Responsibility for sustainability in global supply chains

10The trend toward global sourcing and outsourcing has spawned more complex, dynamic, non-linear and fragmented supply chains (Mena et al., 2013), resulting in the acceleration of environmental and social damage, increasingly discussed as inter-organisational rather than solely intra-organisational issues (Hartmann, 2021). Along with SSCM scholars (Amaeshi et al., 2008; Hartmann & Moeller, 2014), researchers in the fields of political corporate responsibility (Scherer & Palazzo, 2011), global value chains (Gereffi et al., 2005) and strategic management (Kotabe & Murray, 2004) have focused on the extent to which MNCs are expected to be responsible not only for actions within their own boundaries but also for those of others with whom they are socially or commercially connected (Frigant, 2015).

11In the SSCM literature, this question has been mainly investigated through the notions of ‘boundaryless responsibility’ (Amaeshi et al., 2008), ‘responsibility attribution’ (Hartmann & Moeller, 2014) and ‘extended responsibility’ (Spence & Bourlakis, 2009). Despite semantic differences, all these terms imply that focal companies control and dictate economic, environmental and social goals throughout their entire supply chain (Amaeshi et al., 2008; McCarthy et al., 2018). Yet, the meaning of the term ‘responsibility’ is contentious, and different perspectives evince its richness and ambiguity. For instance, assumptions underlying the notion of ‘responsibility’ are viewed as ascribing too much power and autonomy to corporations in setting the terms of their own conduct, without considering potential greenwashing or abuse towards weaker actors in the Global South (Banerjee, 2008, 2018; Newell, 2005). Concurrently, authors adopting a normative perspective on responsibility (Amaeshi et al., 2008) contend that the more powerful party in the relationship is morally accountable for its actions and holds responsibility for the weaker ones (Amaeshi et al., 2008; Touboulic et al., 2014).

12Within the SSCM literature, responsibility tends to be viewed from a rationalist perspective, suggesting that the more powerful par ty in a supplier-customer relationship is responsible for dictating the meanings and practices of sustainability and exer ts control over the economically dependent par tner in a top-down hierarchical manner (Amaeshi et al., 2008; Formentini & Taticchi, 2016; Grimm et al., 2016). Interestingly, whilst focal companies are considered responsible and tend to define sustainability meanings and practices for other supply chain actors, they simultaneously transfer the responsibility for implementation to others (Pizzetti et al., 2019; Touboulic et al., 2014, 2018). Some authors argue that this has led to wrongly blaming the weaker parties in the supply chain (e.g., farmers and workers) for environmental and social problems, rather than the MNCs upon which they are economically and operationally dependent (Banerjee, 2003; Glover & Touboulic, 2020).

13The construction and attribution of responsibility in the context of supply chains cannot be detached from the nature of the relationships in these chains, which are embedded in power dynamics, social norms and political discourse (Black, 2008; Moerman & van der Laan, 2015). Previous research has shown that responsibility can be constructed and attributed through both material and discursive practices (Etchanchu & Djelic, 2019; Glover & Touboulic, 2020) and is, therefore, a political and power-laden notion. Material practices include transactional and contractual mechanisms, such as audits, performance assessment and the inclusion of sustainability requirements in supplier contracts (Glover & Touboulic, 2020; Touboulic et al., 2014). Whilst the SSCM literature has largely focused on such material practices, recent studies have started to explore discursive practices around sustainability in the context of supply chains and inter-organisational networks, evidencing the power dynamics at play (Glover & Touboulic, 2020; McCar thy et al., 2018).

A discursive perspective on power

14In recent decades, discursive approaches have been more widely adopted to explore power dynamics across various topics, including sustainability studies (see, for example, a special issue of Forest Policy and Economics, Giessen et al., 2009). Discourse not only contains linguistic components but also encompasses a wide variety of visual representations: images, videos, photographs, charts and graphs (Hardy & Thomas, 2014; Höllerer et al., 2018). From a Foucauldian perspective, discourse cannot be separated from power, since all power relations ‘cannot themselves be established, consolidated nor implemented without the production, accumulation, circulation and functioning of a discourse’ (Foucault, 1980, p. 93). Discourse can be considered as a ‘subtle, but nonetheless pervasive’ form of exercising power or as manipulation (Hardy & Thomas, 2014, p. 322; Van Dijk, 2006). Focusing on discourse is an effective way to investigate power dynamics within and between organisations (Hardy & Phillips, 2004). Discourse is a central concept in the power literature, which investigates how actors using ‘texts as weapons’ have the potential to skew the power relations to benefit their own interests (Hardy & Philipps, 2004). Considering the growing political role played by corporate actors on the global stage (Scherer & Palazzo, 2011), it makes sense to take a close look at their discursive practices. Arguably, the discursive power of MNCs allows them to define the rules of the game (Banerjee, 2018; Courpasson & Golsorkhi, 2011). In the supply chain literature, Western focal firms are constructed as the primary change agents for sustainability, and the dominant narrative around responsibility gives primacy to their interests and role (Banerjee 2003, 2008; McCarthy et al., 2018). The dominant narrative centred on the actions and interests of large corporations simultaneously makes these actors the villains and heroes of sustainability in global supply chains (McCarthy et al., 2018). Critical scholars have drawn our attention to the instrumentality of corporate-led sustainability initiatives and the disconnect between talk and action, e.g., greenwashing and corporate hypocrisy (Glover & Touboulic, 2020). The disconnect between what is said, i.e., sustainability narratives and discourse, and what is done is well researched in the corporate sustainability literature (Pretorius, 2016). It has been argued that such greenwashing serves as a mechanism through which corporations can maintain their authority and legitimacy (O’Sullivan, 2005; Whelan et al., 2019; Zott & Huy, 2007), often at the expense of less powerful players (Banerjee, 2008). MNCs exercise their power through the deployment of emancipatory sustainability narratives, as well as discourses of risk mitigation, assigning responsibility to other par ties, as a means to protect their corporate image and enhance their legitimacy whilst under taking relatively few actions themselves. This highlights the symbolic rather than substantive nature of corporate sustainability claims (Amaeshi et al., 2008; Banerjee, 2008; Gao & Bansal, 2013; Garriga & Melé, 2004) and shows that it is not possible to detach the construction and attribution of responsibility for sustainability in supply chains from the power dynamics at play.

