Introduction

The responsibility of firms toward their stakeholders and the concrete actions this responsibility requires have become a central managerial concern. We nevertheless observe an ambiguity in the nature of firms’ socially “responsible” activities and the way they are designed. Corporate social actions (CSA) are designed to ensure the stakeholders’ welfare (Marquis et al. 2007) and include a broad range of activities implied by the firm’s corporate social responsibility (CSR). But the true motivations behind CSA have remained a constant topic of inquiry (Buehler and Shetty 1974). Firms may design these actions to mislead their stakeholders about the magnitude of their engagement in order to gain legitimacy (Walker and Wan 2012). In this respect, they are commonly said to engage in what has been termed “greenwashing” (Delmas and Burbano 2011). Greenwashing is defined as giving priority to CSA symbolically rather than substantively (Walker and Wan 2012). They opt for actions aimed at anchoring the firms’ conformity to social responsibility (e.g., green accreditations, CEO speeches, sustainability reporting) in the stakeholders’ mind to the detriment of what is effectively impacting the stakeholders’ welfare (e.g., reduction in CO2 emissions, implementation of safety guidelines at work) (Walker and Wan 2012). In other words, the balance of their actions favors a stated and ostensible commitment to benefit stakeholders rather than a genuine dedication to effectively increase the stakeholders’ welfare (Lim and Tsutsui 2012).

Business scholars have progressively tried to come up with an instrumental justification for CSA (Lee 2008). Several try to understand the consequences of substantive CSA in terms of the firms’ financial performance (McWilliams and Siegel 2001; Waddock and Graves 1997; Walker and Wan 2012). Others suggest that firms could design and frame their social activities symbolically and still maximize their profits (Weaver et al. 1999; Christmann and Taylor 2006; Hawn and Ioannou 2012). Quite a number of studies also attempt to identify genuine motives which lead to CSA (Hemingway and Maclagan 2004). In particular, the literature has largely pointed out the role of cultural beliefs at the country level in pushing firms toward more or less CSA. Populations share cultural beliefs and values that may foster or dampen firms’ implementation of CSA. Matten and Moon (2008) argue, for instance, that the nature of economic systems and rationales at the national level influence the features of CSA, whether they are “implicit” (abiding by social norms) or “explicit” (voluntary programs). The role played by institutional diversity in the approach to CSA is a fruitful area of research (Brammer et al. 2012).

While the impact of liberalism on CSA has often been looked at (see Campbell 2007; Matten and Moon 2008; Jackson and Apostolakou 2010; Lim and Tsutsui 2012; Kinderman 2012), Jackson and Deeg (2008, p. 541) have stressed that institutions should not be analyzed “as unidimensional variables” or operationalized with “summary indicators.” At the country level, economic systems are based on a complex combination of shared beliefs regarding economic rationality (Denzau and North 1994). The beliefs in the virtues of liberalism form one example of these cultural attitudes, which may or may not be shared by populations at the country level. We suggest that two distinct cultural beliefs prevail: the populations’ belief in the central role of individual responsibility rather than that of governments, and the belief in the advantages of competition (Hay 2004; Whitman 1999; Fligstein 2001). Populations which believe in the virtues of liberalism in turn believe in the virtue of individual responsibility and competition rather than governmental responsibility and regulation. Lim and Tsutsui (2012) argue that more liberal economic systems, more competitive and individualist societies proportionally increase the likelihood of firms’ implementing CSA, although these actions remain symbolic rather than substantive in developed countries. These scholars underline how liberal systems lead firms to anticipate criticism and build their legitimacy with impression management tactics such as symbolic CSA (Prechel and Morris 2010; Jones and Nisbet 2011). This argument relies on the idea that firms take advantage of the freedom offered by less institutionalized environments (Julian and Ofori-Dankwa 2013). However, Jackson and Apostolakou (2010) put forward a different argument, suggesting that in liberal economies where the role of government is limited and competition is fierce, a vacuum emerges from the lack of regulation, and firms, therefore, tend to engage in more substantive social behaviors to answer stakeholders’ needs that are not met by the state. The extent to which greenwashing occurs reflects the balance between symbolic and substantive CSA (Walker and Wan 2012): firms that are willing to engage in CSA for the legitimacy benefits it generates will tend to favor symbolic rather than substantive actions. Thus, it is unclear whether beliefs in the virtue of liberalism itself favor or hamper the likelihood to engage in greenwashing.

These two rationales implying that liberalism has a bearing on the propensity to greenwash differ in their propositions, because they make antithetic assumptions regarding firms’ reactions to the individualistic and competitive nature of their environment. Conventional wisdom holds that populations’ cultural belief in the virtues of liberalism leads to more greenwashing. In this sense, it implies that firms, as social entities, are inherently selfish. When embedded in a liberal context, they favor impression management tactics and symbolic CSA which make stakeholders believe a firm is responsible, when it is not, with limited costs. On the contrary, the proponents of voluntarism consider that beliefs in liberalism lead to more effective actions in terms of social impact on stakeholders rather than in terms of benefits for the firm. In that sense, it makes the unspoken assumption that firms are pro-active: When favoring liberal values, firms step into the space left by a weak government by providing stakeholders with beneficial CSA.

In this study, we specifically focus on the impact of two beliefs on how CSA is approached, which play a fundamental role in designing liberal economic systems: the belief in the beneficial role of competition, and the belief in the preeminence of individual responsibility (Hay 2004). We first set up the conceptual building blocks of our study, by defining the key concepts and our theoretical lens. We then begin investigating how those concepts articulate through an exploratory qualitative study. We use a focus group method, commonly used in anthropology (Agar and MacDonald 1995) and in medical research (Kitzinger 1995), but also in business ethics (Vyakarnam et al. 1997; Freestone and Mitchell 2004). We complement this with interview data. Building on these findings, we put forward two alternative sets of hypotheses. Following the traditional view, we expect beliefs in favor of competition and individual responsibility both to induce greenwashing, but we also formulate the opposite hypothesis, that beliefs in favor of competition and individual responsibility will be negatively related to greenwashing.

