How Tight Are the Ties that Bind Stakeholder Groups?

The purpose of stakeholder management is to facilitate our understanding of increasingly unpredictable external environments, thereby facilitating our ability to manage within these environments. We argue that a powerful implicit assumption within the stakeholder literature—that priorities within rolebased stakeholder groups are relatively homogeneous—blurs our understanding of organization-stakeholder relationships. Two important and related areas of concern are presented. The first involves the primacy of role in stakeholder definition. This role primacy approach to stakeholder definition is appropriate if, for a particular issue, role-based stakeholder group members have similar priorities. Individual and collective self-interest provides a rationale for this assumption. However, an important problem with this approach arises in situations in which selfinterest is not the primary motivator of individuals’ priorities. In these instances, subgroups within different role-based stakeholder groups might have more similar priorities than either subgroup has with others within their role-based stakeholder group. In these situations the role primacy approach impedes, rather than facilitates, an understanding of our environment. Our second concern is related to insufficient rigor in the application of stakeholder analysis. Most stakeholder studies, both theoretical and empirical, fall short in the determination of relevant interests and the subsequent subdivision of role-based stakeholder groups into rigorously defined specific stakeholder groups. Having suggested that the role primacy approach to stakeholder definition is less than ideal, we examine the extent to which, and the conditions under which, roles are likely to determine priorities, and thus, the likelihood of relatively homogeneous priorities within role-based stakeholder groups. In addition, we present an illustrative empirical analysis of stakeholder group priorities. The illustrative study is conducted within the context of intercollegiate athletics. Related literature and our empirical results indicate that role-based self-interest frequently is not a sufficient “binding tie” of stakeholder groups. Given this background, we present an alternative approach to stakeholder analysis that borrows heavily from the customer segmentation literature of marketing. Our alternative approach can accommodate heterogeneous priorities within role-based stakeholder groups. (Stakeholder Management; Intercollegiate Athletics; Stakeholder Homogeneity; Collective Self-Interest; Symbolic Predisposition) The stakeholder concept is deceptively simple. It is “simple” because it is easy to identify those groups and individuals who can affect, or are affected by, the achievement of an organization’s purpose. It is “deceptive,” because once stakeholders are identified, the task of managing the relationships with them is enormous (Freeman 1984 p. 246; emphasis added). The purpose of stakeholder management is to facilitate our understanding of, and thereby our ability to manage within, increasingly unpredictable external environments. “Given the turbulence that . . . organizations are currently facing and the very nature of the external environment, as consisting of economic and socio-political forces, there is a need for conceptual schemata which analyze these forces in an integrative fashion” (Freeman 1984, p. 40). The desired result of stakeholder management is to more closely align corporate priorities and actions with stakeholder needs. It is hypothesized that creating this alignment produces a good fit between the organization and its environment, thus increasing the probability of the organization’s success. Understanding the priorities of and dealing with identifiable stakeholders—any group that can affect, or is affected by, the achievement of an organization’s objectives (Freeman 1984, p. 46)—offers strategic and cognitive efficiency advantages over conceiving of an organization’s environment as being composed of innumerable individuals and institutions. The cognitive efficiency advantages attributed to stakeholder management derive from its provision of a systematic approach for conceptualizing, comprehending, and RICHARD A. WOLFE AND DANIEL S. PUTLER How Tight Are the Ties that Bind Stakeholder Groups? ORGANIZATION SCIENCE/Vol. 13, No. 1, January–February 2002 65 analyzing external environments. Focusing on relatively few identifiable stakeholders, as opposed to innumerable individuals and institutions, provides a simplified and more easily comprehended representation of the organization’s world. Consequently, the stakeholder approach to management can be considered a knowledge structure that determines how a manager selectively perceives, evaluates, and interprets attributes of the environment. The use of a knowledge structure can facilitate information processing and decision making (Walsh 1995). However, the use of a knowledge structure is not without risk. As argued convincingly by Allison (1971), fundamental choices among the categories and assumptions of knowledge structures channel our thinking and influence our understanding. The often implicit nature of our conceptual models contributes to a lack of cognizance of what they magnify and reveal as well as what they blur or neglect. Allison surmised that it is important for researchers and practitioners to consider the assumptions inherent in the conceptual models they adopt, how these assumptions channel their thinking, and what other perspectives are available. We believe that much of the stakeholder literature is prone to the magnifying, blurring, and/or neglecting described by Allison. Development of stakeholder theory has concentrated on stakeholder analysis—attempts to classify stakeholders into categories that provide an understanding of how stakeholder groups can influence a firm (Rowley 1997). There is agreement in the literature concerning the major steps involved in stakeholder analysis: (1) identification of stakeholder groups (e.g., employees, owners, communities, customers); (2) determination of the stakeholders’ interests; and (3) evaluation of the type and level of stakeholder power (Wood 1994) or salience (Mitchell et al. 1997). We argue that powerful, though implicit, assumptions of the type suggested by Allison influence both the methods used to define stakeholder groups (the first two steps above) and how rigorously these methods have been applied. Our first and primary area of