families, society, and the overall community (Abbas, Aman, Nurunnabi, & Bano, 2019; Branco & Rodrigues, 2006).Cowan and Guzman (2020) define CSR as voluntary activities that can enhance reputation and brand image for firms (Torelli, Monga, & Kaikati, 2012). Additionally, in some cases, firms may be legally required to engage in CSR activities (Rai, 2020). The Commission of the European Communities defines CSR as “a concept through which companies integrate social and environmental concerns in their commercial operations and in their interactions with their stakeholders on a voluntary basis” (Commission of the European Communities, 2006). Thus, for the purposes of this research, CSR can be defined as organizational behaviors that affect economic, legal, ethical, and/or philanthropic activities that are for the good of multiple stakeholders while creating some positive value for the organization and considering the potential environmental and societal impacts. (See Appendix A for a summary of the CSR-related literature.)

There are generally two types of CSR and CSR-related fraud: Corporate Operating Performance (COP) and Corporate Social Performance (CSP) (Huang, 2015; Huang, 2022). COP activities by firms generally assist those closest to the firm, such as stockholders, employees, and vendors, while CSP activities generally are socially oriented and are aimed at community-related issues or other activities that are not related to the firm’s core business (Huang, 2015). Recent research on CSR can be largely grouped into three categories: 1) firm outcomes, such as brand reputation and competitive advantages 2) management of CSR, such as governance and compliance, and 3) ancillary but related topic dimensions, such as CSR’s relationship with ethics, sustainability, and non-financial benefits (Abbas et al., 2019; Davidson & Griffin, 2000; Ferrell, Harrison, Ferrell, & Hair, 2019; Geng et al., 2022; Jiménez, Dittmar, & Portillo, 2021; Lin, Yang, & Liou, 2009; Pfajfar et al., 2022; Weller, 2017). Recent findings related to CSR’s impact on tangible firm outcomes find that organizations benefit when they actively communicate CSR activities and ethical compliance, as these can result in competitive advantages (Geng et al., 2022). Some scholars state that the goal of CSR is to provide benefits to stakeholders, such as employees, customers, and society, as well as improve the triple bottom line consisting of social, financial, and economic dimensions (Geng et al., 2022).

Relatedly, a positive relationship found between CSR that benefits employees and consumers’ perceptions of CSR usefulness for society, customers, and employees (Geng et al., 2022).Ferrell et al. (2019) find that while consumer perceptions of CSR are important, views of business ethics are more important than perceptions of CSR in creating brand attitudes. This suggests that a firm’s brand could take a substantial hit in goodwill and reputational equity if there is an ethical breakdown related to CSR activities. In order to better conceptualize the dimensions of CSR and ethics,Weller (2017) delved deep into the meaning and relationships managers give to ethics and compliance practices and CSR activities. CSR and business ethics have conceptually been viewed as the same thing but applied to different parts of an organization. The scholar’s research found that ethics influenced the meaning of compliance and CSR work, but ethics is not a distinct set of business practices.Weller (2017) also found informal linkages between compliance and CSR structures. Managers believe that ethics and compliance programs are really more about compliance with rules and regulations, while CSR programs are more about ethics or responsibilities. Largely, these findings state that CSR is a part of business ethics, and ethics and CSR are not interchangeable concepts (Davidson & Griffin, 2000; Joyner & Payne, 2002). Ethics is defined along a continuum from the lack of misconduct to actual positive impact, while CSR activities range from reducing harm to stakeholders to doing positive things for society.

Motivated by the Covid-19 pandemic,He and Harris (2020) suggest that firms should shift towards more genuine and authentic CSR by addressing currently urgent social and environmental challenges. Looking into the dialogue about CSR, (Ferri et al., 2022) examined whether firm motives for CSR affect dialogue with stakeholders when there are institutional voids (lacking governance mechanisms). Meseguer-Sánchez, Gálvez-Sánchez, López-Martínez, and Molina-Moreno (2021) examine the relationship between sustainability and CSR. They state that CSR picks up where the law ends, and while it can be focused on generating a positive public image and/or profits, it is for economic, political, social, and environmental well-being while considering current and future needs. There have been inconclusive results surrounding the relationships between CSR, sustainability, and economic performance; however, a strong positive relationship has indeed been found between CSR and financial performance (Lin et al., 2009).

CSR Fraud

Since fraud is defined as “the deliberate actions taken by management at any level to deceive, con, swindle, or cheat investors or other key stakeholders” Zahra, Priem, & Rasheed,2005, p. 804), CSR-related fraud is the willful mismanagement of the CSR related activities, which often results in either harm to those the firm intended to benefit or the firm defaulting on pledged beneficial actions (Beasley, 1996; Huang, 2022). Just as the old adage states, “no good deed goes unpunished,” oftentimes things that are meant to be positive and beneficial for others will backfire due to either carelessness, mismanagement, or greed, such as how the well-intended TOMS shoes enterprise went out of business, likely from donation mismanagement (Hessekiel, 2021). TOMS shoes enacted the buy-one-give-one business model – one pair of shoes bought equated another pair of shoes given away in a developing country; however, it’s been suggested that TOMS disrupted the shoe industry in targeted nations and distribution partners had to comply with other programs as required by TOMS, often to the local organizations’ detriment. Ultimately, the company failed, and it was uncertain whether the program ever had a positive impact on the economies and people they targeted with donated shoes (Hessekiel, 2021).

There are many instances of firms that engage in mismanagement of CSR that leads to wrongdoing or outright fraud, which consumers can also perceive as greenwashing behaviors or inauthentic CSR activities (Siano, Vollero, Conte, & Amabile, 2017). Greenwashing behaviors are seen as symbolic actions, not functional ones, taken in regard to CSR activities and can even be seen as a type of deceptive manipulation (Siano et al., 2017). Inauthentic CSR is similar in that firms' intended communicated CSR actions do not match what they actually do; where individuals perceive a type of inconsistency or contradiction resulting in an inauthenticity aversion or moral indignation (Silver, Newman, & Small, 2021). While related to greenwashing, but not quite the same, CSR-related fraud is mostly mismanagement of CSR activities – the CSR is meant to garner goodwill and positively benefit others, but when it is mismanaged or goes wrong, it can harm the intended beneficiaries (Huang, 2022). Some scholars call this deceptive manipulation, where a firm deliberately misrepresents corporate sustainability practices (Siano et al., 2017) , and consumers can become even more skeptical of CSR communication as they take it as a sign that something disingenuous is occurring (Prasad & Holzinger, 2013). CSR-related fraud occurs when firms state they are engaging in either COP or CSP activities and then misuse those funds resulting in negligence or even harm to the intended recipients. When there is COP fraud, those who were the intended beneficiaries closer to the firm, such as suppliers or employees, are not helped in a best-case scenario and can even be harmed in the worst case — this ultimately harms the firm though. Relatedly with CSP fraud, the intended beneficiaries in the community are left without their promised “benefits” at best and are harmed at worst — the firm is not necessarily impacted by this. Thus, there can be a stronger backlash or dislike of a firm when fraud or wrongdoing is connected to firms’ specific CSR behaviors.