To date, even if some scholars investigated the impact of liberal economies on the nature of CSA (Jackson and Apostolakou 2010; Lim and Tsutsui 2012) or the determinants of greenwashing (Delmas and Burbano 2011), no study has looked at how fundamental shared beliefs that underpin liberalism can be a potential determinant of greenwashing. By disentangling the differential effects of our two foundational beliefs, we aim at understanding how seemingly similar values can have contrasting effects on corporate action. We construct our measures of country-level beliefs in the role of competition and individualism from the World Value Survey (WVS), assess the nature of CSA at the firm level using Thomson Reuters’ ASSET4 database, and then test our hypotheses on a sample of 2,621 firms spanning from 2002 to 2008. Our results show that the two foundational beliefs we study have opposite effects.

In this sense, our findings confirm Bowie’s challenge of the “egoistic paradigm” (Bowie 1991). They show that even under the assumption of economic rationality, when populations believe in the virtues of liberalism, they do not necessarily screen out actions going beyond shareholder value maximization (Stormer 2003). When beliefs in the virtues of competition are strong, firms are more likely to greenwash, while in individualistic societies firms tend to be less favorable to greenwashing.

Introducing the Concepts: Greenwashing, Economic Liberalism, and Shared Beliefs

The role of business in society has been of central interest for management scholars, since social and environmental concerns have been more and more pressing (Margolis and Walsh 2003). CSR refers to the responsibility of firms toward society. CSA refers to the actual undertakings this responsibility implies (Marquis et al. 2007). CSAs are defined as initiatives that aim at benefiting the welfare of a firm’s stakeholders (Marquis et al. 2007). This distinction is made necessary by situations in which firms might engage in CSAs without being socially responsible (Buehler and Shetty 1974), just because it generates positive spillover effects. However, the institutional determinants of corporate social behaviors remain to be explored (Brammer et al. 2012).

Depending on the country, CSA can be seen as more or less legitimate (Lim and Tsutsui 2012; Julian and Ofori-Dankwa 2013). Firms tend to implement actions which provide them with legitimacy gains (DiMaggio and Powell 1983). Therefore, depending on the firms’ country of origin, CSA is implemented in different ways. This article falls within the scope of studies looking at how a cultural context may affect socially responsible behaviors (Campbell 2007), and tries to explain how and why the nature of CSA may vary across countries (Maignan and Ralston 2002). More specifically, we focus on how country-wide beliefs in the virtues of liberalism may lead to greenwashing, or in other terms, “talking about,” rather than actually engaging with CSA (Delmas and Burbano 2011; Walker and Wan 2012).

What is “Greenwashing?”

Firms are sometimes prone to comply only in appearance with stakeholders’ needs (Westphal and Zajac 2001) when they try to build or repair reputation or gain legitimacy (Prechel and Morris 2010). With the growing concerns regarding their responsibility toward society, firms have learned how to design their CSA strategically rather than implement it substantively (Walker and Wan 2012). Mainstream media and management research have discussed those strategies using the umbrella term “greenwashing” (Laufer 2003; Ramus and Montiel 2005; Delmas and Burbano 2011).

A common way to define greenwashing is to refer to the nature of the firm’s CSA: Walker and Wan (2012) define greenwashing as the gap between “symbolic” and “substantive” CSA. Management scholars have for a long time contrasted different forms of CSA: to respond to institutional pressure, firms may substantively conform to imposed norms, or pretend to do so (i.e., ostensible conformity) (Ashforth and Gibbs 1990; Westphal and Zajac 1994). More specifically, when we talk about CSA, symbolic CSA includes what the company claims to plan to do in terms of socially responsible behaviors, while substantive CSA is what the company is actually doing or has done in terms of CSA (Walker and Wan 2012). Thus, greenwashing is a decoupling strategy that aims at gaining legitimacy and signaling conformity rather than actually conforming (Delmas and Burbano 2011). This usually takes the form of favoring symbolic actions (Weaver et al. 1999) such as policy claims and codes of conduct (Christmann and Taylor 2006), social accreditations, or green labels (Walker and Wan 2012). Russo and Harrison (2005) take the example of the ISO 14001 certification and prove, paradoxically, that it is related to more environmental wrongdoing. Firms can take actions detrimental to the environment but still be certified (King et al. 2005), they can implement substantive measures to protect their employees but stay silent about them. Equally, they can simultaneously engage in both substantive and symbolic actions (Hawn and Ioannou 2012).

Greenwashing has been proven to be strategic in the sense that it may mislead stakeholders about a firm’s actual social performance. This strategy may trigger stakeholder support, but with minimal investment in substantive (and necessarily more costly) CSA (Husted and Allen 2009). However, greenwashing is not necessarily “bought” by stakeholders and when this is the case, it may also destroy value (Walker and Wan 2012). We propose, therefore, that what matters when it comes to engaging in greenwashing are the beliefs and values anchored in the firm’s direct environment. More specifically, we suggest that it is beliefs in the virtues of liberalism which drive the various approaches to CSA. Jackson and Deeg (2008) have explained the benefits of looking at precise sub-elements of the institutional context for international business research, where institutions are often approached as monolithic determinants of corporate behaviors. In this study, we offer a more fine-grained understanding of the effect of economic liberalism on corporate behaviors.

Economic Liberalism and Shared Beliefs

Research has pointed out the role of country-level cultural values in the construction of corporate behaviors (Matten and Moon 2008). Rather than simply trying to act in their own interest, individuals share beliefs that model their behaviors (Denzau and North 1994). Cultural beliefs are norms and values that are shared at the country level and may influence both stakeholders and organizational decision-making. Denzau and North (1994) explain that social agents sharing “common cultural background” make sense of their environment in the same way and have a similar set of reactions at their disposal. Beliefs sustain the rationale of social agents, which in turn enable and constrain their social and economic behaviors (Thornton and Ocasio 1999, 2008).

Country-level institutional contexts, therefore, act as both a constraint and enabler for corporate behaviors, and, more specifically, for CSA. Norms regarding appropriate corporate behaviors—which evolve at a macro-level—act as mediators between economic constraints and engagement in CSA (Campbell 2007). What firms consider acceptable and beneficial in corporate behaviors is inspired by business-relevant subsets of societal-level ideologies (Galvin et al. 2004). Ultimately, the likelihood to greenwash, or in other words to design CSA to maximize legitimacy gains at the expense of societal benefits, is influenced by national contexts and shared cultural beliefs.

