Few phenomena prompt the discussion of legitimacy as notably as multi-stakeholder initiatives (MSIs), that is, institutions bringing together various stakeholders from the political, non-profit, and corporate sectors to address a societal issue in a self-regulatory manner (Grimm, Ruehle, & Reinecke, Reference Grimm, Ruehle and Reinecke2024; Huber & Schormair, Reference Huber and Schormair2021; Rasche, Reference Rasche2012). These initiatives (notable examples being the UN Global Compact or the Extractive Industries Transparency Initiative) deviate from the classic notion of (hierarchical) state governance, and hence urge researchers and practitioners to discuss alternative forms of legitimacy. This urge is mirrored in a wide array of academic disciplines that contribute to the topic, including sociology (in particular, organisational sociology), management science, political philosophy, and ethics (in particular, business ethics), each of which has its own foundations and perspectives (De Bakker, Rasche, & Ponte, Reference De Bakker, Rasche and Ponte2019). Thus, academic research has produced a long list of publications that define theories, frameworks, and criteria for when MSIs are legitimate (e.g., Bäckstrand, Reference Bäckstrand2006; Fuchs, Kalfagianni, & Havinga, Reference Fuchs, Kalfagianni and Havinga2011; Mena & Palazzo, Reference Mena and Palazzo2012; Schouten & Glasbergen, Reference Schouten and Glasbergen2011).
Yet, in the midst of the debate on MSIs legitimacy, there seems to be a growing confusion regarding the term “legitimacy” itself. Most importantly, while some sources understand the legitimacy of an institution or organisation as a descriptive, relative construct based on perception (i.e., a subjective assessment by the people affected, thus a collective social evaluation), others argue that legitimacy is a normative, absolute construct based on rational justification (i.e., an objective assessment granted upon the de facto fulfilment of certain previously substantiated, well-grounded standards) (Applbaum, Reference Applbaum2019: 13–20; Beetham, Reference Beetham1991: 5–6; Hahn & Weidtmann, Reference Hahn and Weidtmann2016; Peter, Reference Peter and Zalta2017; Zürn, Reference Zürn2004). Put differently, in their two quintessential forms, legitimacy can be understood as a social construct that may vary with the people assessing it or as a stable ideal rooted in norms that people can discover but not alter. This difference, while rather simple and straightforward, is often overlooked, leading to three different problems that we identify in the literature on MSI legitimacy. The first, lack of precision, occurs when researchers writing on the topic of MSI legitimacy do not specify whether they draw on the descriptive form, the normative form, or a combination of the two. The second, blurring of boundaries, occurs when researchers combine elements of descriptive and normative legitimacy without specifying the relationship between these two forms. The third, ambiguity of assessment, occurs when researchers employ fine-grained criteria for MSI legitimacy that can be assessed from both the descriptive and the normative perspective but either neglect to specify which perspective better assesses those criteria or specify this inaccurately.
We hold that these three problems are detrimental to the construct clarity of MSI legitimacy. Construct clarity is of vital importance to research because it can “encourage researchers to generate more effective research questions, apply appropriate and epistemologically consistent methods, and identify exceptions to the categories that open opportunities for future research” (Suddaby, Reference Suddaby2010: 355). Our call for more construct clarity in legitimacy research resonates with those voices in management and organisation studies who point out the importance of construct clarity for the quality of theorising, the (interdisciplinary) exchange and accumulation of knowledge, the operationalisation of measurable components, and the falsifiability of hypotheses (Ragins, Reference Ragins2012; Suddaby, Reference Suddaby2010; von Nordenflycht, Reference von Nordenflycht2023). We believe that construct clarity is particularly relevant in an interdisciplinary field such as MSI legitimacy research, as the lack thereof can impede potential scientific advancements due to misunderstandings between researchers. Furthermore, alongside these theoretical points, limited construct clarity can also result in unintended suggestions in practice. For example, a person who adopts a descriptive form of legitimacy without using the term “descriptive” might be misread as suggesting that the legitimacy criterion of transparency could be assessed by objectively analysing the information content of an MSI’s website instead of asking relevant stakeholders about their subjective perceptions of the MSI’s transparency. A purely descriptive understanding of MSI legitimacy would only rely on the latter.
Moreover, we assert that these three problems in the research on MSI legitimacy contribute to a more general phenomenon that we label erosion of normativity and define as the increasing lack of normative grounding or (unintentional) deconstruction of the normative elements in business ethics research. Such a “perceived decline in normativity” (Smith, Donaldson, Pouryousefi, Scholz, & Spence, Reference Smith, Donaldson, Pouryousefi, Scholz and Spence2023: 785) or “loss of the proper ethical perspective” (van Liedekerke & Dubbink, Reference van Liedekerke and Dubbink2008: 273) has been acknowledged in the wider business ethics community. Lock and Seele (Reference Lock and Seele2015: 24) claim that, in business ethics, “the normative and conceptual approach … is slowly being challenged by empirical, positivist paths,” which is supported by the finding that “ethics as a philosophical discipline [is taken over] by nonphilosophers” (Seele, Reference Seele2016: 86). We contribute to this discussion by pointing out what we hold to be one of its crucial drivers—that is, the lack of construct clarity, which often disregards the normative elements of theoretical concepts (see also the point on imprecise language by Scharding and Warren, Reference Scharding and Warren2023). We argue that, if these normative elements are not acknowledged, concepts are not only oversimplified but also lose some aspects of their societal meaning and function. This is the case for legitimacy. With our argument, we do not aim to disregard descriptive legitimacy; on the contrary, the two quintessential forms of legitimacy are equally valid, justified, and important positions within academic research that we should treat as such, rather than allowing the descriptive form to be “parasitic” of the normative (Applbaum, Reference Applbaum2019: 17).
If construct clarity is not maintained and, particularly, if erosion of normativity arises, this can create problematic incentives and may give rise to attempts at deliberate deception. For example, practitioners may start caring more about measures that increase their MSI’s perceived legitimacy instead of actually mitigating wrongdoings. Alamgir and Banerjee (Reference Alamgir and Banerjee2019: 295) manifest this problem in what they call a “legitimacy paradox,” questioning whether a global retailer who is part of a specific MSI could legitimately “sell a branded T-shirt for $10 … when the women who manufactured the shirt are paid $2 a day and continue to work in unsafe and oppressive conditions.” This connects to the larger conversation of MSI greenwashing, that is, the danger of corporations feigning problem-solving efforts without any actual execution, which has been widely discussed in the literature (Berliner & Prakash, Reference Berliner and Prakash2015; Haack & Rasche, Reference Haack and Rasche2021; MSI Integrity, 2020). In this case, even a corporation that appears to be legitimate due to its participation in an MSI might nonetheless be unwilling to prevent questionable business practices. Limited construct clarity of MSI legitimacy may be one theoretical puzzle piece in explaining how greenwashing is possible despite the employment of legitimacy criteria, because it might be the case that only the perceived (descriptive) and not the de facto (normative) fulfilment of said criteria is assessed. Suchman (Reference Suchman1995: 574) also points to the complexity of legitimacy, stating that “[a]n organization may diverge dramatically from societal norms [thereby undermining what we understand as normative legitimacy] yet retain [descriptive] legitimacy because the divergence goes unnoticed.”
To address these three problems, we state this paper’s research question: How can scholars use the concept of MSI legitimacy in a manner that enhances construct clarity and avoids erosion of normativity? We offer a framework with three cornerstones that help create the foundation for solid theory building. First, we suggest that to counter lack of precision, all theories of MSI legitimacy should be positioned within a theoretical frame of reference that clarifies the implicit assumptions on which the theory draws, or, at a minimum, that theories of MSI legitimacy should be clearly defined or described (which can, at the most basic level, already be achieved by using a clarifying term such as “descriptive or “normative”). Second, we argue that legitimacy theories that involve both the descriptive and the normative perspective on MSI legitimacy should indicate the relationship between these two forms to avoid blurring of boundaries. Third, we propose that when a specific set of fine-grained criteria for the assessment of MSI legitimacy is employed, an analysis should be provided on whether assessing these criteria from a descriptive or a normative perspective provides more explanatory power for MSI legitimacy. This prevents ambiguity of assessment.
