Stakeholder Approach: What Effects Should We Take into Account in Contemporary Societies?
In recent years, the stakeholder approach has been widely applied in the debate on corporate social responsibility (CSR). Although many authors of this approach have reviewed many elements of the model, they have unconditionally accepted several criteria assumed by Freeman (1984) to identify stakeholders. In general, stakeholder authors have assumed that (a) the company establishes dyadic relationships with other agents, and (b) decisions made by a company only have foreseen and direct effects on other agents. These criteria have enabled researchers to understand simple processes. However, they have also prevented researchers from explaining how action comes about, and how responsibility is shared, in many complex processes taking place in contemporary societies. Such complex processes involve many agents, and each decision can generate unexpected effects which accumulate or disseminate. Furthermore, the normative structure governing these processes can affect and/or be affected by the actions of agents. In this study, we propose new criteria to expand the stakeholder model and facilitate the study of CSR in such processes.
KeywordsStakeholder approachEffectsRelationshipsCorporate responsibilityContemporary societies
The term corporate social responsibility (CSR) first appeared within the context of business studies in the mid 1960s. Subsequently, the notion of responsibility, traditionally attached to individuals, started to be widely applied to companies (Epstein 1989; French 1984; Soares 2003). In the field of CSR, authors have debated over questions such as what actions should a company be responsible for? On what grounds is this responsibility justified? What agents should be taken into account when making decisions? What objectives should guide the company’s decision-making? In recent years, the stakeholder approach has been at the center of all these debates (for an extensive review, see Laplume et al. 2008). However, a critical study on all the effects that are included and omitted in this approach is yet to be done.
Freeman (1984, p. 46) defines stakeholders as “any group or individual who can affect or is affected by the achievement of the firm’s objectives.” In this definition, the company’s stakeholders are described in very broad terms, with the clear intention of incorporating their interests into the firm’s decisions (Freeman 2005, p. 425). However, in his visual and theoretical model, Freeman (1984) incorporates criteria which are considerably more specific to identify stakeholders. In Freeman’s model (a) all stakeholders are linked to the company through dyadic relationships, and (b) the decisions made by an agent (the focal firm or any of its stakeholders) produce only foreseen and direct effects on other agents. This way of identifying stakeholders has made the application of the model more convenient for managers and academics, but it has also proven to be much narrower than Freeman’s original definition.
Nowadays, companies have simple relationships with other agents. Such is the case of the relationship binding a firm with shareholders who only expect to obtain financial returns, or with a consultant hired by the company to impart technical training to its employees. However, many firms also participate in much more complex interaction processes. A number of serious problems currently affecting many countries (such as environmental degradation, distrust of the financial system, human rights abuse, over-exposure of children to violent images, or the diffusion of an unreal image of physical beauty through advertising) involve numerous agents (companies, governments, NGOs, natural resources, local communities, the media, etc.) engaged in complex processes. In many of these processes, the ways in which effects unfold do not meet the narrow criteria assumed by most stakeholder models. Although we value the stakeholder approach as a good lens to analyze the firm’s responsibility, we deem it necessary to broaden the scope of relationships and effects to be considered.
In this study, we review the image of relationships and effects traditionally accepted by the stakeholder approach, and then propose alternative images which may help stakeholder authors drive CSR research forward. From these proposals, stakeholder authors will be able to study how action comes about, and how responsibility is shared, in many complex processes currently developing. For this purpose, we will first review Freeman’s proposal (1984) for identifying stakeholders. Second, we will analyze the characteristics of many contemporary processes which have made his proposal unsuitable. Third, we will broaden the range of the stakeholders considered, through the study of several images of relationships and effects. Finally, we will highlight a number of potential implications of these ideas in the future development of research.
Who Were Stakeholders Initially?