Among them, cultural beliefs regarding the efficiency of economic systems are crucial. When the cursor moves between believing in the virtues of “liberal” or “coordinated” market economies, economic actors’ assumptions regarding the purpose of a firm—and thus its course of action—vary (Hall and Soskice 2001; Matten and Moon 2008). The ideology of liberalism makes the assumption that through competition firms will make optimal decisions to maximize the wealth of their shareholders (Hay 2004), even when those actions include social actions. A coordinated-markets perspective would by contrast claim that in targeting wealth maximization for shareholders, firms may induce negative externalities on society, and that government should regulate the firms’ actions to limit those negative externalities (Hall and Soskice 2001). As a consequence, country-level beliefs in the virtues of liberalism have been proven to influence the nature of corporate social behaviors (Hemingway and Maclagan 2004; Matten and Moon 2008).

We aim at studying two foundational beliefs in the virtues of liberalism, and their impact on corporate social behaviors. It appears first that liberal economic policies (e.g., deregulation, low-trade barriers, openness to globalization, etc.) are based on the idea that competition has positive societal consequences (Campbell 2007). It appears secondly that liberal social policies (e.g., private social security or educational systems) rest on the belief that individuals should be responsible for themselves, rather than being cared for by a government (Matten and Moon 2008). We propose that beliefs in favor of competition and individual responsibility might be seen as two founding beliefs in the virtues of liberalism (Hay 2004; Whitman 1999; Fligstein 2001). In turn, the degree to which these beliefs permeate economic actors’ cultural environment affect how CSA is approached.

Qualitative Exploration

How are decision-makers in firms influenced by the dominant beliefs regarding the economic system? How does this translate into organizational decision-making when it comes to CSA? What are the most contentious points regarding economic liberalism and the role of business in society? In this first part of the study, we explore those questions qualitatively as a first step toward building our understanding of the relationship between beliefs in the virtues of liberalism and corporate decision-making.

We use a focus group approach for this qualitative exploration, and completed this data collection with follow-up interviews with the participants. Focus groups were originally used to collect feedback on television programs but since then, they have been used in anthropology (Agar and MacDonald 1995), in medical research (Kitzinger 1995), and more recently in business ethics (Vyakarnam et al. 1997; Freestone and Mitchell 2004). Focus groups rely on participants’ interaction around the topic under investigation to generate relevant data by observing the kinds of reactions and exchanges that emerge from this communication process (Morgan 1988; Vyakarnam et al. 1997). We build on a phenomenological approach as we focus on explicitly eliciting individuals’ experience and stories (Gill, forthcoming). Participants are encouraged to reveal their perspective on a topic, which is particularly adapted for preliminary exploration of a research question (Morgan 1988). Focus groups aim to get round participants’ reluctance to be interviewed when they feel they have nothing to say (Kitzinger 1995) and steer clear of any contamination by the researcher (Vyakarnam et al. 1997). Group dynamics “help people to explore and clarify their views in ways that would be less easily accessible in a one to one interview” (Kitzinger 1995, p. 299) and gives the researcher the opportunity to identify agreement or dissension around the topics of interest. The drawbacks with focus groups are that they can be dominated by stronger personalities (Jenkins and Harrison 1990) leading to biases in the discussion, and that the focus group setting can be seen as “too artificial” (Vyakarnam et al. 1997, p. 1629). As a consequence we followed up with individual conversations with some of the participants to get a better sense of what they wanted to express. These interviews are particularly useful for the participants that were less involved in the focus group.

We use a theoretical sampling model (Kitzinger 1995). Our nine chosen participants were selected to maximize diversity in terms of country of origin, as our main focus of interest is around country-level values. We are confident that our nine participants comprise a sufficiently diverse panel, without being too large to preclude sufficient interaction between them. The participants are established businesspersons based in 8 distinct countries and hold high-level responsibilities in medium-sized business from various industries (see Table 1 for a summary of their characteristics). The common criteria of our participants is that they have been involved in CSA for their firm or for major clients, including pro-bono work for non-profit organizations, initiatives to reduce the environmental footprint, philanthropic engagement, fair trade business, or community support. The participants had not met each other prior to the focus group, but all were known to the mediator and researcher who invited them to participate in this study. Considering the difficulties of bringing together nine participants from distant part of the world, the focus group and the follow-up interviews were conducted through video conferencing. The focus group lasted for around 60 min, and the follow-up interviews 20–35 min.

Table 1 Characteristics of the focus group participants
Full size table

Although we were interested in the beliefs in the virtues of economic liberalism (i.e., what judgment is expressed regarding the value of economic liberalism), we wanted participants to express their own views on liberalism. As a consequence, the topic of the discussion was framed around liberalism and the role of business in society. The participants were unaware of this articulation and were progressively steered to debate this relationship. The design of our focus group was inspired by Vyakarnam et al. (1997), whereby the mediator is given the role of facilitator and ensures that all participants have the chance to express their view—this way, participants are free to formulate their own questions using their own words, and forms of expression (anecdotes, jokes, arguments, etc.) (Kitzinger 1995). The session was introduced with questions regarding liberalism, then the mediator allowed the debate to unfold between the participants, progressively leading them to a discussion around the relationship between firms and liberalism, and the role of business in society. The coding of the data relied on common themes and patterns in the discussion. Arguments in favor of and against economic liberalism were distinguished and highlighted. In our analysis, we use the group dynamics to identify consensus opinion (consensus was almost never reached in the debate) and “deviant” points of views, when individuals express a different perspective on the debated topic.

