Consumer skepticism towards Corporate-NGO partnerships: the impact of CSR motives, message frame and fit

Abstract

Purpose

Corporate-NGO partnerships are gaining increasing importance as part of a company's CSR effort. This study aims to understand which communication tactics (CSR motive, CSR message frame, CSR fit) lead to more positive consumer outcomes in the context of corporate-NGO partnerships, and whether consumer skepticism and consumer trust mediate the proposed relationships.

Design/methodology/approach

An online experiment was conducted (N = 298) to examine the theoretical predictions, involving a 2 (CSR motive: firm-serving/public-serving) x 2 (CSR message frame: narrative/expositive) x 2 (CSR fit: high/low) between-subjects design.

Findings

The results confirmed that consumer attitudes and electronic Word-of-Mouth (eWOM) can be affected by CSR motives and CSR fit. Also, CSR skepticism and consumer trust both mediate the relationship of CSR motives and consumer outcomes.

Practical implications

The results of this study make a strong case for expressing public-serving CSR motives and refraining from firm-serving CSR motives when communicating about a corporate-NGO partnership to consumers.

Originality/value

Focusing on the communication tactics of corporate-NGO partnerships extends existing literature by uncovering whether and how the factors driving effective communication in other CSR activities can be applied to the context of corporate-NGO partnerships.

Citation

Schade, J., Wang, Y. and van Prooijen, A.-M. (2022), "Consumer skepticism towards Corporate-NGO partnerships: the impact of CSR motives, message frame and fit", Corporate Communications: An International Journal, Vol. 27 No. 5, pp. 39-55. https://doi.org/10.1108/CCIJ-04-2022-0048

1. Introduction

Societal power dynamics and relationships among key agents of society are subjected to constant change. One of the societal actors which gained significant importance in recent years is non-governmental organizations (NGOs) (Arenas et al., 2009; Burchell and Cook, 2013). NGOs have not only obtained a pronounced voice within general societal debates, but also became an increasingly salient stakeholder group for corporations (Burchell and Cook, 2013; Esposito and Antonucci, 2022). Furthermore, as part of extensive campaigning of NGOs, public awareness about corporations' unethical business practices has increased, thereby pressuring companies to adopt Corporate Social Responsibility (CSR) (Burchell and Cook, 2013; Brzustewicz et al., 2022). The former idea of NGOs and corporations functioning as opponents has been replaced by a new understanding that the two agents can also collaborate and join forces to achieve societal change (Arenas et al., 2009). Accordingly, the last decade has been characterized by the ever-increasing importance of partnerships between corporations and NGOs (Austin and Seitanidi, 2012; Seitanidi and Crane, 2009; Yaziji and Doh, 2009). Since there is a general tendency towards problem-oriented and strategic collaborations (CandE Advisory Services Limited, 2019), this research focuses on partnerships that go beyond philanthropism and are based on mutual exchange of expertise, thereby conforming to what Austin (2000) defines as transactional partnerships.

However, the accumulation of CSR-related corporate scandals has increased consumers' caution against CSR activities (Connors et al., 2017). Simply engaging in CSR and reaping the benefits is not possible anymore. The success of CSR activities and, hence, also corporate-NGO partnerships is strongly dependent on the communication with external stakeholders such as consumers (Du et al., 2010). Yet, research suggests that consumer skepticism is a major factor for the decreasing effectiveness of CSR communication (Connors et al., 2017; Du et al., 2010; Skarmeas and Leonidou, 2013). Higher consumer skepticism has been shown to lead to more negative consumer attitudes (Lim, 2019)—which provide the basis for subsequential behavioral processes (Petty and Cacioppo, 1986). This can potentially lead to more negative electronic Word-of-Mouth (eWOM)—which nowadays is a key determinant of corporate reputations (Van Norel et al., 2014). As such, it is important to understand how different communication strategies regarding the promotion of corporate-NGO partnerships can impact consumer responses.

Previous research tackling CSR communication strategies mainly focused on unspecific CSR activities such as companies supporting a cause (e.g. Bae, 2018; Becker-Olsen et al., 2006; Dhanesh and Nekmat, 2019; Ellen et al., 2006; Kim, 2014; Lim, 2019; Pérez et al., 2020; Shim et al., 2017), or CSR sponsorships (e.g. Elving, 2013; Forehand and Grier, 2003; Kim and Ferguson, 2019; Rifon et al., 2004). In the current research we use an experimental design to investigate the extent to which communicated CSR motives (firm-serving vs public-serving), CSR message frame (expositive vs narrative), and CSR fit (high vs low) affect consumer attitudes and eWOM when communicating about corporate-NGO partnerships. In addition, we focus on how consumer skepticism and consumer trust mediate the proposed relationships.

