What We Talk About When We Talk About Stakeholders

Data Collection

We collected data using the Web of Science (a product of Clarivate Analytics Web of Knowledge® database), which includes the Science Citation Index-Expanded (SCI-Expanded), the Social Science Citation Index (SSCI), and the Arts and Humanities Citation Index (A&HCI). This facilitated the retrieval of articles based on the search constraints discussed earlier and yielded a global dataset that was further segregated into smaller datasets based on journal type and sub-periods. The resulting article counts are summarized in Table 1.

OPEN IN VIEWER

Table 1.

Number of Articles in the Dataset.

Journal	1984–1998	1999–2008	2009–2020	Total
AMR	8	18	13	39
AMJ	0	7	13	20
SMJ	0	7	24	31
Management (Total)	8	32	50	90
JBE	18	84	198	300
BEQ	0	27	26	53
BAS	0	5	29	34
SIM (Total)	18	116	253	387
Total	26	148	303	477

Source. Web of Science.

Note. AMJ = Academy of Management Journal; AMR = Academy of Management Review; BAS = Business & Society; BEQ = Business Ethics Quarterly; JBE = Journal of Business Ethics; SIM = social issues in management; SMJ = Strategic Management Journal.

Initially, a limited number of Business & Society articles featured in the first and second periods of interest led us to look further into the discrepancy. We attribute this limitation to the lack of Business & Society entries in the Web of Science prior to its inclusion in 2008. To analyze the Business & Society publications, we manually extracted the entries for the articles with “stakeholder” in their title in the SAGE Publications database and manually populated the relevant details. The number of non-Web of Science Business & Society articles are tabulated. It should be noted that we chose to integrate the Web of Science datasets in Table 1 with the manually extracted Business & Society dataset in Table 2 to generate a more representative view, especially given our interest in this journal’s distinctive role in the field’s development.

OPEN IN VIEWER

Table 2.

Number of Articles in Business & Society.

	1984–1998	1999–2008	2009–2020	Total
BAS	6	25	33	64

Source. SAGE Publications.

Note. BAS = Business & Society.

Among the parameters in the dataset (e.g., author[s], title, publishing year, journal, abstract, etc.) is the list of references cited in the said articles. These permitted the next step of the analysis. There were, however, inconsistencies in the conventions adopted in the dataset by different journals over time. As bibliometric software recognizes only exactly matching strings of characters, a manual normalization process is required to maintain accuracy, especially in the spelling of cited authors’ names. For example, references to R. Edward Freeman’s cited publications in the dataset appeared in multiple variations depending on the references and representation styles (e.g., “Freeman, R.E.,” “Freeman, RE,” “Freeman, R.,” “Freeman, R”). Thus, the normalization entailed using unified naming conventions within the dataset to avoid misrepresentations in the frequency of citation or co-citation.

The resulting dataset included 21,521 cited publications in 508 articles. Broken down according to the periods identified earlier, Phase 1 comprised 32 articles published in the journals between 1984 and 1998. Phase 2 included a total of 168 articles published between 1999 and 2008. Phase 3 included 308 articles published between 2009 and 2020.

Data Analysis

Bibliometric analysis typically occurs in two stages (Nerur et al., 2008; Ramos-Rodríguez & Ruíz-Navarro, 2004; White & McCain, 1998). In this study, the preliminary stage comprised a citation analysis to identify those publications that have been cited most frequently by the articles within our dataset. The purpose of this step is to highlight influential ideas in the development of specific periods of stakeholder theory. In the second step, co-citation analysis identified dyadic (and subsequently higher order network) relationships between and among cited references.

To illustrate this form of analysis, a co-citation occurs between two works, say Freeman (1984) and Rowley (1997), when an article in the dataset, published subsequently, includes both works in its list of references. Across the dataset, the total number of co-occurrences of Freeman (1984) and Rowley (1997) represents their co-citation frequency and is presented graphically in network maps. At this aggregate level, one can identify associational themes or clusters of works that are frequently cited together. To analyze these patterns, we used VOSviewer, a bibliographic mapping software program that creates co-citation lists based on text mining of the reference lists of the articles within the dataset. VOSviewer has been increasingly used in research (Barnett et al., 2020) and provides functionality through which it generates co-citation network maps. These maps enable the visualization of the intellectual structure of a discipline. They also allow us to see changes in how different publications have been invoked over time (Ding et al., 2001; Linnenluecke et al., 2020).