The paper contributes to the extant literature in two ways. First, our framework helps enhance construct clarity in research on MSI legitimacy by specifying the distinction between descriptive and normative legitimacy, thereby preventing fragmentation of the field. While the paper focuses on MSI legitimacy, the final section also demonstrates the implications for other fields of legitimacy research. Second, we raise our concern about an increasing erosion of normativity in scholarship on MSI legitimacy. Our paper not only demonstrates how values and hidden assumptions are prone to affecting research, as has also been discussed in the debates on both the “value-ladenness” of management and organisation studies (Hodgson, Reference Hodgson1988; Seeck, Sturdy, Boncori, & Fougère, Reference Seeck, Sturdy, Boncori and Fougère2020; Vallentin & Murillo, Reference Vallentin and Murillo2022) and the “separation thesis” in business ethics (Sandberg, Reference Sandberg2008; Singer, Reference Singer1998). It also shows how our framework helps to avoid erosion of normativity in research on MSI legitimacy and beyond.
The structure of the remainder of the paper is as follows. We first address the distinction between the descriptive and the normative forms of MSI legitimacy. Then, we highlight three problems in research on MSI legitimacy and discuss erosion of normativity. We subsequently introduce our framework and its three cornerstones for solid theory building. The paper concludes with a discussion of implications for research and practice.
Forms and problems of MSI Legitimacy
Two Quintessential Forms of MSI Legitimacy
The legitimacy of MSIs can be grounded in two quintessential forms of legitimacy that are prevalent in research: descriptive and normative legitimacy (Applbaum, Reference Applbaum2019: 13–20; Beetham, Reference Beetham1991: 5–6; Hahn & Weidtmann, Reference Hahn and Weidtmann2016; Peter, Reference Peter and Zalta2017; Zürn, Reference Zürn2004).Footnote 1 Descriptive legitimacy derives from a subjective assessment, reflecting the degree to which citizens approve of an institution or organisation (Applbaum, Reference Applbaum2019: 13–20; Beetham, Reference Beetham1991: 5–6; Peter, Reference Peter and Zalta2017). If an individual argues that an MSI has descriptive legitimacy, they mean that people perceive the MSI to meet certain standards and requirements. For example, in his widely used definition, Suchman (Reference Suchman1995: 574) characterises descriptive legitimacy as “a generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions.” Such references to social constructivism or people’s perceptions are clear indicators of descriptive legitimacy. Importantly, descriptive legitimacy depends only on what people hold to be legitimate. This means that when the people in question change, their granting of descriptive legitimacy can be assumed to change as well. In this sense, standards of legitimacy are flexible. Hence, we understand descriptive legitimacy as relative because it is the result of a subjective assessment that depends on a specific reference group.
Conversely, normative legitimacy relies on an objective assessment based on the reasons justifying the existence of an institution or organisation (Applbaum, Reference Applbaum2019: 13–20; Beetham, Reference Beetham1991: 5; Peter, Reference Peter and Zalta2017). Normative legitimacy puts the validity of an institution or organisation (often a political system) under scrutiny. If an individual argues that an MSI is legitimate in the normative understanding of the term, they claim that it meets certain previously substantiated, well-grounded standards and requirements, independent of the personal attitudes that people affected by the system may have. This means that even if people change, the status of legitimacy remains the same. In this sense, the standards of legitimacy are stable. Thus, normative legitimacy refers to an ethically grounded rational justification of, for example, power or authority (Peter, Reference Peter and Zalta2017). It is not enough for people to believe that the conditions for legitimacy are met—they must de facto be met. Hence, we understand normative legitimacy as absolute because it is the result of an objective assessment independent of a reference group.Footnote 2 Table 1 distinguishes between descriptive and normative legitimacy.
Table 1: Descriptive Legitimacy and Normative Legitimacy
These two quintessential forms of legitimacy have developed from two different research traditions. Research on normative legitimacy dates back to at least ancient Greece; one might even refer to normative legitimacy as “one of the oldest [problems] in political theory” (Bodansky, Reference Bodansky2008: 705). Descriptive legitimacy, on the other hand, is relatively young, with its origins traceable to the beginnings of sociology, in particular the works of Max Weber. Despite its relative youth, descriptive legitimacy has relegated normative legitimacy approaches to the background, especially in the context of management and organisation studies. Nonetheless, we hold both quintessential forms of legitimacy to be equally valid, justified, and important within academic research—as are theories drawing on the interdependencies between the two forms.
Three Problems in Research on MSI Legitimacy
While, on an analytical level, the distinction between descriptive and normative legitimacy will be familiar to most researchers, it is prone to cause three problems that we see as particularly relevant in research on MSI legitimacy: lack of precision, blurring of boundaries, and ambiguity of assessment. Table 2 lists the twenty most-cited journal articles discussing MSI legitimacy according to a Web of Science search.Footnote 3 The table shows that the three problems arise in several papers on the list.Footnote 4
Table 2: 20 Most-Cited Relevant Articles on MSI Legitimacy
Note. The term “not applicable” signifies that the respective problem did not arise in this paper due to only one form of legitimacy being addressed (relevant for blurring of boundaries) or due to no fine-grained criteria being employed (relevant for ambiguity of assessment).
Lack of Precision
We identify a lack of precision in works that do not specify whether their concept of MSI legitimacy draws on a descriptive form, a normative form, or any combination of the two. Lack of precision occurs when scholars omit definitions or describe their concepts vaguely, leaving readers to make their own assumptions about what the authors have in mind. This does not necessarily mean that the authors made a mistake; however, a lack of precision is prone to cause scholars to talk past each other.
Table 2 shows that ten of the twenty articles are subject to lack of precision, that is, they neglect to make visible to readers whether they understand legitimacy as descriptive, normative, or a combination of the two (see Alamgir & Banerjee, Reference Alamgir and Banerjee2019; Anner, Reference Anner2012; Balzarova & Castka, Reference Balzarova and Castka2012; Boström, Jönsson, Lockie, Mol, & Oosterveer, Reference Boström, Jönsson, Lockie, Mol and Oosterveer2015; Fuchs et al., Reference Fuchs, Kalfagianni and Havinga2011; Klooster, Reference Klooster2010; Ponte, Reference Ponte2014; Ponte & Cheyns, Reference Ponte and Cheyns2013; Raymond & DeNardis, Reference Raymond and DeNardis2015; Schouten, Leroy, & Glasbergen, Reference Schouten, Leroy and Glasbergen2012). This means that they provide neither definitions nor clear descriptions of MSI legitimacy with respect to its descriptive or normative form.
Some of these publications, while neglecting to define or describe MSI legitimacy, provide contextual indicators from which readers may assume which form of legitimacy the authors mean. For example, some authors use phrases that put legitimacy in the context of perception—for example, Anner (Reference Anner2012: 618), who writes about an MSI: “If there is too much perceived corporate influence, the program loses legitimacy.” Similarly, Alamgir and Banerjee (Reference Alamgir and Banerjee2019: 292) state, regarding the rationale of corporations participating in an MSI, that it “provides an additional sheen of legitimacy to their glossy sustainability and CSR reports.” Such contextual indicators can be read as signs that the authors have a certain form of legitimacy in mind—in the noted cases, the descriptive form. Still, in the end, they do not provide readers with certainty, leaving them to infer which form of legitimacy the authors draw on. Hence, they do not sufficiently overcome the lack of precision.