Social Media Communications

It is important for brands to consider their online presence across social media platforms when generating content, as perceptions of brand authenticity in online engagements impacts consumer evaluation of the brand (Eigenraam, Eelen, & Verlegh, 2021). Consumers react to positive and negative firm behaviors through social media posts such as Twitter (Araujo & Kollat, 2018; Brown et al., 2007; Huang et al., 2019; Okazaki et al., 2020; Silva, Hruschka, & Jr., 2014). Consumer sentiments that are posted on social media are often coded by researchers for valence (positive vs. negative) and character (emotional vs. rational) (Coletta, Silva, Hruschka, & Hruschka, 2014; Dwyer, 2007; Huang, 2010; Huang et al., 2019). Importantly, individuals who organically spread Word of Mouth (WOM) independently of the business they are sharing information about, are often considered more trustworthy (Arndt, 1967; Brown et al., 2007; Huang, 2010; Schiffman & Kanuk, 1995). Thus, their opinions are worth more – so the information they post is valuable for businesses in understanding consumer sentiment (Barlas & Huang, 2009; Huang & Barlas, 2009). Understanding what consumers really mean in their online content can be incredibly enlightening for firms (Dwyer, 2007). Even now, firms are unsure if digitally engaged consumers will increase purchase behaviors (Halloran & Lutz, 2021), but firms do see the importance of using analytical tools on consumer WOM (Huang, 2010; Verma & Yadav, 2021).

Sentiment analysis is the methodology often used to classify tweets (Huang et al., 2019) in order to determine consumer opinions, emotions, and attitudes (Coletta et al., 2014; Huang, 2010; Li, Shah, Fang, Nourbakhsh, & Liu, 2016; Liu et al., 2020). Sentiments consist of the target, like a company’s activities, and their opinion about it, such as positive, negative, or neutral sentiment orientations (Liu, 2012). Additionally, a statement can be objective if it is rational and fact-based or subjective with emotional wording based on feelings, views, or beliefs (Chaudhuri, 2006; Huang, 2010; Liu, 2012). Thus, the two-way communication and opinions expressed by consumers via social media can be coded based on the valence of the sentiment and the tone or type of message used (rational vs. emotional). Knowing this can help businesses understand how consumers think and feel about the firm.

Theoretical Framework and Hypotheses Development

As previously mentioned, tweets and other consumer social media posts are often analyzed based on the valence (positive or negative) and the character (rational or emotional) that posters use (Chaudhuri, 2006). A factual character is rational and objective, while an emotional tone of voice focuses on the feelings generated from an experience and is generally subjective and abstract (Keyzer, Dens, & Pelsmacker, 2017). In order to explain why individuals make rational or emotional statements based on ethical firm behavior, ethical ideological theories were referenced. The ethical ideologies of idealism and realism have historically formed an axis of beliefs and are tangentially related (Crawford, 2005). Realists believe in rational actors, while Idealists desire to overcome perceived injustices or to “do the right thing” (Crawford, 2005). Additionally, ethics scholars often classify decisions based on deontological or teleological evaluations (Friesdorf, Conway, & Gawronski, 2015), where a deontological perspective is more rules and process-based, and a teleological perspective is focused on outcomes and consequences of actions (Mcmahon & Harvey, 2007). (See Appendix B on teleological vs. deontological ethical perspectives.)

Kolodinsky et al. (2010)’s research on ethical ideologies and CSR attitudes adopted two distinct views of ethical ideology of idealism vs. relativism to understand individual orientations towards CSR usage in firms. Trautwein and Lindenmeier (2019)’s research considered whether there was a response to Corporate Social Irresponsibility through affective (direct) and cognitive channels (indirect with ethical judgment as a mediator) contingent on the degree that a consumer was ethically conscious in product preferences. Ethical consciousness in product preferences is the degree that consumers consider ethical factors when selecting products or services (Trautwein & Lindenmeier, 2019). Generally, cognitive processes are used for those with higher levels of ethical conscious consumption, and affective routes are used for those with lower levels of ethical preferences for the products bought and consumed (Trautwein & Lindenmeier, 2019). Moral equity and contractualistic judgments are related to those with higher ethical product preferences, and utilitarian judgments are related to those with lower ethical product preferences (Cohen, Pant, & Sharp, 1993; Reidenbach & Robin, 1990). Thus, higher ethical consciousness about products is equated with cognitive processes and a moral lens that implies that the individual follows rules, which is akin to a deontological ethical belief system (Mcmahon & Harvey, 2007). Lower ethical consciousness towards products is equated with emotions and utilitarian/teleological judgment, which implies a belief in the greatest good for the greatest number (Reidenbach & Robin, 1990). Thus, because of the connections found by Trautwein and Lindenmeier (2019), deontological ethical beliefs are more likely to be associated with cognitive processes, and teleological ethical beliefs are more likely to be associated with emotions.

Per Trautwein and Lindenmeier (2019), some consumers are driven by more negative emotions when hearing messages about corporate misconduct (lower preferences for ethical products). Because of this negative emotional reaction to firm outcomes, these consumers are more teleological, where actions taken are based on an analysis of consequences (Reidenbach & Robin, 1990). Relativism is the degree that each specific situation is taken into account when making a moral decision, that is, there are no universal rules or codes of conduct, and evaluations of ethics are context-specific (Forsyth, 1992; Mcmahon & Harvey, 2007). Higher levels of ethical sensitivity equate to higher levels of relativism – that is, a relativist may think ethical issues are less important and are more concerned about the ends justifying the means (Sparks & Hunt, 1998). Relativists’ beliefs appear to be utilitarian, which promotes the greatest good for the greatest number, which is a teleological or consequentialist ethical viewpoint (Forsyth, 1992). Consequently, moral relativism is indeed moral realism (Harman, 2014), and relativists’ beliefs align most closely to a teleological ethical lens (Cohen et al., 1993; Mcmahon & Harvey, 2007; Trautwein & Lindenmeier, 2019). Therefore, because of the connections found by Trautwein and Lindenmeier (2019), it is possible that those who hold relativists’ beliefs are likely to respond to ethical violations with emotions and are more likely to see things through a teleological lens.