Corporate-NGO partnerships have received little empirical research attention (Kim et al., 2012; Rim et al., 2016). Moving beyond sponsorships which are often short-run oriented, partnerships demonstrate a business's critical ability to collaborate across profit—non-profit boundaries (Porter and Kramer, 2011; Seitanidi and Crane, 2009), and have been recognized for its great potential for value-creation through developing a long-term relation (Yaziji and Doh, 2009). Given its long-run orientation, the mechanism underlying the impact of communicating corporate-NGO partnerships on consumer attitudes may vary from other CSR activities. Therefore, focusing on the communication tactics of corporate-NGO partnerships is of particular importance as it extends past studies by examining whether the factors driving effective communication in other CSR activities can be applied to the context of corporate-NGO partnerships.

2. Theoretical framework

2.1 Corporate-NGO partnerships and CSR communication

According to Austin (2000), cross-sector partnerships such as corporate-NGO partnerships can be classified across a collaboration continuum, where partnerships are distinguished according to their level of involvement, resource commitment and strategic importance increases (Austin, 2000). The stages on the continuum encompass philanthropic collaborations, transactional partnerships, integrative partnerships and transformational partnerships (Austin and Seitanidi, 2012). In transactional partnerships, which are the type of partnerships this study focuses on, a “reciprocal exchange of more valuable resources through specific activities” (Austin and Seitanidi, 2012, p. 736) takes place. Furthermore, this research adopts Teegen et al.’s (2004) rather loose definition of NGOs as “private, not-for-profit organizations that aim to serve particular society interests by focusing advocacy and/or operational efforts on social, political and economic goals, including equity, education, health, environment protection and human rights” (Teegen et al., 2004, p. 466).

To maximize the value of corporate-NGO partnerships, an effective communication of these activities to stakeholders is crucial (Du et al., 2010). The two key challenges that inhibit effective CSR are low levels of awareness across stakeholders as well as negative attributions and skepticism towards CSR (Du et al., 2010). Even though external stakeholders such as consumers state that CSR activities are important to them, many have a low level of awareness regarding a company's CSR efforts (Pomering and Dolnicar, 2009). However, if people are exposed to information about a company's CSR activities, thereby increasing their CSR knowledge, attitudes regarding the company and purchase intentions are positively influenced (Sen et al., 2006). Hence, CSR communication is essential for companies to raise awareness and thereby creating the foundation to be able to maximize potential business returns (Du et al., 2010). A large number of studies show that CSR communication can indeed impact the effectiveness of CSR, but that this impact depends on how the CSR communication is structured (e.g. Bae, 2018; de Vries et al., 2015; Kim, 2019; Kim and Ferguson, 2018). This demonstrates that people process information differently depending on the communication tactics.

2.2 Consumer attitudes and eWOM

Attitudes can be understood as “general evaluations people hold in regard to themselves, other people, objects, and issues” (Petty and Cacioppo, 1986, p. 4). In accordance with general consistency theory, once established attitudes are further reinforced through selective perception of information (Fiske and Taylor, 2017). Hence, the concept of attitudes becomes important for corporations and their CSR efforts not only because attitudes form the basis for subsequential behavioral processes (Petty and Cacioppo, 1986), but also because once established attitudes are rather difficult to change (Fiske and Taylor, 2017). The way CSR activities are communicated to stakeholders has an impact on stakeholders' attitudes towards the company (Becker-Olsen et al., 2006; Belas et al., 2022). Given the importance of consumer attitudes for companies, this study takes them into account when evaluating the effectiveness of CSR communication tactics.

With the rise of the Internet and the steep increase in the usage of social networking sites (SNS), the concept of eWOM has gained significant importance. eWOM is defined as “any positive or negative statement made by potential, actual, or former customers about a product or company, which is made available to a multitude of people and institutions via the internet” (Henning-Thurau et al., 2004, p. 39). It has been shown that eWOM can influence consumers' attitudes and purchase intentions (Chen et al., 2016). An underlying reason of this impact is that eWOM is characterized by high source credibility, especially in comparison with information released from companies (Bickart and Schindler, 2001). In the context of CSR communication, Connors et al. (2017) show that different CSR communication tactics such as message concreteness impact (offline) Word-of-Mouth (WOM). When specifically looking at eWOM, empirical evidence suggests that companies can stimulate eWOM by communicating their CSR activities on their social networking sites such as Facebook (Fatma et al., 2020). Therefore, this study focuses on the effects of CSR communication tactics as potential triggers for eWOM behavior.