A similar case arises when authors only link their concept of MSI legitimacy to particularly established publications known for how they understand legitimacy—for example, to Suchman’s (Reference Suchman1995) work on descriptive legitimacy. On the one hand, such links are suitable for designating the form of legitimacy an author has in mind, yet, on the other hand, they raise the question of the point at which a reference publication qualifies as sufficiently known to help overcome a lack of precision. Indeed, three of the ten articles that we hold to be subject to lack of precision described MSI legitimacy mainly by referring to seminal publications (Balzarova & Castka, Reference Balzarova and Castka2012; Fuchs et al., Reference Fuchs, Kalfagianni and Havinga2011; Ponte, Reference Ponte2014). Following our analysis, these articles are equally subject to lack of precision, as deciding otherwise would increase subjectivity.
Table 2 shows all cases of lack of precision among the twenty articles. It also specifies the form(s) of legitimacy the authors use. In cases in which lack of precision exists—yet where there are contextual indicators suggesting which form the authors might mean—the assumed form is indicated alongside the term “presumably.” In only two publications (Raymond & DeNardis, Reference Raymond and DeNardis2015; Schouten et al., Reference Schouten, Leroy and Glasbergen2012) did we not find any contextual indicators.
On a positive note, many publications define MSI legitimacy clearly with respect to its descriptive or normative form by either referring to established definitions (e.g., Schouten & Glasbergen, Reference Schouten and Glasbergen2011) or using precise definitions of their own (e.g., Roloff, Reference Roloff2008). Some even discuss both forms when defining MSI legitimacy (e.g., Fransen, Reference Fransen2012).
Blurring of Boundaries
In addition to lack of precision, we also found research on MSI legitimacy to be affected by blurring of boundaries, which occurs when researchers combine elements of descriptive and normative legitimacy without specifying their relationship. Thus, while lack of precision occurs when authors do not clearly indicate which concept of legitimacy they use, blurring of boundaries occurs when they use both forms without explaining the link between them. So, while a lack of precision may arise in any publication on legitimacy, blurring of boundaries may only arise when the authors intentionally or unintentionally combine both forms.
Including elements of both descriptive and normative legitimacy in one theory does not necessarily lead to blurring of boundaries; on the contrary, it can help researchers craft comprehensive theories of legitimacy. While the two quintessential forms of legitimacy are conceptually distinct (meaning that they result from either a purely subjective or a purely objective perspective), many theories of legitimacy are indeed situated between those two opposites, drawing on the interdependencies between them (Fransen, Reference Fransen2012; Peter, Reference Peter and Zalta2017; Scharpf, Reference Scharpf1997: 13–15). Descriptive legitimacy can be influenced by normative legitimacy because, for example, the public support for such an institution as an MSI (a criterion whose result is relative to the reference group) usually hinges on whether it can be justified based on previously substantiated, well-grounded standards. In turn, normative MSI legitimacy can incorporate descriptive MSI legitimacy when the objective standards demand that the MSI be supported by those affected by it. However, as fruitful as the blend of descriptive and normative legitimacy can be, it requires theories to define and describe precisely the relationship between them. Neglecting to do so blends the two forms of legitimacy in a way that glosses over the important differences discussed above. Of the analysed twenty publications, four involve blurring of boundaries.
Mena and Palazzo define MSI legitimacy as descriptive (referring to it as a “socially shared belief,” Reference Mena and Palazzo2012: 528) and draw on Fritz Scharpf’s distinction between input and output legitimacy (Scharpf, Reference Scharpf1972, Reference Scharpf1997, Reference Scharpf1999), according to which research on the input side observes a political system’s opportunities for citizen participation (Scharpf Reference Scharpf1997: 153–155, Reference Scharpf1999: 7–10), while research on the output side concentrates on result quality as well as the avoidance of nuisances (Scharpf Reference Scharpf1997: 153–155, Reference Scharpf1999: 10–13). This distinction commonly appears in descriptive (Scharpf, Reference Scharpf1997, Reference Scharpf2003, Reference Scharpf2009) or normative (Scharpf, Reference Scharpf1972, Reference Scharpf1999) research. However, Mena and Palazzo seem to understand input legitimacy as descriptive (“to what extent regulations are perceived as justified or credible,” Reference Mena and Palazzo2012: 528, emphasis added), while their definition of output legitimacy sounds rather normative (“to what extent they effectively solve the issues that they target,” Reference Mena and Palazzo2012: 528, emphasis added). They do not explain this blend further, thus creating blurred boundaries.
Schouten and Glasbergen (Reference Schouten and Glasbergen2011) also draw on a descriptive understanding of MSI legitimacy, which they then subdivide into three different aspects. The first, “legality,” is understood descriptively (“What is perceived as legitimate differs across time, place, and organizational context,” Reference Schouten and Glasbergen2011: 1892). The second, “moral justifications,” focuses on “normative principles for rights to govern” (Reference Schouten and Glasbergen2011: 1892), thus using normative legitimacy as a basis. The third, “consent/acceptance,” is also understood as a descriptive form of legitimacy (“it is widely believed that it [an institution] has the right to rule,” Reference Schouten and Glasbergen2011: 1893). While the authors then examine “each of these three components and their interactions” (Reference Schouten and Glasbergen2011: 1893), they leave open how these three different forms of legitimacy come together in their concept of descriptive legitimacy. Accordingly, blurring of boundaries arises.
Bäckstrand (Reference Bäckstrand2006: 473) conceives legitimacy as entailing descriptive and normative elements, defining it as a “belief held by actors,” but explaining that “[n]ormatively, global governance is the process of creating a legitimate political order.” The paper interweaves and discusses both the descriptive and normative forms of legitimacy. For example, the author asks whether “participation … without real decisional power, is … meaningless,” reflecting on the discrepancy between the perceived meaningfulness of a dialogue without decision-making power and the power that ideally should follow participation (Bäckstrand, Reference Bäckstrand2006: 484). Such allusions to interdependencies without further explanation of how the descriptive and normative elements come together in one concept of legitimacy create blurring of boundaries.
Finally, Voegtlin and Pless (Reference Voegtlin and Pless2014: 187) identify three theoretical perspectives, of which two—the “socio-historical” and the “normative” perspectives—are relevant for legitimacy. While the “socio-historical” perspective focuses on “perceived legitimacy,” the normative perspective on legitimacy aims to “determine and justify the specific values and norms of socially responsible business by drawing on ethical theories” (Voegtlin & Pless, Reference Voegtlin and Pless2014: 188). However, the paper separates the two forms of legitimacy without further explaining their relationship.
On a positive note, both the articles by Fransen (Reference Fransen2012) as well as by Schouten and Bitzer (Reference Schouten and Bitzer2015) mention both forms of legitimacy and adequately explain the relationship between them. No blurring of boundaries occurs in these publications.
Table 2 shows all cases of blurring of boundaries as well as the cases in which the problem was overcome. Additionally, it marks all remaining cases as “not applicable”—those in which only one form of legitimacy is used as well as those in which lack of precision existed and the form of legitimacy can be presumed.Footnote 5
Ambiguity of Assessment
Finally, we argue that a third problem in research on MSI legitimacy arises from the ambiguity of assessment when legitimacy is assessed with the aid of fine-grained criteria. These are sets of items into which the broad concept of legitimacy is divided—for example, “transparency,” “consensual orientation,” “inclusion,” “efficacy,” and “accountability” (Bäckstrand, Reference Bäckstrand2006; Mena & Palazzo, Reference Mena and Palazzo2012; Schouten & Glasbergen, Reference Schouten and Glasbergen2011). Throughout this paper, we use the term “lenses” to explain that most criteria can be assessed from both a descriptive and a normative perspective on legitimacy. Importantly, although both lenses can be applied to a criterion, we argue below that one of the two lenses often provides a more relevant answer than the other. When we speak of the “descriptive lens,” we refer to the subjective assessment of a criterion through the perceptions of a specific reference group, making the result relative. Analogously, when we speak of the “normative lens,” we refer to the objective assessment of a criterion according to previously substantiated, well-grounded standards and independent of any perceptions, making the result absolute. For instance, for the criterion of “transparency,” we can ask people whether they think an institution is transparent (which is then relative to their beliefs), and we can objectively analyse the information that institution has shared with the public. The lens through which we assess the criterion of transparency may lead to different results.