15The performative character of corporations’ sustainability discourses is well researched (Banerjee, 2018; Carlos & Lewis, 2018; Christensen et al., 2013; Glover & Touboulic, 2020). These discourses serve to construct and enact socio-environmental practices, i.e., self-regulation, that can be passed onto others, such as large buying firms imposing sustainability requirements on their suppliers (Grimm et al., 2016). Several examples of self-regulation can be found in the SSCM literature, including the increasing number of codes of conduct, through which focal companies establish and enforce their own environmental and social standards on suppliers (Egels-Zandén, 2014; Gualandris et al., 2015). Etchanchu and Djelic (2019) show how Tesco (a major retailer) explicitly uses discursive power to force its suppliers (farmers) to adhere to the company’s standards of responsible behaviour. In another study, Spencer and Bourlakis (2009) have shown how Waitrose, another UK retailer, defines sustainability standards and expectations for their suppliers. These studies illustrate how the construction and attribution of responsibility for sustainability in the supply chain through the discursive strategies of leading actors is a manifestation of their power which translates into unilaterally imposed practices. Suppliers find themselves with limited agency with regard to environmental and social practices and are obliged to comply with the norms defined by the focal companies.

MS-MOs as an alternative: Towards collective forms of responsibility

16The growing importance of corporate-led sustainability practices and self-regulation identified earlier is evidence of the growing political role of MNCs in the globalised economy (Scherer & Palazzo, 2011). Most self-regulation mechanisms operate through collective organisations, such as industry unions (Marques, 2017), trade associations (Lawton et al., 2018), multi-stakeholder initiatives (Scherer & Palazzo, 2011; Whelan et al., 2019) and standards organisations (Berkowitz & Souchaud, 2019), all of which can be classified as MOs, i.e., organisations whose members are other organisations (Ahrne & Brunsson, 2005). The MO concept has emerged to distinguish the diverse dynamics of collective versus individual organisations (Ahrne & Brunsson, 2005).

17Proponents of MOs argue that collective action is not only necessary but also critical in order to address societal grand challenges, which call for the par ticipation of a wide range of actors (Berkowitz et al., 2017). Whilst most of the literature to date has been focused on MOs composed of a single type of member, e.g., organisations in the same sector with some notable exceptions (Berkowitz et al., 2017, 2020; Carmagnac & Carbone, 2019; Laurent et al., 2020), there is growing evidence that societal grand challenges call for a MS-MO approach integrating a wide-range of heterogeneous actors, such as businesses, NGOs, and public and scientific institutions (Berkowitz et al., 2017, 2020). Bringing together diverse exper tise and capabilities has been shown to be more effective than solo strategies to address sustainability challenges on a global scale (Ferraro et al., 2015; George et al., 2016). Recent studies have highlighted how MS-MOs can create and sustain an innovative space to tackle sustainability challenges related to oceans and to the energy and palm oil sectors (Berkowitz et al., 2017, 2020; Carmagnac & Carbone, 2019).

18Considering the emphasis placed on MS-MOs as central actors striving for global sustainability, one may assume that they have a key role in reshaping not only the attribution of responsibility for sustainability but also the power dynamics between the different actors involved and the shift toward collective and collaborative approaches. For instance, it has been argued that responsibility attribution is socially constructed and, thus, more likely to be discussed and collectively negotiated within a multi-actor entity that is invested with a societal mission (Moerman & van der Laan, 2015; Reinecke & Ansari, 2016). In a space where heterarchy prevails and decisions are made by consensus, hierarchical power has little strength (Berkowitz et al., 2020). As consensus-based deliberation formally attributes an equal voice to all MS-MO members, they would have a comparable contribution to the construction and attribution of responsibility for sustainability; they would become the co-authors of such collective responsibility.