Network maps communicate three important features: the size of the nodes, the links between the nodes, and the color-coded clusters:

To continue our illustration from above, consider again Freeman (1984). Unsurprisingly, it is the most commonly cited reference in the dataset, and it has been co-cited repeatedly along with other publications. As a result, the node representing Freeman (1984) is among the largest nodes in the network map. The lines between Freeman (1984) and Rowley (1997) as well as their proximity correspond to the frequency with which articles in our dataset cite both works in conjunction. Finally, in each map, these two citations are colored according to their membership in the clusters. This is to say that these citations are more closely associated with similarly colored nodes (i.e., contributions in the same cluster) and less closely associated with different colored nodes (i.e., contributions in distant clusters). It is very possible that these two citations will appear in different colors (clusters) or in a different relationship with each other across the time periods, as the stakeholder theorists citing them ascribe different meanings to them and, thus, associate them with different themes.

Using these maps and the results of the co-citation analysis, we conducted an iterative analysis to identify similarities, patterns, and differences. Individual observations were resolved in weekly conferences to yield interrater consensus (Cascio et al., 2019). Our interpretation of these results forms the basis for the next section.

Consolidated Journals

Phase 1 (1984–1998)

The first network map (Figure 1) isolates articles with five or more citations in the first phase. This relatively low-density network foreshadows nascent patterns that emerge in later phases. We see, for example, the early prominence of scholars who go on to define the associational themes in future periods. During this 14-year period, scholarly conversations about stakeholders were defined by works from the pioneers of the field. Carroll (3), Freeman (2), Jones (2), and Preston (2) are collectively responsible for 9 of the 14 most influential works in this first phase. Not represented on this image are two other Freeman works with four co-citations each—just below the cutoff for inclusion in the map. This concentration is not unusual and accords with the basic assumption that new ideas are frequently associated with institutional entrepreneurs (Battilana et al., 2009).

OPEN IN VIEWER

Figure 1.

Consolidated journal output for Phase 1 (publications with 5 or more citations).

We also note that corporate social responsibility (CSR) and stakeholder theory are closely related concepts in this early stage. While the two associational themes (clusters colored red and green) remain only loosely defined, they hint at an emerging separation between those references associated as contributing to stakeholder theory defined in itself (in se) and those advancing the stakeholder concept as a handmaid construct for further development of CSR. Although Freeman (1984) explicitly attempts to disentangle his work from CSR (Elms et al., 2011), early readers were having none of it and associated Freeman (1984) with key works in that domain. We further note that Freeman and Preston each found their works contributing to both conversations and that Freeman and Gilbert (1988) is the only node in the green network that was not published in AMR. ³

Finally, these nascent associational themes—both the stakeholder theory in se and stakeholders qua CSR—remain largely conceptual. Only one of the studies influential in the foundational phase is quantitative (Aupperle et al., 1985). This pattern of emergence is not unusual in institutional terms. After all, CSR had a well-established history in the field. Emerging logics rarely replace existing ones fully or instantly; instead, they emerge in relation to them, prompt a renegotiation of identities, and coexist in ideological tension and symbiosis (Thornton et al., 2012).

Phase 2 (1999–2008)

In the second phase (Figure 2), we witness the rising influence of a number of additional papers published in AMR (Frooman, 1999; Jones & Wicks, 1999; Mitchell et al., 1997; Rowley, 1997). These nodes join Clarkson (1995), Donaldson and Preston (1995), and Jones (1995) in positions of centrality in the network. The impact of that journal (AMR) on the trajectory of stakeholder theory is clear, as seven of the eight most frequently cited works in this phase were in AMR. Although the influence of AMR as a venue for stakeholder theorizing was already present in Phase 1, its role in Phase 2 appears to suggest that the underlying causal arguments being advanced around stakeholder theory were assuming a level of rigor and clarity to reach the most selective outlets for conceptual work in the field with increasing frequency. Also of note is the diminished role of three other AMR publications—Wood (1991), Carroll (1979), and Swanson (1995)—indicating that the differences in influence are not merely artifacts of the journal’s reputation or status. Rather, stakeholder theorists publishing in Phase 2 appear to ascribe a growing conceptual separation between stakeholder theory and CSR (in contrast to Phase 1). We consider the three associational themes:

OPEN IN VIEWER

Figure 2.

Consolidated journal output for Phase 2 (publications with 15 or more citations).

Theme 1: The question of integration

The smallest of the three clusters appears to reflect the impact of Jones and Wicks (1999) and the “Dialogue” that accompanied it. Five of the seven papers in this cluster are from that same AMR issue. Jones and Wicks (1999) argue in favor of an integrative approach to stakeholder theory that incorporates both social science- and ethics-based arguments. This impulse (and the debate it gives rise to) also represents a new, stakeholder-centric way of framing the old logic of the relationship between social and financial performance. Even as CSR theorists begin to move to the network’s periphery, they (Carroll, 1979; Waddock & Graves, 1997; Wood & Jones, 1995) continue to cluster as a conversation among themselves. Also at issue in that dialogue was the gatekeeping role of ethics in stakeholder theory and whether stakeholder theory can bear the mantle of theory absent being tethered to another existing theory of management. Conspicuously, Frooman’s (1999) article immediately precedes Jones and Wicks (1999) in the same issue of the journal. Nonetheless, the editor did not consider it part of the dialogue, and it is not associated with the question of integration by subsequent theorists. It appears far off in the green cluster in the Phase 2 map. The article explicitly draws on resource dependence theory (Pfeffer & Salancik, 1978) and avoids reference to the moral status of stakeholders.