In his seminal study, Strategic Management: A Stakeholder Approach, Freeman (1984) presents his model as a tool for managers to analyze the environment and plan strategies, thus avoiding problems with other agents. The model would also facilitate access to resources and consequently improve the performance of the company. According to Freeman, changes in the business environment during the 1980s made the managerial models available at the time obsolete. These changes included foreign competition, increasing takeovers, activism, new industrial relations, a worldwide resource market, government reform, supranational agencies, a rising consumer movement, increasing environmental concerns, and changes in communication technology. These developments had affected the relationships between companies and their internal stakeholders (owners, customers, employees, and suppliers) and their external stakeholders (governments, competitors, consumer advocates, environmentalists, special interest groups, local community organizations, and the media). Consequently, American managers of the 1980s had to consider “all of those groups and individuals that can affect, or are affected by, the accomplishment of organizational purpose” (p. 25).
As many authors have pointed out, this means of identifying stakeholders has made it possible to develop two branches of the literature: strategic and normative (Donaldson and Preston 1995; Goodpaster 1991). Models included in the strategic branch have focused on agents which may affect the firm’s objectives (Freeman 2000; Moore 2001; Waxenberger and Spence 2003), while models in the normative branch have focused on agents who may be affected by the firm’s decisions (Bowie 1991; Evan and Freeman 1988). Researchers of both branches have raised a number of issues, such as which normative basis should be adopted for this approach? (Beekun and Badawi 2005; Burton and Dunn 1996; Etzioni 1998; Lea 2004; Wijnberg 2000), which agents are, or should be, given preferential treatment by the company? (Clarkson 1995; Cragg and Greenbaum 2002; Driscoll and Starik 2004; Mitchell et al. 1997; Phillips et al. 2003), how should the company balance the interests of its stakeholders? (Jensen 2002; Lampe 2001; Reynolds et al. 2006; Schwarzkopf 2006), or what are the potential consequences for the company who implements a stakeholder policy? (Coleman 2011; Harrison and St. John 1996; Heugens et al.2002; Hillman and Keim 2001; Roome and Wijen 2005). These debates have propelled our understanding of strategy and CSR, but there is a key question that remains unanswered: what should the expressions “can affect” and “can be affected” mean in contemporary societies? In other words, what kind of social relationships, and what kind of effects ought to be considered nowadays when identifying stakeholders? The analysis of this question is the focus of this study.
In recent years, this model has been thoroughly revised. Some authors have added or removed circles surrounding the company (Frooman 1999; Mitchell et al. 1997; Orts and Strudler 2002; Phillips 2003). Others have asserted that not all stakeholders have the same influence or represent the same risks for the company (Agle et al. 1999; Eesley and Lenox 2006; Mitchell et al. 1997; Pajunen 2006). Other authors have identified subgroups within each stakeholder category (Fassin 2009; Ryan and Schneider 2003). Others have argued that agents may belong to more than one group of stakeholders at the same time (Jansson 2005; Pesqueux and Damak-Ayadi 2005). Others have placed stakeholders in different levels, depending on their distance from the focal firm (Post et al. 2002), etc. Although all these studies have shown the openness and flexibility of this approach, they have unquestioningly accepted Freeman’s criteria when it comes to identifying stakeholders. For decades, widespread acceptance of these criteria has enabled authors to observe many simple processes closely. However, this same acceptance has also become an obstacle when it comes to explaining, or even being aware of some complex processes which are frequent in current societies. In the “Interaction Processes in Contemporary Societies” section, we will review distinctive features of such complex processes.
Interaction Processes in Contemporary Societies
Important developments have taken place in the most advanced societies over the last decades. Numerous authors have highlighted the significance of phenomena, such as increased competition, the globalization of markets, the acceleration in technological development, convergence across industries, the relevance of knowledge within companies, the search for more flexible organizational structures, changing preferences among consumers, environmental concerns, or cultural diversity within states (Eisenhardt and Martin 2000; Prahalad and Hamel 1994; Rowley 1997; Scherer and Palazzo 2007; Shankar and Bayus 2003). All these changes have driven developed societies into a new era which has been designated and described in different ways, depending on the feature highlighted by each author. Thus, we can frequently find this new era described as postindustrial society (Bell 1973), global economy (Levitt 1983), reflexive modernity (Giddens 1991), second modernity (Beck 1992), knowledge society (Drucker 1993), digital economy (Tapscott 1996), network society (Castells 2000), postnational society (Habermas 2001), etc.