Hence, we consider it important to inquire how to adequately assess legitimacy criteria to obtain a relevant answer on legitimacy—that is, whether the descriptive or the normative lens on these criteria has more explanatory power for legitimacy. Thus, ambiguity of assessment arises when a theory suggests fine-grained criteria of MSI legitimacy that can be assessed through both the descriptive and the normative lens but neglects to specify whether the former or the latter better assesses those criteria. Ambiguity of assessment may also arise when this specification exists but does not align with the authors’ definition of legitimacy or frame of reference.
Ambiguity of assessment differs from blurring of boundaries in that blurring of boundaries refers to a blend of descriptive and normative elements of MSI legitimacy on a higher level—that is, for MSI legitimacy in general—while ambiguity of assessment refers to the level of criteria for MSI legitimacy, thus, a more fine-grained level of analysis. Still, these two problems can affect each other. If the relationship between descriptive and normative legitimacy is unclear (blurring of boundaries), the assessment of fine-grained criteria is even more difficult to determine, making ambiguity of assessment more likely.
Scholars have suggested many sets of criteria for assessing MSI legitimacy (e.g., Bäckstrand, Reference Bäckstrand2006; Fuchs et al., Reference Fuchs, Kalfagianni and Havinga2011; Mena & Palazzo, Reference Mena and Palazzo2012; Schouten & Glasbergen, Reference Schouten and Glasbergen2011). However, during this process, criteria may be employed without any specification or reflection on whether they are better assessed through a descriptive or a normative lens. In Table 2, four of the twenty articles are subject to ambiguity of assessment.
Mena and Palazzo (Reference Mena and Palazzo2012), drawing on Habermasian deliberative democracy, provide a set of seven criteria: “inclusion,” “procedural fairness,” “consensual orientation,” “transparency,” “rule coverage,” “efficacy,” and “enforcement.” As noted above, they define their concept of MSI legitimacy as descriptive. However, despite their reflected use of the concept of MSI legitimacy, we hold that some of their criteria are better assessed through a normative lens, which challenges their understanding of descriptive legitimacy. This will be illustrated later.
The paper by Bäckstrand (Reference Bäckstrand2006) provides three criteria: “representation,” “accountability,” and “effectiveness.” While the author defines legitimacy as comprising both descriptive and normative elements, she provides no further discussion on whether her criteria should be rather assessed from a descriptive or a normative perspective. This creates ambiguity of assessment.
The same holds true for Schouten and Glasbergen (Reference Schouten and Glasbergen2011), who draw on a descriptive form of legitimacy and offer a set of nine criteria but remain silent regarding the lenses through which to assess those criteria.
Lastly, Fuchs et al. (Reference Fuchs, Kalfagianni and Havinga2011) appear to employ a normative form of legitimacy, offering a set of three criteria (“participation,” “transparency,” and “accountability”), which, we assume, are thus to be assessed through the normative lens. We hold that, despite this view through the normative lens, the criterion of “participation” should ideally be assessed through both lenses. Neglecting this creates ambiguity of assessment (see our discussion of Mena and Palazzo’s criterion of “inclusion” in this paper).
Table 2 shows all cases of ambiguity of assessment. Additionally, it marks all remaining cases, in which no criteria are employed and in which hence ambiguity of assessment was not possible, as “not applicable.”Footnote 6
Erosion of Normativity
We assert that all three problems described above, due to their capacity to harm construct clarity, also contribute to a more general phenomenon that we call erosion of normativity, by which we mean the increasing lack of normative grounding or (unintentional) deconstruction of the normative elements of legitimacy.
As we have pointed out above, the two quintessential forms of legitimacy are equally valid, justified, and important positions within academic research. Hence, when we write critically on erosion of normativity, our argument is not that we hold one quintessential form of legitimacy to be in any way superior to the other. Instead, our argument is that the three problems lead to an erosion of normativity due to researchers systematically neglecting the normative form of legitimacy in their writing. As Barlow (Reference Barlow2022: 546) puts it: “The risk … is to conflate the appearance of normative justification or the conviction among members of the population that a governance arrangement is justified with its actual justification.”
Erosion of normativity particularly occurs due to the three problems we identified in the literature. First, lack of precision contributes to erosion of normativity because the confusion between the two forms usually arises at the expense of the normative form. When researchers do not clearly define or describe their concept of MSI legitimacy yet presumably have a descriptive form in mind, they suggest, whether intentionally or not, that descriptive legitimacy is the default form of legitimacy, thereby contributing to the hollowing-out of the normative form. For example, Anner (Reference Anner2012: 618) does not define legitimacy in his paper, yet claims that “[i]f there is too much perceived corporate influence, the program loses legitimacy.” This makes it seem as if the de facto corporate influence holds no significance for legitimacy.
Second, blurring of boundaries contributes to erosion of normativity because it is mostly the normative perspective on MSI legitimacy that is incorporated into the descriptive form without further explanation (rather than vice versa). Consequently, researchers, whether intentionally or not, suggest that normative legitimacy becomes a subcategory of descriptive legitimacy, as if it were on a lower conceptual level. For example, when Mena and Palazzo (Reference Mena and Palazzo2012: 528) combine their descriptive concept of input legitimacy with their normative understanding of output legitimacy in an overall descriptive concept of legitimacy, this creates the impression that normative legitimacy is solely a subcategory of descriptive legitimacy.
Finally, ambiguity of assessment contributes to erosion of normativity because it is often papers with a descriptive understanding of legitimacy that neglect to reflect through which lens fine-grained criteria of MSI legitimacy should be assessed, or that propose a certain lens while overlooking the criterion’s normative foundations. In these cases, the authors, whether intentionally or not, suggest that the normative elements of MSI legitimacy can be reduced to mere attitudes and sentiments that could be summed up along with descriptive elements into one measure of descriptive MSI legitimacy. For example, Schouten and Glasbergen (Reference Schouten and Glasbergen2011: 1897) define their criterion of “broader societal acceptance” as descriptive and point out the negative sentiments some non-governmental organisations (NGOs) as well as the public hold towards the MSI. Yet, they do not link these negative sentiments to the reasons previously discussed by the authors, such as the fact that membership “does not guarantee compliance with the RSPO principles and criteria” (Schouten & Glasbergen, Reference Schouten and Glasbergen2011: 1897, emphasis added). This reason, however, demonstrates that the criterion of “broader societal acceptance” actually hinges on the de facto criteria fulfilment of compliance—that is, the normative lens, and not solely the perception thereof (descriptive lens). Yet, this normative element of MSI legitimacy is reduced to a mere attitude or sentiment and then summed up along with descriptive elements into one criterion of descriptive MSI legitimacy.