Theme 2: Adopting a sociological approach

The second cluster (green) most prominently anchors around Mitchell et al. (1997). Their theory of stakeholder identification and salience, like many of the co-citations nearby, prominently features extant theories from organization theory and sociology as applied to stakeholder-relevant questions. Where the meta-theoretical conversation described in the first theme is concerned with the relationship between ethics and social science, this cluster glosses over that concern and draws alternately on institutional theory (Mitchell et al., 1997), resource dependence theory (Frooman, 1999), and social network analysis (Rowley, 1997). This cluster also features direct co-citations with sources in those underlying domains (Pfeffer & Salancik, 1978; Suchman, 1995). The interpretation of this theme also draws strength from the inclusion of Berman et al.’s (1999) quantitative test of stakeholder theory. Many features of that study suggest a close relationship with Theme 1. It derives its hypotheses from Jones’s (1995) propositions and represents an early use of Kinder, Lydenberg, & Domini (KLD) data to test those hypotheses. Nonetheless, its reliance on adapted social science theory without explicit reference to ethical theory lands it squarely in the green cluster. (In Phase 3, Berman et al., 1999 take on a different structural role altogether.)

Of course, Theme 2 is also defined in contradistinction to the third cluster (red). Here, we begin to see stakeholder theorists making a clear distinction between work largely focused on the nature and behavior of stakeholder groups (Theme 2) and those in the other two clusters continuing Freeman’s (1984) emphasis on the firm, its managers, and its approach to stakeholder management. This is not to imply that these works ignore managers; indeed, Mitchell et al. (1997) explicitly theorize about how managers will prioritize stakeholder claims. However, all of the notable appeals to firm-centric theory (Jones, 1995) or managerialism (Donaldson & Preston, 1995) are more closely associated with work outside Theme 2.

Theme 3: Ethical stakeholder management

The third cluster of associations contains important normative contributions (Goodpaster, 1991; Phillips, 1997). In addition, it contains co-citations with foundational works of ethics, political philosophy, political economy, and economics (Friedman, 1962; Jensen & Meckling, 1976; Rawls, 1971; Williamson, 1985). Here, as in Phase 1, we observe, within the narrower confines of this associational theme, the same concentration around institutional entrepreneurs. More than half (seven out of twelve) of the contributions in this cluster were (co-) authored by Freeman or Donaldson. Donaldson and Dunfee’s (1994) work on “integrative social contracts theory” contributes prominently to this theme as does Freeman’s own normative work (Freeman, 1994; Freeman & Gilbert, 1988).

Taken together, the three clusters of Phase 2 suggest similar moves afoot to legitimate stakeholder theory’s cognitive frames and normative claims. Each associational cluster represents a common attempt by subsequent authors during the period to associate stakeholder theorists and their work with a more established discipline: CSR, sociology, and philosophy. Yet, at the same time, as the works in the integrative (red) cluster make clear, stakeholder theorists continue to advance synthetic arguments that no one fundamental discipline would or could. Of course, these disciplines also make demands in return. For example, work geared toward more practitioner audiences (e.g., through advocacy of the so-called Clarkson Principles; Clarkson Centre for Business Ethics, 1999; Donaldson, 2002) narrowly but conspicuously falls below our minimum co-citation threshold during this period. Within the nested subdivisions of the growing field, the institutional work of outreach to practicing managers became secondary, crowded out by the rigors of academic legitimacy-seeking.

Phase 3 (2009–2020)

The network map of co-citations in the third phase (Figure 3) also divides into three clusters. Two of these three associational themes are recognizable as evolving from the prior period, but there are important differences. The first is the marked decrease in citations from philosophy. Works by Goodpaster, Rawls, and even Donaldson (outside of Donaldson & Preston) no longer number among the commonly cited publications. Those that remain (Freeman, 1994; Phillips, 2003a) appear in a somewhat different structural role. The second difference is the growing presence of mainstream strategic management scholars. The Strategic Management Society formed an interest group concerned with “Stakeholder Strategy” in 2010, and the Strategic Management Journal becomes an increasingly important outlet for stakeholder-native work during this phase. Hillman and Keim (2001) and Waddock and Graves (1997) continue to influence the conversations, and Eesley and Lenox (2006), Bosse et al. (2009), and Harrison et al. (2010) are also widely cited.