Two themes are at the center of our analysis and each of them reflects one of the two perspectives of participants, who are at the same time business managers and citizens. These themes can be broadly captured by the two questions: How can economic liberalism be defined from a business perspective? And what is the role of business in a liberal society? “Freedom of enterprise” and a “limited role of the government” are clearly associated with economic liberalism by the majority of the participants. As stressed by one of them, these elements are connected to low fiscal pressure: a low level of taxes is associated with greater freedom for private firms. The impact of governments and institutions on business was a central concern. Some participants recognized that “hybrid systems” have emerged as the principles of economic liberalism are applied in ”different scope and ways depending on local perception.” This way “political power can still have a say on economic concerns.” Participants also stress the importance of the “rule of law” and the fact that a liberal society is supposed to ensure “everybody [gets] a chance,” although some participants attribute this responsibility to the state, or present it as a consequence rather than a necessary condition for a liberal society.

A clear area of interest but also dissension between participants was whether economic liberalism is an ”internally coherent system”. While some participants defend this view, others point out the failure of economic liberalism to fulfill the mission of offering opportunities to the widest number “by over-favoring those that are already at an advantage” while “some effects might contradict each other.” Participants discussed the example of monopolies: do they emerge as a consequence of economic liberalism, because the state is divorced from the economic system? If yes, some contributors explain that the system contradicts itself, as it is supposed ensure fair competition. Although liberty is perceived as the key foundational value of economic liberalism, it relies on the assumption that liberty can express itself and that the situation is fair for everybody. The role of the government must remain limited but it must still “ensure that everybody’s get the same chance,” which is stated as “paradoxical” by participants.

How did participants view the role of business in society in relation to their perspective on economic liberalism? Participants were mainly in favor of a limited impact of government on business. On the other hand, the role of business in society was less clearly seen. When defining the role of the government in a liberal society, it was unclear for participants whether the state should or could ensure a principle of fair competition between social actors, which was perceived as essential to economic liberalism. For some participants “there is a clear cut” between the role of governments and the role of businesses, because liberalism relies on a delegation of power to the government. This implies that the role of business in society is expected to be limited in a liberal society, as their primary focus is to “work for themselves”. This in line with previous research implying that entrepreneurs in developing economies are focused on short-term profitability and survival in contexts of strong competition, and weak governmental institutions (Touboul and Roulet 2011; Julian and Ofori-Dankwa 2013). However, this view was nuanced by some participants, whose arguments unfold from the internal contradictions associated with economic liberalism. Because of weak governments (“no money, no legitimacy”), some businesses end up having more power than the state. One example is the case of multinationals being involved in the development of infrastructures, where firms end up substituting for governments. The paradox relies on the fact “the smaller the government, the more other people have to step in” as one participant pointed out. Is extreme liberalism to be confused with anarchy? One suggestion is to distinguish things for which firms are “really responsible” such as their own pollution, and, on the other hand, things they engage in which are not necessarily part of their duty. One participant used the example of Henry Ford doubling the wage of his employees so that they would be more likely to buy the Ford cars (the idea that “what you give, you get it back somehow”). CSAs are not necessarily as “genuine,” but rather can be used more cynically, for example as a “marketing tool.”

Table 2 provides a collection of quotes to illustrate the main insights drawn from this qualitative exploration. Economic liberalism is mainly associated with a limited role of government, in conjunction with fair competition between social actors. However, there is an inherent contradiction in the limited role the government is expected to play in a liberal society and the fact that economic liberalism is associated with fair competition. In other words, can government both be small and ensure that economic competition is equitable? For some participants, businesses can participate in rebalancing the system and substitute for the government when required. As suggested by Kinderman (2012), CSR and liberalism are not necessarily antithetic.

Table 2 Data exemplars on the three main topics of dissension
Full size table

Hypotheses Development

Building on these first insights, we set out to determine how country-level beliefs in the virtues of liberalism might influence the likelihood of firms to greenwash. The findings of our qualitative exploration confirm that economic liberalism is associated both with limited government and fair competition. It also reveals the ambiguity surrounding the role of businesses in liberal societies: a limited government implies that other stakeholders need to step into ensure the founding principles of liberalism are respected.

We look first at the literature on the established link between actual economic systems and corporate actions. Some of this literature adopts what we might call a conventional view. As an example, Lim and Tsutsui (2012) use Prechel and Morris’ (2010) argument that a liberal economic system leads firms to focus solely on profit-maximizing activities, and to “gain legitimacy as social actors by working with policymakers to create voluntary CSA frameworks” that are created mostly in name only in developed countries (Lim and Tsutsui 2012, p. 79). In a context where firms are pressurized to focus on profitability, firms are more likely to frame their CSA to maximize financial gain (Julian and Ofori-Dankwa 2013), and thus to greenwash. But some scholars adopt an alternative view on the relationship between economic liberalism and corporate behaviors. Jackson and Apostolakou (2010) make the opposite claim to the conventional view. They show that in liberal market economies, due to the absence of government protection, stakeholders are more likely to pressurize firms into social engagement. In turn, firms answer these pressures positively with substantive CSA, which in this context becomes a way to differentiate themselves from their competitors. Kinderman (2012) makes a similar argument by explaining that CSA substitutes for institutionalized social solidarity, and that CSR and neoliberalism have co-evolved and complemented each other. In these contexts, there is more reliance on voluntary action (Kang and Moon 2012). We call this alternative perspective the “pro-active” view because it assumes the existence of a form of corporate voluntarism.

Institutional perspectives oscillate between looking at CSA as “voluntary engagement” or as “binding responsibility” (Brammer et al. 2012, p. 3). These two perspectives would imply conflicting results when it comes to the likelihood to greenwash. They in fact only differ in the assumption they make regarding the behaviors of economic actors. The conventional view implies that firms behave selfishly. In more liberal contexts, they opt for symbolic CSA that aims at gaining the stakeholders’ support without bearing the cost of substantive actions, thereby greenwashing. On the other hand, the pro-active view assumes that firms are willing to step into contribute to the coherence of the institutional system. Therefore, when the government is not thought to be responsible for stakeholders’ welfare, firms consider that it is their responsibility to cope with their stakeholders’ social issues. Because beliefs in favor of competition and individual responsibility are micro-foundations of economic liberalism, we develop hereafter two sets of hypotheses that follow these two perspectives (conventional and pro-active).