For many countries, these phenomena have generated a social, economic, and political reality that bears little resemblance to their situation 40 years ago. However, some authors have assumed that such phenomena affect all societies equally (see, critically, Bettis and Hitt 1995; Castells 2000; Sampler 1998; Scherer and Palazzo 2007, 2011; Stehr 1994). Still immersed in this process of change, we lack the necessary distance to rigorously evaluate the magnitude and meaning of all these phenomena, but it already seems to be clear that their impacts vary widely in different countries, and even within the social and economic structure of each country (De Lombaerde and Iapadre 2008; Douglas and Wind 1987; Løwendahl and Revang 1998).
Numerous agents participate.
The decision of one agent affects other agents whose responses may trigger a long chain of effects which accumulate or spread out in unpredictable ways.
There are no formal coordination mechanisms.
The decision of one agent may affect and/or be affected by the social norms and values accepted by the agents who participate in the process.
Not all interaction processes develop in the same way in contemporary societies.
The image of interaction assumed by a model conditions the scope of social relationships considered by authors and managers.
New Insights into the Meaning of “Can Affect” and “Is Affected” in Contemporary Societies
Let us imagine Company A is engaged in producing paper in a developed country. The company produces toxic waste arising from the manufacturing process, and dumps it in a nearby river (Action 1), extracts water from a subterranean aquifer to supply to its factory (Action 2), or buys a consignment of wood pulp from Distributor B, located in the same country (Action 3). The managers of Company A are well aware of the negative (however minimal) effects of their decisions to dump waste in the river or to extract water from the aquifer. However, they have very little idea of the potential final effects of Action 3. Distributor B usually buys wood pulp from Exporters C and D, who are located in a developing country. These, in turn, buy pulp from local Companies E, F, G, and H, who illegally log in protected areas suffering from progressive deforestation, thus harming local natural resources as well as the people who live in the area. These sorts of actions link the company to other agents and may trigger a range of effects. In the following subsections, we analyze the different types of effects and to what degree they have been recognized in the stakeholder literature.
Effects on Other Agents
Human resources and production policies applied by companies have direct effects on their employees. These policies determine their salary, work and recess hours, their level of effort and stress, etc. Besides, each company offers certain products, or services, which have direct effects on the needs and satisfaction of their customers. These dyadic relationships between the firm and its employees and clients were already present in Freeman’s original model (1984) and have been well underlined by stakeholder authors. In reality, numerous companies have shown great awareness of these effects and tried to improve them, by including their commitments in their ethical codes (Adams et al. 2001; Kaptein 2004). This is the case of companies, such as H&M, Barklays, or IKEA, ethical codes of which comprise a commitment to maintaining high quality and safety standards in their relationships with employees and customers. The visibility of such effects has made other companies (for example, McDonald’s) to be strongly criticized for the conditions under which it hires its employees as well as the potential negative effects of its products on the clients’ health. However, these policies frequently bring significant indirect effects as well. The working conditions set by the company can have serious effects on the employee’s attention to his/her family, while the consumption of harmful products generates additional burden on the health system.
In reality, it is easy to relate these effects to the over-exploitation of natural resources. For instance, the intensive extracting activity developed by oil, timber, and mining companies in the island of Borneo has destroyed extensive areas of tropical forest in a short period of time (Sears et al. 2001). A similar case can be found at Las Tablas de Daimiel, a Spanish wetland of great environmental value, where the water and biodiversity levels have dramatically dropped because of continued extraction of water by local farmers (Álvarez-Cobelas et al. 2001). However, cumulative effects are not always negative, and may affect other agents than natural resources. In their study of multinational corporations (MNCs) in developing areas, Bird and Smucker (2007) asserted that some MNCs interpret their direct investments as another element of their cost-minimization policies, while other companies contribute to the development of local actives with their foreign investments. According to these authors, the investment of certain companies in China has helped improve local workers’ skills, as well as accelerating the development of infrastructure, ameliorating the credit system, and strengthening the local structure of suppliers and trade.
