Introduction

In the field of stakeholder research, and in particular within the business ethics and general management literature, it is widely acknowledged that there is conceptual confusion over the expression ‘stakeholder’ (Fassin 2009; Friedman and Miles 2006; Mitchell et al. 1997; Phillips and Reichart 2000; Sternberg 1997; Stoney and Winstanley 2001). Kaler (2003) and Waxenberger and Spence (2003) call for the refinement of the stakeholder concept, arguing that this is central, and fundamental to the development of stakeholder theory and practice. This has led to some critiques rejecting stakeholder theory. For example, Freeman et al. (2010, p. 63) state that ‘others have suggested that there is just too much ambiguity in the definition of the central term for it ever to be admitted to the status of theory’. This conceptual confusion is indicative in the wide array of definitions of the expression ‘stakeholder’ in the literature. Mitchell et al. (1997) were early explorers of definitional variety. They analysed 38 definitions in their preliminary discussions to establish the criteria for stakeholder identification and salience. Almost a decade later Friedman and Miles (2006) identified 55 definitions covering 75 publications and Laplume et al. (2008) reviewed 179 stakeholder definitions. Miles (2011) suggests that the current situation is far worse, with an analysis of 435 different definitions from 493 articles: a new definition every 1.13 articles published.

Every concept is liable to be contested, as the kind of use of a concept varies according to situation and circumstance. This fosters debate which in turn, is expected to result in further elucidation once the varieties of uses of a concept are acknowledged. Such debate surrounding conceptual clarity is needed in the development of theories and policies which encompass such concepts. There are, however, a number of concepts which are endlessly debated as each party maintains that the special situation in which they are defining the term is correct, or the only acceptable function in which the term in question can be said to fulfil (Gallie 1956). The lack of agreement can lead to conceptual confusion which is a major source of difficulty in theory and empirical analysis (Collier et al. 2006). Conceptual clarity helps to create a common frame of reference, enabling concepts to be discussed in the knowledge that there is a shared understanding of what they mean and prevents the field under investigation developing into an accumulation of disparate ideas. If conceptual confusion persists within a field of research, the development of the field is hampered, as researchers spend efforts debating these fundamental elements. This also hampers cross-discipline debate as different literatures adopt different definitions and conceptual understandings. The credibility of the associated adopted conceptual frameworks are called into question if fundamental concepts are thought of as vague or ambiguous, providing ammunition for skeptics to discount the concept and approach.

Why is there so little consensus over the concept of a stakeholder? Is this an insurmountable problem? Will stakeholder theorists continue to debate this concept and generate new definitions and interpretations with every (or almost every) article or theoretical advancement? Can a universal definition be identified, accepted and adopted, or is this a thankless task considering the diversity of philosophical bases that have been advocated and contributed to the stakeholder field of research? Kaler suggests that, at least for the purposes of business ethics, only ‘claimant’ definitions will do, ‘requiring some sort of claim on the services of a business’ (2002, p. 91). Does it then follow that, for the purpose of strategy and operations, only ‘influencer’ definitions, ‘requiring only a capacity to influence the workings of the business’ (2002, p. 91) are appropriate? Could a singular definition be accepted and adopted per discipline or school of thought? One way that these questions can be assessed is to consider whether the stakeholder concept is an essentially contested one, according to the philosophy of Gallie (1956) that will inevitably involve perpetual arguments about its proper use, rendering it inappropriate and unfeasible to seek such a singular definition. Or is the stakeholder concept radically confused and in need of further structured debate to add clarity and consent? This article attempts address some of these questions through an assessment of whether the lack of consensus surrounding the stakeholder concept is due to a radical confusion, or whether the stakeholder concept is essentially contestable or, indeed, essentially contested.

Essentially Contested Concepts

Essentially contested concepts (ECC) have been described as ‘vague, ambiguousFootnote 1 and general’ (Kekes 1977, p. 71) and ‘complex and normative’ (Jacobs 2006, p. 25). It relates to situations where there is a variety of meaning employed for key terms in an argument and is widely used to signify imprecise use of terminology. Hence ECCs are used to indicate situations where there is widespread acceptance of a concept but disagreement on the best instantiation of it. It is important to recognize that, in situations where a concept is essentially contested, each party recognizes that its own use is contested by other parties and that each party has some consideration of the different criteria applied by others. Not all concepts that are debated are ECCs, as some disputes, according to Gallie (1956), may be resolved through either discovering a new meaning of a term to which all disputants could thenceforward agree; convincing all the disputants to conform to one meaning; or by declaring the concept to be a polyseme, or a number of different concepts employing the same name.

Gallie dealt exclusively with abstract, qualitative and positively regarded notions such as art, religion, science, democracy and social justice. In the intervening decades, it has been applied to a range of concepts including, corporate social responsibility (Okoye 2009), liberalism (Abbey 2005), love (Hamilton 2006), powerFootnote 2 (Lukes 1974), rule of law (Collier et al. 2006) and sustainable development (Jacobs 2006).

Jacobs (2006) argues that ECCs have two levels of meaning: At the first level of meaning there is agreement, in that, even people holding widely different views agree what the subject is when using a certain term, as there are no other terms expressing the same set of core ideas. The debate occurs at the second level of meaning: “political arguments over how the concept should be interpreted in practice” (Jacobs 2006, p. 25). This conforms to Swanton’s (1985) idea that an ECC has a common core (the concept) but that detailed specification (interpretation) of this core is essentially contested. Consequently there is no point in trying to develop a singular universally acceptable definition because those with different political values and interests will only agree at the first level. Jacobs argues that such disagreements are not semantic disputations but substantive political arguments.

Gallie (1956) argued that if none of these resolutions are feasible it is likely that the concept is an ECC. An ECC must meet five necessary conditions (1956, pp. 171–172):

  1. 1.

    “It must be appraisive in the sense that it signifies or accredits some kind of valued achievement;”

  2. 2.

    “This achievement must be of an internally complex character, for all that its worth is attributed to it as a whole;”

  3. 3.