The Conventional Perspective

From the traditional perspective, what can we infer regarding the relationship between country-wide beliefs in liberalism and the implementation of symbolic rather than substantive CSA? Lim and Tsutsui (2012) found that in developed countries, the more liberal the economic system is, the more the adoption of CSA will be gestural. They argue that in liberal contexts in developed countries “major corporations in rich countries use voluntary CSR frameworks to deflect criticisms and to circumvent stringent regulations” (Lim and Tsutsui 2012, p. 87). Julian and Ofori-Dankwa’s (2013) argument is in line with this perspective as they suggest that weakly institutionalized systems offer a lot more freedom to firms and make them less likely to invest in CSA. Our qualitative exploration suggests that liberalism is often associated with limited government and more reliance on individual responsibility. Participants stressed a clear separation of roles and responsibilities between business and government, which suggests that firms, including in their use of CSA, should focus on building up a competitive advantage.

Campbell (2007) argues more specifically that businesses are less likely to engage in substantive CSA if there is too much competition. If competition is too high, firms will cut expenses that apparently fall outside direct profit maximization, and thus focus on impression management technique and favor symbolic over substantive CSA. This view implies that firms adopt opportunistic approaches, and, based on this assumption, in a context in which people believe that competition is beneficial, firms would be expected to elbow their way through. Thus, beliefs in the virtue of competition are more likely to trigger the design of CSAs that build a competitive advantage, rather than to focus on social and environmental impact. Therefore, when they are embedded in cultural beliefs in favor of competition, firms tend to greenwash, or in other words try to generate the benefits of stakeholders’ support with limited costs. Our first hypothesis is, therefore:

Hypothesis 1a

When cultural beliefs are in favor of competition, firms are more likely to greenwash.

We next investigate the role of beliefs in favor of individual responsibility. Prechel and Morris (2010) have shown that in liberal contexts the presence of loopholes in institutional arrangements prompts individualistic actors to misbehave. This research suggests that individuals are inherently selfish. Consequently in context with strong beliefs in the virtues of individual responsibility, economic actors naturally act in their own interest at the expense of others. In this traditional view, substantive corporate social behaviors conflict with individual responsibility (Jo 2012). Kim and Kim (2010) confirm this perspective and show on a sample of public relations practitioners that individualism is negatively related to substantive CSA. They argue that collectivism, as opposed to individualism, promotes societal values, as the utility of the group is ranked above individual interests. However, when the utility of the group is below personal interests, individuals tend to neglect societal values. As a consequence, we can hypothesize that a country-level belief in the virtues of individualism will have adverse behavioral consequences on firms’ commitment to CSA, and thus will be more related to greenwashing. Our next hypothesis can be stated as follows:

Hypothesis 1b

When cultural beliefs are in favor of individual responsibility, firms are more likely to greenwash.

The Pro-Active View

In opposition to the conventional view, some scholars have shown that companies in fierce liberal environments are also likely to engage in sincere CSA (Matten and Moon 2008; Jackson and Apostolakou 2010) and even that liberalism co-evolved with the movement for CSR (Kinderman 2012). This argument is less present in the literature. In liberal market economies, there is a lack of any kind of institutionalized interaction with stakeholders. In this situation, firms—because they are “pro-active”—are less likely to use CSA as an impression management tool, as they fill the vacuum that weak government generates. Some elements of the focus group suggest an inherent contradiction between the idea of a small government and a system that ensures fair competition. This contradiction may imply a greater role for business to complement the role of the state in a liberal economy.

In coordinated market economies, regulations are stricter and competition is limited. In such a situation, interactions with the environment are subjected to institutionalized constraints and regulations. There is less scope for voluntary CSA (Brammer et al. 2012). However when competition is fierce, stakeholders exert a greater pressure on firms toward CSA, as no regulation is there to impose substantive actions on those firms. In this situation, engaging in CSA is a way for firms to differentiate themselves from their competitors. In this context, it implies that firms will be “pro-active”: they are less likely to greenwash as they will focus on substantive CSA, which not only differentiates their firm but impacts the welfare of stakeholders. In a context of strong beliefs in the virtue of competition, acting substantially in favor of stakeholders is a way to remain competitive. Therefore, in the presence of country-level cultural beliefs in the beneficial role of competition, stakeholders, by valuing the firms that differentiate themselves, have a tendency to pressure them toward substantive CSA, and thus, away from greenwashing. Our second set of hypotheses, therefore, begin thus:

Hypothesis 2a

When cultural beliefs are in favor of competition, firms are less likely to greenwash.

Matten and Moon (2008) argue that in liberal market economies, because the role of individual responsibility is prominent, firms are more likely to engage in “explicit” CSA (i.e., voluntary and concrete actions in favor of their stakeholders). Their argument is that when individual responsibility is favored instead of government responsibility, firms feel the pressure to tackle social issues since no government or institution will do it. In this view, Christie et al. (2003) found that in an individualistic society such as the United States, people are more sensitive to unethical behaviors than in a society with a protecting government. As for the UK, Jackson and Apostolakou (2010) note that the government’s disengagement from the mid-1980s to the end of the 1990s corresponded to a substantive increase in managers taking responsibility toward social and environmental issues. This evidence suggests that when a majority of the population supports a system in which responsibility is vested in individuals rather than in a government, substantive CSA increases. Firms fill the “institutional void” created by loose regulation by favoring a substantive impact on the welfare of their stakeholders, rather than tending toward greenwashing. Consequently, our final hypothesis is as follows:

Hypothesis 2b

When cultural beliefs are in favor of individual responsibility, firms are less likely to greenwash.

There are thus two perspectives on the impact of cultural beliefs in favor of competition and individual responsibility on the propensity of firms to greenwash. Our empirical analysis aims at validating, invalidating, or confounding these two perspectives.