Additionally, realism beliefs are generally congruent with the use of COP activities by firms, which has previously been associated with realistic belief systems (Huang, 2022; Kolodinsky et al., 2010). Thus, consumers who process ethically-based messages with more emotions are more likely to be realistic/relativist as they use a utilitarian lens to view issues (the opposite of idealism) (Forsyth, 1992; Sparks & Hunt, 1998; Trautwein & Lindenmeier, 2019).

On the other hand, consumers who have higher levels of ethical consciousness in product preferences are more likely to use their cognitive channels for evaluations (Trautwein & Lindenmeier, 2019). It was also found that these consumers put less emphasis on utilitarian judgments, which is consistent with idealism (Forsyth, 1980), where it is seen that a good outcome does not justify unethical means. Idealism is defined byForsyth (1992) as the degree that someone has genuine concern for others and takes action to avoid harm. Idealists assume positive outcomes are reached through moral actions (Henle, Giacalone, & Jurkiewicz, 2005). That is, a positive input should result in a positive outcome. In this way, an individual rationalizes what “should” be.Kolodinsky et al. (2010) found that more idealistic individuals have a more positive attitude toward firms’ CSR use, while those who have higher levels of ethical relativism hold negative attitudes toward CSR. Additionally, idealism beliefs are generally more congruent with CSP firm activities (Huang, 2022). SinceTrautwein and Lindenmeier (2019) found that cognitive processes and moral-equity/contracturalism (deontological) judgments are related and that these consumers are more likely to be idealistic, it is likely that there is a relationship between cognitive responses and an idealistic belief system. Therefore, those who are more idealistic are also deontological and are more likely to process messages about ethics cognitively.

Ethical idealists are more likely to hold positive attitudes toward firms using CSR, while ethical realists hold the opposite view (Kolodinsky et al., 2010). CSP is designed to help those outside of a firm’s direct business channel and is more value-based and idealistically driven. COP is for the benefit of those inside the firm and is more realistically based and economically driven (Huang, 2015; Huang, 2022; Maignan, 2001). Individuals who are more likely to criticize a firm after a fraud or scandal are more likely to be idealistic, while those who seem to think a firm can recover after a wrongdoing or fraud are more realistically driven (Forsyth, 1992; Kolodinsky et al., 2010). Thus, emotional tweets may infer that participants have less of a preference for CSR but are more realistic and, thus, are more likely to “recover” from frauds or scandals related to CSR. Cognitive tweets will have mixed results based on someone’s deeper moral lens – but more likely, they are idealistic due to the connections found by Trautwein and Lindenmeier (2019). Therefore, it is theorized that emotional responses are associated with a relativist ethical lens, while cognitive responses are associated with an idealistic ethical lens (Forsyth, 1992; Kolodinsky et al., 2010; Trautwein & Lindenmeier, 2019). Additionally, realism is associated with COP firm behaviors, and idealism is associated with CSP behaviors (Huang, 2022). Therefore, how individuals perceive firm CSR behaviors and CSR fraud will be impacted by their ethical lens (relativist or idealist), where emotional reactions are relativist and cognitive reactions are idealistic.

CSR Beliefs

CSR beliefs are how consumers feel about the use of CSR activities at a corporation and whether they believe the firm should be engaging in those activities at all (Sen & Bhattacharya, 2001). Some consumers believe that socially responsible behavior can detract a firm from its core mission, while some consumers feel that a firm can effectively engage in socially responsible behavior while creating strong offerings and value. If a consumer believes that a firm should not be engaging in CSR, they will not support a firm doing so, which can lead to decreased purchase intentions (Sen & Bhattacharya, 2001). Analyzing what consumers are saying about the firm can help the firm understand how a consumer thinks and feels about them – some firms are conducting content analyses of tweets, which can elucidate consumer sentiment and evaluations (Chaudhuri, 2006; Gomez-Carrasco & Michelon, 2017). The sentiment expressed is coded as either positive or negative, and the content or character of the tweet is understood as rational or emotional (Chaudhuri, 2006; Huang et al., 2019). Rational evaluations are tangible beliefs and utilitarian attitudes, while emotional evaluations come from non-tangible and affective reactions. Thus, the valence and the character of the tweet are interrelated, and understanding what consumers are saying will be influential in knowing their beliefs about CSR.

Additionally, per the taxonomy provided byCarroll (1979); Carroll (1991) and the findings by Maignan (2001), some types of socially responsible behavior are expected by businesses before others. Thus, firms that are not doing the “bare minimum,” such as not abiding by laws or ethics, will be judged harshly, and it is likely that consumers will believe that they should not be engaging in CSR. This conclusion is likely to be reached via a rational view or through a cognitive appraisal (Lazarus & Folkman, 1984). Therefore, it is also likely that the more rational or utilitarian tweets (Chaudhuri, 2006) will be significant and have stronger effects on this relationship due to the congruence between rational statements and cognitive appraisal (Nadler & Tushman, 1980). Thus, it is likely that the valence (positive or negative) and the character (emotional vs. rational) will interact with each other (Chaudhuri, 2006; Huang et al., 2019) and will do so in a way that reflects beliefs on whether a firm should be engaging in CSR or not (Carroll, 1979; Carroll, 1991; Maignan, 2001). See Figure 1 below for the conceptual model. The following hypothesis is thus offered:

H1: The relationship between the valence of the tweet on CSR and the strength of CSR beliefs is moderated by the character of the tweet.