OPEN IN VIEWER

Figure 3.

Consolidated journal output for Phase 3 (publications with 30 or more citations).

Theme 1 (Blue): Stakeholders and KLD

The first cluster from the third phase features quantitative studies of the corporate social performance–financial performance relationship—largely framed in stakeholder theoretical terms. Among these studies Berman et al. (1999) emerges as the most widely cited and prominently co-cited. Although a relatively small cluster, this theme presents a curious collection of overlapping observations. Three of the five studies in this conversation employ data from the KLD database, and two of the five explicitly acknowledge normative claims. (Orlitzky et al., 2003, lists “business ethics” and “stakeholder theory” as keywords, but neither ethics nor norms are discussed in the meta-analysis.)

Theme 2 (Green): Focus on stakeholder behavior

This cluster is the most stable of the clusters in the transition from Phase 2 to Phase 3. Anchored once again by Mitchell et al. (1997), 7 of the 11 works in this associational theme carried over from the prior phase. The centrality of two nodes—Agle et al. (1999) and Eesley and Lenox (2006), both quantitative studies that motivate their studies with reference to Mitchell et al. (1997)—suggests that the main difference between Themes 1 and 2 is not empirical versus theoretical but rather a focus on firm behavior as the independent variable (blue) versus a focus on stakeholder behaviors and claims. Another feature of this cluster is its relative disconnection from those works advancing the normative claims of stakeholder theory. In most cases, even where our maps distinguish clusters, there is a strong connective pattern across cluster boundaries. In this map, Theme 2 stands out in the relative lack of ties to the other two themes, both of which press for a deeper connection between normative and empirical work.

Theme 3 (Red): Taking integration seriously

Perhaps the most fascinating transformation among the three associational themes is the emergence of this third (red) cluster in Phase 3. This cluster suggests new patterns of association based neither on the fundamental discipline nor on the target audience (i.e., philosophers and sociologists) as in the prior phase. Rather, what the more prominent works that populate this cluster share is a common commitment to Jones and Wicks’s (1999) integrative project. (The critical responses from that Dialogue have disappeared altogether.) Having rejected the separation thesis (Freeman, 1994; Wicks, 1996), this cluster associates the normative claims of Freeman (1994) and Phillips (2003a) with the strategic and instrumental claims of Jones (1995), Coff (1999), and others. Indeed, this cluster is the least explicitly dependent on prior, non-stakeholder-native theorizing than any other. Many of the contributions represent a natural synthesis of the normative and cognitive by a new generation of pro-integration scholars. For example, we note Phillips’s influence on this conversation. He figures in the production of 5 of the 17 works in this cluster (Bosse et al., 2009; Freeman, 2011; Harrison et al., 2010; Phillips, 2003a; Phillips et al., 2003) with four of his five works in this cluster representing collaborations with social scientists (strategists). In sum, we contend that the resulting associational theme might well represent the logical outcome of the original stakeholder theory in se logic which originated in Phase 1.

There is, however, a twist. The only nodes in this cluster that are not native to stakeholder theory are Barney (1991) and Jensen and Meckling (1976). This is consistent with the increasing rapprochement in Phase 3 between stakeholder theory and mainstream strategic management. While some scholars in this cluster may have long believed that their conversation was about the strategic management of organizations—recall the title of Freeman’s (1984) book—only recently has this view come to be widely shared in the field. This demonstrates the value of co-citation structures over traditional citation counts. After all, somewhere in the more than 44,000 citations of Freeman (1984) are economists citing the concept; however, only with the growing co-citation of stakeholder theory in se and mainline strategic management contributions does the association become a meaningful part of the literature, laying the foundation for further work in this theme (e.g., McGahan, 2021).

Business & Society

Having traced the long story of stakeholder theory in the consolidated maps, we return to one of our original framing questions: What is the unique role of Business & Society in the development of stakeholder theory? We trace the evolution during the same three periods.

Phase 1

This map contains only one unitary network (see Figure 4). Comparing it back to the consolidated Phase 1 map, we see echoes of its two distinct themes. Many of the more prominent nodes are equally prominent here. There are directional subtleties, such as the close clustering of Wood (1991), Freeman (1984), and Carroll (1979) along one side of the map. This proximity suggests that Freeman (1984) stands, in the minds of authors publishing their work in Business & Society during the period, in direct association with two articles expressly about corporate social performance. Also, interesting is the very different position for the so-called stakeholder theory in se citations: Jones (1995), Donaldson and Preston (1995), and Clarkson (1995). Their centrality in the Business & Society-specific map suggests the relatively quick acceptance of stakeholder thinking in this journal. Finally, there are four unique articles that are not on the consolidated Phase 1 map: Starik (1995), Waddock (1996), Carroll (1995) and Wood and Jones (1995). These can be viewed as suggesting a nascent conversation about alternative “normative cores” for stakeholder theory (Donaldson & Preston, 1995; Freeman, 1994) and in other meta-theoretical discourse. In sum, the first phase co-citations from Business & Society might, at first glance, appear amorphous. However, we suggest that the relative positioning of CSR-, in se-, and normative-relevant citations foreshadow substructures that would ultimately become the clusters in Phase 2 of the consolidated map.