concern involves the methods used to define stakeholder groups in stakeholder analysis. The basis of this concern is the primacy of role in stakeholder definition. As described above, stakeholder analysis begins with identifying relevant role-based stakeholder groups and then determining the relevant interests (in other words the “stakes”) of individuals within each identified stakeholder group. This approach is appropriate if, for a particular issue, members of a rolebased stakeholder group have similar priorities. Not all stakeholder group members would have to have identical interests or perceived stakes; variations in interests are the reason why further refinement (i.e., Step 2) is part of a rigorous stakeholder analysis. Self-interest provides a natural reason to assume that individuals within a rolebased stakeholder group will employ a similar lens when perceiving their stakes, and thus be fairly homogenous with respect to their views on a particular issue. Presumably, so the thinking goes, employees will view their stakes in a particular issue through a lens of their selfinterest concerns with wages and job security; shareholders will view the stakes for the same issue through a lens of their own self-interest concerns with earnings and dividends; and so on. An important problem with what we call the “role primacy” approach to stakeholder analysis arises in situations in which self-interest does not constitute the primary motivator of individuals’ attitudes and priorities. In these instances it may well be that individuals or subgroups, each within different role-based stakeholder groups, have more similar priorities with respect to a particular issue than they have with others within their own role-based stakeholder group. In these situations the role primacy approach adds nothing to understanding our environment, and, in our opinion, actually impedes understanding. The basis for our position is that instead of creating a parsimonious and more cognitively efficient knowledge structure that classifies similar individuals into a single group, individuals with different priorities are classified within their respective (but uninformative) role-based stakeholder groups. Our second concern is related to insufficient rigor in the application of stakeholder analysis. Most stakeholder studies, both theoretical and empirical, fall short in executing the second step of stakeholder analysis: the determination of relevant interests and, when necessary, the subdivision of generic role-based stakeholder groups into rigorously defined specific stakeholder groups. As we present in some detail below, stakeholders are defined solely by role in essentially all of the literature. This is of great importance because stakeholder theory and management are used to postulate and create an alignment between organizational and stakeholder priorities. In order for an organization to align its priorities and actions with those of a stakeholder group, preferences concerning the organization must be relatively homogeneous within that stakeholder group. Otherwise, with what would the organization align its priorities and actions? Inherent in extant stakeholder research, therefore, is the implicit assumption of strong homogeneity of interests concerning an organization within stakeholder groups, and thus within roles. At the heart of both of our concerns is the issue of stakeholder group homogeneity. The relevance and importance of group homogeneity/heterogeneity is well established in the organizational studies literature; lack of RICHARD A. WOLFE AND DANIEL S. PUTLER How Tight Are the Ties that Bind Stakeholder Groups? 66 ORGANIZATION SCIENCE/Vol. 13, No. 1, January–February 2002 sensitivity to it implies important pragmatic problems for researchers and managers. To the extent that homogeneity exists, only a small number of differentiated preferences and expectations within a stakeholder group will impinge upon an organization. Homogeneity thus simplifies organizational activities necessary to satisfy a stakeholder. It is difficult for an organization to respond to the diverse needs and expectations of heterogeneous groups because a diverse set of actions are required (Tsui 1990). Although assuming homogeneity simplifies stakeholder theory, this assumption can have powerful, unanticipated, and undesirable consequences. Given its major influence—movement from management for stockholders to management for stakeholders (Meznar et al. 1994, Wang and Dewhirst 1992)—it is understandable that the focus of the stakeholder management literature has been on heterogeneity across, rather than within, stakeholder groups. Having suggested that the role primacy approach to stakeholder definition can be less than ideal, related questions emerge concerning the extent to which, and the conditions under which, roles are likely to determine priorities, and thus the likelihood of relatively homogeneous priorities within role-based stakeholder groups. To address these questions we turn to the findings of a related literature on the function of self-interest in explaining social and political attitudes and behavior that have developed in political science, social psychology, and sociology. However, before doing this we examine how stakeholder groups are defined in the existing literature, and explicitly examine the extent to which stakeholder classification has gone beyond the use of simple generic roles. In addition, we present an illustrative empirical example to examine several aspects of this issue. The example is a stakeholder analysis of intercollegiate athletics within the context of a university community. The review of the related literature and our empirical results suggest that role-based self-interest is frequently not a sufficient “binding tie” of stakeholder group priorities. In fact, the literature on social and political attitudes and behavior suggests that in many circumstances role-based selfinterest is not likely to be the key underlying driver of priorities and behavior. Given this background, we present a new approach to define stakeholder groups that borrows heavily from the customer segmentation literature of marketing. The paper concludes with a discussion of the implications of our findings and suggests avenues for future research. How Are Stakeholder Groups Defined in the Literature? In a systematic review of how stakeholder groups have been defined in the literature, we find unanimous adoption of the role primacy approach. A few authors have recognized that members