CSR Type

As previously discussed,Huang (2015) categorizes CSR activities as either Corporate Operating Performance (COP) dominated or Corporate Social Performance (CSP) oriented. The outcomes of COP primarily benefit stakeholders closest to the firm, such as employees or vendors; CSP primarily benefits stakeholders farther away from the firm, such as communities or higher-level societal causes. Because of these differences in CSR types and the associated consumer thoughts about these (Huang, 2022), it is likely that CSR type will moderate the relationship between the valence of the tweets and the CSR beliefs. Additionally, the different types of CSR can lead to various conceptualizations in consumers’ thoughts via a categorization process (Fiske & Pavelchak, 1986). Since CSP primarily benefits those outside a firm, individuals may then categorize this behavior as sincere, altruistic, or philanthropic (Carroll, 1979; Carroll, 1991) and then transfer those beliefs to the firm’s values. COP behaviors generally are for the good of the firm, which may lead consumers to categorize these behaviors as economically motivated (Carroll, 1979; Carroll, 1991) and thus project those beliefs onto the firm. As previously mentioned in Carroll’s taxonomy, economic behaviors should be the bare minimum that a firm engages in before they start with philanthropic behaviors. Thus, it is expected that consumers would strongly believe that a firm should not be engaging in CSR when they cannot effectively engage in COP behaviors. Therefore, the COP should be a stronger moderator of negative CSR beliefs than CSP, but overall, the CSR type is expected to moderate the positivity or negativity of the tweet on CSR beliefs. The following hypothesis is offered:

H2: The relationship between the valence of the tweet on CSR and the strength of CSR beliefs is moderated by the type of CSR activities.

https://typeset-prod-media-server.s3.amazonaws.com/article_uploads/ae88e1d3-008b-46af-a8d7-736ac5906571/image/88a7fe77-ca9e-48b7-8978-efb346d63e68-uscreenshot-2023-04-26-020146.png

Firm Confidence

Organizational confidence involves seeing value in an organization’s products and services, holding optimism toward the future, having a sense of purpose, and fostering a belief that a firm can take on challenges (Liozu & Hinterhuber, 2013). Thus, organizational confidence is a key competency that drives firm performance (Liozu & Hinterhuber, 2013). Additionally, social cognitions can be formed about a firm’s capabilities (Bandura, 2000; Bohn, 2002), and confidence is defined as positive expectations for desired outcomes and results (Hoover & Valenti, 2005). Therefore, organizational confidence is a firm’s belief that it can accomplish a mission or a purpose. This is measured from a consumer’s perspective about their belief in a firm’s positive ability to meet their sales goals. Thus, firm confidence is a consumer’s confidence in the firm based on a belief in a firm’s ability to continue to meet its sales goals. Because the valence and character of a tweet can be so influential in understanding a consumer’s opinion about a firm (Chaudhuri, 2006), it is likely that tweet character will moderate the relationship between tweet valence and confidence in a firm (Chaudhuri, 2006; Huang et al., 2019). Since there is an order of activities that consumers expect firms to engage in before they become philanthropic, it is also expected that confidence in a firm to meet economic goals will be impacted by how a consumer sees and interprets fraud-related CSR (Carroll, 1979; Carroll, 1991; Maignan, 2001). Just as how orientations towards CSR will vary based on various beliefs in CSR’s role in business (Maignan, 2001), confidence in a firm is reflective of how firm CSR activities and related frauds are interpreted by consumers. Additionally, the positive or negative interpretations of the events should interact with the character of the tweets (Chaudhuri, 2006; Huang et al., 2019). Therefore, the positivity and negativity of the tweets, coupled with their character (emotional or rational), will reflect the confidence consumers see in a firm. The following hypothesis is thus offered:

H3: The relationship between the valence of the tweet on CSR and the confidence in a firm is moderated by the character of the tweet.

CSR Fraud

Just as there are different CSR types, there are different CSR frauds. Additionally, because of the cognitive-appraisal process of the evaluation of the firm’s behavior, it is likely that rational tweets will have a stronger moderating effect size on firm confidence (Lazarus & Folkman, 1984). As discussed in H2, COP primarily benefits stakeholders close to the firm, and CSP benefits those further away, such as those in the community or large collective causes (Huang, 2015). Therefore, the fraud related to the specific CSR types will be specific to the type of CSR the firm was attempting to implement. COP fraud is related to CSR that would have helped those closest to the firm, such as a CFO stealing money that would have gone for employee education. CSP fraud is related to CSR that would have helped those in the community, such as a bank CFO stealing money from a fund that would have gone to a community clean water initiative. Just as how the types of CSR were categorized by consumers as values to the firm, these types of frauds will also be projected onto the firm through an inferred congruence of values via cognitive appraisals (Lazarus & Folkman, 1984).

Frauds that have been made public are generally known to negatively impact consumer confidence in firms, shown through stock price drops or overall negative attitudes to the firm (Thiesen, 2020; Wolfe & Hermanson, 2004). Because of these differences in CSR fraud types and the associated consumer thoughts about these, it is likely that CSR fraud type will moderate the relationship between the valence of the tweets and confidence in the firm. Since CSP primarily benefits those outside a firm, fraud associated with this can be construed as neglecting or harming those who would have been helped. Individuals may then categorize this type of CSR fraud as particularly negative since that CSR type is generally seen as sincere, altruistic, or philanthropic (Carroll, 1979; Carroll, 1991). There could be a type of backlash effect for this type of CSR fraud, such as how firms are punished worse for being seen as “good” at first, and then when negative news comes out about them, are thought of as worse than if they had never appeared “good” to begin with (Becchetti, Ciciretti, & Hasan, 2009). On the other hand, individuals could see CSP fraud as the firm attempting to do good for others outside of their network yet struggling with it. This could lead to the individual holding some understanding, compassion, or hope that the firm will do better in the future. Additionally, COP behaviors generally are for the good of the firm, and thus COP fraud may lead individuals to think that the firm has only harmed themselves economically. Therefore, COP-type fraud may not as negatively impact firm confidence as CSP-type fraud. Thus, CSR fraud type will moderate the relationship between the valence of the tweet and confidence in a firm, with CSP fraud worse than COP fraud. The following hypothesis is offered:

H4: The relationship between the valence of the tweet on CSR and the confidence in a firm is moderated by the type of CSR frauds.

Methodology

Participants and Procedures

One thousand one hundred and fifty-four participants (46.9% males; Mage = 30.7 years, SD = 13.01) were recruited from Amazon Mechanical Turk (MTurk), which was collected between November and December 2019. An MTurk panel was created with broad parameters as a cross-section of society was desired – the only requirement was that they were above the age of 18 and have interacted with Twitter sometime in the past and/or present. Attention checks were also utilized in the data collection to ensure that participants were thoughtfully answering the questions and prompts. Thus, the data was first cleaned and filtered based on passing the attention checks. Then, the data was cleaned based on survey response time –those whose survey duration time fell below Q1-3*IQR and above Q3+3*IQR were removed. Using theArmstrong and Overton (1977) technique, no significant differences in responses based on time “wave” data collection periods were found; thus, it is likely that this sample does not have a non-response bias or issues with time collection periods. To trigger a tweet, each participant (PP) was randomly presented with a scenario description about a fictional financial institution, ABC Firm, who promised to engage in different types of CSR activities (COP vs. CSP) (Harrison & Huang, 2022). The scenarios are below:

COP - ABC firm has stated that they will match donations at 150% given to a non-profit organization that is committed to providing small business entrepreneurs with loans at a fair interest rate that is typically lower than most APRs (annual percentage rate) from the financial institutions in the market. 