2.3 CSR communication tactics

2.3.1 CSR motives

The first CSR communication tactic, that is of interest in this study, is communicated CSR motives. Forehand and Grier (2003) distinguish between firm-serving and public-serving motives. While firm-serving motives focus “solely on the needs of the firm itself” (Forehand and Grier, 2003, p. 350), such as reputation enhancement or profit maximization (de Vries et al., 2015), public-serving motives “refer to any motive that includes attention to the well-being of individuals outside of the firm” (Forehand and Grier, 2003, p. 350). The distinction between firm- and public-serving motives is widespread in consumer and communication research (e.g. Ham and Kim, 2020; van Prooijen et al., 2020; van Prooijen, 2019). However, findings on how perceived CSR motives relate to consumer outcomes are ambiguous. While some studies show that admitting firm-serving motives results in more positive consumer outcomes (Bae, 2018; de Vries et al., 2015; Kim, 2014), other studies report the opposite effect (Shim et al., 2017; van Prooijen et al., 2020; van Prooijen, 2019). For instance, Kim (2014) shows that news reports that present firm-serving motives in addition to public-serving motives about a company's CSR activities result in an increase in purchase intentions and make the company more attractive as an employer. Conversely, Shim et al. (2017) shows that news reports framing CSR activities of a company as firm-serving instead of public-serving decrease communication intentions and lead to higher levels of perceived corporate hypocrite.

Whether or not communicating public or firm-serving motives lead to better consumer outcomes is context-dependent (Forehand and Grier, 2003; van Prooijen et al., 2020; van Prooijen, 2019). Forehand and Grier (2003) show that expressing firm-serving motives in addition to public-serving motives only lead to more positive outcomes when consumers are activated to engage in causal reasoning beforehand. Causal reasoning describes the elaboration and reflection about underlying ulterior reasons for a company to follow a specific activity (Kelley, 1972; Kelley and Michela, 1980). Conversely, in situations in which such causal reasoning is not evoked, only expressing public-serving motives leads to better consumer outcomes (Forehand and Grier, 2003). Therefore, if there is no specific reason for consumers to actively engage in causal reasoning, then stating public-serving motives should lead to better consumer outcomes (van Prooijen, 2019).

Since corporate-NGO partnerships have experienced a momentum in recent years (Austin and Seitanidi, 2012; Brzustewicz et al., 2022; Seitanidi and Crane, 2009), consumers should be generally knowledgeable about this type of CSR activity. Furthermore, the majority of consumers rate partnerships between companies and NGOs as rather positive, reasonable and as a benefit for society (Rohwer and Topić, 2018), which is why it is expected that consumers do not automatically elaborate on underlying ulterior motives when confronted with a message about a corporate-NGO partnership. Additionally, NGOs are generally perceived as trustworthy and credible (Arenas et al., 2009), which is why positive halo effects on the company might occur when they engage in a transactional partnership. Therefore, it is expected that the communication about a corporate-NGO partnership does not immediately trigger causal reasoning and hence elaborative thoughts about ulterior reasons. Consequentially, the expression of public-serving motives which resembles an acknowledgment of the company's aim to support the NGO with the partnership should lead to better consumer outcomes. Hence, it is hypothesized that:

H1.

The communication of a public-serving CSR motive leads to more positive (a) consumer attitudes and (b) eWOM intentions than a communicated firm-serving CSR motive when communicating corporate-NGO partnerships.

2.3.2 CSR message frame

The next CSR communication tactic is CSR message frame. A distinction is made between an expositive and narrative CSR message frame. Expositive message frames are rational as well as non-narrative (Pérez et al., 2020) and “present propositions in the form of reasons and evidence supporting a claim” (Kreuter et al., 2007, p. 222). Contrarily, in a narrative message, the information about the CSR activity is embedded in a story and “describes the cause-and-effect relationships between events that take place over a particular time period that impact particular characters” (Dahlstrom, 2014, p. 13614). However, research that focuses on CSR message frame shows ambiguous results (e.g. Dhanesh and Nekmat, 2019; Pérez et al., 2020). Pérez et al. (2020) indicate that a narrative CSR message frame generally leads to a more positive perception of the company's CSR commitment, the impact of the CSR activity, and its fit. However, purchase intentions and advocacy are higher when an expositive CSR message frame is applied (Pérez et al., 2020). Furthermore, Dhanesh and Nekmat (2019) suggest that expositive CSR messages lead to more positive attitudes, recommendations, and purchase intentions. However, this only holds true for involved audiences. The researchers propose to use narrative CSR message frames when targeting audiences with low involvement (Dhanesh and Nekmat, 2019). In this research, a twofold argumentation is used to explain why a narrative message frame in the context of CSR is expected to increase consumer outcomes.