OPEN IN VIEWER

Figure 4.

Business & Society output for Phase 1 (publications with 3 or more citations).

Phase 2

This map seems to have a more direct correspondence with the consolidated map of the same period (see Figure 5). Many of the same associations populate what is the red theme in this map and green in the consolidated; these include Frooman (1999), Rowley (1997), Mitchell et al. (1997), Suchman (1995), and several others. This theme is similarly oriented to prior theorizing in organization theory and sociology. It does, however, offer an interesting twist on the familiar story of assimilation and appropriation common to the development of institutional logic (Thornton et al., 2012). Unlike fields in which borrowing reflects the infusion of new associations from the outside into a monolithic logic, we suggest that the presence of these associations in works intended for Business & Society suggests that the audience was more, rather than less, heterogeneous. In other words, because this journal’s readership was so interdisciplinary, authors had to make the links between stakeholder theory and fundamental work even more explicit. Thus, in the Phase 2 map for Business & Society, we find inclusion of even more overtly sociological works (Aldrich & Fiol, 1994; DiMaggio & Powell, 1983; Oliver, 1991) that are missing from the same conversation on the consolidated map. This prerogative seems to take on a greater level of urgency in Phase 2 Business & Society work, considering that in Phase 1 the map includes no citations at all to work outside of the domain.

OPEN IN VIEWER

Figure 5.

Business & Society output for Phase 2 (publications with 5 or more citations).

Phase 3

Here, we see a growing disconnect between the association themes in Business & Society and those in the consolidated journals. What we describe earlier as the “Stakeholder Behavior” theme (anchored by Mitchell et al., 1997) fragments across all three clusters in the Business & Society map (see Figure 6). Some of these citations find a more coherent association with relevant pieces. For example, Business & Society articles associate Pfeffer and Salancik (1978), Rowley (1997), Frooman (1999), and Eesley and Lenox (2006) more directly with Oliver (1991) and den Hond and de Bakker (2007), neither of which figures in the consolidated Phase 3 map. Articles in this cluster feature discussions of power between the firm and its stakeholders as a unifying theme and prefigure a great deal of later work on stakeholder influence and activism.

OPEN IN VIEWER

Figure 6.

Business & Society output for Phase 3 (publications with 5 or more citations).

Similarly, where the consolidated map contains a rather wide-ranging red cluster (Theme 3), the red cluster in the Business & Society map contains Freeman (1984), Phillips (2003b), and Harrison et al. (2010), among several others, but also closely associates these with DiMaggio and Powell (1983), Freeman et al. (2010), and Mitchell et al. (1997). What ties these together is a bit more subtle: a shared resort to the concept of legitimacy. At some risk of oversimplification, one could interpret “legitimacy” in the hands of the sociologically inspired as an effort to import norms and ethics while maintaining scientific purity. To wit, Mitchell et al. (1997) rely on Suchman’s taxonomy of legitimacy but juxtapose it with “power”—contrary to Suchman’s own elaboration (Phillips, 2003a). Their recognition that norms matter when identifying who is a stakeholder shares a place in this cluster with normative theorizing about why those claims might be morally legitimate.

What are we to make of these subtle differences where Business & Society is concerned? These differences hint at the possibility that articles published in this journal are engaging in slightly different conversations than those present in the broader domain of management journals that helped to shape the consolidated map. These conversations permit a more focused examination of constructs like stakeholder legitimacy. They also permit more nuanced and specialized conversations, like that of stakeholder behavior, than can be found in the broader management journals. As discussed in the next section, this introduces important questions about future research and the role Business & Society can play in shaping the ongoing development of stakeholder theory.

A Return to Managerialism

One of the central challenges inherent in institutional change is the possibility that those agents who would create change are themselves embedded in the established logics that prescribe the status quo (Friedland & Alford, 1991). This notion of embedded agency (Battilana, 2006; Seo & Creed, 2002) applies to both managers and to the stakeholder theorists who would urge those managers to treat stakeholders with respect. In either case, the question arises: where, in what form, and to what extent do those trying to alter how firms treat their stakeholders have the agency to effect change? This question crosses levels of analysis because the micro-foundations of stakeholder influence, the contest among organizations, coalitions and social movements, and the gaps between and among institutional logics at the macro-level all come to bear on whether and how change occurs (Thornton & Ocasio, 2008).