CSP - The bank, ABC Firm, pledges that it will donate money to the community in the form of scholarships for people to go back to school and complete their degree or obtain advanced degrees so that they can enhance their competency in the job market.

After that, the PP was asked to write a simulated tweet about their opinions on the ABC Firm’s CSR activities, followed by rating their overall CSR beliefs. Then, the PP read the second part of the scenario description where ABC Firm turned out failing to keep their word, which led to CSR fraud (Huang, 2022), followed by writing a simulated tweet about their opinions on the ABC Firm’s wrongdoing (or fraud). Finally, the PP was asked to assign a numeric point (between 0 and 100) to indicate their confidence of ABC Firm in the market. The fraud scenarios are below.

COP Fraud - Recently, a news source that you regularly read and trust reported that ABC firm was not at all matching the promised donation amounts the non-profit.

CSP Fraud - Recently, a news source that you regularly read and trust came out with a story about ABC Firm. Fraud has been discovered - the firm was only giving the scholarship money to the friends and family of the top executives.

Coding

Two research assistants who were unfamiliar with the purposes of the study and blind to its hypotheses coded: (1) the character of tweets (emotional vs. rational); (2) the valence of the information in each tweet (positive vs. negative); and (3) other by following the procedures as outlined by Huang (2010). The coders agreed on 1,745 of the 2,308 Tweets for an inter-rater agreement of .756. The overall inter-rater reliability Cohen’s Kappa was 0.674 with a 95% CI [0.671, 0.702]. Reliability was calculated without resolving disagreements with coders (Weber, 1990). The agreed-upon codes were used in the subsequent content analyses.

A PP was asked, “If you were to write a tweet about ABC's CSR activities that you just read, what would you say?” The following tweets are examples coded as positive rational:

So glad to see businesses make decisions that can benefit others as well as themselves... #nocoincidence

Great job ABC to promote clean water for their community.

[I think] CSR is great because it improves both professional and personal development.

Get some money; increase your education; make more money…

A PP was asked “If you were to write a tweet about the ABC firm's wrongdoing (or fraud), what would you say?” The following tweets are examples coded as negative emotional:

This is where customers money is going?! This is awful! I don't blame employees for quitting after this.

Take the shame, exploit the blame!

Hate what you guys did to us This isn't right!

It is a shame that a corporation that prides its self with treating people with respect is committed of fraud, I am highly disappointed.

Analysis and Results

Validity and Reliability

A first regression was used to analyze the relationships between the dependent variable, CSR beliefs, and the independent variable, the tweet valence of CSR activities (0 = negative, 1 = positive). The tweet character (0 = rational, 1 = emotional) and the types of CSR activities (0 = CSP, 1 = COP) are treated as moderators, while age and gender are covariates. All the items of CSR beliefs are measured by a seven-point Likert scale with anchors of 1 = strongly disagree and 7 = strongly agree (Maignan, 2001). The correlation matrix presented in Table 1 (n = 1,154) suggests that most correlations, except for Item 7, are found in excess of .30. Additionally, none of the correlation coefficients is larger than .80, which reveals no evidence of multicollinearity (Field, 2019).

Table 1: Correlations among the CSR Belief Measures

1

2

3

4

5

6

7

8

9

1. Socially responsible behavior detracts from ABC Firm’s abilities to provide the best possible products

1

2. Socially responsible behavior is a drain on ABC Firm’s resources

.560**

1

3. Socially responsible behavior by firms such as ABC is often a cover up for inferior product offerings or inferior service

.421**

.390**

1

4. Socially responsible firms like ABC Firm produce worse products than do firms that do not worry about social responsibility

.429**

.444**

.486**

1

5. All else equal, a socially responsible firm like ABC Firm is likely to have lower technological expertise than a firm that is not socially responsible

.400**

.425**

.406**

.633**

1

6. Firms like ABC Firm that devote resources towards socially responsible actions have fewer resources available for increasing employee effectiveness

.341**

.470**

.329**

.525**

.585**

1

7. A company like ABC Firm can be both socially responsible and manufacture products of high value

.062*

.171**

.063*

.230**

.244**

.184**

1

8. Firms like ABC engage in socially responsible behavior to compensate for inferior product offerings

.346**

.363**

.493**

.533**

.464**

.455**

.165**

1

9. Resources devoted to social responsibility come at the expense of improved product offerings for ABC Firm

.376**

.435**

.410**

.503**

.456**

.557**

.149**

.588**

1

**. Correlation is significant at the 0.01 level (2-tailed).

*. Correlation is significant at the 0.05 level (2-tailed).

The construct validity is assessed by conducting a principal component analysis (PCA). The exploratory factor analysis determinant is .004, the Kaiser-Meier-Olkin Measure of Sampling Adequacy (KMO MSA) is .878, and Bartlett’s χ2 (36) = 3,887.34 (p < .001). The component matrix in Table 2 suggests that Item 7 should be removed by using a loading criterion of .40 (Pituch & Stevens, 2015). The reliability Cronbach’s Alpha is increased from .852 to .868 with the removal of Item 7. PCA was used as it has been shown to be effective in a variety of contexts and disciplines as it improves performance at a ‘low cost’ of model accuracy, reduces data noise, and can produce independent, uncorrelated features of the data (Jolliffe & Cadima, 2016).

Table 2: Component Matrix of CSR Belief Measures

Items

Mean

Std. Deviation

Component

1. Socially responsible behavior detracts from ABC Firm’s abilities to provide the best possible products

3.57

1.833

0.646

2. Socially responsible behavior is a drain on ABC Firm’s resources

3.37

1.740

0.696

3. Socially responsible behavior by firms such as ABC is often a cover up for inferior product offerings or inferior service

3.95

1.595

0.661

4. Socially responsible firms like ABC Firm produce worse products than do firms that do not worry about social responsibility

3.21

1.532

0.796

5. All else equal, a socially responsible firm like ABC Firm is likely to have lower technological expertise than a firm that is not socially responsible

3.09

1.484

0.765

6. Firms like ABC Firm that devote resources towards socially responsible actions have fewer resources available for increasing employee effectiveness

3.34

1.512

0.741

7. A company like ABC Firm can be both socially responsible and manufacture products of high value

2.50

1.440

8. Firms like ABC engage in socially responsible behavior to compensate for inferior product offerings

3.66

1.563

0.733

9. Resources devoted to social responsibility come at the expense of improved product offerings for ABC Firm

3.70

1.531

0.747

Moderations on CSR Beliefs

As H1 predicted, the relationship between the valence of the tweet on CSR activities and the strength of CSR beliefs is moderated by the tweet character (ß = .29, p = .02, refer to Table 3).