In recent years, many companies have outsourced part of their activities to lighten their structure and minimize their costs. These policies are common among industrial companies (such as Ford or IBM) and service companies (such as Vodafone or ING). As a result, the map of relationships binding a firm to other agents has become more complex. In the fashion industry, companies have applied such policies in different ways. H&M manufactures the entirety of its products through external suppliers. On the other hand, Inditex completes part of the production in its own factories and subcontracts the rest to external suppliers, some of which are located close to their own factories (in Arteixo, Spain) and others abroad (Morocco and Asia). This model has enabled Inditex to reduce its costs while maintaining a certain ability to change their product decisions rapidly. According to the image proposed in Fig. 5, Inditex’s manufacturing policy may have a direct effect on its own employees (direct effect 1) and on its external suppliers (direct effect 2), as well as an indirect effect on the employees of those external suppliers (indirect effect 1). If one of the latter were to subcontract to another supplier, then Inditex’s decisions could affect that new supplier (indirect effect 2) and its employees (indirect effect 3).
As illustrated in Fig. 6, the disseminative effects may come back to the focal firm. Fashion companies GAP (November, 2007) and Inditex (June, 2008) were respectively accused of manufacturing their products in developing countries making use of child labor and abusive working conditions. Interestingly, the workers had not been hired by the external suppliers of these MNCs, but by other companies outsourced by those suppliers. Despite the causal distance between the MNCs’ decisions and the effects on the workers, the accusations against GAP and Inditex had an immediate influence on the public opinion (feedback effect 1), negatively affecting the reputation of both companies (feedback effect 2). Their managers reacted rapidly canceling all contracts with these suppliers and introducing explicit conditions in the ethical codes of their companies to limit their suppliers’ ability to subcontract.
Disseminative effects have become more evident in companies which have decentralized their decisions. However, their impact on the reputation of many MNCs has put their managers in alert and drawn the attention of many authors toward this kind of companies. Although not using our terminology, numerous researchers in this field have analyzed questions such as: How should responsibility be shared out in MNCs’ supply chains (Amaeshi et al. 2008)? what mechanisms can MNCs apply to control these effects (Detomasi 2007; Krueger 2008)? what factors influence the way in which MNCs assume their responsibilities (Özen and Küskü 2009)? or what responsibilities toward local communities should be assumed by MNCs operating in developing countries (Bird and Smucker 2007; Idemudia 2009)?
Some stakeholder authors have included indirect effects in their models (Key 1999; Neville and Menguc 2006; Rowley 1997; Welcomer 2002). In their view, each stakeholder group can be the center of a new set of relationships with other agents. Thus, the relationships between a firm and its stakeholders and the latter’s relationships with other agents form an extensive network which needs to be taken into account by managers. These versions of the stakeholder approach have correctly emphasized the potential indirect (feedforward) effects, but are still far from acknowledging the wide range of relationships a company can potentially have.
Effects on the Normative Structure
Besides producing effects on other agents, a company’s actions may also affect and/or be affected by the structure of norms and values regulating the process. Depending on the specific features in each case, the isolated action of an agent can influence and/or be influenced by the content, and the binding force, of a socially established normative structure. In our hypothetical example, when Company A buys wood pulp from Distributor B without making any attempt to establish its origin, it is following, and nurturing, a common rule shared by many companies, which could be summarized as “out of sight, out of mind.” Subsequently, any company in the industry operating along these lines will feel legitimized by this socially accepted norm and, in turn, contribute to the continuity of that practice.
The social system of norms can be seen as the result of past actions of individual agents, but also as the normative base for their future behaviors (Fig. 7). Consequently, in addition to being affected by a set of socially accepted rules, the isolated actions of a company can contribute to establish, reinforce, undermine, or eliminate certain socially accepted rules governing actions and relationships.
Structural properties of social systems are both medium and outcomes of the practices and activities that comprise those systems. People in an environment behave in common ways to produce customs. Customs are organized into moral systems. Moral systems are institutionalized, and influence the way people learn to behave. Conforming behavior validates moral systems, and deviant behavior puts pressure on those systems to change in addition to providing negative comparison. When the pressure is sufficient, the systems are modified along with the conventional ideas that support them.