19Despite the growing body of work on MS-MOs’ role in tackling sustainability issues (Berkowitz et al., 2017, 2020; Carmagnac & Carbone, 2019), there is still a dearth of research looking to unveil subtle forms of influence and power within these MS-MOs. In this study, we are interested in challenging the prevailing assumption that MS-MOs should be considered as the best way to achieve the sustainability transformation. We need to unpack MS-MOs’ discursive practices in order to fully understand the political and power dimensions at play in shaping the construction and attribution of responsibility for sustainability on a global scale.

Approach: M-CDA

21Critical Discourse Analysis (CDA) is an interdisciplinary methodology used to critically investigate the linguistic aspects of social phenomena and expose power implications (Balogun et al., 2014). CDA problematises discourse as an instrument and effect of power: ‘discourse transmits and produces power; it reinforces it, but also undermines it and exposes it, renders it fragile and makes it possible to thwart it’ (Foucault, 1978, p. 101). This approach prioritises social relations over individual actions or entities, making it highly appropriate for our study. CDA consists of the systematic unpacking of often opaque relationships between (1) discursive strategies, events and text, and (2) the broader social structure, relations and processes. In line with Phillips et al. (2008), we argue that a discursive strategy is a struggle for power within and between organisations, with the capacity to define concepts and subject positions and to control how the objects are perceived and handled. Like McCarthy et al. (2018), we adopt a multimodal approach to CDA (M-CDA) to include multiple types of sources and expand the analysis beyond text to language, images, sound and gestures (Höllerer et al., 2018). With this approach, we explore the interactions, relationships and contradictions between these elements in order to unravel the narratives and counter-narratives associated with different stakeholders.

Research context

22We focus on a single MS-MO, the RSPO, which is conducive to rich data collection for greater depth of analysis and increased precision, revealing unique insights that other collective arrangements would not have provided (Siggelkow, 2007). In this field, the production, processing and trading of palm oil have all sparked controversies related to intertwined negative environmental and social impacts (Schouten & Glasbergen, 2011). The RSPO is a voluntary membership organisation, whose aim is to tackle the sustainability challenges in palm oil supply chains through a single certification scheme. It is the most established and successful MS-MO in the sector (De Man & German, 2017) and one of the fastest-growing MS-MOs when compared to similar initiatives developed in other commodity sectors (De Man & German, 2017). In its 13 years of existence, the RSPO has certified a 19 % share of the total global palm oil market and brought more than 5000 heterogeneous members together. As shown in Figure 1, its membership comprises three categories: associate members (organisations that purchase less than 500 metric tons of palm oil products annually), affiliate members (universities, research centres and sponsors) and ordinary members (retailers, manufacturers, traders, growers, financial institutions, environmental and social NGOs) (RSPO, 2021). The RSPO’s governance structure consists of a General Assembly, an Executive Board and several working groups and task forces. The General Assembly is the highest decision-making authority. The General Assembly meetings are held annually, and the decisions are made by a simple majority vote of only the ordinary members. As manufacturers are the most numerous amongst RSPO ordinary members, they easily outvote the others and dominate the decisions in the General Assembly (De Man & German, 2017; Schouten & Glasbergen, 2011). The executive board is composed of 16 representatives of the ordinary members. The formation of working groups and task forces is designed to address specific, emergent issues. Decision-making within the board of governors, working groups and task forces is based on consensus, which has been widely explored in the literature and criticised by some RSPO members. On the one hand, consensus is viewed as a strength of the RSPO, as it gives equal voice to all members. On the other hand, to meet the expectations of all the different stakeholders, decision-making based on consensus often becomes a long process. This leads to lowest common denominator decisions, which result in less strict standards (Schouten & Glasbergen, 2011). This MS-MO constitutes a fascinating context to explore subtle forms of power, where discourse can shape roles and attribute different degrees of responsibility to the actors involved.

Figure 1

RSPO members in numbers (RSPO, 2021) RSPO, Roundtable on Sustainable Palm Oil; NGOs, non-governmental organisations.

RSPO members in numbers (RSPO, 2021) RSPO, Roundtable on Sustainable Palm Oil; NGOs, non-governmental organisations.

Table 1

Data sources

Type of source	Data source	Total	Detail	N	Further information
Primary	Inter views	22	Oil palm growers	4	Malaysian and Brazilian companies
			Traders	5	American and French companies
			Manufacturers	9	Food, cosmetic and consumer goods MNCs
			RSPO staff	4	Located in Indonesia
Secondary	Videos	69	Inter views (VW)	39	Interviews conducted with RSPO members during annual meetings (2014–2017)
			Institutional (VC)	18	Videos that explain the RSPO’s operating mode in order to engage and recruit new members
			Smallholders (VS)	12	Testimonials of smallholders located in Indonesia, Malaysia and Thailand

Data sources

MNCs, multinational companies; RSPO, roundtable on sustainable palm oil.