The nature of managerialism is not a new concern for stakeholder theorists. One interesting storyline in the evolving co-citation patterns around Donaldson and Preston (1995) involves its place in the network relative to which clusters accept or question the possibility of agency. Often overlooked by readers is the thesis that stakeholder theory needs to be “managerial,” in the sense that it “does not simply describe existing situations or predict cause-effect relationships; it also recommends attitudes, structures, and practices that, taken together, constitute stakeholder management” (Donaldson & Preston, 1995, p. 67; see also Phillips et al., 2019). Attention to stakeholders should permeate the policies and judgments of those whose decisions alter the relationship between firms and stakeholders. While clusters of research have ascribed more or less agency to managers, there is more research to be done.

We imagine this taking at least two possible directions, each of which is rooted in some thread of extant stakeholder theory. The first would explore the extent and boundaries of the managerial agency. This is the question of managerial discretion. As Phillips et al. (2010, p. 181) argue, stakeholder theory “more richly describes stakeholders’ role in shaping managerial actions” relative to other perspectives. Critics have argued that managerial discretion as a construct has not advanced far beyond this point in the years since Shen and Cho’s (2005) distinction between managerial discretion as the latitude of actions and objectives (Parker et al., 2019). Elaborating on what “managerial” means for stakeholder theory in light of recent empirical research on managerial discretion (Gupta et al., 2019; Sandhu & Kulik, 2019; Wangrow et al., 2015) is an important avenue for future research.

The second possible direction would explore the individual- and group-level dynamics nested within firm–stakeholder interactions; it would examine how these factors facilitate or constrain the efficacy of the stakeholder ideology in practice. Our bibliometric analysis suggests an under-explored gap here, although we suspect that, when these analyses are replicated in the future, the micro-foundational work of Bridoux and Stoelhorst (2014, 2016) will prove seminal in bridging the gap. What comes of this strand of managerialism will depend on how rich the co-citation network becomes between these citations and new subdomains in social psychology and behavioral economics. A richer interchange with psychological studies on cognition, personality, and values would relax behavioral assumptions that constrain or catalyze whether and how firms manage for stakeholders. For example, questioning the micro-foundational assumption of managerial self-interest sheds light on how fairness judgments and distributive justice shape firm–stakeholder interactions (Bosse et al., 2009; Bosse & Phillips, 2016). Likewise, how do assumptions about risk aversion among managers or the disutility of work among stakeholder group members shape the relationships between firm–stakeholder interactions, competitive advantage, and stakeholder value creation?

Future research might also connect stakeholder theory with the study of group behaviors and process (Felin et al., 2015; List & Pettit, 2011). Surely, the fact that many impactful decisions about stakeholders are made by groups occasions a role for group dynamics in shaping stakeholder-related outcomes. What group processes affect whether managers treat stakeholders with dignity or instill intrinsic stakeholder cultures (Bundy et al., 2013; Jones et al., 2007)? And how would this reshape how we think about processes for stakeholder management and engagement? In either case, the end result of such work might be a better micro-foundation for stakeholder theorists’ understanding of managerialism.

Toward a Vital Normative Vision

From its earliest phases, the field (and in particular, Business & Society) has welcomed philosophical arguments on normative stakeholder theory. Important contributions came with some regularity (Derry, 1999; Gilbert, 1995; Wicks, 1996), and the early ethics-based contributions in Business & Society during Phase 1 presaged the heyday of the normative theme in Phase 2 across all journals. Today, it appears that many active ethicists in the domain have moved on to equally worthy undertakings: intensifying the exchange with mainstream strategy (Harrison et al., 2010), propagating stakeholder ideology in more practical terms (Freeman, Parmar, & Martin, 2020), and contextualizing the notion of value at a broader level (Donaldson & Walsh, 2015). Yet, there is still work to be done. Guided by the institutional perspective adopted at the outset, we contend that the creation of a compelling new institutional logic for business occurs, in no small part, because it marries the cultural-cognitive features of an ideology with explicitly normative claims. What invigorates stakeholder theory is its consistent challenge to moral consensus. Consider these possible lines of normative inquiry.