Firstly, narrative CSR messages are expected to lead to more positive consumer outcomes. According to the narrative paradigm theory, “humans are essentially storytellers” (Fisher, 1984, p. 7). As “narrative beings” (Fisher, 1984, p. 8), human rationality is guided by stories (Fisher, 1984). Hence, narrative CSR message frames are expected to create more positive responses since they conform with individuals' nature of processing information. Furthermore, Escalas (2004) argues that consumers who face narrative messages process these by creating links between themselves and the company by establishing self-brand connections. This process results in more positive attitudes and behavioral intentions (Escalas, 2004).

Secondly, expositive CSR message frames are expected to decrease consumer outcomes. Lundqvist et al. (2013) show that narrative advertisements promote more positive attitudes and increase purchase willingness. Conversely, consumers who are exposed to an expositive advertisement are inclined to be more critical and develop more negative attitudes (Lundqvist et al., 2013). Furthermore, when consumers face narrative CSR message frames, the elaborative effort to process the information is reduced compared to expositive CSR message frames (Pérez et al., 2020). Hence, consumers are more likely to use heuristic processing for narrative messages (Pérez et al., 2020). In heuristic processing, consumers adhere to heuristics and put “little effort in judging message validity” (Chaiken, 1980, p. 752). Conversely, Pérez et al. (2020) argue that expositive messages are systematically processed. In systematic processing, consumers “actively attempt to comprehend and evaluate the message's arguments” (Chaiken, 1980, p. 752), meaning that more cognitive effort is used to engage with the message content. However, Alcañiz et al. (2010) argue that systematic processing can increase suspicion and can ultimately result in more negative consumer outcomes. Taken together, it is expected that narrative CSR message frames lead to more positive outcomes as they resemble a consumer's natural information processing. Therefore, it is expected that:

H2.

A narrative CSR message frame leads to more positive (a) consumer attitudes and (b) eWOM intentions than an expositive CSR message frame when communicating corporate-NGO partnerships.

2.3.3 CSR fit

CSR fit is an extensively researched topic in the field of CSR communication (Kim et al., 2012). Research that specifically focuses on fit in the context of CSR partnerships generally agrees that high-fit CSR partnerships result in more positive consumer outcomes (e.g. Alcañiz et al., 2010; Elving, 2013; Rim et al., 2016). For instance, Rim et al. (2016) reveal that high-fit results in an increase in supportive CSR outcomes, higher consumer-company identification, and higher levels of perceived altruism.

Some studies go a step further and differentiate between different types of fit for cross-sector partnerships (e.g. Kim et al., 2012; Simmons and Becker-Olsen, 2006). Kim et al. (2012) argue that this is necessary because conventional CSR fit definitions often confuse the partner organization and the supported purpose and thus do not do justice to the complexity of the fit concept. Hence, Kim et al. (2012) distinguish between three types of fit – the familiarity, business and activity fit. This study focuses on business fit which is defined as “the degree that the business domain of one organization matches with the business domain of another” (Kim et al., 2012, p. 164). The reason why this type of fit is examined in this research is that it allows getting more insights for the partner selection process. Furthermore, this type of fit is also relevant from a cost-efficient perspective since the partner organizations do not have to spend money on creating an actual fit between each other (Simmons and Becker-Olsen, 2006).

The decisive factor for these different consumer reactions towards CSR fit is the varying way in which the information about the partnership is processed (Rifon et al., 2004; Simmons and Becker-Olsen, 2006). High-fit scenarios are evaluated more positively because the new knowledge is linked to existing expectations thereby decreasing the need for in-depth elaboration (Rifon et al., 2004; Simmons and Becker-Olsen, 2006). Conversely, low-fit scenarios generate more elaborative thoughts about the partnership because consumers might start questioning the underlying motives for setting up the partnership, which leads to more negative consumer outcomes (Rifon et al., 2004; Simmons and Becker-Olsen, 2006). Simmons and Becker-Olsen (2006) confirm that low fit between partners leads to more negative attitudes due to the increase in elaboration which is biased towards negativity. In addition, Alcañiz et al. (2010) argue that in low-fit scenarios, consumers engage in systematic information processing (as defined by Chaiken, 1980) and teleological reasoning. Furthermore, in high-fit scenarios, consumers engage in heuristic information processing (as defined by Chaiken, 1980) and deontological reasoning, thereby increasing consumer outcomes. Taken together, low CSR fit triggers elaboration about the partnership since the lack of congruence does not fit with existing knowledge patterns. As explained above, the involvement of elaboration is associated with systematic information processing as well as the use of persuasion knowledge which both hold the potential to decrease consumer outcomes. Hence, it is hypothesized that:

H3.

A high CSR fit leads to more positive (a) consumer attitudes and (b) eWOM intentions than a low CSR fit when communicating corporate-NGO partnerships.