Data sources

23Our analysis relies on two main sources of data: 69 videos and 22 inter views (Table 1). The videos represent the main source of empirical evidence, comprising interview (VW), institutional (VC) and smallholder (VS) videos. Inter view videos are produced during RSPO meetings and showcase RSPO members’ opinions about the achievements and challenges of the MS-MO. Institutional videos promote the RSPO’s activities, by presenting the organisation’s operating mode, raising awareness about palm oil, engaging stakeholders and recruiting new members. Smallholder videos are a set of short testimonials by RSPO smallholders, demonstrating the importance of their engagement in the sustainable palm oil supply chain.

24In addition, 22 semi-structured interviews were conducted with RSPO staff and members from 2017 to 2019. Each interview lasted approximately 90 min. The interview protocol covered questions that explore informants’ perceptions regarding the role of the MS-MO in addressing sustainability issues. The selection criteria for interviewees attempted to reflect the diversity of members.

Data analysis

25Our analysis followed an iterative process, looking into the differences, similarities and complementarities present throughout the different materials (Höllerer et al., 2018). We conducted our analysis in three stages (Elliott & Stead, 2018). First, all interviews and videos were carefully recorded and transcribed. Second, we carried out an open coding process to familiarise ourselves with the data. Working individually, each researcher coded 46 videos and 14 inter views, with each video and interview being coded by two people independently to ensure greater reliability. In this stage, we took notes on the main discourse features, including the context, e.g., main characters and the positioning of the audience, as well as the content such as the recurring themes and narrative patterns present in the discourse of both the RSPO and its members.

26In the third stage, we discussed our individual coding as a group before agreeing on and coding the recurrent themes considered to be relevant. Four categories of themes emerged in this stage: (1) the construction of sustainability discourse; (2) the main actors identified as responsible for sustainable practices; (3) the initiatives/mechanisms adopted by the RSPO to manage such practices; (4) the RSPO’s outcomes and promises. In order to adequately explore the visual and kinetic aspects of the videos, we also included six other categories of analysis: (1) video format/forms, e.g., cartoon, documentary and whiteboard animation; (2) sound, e.g., soft background music, noise and people talking; (3) landscape and settings, e.g. exterior/interior, countryside/city and private/commercial; (4) people’s appearance, e.g. female/male, young/old and clothing; (5) postures and gestures, e.g. eye contact and physical distances; (6) who is speaking. During this analysis, we carefully observed how imagery may reflect social patterns and reaffirm power dynamics. As Elliott and Stead (2018) argue, the position of an individual in the image may imply/indicate specific power imbalances or suggest a relative social position, power, reputation and renown. In this stage, we also paid special attention to latent elements in the discourse that were not evident at first glance. To detect these elements, we looked for expected and unexpected ‘absences’ in the discourse using questions, such as who is speaking to whom, in the name of whom, and about what (Van Dijk, 2006). Table 2 provides an illustration of the complete coding scheme adopted.

Discursive construction of smallholders and associated sustainability issues

29Smallholders are small-scale farmers with less than 50 hectares who rely on family labour and depend on the oil palm crop as a source of subsistence (RSPO, 2021). They account for approximately 40 % of global palm oil production, making them ‘significant contributors towards a sustainable oil palm industry’ (RSPO, 2021).

30Smallholders appear in three different roles in our data: (1) as the main characters in a series of videos showcasing personal life-stories of smallholders (VS videos); (2) as the target audience for several institutional videos (e.g., ‘Certification for Smallholder’); (3) as the primary subject of the narratives in videos and interviews, being positioned as the central element of the discourse and the main actor to which the responsibility for sustainability is attributed.

31Overall, smallholders are characterised in text and imagery as poor and uneducated, conveying their dependence on the RSPO certification, and as a remote and exotic group of isolated actors, located far from all other supply chain actors (Table 3). The data show how these associations are crafted, and how the videos have been carefully edited to reinforce such representations. For instance, the representation of smallholders as poor and uneducated is recurrent throughout all smallholder series videos (VS). Most of these videos emphasise the improvements that joining the MO has brought to smallholders’ lives. The videos depict a dichotomy of before and after the RSPO. The narrative reinforces the idea that smallholders had insufficient income, information and knowledge; improper housing and inadequate educational resources for their families prior to joining (e.g., ‘Before, my house didn’t have walls and I only had a bicycle’, ‘Before joining RSPO, it was very difficult for us to sell fruit’, ‘Before, our income was modest’. VS_01;VS_05;VS_09). In contrast to the ‘before’ narrative, the images portray the ‘after RSPO’, showing smallholders getting training, new houses, education for their families and new vehicles. Interestingly, these videos are an illustration of how reality is artificially crafted by the RSPO; even though each one is supposed to feature the story of a specific smallholder, many videos contain identical footage: same home, same children, same school and same training (Figure 2).