There remains a great deal of debate on whether stakeholder theory has or has not arrived at a normative consensus (Freeman, 2017; Ingerson et al., 2015). The Phase 2 consolidated map highlights the primary philosophical influences on stakeholder theory as Kantian and Rawlsian (Agle et al., 2008; Freeman & Phillips, 2002). Cultivating a sufficient range of alternatives might take several forms. One would be the continued search for alternative justifications to contrast with the growing libertarian consensus. Is there, for example, value in advancing a more communitarian account of a firm’s stakeholder obligations (Ray et al., 2014)? Alternatively, while the current appeal to universal moral claims on human behavior is attractive to foundationalist philosophers, they might be less compelling as a spur to ethical treatment of stakeholders in practice. Donaldson and Dunfee (1994) argue persuasively that local community norms often exert a powerful influence on business behavior, and stakeholder research offers an ideal stepping stone from the present preoccupation with the universal to a more immediate and particular worldview (McVea & Freeman, 2005). Thus, we might ask: What other sources of normativity (e.g., community standards, patriotism, culture, species membership) might prove more compelling as the basis for stakeholder influence in the increasingly heterogeneous field of actors whom stakeholder theorists seek to influence?

Ultimately, the most generative basis for challenging moral consensus around stakeholders may be to reject foundationalist notions altogether. A growing body of work invokes pragmatism in connection to stakeholder theory (Godfrey & Lewis, 2019; Parmar et al., 2015; Wicks & Freeman, 1998). The upshot, captured best by Pouryousefi and Freeman (2021), is that a pragmatist turn in stakeholder theorizing might, following Rorty (1998, 1999), consist of conversations intended to try normative claims about firm–stakeholder interactions, weigh the relative usefulness of holding to such claims, and frame this enterprise in service to values. Viewed as the basis for future research, stakeholder pragmatism (or pragmatic stakeholderism) moves closer to the experimentalism of practice and interrogates potential future systems for human enterprise: How might we bring managers’ individual moral practice more nearly into line with humanistic and democratic values such that daily practice and momentous decisions might both incline toward their advancement (Dmitrieva et al., 2020)? How might stakeholder theorists engage more closely with conversations on all manner of business topics (from the nature of employment to sustainable supply chains) that are only rarely labeled moral deliberations?

A second path for reinvigorating this debate leads through a parallel conversation concerning the processes of engagement and influence in firm-stakeholder relationships (de Bakker & den Hond, 2008). There is a persistent intuition in the literature that firm–stakeholder processes can and should amount to more than sequential political moves (Calton & Payne, 2003; Crane & Livesey, 2003). Drawing on Habermas (1984, 1987, 2015) and his theory of communicative action, a growing cadre of primarily European scholars has sought to give some normative content to this procedural intuition. We expect that work building on the insights of Scherer and colleagues (Patzer et al., 2018; Scherer & Palazzo, 2007; Scherer & Patzer, 2011) will comprise a significant new branch of normative stakeholder debate. As with pragmatism, the ground (i.e., summarizing and invoking Habermas as the basis for future work) has been broken. The current measure of a contribution requires substantive engagement with more specific questions: What would it look like to hold up managerial practice and stakeholder engagement processes to the standard of communicative action and democratic deliberation? What are the limits of deliberative democratic norms within and outside formal governance processes relative to stakeholders? How might these expectations shift as managerial responsibilities are contextualized in relation to the proper function of macro-societal and political deliberation (cf. Habermas, 2018)? Moreover, if we evaluate responsible treatment of stakeholders relative to managers’ ability to engender discourse, what descriptive work on firm–stakeholder deliberations would produce the necessary insights into current practice to ground normative critiques?

Third, while much has been written about the consensus around managerial responsibility, the field has paid less attention to stakeholder responsibility. Notably missing from our co-citation analysis is Goodstein and Wicks (2007). Cited 170 times since its publication, that article raises the possibility that normative stakeholder theorists might have as much to say about the moral conduct of stakeholders themselves as participants in the cooperative scheme as they do about firms’ obligations. Building on this insight, Freeman and Elms (2018) write,

The stakeholder approach aims to create a new narrative about business—a new story—that enables great companies to make our communities and our lives better through the creation of stakeholder value, rather than simply profit to shareholders. The story includes a recognition that if we want the outcome of business to be a more responsible capitalism, it requires stakeholders to value business responsibility.

Specific questions might include: What would normative stakeholder theory look like if we accept that moral conduct may be the interaction effect of both parties observing moral norms? How should stakeholders value responsible business behavior? What influence tactics should stakeholders refrain from? What are the implications of breaching the strictures of stakeholder responsibility (Martin & Phillips, 2021)? To what degree do stakeholder responsibilities vary based on industry-specific practices (Elms & Phillips, 2009)?