    “An explanation of its worth must therefore include reference to the respective contributions of its various parts or features…the accredited achievement is initially variously describable;”

  4. 4.

    “The accredited achievement must be of a kind that admits of considerable modification in the light of changing circumstances; and such modification cannot be prescribed or predicated in advance. For convenience I shall call the concept of any such achievement “open” in character”.

  5. 5.

    “That each party recognizes the fact that its own use of it is contested by those of other parties, and that each party must have at least some appreciation of the different criteria in the light of which the other parties claim to be applying the concept in question. More simply, to use an essentially contested concept means to use it against other uses and to recognize that one’s own use of it has to be maintained against these other uses. Still more simply, to use an essentially contested concept means to use it both aggressively and defensively”.

There are also two justifying conditions (1956, p. 180), which are required to differentiate ECCs from those concepts that are merely “radically confused”:

  1. 6.

    “The derivative of any such concept from an original exemplar whose authority is acknowledged by all contestant users of the concept, and”

  2. 7.

    “The probability or plausibility, in appropriate senses of these terms, of the claim that the continuous competition for acknowledgement as between contestant users of the concept, enables the original exemplar’s achievement to be sustained and/or developed in optimum fashion”.

In partial response to Garver’s (1990) claim that ECCs have been theorised about more than used, these criteria will now be applied to the concept ‘stakeholder’ to ascertain whether this concept is essentially contestable, or indeed essentially contested.

Is the Stakeholder Concept Essentially Contested?

The stakeholder concept could be regarded as a contested concept because it has been the focus of continued debate about its proper meaning from different users. An ECC goes beyond the idea that a concept is vague and ambiguous, but as a starting point it is clear that the ‘stakeholder’ concept is vague. Many authors have highlighted the wide range of meanings ascribed to this concept, and the ensuing confusion. It has been clearly argued that this has hampered intellectual enquiry and that calls have been made to address this central issue to stakeholder theory, most notably by Kaler (2002). Freeman et al. (2010) argue that ‘a recurring issue for stakeholder theory has been how to understand who stakeholders are’ (p. 206) and that ‘work needs to be done to pare down and refine what we mean by stakeholders if the term is to prove helpful at a conceptual level or at a practical level’ (p. 208). The range of definitions is quite frankly unhelpful. Miles (2011) identified 435 different definitions of the expression, with clusters around just a few definitions such as Freeman (1984, pp. vi, iv, 25, 54) “affects and affected by”; Clarkson’s (1995) ‘primary stakeholder’; Mitchell et al. (1997) ‘power-legitimacy-urgency’; Clarkson’s (1994) ‘secondary stakeholders’ Clarkson (1994) ‘at risk’, and SRI (1963) “without whose support …”

It is also clear that to date, such resolutions to disagreement, as suggested by Gallie, have not been found and, whilst one can recognize different forms of stakeholders, the first level meaning is agreed upon, negating the idea that it is polysemous. This does not seem surprising considering, sustainability (Jacobs 2006) and corporate social responsibility (Okoye 2009) have been found to be ECCs; concepts with which the ‘stakeholder’ concept holds many similarities. All of these concepts are widely accepted, supported by a burgeoning literature, subject to dissimilar interpretations from diverse social contexts and ideologies, relate to varied practice and theory, attract continual arguments about ‘proper meaning’, and, suffer from ambiguity due to the wide array of definitions attributed.

Appraisive in Nature

Gallie simply refers to the first criteria in that ‘it must be appraisive in the sense that it signifies or accredits some kind of valued achievement’ (1956, p. 172). He is also clear in stating that an ECC can exist if the term is mainly, if not exclusively, an appraisive term. Collier et al. (2006) state that ‘for some major concepts the normative valence may be unclear and may depend on the theoretical framework employed, or on the specific context in which the concept is applied (2006, p. 216). They refer to Lukes (1974) assessment of power, and Baldwin’s (1997) assessment of security, whilst neither being directly and overtly appraisive, security is a desirable outcome and power can be used for good (and bad). So an ECC does not have to be exclusively appraisive, and the appraisive nature may stem from the theoretical framework or context applied (see Kekes 1977). The concept of a ‘stakeholder’ per se cannot be considered appraisive or as accrediting a vital element to corporations, but the process of stakeholder management (specific context), or adopting a stakeholder theory approach (ideological framework) to running a corporation is appraisive, as there are many benefits associated with it. For example, Freeman (1984, p. 26) argues that the consequences of not adopting a stakeholder approach is to suffer from law suits, damaging regulation, and a loss of markets to foreign competitors who can satisfy a range of stakeholder needs. Some argue to the contrary, for example, Cennamo et al. (2009, p. 492), in support of Greenwood (2007), state that stakeholder management is ‘morally neutral – engaging stakeholders and committing to them does not automatically imply a responsibility of the firm to act in their interests: A stakeholder-committed organization may still act out of self-interest’. However, as Okoye (2009) argues, the emphasis is on the evaluative nature of the concept and not on the response. Stakeholder management can result in a valued achievement regardless of whether this is predicated through the genuine ethical consideration of all stakeholders, or via a public relations exercise: arguably in either case stakeholder management is a positive attainment. Normative consequences may stem from strategic actions. Likewise from the stakeholder’s perspective, recognition of their voice, claim, interest or legitimacy by an organisation is also appraisive.

Not all commentators agree on the inclusion of ‘appraisive in nature’ as a criterion, for example, Freeden (1998, p. 56) argues that ‘not all essentially contested concepts signify valued achievements; they may equally signify disapproved and denigrated phenomena’.