The further simple slope analyses in Figure 2 suggests that negative rational tweets have a stronger impact on the CSR beliefs than the positive rational tweets (B = -.77, t (1,148) = -2.79, p = .005), although the emotional tweets have no statistically significant influences (B = -0.01, t (1,148) = -0.09, p > .5). However, the moderation of the CSR activities’ tweet character is not significant (ß = .08, p = .48, refer to Table 3 ). Thus, H2 is not supported (Figure 3). These results combined suggest that those who post negative rational tweets do not much like CSR use in general, while those who post positive rational tweets do like businesses using CSR. Rational tweets are more evaluative and demonstrate a belief that CSR is consequence focused and impacts society (Rawwas, 2001).

Table 3: The Moderations on CSR Beliefs

DV = CSR Beliefs

Predictors

ß

t

p

Intercept

15.12

<.001

Tweet Valence

-0.19

-3.07

0.002

CSR Type

-0.01

-0.10

0.994

Tweet Character

-0.18

-1.57

0.116

Tweet Valence × CSR Type

0.08

0.71

0.480

Tweet Valence × Character

0.29

2.33

0.020

Model Summary

R2 = 0.018

F (5, 1146) = 4.23

0.001

https://typeset-prod-media-server.s3.amazonaws.com/article_uploads/7b0b30e4-4780-4d84-b3b6-94a25cea5ec4/image/f7d371cf-4c9e-4516-81cd-3db1387da8f4-ufigure-2a.png
Figure 2: The Moderation of CSR Tweet Character on CSR Beliefs
https://typeset-prod-media-server.s3.amazonaws.com/article_uploads/7b0b30e4-4780-4d84-b3b6-94a25cea5ec4/image/adb3d3d5-4c8a-450a-a577-5a20e68807ef-ufigure-2b.png
Figure 3: The Moderation of CSR Activity Types on CSR Beliefs

Moderations on Firm Confidence

A second regression is used to analyze the relationships between the dependent variable, the market confidence, the independent variable, and the tweet valence of CSR frauds (0 = negative, 1 = positive). The tweet character and the types of CSR frauds (0 = CSP, 1 = COP) are treated as the moderators, while age and gender are covariates. The market confidence is measured by a numeric number between 0 and 100 to the question, “After reading the scenario and knowing what you have learned about ABC, how confident are you in ABC Firm to keep up its annual sales goal this year?” This question was used to reflect the consumer’s confidence in ABC Firm after its CSR frauds were released.

As H3 predicted, the relationship between the valence of the tweet on CSR and the market confidence is moderated by the tweet character (ß = -.09, p = .015, refer to Table 4). Positive emotional tweets suggests that can individual sees how CSR can impact them at almost a personal level (Macdonald, 2001), which means that the individual does have confidence in the firm. The further simple slope analyses (Figure 4 ) imply that the positive rational tweets have a stronger impact on the company’s market confidence than the negative rational tweets (B = 24.57, t (1,148) = 2.98, p = .003). A similar pattern is also presented on the emotional tweets (B = 6.91, t (1,148) = 2.38, p = .017). The high dislike of CSR used by those who make emotional positive tweets suggests that the theorized relationship between those with low ethical consciousness and subsequent emotional processing is correct (Trautwein & Lindenmeier, 2019).

Generally, these results suggest that rational tweets predict higher firm confidence than emotional tweets. This is relevant to rationalizing outcomes via ethical, ideological theories and the perspective that different levels of ethical consciousness can impact confidence in firms based on beliefs in CSR activities used. As theorized, those with lower levels of ethical consciousness are more likely to process firm events with emotions and are likely also relativists, while those with higher levels of ethical consciousness are more likely to process firm events with cognitions and are more likely to be idealists (Forsyth, 1980; Kolodinsky et al., 2010; Trautwein & Lindenmeier, 2019). Moreover, negative rational tweets have higher firm confidence than all the tweets. Particularly, negative rational tweets predict higher firm performance than positive rational tweets. Interestingly, positive emotional tweets predict higher firm confidence than negative emotional tweets. These individuals may be more critical and cynical because they expect the best from the firm – perhaps they are relativists and see a firm’s potential but criticize it more strongly for its current status.

Table 4: The Moderations on Market Confidence

DV = Confidence in Firm

Predictors

ß

t

p

Intercept

33.40

<.001

Tweet Valence

0.38

3.78

<.001

CSR Fraud Type

0.02

0.50

0.616

Tweet Character

0.06

2.02

0.044

Tweet Valence × CSR Fraud Type

-0.09

-2.45

0.015

Tweet Valence × Character

-0.24

-2.63

0.009

Model Summary

R2 = 0.021

F (5, 1136) = 4.83

<.001

https://typeset-prod-media-server.s3.amazonaws.com/article_uploads/ae88e1d3-008b-46af-a8d7-736ac5906571/image/02f31610-71fb-4a1b-8348-2f14af605a04-upicture3.png
Figure 4: The Moderation of CSR Tweet Character on Firm Confidence

H4 is also supported, as the relationship between the tweet valence on CSR, and the market confidence is also moderated by the type of CSR frauds (ß = -.24, p = .009, refer toTable 4 ). The further simple slope analyses (Figure 5 ) suggest that positive tweets about CSP have a stronger impact on the company’s market confidence than the negative tweets (B = 12.64, t (1,148) = 3.72, p < 0.001) although the tweets about COP have no statistically significant influences on the change of market confidence (B = 2.13, t (1,148) = .46, p > .5). With CSP related fraud, therefore, the business was not seeking to improve their bottom line in any way, since CSP is related to helping society and any type of CSP related fraud might imply that the firm just made a mistake. With COP fraud, however, the business sought to benefit before they engaged in their activities, which could be why this type of CSR could be viewed as worse or result in less positive firm confidence.

https://typeset-prod-media-server.s3.amazonaws.com/article_uploads/7b0b30e4-4780-4d84-b3b6-94a25cea5ec4/image/a8b91564-4c11-4968-b1c5-243990eccf08-ufigure-3b.png
Figure 5: The Moderation of CSR Fraud Types on Firm Confidence

In sum, both positive rational and emotional tweets predict more positive firm confidence. Positive tweets, when there is CSP related fraud, can predict more firm confidence than negative tweets with CSP fraud. But COP fraud is not significant. Thus, only CSP fraud is important. These findings suggest that perhaps CSP related fraud is not “as bad” as other frauds, perhaps because it is more “idealistically” oriented (Huang, 2022). This could mean not all fraud is harmful, especially with this type of fraud, since the firm was not seeking to improve their business before they set out to engage in this type of CSR activities.