Quantitative Testing

Data and Method

Data in this study were extracted from three different sources. To compute a measure of greenwashing, we first extracted evaluations of firms’ CSA from the Asset4 Database. Asset4 is a Swiss-based extra-financial rating agency subsidiary of Thomson Reuters, which provides several ratings related to firms’ CSA. Asset4 analysts compile their ratings by gathering raw data on firms’ CSA (e.g., number of employees’ health and safety issues at work, definition of a strategy to reduce CO2 emissions, signing of the United Nations Global Compact, etc.). They collect such information from all sources available in the public domain (Annual reports, social responsibility reports, Newspapers, NGO websites, etc.). These raw data are then transformed through a proprietary algorithm in several ratings and sub-ratings. Asset4 covers firms from 67 different countriesFootnote 1 belonging to major financial indexes worldwide.Footnote 2 By covering a set of 958 firms in 2002 to 2,920 firms in 2008, Asset4 was the world’s largest database on firms’ extra-financial information in the frame of our study. Asset4 ratings on firms’ CSA are all the more relevant to our study as Asset4 analysts not only examine firms’ degree of CSA, but also the detailed nature of these actions. More specifically, they differentiate firms’ substantive CSA (e.g., amount of CO2 emissions reduced in the past year, number of injuries and fatalities at work), from their symbolic CSA (e.g., does the firm claim to have a policy for reducing environmental emissions? Does it claim to strive to improve its employee health and safety?). Asset4 claims that its ratings are able to measure firms’ “talk” versus their “walk” about CSA. Following Walker and Wan (2012), we combine these two measures by computing the ratio of symbolic to substantive action as a way of capturing the extent to which firms greenwash.

We then completed our initial sample of firm-specific measures of the degree to which they engaged in greenwashing, with country-specific measures of cultural beliefs in the virtues of competition and individual responsibility. We are interested in the net effect of these two beliefs, everything else being equal, and how they are related to firms’ propensity to greenwash. These beliefs are related to each other (as they are both antecedents of economic liberalism) but not with other independent variables, considering they are at the firm level. The objective of this research is to disentangle the impact of these beliefs instead of establishing a conjunctional causation by looking at the effect of liberalism as a whole (with the conjunction of multiple antecedents). We ensure below that despite those measures being related, they are distinct enough so as not to contaminate our results, as we stress in the analysis section. We extracted such measures from the WVS. The WVS is a public opinion survey, which has been conducted since 1981 on a regular basis (every 3 or 4 years depending on the country) by a worldwide network of social scientists. In 2008, the WVS covered a set of 83 countriesFootnote 3—including countries whose data are less accessible. Data are collected through a rigorous process of interviewing—mostly face-to-face, except in remote areas—representative samples of national populations (on average 827 individuals are interviewed per wave and per country). Sociology scholars in particular have used the WVS to monitor and capture the evolution of values among populations (Inglehart and Baker 2000). Following Berry et al. (2010), we use interpolation to obtain a measure of cultural beliefs for the 83 countries over the 2002–2008 period, including for years between waves. For each year between 2002 and 2008 and for each of the 83 countries, we obtained a measure of a population’s cultural beliefs in the virtues of competition and individual responsibility. We matched this country-specific data with our initial dataset of firm-specific measures considering the firm’s country of incorporation. In this sense, we considered the most salient stakeholders for a firm to be those from its country of incorporation. Regarding the weight of the influence of the country of incorporation’s government on a firm’s operations, we formulate the reasonable assumption that this share is significant. This assumption also relies on a number of elements that confirm the primary influence of the country of incorporation. First, NGOs tend to target firms that they already know, and are thus incorporated in their own country. In addition, most firms are incorporated in the country where they were historically established, and thus where they make the highest share of their revenues.

We finally completed our sample with firm-specific financial measures from Bureau van Dijk’s ORBIS database to compute our control variables.

Our final sample is a panel that contains 10,232 observations from 2,621 firms over 7 years (2002–2008). It is an unbalanced dataset due to limited availability of some data related to firms’ CSA or countries’ cultural beliefs over the whole period. Our sample includes, on average, 1,462 observations per year and the 2,621 firms from 38 different countriesFootnote 4 appear on average 3.9 years in our dataset.

Dependent Variable

Distinguishing symbolic from substantive CSA is a well-documented challenge (see Hawn and Ioannou 2012). For each firm in its dataset, Asset4 first provides 18 ratings related to the firms’ degree of implementation of substantive CSA. Those 18 ratings reflect the different components of CSA (e.g., emission reductions, community impact, board structure, client loyalty) and are grouped in four general categories (Economic, Environmental, Social, and Governance responsibility).Footnote 5 More specifically, these 18 ratings measure the substantive CSA for each firm (implementation of care services for employees, amount of philanthropic donations, number of green facilities, etc.). To measure the firms’ Substantive CSA, we computed the average of those 18 ratings equally weighed per general category (Economic, Environmental, Social and Governance). We averaged the 18 ratings with equal weightings for each category so that categories with a higher number of ratings did not over-influence the final measure (as an example, the Social category contained seven ratings while the Governance category contained five sub-ratings). This measure of Substantive CSA increases when firms implement a higher degree of CSA, and with the same magnitude whether actions were implemented in terms of Economic, Environmental, Social or Governance responsibility. Second, Asset4 also provides 18 ratings relating to the degree of implementation of symbolic CSA. These 18 ratings cover the same 18 topics covered by substantive CSA ratings and are grouped in the same 4 general categories (Economic, Environmental, Social and Governance responsibility). Compared with substantive CSA ratings, symbolic CSA ratings measure symbolic CSA for each firm (e.g., does the firm claim to have a strategy to reduce its pollutant emissions? Does the firm claim to monitor its impact on local communities?). In the same way as we computed a measure of firms’ Substantive CSA, we compute a measure of firms’ Symbolic CSA by calculating the average of Asset4 18 symbolic CSA ratings equally weighted per general category.

Our measure of greenwashing is aimed at capturing whether firms favor symbolic instead of substantive CSA: we, therefore, computed the ratio of our measure of firms’ Symbolic CSA to our measure of firms’ Substantive CSA. Thus, our greenwashing ratio measures the propensity to implement symbolic instead of substantive CSA.

Independent Variables

As we wanted to examine how populations’ cultural beliefs in favor of competition and individual responsibility would affect greenwashing, we concentrated on these two specific variables produced by the WVS. We also concentrated solely on the answers from the 201,675 interviews that were conducted in the 38 countries covered by both the WVS and Asset4 during the 2002–2008 period.