32Representations of smallholders as exotic and remote are also recurrent. Abstract terms such as‘on the ground’ (VW_01,VC_03, interviews with traders) or the ‘rest of the world’ (VW_16) are frequently employed to convey ‘remoteness’. The representation of smallholder exoticism clearly emerges from the video ‘Sabah commits the entire state to sustainable palm oil practices by 2025’ (VC_15), in which a multi-stakeholder group is seen crossing a narrow bridge to visit a palm oil plantation in a 4WD vehicle, implying that the plantation is remote and difficult to access. In the video, the group ‘exploring’ the field is portrayed as ‘tourists’, wearing backpacks and taking pictures, like on a safari.

33The sustainability topics associated with smallholders are constructed through references to negative phenomena, such as deforestation, fire, biodiversity destruction, use of chemicals, pover ty and human exploitation. Such associations implicitly convey a relationship between smallholders and poor social and environmental practices in the supply chain (e.g., ‘today it is mainly because of smallholders that we have deforestation… they represent 40 % of production’ – Interview with a trader). This construction of (un) sustainability in relation to smallholders is strongly reinforced by the images used (Table 4).

Table 3

Main representations of the smallholder subject position

	Visual representation	Usual context	Illustrative quotes
Exotic/remote	VC_04: RSPO Smallholder Suppor t Fund VC_02: RSPO Cer tification for Smallholder	VC_02 and VC_04: Farmers are homogeneously represented, all wearing the same work uniform (blue overalls and black boots). Moreover, they are represented inside a box or circle (which conveys that they are separated from the rest of the world or from the other actors).	VC_02: ‘currently they are millions of smallholders around the world who depend on palm oil for their livelihood’.‘They are farmers with less than 50 hectares of oil palm, their principal source of income. And their family is the main source of labour’.
Poor and uneducated	VS_03: RSPO Smallholder Series - The Story of Darus	VS_03: The smallholder, called Darus, is sitting on a carpet on the floor somewhere that seems to be his house. The house is very simple (it looks unfinished, a broom can be seen in the background). He is barefoot and wearing a sweaty shirt. His wife is sitting directly on the floor behind him.	VS_03: ‘From working in these plantations, I can fulfil my family’s daily needs, renovate my house’...
	VS_04: RSPO Smallholder Series - The Story of Namo	VS_04: Smallholders are being trained by a member of RSPO staff. The RSPO staff member is wearing a polo with the RSPO logo and a big wristwatch. He is the only person who is not wearing a security helmet in the PO field. Smallholders, wearing white security helmets, look concentrated and are paying attention to what is being said by the RSPO staff member.	VS_04: ‘When independent smallholders get financial suppor t and coaching by RSPO they are able to implement environmentally friendly practices’.

Main representations of the smallholder subject position

Discursive construction of consumers and associated sustainability issues

34Whilst we have smallholders at one end of the supply chain, consumers constitute the other extreme. Consumers comprise all individuals who, consciously or not, buy products containing palm oil. Despite the share of responsibility attributed to consumers as subjects, their voices are completely absent. These actors consistently appear in two different forms: either as the target audience for a set of institutional videos (VC_05, 06, 07, 13 and 14) or as the object of the discourse created by both the RSPO and its members. They are visually and textually por trayed uniformly as greedy, unaware and naïve (Table 5). The ‘greedy’ representation is apparent in text associating consumers with increasing demand and consumption of products containing palm oil and strongly reinforced by a range of visual elements. For instance, in a series of over-simplified cartoon videos (VC_05, 06, 07), products containing palm oil (noodles, ice cream and cookies) disappear from the plate, suggesting that they had been consumed regardless of whether or not they were truly sustainable. Consumers are also represented as unaware and naïve, especially in their lack of knowledge of the power they hold through consumption. Subtitles in the car toon video ‘Episode 5 – Palm Oil What Can You Do’ explicitly address consumers: ‘Did you know that with every swipe and tap comes great power? Your ever yday choices can have a positive impact on the planet and reduce deforestation and climate change. Choosing sustainable palm oil is an easy way to make a big difference’, VC_18).

35In line with the naïve and unaware characterisation, the sustainability topics associated with these actors are framed around products with ‘bad’ palm oil or non-certified palm oil products (Figure 3). Some par ts of the narrative suggest that consumers will avoid spending more for ‘good’ sustainable products even after becoming aware of these issues (e.g., ‘Consumers are still not willing to pay for the RSPO premium’ – inter view with a trader). Similarly, the representation of ‘bad’ palm oil products is often constructed using a mix of complementary and contradictory messages in text and images. Whilst the images of one set of videos simply show everyday products that contain palm oil, the accompanying texts question whether these products come from a ‘good’/cer tified source, leaving the audience, i.e., consumers of these products, in doubt (e.g., ‘how can you be sure that the palm oil product sold as RSPO certified comes from RSPO certified sources?’, VC_10). The video ‘Episode 3 – Let’s Transform the Palm Oil Sector’ (VC_16) also illustrates this dialectical link between text and imagery. Whereas the image presents a variety of familiar products (bread, soap and ice cream) that contain bad palm oil, the text suggests that reducing the consumption of such products may not be a good solution: ‘You are probably thinking ‘I should just boycott palm oil’. Not so fast. Boycotting seems like a quick fix, but it could actually lead to a worse outcome’ (VC_16).