In the fashion industry, there is a group of well-known companies (including H&M, GAP, Inditex, Topshop or Mango) which see one another as similar. These firms interact frequently in national markets and share a series of norms which are well accepted at the industry level. In general terms, all these companies use similar strategies in relevant CSR areas such as environmental care, formulation of ethical codes or labor rights protection (Graafland 2002; Rudell 2006; Wolfe and Dickson 2002). However, these processes of mutual observation and imitation have also affected the canon of beauty these companies project on their customers. The product design and advertising of these companies have converged over time into a very narrow canon of female beauty. Many women compare their bodies to the images projected in this industry, and such comparison has generated body image disturbances, eating disorders, or low self-esteem, especially in young women (Martin and Kennedy 1993; Park et al. 2007). However, not all companies are obliged to follow this canon. For instance, in the cosmetic industry (ruled by a similar canon of beauty), Dove launched a marketing campaign in 2004 (“Dove Campaign for Real Beauty”) featuring women of different ages, shapes, and sizes. By showing a more real and varied images of the female body, the firm diverged from the norm and eased the pressure on its customers for getting a perfect body.
Because of both indirect and direct imitative tendencies over time, the mental models of competing strategists become similar, thereby creating “group-level” beliefs about the marketplace. (Porac et al. 1989, p. 400)
This process of mutual influence has been partly acknowledged in the stakeholder literature. Some authors have suggested that the focal firm and its stakeholders may constitute a genuine social community (Donaldson 1982; Donaldson and Dunfee 1994, 1999; Etzioni 1998). From a contractualist point of view, these authors have argued that companies connect with society by means of implicit contracts on two different levels. On a macro-social level, all economic agents are connected via a hypothetical contract which gives rise to several norms (hypernorms) binding all participants. On a micro-social level, agents implicitly participate in contracts which bind them in small communities. According to these authors, the company and its stakeholders are linked at the second level, forming a stakeholder community where agents share a system of norms, values, and expectations (stakeholder norms) which guide their behaviors. From our point of view, the concepts of stakeholder community and stakeholder norms correctly highlight the normative context in which companies operate. However, these concepts—based on hypothetical contracts, rather than actual interactions—do not help us understand the mutual influence between agents’ behavior and the normative system.
Foreseen and Unforeseen Effects
Finally, depending on an agent’s anticipation of the potential effects of his/her action, we can talk about foreseen or unforeseen effects. This distinction can be applied both to the effects on agents and the effects on normative structures. The predictability of an action’s consequences will depend on the cognitive ability and the interest shown by the agent, as well as the information available in the environment regarding potential effects of that type of action. This concept helps us identify an area in which certain social agents (NGOs, consumer associations, the media, governmental agencies, etc.) could play a significant role, by revealing barely known effects resulting from the actions of a company. In our hypothetical example, an NGO could make public the potential negative effects of Company A’s supply process on both natural resources and people, or disclose the names of other companies involved in the process.
In reality, many unforeseen effects have been revealed to the public opinion, and to the companies themselves, by NGOs and the media. Journalists working for The Observer and the BBC uncovered the labor abuses committed in GAP’s and Inditex’s supply chains, respectively. In 2004, Amnesty International’s reports revealed the fact that several vans manufactured by IVECO had been sold to the Chinese government and were being used for mobile execution in the province of Yunnan. In the case of Las Tablas de Daimiel, several NGOs (including Greenpeace, SEO/Birdlife and WWF) disclosed the serious cumulative effects deteriorating the wetland, thus putting pressure on politicians for the area to be protected.
Discussion and Conclusions
In recent years, many authors have reviewed the stakeholder model. As Freeman admitted in his seminal book, “the stakeholder approach offers no concrete, unarguable prescriptions for what a corporation should stand for… Instead, it presents a framework for discussing a host of differing moral views” (1984, p. 210). Despite defining stakeholders in very broad terms (as those agents who can affect or be affected by a company’s decisions), in his model, Freeman assumed two narrower criteria for the identification of these agents. Most authors have accepted his proposal without debating what kind of relationships and effects may link a company with other agents. As a consequence, stakeholder authors have yet to discuss a crucial question in this approach: what might the terms “can affect” and “can be affected” mean in contemporary societies? In our opinion, the stakeholder model is still a good framework to discuss CSR-related issues. Nonetheless, we also believe that the concept of stakeholder should be reviewed.