Internally Complex and Variously Describable

Gallie (1956, pp. 171–172) proposed that ‘This achievement must be of an internally complex character; for all that its worth is attributed to it as a whole.’ Kekes (1977, p. 79) clarifies this stating that ‘although the activities exemplifying ECCs are internally complex, the normative character of ECCs is due to the performance as a whole’. In illustrating the application of ‘internally complex’ to the concept of democracy Gallie states that it ‘is internally complex in such a way that any democratic achievement (or programme) admits of a variety of descriptions in which its different aspects are graded in different orders of importance’ (1956, p. 184), thereby suggesting that different appraisers afford different weightings to various features of the concept. This criterion is very closely linked to criterion III, ‘variously describable’. This may involve an exclusive emphasis on one or another aspect of the concept, or different facets may be emphasised to varying degrees, involving contrasting relative importance (Collier et al. 2006, p. 217). Gallie illustrates this with respect to democracy, by providing a range of definitions which focus on various aspects such as: the power of the majority of citizens to choose (and remove) governments; equality of all citizens to attain position of political leadership, and; continuous active participation of citizens in political life. Whilst a range of definitions can emerge from the weightings given to various components of a concept, according to the ideology and social positioning of the definer, it should still be possible to judge some meanings as better than others (Swanton 1985).

Most authors agree on the importance of both of these criteria (for example, Collier et al. 2006; Grafstein 1988). Gallie frequently combines the terms internally complex and variously describable, and I shall deal with these concomitantly.

Prior to the assessment of whether the stakeholder concept is internally complex and variously describable, I agree with Collier et al. (2006) in that it is import first to assess whether the concept under investigation is over aggregated or a ‘cluster concept’. They argue that labelling a concept as diversely describable is only meaningful if, indeed, the different components are actually part of the same concept (cluster concept) rather that one which is over aggregated, combining elements that are only loosely related to one another. In the latter case it is appropriate to disaggregate the concept, thereby drastically reducing or even eradicating the internal complexity. Hence the concept cannot be essentially contested. Is the stakeholder concept over aggregated? It is widely recognized that the concept is muddled and confused and there has been a lot of focus on trying to organize and add structure to stakeholder definitions. For example Kaler’s separation of influencer, claimant and combinatory definitions and Fassin’s (2009) differentiation between a “stakeholder”, “stakewatcher” and “stakekeeper”, which could be interpreted as an attempt at disaggregation. Even if this is so what remains is clearly a cluster as there is sufficient commonality across all interpretations to warrant that the components are actually part of the same concept.

With respect to the concept ‘stakeholder’, there are a number of different components that can be considered when defining the term, and different appraisers have notably given more weighting to one component over another. The components emphasised can relate to:

Who?

  • Who are the stakeholders? Any human agency (Gray et al. 1996), any naturally occurring entity (Starik 1993), citizens (Crane et al. 2004), coalition of people (MacMillan and Jones 1986), constituents (Ruf et al. 2001), employees; suppliers; government; local, regional and national communities; banks and shareholders (Gamble and Kelly 2001), groups (SRI 1963), individuals (Donaldson and Preston 1995), moral actors (Hendry 2001), participants (Orts and Strudler 2002), parties (Lampe 2001), socio-political actors (Mattingly 2004), voluntary members of a co-operative scheme (Phillips 1997).

  • Who is identifying the stakeholder? Corporation (Wicks et al. 1994), enterprise (Nuti 1995), firm (Donaldson and Preston 1995), organisation (Freeman 1984), management (Reed 2002).Footnote 3

How?

  • How the stakeholder impacts the organisation? affects (Freeman 1984); benefits (Ojala and Luoma-aho 2008), continuing participation (Clarkson 1995), gives meaning and definition to (Wicks et al. 1994), harms (Ojala and Luoma-aho 2008); helps (Miller and Lewis 1991), hurts (Miller and Lewis 1991); impacts (Post 1989), influences (Starik 1994); interacts (Näsi 1995), supports (SRI 1963) and threatens (Polonsky 1996).

  • How the organisation impacts the stakeholder? Affected by (Freeman 1984), benefit from (Donaldson and Preston 1995), gain (Rowley 1997), harmed by (Donaldson and Preston 1995), influence by (Starik 1994); lose (Rowley 1997), placed at risk (Clarkson 1994), rights are violated or respected (Evan and Freeman 1988)

Why?

  • Why are the stakeholders being identified? Exchange relationship (Hill and Jones 1992), exchange transactions (Brenner 1993), for existence (Rhenman 1964), joint value creation (Freeman 1994), moral obligation (Phillips 2003), resources (Hill and Jones 1992), responsibility (Alkhafaji 1989), survival (SRI 1963).

What?

  • What is the form of the stake? Claim (Evan and Freeman 1988), contract (Cornell and Shapiro 1987), interest (Carroll 1993), investment (Blair 1998), ownership (Madsen and Ulhøi 2001), right (Carroll 1993), risk (Clarkson 1994).

  • What is the nature of the stake? Clear (Bendheim et al. 1998), direct (Schneper and Guillén 2004), future (Galai and Wiener 2008), implicit or explicit (Donaldson and Preston 1995), legal (Langtry 1994), legitimate (Donaldson and Preston 1995), moral (Langtry 1994), non-trivial (Brenner 1993), normative (Reed 1999), urgent (Mitchell et al. 1997), vital (Murphy et al. 1997).

  • What does the stake relate to? accomplishment of organisational purpose (Foster and Jonker 2007), achievements of the organisation (Freeman 1984), firm behaviour (Eesley and Lenox 2006), firm’s actions (Burton and Dunn 1996), firm’s performance (Jones 1995), institutions policies (Mitroff and Linstone 1993), operations (Hart and Sharma 2004), organisation’s activities (Driscoll and Crombie 2001), organisation’s mission (O’Riordan and Fairbrass 2008), organisational decisions (Wartick 1994), organisational outcomes (Ferrell and Ferrell 2008), organisational performance (Starik 1995), procedural and/or substantive aspects of corporate activity (Donaldson and Preston 1995), value of the firm (Jensen 1983).