2.4 The mediating role of situational CSR skepticism and consumer trust

Forehand and Grier (2003) identify two types of consumer skepticism – dispositional and situational skepticism. While dispositional skepticism refers to an individuals' general tendency to challenge given information, situational skepticism is induced through “situational variables that direct consumers attention” (Forehand and Grier, 2003, p. 349). Both types of skepticism can be applied to the CSR context (Connors et al., 2017; Ham and Kim, 2020; Skarmeas and Leonidou, 2013). This research focuses on situational CSR skepticism, as it refers to a direct response towards a CSR activity induced through a company's communication (Ham and Kim, 2020).

If an event requires increased elaboration and cognitive effort, it is more likely that persuasion knowledge is used so that the persuasion attempt can be corrected (Campbell and Kirmani, 2000; Friestad and Wright, 1994). The use of persuasion knowledge is associated with skepticism (Friestad and Wright, 1994). This idea is exemplarily illustrated by Lim (2019) whose results indicate that low CSR fit leads to higher levels of situational CSR skepticism. However, situational CSR skepticism is not only influenced through CSR communication tactics, but also influences consumer outcomes. Studies indicate that situational skepticism leads to more negative consumer outcomes (e.g. Elving, 2013; Ham and Kim, 2020; Skarmeas and Leonidou, 2013). For instance, Skarmeas and Leonidou (2013) show that in the context of CSR, situational skepticism leads to lower company equity, more negative word-of-mouth as well as lower resistance to negative information. Hence, situational skepticism is not only influenced by CSR communication tactics but also itself negatively impacts consumer outcomes. In other words, situational skepticism functions as a mediator between CSR communication tactics and consumer outcomes (Ham and Kim, 2020; Lim, 2019). Taken together, situational CSR skepticism is predicted to be triggered when consumers engage in more extensive cognitive elaboration about a CSR activity. As explained in previous sections, this is expected to be the case for perceived firm-serving CSR motives, expositive CSR message frame, and low CSR fit. This higher level of situational CSR skepticism is then expected to lead to lower consumer outcomes, thereby making situational CSR skepticism a mediator in the relationship between CSR communication tactics and consumer outcomes.

Trust within the context of CSR is defined as “an individual's firm belief in the reliability of a corporation's promise regarding its CSR activities” (Kim, 2019, p. 1147). The concept has been identified as a mediator between the perception of different CSR communication factors and their consumer outcomes (Kim, 2019; Yu et al., 2021). For instance, a promotional-tone message about a CSR activity results in lower levels of trust which then decreases the reputation of the company (Kim, 2019). Interestingly, Kim et al. (2019) showed that CSR messages which use textual and visual elements result in higher trust levels while, at the same time, minimize skepticism. The findings in previous literature indicate that trust and skepticism may function as antagonists in mediating the relationship between CSR communication tactics and consumer outcomes. While skepticism is a key construct that inhibits the effectiveness of CSR activities (Du et al., 2010), trust is key for explaining the success of CSR activities (Kim et al., 2019). Hence, it is hypothesized that:

H4.

Situational CSR skepticism mediates the relationship between CSR motives and (a) consumer attitudes and (b) eWOM intentions when communicating corporate-NGO partnerships.

H5.

Situational CSR skepticism mediates the relationship between CSR message frame and (a) consumer attitudes and (b) eWOM intentions when communicating corporate- NGO partnerships.

H6.

Situational CSR skepticism mediates the relationship between CSR fit and (a) consumer attitudes and (b) eWOM intentions when communicating corporate-NGO partnerships.

H7.

Consumer trust mediates the relationship between CSR motives and (a) consumer attitudes and (b) eWOM intentions when communicating corporate-NGO partnerships.

H8.

Consumer trust mediates the relationship between CSR message frame and (a) consumer attitudes and (b) eWOM intentions when communicating corporate-NGO partnerships.

H9.

Consumer trust mediates the relationship between CSR fit and (a) consumer attitudes and (b) eWOM intentions when communicating corporate-NGO partnerships.

3. Method

3.1 Research design

A 2 (CSR motive: firm-serving/public-serving) x 2 (CSR message frame: narrative/expositive) x 2 (CSR fit: high/low) between-subjects design was deployed to examine our hypotheses. The three independent variables were operationalized across eight experimental scenarios, each consisting of a short company description on the company's Facebook page and two Facebook posts about the corporate-NGO partnership posted on the company's Facebook page and a newspaper's Facebook page. Facebook posts were chosen as the medium to communicate the corporate-NGO partnership because eWOM was measured as an outcome variable. Furthermore, Facebook is one of the most widely used social networking sites worldwide as it reaches 1.8 billion users daily and 2.7 billion monthly users (Facebook, 2020).