Finally, students of the normative realm would do well to remember that, while the field occasionally conflates normative theorizing with solely philosophical arguments, stakeholder theory requires additional social science-based inquiry on norms as well. Here, a natural disciplinary connection is with anthropology, a field rich with interpretive understandings of normative systems. In particular, we highlight one potential stepping stone to a richer dialogue between these two fields: the perceptual mapping work of Doh and Lucea (2013; Lucea & Doh, 2012). Utilizing techniques first developed for cultural and cognitive anthropology (Bernard, 2017; Weller & Romney, 1988), they demonstrate the power of such tools for mapping perceptual differences within and across groups of companies and their stakeholders. To build on this foundation, future researchers might inquire about how normative preferences, expectations, and values engender similar structures and map across stakeholder networks. Adopting these more systematic techniques for understanding culture would help ethicists to train their critique on widely held but largely implicit normative assumptions that constitute an in situ approach to stakeholder culture (Jones et al., 2007), remembering that some cultural norms may not be supportive of stakeholder theory’s insistence on inclusion and dignity.

A Place in History

Any account of the development of a field over time quickly runs up against the role of history. In no small part, the institutional logics we are describing and altering through stakeholder theory are the residue of historical forces and events, yet we have only started to scratch the surface on how these processes unfold (Stutz & Sachs, 2018). The adoption even of the rudimentary principles of stakeholder theory by practitioners has been a long time in the making, and the evolution of individual companies’ approaches to stakeholder management seem equally time-bound (cf., Royal Dutch Shell in Nigeria). One fruitful direction for future research would explore alternative conceptions of the time-cycles within which firm–stakeholder dynamics unfold. We have ample evidence of the potential for observing firm–stakeholder phenomena over time (Crilly et al., 2012; Shropshire & Hillman, 2007). Expanding the scope, we ask: What can today’s managers and scholars learn about stakeholder relationships from studying historical cases? How does history shape the complex patterns of mutual perceptions held by managers and stakeholders (Lucea, 2010)? How does shared historical memory (or conversely, shared forgetting) affect the nature of firm–stakeholder collaboration (Mena et al., 2016; Schrempf-Stirling et al., 2016)?

Phillips et al. (2020) present one way to organize our understanding of the past to inform stakeholder scholarship. In considering the relationship between history and CSR, they isolate:

Stakeholder researchers may employ a similar taxonomy to consider, as we have done here, the past of stakeholder theory as a concept (“past-of-stakeholder theory”). Future research might also examine the “past-in-stakeholder theory.” How do past stakeholder practices affect current stakeholder relationships in specific companies and industries? How far back can and should managers look for sources of the perceived limits on their managerial discretion? What historical legacy is catalyzing or dampening current organizational initiatives? How long is the shadow of today’s decisions on the future?

The Stakeholder-System Divide

There is a growing demand in the business and society field for work addressing business’s role in tackling societal grand challenges (Brammer et al., 2019; Walsh, 2005; Wasieleski et al., 2020). Yet, despite Wood’s (1994) early call for a “system-level stakeholder theory” developed in parallel with the firm-centric model, the disconnection between firm- and system-level needs has become entrenched. For many reasons, not least being stakeholder theory’s long-term relevance, this disconnect may not be viable. But how could future stakeholder research connect firm-stakeholder level insights with large system change (Waddock & Kuenkel, 2019)? Reviewing recent contributions for citations likely to anchor this theme, we find few prospects. Indeed, the authors of this current article are, themselves, divided as to the potential contribution of such work. Our views fall along a spectrum whose endpoints suggest starkly different ambitions for stakeholder theory at the system level.

One direction proceeds from the view that stakeholder theory itself is not designed to offer insight into system-level outcomes. Much intellectual energy has, for almost four decades, gone into the work of identifying which actors should be considered stakeholders (Mitchell et al., 1997; Phillips, 2003a), distinguishing stakeholder theory from related constructs with more systemic intent like CSR (Dmytriyev et al., 2021; Elms et al., 2011) and contributing to the more well-defined expectations of strategic management’s narrower search for causal relationship between firm conduct and sustainable competitive advantage (Bosse et al., 2009; Harrison et al., 2010). Ultimate success, viewed from this perspective, would be to replace the traditional emphasis on shareholder wealth maximization with a new aggregate measure of social welfare (Jones et al., 2016), and the best stakeholder theory can do is to urge more responsible management practices that, in turn, enhance total stakeholder happiness (Jones & Felps, 2013a, 2013b). Future researchers might ask interesting questions along these lines: Do different forms of incommensurability across stakeholder interests affect managerial decision-making differently in its pursuit of aggregate welfare? Are some managers better at getting around incommensurability (Freeman, Phillips, & Sisodia, 2020)?

The other possibility is that stakeholder theory can begin to help account for how and why firms contribute to, impede progress on, and in some cases advance solutions to societal problems. As one prominent participant in a recent stakeholder mini conference asked: What can stakeholder theory say to help the homeless? A contribution of this sort likely requires research to proceed from a radically different basis, involving revision of dependent variables, sorting logic, causal mechanisms, and the population of firms being studied. We explore each in turn.