In an essentially contested concept, the participants to the argument disagree about the significance and, consequently, the weighting given to each element so that they may assess the situation differently according to their own agenda (Kekes 1977). The complexity of the stakeholder approach is due, in part, to its diverse origins. In tracing the history of stakeholder theory, Freeman et al. (2010, p. 30) state that there are many different versions and it is unhelpful to synthesize all of them into something approximating the ‘correct version’. The complexity of the stakeholder concept and the diversity of approach have been highlighted by those attempting to organize or structure the field of research. Donaldson and Preston advocated that stakeholder theory is unarguably descriptive, instrumental and its fundamental basis is normative. They also state that it is managerial. This basic classification is widely accepted by stakeholder theorists and, broadly speaking, research falls into these categories, although it is not without its critics. Freeman argues ‘As a pragmatist, I tire of the debate between what’s normative, what’s descriptive and what’s instrumental. It really leads nowhere for lots of good reasons’ (cited in Agle et al. 2008, p. 163). Friedman and Miles (2006) also attempt to classify stakeholder theories. They suggest that ‘analytic’ stakeholder theories can be ordered on the basis of the strategy formulation process (whether theories are concerned with issue identification, stakeholder salience, stakeholder identification, analysing or explaining behaviour, etc.) and view of the field of stakeholders (whether distinct groups or interacting groups). For Friedman and Miles (2006) normative theories can be distinguished on the basis of the ‘degree of normativity’ involved:

  1. 1.

    Third kind theories: those theories which are easily achievable in practice, for example by aiming to change managerial attitudes, and can therefore have a direct impact on business in society. Examples include stakeholder theory based on: Fiduciary relationships (Boatright 1994, 2002; Goodpaster 1991); Property Rights (Donaldson and Preston 1995) and Aristotelian ethics (Wijnberg 2000).

  2. 2.

    Second kind theories: those theories that are less readily realizable and call for changes at societal levels, for example through modifications in law. Examples include: Feminist ethics (Wicks et al. 1994; Burton and Dunn 1996); Stakeholders viewed as investors (Etzioni 1998; Blair 1998)

  3. 3.

    First kind theories: those theories that represent idealized levels of normativity and are therefore unrealizable. Examples include Ecology (Starik 1994); Kantian ethics (Evans and Freeman 1993; Bowie 1998); Rawlsian ethics (Phillips 1997; Freeman 1994) and The Common Good (Argandoña 1998)

Using this continuum as a framework to examine the significance and weighting given to each element by contributors, it is clear that even within normative stakeholder theory the range of normative cores adopted is diverse, with clear disagreements over the hierarchical ranking of the elements. For example, to facilitate the acceptance of third kind theories one might expect to see narrow stakeholder definitions as the narrower the definition the more operational stakeholder management becomes. First kind theories are, in contrast, likely to be associated with very wide, all inclusive, but perhaps impractical definitions. This is clearly seen in the following: Starik (1994, p. 90), a first kind theory, defines stakeholders as those who “can and are making their stakes known…those who “are or might be influenced by, or are, or potentially are, influencers of, some organization”. This definition could include just about anybody, or anything, which is what Starik intended as he was arguing that trees should have managerial status. Boatright (2002, p. 1839), a third kind theory, defines stakeholders as “…groups [other than shareholders] have a legitimate claim or ‘stake” in a corporation”, narrowing the remit of managerial attention to just those groups (not individuals) with legitimate claims or stakes in the organisation. This is interesting as one might expect to see large variations in the emphasis between strategic and normative theories, with respect to the form of stakeholder definitions, but not necessarily such differences within normative stakeholder theory. Even within first kind theories there are noticeable difference between scholars. For example, Phillips’ (1997) stakeholder fairness, based on Rawlsian principles, permits the separation of those who have a stake in an organisation from those that can influence an organisation, unlike Starik, who treats the two as interchangeable. Consequently Starik’s definition includes both non-humans and antagonists, such as terrorists or activists, which would all be excluded according to Phillips. Other weightings vary according to approach. Wijnberg (2000), for example, focuses on the individual responsibility of managers in his exploration of how stakeholder theory can be informed from Aristotelian ethics, whereas most other contributors focus on corporate responsibility. Another interesting contrast is the degree of symmetry perceived in the organisation–stakeholder relationship: whether the relationship is multidirectional or one directional, and if so which direction? For example, Argandoña (1998) presents an approach to stakeholder theory based on the common good and Christianity. He suggests that the common good of a company is to enable its members (stakeholders) to achieve their personal aims and that the company facilitates this through the achievement of its own aims. In line with this, Argandoña (1998, p. 1099) defines stakeholders as “those who have an ‘interest’ in the company (so that the firm, in turn, may have an ‘interest’ in satisfying their demands”. This rather symmetrical perspective of a stakeholder and firm contrasts with the feminism and virtue ethics approach of Burton and Dunn (1996) and Wicks et al. (1994), which emphasise the co-operative and caring relationship that firms should adopt to promote stakeholder satisfaction. For example, Wicks et al. (1994, p. 483) define stakeholders as “groups who interact with and give meaning and definition to the corporation”.

Likewise, within strategic stakeholder theory, numerous approaches have been purported, depending on the purpose of the theory and aims of the researcher. In order to try to explain and analyse stakeholder behaviour researchers have looked to critical realism (Friedman and Miles 2002), resource dependency theory (Frooman 1999; Kassinis and Vafeas 2006), network theory (Rowley 1997), social movements theory (King 2008) and communication theory (Kuhn 2008). Consequently stakeholder theory is variously describable as the focus changes from whether stakeholder–organisation relations are necessary or contingent and compatible or incompatible (Friedman and Miles 2002), to the degree of dependency within the firm-stakeholder relationship (Frooman 1999), to the degree of interest overlap and interest identity within stakeholder groups (Rowley and Moldoveanu 2003). In addition, as noted by Friedman and Miles (2006, p. 84), further internal complexity arises depending on whether the emphasis of the theory is organisation-centric, for example Freeman (1984) and Savage et al. (1991), stakeholder-centric (Frooman 1999; Rowley and Moldoveanu 2003) or looking more specifically at the actual organisation–stakeholder relationship (Friedman and Miles 2002).