Figure 2

Screenshot of four different videos illustrating the same training (VS_03,VS_05,VS_01 and VS_04)

Screenshot of four different videos illustrating the same training (VS_03,VS_05,VS_01 and VS_04)

Discursive construction of traditional supply chain actors and associated sustainability issues

36Here, ‘traditional SC actors’ refers to large-scale growers, traders, manufacturers and retailers that commercialise more than 500 metric tonnes of oil per year. Examples include Unilever, one of the co-founders of the RSPO, which is recognised as the largest palm oil manufacturer, consuming approximately 3 % of global supply (Von Geibler, 2013) and five Indonesian traders, which together control about 90 % of the palm oil trade market (Pacheco et al., 2020). These actors, as explained before, are the only players with the right to vote and to hold seats on the board of governors. Moreover, due to their over-representation (see Figure 1), they outvote other categories of members (e.g., financial institutions and environmental and social NGOs) in the General Assembly Meetings (Carmagnac & Carbone, 2019; Schouten & Glasbergen, 2011). This clearly creates an imbalance of power amongst RSPO members. Indeed, several studies have criticised the RSPO, saying that despite its international scope, it is still recognised as a European initiative to address the demands of Western MNCs and does not attend to the needs of the smallholders (Pacheco et al., 2020; Schouten & Glasbergen, 2011).

37Traditional SC actors are the main characters in the interview videos. However, they rarely talk about themselves or about their own responsibility in addressing sustainability issues (e.g., ‘sustainability is always long-term commitment, but it is essential that the demand comes from the consumer’ – interview with a food manufacturer). The only exception is when they refer to their active role in purchasing RSPO cer tified PO, highlighting a positive association between these traditional actors and the RSPO certification as the main solution for a sustainable transition.

38Beyond their ‘endorsement’ responsibility in growing the RSPO certified market, they are not explicitly represented in the RSPO’s sustainability narrative. An example of this occurs in the video ‘United Nations Environment Programme (UNEP) and RSPO partner to raise profile of Sustainable Palm Oil’, where SC actors are not even alluded to: the RSPO brings together ‘producers, consumers, environmental NGOs, and also national governments’ and ‘ [the RSPO] allows consumers and producers to be rewarded for a more sustainable consumption and production pattern’ (VC_03). We refer to them as hidden actors, or actors in the shadows, seemingly shielding themselves behind the RSPO ‘black-box’. These actors are depicted in the text and imager y as a simple link between the two extremes of the supply chain (Table 6). Despite their specific role in the supply chain (as refiners, traders, manufacturers and retailers), they are often referred to as an undifferentiated group of entities, i.e., ‘organisations along the supply chain’ (VC_17), ‘brands that are buying palm oil’ (interview with RSPO staff) and ‘the whole supply chain’ (VW_14).

Table 4

(Un) sustainable topics associated with smallholders

Unsustainable element	Visual representation
Deforestation	VC_03	VC_15
Fire	VC_03	VC_03
Pover ty	VC_03	VC_09
Biodiversity destruction	VC_11	VC_11
Use of chemicals	VC_15	VC_07
Human exploitation	VC_15	VC_15

(Un) sustainable topics associated with smallholders

39The complexity of the supply chain is used rhetorically to reduce these traditional actors’ involvement to ‘a simple link’. This is well-exemplified by a trader, who states ‘We are all suppliers, customers, and growers. It is a very spider [web-like] market... And it is complicated for people who are not part of the business to understand the supply chain behind that’. There is an interesting contrast between their conspicuous absence from the RSPO discourse and the fact that these actors not only fill most seats on the board of governors but also outvote other members and dominate the decision-making process (Schouten & Glasbergen, 2011). We can infer that traditional SC actors behind the RSPO black-box are the key players who create the rules for the other members, mainly smallholders, to follow (e.g., ‘the aim [of RSPO certification] is to ensure smallholders improve their practices by complying with the RSPO P&Cs [principles and criteria]’,VC_02).

Figure 3

(Un)sustainable topics associated with consumers

(Un)sustainable topics associated with consumers

Table 5

Main representations of the consumer subject position

	Visual representation	Usual context	Illustrative quotes
Naive and unaware	VC _ 09: RSPOTrademark App	VC_09: The protagonist of this video is a young Asian female consumer wearing a yellow sweater, jeans and a Pokémon face mask. She is using her cell phone to verify which products have the RSPO label in a supermarket (where all products and signs are written in English). She demonstrates surprise when she finds a certified PO product. All other young Asian clients in the supermarket are holding cell phones and searching for Pokémon. At the end of the video, two messages are directly addressed to these unaware consumers: ‘download the app’ and ‘always choose RSPO-cer tified goods’.	VC_09:‘Cool’;‘Hey did you catch any Pokemons? Actually, I’m catching something else’.
Greedy	VC_06: Ice cream contains palm oil	VC_06: The video has very funny background music, similar to a video game, and informs the audience that ice cream contains palm oil before asking whether this is good or bad. In the video, we can see on the table a glass of juice, a spoon, two bottles of sauce and the ice cream. Interestingly, everything is real, except the ice cream, which is shown as balls of wool and is the only object that disappears. This may suggest that, whether the ice cream is made from a good or bad palm oil, the product is consumed.