Discussion

The main point of this research is to understand what consumers say about CSR activities and CSR fraud and what this means for their attitudes toward firms. This not only fills a gap in the literature for delving deeper into CSR and CSR fraud aside from the correlations found in past research (Harjoto, 2017; Hu et al., 2019; Liao et al., 2019), but the findings also offer very practical recommendations for firms on how to handle their communications based off what is said about them over social media. This research utilized an experimental design, which created a controllable environment and eliminated confounds, such as prior beliefs about a firm or other pre-existing factors. Because of the methodology, the researchers were able to isolate what someone really means when they make a statement about a firm based on either the positive or negative tone and the emotionality or rationality behind the statement. The findings offer both theoretical and practical implications.

Theoretical implications. This research sheds light on how consumers evaluate and think of firms engaging in CSR and firms with CSR-related frauds. This is conducted through understanding ethical ideologies as they relate to emotional and rational “processing” through writing simulated tweets. While past research has considered moral philosophies and consumers’ attitudes towards CSR (Kolodinsky et al., 2010) as well as examined discussions around CSR activities via social media (Gomez-Carrasco & Michelon, 2017; Kollat & Farache, 2017), this research adds on to these studies by analyzing tweets made about CSR activities and related CSR-fraud by coding those statements for an understanding of their views. The nature of the experimental design, while not without its limitations, created a controllable scenario where participant tweets could be collected in a simulated environment, thus eliminating numerous confounds. Thus, this research is able to pinpoint why a consumer says what they say and what it means for a firm. Theoretically, emotional statements are likened to lower levels of ethical consciousness and likely relativist ethical view (Forsyth, 1980; Trautwein & Lindenmeier, 2019). Rational statements are related to higher levels of ethical consciousness and likely an idealist ethical view (Forsyth, 1980; Kolodinsky et al., 2010; Trautwein & Lindenmeier, 2019).

Practical implications. Consumers generally have positive feelings towards CSR use, or in other words, positive CSR beliefs (Brown & Dacin, 1997; Mohr & Webb, 2005; Walker & Kent, 2009). However, past research has not explored how consumers perceive CSR and what consumers consider when evaluating CSR enough (Öberseder, Schlegelmilch, Murphy, & Gruber, 2013). This research extends existing findings around CSR-related moral transgressions by Kolodinsky et al. (2010), who identified which types of people like CSR based on their ethical ideologies. Their research was extended by delving deeper into the meaning of individual tweets or statements about firm CSR used in the context of related CSR fraud by seeking to understand how tweets are evaluated based on tone (valence and character) coupled with the CSR type, and the related CSR fraud can subsequently impact individual CSR beliefs and confidence in the firm.

This study suggests that tweet valence (positive or negative) impacts the beliefs that firms should engage in CSR. If potential consumers are tweeting very positive things about a firm’s CSR activities and then there is a subsequent related fraud that has to do with the CSR, there is a stronger belief by individuals that firms should not be engaging in CSR activities. It is also found that the valence of the tweets moderated by the tweet’s character or tone impacts the consumer’s belief in firms engaging in CSR activities. Rational and negative tweets lead to a higher belief that the firm should not engage in CSR compared to rational and positive valence tweets. This finding supports both the ranking of importance of business responsibility and even ethical philosophies and how individuals interpret and process moral transgressions. Individuals evaluated the CSR-related fraud via rational or emotional statements, where rational is associated with idealistic ethical views, and emotional is associated with realistic ethical views. They then appraised the cause as potentially the fault or negligence of the firm, then categorized those behaviors as being intrinsic to the firm, and then determined that the firm was unable to meet basic business responsibility requirements and subsequently had higher beliefs that the firm should not be engaging in CSR activities. Therefore, in some situations, especially after a fraud and related negative buzz, firms may want to back off from broadcasting their CSR activities.

While a business reality, few past studies have examined the effects of CSR-related fraud, specifically in this context where a firm commits fraud related to their promised CSR activities (Cumming, Hou, & Lee, 2016; Harrison & Huang, 2022; Huang, 2022). This research addresses this gap and finds that tweet valence increases confidence in firms while tweet valence moderated by the fraud type (CSP fraud vs. COP fraud) decreases confidence in firms. Negative rational tweets lead to less confidence in firms, while positive rational tweets lead to more confidence in firms. Negative tweets with CSP fraud lead to less confidence in firms, and positive tweets with CSP fraud lead to more confidence in firms. CSP appears to be a strong moderator, possibly because of the nature of CSP frauds since the firm was not seeking to improve their business with it in the first place – individuals here may really believe in the firm’s ability to recover. CSP is seen as the more “genuine” and “sincere” of the CSR types (Huang, 2022), and individuals may categorize firms engaging in this as more sincere than other firms. Additionally, CSP is associated with an idealistic ethical viewpoint (Huang, 2022), and anything that goes wrong with firm ethics after using this more idealistic type of CSR could result in stronger firm beliefs in either direction.

The findings further suggest that tweet valence moderated by tweet character decreases confidence in firms. Negative valence rational tweets and negative valence emotional tweets lead to lower confidence in firms compared to positive valence rational and positive emotional tweets. Additionally, rational tweets have a stronger moderating effect with negative valence leading to lower firm confidence than emotional tweets, and positive valence rational tweets with higher confidence than those emotional tweets. There is something important about the rational character of the tweets - rational evaluation uses high-involvement processes and cognitive information processing (Chaudhuri, 2006; Huang et al., 2019). Rational sentiments are also associated with high levels of ethical consciousness (Trautwein & Lindenmeier, 2019). Thus, those making rational tweets likely are more cognizant of ethical consciousness and will react more to firms’ ethical violations. Therefore, understanding the nature of the tweets will explain the confidence consumers see in the firm and their overall belief in the firm’s CSR activities.