Internal complexity has also arisen because of the wide range of subject fields that have adopted a stakeholder approach. Freeman et al. (2010) explore the extent to which stakeholder theory has been used within finance, accounting, management, (including human resource management, operations and strategy) and marketing. They argue that researchers have “selected the portions of the theory that are most applicable to the questions that they are trying to answer” (2010, p. 121). An example cited is that “finance scholars traditionally ignore the moral foundation of stakeholder theory” (2010, p. 124) as the objective of the firm, according to finance theory, is to maximise shareholder value, which would suggest that definitions would focus on narrow ‘influencer’ definitions. For example, Jensen (2001, p. 299) defines stakeholders as “all individuals or groups who can substantially affect the welfare of the firm”. This contrasts with the application of the stakeholder concept to the field of accounting research which has been dominated by the need to widen both the accountability of managers in terms of what they report on (to include corporate social and environmental impact), and the range of stakeholders to whom they report. Consequently accounting scholars are expected to adopt wider stakeholder definitions and to prefer claimant to influencer forms, such as Gray et al. (1997, p. 330) who define stakeholders as “those with rights to the account”. Whilst one might expect the marketing literature to adopt a narrow definition of a stakeholder, focusing on the corporate–customer relationship, there is a call for a widening of this remit to consider the impact of other stakeholders, such as employees, suppliers, distributors, dealers and advertising agencies on the customer experience (Jackson 2001; Kotler 2003; Polonsky and Ottman 1998). Other disciplines embracing stakeholder theory include corporate governance, information systems, corporate environmental management and business ethics. Kaler (2002, p. 91) expresses alarm that “even within the context of the very same area of business studies (and for the purposes of this paper that means business ethics) and, quite independently of variations in theoretical perspective (be they pro or anti), there is a deep divide in definitions of what it is to be a stakeholder in a business”. According to Kaler, three types of stakeholder definitions are prevalent within business ethics: influencer definitions “requiring only a capacity to influence the workings of the business”, claimant definitions “requiring some sort of claim on the services of a business” and combinatory definitions “allowing for either or both of these requirements”. Whilst it is beyond the scope of this paper to map the range of, and reason for, the diversity of approaches that exist both within and beyond management subjects, it is reasonable to assume that similar ‘stories’ exist in other disciplines across the management literature.

Looking wider that just management literature, it is easy to establish, through the search engines of electronic library databases, that the adoption of the stakeholder concept has been widespread throughout many literatures from the environmental sciences, through to health and medical studies, language studies, law, pedagogy and education, planning and architecture, public sector studies, social studies and tourism. Again, it is not the intention of this research to investigate the diversity of approaches across all these disciplines but it would be highly surprising if there was any degree of coherence and consistency. In practice, one would expect wide definitions to be adopted or developed in the natural and environmental sciences, but narrow definitions for the purpose of law, education or language studies.

A further factor that influences the type of definition adopted and the relative weightings given to the various attributes of what constitutes a stakeholder is who is the user. It can be seen from the earlier discussion that academics researching within a single field of study cannot agree on an approach. It is, therefore, unlikely that a consensus can be agreed upon between practitioners with divergent political motives and personal agendas such as corporate managers, governors of charities, directors of non-government organisations, political activists, management consultants, public sector managers or politicians.

Open in Character

The fourth criterion requires that essentially contested concepts are subject to unpredictable periodic revision or considerable modification in new situations or changing circumstances. Gallie (1956) provides two illustrations of ‘openness’: ‘Artistic achievement, or the persistence of artistic activity is always “open” in character in the sense that, at any one stage in its history, no one can predict or prescribe what new development of current art-forms may come to be regarded as of properly artistic worth.’ (1956, p. 182) and ‘Politics being the art of the possible, democratic targets will be raised or lowered as circumstances alter, and democratic achievements are always judged in the light of such alterations’ (1956, p. 186).

For Gallie, advocates of the conceptual debate must consider the changing situations when defending their position, and they may effectively adapt to preserve their preferred conceptualisation. MacIntyre (1973) views openness as a key factor for essential contestability, and according to Collier et al. (2006) most commentators have affirmed the importance of openness. Openness, according to Care (1973) may, however, be partially overcome. He suggests the possibility of practical or temporary closure whereby conceptual competitors reach temporary and provisional settlements of debate.

The stakeholder concept has developed over time as diverse fields have embraced it. The first conception of a stakeholder in the management literature was in an internal memo at the Standford Research Institute (SRI) (1963): “those groups without whose support the organization would cease to exist”. This definition was central to the SRI corporate planning process. It focused on the need to formulate strategic objectives that considered the support of shareholders, employees, customers, suppliers, lenders and society as necessary for the survival of the firm. Early conceptions of stakeholders focused on influencer definitions, as the subject emerged from the field of strategy, thereby focusing on how to manage primary or core stakeholders. Freeman et al. (2010, p. 38) argue that ‘since the major concern was with forecasting the future environment and not with changing specific stakeholder behaviour, there was no need to go beyond this generic stakeholder analysis’. This was sufficient in the 1950s–1970s as the business environment was relatively stable. Freeman (1984) spoke of the turbulence that corporations were facing in the 1980s and the need for a new management paradigm. The changing circumstances included the impact of the social movements of the 1960s and 1970s with respect to environmentalism, women’s rights, civil rights and anti-war movement. Other modifications to business, which have impacted the conceptualisation of the stakeholder include a rise in stockholder activism following a change in resolution laws, a loss of markets to foreign competitors, a younger workforce with different values and a need for participative management styles, politicization of international supply of raw materials, expansion of federal, state and local government activities, the growth of special interest groups and the consumer movement. Freeman’s corresponding conceptualisation of a stakeholder was “any group or individual who can affect or is affected by the achievements of the organization’s objectives” (1984, p. 46). He warned management that the consequence of not adopting this approach would be to suffer from law suits, damaged reputations and a loss of markets to competitors who can satisfy stakeholder needs. This required a change from merely forecasting the future environment to understanding how stakeholder groups arise, the key issues that they perceive and their willingness to help or hinder the corporation. Hence the concept needed to reflect a symmetric situation in which the organisation can affect a stakeholder and the stakeholder can affect the organisation.