It is important to note that on a theoretical level, the three problems could also work in the opposite direction. This means they could equally occur in publications on normative MSI legitimacy that fail to duly acknowledge the descriptive form. However, we see far fewer examples of this, as our list of papers on MSI legitimacy (Table 2) demonstrates. Of the fourteen papers in which at least one of the three problems occurs, only one features a (presumably) normative concept of MSI legitimacy (i.e., Fuchs et al., Reference Fuchs, Kalfagianni and Havinga2011) and two others consciously blend descriptive and normative legitimacy (i.e., Bäckstrand, Reference Bäckstrand2006; Voegtlin & Pless, Reference Voegtlin and Pless2014). Of the remaining eleven papers, two are (and seven presumably are) descriptive concepts of MSI legitimacy and two provide no indication of which form of legitimacy is employed.
These signs of the normative perspective’s loss of significance are noted repeatedly in various areas of research (Smith et al., Reference Smith, Donaldson, Pouryousefi, Scholz and Spence2023). For example, Habermas (Reference Habermas1996: 333) points out that some approaches in political and economic theory “have pushed the normative weight reduction too far.” Applbaum (Reference Applbaum2019: 17), writing on descriptive and normative legitimacy, argues even more strongly: “[I]t is a conceptual confusion to hold that ‘legitimate’ simply means ‘believed to be legitimate,’ for descriptive legitimacy is parasitic on the conceptually prior idea of normative legitimacy.” Moreover, the signs of this loss of significance can also be connected to the revived debate on value-ladenness in research, which questions whether “there can be such a thing as value-free theorizing within the realm of the social sciences” (Vallentin & Murillo, Reference Vallentin and Murillo2022: 638). In sum, we hold that erosion of normativity is particularly relevant in research on MSI legitimacy, given the interdisciplinary nature of the topic and the frequent occurrence of the three problems in this context.
A FRAMEWORK TO BETTER SITUATE MSI LEGITIMACY
Our aim in this paper is to present a framework that helps to prevent lack of precision, blurring of boundaries, and ambiguity of assessment, thereby enhancing construct clarity and avoiding erosion of normativity. The framework, illustrated in Figure 1, shows how the three problems are detrimental to construct clarity and, as a consequence of this, create erosion of normativity. The figure also shows how the cornerstones for solid theory building in MSI legitimacy research (on the right) enhance construct clarity, thereby preventing erosion of normativity. The framework will be explained with the help of a concrete illustrative example throughout the paper.
Figure 1: Three Problems, Construct Clarity, Erosion of Normativity, and the Cornerstones for Solid Theory Building
First Cornerstone: Selection of a Theoretical Frame of Reference
On the basis of the above analysis, we argue that using the term “legitimacy” alone does not sufficiently specify what form of legitimacy is meant. Rather, additional information is needed to prevent lack of precision.
We believe that the most precise way to ground a theory of legitimacy is to select a theoretical frame of reference. This compels authors to not only make use of certain terms but also to engage with a theoretical concept more deeply and make themselves aware of its implicit assumptions. Such assumptions are prevalent in theories of legitimacy and often depend on the author’s chosen philosophical or political convictions. For example, someone who uses a descriptive form of legitimacy (thus, a relative perspective), may implicitly assume a position such as moral relativism, which holds all moral norms as necessarily relative because they depend on cultural, societal, or individual perspectives rather than objectivity and universality (Gowans, Reference Gowans2021). This would mean that one assumes legitimacy to exist solely in its descriptive form, with no conceivable objective normative position. However, many researchers working on descriptive legitimacy might not hold such a strong position. Needless to say, a descriptive form of legitimacy can also be used with less demanding assumptions. Yet, without making this transparent, the conversation on MSI legitimacy might draw on different understandings.
We hold the selection of a theoretical frame of reference to be the most viable means of overcoming lack of precision because it enables grounding one’s work within a comprehensive and established theory—for instance, the seminal works by Rawls (Mäkinen & Kourula, Reference Mäkinen and Kourula2012) or Habermas (Gilbert & Rasche, Reference Gilbert and Rasche2007).
We also are aware that in some contexts and debates, it may not be appropriate to devote excessive space to assumptions and backgrounds, but rather to proceed directly to the content. This is true, for example, when one’s focus lies on the presentation of empirical results without overly emphasising theoretical embedding. In such cases, minimal clarity might suffice, so the use of a clear definition of legitimacy, a description, or even just a clarifying term like “descriptive” or “normative” (which could be viewed as a minimal definition), is enough. This may not be the most precise way to overcome lack of precision, but it can at least provide sufficient guidance.
Hence, in order to avoid lack of precision, scholars should disclose their chosen theoretical frame of reference or provide unequivocal definitions for the concepts they use.
In line with many experts in the field, we chose Habermasian discourse ethics as a theoretical frame of reference for the illustrative example in this paper (Habermas, Reference Habermas1991). Habermas’s theory of discourse ethics provides a solid foundation for how MSIs can strive towards these goals in a way that meets democratic standards (Gilbert & Rasche, Reference Gilbert and Rasche2007; Rasche & Esser, Reference Rasche and Esser2006). While most theories lean strongly towards either descriptive or normative legitimacy, Habermas’s theory seeks to bridge the trenches between sociology and philosophy, making it well-suited to demonstrating how both perspectives can be integrated into a single theory. Researchers working with Habermas’s discourse ethics would therefore find themselves in the middle of the descriptive-to-normative continuum. According to Habermas, valid moral norms and rules are the product of an all-embracing discourse between equals that excludes no stakeholder and in which participants only make statements that they themselves believe to be true (Habermas, Reference Habermas2005, Reference Habermas1991: 87–88). Participants in such discourse genuinely seek a morally sound solution or answer to the problem at hand. Thereby, they engage in communicative action (reasoning and sharing of arguments) instead of strategic action (bargaining and pursuing one’s own interests) (Habermas, Reference Habermas1984: 285–286). While a discourse must objectively follow certain procedures (incorporating elements of normative legitimacy), its goal is to bring together the subjective perspectives of participating individuals (incorporating elements of descriptive legitimacy).
Second Cornerstone: Consideration of Interdependencies
We argue that when the descriptive and normative forms of legitimacy are combined into one concept, the relationship between the two forms should be explained to prevent blurring of boundaries.
After the selection of a theoretical frame of reference, it is important to analyse whether the chosen concept of legitimacy also entails elements of the other form. In some cases, the frame of reference might already imply certain presuppositions. Even if not, it is conducive for solid theory building to disclose how the two forms of legitimacy affect each other. This means that if researchers predominantly use a descriptive form of MSI legitimacy but also mention normative elements, they should demonstrate how the descriptive form incorporates the normative elements within the scope of their argument. For example, when they use the descriptive form—discussing the perceived legitimacy of an MSI among those whom it affects—they may conceive this perceived legitimacy to comprise what those affected deem objectively legitimate. Researchers should then show how people’s objective (i.e., normative) assessments feed into the descriptive form. If they do not explain this link, readers might assume that the authors consider normative legitimacy a subcategory of descriptive legitimacy.
In turn, if researchers use a normative form of MSI legitimacy and then also mention descriptive elements, they should show how the normative form incorporates these descriptive elements within the scope of their argument. For example, many researchers who discuss the normative legitimacy of an MSI nevertheless build on the consent or acceptance of those affected by it. Consent usually relies on people’s perceptions, an inherently subjective factor, creating a need for clarification. Researchers should demonstrate how this descriptive assessment of those affected influences the normative form. If they do not explain this link, readers might assume that the authors consider descriptive legitimacy a subcategory of normative legitimacy.
Hence, to avoid blurring of boundaries, we recommend that scholars elucidate how the descriptive and normative elements of their concepts are related.