Main representations of the consumer subject position

Table 6

(Un) sustainable topics associated to traditional supply chain actors

Unsustainable element	Visual representation
Climate change	VC_03
Ecosystem degradation	VC_03  VC_03

(Un) sustainable topics associated to traditional supply chain actors

40Sustainability topics associated with traditional SC actors highlight global challenges, such as climate change and ecosystem degradation. Interestingly, a closer look at the visual elements reveals that the construction of these global environmental challenges is associated exclusively with upstream actors of the supply chain, i.e., millers and refineries (Table 7). There are no images or texts linking these issues to the large focal firms in the supply chain, i.e., manufacturers and retailers.

Discursive construction of the RSPO’s responsibility for sustainability

41In exposing the ways in which subject and object positions are textually and visually framed in relation to each other, we are able to tease out the implications of responsibility construction and attribution for sustainability concerning the MS-MO itself (Table 8). We show that the RSPO as a subject position is constructed around two main representations, which, in turn, relate to several promises and practical solutions.

42First, to remedy smallholder ‘poverty’ and exploitation, the RSPO promises fair working conditions and economic prosperity by providing funds and price premiums to smallholders. According to the RSPO, these mechanisms help smallholders by increasing their incomes, hence giving them the opportunity to expand their plantation areas, build homes and educate their children. Next, the RPSO addresses the question of ‘uneducated’ smallholders by promoting sustainable agricultural practices and professionalisation through control and management systems and technical training. Finally, despite their ‘remoteness’, smallholders’ access to the market is ensured through their participation in the RSPO’s certification scheme.

43In order to tackle the (un) sustainability of ‘bad’ palm oil products at the other end of the supply chain, where consumers are depicted as ‘unaware’, ‘naïve’ and/or ‘greedy’, the RSPO promotes increased sustainability awareness and stimulates demand for sustainable products. This is done by providing education on responsible decision-making (‘the right thing to do’), empowering consumers through mobile applications and creating an easily identifiable ‘sustainable palm oil’ label to facilitate sustainable purchases.

44Finally, the RSPO is constructed as erasing the complexity of the supply chain, representing traditional members as a ‘simple link’ or ‘hidden actors’ in the chain, by providing them with tools to ensure the traceability of PO products and different types of cer tification schemes. Moreover, the RSPO addresses the hidden facets of traditional supply chain actors by protecting their reputational image, addressing risks and aligning with the companies’ identities.

45The RSPO’s subject position is carefully crafted, notably around promises and solutions to tackle the (un) sustainable issues related to different actors in the supply chain. In all text and visual elements, the MS-MO is constructed as ‘the only practical and legitimate solution’, as a credible, global standard and as an ‘all powerful’ entity (Table 9).

Table 7

Main representations of the traditional supply chain actor subject position

	Visual representation	Usual context	Illustrative quotes
A simple link	VC_10: RSPO PalmTrace	VC_10: The video details the Palm Oil Trace, a tool that enables certified palm oil traceability from mills to retailers. The video shows the role of each supply chain actor in inputting data to enable PO traceability. Focal companies are not especially highlighted, but described as a link between other players in a linear chain.	VC_10: ‘Manufacturers and retailers purchase certified sustainable oil palm products, manufacturers then produce goods containing certified sustainable oil palm products, manufacturers and retailers can purchase RSPO credits in RSPO palm trace and thereby directly incentivize the production of cer tified sustainable oil palm products’.
Hidden	VC_17: Episode 4 - What Is Sustainable Palm Oil	VC_17: A cartoon video with soft music in the background. In this video, focal firms (manufacturers and retailers) are described as ‘companies along the supply chain’. Emphasis is given to NGOs and financial institutions. All the characters in the video are of different ethnicities and from different professions (e.g., a tan Asian man wearing a farming hat and a Black businessman wearing a suit and tie and using a phone)	VC_17: ‘companies along the supply chain... they have teamed up under the RSPO’

Main representations of the traditional supply chain actor subject position

Whose responsibility? Challenging the shift towards a collective narrative in SSCM

47Research in SSCM is predominantly based on assumptions of linearity, hierarchy and power exerted by focal firms over their direct and indirect suppliers (Rauer & Kaufmann, 2015; Touboulic et al., 2018). However, recent studies have highlighted how collective bodies such as MS-MOs can assume an orchestrating and pivotal role in driving sustainability along the supply chain (Carmagnac, 2021). We focus on this under-explored level of analysis in the SSCM literature and show how, rather counter-intuitively, the long-standing assumptions of linearity, hierarchy, control and power-led dynamics persist and are even exacerbated by the presence of an MS-MO. Our findings shed light on subtler forms of power, influence and persuasion, based on the discursive construction of responsibility for some actors (i.e., smallholders and consumers) paired with the concealment of others, i.e., the powerful SC actors (Etchanchu & Djelic, 2019).