Research by Brown and Dacin (1997) shows that prior knowledge and associations can influence individual's attitudes and beliefs towards an organization. Hence, this experiment uses fictitious organizations (company and NGO) to rule out any company-related biases, thereby increasing the internal validity of the experiment (Kim and Ferguson, 2019). The fictitious company was chosen to be a restaurant chain. Previous studies on CSR communication already used this type of fictitious company in their research design (e.g. Lim, 2019; Pérez et al., 2020) because the food industry is a crucial part of the overall economy and is present in consumers' daily life (Maloni and Brown, 2006). The name of the fictitious restaurant chain was chosen to be Nutric Food. A fictitious NGO that teaches healthy eating habits and another one that advocates against bullying were chosen for the high and low CSR fit conditions, respectively. To avoid any confounding effects due to different names of the fictitious NGOs, both were named StrongTogether. To make the experimental scenarios as realistic as possible, the German restaurant chain deananddavid Superfood GmbH (2020), and the organizations STOMP Out Bullying (2021, 2020), Common Threads (2020), The Food Trust (2012), and their paper on “One healthy breakfast toolkit” (The Food Trust, n.a) served as informative sources to create the stimulus material.

The first stimulus material was a short description of Nutric Food on the company's Facebook page to provide some basic knowledge about the business activities of the fictitious restaurant chain. Afterward, participants were shown a post on Nutric Food's Facebook page. The post contained an announcement of the newly established corporate-NGO partnership of the company. With these Facebook posts, CSR fit and CSR message frame were operationalized. Furthermore, a post of a fictitious newspaper named Today Corporate was used as a third stimulus material to convey the communicated CSR motives from the company to enter the corporate-NGO partnership.

3.2 Measurements

Validated scales which are well recognized in the CSR literature were used to measure the dependent and mediation variables. The dependent variable consumer attitudes was operationalized with a scale from Nan and Heo (2007) (M = 5.85, SD = 0.99, α = 0.91). The four-item scale measured consumer attitudes with a seven-point bipolar scale across four dimensions – negative/positive, dislike/like, unfavorable/favorable and socially irresponsible/socially responsible. Higher scores indicated more positive consumer attitudes.

The dependent variable eWOM intentions was measured with a seven-item scale developed by Eelen et al. (2017) (M = 3.67, SD = 1.46, α = 0.91). To quantify eWOM, participant's likelihood to do certain eWOM activities was measured on a seven-point Likert scale (1 = extremely unlikely, 7 = extremely likely).

The first mediator—Situational CSR skepticism—was operationalized with a four-item scale developed by Skarmeas and Leonidou (2013) (M = 2.70, SD = 1.24, α = 0.88). Each item was measured with a seven-point bipolar scale anchored on doubtless/doubtful, certain/uncertain, sure/unsure and unquestionable/questionable. Higher scores indicate higher levels of skepticism.

The second mediator, consumer trust towards the company's CSR activities, was measured with a four-item scale developed by Kim (2019) (M = 5.36, SD = 0.87, α = 0.89). The items were slightly adapted so that instead of focusing on general CSR activities, the items specifically referred to NGO partnerships. All items were measured using a seven-point Likert scale (1 = strongly disagree, 7 = strongly agree). Table 1 shows the correlation matrix of the variables.

To be able to account for possible confounding effects, two questions about participants' frequency of Facebook usage and the average length of stay on Facebook from Mazman and Usluel (2010) were included.

3.3 Manipulation checks

In the manipulation check on CSR motives, participants were asked to indicate Nutric Food's CSR motive they were exposed to and were able to choose between the two answer options, firm-serving and public-serving. The results of the Chi-Square test of independence revealed that 70.8% of the participants answered this manipulation check correctly. With 95% certainty, the manipulation check on CSR motives was successful, χ² (7, N = 298) = 55.78, p < 0.001.

Next, the manipulation check on CSR message frame asked the participants about the writing style of Nutric Food's Facebook post. Again, participants had to choose between two answer options that replicated the two CSR message frame conditions. The Chi-square test of independence indicated that 85.4% of the participants answered the manipulation check correctly. Within a 95% confidence interval, the CSR message frame manipulation check was successful, χ² (7, N = 298) = 151.91, p < 0.001.

The last manipulation check focused on CSR fit. Participants had to evaluate the fit between Nutric Food and the partner NGO and were able to choose between high-fit and low-fit. The Chi-square test of independence indicates that 88.6% of the participants answered the manipulation check correctly. Within a 95% confidence interval, the CSR fit manipulation check succeeded, χ² (7, N = 298) = 182.88, p < 0.001.