In our illustrative example, which follows Habermasian discourse ethics, the interdependencies of descriptive and normative legitimacy are particularly distinct. Like many other approaches in the field of deliberative democracy, Habermasian discourse ethics bases its normative propositions on a descriptive element, namely, what those affected by an issue can agree upon in an ideal deliberative discourse (Bächtiger, Dryzek, Mansbridge, & Warren, Reference Bächtiger, Dryzek, Mansbridge, Warren, Bächtiger, Dryzek, Mansbridge and Warren2018; Bohman & Rehg, Reference Bohman, Rehg and Zalta2017). Thus, these normative propositions derive from the reasonable opinions, values, and judgements of discourse participants. In the context of MSI legitimacy, this means that if, after a rational exchange of arguments, a proposition is perceived appropriate by all of those affected (giving it descriptive legitimacy), then this proposition also has normative legitimacy, as it was the result of an ideal discourse (Habermas, Reference Habermas2005, Reference Habermas1991: 66, 87–88). Habermas emphasises this strong link between descriptive and normative legitimacy throughout his work.
Third Cornerstone: Choice of Lens to Assess Legitimacy through Fine-Grained Criteria
MSI legitimacy theories often substantiate legitimacy through sets of fine-grained criteria to assess the legitimacy of MSIs. First, it is important for solid theory building to select criteria that are aligned to the theoretical perspective that has been chosen before, that is, the respective theoretical frame of reference. When adopting Habermasian discourse ethics, it would be apt to choose a criterion such as “balance of power structures,” that is, the degree to which power is equally distributed, in contrast to the criterion of “legacy,” that is, historical claims of power, as the latter criterion should ideally play no role in Habermas’s discourse ethics. Second, we argue that there is a need to analyse whether the descriptive or the normative lens on these criteria provides more explanatory power for MSI legitimacy and explain how the chosen lens fits the theoretical frame of reference. Otherwise, ambiguity of assessment might occur.
When we speak of explanatory power, we refer to the fact that in some cases, it is the perceived fulfilment of a criterion (assessed by the descriptive lens) that is key to achieving legitimacy, while in other cases, it is the de facto fulfilment of a criterion (assessed by the normative lens). Confusion regarding how to assess legitimacy may lead to inconsistencies and contradictions, limiting the validity of one’s arguments.
Thus, to avoid ambiguity of assessment, we propose three guiding principles—measurability, consistency, and meaningfulness—that can help to determine whether a criterion for legitimacy should be preferably assessed through the descriptive or normative lens.
-
• Measurability: In some cases, there are epistemic limitations on what a lens can reasonably measure. For example, objectively assessing the true intentions of stakeholders through a normative lens may be difficult because it relies on observed behaviour or testimony. Thus, assessing people’s perceptions (the descriptive lens) might be more feasible.
-
• Consistency: Sometimes, a lens is more or less consistent with the assumptions of a theoretical frame of reference. Political and philosophical theories that engage with legitimacy often come with certain (hidden) assumptions. Thus, the assessment of a criterion should be consistent within the system and rationales of the chosen frame of reference. For example, the absolute fulfilment of inclusion, objectively assessed through the normative lens, is a precondition for some theories of discourse or deliberation.
-
• Meaningfulness: In some cases, discrepancies arise between the descriptive assessment of people’s perceptions of a criterion’s fulfilment and the normative assessment of objective standards for such fulfilment. Deciding which of the two lenses has more explanatory power requires asking whether it would be worse if a criterion were perceived to be satisfied but de facto not satisfied, or de facto satisfied but perceived as not satisfied. For example, perceived transparency might be less meaningful than actual transparency, and having the former without the latter might produce a superficial argument.
Consequently, we must inquire into each criterion and observe which lens provides better answers or has more explanatory power regarding the legitimacy of an MSI. When we ask which lens is better suited, we do not aim to deprecate the other lens; rather, we argue that sometimes one lens provides more explanatory power for legitimacy than the other. Our goal is to illustrate how the third cornerstone of our framework, like the other two, can contribute to situating a model of legitimacy more precisely between descriptive and normative legitimacy.
The following features an application of this third cornerstone. To illustrate, we discuss three criteria: transparency, consensual orientation, and inclusion. These three criteria are not only part of the set of criteria that Mena and Palazzo (Reference Mena and Palazzo2012) offered in arguably one of the most influential publications on MSI legitimacy. They also reappear in many other publications on MSI legitimacy. An overview of all seven of Mena and Palazzo’s criteria can be found in Table 3, which provides MSI examples as well as further illustration of how to make use of both the descriptive and the normative lens on legitimacy.
Table 3: Mena and Palazzo’s Seven Criteria and the Suitable Lens from the Perspective of Habermasian Discourse Ethics
We embed these three criteria into the illustrative example of Habermasian discourse ethics. Thereby, we aim to explain when and why one ought to prioritise one lens over the other to prevent ambiguity of assessment. To do so, we first discuss the three guiding principles and then add concrete examples of MSIs for each criterion.
Transparency
We understand transparency as an adequate level of information-sharing that demonstrates openness and allows for accountability (Hale, Reference Hale2008). Many scholars, such as Mena and Palazzo (Reference Mena and Palazzo2012), argue that MSIs should provide the public with sufficient information about the MSI’s internal and external processes, decisions, and performance. Utilising our three guiding principles, we can determine whether a descriptive or a normative lens is more suitable for assessing an MSI’s transparency.
First and foremost, we should ask whether any limitations with respect to measurability affect either the descriptive or the normative lens of assessment. In the case of transparency, we believe that both lenses can reasonably measure the criterion. Researchers can ask people about their perceptions of an MSI’s transparency, or they can draw on objective indicators for transparency, such as the quality and quantity of information on the MSI’s website and in its annual reports (Mena & Palazzo, Reference Mena and Palazzo2012). Thus, the first guiding principle does not provide clarification on ambiguity of assessment.
The second guiding principle asks for consistency with the theoretical frame of reference, which, in the case of our example, is Habermasian discourse ethics. Habermas demands that participants not deceive each other within a deliberative discourse, as this would be a form of strategic action (Habermas, Reference Habermas1984: 285–286). Conversely, deliberation only succeeds when stakeholders in the process of finding joint solutions are willing to let the best argument win (communicative action) (Habermas, Reference Habermas2005, Reference Habermas2006). Following our interpretation of Habermas’s discourse ethics, de facto transparency (in the form of valid information about decision-making processes or performance indicators, honest sharing of arguments, and reduced information asymmetries) is essential for deliberation (Habermas, Reference Habermas1991, Reference Habermas2006). In this case, legitimacy hinges not on subjective perception, which depends on the respondents’ view, but rather on the actual degree of transparency within an MSI, assessed objectively. Thus, the guiding principle of consistency suggests that the normative lens is more suitable for assessing the transparency of an MSI within the Habermasian frame of reference.
Third, for an additional indicator, we can assess the meaningfulness of the outcome of each lens. Two extremes are possible: an MSI that is perceived as transparent but is in fact not transparent, and an MSI that is not perceived as transparent but is actually transparent, according to previously substantiated, well-grounded standards. We argue that there is good reason to prefer the latter to the former: A de facto non-transparent MSI is detrimental to democratic procedures if it is perceived as transparent. As some MSIs aim to occupy a role akin to that of state legislation, one can argue that they also must answer to the people as quasi-democratic institutions. However, the public often cannot judge whether information is being kept from them, which obstructs their subjective assessment. Therefore, having objective transparency indicators is more meaningful than asking for the public’s subjective opinions.
Consequently, we argue that transparency should be assessed through a normative rather than a descriptive lens when following Habermasian discourse ethics.
Consensual Orientation
Mena and Palazzo (Reference Mena and Palazzo2012: 537) define consensual orientation as a “[c]ulture of cooperation and reasonable disagreement” among participants. Although in many MSIs, dissensus is more likely than consensus, the emphasis is on “orientation.” Thus, in contrast to “transparency,” “consensual orientation” does not refer to the properties of the MSI but rather to the mindset of the various stakeholders. It asks the extent to which the participants within an MSI want to achieve consent rather than seeking quarrels by tenaciously sticking to their positions.