Classification of the possible effects of a company’s actions
Type of effect
Number of causal connections
Focus of the effect
Orientation of disseminative effects
Element that receives the effects
Effects on other agents
Effects on normative systems
Degree of anticipation
As discussed above, processes involving many agents can lead to serious cumulative and disseminative effects. In such processes, a company may have direct responsibility for the direct effects of its actions. However, the company in question may also have indirect responsibility, in the case of disseminative effects, or co-responsibility, in the case of cumulative effects. Stakeholder models have, to some extent, identified these kinds of effects. As we have already said, some models have acknowledged the possibility of an agent’s actions having indirect feedforward effects on other agents. However, there are other disseminative and cumulative effects yet to be studied.
Furthermore, the actions of a company can give rise to serious effects unforeseen by managers. This possibility makes it necessary to consider a wide range of relationships which were not included in Freeman’s original model. Acknowledging these unpredicted effects should lead companies to make further efforts to detect and counteract them. In order to improve our understanding of these strategies, it may be helpful to employ several notions developed within the literature on management under uncertainty (see De Meyer et al. 2001; Jafaari 2001; Pich et al. 2002), and on participatory planning, monitoring, and evaluation (see D’Aquino 2007; Estrella 2000; Estrella and Gaventa 1998; Nichols 2002). Research on management under uncertainty mostly stems from the project management field, and may suggest interesting ideas about the ways in which managers cope with unexpected events (Geraldi et al. 2010; Söderholm 2008). On the other hand, participatory planning, monitoring, and evaluation are typically used in development programs to involve all stakeholders, particularly local communities, in all aspects of a program (Holte-McKenzie et al. 2006). These techniques allow program designers to detect and manage unforeseen effects. However, their ethical dimension is yet to be explored.
In this study, we have emphasized the relationship binding the company to the normative structure. Our aim was not reviewing the full range of factors affecting the moral decisions of an agent. We have simply applied a logic which is characteristic of the stakeholder approach to the relationship between company and normative structure. The actions of a company may affect and/or be affected by the normative and value structures regulating relationships. Consequently, a company may be responsible—or co-responsible—for socially accepted patterns, especially if it holds an outstanding position within its environment. Although this idea has not been taken into account by the stakeholder approach, it could easily be incorporated into its models. If we all aspire to attain certain goals (related to the common good, progress in human rights, natural resource sustainability, etc.), then we should ensure that these normative systems—and not only the isolated actions undertaken by a firm—are truly sustainable.
In our opinion, all the effects and forms of responsibility presented in this essay may be included in the stakeholder approach. This task, however, would mean reviewing the stakeholder notion itself. In other words, this task would require expanding the range of relationships and effects considered to date. With a wider array of effects, stakeholder authors would be able to study how action comes about, and how responsibility is shared, in many complex processes traditionally left out of this approach.
The different kinds of effects included in this study may help authors to develop new models. These new models should, in turn, help us identify the agents involved, as well as understanding the processes they trigger, the normative structures affecting or/and affected by the agents’ actions, the mechanisms which allow agents to coordinate their actions, the effects generated by their decisions, etc. With these new models, it should be possible to answer questions which are still unanswered—and, in some cases, yet to be issued—such as can we limit the responsibility of a company to the direct and expected effects of its actions on other agents? What is the responsibility of a company which, despite being aware of the negative effects of a certain process, does not take active measures to avoid them? How should we evaluate the action of a company which contributes to creating, strengthening, maintaining, undermining, or eliminating norms and values that regulate the process? What are the conditions to be met to assert that a company has indirect responsibility or co-responsibility? With regard to cumulative and disseminative effects, which of the final effects should be attributed to each of the agents involved in the process? etc. From our point of view, the effects and relationships presented in this article could help authors understand social interaction—and consequently CSR—in a more complex yet realistic way.