3.4 Sample

As the experiment referred to the social networking platform Facebook it was a prerequisite for participants to use Facebook at least once a month. This pre-screening was performed via Prolific. For successfully completing the experiment, each participant received £1.13. The final sample consisted of 298 participants (61% male, 36.6% female, 2.3% non-binary/third gender). Participants' age ranged from 18 to 71 years (M = 27.96, SD = 9.60). A total of 37 nationalities was represented in the sample with the majority of participants from Poland (20.5%) and the United Kingdom of Great Britain and Northern Ireland (20.1%). The majority of the participant were either students (37.2%) or full-time employed (30.5%).

4. Results

4.1 Effects of communication tactics on consumer attitudes and eWOM

The effect of each CSR communication tactic as well as their interaction effect on consumer outcomes was tested with a three-way ANOVA. The ANOVA for consumer attitudes yielded a significant main effect of motives, F(1, 290) = 21.58, p < 0.001, η² = 0.07. In line with Hypothesis 1a, attitudes were more positive in the public-serving CSR motives conditions (M = 6.08, SD = 0.79) than in the firm-serving CSR conditions (M = 5.58, SD = 1.13). The main effect of CSR fit on consumer attitudes was also significant, F(1, 290) = 6.16, p = 0.014, η² = 0.02. A high CSR fit resulted in more positive attitudes (M = 5.96, SD = 1.00) than a low CSR fit (M = 5.72, SD = 0.98), which supported Hypothesis 3a. No main effect of message frame was found, p = 0.08. Thus, Hypothesis 2a was rejected. Furthermore, none of the interaction effects were significant, all p's > 0.15.

A main effect of motives on eWOM intentions was observed, F(1, 290) = 10.28, p = 0.001, η² = 0.03. Participants reported more eWOM intentions in the public-serving CSR motive conditions (M = 3.92, SD = 1.49) than in the firm-serving CSR motives conditions (M = 3.38, SD = 1.38), which supports Hypothesis 1b. The main effect of CSR fit on eWOM intentions was also significant, F(1, 290) = 4.491, p = 0.035, η² = 0.02. In line with Hypothesis 3b, participants had stronger intentions to engage in eWOM in the high CSR fit conditions (M = 3.81, SD = 1.49) than in the low CSR fit conditions (M = 3.50, SD = 1.41). No main effect was observed for message frame, p = 0.318. Consequently, Hypothesis 2b had to be rejected.

The interactions effects between the independent variables were also not significant, all p's > 0.09.

4.2 Mediation effects of situational skepticism and consumer trust

To test the mediation effects, PROCESS model 4 developed by Hayes (2021) was utilized. The indirect effect of CSR motives on consumer attitudes via skepticism was found to be statistically significant, b = −0.22, CI95% [−0.41, −0.06]. Furthermore, an indirect effect of CSR fit on consumer attitudes via skepticism was found, b = −0.20, CI95% [−0.37, −0.04]. No significant indirect effect was observed of CSR message frame on consumer attitudes via skepticism, b = −0.13, CI95% [−0.32, 0.03]. We therefore accepted hypotheses 4a and 6a, but rejected Hypothesis 5a.

A similar test was conducted considering eWOM intentions as the dependent variable. An indirect effect through skepticism was observed for the relation between CSR motives and eWOM intentions, b = −0.15, CI95% [−0.28, −0.04]. However, no indirect effects were found on eWOM intentions through skepticism for message frame, b = −0.09, CI95% [−0.21, 0.02] or for CSR fit, b = −0.14, CI95% [−0.28, 0.02]. Thus, we accepted Hypothesis 4b, while we rejected Hypotheses 5b and 6b.

In support of Hypothesis 7a, the effect of CSR motives on consumer attitudes was mediated by trust, b = −0.23, CI95% [−0.39, −0.09]. However, we rejected Hypotheses 8a and 9a, as trust did not function as a mediator for message frame, b = 0.08, CI95% [−0.24, 0.06], or for CSR fit, b = −0.14, CI95% [−0.29, 0.01].

Trust mediated the relation between CSR motive on eWOM intentions, b = −0.22, CI95% [−0.36, −0.08], which supported Hypothesis 7b. No mediation effects of trust were found for the relation between eWOM intentions and message frame, b = 0.08, CI95% [−0.22, 0.07], or CSR fit, b = 0.13, CI95% [−0.28, 0.00]. Hypotheses 8b and 9b were rejected.

5. Discussion and conclusion

The findings showed that communicated public-serving CSR motives led to more positive consumer attitudes and eWOM intentions than firm-serving CSR motives. Since studies on this topic generally have come to contradicting results (e.g. Shim et al., 2017; van Prooijen et al., 2020; van Prooijen, 2019), this study followed the line of argumentation that the outcomes of perceived CSR motives are context-dependent (Forehand and Grier, 2003). It seems that the decisive factor for communicated public-serving motives to result in more positive consumer outcomes is the active involvement of the NGO in the partnership. Furthermore, it appears feasible that NGOs' trustworthiness and credibility (Arenas et al., 2009) influence the perception of the entire corporate-NGO partnership and thereby also positively impact the image of the corporation, which makes consumers unlikely to elaborate on ulterior motives when being confronted with a message about a company's partnerships.