More recently there have been growing pressures on organisations to become more socially responsible, thereby widening the concept of the stakeholder. The impact of the internet cannot be understated as corporate misdemeanors are disseminated, in real time, 24 h a day, to a global audience. Corporate disasters, from Union Carbide’s explosion in Bhopal in 1984 through to BP’s deepwater Horizon drilling rig explosion in the Gulf of Mexico in 2010 have raised questions over the environmental and social cost of profits. The impact of globalization has unveiled a number of ethical and environmental issues such as the exploitation of labour and the adoption of ethical or environmental standards in host nations which are inconsistent and substandard in relation to head office policies. Ethical consumerism is now an important consideration. Even shareholders are concerned about the level of social responsibility in corporations, as evidenced by the exponential rise in socially responsible investment (SRI) funds and the emergence of SRI indices such as the Dow Jones Sustainablity index, FTSE4Good, and the Ethibel Sustainability Index, as alternative listings on stock exchanges. This is set against a general growth in social awareness and improving education on corporate social responsibility (CSR) issues. During this period corporate and public policy in many nations endorsed the stakeholder concept. Anecdotal evidence shows a trend amongst corporations to actively manage stakeholder relations by appointing stakeholder or community affairs managers and increased instances of ‘stakeholder reporting’ as an adjunct to environmental and social reporting on corporate websites. Friedman and Miles (2006, p. 28) state that in the UK, the term stakeholder was associated with the Blair Labour government and their idea of the ‘Third Way’ and specific government policies, such as stakeholder pensions. A wide range of stakeholder public policy also emerged:

  • Stakeholder laws are evident in 32 US states.

  • The Sarbanes–Oxley Act, a broad law covering corporate responsibility, fraud, auditor independence and financial disclosures, was enacted in 2002 in response to high profile corporate scandals involving WorldCom, Adelphia, Enron and Arthur Andersen, Global Crossing and Tyco amongst others.

  • In Europe, legislation requiring pension institutions to report how CSR issues are addressed in investment portfolios was passed in Belgium, Germany, Sweden and the UK in 2001–2002.

  • Enacted in 2001, the New Economic Regulations Law requires all French listed companies to report on stakeholder issues.

  • An EU green paper on CSR (July, 2001) emphasised that an important aspect of CSR is how corporations interact with stakeholders. A related multi-stakeholder forum on CSR was established in 2002.

These changing circumstances in the business environment have resulted in more claimant definitions as business ethics research into the stakeholder concept has burgeoned, whereby the range of stakeholders identified now includes those with informal or implicit contracts. For example, Kaler (2002, p. 91) defines stakeholders as ‘claimants towards whom businesses owe perfect or imperfect moral duties beyond those generally owed to people at large’.

Aggressive and Defensive Uses

Gallie (1956, p. 172) states that ‘each party recognizes the fact that its own use of it (the contested concept) is contested by those of other parties, and that each party must have at least some appreciation of the different criteria in the light of which the other parties claim to be applying the concept in question. More simply, to use an essentially contested concept means to use it against other uses and to recognize that one’s own use of it has to be maintained against these other uses. Still more simply, to use an essentially contested concept means to use it both aggressively and defensively.’ Collier et al. (2006) refer to this as ‘reciprocal recognition’. Although this criterion has widespread acceptance (e.g. Smith 2002; Grafstein 1988), it is dismissed by Freeden (1998) who argues that this criterion is not always present as parties may employ concepts without explicitly acknowledging alternative conceptualisations. Consequently, he argues that the violation of this criterion is insufficient to dismiss the essentially contested character of a concept. This is very true of individual papers arguing for and against a stakeholder approach. There are, however, a number of clear opponents of the stakeholder approach, such as Friedman (1970), Jensen (2002), and Sternberg (1996, 1997) who argue that managers act as agents to shareholders and they are imparted with a fiduciary duty to maximize the wealth of shareholders. Attending to other stakeholder interests is therefore a breach of their fiduciary duty and clear example of property theft. On the other hand, the list of supporters appears almost infinite. For example, Post et al. (2002, pp. 16–17) state that “The corporation cannot, and should not, survive if it does not take responsibility for the welfare of all its constituents and for the well-being of the larger society in which it operates”.

There are a number of definitions that are frequently discussed when authors define a stakeholder, the most popular includes;

  • Stanford Research Institute (1963): “those groups without whose support the organization would cease to exist”;

  • Clarkson (1995, p. 106) “A primary stakeholder group is one without whose continuing participation the corporation cannot survive as a going concern”.

  • Clarkson (1995, p. 107) “secondary stakeholder groups are defined as those who influence or affect, or are influenced or affected by, the corporation, but they are not engaged in transactions with the corporation and are not essential for its survival. …however such groups can cause significant damage to a corporation”

  • Clarkson (1994, p. 5) “bear some form of risk as a result of having invested some sort of capital, human or financial, something of value, in a firm”.

  • Mitchell et al. (1997, p. 854) “classes of stakeholders can be identified by their possession or attributed possession of one, two, or all three of the following attributes: the stakeholder’s power to influence the firm, the legitimacy of the stakeholder’s relationship with the firm, and the urgency of the stakeholder’s claim on the firm”.