Following our guiding principles, we first ask whether consensual orientation has limits in terms of measurability. Mena and Palazzo (Reference Mena and Palazzo2012: 547) explain their understanding of consensual orientation by proposing two ways of evaluating this criterion through what appears more like a normative lens (despite their descriptive concept of legitimacy): the frequency with which stakeholders consider opposing arguments and the observation of other “signs of cooperation.” Although it may at first appear that a normative lens can help objectively assess consensual orientation, we argue that neither the proposed assessment nor any other objective assessment could reveal the discussants’ true intentions. Our argument relies on the conceptualisation of consensual orientation as the stakeholders’ mindset, that is, the intention to listen to each other’s arguments and agree upon a reasonable solution (i.e., Habermasian “communicative action”), in contrast to self-interested bargaining (i.e., Habermasian “strategic action”) (Habermas, Reference Habermas1984: 285–286). Showing “signs of cooperation” can also be a tactical move to appear consensus-oriented while holding on to one’s goals and improving one’s bargaining position. In fact, strategic and communicative action “often go together in actual political discussion” (Thompson, Reference Thompson2008: 504). Hence, we argue, objectively measuring the sincerity of consensual orientation is challenging, if not impossible.
While the second principle, consistency, cannot assist us any further due to the epistemic limits of knowing whether this ideal has de facto been fulfilled, the third guiding principle might provide some additional support. We argue that it is also more meaningful to inquire about the perception of consensus orientation (through the descriptive lens) within the MSI. This does not necessarily reveal true stakeholder intentions, but it can at least provide insights into whether discourse participants believe that their fellow discussants are bargaining or deliberating. Such a belief is relevant because when stakeholders perceive others as interested in finding a joint solution (even if they are not measured by a normative lens), it may nevertheless change their own mindset for the better in the long run, thereby increasing the likelihood of communicative action. However, if stakeholders perceive their fellows as acting selfishly (even though they are not), then the atmosphere of the discourse likely suffers, and a reasonable and trustful exchange of arguments is not set in motion. Thus, we argue that beliefs—and thereby the descriptive lens on legitimacy—matter more for deliberation than the normative lens because assessing stakeholders’ perceptions can provide more meaningful insights into the deliberative qualities, and therefore also the legitimacy, of an MSI.
Consequently, in the context of Habermasian discourse ethics, we deem a descriptive lens more suitable than a normative lens for assessing “consensual orientation.”
Inclusion
Inclusion can be defined as the “[i]nvolvement of stakeholders affected by the issue in the structures and processes of the MSI” (Mena & Palazzo, Reference Mena and Palazzo2012: 537). It can take different forms, and we see researchers using different variants of this criterion—for example, demanding participation (an active form of inclusion) or representation (inclusion by proxy). While most criteria are better assessed through either the descriptive or normative lens when adopting Habermasian discourse ethics, the criterion of “inclusion” is more complex.
Following the principle of measurability, we see that inclusion can be assessed through both the descriptive and normative lenses, depending on how we define “inclusion.” On the one hand, inclusion involves the properties of the MSI itself, as in the case of “transparency.” The Habermasian ideal demands that all relevant stakeholders take part in a discourse (Habermas, Reference Habermas2005), which is a requirement that could be assessed objectively, that is, without querying the participants’ perceptions. In that case, we would need to ask whether all individuals are included de facto (normative lens). On the other hand, inclusion also parallels the criterion of “consensual orientation,” asking about stakeholders’ inner state, that is, the feeling of being part of the process. This would be assessed through a conversation with relevant stakeholders (descriptive lens). Thus, neither lens is limited regarding measurability.
From the perspective of consistency, de facto inclusion of all affected parties would fulfil the requirement of Habermasian discourse ethics. Thus, the normative lens would usually suffice to provide an answer. However, there is a feasibility problem. Inclusion in MSIs is realistically only possible via representation (e.g., through NGOs), but such organisations might fail in their task of being a suitable proxy for the groups they aim to represent (Reinecke & Donaghey, Reference Reinecke and Donaghey2022). Consequently, we argue that the normative lens provides relevant but limited information.
From the perspective of meaningfulness, we argue that the ability to actively participate in a discourse makes participants feel included. Feeling part of a discussion concerning one’s own interests is an important aspect of how one engages in the discourse and whether one might be willing to accept the negotiated solutions. It adds something significant: a shared understanding. In this sense, it might, under certain conditions, be even more important for legitimacy to feel included (descriptive lens) than to be de facto included (normative lens).
Consequently, we claim that for the criterion of inclusion, both the normative lens (whether stakeholders are actually included) and the descriptive lens (if stakeholders feel included) provide important information about the legitimacy of an MSI.
Discussion
This paper provides insights into the muddy waters of MSI legitimacy research and offers guidance to researchers and practitioners by asking how the concept of MSI legitimacy can be used in a manner that enhances construct clarity and avoids erosion of normativity.
Enhancing Construct Clarity in MSI Legitimacy Research and Beyond
Our argument aims to contribute to the field of MSI legitimacy by not only raising awareness of the importance of construct clarity but also offering an approach to solid theory building. An interdisciplinary field like MSI legitimacy, in which scholarly traditions and concepts blend (if not clash), requires us all to specify the constructs with which we work in order to maintain the quality of theorising, the (interdisciplinary) exchange and accumulation of knowledge, the operationalisation of measurable components, and the falsifiability of hypotheses (Ragins, Reference Ragins2012; Suddaby, Reference Suddaby2010; von Nordenflycht, Reference von Nordenflycht2023). This holds true particularly in light of more recent articles that embed MSI legitimacy in other theoretical frames of reference, such as the theories of agonistic pluralism (Dawkins, Reference Dawkins2015) or pragmatism (Barlow, Reference Barlow2022). Future research could compare different legitimacy theories and point out how assuming a certain theoretical frame of reference affects the expectations, purpose, and success factors of MSIs.
Moreover, we hold that the three problems, while particularly relevant in the context of MSI legitimacy, also occur in other contexts of legitimacy research. Our analysis and framework may thus contribute more widely to construct clarity.
Lack of precision appears in various works discussing the legitimacy of other institutions or organisations. One typical example is the reception of the seminal works by Fritz Scharpf and his distinction between input legitimacy and output legitimacy. As explained above, Scharpf used this distinction in the context of both descriptive and normative theory. While in his earliest rendition (Scharpf, Reference Scharpf1972) and in other major works (e.g., Scharpf, Reference Scharpf1999), Scharpf clearly attributed this distinction to normative theorising, his later works featured legitimacy as a descriptive concept (Scharpf, Reference Scharpf1997, Reference Scharpf2003, Reference Scharpf2009), describing how “socially shared legitimacy beliefs” (Scharpf, Reference Scharpf2009: 173) can be subdivided into an input and an output perspective. Yet, even though Scharpf took great care to analytically separate the two perspectives—for example, reserving the term “legitimacy” for descriptive legitimacy (Scharpf, Reference Scharpf1997: 153, 2003: 1) while referring to normative perspectives as “legitimation” (Scharpf, Reference Scharpf1998: 1) or “legitimization” (Scharpf, Reference Scharpf1999: 7, 10)—these clues were rather subtle. This led some scholars to interpret both input and output legitimacy as exclusively normative categories (Fuchs et al., Reference Fuchs, Kalfagianni and Havinga2011; Schleifer, Reference Schleifer2019; Schneiker & Joachim, Reference Schneiker and Joachim2018), while others deemed them exclusively descriptive (Krasner & Risse, Reference Krasner and Risse2014; Kruuse, Reming Tangbæk, Jespersen, & Gallemore, Reference Kruuse, Reming Tangbæk, Jespersen and Gallemore2019; Mena & Palazzo, Reference Mena and Palazzo2012; Moratis, Reference Moratis, Aluchna and Idowu2017). This lack of precision regarding Scharpf’s work occurs not only within research on MSI legitimacy but also in other areas of legitimacy research. Thus, following the first cornerstone of our framework can similarly increase terminological precision in these fields. We see a need for future research to examine and overcome lack of precision in similarly established concepts, such as through literature reviews.