Additionally, based on a large body of previous literature which offers evidence that high-fit CSR activities result in more positive consumer outcomes than low-fit CSR activities (e.g. Alcañiz et al., 2010; Elving, 2013; Rim et al., 2016; Simmons and Becker-Olsen, 2006), this study contemplated that high CSR fit results in more positive consumer attitudes and eWOM when communicating corporate-NGO partnerships. The results aligned with the existing literature, and resembled Simmons and Becker-Olsen's (2006) result who also specifically focused on natural fit and confirmed that high fit leads to better consumer outcomes.

This study did not find support for the widespread conception that humans think in stories, thereby making narrative messages superior to expositive ones (Escalas, 2004; Fisher, 1984). A potential explanation could be that consumers' characteristics and beliefs about CSR might impact the effectiveness of the message frame. In detail, Pérez et al. (2020) argue that consumers' general level of skepticism concerning CSR might impact the effectiveness of narrative CSR message frames. They argue that consumers with higher levels of dispositional skepticism might respond more positively to expositive messages because they value facts and rational arguments. This explanation is also supported by Du et al.’s (2010) CSR communication framework which proposes that next to company characteristics also stakeholder characteristics determine the effectiveness of CSR communication by functioning as moderators. Empirical findings confirm that dispositional CSR skepticism can function as a moderator in the relationship between CSR communication tactics and consumer outcomes (de Vries et al., 2015) and that the level of dispositional CSR skepticism influences how CSR messages are perceived (Connors et al., 2017).

Our study also showed that the relations between CSR motives and both consumer outcomes were mediated by situational skepticism and consumer trust. In addition, the relation between CSR fit and eWOM was mediated by situational skepticism. Based on Becker-Olsen et al. (2006), Simmons and Becker-Olsen (2006), and Rifon et al. (2004), we expected that high CSR fit relates to increased elaboration and use of persuasion knowledge. Furthermore, Campbell and Kirmani (2000) as well as Friestad and Wright (1994) argue that the use of persuasion knowledge is associated with situational CSR skepticism. It indeed seems like such an increase in elaboration and use of persuasion knowledge was triggered through high CSR fit, and therefore situational CSR skepticism was also stimulated.

While managers can choose from a variety of CSR communication tactics, this research uncovered that not all of them influence consumers in equal ways. For improving the effectiveness of corporate-NGO partnerships, stated CSR motives and CSR fit should be the primary focus of communication managers. The results of this study make a case for expressing public-serving CSR motives and refraining from firm-serving CSR motives when communicating about a corporate-NGO partnership. It was also shown that situational CSR skepticism and consumer trust mediate this relationship – public-serving CSR motives decreased situational CSR skepticism and increased consumer trust which then impacted consumer outcomes. Communication managers should use any opportunity to decrease situational CSR skepticism. This is important since Forehand and Grier (2003) suggest that situational and dispositional skepticism might be associated and because Connors et al. (2017) argue that dispositional skepticism towards CSR is generally increasing.

Our findings on the effectiveness of different CSR communication tactics and the deployed underlying concepts open up the opportunity for future research. Future studies could shift the focus to the NGOs involved in the partnerships. Studies could focus on the role CSR communication plays for the success of a partnership from the perspective of an NGO as well as confirm and investigate which CSR communication tactics lead to more positive outcomes from an NGO perspective. Since the findings of perceived CSR motives allow the assumption that part of the NGO's credibility is transported in the CSR partnership, more research is needed to understand how the active involvement of an NGO in a corporations' CSR activity changes the perception of CSR activities. In this regard, a comparison between different CSR activities and how consumers respond to them could be interesting.

Table 1

Correlation matrix

	Consume attitudes	eWOM	Situational CSR skepticism	Consumer trust	Age	Gender	Frequency Facebook usage	Length of stay on facebook
Consumer attitude	1
eWOM	0.44***	1
Situational CSR skepticism	−0.73***	−0.37***	1
Consumer trust	0.64***	0.41***	−0.64***	1
Age	0.04	−0.09	−0.08	−0.08	1
Gender	0.09	0.04	−0.08	0.04	0.09	1
Frequency Facebook usage	−0.14*	−0.21***	0.14*	−0.10	0.00	−0.01	1
Length of stay on Facebook	0.09	0.24***	−0.12*	0.12*	−0.07	0.05	−0.52***	1

Note(s): Significance levels ***p < 0.001 (2-tailed), **p < 0.01 level (2-tailed), *p < 0.05 level (2-tailed)