However, the most widely cited conceptualisation of a stakeholder is that of Freeman (1984, pp. iv, vi, 25, 46): ‘any group or individual who can affect or is affected by the achievement of the organization’s objectives’(1984, p. 46). Freeman et al. (2010) recognize the problem of conceptual breadth and state that ‘on this definition one could image virtually anyone, or any organization- including groups who are only incidentally linked to the form, or whose purposes are explicitly directly at odds with the firm…Given such a wide view of what the term might mean, the notion of stakeholder risks becoming a meaningless designation’ (2010, p. 208). Others have criticized and rejected this definition on the basis that it is too wide (Mahon 2002; Mason and Gray 1999; Reed 1999) as “…it counts the whole human race as stakeholders in just about every form” (Langtry 1994, p. 432). The context in which the conceptualisation is being created needs to be considered. Mason and Gray (1999) for example are interested in identifying stakeholders in advance of the purchase decision in hybrid markets and argue that this is not possible with a broad conceptualisation: alternatively they suggest ‘those that have a vested interest in the decision’ (1999, p. 849). Mahon (2002), suggests that anyone looking to develop a schema for weighing stakeholders needs to narrow down the range of stakeholders to focus on the critical ones on which the company depends on for survival. Likewise Mahon and Wartick (2003) considered ‘affects and can affect’ untenable if attempting to define stakeholder to develop a strategy for using reputation in issue management. Kochan and Rubinstein (2000, p. 369) suggest that Freeman’s definition is “too broad to be of analytical value”, preferring a narrower conceptualisation that captures saliency or level of influence of potential stakeholders. For scholars concerned with the question of to whom the firm has obligations, the affects/affected by definition says nothing about what duties management have towards stakeholders and whether such duties are equal. Argandoña (1998, p. 1098), for example, states that there is “a lack of any normative rationale or criteria for identifying who the stakeholders are or for allocating the rights corresponding to each one” and that it ‘cannot be used as a foundation for moral duties: it is “prudent” to take them into account but this does not mean that any duties are created towards them’. Sternberg (1997, p. 4) states: “The widely used Freeman characterization of stakeholder…transforms everyone into a stakeholder. It not only includes those who have a stake in the organization as well as those in whom the organization has a stake, but it excludes all criteria of materiality, immediacy and legitimacy”. This is echoed by Phillips (1997, p. 63) “one example that has troubled some is the problem of whether terrorists are a stakeholder group. Although many of our considered judgments lead us to say “no”, earlier versions of the theory would have to say “yes” due to the fact that they can certainly affect the firm”. Sternberg suggests the alternative definition ‘the various groups and individuals that organizations need to take into account when pursuing their substantive objectives’ (Sternberg 1997, p. 9).

Original Exemplar or Core Concept

Gallie introduces an historical dimension with this criterion, suggesting that there is a past conceptualisation whose authoritative status is accepted by all. Collier et al. (2006) state that the role of exemplars in Gallie’s framework has led to confusion, as Gallie uses the term both narrowly, to mean an original exemplar, and broadly, to mean a common core, or a notion that anchors the concept, so that the contestation is over different manifestations of the same concept. Consequently an exemplar should not be interpreted as a “platonic heaven, a non-contested, unambiguously defined and fully determinate concept” (Gellner 1967, p. 53, cited by Collier et al. 2006, p. 220). The insistence of an exemplar may appear counterintuitive to the very core of an essentially contested concept, as by recognizing the endless disputes over the conceptualisation of a concept, one is acknowledging that there is no best instantiation of an essentially contested concept, or, at least, none knowable to be the best. However, it is clear that some instantiations will be considerably better than others; Swanton (1985, p. 815). In addition, due to the open context, and instantiation deemed to be best at any one time (causing temporary closure of the debate), changing circumstances may lead to better instantiations emerging in the future. Abbey (2005, p. 467) observes that the notion of an exemplar is rejected by Freeden as implausible and even “inimical to the very notion of essential contestability, as it presumes an agreed or correct position from which deviations have occurred’. Waldron (2002) views the achievement of an exemplar too narrow to distinguish a contested concept and Kekes (1977) also questions whether the recognition of an exemplar is good grounds or differentiating between essentially contested concepts and those that are merely radically confused. He contests that it is not always possible to find an exemplar, that there may be exemplars from different fractions and why would past instances be superior to present or future ones? Kekes suggests replacing the notion of an exemplar with problem solving, and Swanton (1985, p. 822) suggests that ‘the idea of a common core concept should be abandoned by essential contestedness theorists’.

Freeman was not the first attempt to define a stakeholder in 1984 but it is without question the original exemplar. Friedman and Miles, in their 2006 book which attempts to map the entire stakeholder field start it with “the classic definition” (2006, p. 1) and state that “Freeman is generally credited with popularizing the stakeholder concept” (2006, p. 19). Mahon and Wartick (2003, p. 24) state “Freeman’s work in 1984 is the starting place for any discussion of stakeholder management”, whilst Walsh (2005, p. 427) clearly states “Edward Freeman’s 1984 text is clearly a classic in our field… ask any member of the Academy of Management about the origins of the stakeholder construct and I wager that 99 out of 100 colleagues will point you to this book”. Phillips calls it a “groundbreaking book” saying that it “took the stakeholder idea to a higher level of theoretical sophistication” (2003, p. 66). “Indeed, a citation count in July 2004 revealed that it had been cited 485 times. That is approximately eight times the average number of citations a typical 1984 Academy of Management Journal paper has received” (Walsh 2005, p. 427). It is clear that his definitions are the most widely cited of all definitions (Miles 2011). In addition many authors acknowledge this as an exemplar before rejecting it (for example, see Argandoña 1998; Humber 2002; Kaler 2003; Langtry 1994; Mahon and Wartick 2003; Mahon 2002; Mason and Gray 1999; Kochan and Rubinstein 2000; Reed 1999).

To end this, I suggest that the stakeholder concept is essentially contested rather than radically confused as the continued use of the ECC is justified by the acknowledgement that, despite ongoing debates, disputants acknowledge that it is derived from Freeman’s singular common exemplar. In addition, there is also general acceptance of the normative core within stakeholder research thereby supporting the broad interpretation of ‘an exemplar’: ‘stakeholder theory as a genre, to recognize the value of variety of uses that can be made from the common set of ideas. There is sufficient commonality across these uses to see them as part of the same genre, but enough diversity to allow them to function on an array of settings and serve different purposes’ (Freeman et al. 2010, p. 64).

Optimal Development

A consequence of an ECC is that there is a marked improvement in the quality of arguments between the disputants over time. This criterion has met with substantial skepticism (see Collier et al. 2006). Abbey (2005), for example, states that Connolly (1983) presents this information as if it were a gloss on the idea of essentially contested concepts, rather than one of its necessary criteria. Gallie provides two interpretations. In the narrower sense this relates to the achievement of more complete acceptance of the original exemplar. To this extent, the acknowledged seminal contribution from Freeman is sustained and has not fallen out of fashion. Indeed Freeman’s (1984) ‘affects and is affected by’ stakeholder definition is the most widely adopted of all definitions within high quality management journals (Miles 2011) with almost 20% of articles (105/563 definitions identified) providing a definition of a stakeholder adopting one of the 1984 variants from Freeman’s seminal book.