Blurring of boundaries also appears in various works that discuss the legitimacy of other institutions or organisations. Our framework contributes to the broader legitimacy literature by motivating scholars using both forms of legitimacy to explain how exactly the two forms are related. One typical example of how our framework could contribute is the reception of Mark Suchman’s influential publication on organisational legitimacy (1995), in which he distinguishes between three types of descriptive legitimacy: “pragmatic legitimacy,” the extent to which stakeholders perceive an organisation to be useful from a self-interested point of view; “cognitive legitimacy,” the extent to which the organisation is passively accepted as taken-for-granted; and “moral legitimacy,” the extent to which they deem the organisation’s moral activities to be valuable. Thus, while we can intuitively qualify pragmatic and cognitive legitimacy as descriptive legitimacy, moral legitimacy is “based on normative approval” (571), which may create a semblance of normativity. Yet, Suchman was clear that moral legitimacy merely “reflect[s] beliefs about whether the activity effectively promotes societal welfare, as defined by the audience’s socially constructed value system” (1995: 579, emphasis added), meaning that it solely incorporates perceptions of normativity into the descriptive lens on legitimacy. However, this nuance is not always captured in due precision when moral legitimacy is equated with “normative legitimacy” (Díez-de-Castro, Peris-Ortiz, & Díez-Martín, Reference Díez-de-Castro, Peris-Ortiz, Díez-Martín, Díez-De-Castro and Peris-Ortiz2018: 7) or described as a “normative concept” (Lock & Schulz-Knappe, Reference Lock and Schulz-Knappe2019: 3). Consequently, the descriptive and the normative lens are blurred. Following the second cornerstone of our framework, we hold that good theory building asks for information about how the relationship between descriptive and normative legitimacy is constructed. While it is well-researched how elements of normative legitimacy can be integrated into concepts of descriptive legitimacy (as illustrated by Suchman’s work and its widespread reception), future research should dive deeper into how descriptive legitimacy can be integrated into concepts of normative legitimacy.
Ambiguity of assessment also appears in various works discussing the legitimacy of other institutions or organisations. The literature features many examples in which fine-grained criteria help to assess the legitimacy of institutions or organisations, such as corporations (e.g., Díez-Martín, Blanco-González, & Díez-de-Castro, Reference Díez-Martín, Blanco-González and Díez-de-Castro2021), governments (e.g., Gilley, Reference Gilley2006), or NGOs (e.g., Baur & Palazzo, Reference Baur and Palazzo2011). Often, these works do not reflect on whether the criteria they suggest provide more explanatory power through the descriptive or the normative lens. This demonstrates that the third cornerstone of our framework is also relevant in other areas of legitimacy research. Future studies could compare the proposed legitimacy criteria for different institutions to learn from their overlaps and differences.
Avoiding Erosion of Normativity
While we see erosion of normativity as particularly relevant in MSI research, we also contribute to the more general debate of value-ladenness of management and organisation studies (Hodgson, Reference Hodgson1988; Seeck et al., Reference Seeck, Sturdy, Boncori and Fougère2020). This debate has called into question whether value-free theories can even exist (Vallentin & Murillo, Reference Vallentin and Murillo2022), arguing that scholars can hardly let go of their convictions or values. Our analysis delves deeper by emphasising the risk of researchers overlooking normative elements in their constructs. This neglect may create a false sense of value-neutrality. Normative elements most likely affect research, even in seemingly value-free publications, which is why they must be acknowledged. Thus, we see a need for future management research exploring such assumptions.
The debate of value-ladenness bears resemblance to what has been discussed as the “separation thesis” of business ethics, which questions whether it is possible to study descriptive and normative approaches or methodologies in isolation (Sandberg, Reference Sandberg2008; Singer, Reference Singer1998; Smith et al., Reference Smith, Donaldson, Pouryousefi, Scholz and Spence2023). In this regard, our analysis demonstrates that any attempt to separate descriptive and normative approaches of legitimacy would only be possible with an adequate degree of construct clarity. Still, as has been illustrated by our example of Habermasian discourse ethics, drawing on the interdependencies between descriptive and normative legitimacy can lead to more comprehensive theories of legitimacy.
Moreover, this paper contributes to the debate “about a possible deficit of normativity in management research, practice, and education” (Smith et al., Reference Smith, Donaldson, Pouryousefi, Scholz and Spence2023: 785; Lock & Seele, Reference Lock and Seele2015; Seele, Reference Seele2016). By identifying the lack of construct clarity as a driver of erosion of normativity, we provide a starting point for addressing this deficit through the three cornerstones of our framework. Many topics entail descriptive as well as normative elements. For example, researchers who empirically assess the concept of “trust” might do so by first pointing out the normative basis that explains why “trust” is deemed crucial (first cornerstone); second, explaining the relationship between the subjective perception of “trust” and objective reasons for “trust” (second cornerstone); and finally, asking whether it is more insightful to ask people about their perceptions of “trust” or to search for objective ways to assess it (third cornerstone). In a similar vein, in their study on descriptive and prescriptive norms, Scharding and Warren (Reference Scharding and Warren2023: 1) point out how imprecise language can lead to “the naturalistic fallacy … wrongly blurring the distinction between ‘what is … and ‘what ought to be’.” Relatedly, our three cornerstones for solid theory building in MSI legitimacy research could help in situating different kinds of norms between the descriptive and the normative form. We conclude that an analytical approach to any study that concerns concepts with descriptive and normative elements, be they values, norms, or—as in our case—legitimacy, can help avoid a possible erosion of normativity in business ethics research.
Acknowledgements
We are grateful to Jill Brown (Bentley University), Gonzalo Conti (MLU Halle-Wittenberg), Andreas König (University of Passau), Thomas Paul Moliterno (VU Amsterdam), Raphael Ng (HHL Leipzig), Ingo Pies (MLU Halle-Wittenberg), Katinka Quintelier (VU Amsterdam), Omar Solinger (VU Amsterdam), and Pieter de Wit (VU Amsterdam) for their valuable feedback on earlier versions of this paper. Also, we would like to thank the participants of the International Association of Business and Society 2020, the Business and Society Seminar 2021 at the University of Namur, the Academy of Management 2021, the Society of Business Ethics 2021, the VU Research Seminar 2021, and the Academy of Management Review PDW at the Vrije Universiteit Amsterdam in 2022 for their helpful comments. Finally, we extend our gratitude to the editor Andreas Rasche and three anonymous reviewers for their constructive feedback throughout the review process.
Adrian Gombert’s research was partly funded by a PhD scholarship of the Karl Schlecht Stiftung, Germany.
Adrian Gombert is a PhD candidate at the Martin-Luther-University Halle-Wittenberg and the doctoral programme of the Wittenberg Center for Global Ethics. His research focuses on the political role of corporations and the legitimacy of multi-stakeholder initiatives.
Rebecca C. Ruehle (r.c.ruehle@vu.nl, corresponding author) joined the Vrije Universiteit Amsterdam as an assistant professor of business ethics in 2020 after she completed the doctoral programme of the Wittenberg Center for Global Ethics with a PhD from the Martin Luther University Halle-Wittenberg. Her research is situated at the intersection of management and ethics. She studies questions of justification, legitimation, autonomy, manipulation, nudging, and technology in the organisational context.
Open access