The WVS first asks interviewees whether competition is good, and if it stimulates people to work hard and develop new ideas, or whether it is harmful, and brings out the worst in people. For each of the interviewees, the strength of their beliefs are coded on a scale from 1 to 10—with 1 = “Competition is harmful,” and 10 = “Competition is good” at opposite ends. As we were trying to compute a country-specific and not an individual measure of cultural beliefs in favor of competition, we computed the weighted average of interviewees’ beliefs in the benefits of competition per country. We thus obtained a measure of Cultural Beliefs in favor of Competition for each of our 38 countries. The individual weightings we deployed were provided by the WVS so that samples of individuals per country were statistically representative of their country’s population in terms of age, gender, or socio-professional category. As a result, our measure of Cultural Beliefs in favor of Competition is specific to each country, and increases when larger sections of this country’s population consider that “Competition is good” rather than “Competition is harmful.”

The WVS then asks interviewees whether people should take more responsibility to provide for themselves, or if the government should take more responsibility to ensure that everyone is provided for. Here, we coded each of the 201,675 respondents answers on a scale from 1 = “Government should take more responsibility” to 10 = “People should take more responsibility.” We then computed the weighted average per country of respondents’ answers to obtain one measure per country of its population’s Cultural Beliefs in favor of Individual Responsibility. Consequently, this measure is high when a country’s population considers that “People should take more responsibility,” and low when people consider that “Government should take more responsibility.” To provide an overview of our independent variables, we present in Fig. 1 below the average country-level Cultural Beliefs in favor of Competition and Cultural Beliefs in favor of Individual Responsibility for the 2002–2008 period.

Fig. 1
figure 1

Average country-level beliefs in favor of competition and individual responsibility, 2002–2008

Full size image

We can observe in Fig. 1 that countries with higher Cultural Beliefs in favor of Individual Responsibility are developed market economies such as Switzerland, the United States, New Zealand or Canada. Whereas these countries also score high in terms of Cultural Beliefs in favor of Competition, it appears that countries with higher scores on this variable are fast developing countries such as India, China, or Morocco.

Control Variables

Taking as our dependent variable the propensity of firms to engage in greenwashing, we controlled our models for major factors that have been shown in literature to influence firms’ CSA and greenwashing such as firms’ Size, Profitability, Risk Exposure, Industry, Country of origin or the Year of the observation (Walker and Wan 2012; Surroca et al. 2010; Waddock and Graves 1997). We operationalized firms’ Size by the logarithm of their revenues, Profitability by their return on assets (ROA), and Risk Exposure by the inverse of their solvency ratio. We controlled for unobserved Industry specific factors with a set of 24 dummies based on the first four digits of firms’ Thomson Reuters Business Classification code, and for Year, specific events with a set of 7 year dummies. Most importantly, we controlled for any country-specific factor other than populations’ Cultural Beliefs in favor of Competition or Cultural Beliefs in favor of Individual Responsibility, such as governmental and other institutional pressures, populations’ receptivity, or foreign trade relations (Lim and Tsutsui 2012), with a set of 38 country-specific dummies based on country of incorporation. We also run our models separately for developing and developed countries (using Lim and Tsutsui’s (2012) dichotomy of OECD vs non-OECD countries), while including the previously mentioned country-fixed effects.

Analysis

Finally, we standardizedFootnote 6 all our non-dummy variables to be able to compare their relative effect on firms’ propensity to favor substantive rather than symbolic CSA. In Table 3 below we provide summary statistics and list the Pearson correlations of our dependent and independent variables. Most correlations lie between −0.18 and 0.56, which suggests limited risks of multicollinearity issues. In addition, we computed the Variance Inflation Factor (VIF) of our models, which ranges from 1.45 to 2.69 and thus suppresses any remaining doubt about the reliability of the finding. We find, naturally, that the two cultural beliefs in the virtues of liberalism are 56 % correlated, considering they are antecedents of liberalism at a more macro-level.

Table 3 Pairwise correlation between dependent and independent variables
Full size table

Our final models are random effects panel models with standardized variables, which were estimated with a Generalized Least Squares (GLS) method and heteroskedasticity-robust estimations of coefficients. We estimated random instead of fixed effects models. Fixed effects models did not allow us to fully control for unobserved country-specific effects on firms’ corporate behaviors, whereas the literature on the topic show those effects to highly impact the nature of firms’ CSA (Lim and Tsutsui 2012). We tested for the relevance of a random effects model with a Breusch-Pagan Lagrange multiplier test, which validated our choice.

Results

Table 4 presents 6 models we used to test how country-specific Cultural Beliefs in favor of Competition and Cultural Beliefs in favor of Individual Responsibility impact firms’ propensity to greenwash.

Table 4 Random effects regressions of beliefs in favor of competition and individual responsibility on firms’ propensity to greenwash
Full size table

Model 1 only includes the control variable and provides us with a base R 2 of 23.2 %. Model 1 shows that bigger, less profitable, and more risk-taking firms tend to greenwash.

Model 2 includes country-level Cultural Beliefs in favor of Competition as an independent variable and estimates a positive (0.21) and significant (at 1 %) impact of Cultural Beliefs in favor of Competition at the country-level on firms’ propensity to engage in greenwashing. In this sense, Model 2 validates Hypothesis 1a and supports an hypothesis built on the traditional view of liberalism and corporate behaviors: when there are country-level Cultural Beliefs in favor of Competition, firms tend to favor symbolic instead of substantive CSA. For instance, in a country with high Cultural Beliefs in favor of Competition such as India, firms will be more likely to greenwash.

Model 3 includes control variables and country-wide Cultural Beliefs in favor of Individual Responsibility. It shows that Cultural Beliefs in favor of Individual Responsibility significantly (0.1 %) and negatively (−0.14) impact firms’ propensity to greenwash. Model 3 provides support for Hypothesis 2b and validates a selective part of the pro-active view on the relationship between the beliefs in the virtues of liberalism and greenwashing. For example, in a country like Canada where the notion of individual responsibility is strong, firms tend to invest in the implementation rather than the signaling of CSAs. In other words, they are less likely to greenwash.