Practically, these findings can assist firms in monitoring their presence on social media and altering their messages based on the tone of the tweets about a firm’s CSR news (Huang et al., 2019). See Table 5 and Table 6 below. These findings can help firms understand why they should be monitoring tweets and using technology to code valence/character of the words to anticipate consumer changes in sentiments/behaviors. Firms can use the understanding of consumer sentiments to change messaging and behavior based on valence and character of tweets. Additionally, if firms decide that they want to nudge tweets to be more rational, they can do so by creating engaging in dialog that is more cognitive based. Also, in some circumstances, it may be useful to stop communicating CSR information altogether, such as when consumers think firms should not be engaging in CSR due to their negative CSR beliefs. Specifically, when consumers are tweeting rational sentiments, firms should know that these consumers will not want to see CSR information as they believe that the firm should not be engaging in any CSR activities. Additionally, these consumers may be idealistic and react more strongly when the firm behavior does not match what they think it should be. The type of CSR fraud matters in how consumers see the firm – if a firm has a CSP fraud, then that means they were attempting to engage in CSP activities, even though they failed. This was a strong moderator because of the associated “genuineness” or “goodness” of the type of CSR activity. Thus, giving to help those outside the firm, as opposed to those close to the firm, may be seen as a positive. Therefore, consumers want to give these types of fraudulent activities more leeway since the firm was at least attempting to do good for those outside of their circle. Consequently, firms may decide that they want to communicate their CSP activities more so than their COP activities, as this may be a more favorable type of CSR.

Table 5: Recommended CSR Communications

Rational

Emotional

Pos. tweets

Downplay CSR

Inconclusive and needs future research. Likely downplay CSR

Neg. tweets

Do not communicate CSR

Inconclusive and needs future research. Likely do not use at all

Table 6: Recommendations Based on the Type of CSR Activities

Stakeholders Close to Firm (COP)

Stakeholders Far from Firm (CSP)

Pos. tweets

Downplay COP

Promote CSP

Neg. tweets

Needs future research

Needs future research

Conclusion

The main point of this research is to understand what consumers say about CSR activities and CSR fraud and what this means for their attitudes toward firms. This not only fills a gap in the literature for delving deeper into CSR and CSR fraud aside from the correlations found in past research (Harjoto, 2017; Hu et al., 2019; Liao et al., 2019), but the findings also offer very practical recommendations for firms on how to handle their communications based off what is said about them over social media. This research utilized an experimental design, which created a controllable environment and eliminated confounds. Because of the methodology, the researchers were able to isolate what someone really means when they make a statement about a firm based on either the positive or negative tone and the emotionality or rationality behind the statement. The findings offer both theoretical and practical implications. Through an ideological ethical lens, it was surmised that emotional statements are likened to lower levels of ethical consciousness and a relativist point of view, while rational statements are related to higher levels of ethical consciousness and an idealistic ethical point of view. Understanding someone’s ethical point of view will help firms understand what their behaviors should be after a CSR fraud. What is said about a firm (positive or negative) can help firms understand consumer beliefs about whether firms should be engaging in CSR at all. If potential consumers are tweeting very positive things about a firm’s CSR activities and then there is a subsequent related fraud that has to do with the CSR, individuals strongly believe that firms should not be engaging in CSR activities. It is also found that the valence of the tweets moderated by the tweet’s character or tone impacts the consumer’s belief in firms engaging in CSR activities. Rational and negative tweets lead to a higher belief that the firm should not engage in CSR compared to rational and positive valence tweets. Therefore, in some situations, especially after a fraud and related negative buzz, firms will not want to advertise their CSR activities.

If the CSR fraud is related to a specific type of CSR activity, Corporate Social Performance (CSP), then idealistic views are likely to be triggered, and consumers may be more likely to hold stronger beliefs in either direction. A rational negative statement will lead to stronger negative CSR beliefs compared to emotional negative statements. A positive rational statement compared to positive emotions will lead to stronger positive CSR beliefs. Thus, when idealistic ethical points of view are encountered, firms will want to pay attention to the valence of those statements more so than relativistic ethical points of view. When an idealist feels negative about CSR, firms will want to not advertise CSR activities, and when an idealistic feels positive about CSR they will want to highlight CSR activities more so than for someone who feels the same way but holds a relativistic ethical belief. Thus, the ethical belief will matter in deciding how to target and communicate CSR to consumers, with idealists needing more ‘care’. If an idealist feels negative, firms should not talk about CSR, but if they feel positive, firms should highlight CSR. Thus, it was found that firms can target their CSR activities based on the consumer’s ethical viewpoints, which can be determined based on their statements' character (rational vs. emotional). Firms can use these results to decide how to communicate with consumers based on what they are saying over social media about them.

Limitations and Future Research

There are numerous avenues for future research based on the findings from this study. For instance, it could be useful for businesses to understand when consumer CSR beliefs “reset,” such as when the Twitter data and associated thoughts and emotions dissipate. Also, what else can these tweets (valence and character) mean, and what else can they predict concerning what consumers feel about a firm? Future researchers can also look into the “tipping point” for when the tone of tweets is predominantly more rational or emotional and investigate what causes this – is this individual consumer behavior that is based on personality, or is there another factor? Additionally, future research can include other moderators, such as individual attitudes toward CSR in general. Additionally, emotional tweets need to be studied more in order to understand what consumers really mean as well as negative tweets with either COP or CSP activities.

Just as in all research, this study is not without limitations. This is an experimental design, so it has limited external validity – future researchers can glean real-time Twitter data during CSR fraud announcements in order to expand the generalizability. Due to Twitter’s recent 2022 buyout, other social media platforms should be considered for obtaining consumer social media information. Additionally, responses were not coded based on opinion type, such as comparative, explicit, implicit, direct, or indirect opinions (Liu, 2012) – future researchers may also want to consider this when analyzing Twitter data. This study should also be replicated on other social media platforms, such as LinkedIn (Huang et al., 2019), as well as use a real firm in future experiments and other industries aside from banking, such as consumer products, hospitality, and transportation.

Funding Statement

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Conflict of Interest

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Cite

×

Harrison, K., & Huang, L., (2023) . CSR Fraud and Consumer Opinions: What Simulated “tweets” Can Explain about Consumer Beliefs in CSR . Journal of Sustainable Marketing , 4 (1) 15-43 , https://doi.org/10.51300/JSM-2023-85

Share

×