of a role-based stakeholder group can have heterogeneous interests and priorities. Freeman (1984) suggests that: each category of stakeholder group . . . can be broken down into several useful smaller categories (pg. 25). . . . [while a] generic stakeholder map . . . can serve as a starting point . . . for stakeholder analysis to be meaningful . . . [s]pecific stakeholder groups must be identified (p. 54). Carroll (1996, p. 82) and Jones (1995, p. 408) make a similar differentiation between generic and specific stakeholder groups; however, our review of the literature indicates that very few studies have gone beyond generic role-based definitions. Consequently, it appears that most authors, probably implicitly, assume homogeneity of interests and priorities within role-based stakeholder groups. Put another way, there is an assumption that the ties that bind role-based stakeholder groups are very strong. Our assessment of the literature is based on a review of five leading general management journals (Academy of Management Journal, Academy of Management Review, Journal of Management, Organization Science, Strategic Management Journal) and of two journals in the social issues in management field (Business & Society, Business Ethics Quarterly). Title and subject indices for each of these publications were reviewed for articles that address stakeholder theory, management, relationships, and so on. Between 1990 and 1999 inclusive, 76 articles that address the stakeholder concept appeared in the seven publications. Of these, 48 are conceptual and 28 are empirical articles. Generic-specific stakeholder differentiation and/or homogeneity (heterogeneity) of interests within stakeholder groups are minimally addressed in the reviewed articles: Nine articles (six conceptual and three empirical) address one or the other of these issues. While mentioned in nine of the 76 articles, homogeneity/ heterogeneity of interests within stakeholder groups is not directly related to the theoretical development, research design, or empirical analysis in any article. Full results of our review can be found at http://www.umich.edu/ wolfer/review.htm . It appears, therefore, that by and large researchers have overlooked the specific-generic stakeholder differentiation as well as the possibility of heterogeneity of interests within stakeholder groups, as “[m]ost of us have a tendency to assume far too quickly that a group has a particular attitude or set of values” (Freeman 1984, p. 132). This tendency leads to the (perhaps implicit) assumption that members within a stakeholder group share homogenous priorities with respect to a focal organization. We RICHARD A. WOLFE AND DANIEL S. PUTLER How Tight Are the Ties that Bind Stakeholder Groups? ORGANIZATION SCIENCE/Vol. 13, No. 1, January–February 2002 67 find that generic role-based terms such as “customers, employees, suppliers, communities, etc.” (Clarkson 1995, p. 106) are those that are used when discussing stakeholder groups, and that these terms are used in a manner that implies that stakeholder groups are homogeneous units. Research Implications of Not Addressing Group Homogeneity. We believe that for stakeholder theory to accomplish its objectives of (1) prescribing how managers should deal with stakeholders, (2) explaining what managers actually did with respect to stakeholder relationships, and/or (3) predicting what would happen if managers adhered to stakeholder management principles (Jones 1995, p. 406), it is necessary to have knowledge of the underlying priorities within stakeholder groups. Consider the case of United Way of America (UWA) during the aftermath of allegations of financial abuse, and the resignation of UWA President William Aramony (Rowley 1997). How could one prescribe, explain, and/ or predict UWA actions relating to local chapters without knowing the chapters’ preferences concerning autonomy, and the extent to which these preferences differed across chapters? Likewise, how could one prescribe, explain, and/or predict UWA actions relating to donors without knowing donors’ preferences concerning local (versus national) handling and disbursement of donations, and the extent of agreement concerning these preferences? Mitchell et al. (1997, p. 854) present a “theory of stakeholder salience that can explain to whom and to what managers [should] pay attention.” The authors depict Alaskan citizenry as an example of a salient stakeholder of Exxon subsequent to the Exxon Valdez oil spill. While Alaskan citizens are salient stakeholders according to the authors’ three criteria for salience (possession of power, legitimacy, urgency of claim), those attributes, in and of themselves, do not “explain to whom and to what” Exxon managers should pay attention. Alaskan citizens’ interests could have been very heterogeneous concerning such issues as timing and amount of compensation for the spill, reclamation of wildlife habitats, and pollution controls required prior to resumption of oil shipping. Prescribing, explaining, and/or predicting Exxon managers’ actions would depend on the extent to which Alaskan citizens had common priorities concerning such issues. The Nature of Stakes and the Role of Self-Interest The Relevance of “Stakes.” Stakeholder theory is based on the concept of “stake,” or “interest” (Freeman 1984, p. 60). Though it is central to stakeholder theory, discussion and development of the concept of “stake” subsequent to Freeman (1984) is minimal. Consequently, a primary challenge to further developing stakeholder theory is the development of a broadly acceptable definition of the term “stake” (Rowley 1997). This is of importance to us because stakes are what motivate stakeholder groups, and thus are important determinants of stakeholder group priorities and the degree to which members of a stakeholder group are likely to have common priorities with respect to a given issue. To the extent that the concept of “stake” is addressed in the literature, it is done in a manner consistent with Freeman (1984, p. 60) who uses the term interchangeably with “interest.” An author who has expanded on the concept is Carroll (1996, p. 73): To appreciate the concept of stakeholders, it helps to understand the idea of a stake. A stake is an interest or a share in an undertaking. . . . The idea of a stake . . . can range from simply an interest in an undertaking at one extreme to a legal claim of ownership at the other extreme. In between these two extremes . . . might be a legal right . . . (or) a moral right. (emphasis