Model 4 includes all of our independent variables. It highlights no significant changes in sign or value of the estimated coefficients for our Cultural Beliefs in favor of Competition and Cultural Beliefs in favor of Individual Responsibility variables. In this sense, Model 4 provides further support for Hypotheses 1a and 2b. It also confirms that our estimates are not affected by multicollinearity. Model 4 confirms that firms will be more likely to greenwash when populations’ beliefs in individual responsibility are predominant, and when their beliefs in the virtue of competition are less prominent. Therefore, in a country like Morocco, where beliefs in the virtue of individual responsibility are low, but in the virtue of competition are high, firms are more likely to greenwash. Conversely, in a country like France, where the population believes in the virtue of individual responsibility but prefers an absence of competition, firms are less likely to greenwash as they tend to effectively implement CSA, without specifically signaling those actions.

Models 1–6 include country-specific effects: in other words, we control for omitted variables that might be related to the country of origin, for example the level of development. To better understand the dynamics behind the impact of economic development, Models 5 and 6 split the samples by developed and developing countries, while also keeping country-level fixed effects in the model. Model 5 exhibits similar results to those obtained on the larger sample. We have far fewer observations for developing countries (only 220 firms and 493 observations) and thus, the results of Model 6 have to be taken with caution. When we limit our sample to developing countries, beliefs in favor of competition are negatively related to greenwashing. This suggests that firms in developing countries, when they do invest in CSA, will tend to focus in actual implementation. This is in line with Lim and Tsutsui’s (2012) findings, that “faced with pressures to signal commitment to CSR, corporations in developing countries are more likely to make serious efforts” (Lim and Tsutsui 2012, p. 88). This is also explicable by the fact that there are probably fewer legitimacy gains to be made in developing countries as suggested by Julian and Ofori-Dankwa (2013)—stakeholders are less sensitized to social issues and/or have less impact on business, and thus CSAs are designed to maximize social or environmental impact rather than convey a positive image of the firm.

It is interesting to note that as we initially standardized our variables, it is now possible to investigate the relative impact of each independent variable on the probability of firms’ greenwashing. Model 4 highlights that the positive impact of Cultural Beliefs in favor of Competition (0.37) is higher than the negative impact of Cultural Beliefs in favor of Individual Responsibility (−0.21). If we assume that believing in the virtues of liberalism means believing equally in the beneficial role of competition and in the prominence of individual responsibility, then this finding underlines that even if country-level beliefs in individualism push firms toward less greenwashing, in general, cultural beliefs in the virtues of liberalism favor firms’ greenwashing. Finally, it is also interesting to note that the impact of country-level Cultural Beliefs (0.37 and −0.21) is similar in magnitude to firm-level variables such as firms’ Size or Profitability (−0.23 and 0.24). This finding shows that firms adjust their behaviors in terms of CSA not only depending on their own characteristics, but also depending on the normative context of the surrounding populations’ beliefs.

Discussion

Building on the existing literature, we initially assumed that beliefs in favor of competition and individual responsibility would affect corporate behaviors in the same way. Scholars considering that firms are naturally prone, as profit driven social actors, to selfish decision-making oppose those who think that they can also act pro-actively in a context where the interaction with stakeholders is weakly institutionalized. Our results show a more complex relationship between distinct beliefs in the virtues of liberalism and the nature of CSA. We show that, paradoxically, beliefs in favor of individual responsibility tend to push firms toward less greenwashing and with a true aspiration to do good, while beliefs in favor of competition lead managers toward egoistic greenwashing strategies based on a gestural commitment only.

Although our study sheds light on several assumptions and mechanisms pertaining to the relationship between beliefs in the virtues of liberalism and the nature of CSA, we recognize that certain limitations exist. We focused, for example, on two beliefs which lie at the heart of the liberal creed—competition and individual responsibility—and in so-doing have no doubt omitted some other related convictions. We made this decision for several reasons. First, the other founding values of liberalism are derived from the two beliefs we have focused on. As an example, confidence in free markets rather than in centralized and organized exchanges (Hay 2004) relies on the idea that individuals rather than governments should be responsible, and that they should be able to compete freely. Second, the WVS from which we draw our measures of cultural beliefs in favor of competition and individual responsibility offered us a unique and powerful way to measure how central these values were in a broad variety of national contexts. Items used are unequivocal and enable us to compare cross-country attitudes in a consistent manner. This would not have necessarily been possible with a more complex set of beliefs. We nevertheless recognize that future research could take an interest in the impact on greenwashing of other liberal values than competition and individualism, such as the importance given to family, the social role of religion, tolerance, or compassion.

As stressed by Kinderman (2012, p. 50), “neo-liberal sympathies do not rule out strong moral convictions and CSR engagement.” Firms, as social agents, may perceive other stakeholders as laborious, self-interested entities, but still feel the urge to engage in goodwill gestures to protect the individual rights of those stakeholders, the right to be autonomous and independent. These stakeholders might be the firm’s clients, business partners, and might play a crucial role in the future. In return, these managers expect stakeholders to act in a similar way. We, therefore, argue that substantive CSA is a way for firms to step in, protect, and support weaker actors when there is nobody else to do so. Consequently, when firms are embedded in a culture of strong beliefs in the virtue of individual responsibility but with a feeble government, they are relied upon in the absence of stronger collective forces for the protection of stakeholders’ and individual rights. This finding also emerged in the focus group we organized. When governments are weak (a factor implied by economic liberalism), businesses may have more power to ensure fair competition. This view is coherent with the results of the pro-active perspective on liberalism and greenwashing, revealing how pro-individualistic societies make social actors more sensitive to the treatment of others (Christie et al. 2003). Our results show that firms with a competitive mindset might still engage in greenwashing to construct a competitive advantage. Respecting others does not mean that the quest for survival should be brushed aside altogether. When beliefs in the virtue of competition are prominent, firms tend to greenwash and use CSA as a strategic tool to manage their image.

Our results show that beliefs in the beneficial role of competition is strongly related to firms’ greenwashing, and it more than compensates for the positive effect of beliefs in favor of individual responsibility. This finding suggests that even if firms are benevolent, social welfare cannot fully rely on their behavior, as predatory moves still prevail in liberal contexts.