In the broader sense optimal development relates to the progressive clarification of concepts, as demonstrated by Mitchell et al. (1997) in the development of their stakeholder salience model to determine ‘who really counts’ and Kaler’s (2002, 2003, 2006) work on stakeholder identification. Optimal development can also apply to the sharpening of the analytical conceptual framework, the ability to respond to changing circumstances and the incorporation of these new directions whilst building on early foundations. Such developments are clearly discernible trends in the stakeholder literature and, where not yet achieved, are being called for. For a summary of such development over the past 30 years a good starting point would be Friedman and Miles (2006) and Freeman et al. (2010). Both texts provide a sharpening of the analytical conceptual framework: Friedman and Miles through various classifications and ordering of the burgeoning literature and theoretical developments, and Freeman et al., through an examination of the basic tenets of the theory and how these have been applied across a range of business disciplines, calling for a re-framing of capitalism in terms of stakeholder theory.

Conclusions

Drawing upon the work of a renowned philosopher W. B. Gallie, this article has demonstrated, through an exploratory application of his ‘hypotheses’ that the stakeholder concept is an essentially contested concept, rather than merely ambiguous, or confused. The field of stakeholder research suffers from conceptual contestation in which contributors defend their own usage of a concept, whilst others contend that their version and usage is correct, thereby attracting persistent and repeated argument about the use and meaning of the stakeholder concept. In these circumstance, “no amount of discussion can possibly dispel” such head on conflicts (Gallie 1956, p. 169).

Gallie’s framework has, inevitably, undergone some criticism. Collier et al. (2006) suggest that it might be implausible to expect that all criteria are present all of the time. Indeed, Abbey (2005) points to the fact Gallie admits that none of his four examples of art, democracy, social justice and a religious way of life, fulfil his seven criteria perfectly. Rather ‘they approximate them sufficiently for the notion of essential contestedness to explain the roles they typically play in aesthetic, political and religious debates respectively’ (2005, p. 468). Bearing this in mind, this study has demonstrated that there is significant evidence of the existence of all criteria to warrant classifying the stakeholder concept as both essentially contestable and essentially contested. Some authors have rejected or marginalized individual criteria (Connolly 1983), some have proposed alternative conceptions (Swanton 1985; Kekes 1977; Gray 1977), whilst other have rejected the notion of essentially contested as optimistic (Care 1973) or radically mistaken (Clarke 1979), arguing that ‘In view of these inconsistencies it might be useful in future to delete all references either to essentially contestable or essentially contested concepts’ (Clarke 1979, p. 126).

Nevertheless, not withstanding these criticisms, it is widely acknowledged that Gallie’s framework still provides a major set of tools for understanding and analysing concepts (Collier et al. 2006). What is the significance of treating a dispute as involving essentially contested concepts? Garver (1990) argues that a desirable consequence of the recognition of essential contestedness might be an improvement in the quality of arguments, promoting ‘a more sophisticated, more intellectually and morally advanced understanding of one’s arguments and opponents’ (1990, p. 264) and Waldron (2002, p. 142) states that it ‘deepens and enriches our sense of what is at stake in a given area of value’. Such debates can only be achieved through argument. The arguments are endless but are capable of rational resolution (Kekes 1977), but the nature of the disputes means that they cannot be settled ‘by appeal to empirical evidence, linguistic usage, or the canons of logic alone’ (Gray 1977, p. 344).

Gray (1978, p. 392) observes that “the strong normative resonance of essentially contested concepts results in divergent, intractable positions and values, precisely those not conducive to standardization or harmonization through continued debate.” Hence, if a concept is essentially contested there is little chance that a universal singular definition would, or could, be accepted. Different stakeholder definitions highlight core themes and give weighting to components that are relevant to the context and situation in which they are developing or using the construct. It is acknowledged that the stakeholder concept is a complex construct and that it has been widely adopted by a plethora of diverse academic fields of research, different organisational forms and is dynamic, responding to changes in society over time. Consequently, the justifications for defining the concept emerge from heterogeneous needs, whether normative, strategic or simply descriptive or operational.

However, in discussing corporate social responsibility, Okoye (2009) argues accepting that a concept is essentially contested does not entirely occlude the definitional debate as the central core must be identified and defined. As Freeman et al. (2010, p. 208) argue “work needs to be done to pare down and refine what we mean by stakeholders if the term is to prove helpful at a conceptual level or at a practical level.” They go on to argue that “From the standpoint of thinking about the definitional problems as singular and fixed, we face a seemingly intractable difficulty. No matter how many positive attributes any given candidate may have, no single definition seems to work for all purposes in all situations. All have limitations and weaknesses…rather than seeing the definitional problem as singular and fixed, admitting of one answer, we instead can see different definitions serving different purposes” (2010, p. 211). I agree, and propose that whilst a universal definition is not necessary, there is a need to refine the common core that ensures arguments do not become over aggregated. The essence of an essentially contested concept is that there will continue to be various valid conceptions that exist and emerge in the future, but such conceptions need to be refined. Kaler (2003) has gone some way in this direction in terms of his analysis of influencer and claimant definitions, arguing that only claimant definitions will do for the purpose of business ethics, although he does suggest his own conceptualisation, in addition to setting out the distinguishing factors that identify claimant definitions. With an essentially contested concept, arguments about the best instantiation of the construct can be perfected but never resolved. These search for the best instantiation needs to consider the divergent voices with their varying interest and authors need to clearly acknowledge their conceptual basis, so that contributors with similar interests can work towards creating a similar voice. In agreement with Okoye (2009) I suggest that articles which map theories and definitions over time are an invaluable tool in predicting the future direction of a concept. I believe it is fundamental to this process to revisit old arguments and old debates to establish the reasoning behind why authors create particular conceptualisations, to establish the core elements of their conceptualisation and to work towards establishing a common core which can be used to develop better instantiations of the stakeholder concept for specific circumstances.