48Our findings enrich our understanding of the notions of ‘boundaryless responsibility’ (Amaeshi et al., 2008) or ‘extended responsibility’ (Spence & Bourlakis, 2009) for focal firms, which translate into attempts to control the entire supply chain and to define and impose mechanisms to transfer the responsibility for sustainability onto their suppliers (Amaeshi et al., 2008; Formentini & Taticchi, 2016; Grimm et al., 2016). Our results suggest that joining an MS-MO reinforces the control of focal firms over sustainability issues in supply chains, especially in the context where the MS-MO uses market logic and presents a business case to construct the meaning and practices of sustainability.

49Our work also connects to the emergent stream of SSCM literature on voiceless actors in global supply chains (McCarthy et al., 2018). We show that although MS-MOs are meant to empower voiceless actors (smallholders in our case) by providing them with a platform and through consensus-based deliberative processes, these actors are still objectified and remain in a position of limited agency where they have to accept and follow rules imposed by a minority of powerful players (Banerjee, 2018; Carlos & Lewis, 2018). This means that despite the MS-MO discourse of inclusiveness, ‘every consensus exists as a temporary result of a provisional hegemony, as a stabilization’ (Mouffe, 1999, p. 756). MS-MOs tend to reproduce and even exacerbate domination and paternalistic patterns (Etchanchu & Djelic, 2019) and, in so doing, continue to give primacy to the interests and role of focal firms, exerting control over marginalised stakeholders.

Table 8

Promises and solutions of the RSPO in relation to other subject positions

		Smallholders
Promises of the RSPO	Associated solutions/ practices	Discursive evidence
Economic prosperity and access to the market	Cut out the middleman	‘We can now sell directly without relying on a middleman’ (VS_11)
	Funds	‘Now there is a good news for the smallholders who need financial help to get RSPO certification. In line with that commitment to support smallholders, RSPO has created a special fund for them’ (VC_04)
	Price premium	‘As global demand for SPO increases, certified SPO can be sold for a premium price’ (VC_15)
	Higher incomes	‘RSPO certification can support smallholders in… raising levels of income among poor farmers’ (VC_02)
	Economic gains	‘It generates income, it generates employment, and it is how economic growth has been established’ – inter view with a manufacturer
Improvement of smallholder living and working conditions	Land ownership	‘The benefits are that we have doubled our plantation from two to four hectares, and we are rebuilding our house as a permanent structure’ (VS_06)
	University/house	‘From working in these plantations, I can fulfil my family’s daily needs, renovate my house and send my daughter to university in Pekan Baru’ (VS_03)
	Wellbeing	‘Wilmar ensures the wellbeing of workers are taken care of. An example of this is providing a safe hold for workers’ passports’ (VC_15)
	Protect work	‘So a certified mill or a certified plantation should not have child labour or slave labour’ – interview with a manufacturer
	PPE	‘Those of us who have had training understand more and so we always use PPE (Proper use of Personal Protective Equipment)’ (VS_07)
Managerial and environmental training for smallholders	Technical training	‘When independent smallholders get support and coaching… they are able to preserve the environment, and implement sustainable and environmentally friendly practices in their farming methods’ (VS_04)
	Proper use of pesticides/ fer tiliser	‘Since joining Amanah, I apply fertilizer more regularly, and I’m also well informed about the appropriate dosage’ (VS_01)
	Higher yields Methods of work	‘I think that our RSPO members have a higher rate of production’ – interview with a trader ‘After we joined the RSPO, we learned the best practices to manage our plantation.’ (VS_09)
	Control and management systems	‘Ifyou look at the P&C,at the RSPO P&C standards,it is in principle no more than management systems standards’ – interview with a manufacturer
Consumers
Promises of the RSPO	Associated solutions/ practices	Discursive evidence
Increased sustainability awareness	Education for responsible choices	‘Is it good palm oil or is it bad palm oil? Click to find out more’ (VC_05,06,07)
	Empowerment through an app	‘Supporting brands that use sustainable palm oil helps to protect wildlife, people, the environment… Tell your favourite brands and retailers that you only want to see sustainable palm oil in your favourites products’ (VC_18)
	Creating a label	‘I’m using this RSPO trademark app to snap and add RSPO certified products’ (VC_09)
Stimulated SPO demand	Pushing actors towards demand for SPO Right thing to do	‘Make the RSPO certifications more robust for the consumers who are the ones demanding sustainable palm oil’ (VW_39) ‘How we can convince consumers it is only the right thing to do, and that is where NGOs, buyers, producers need to join hands, it is the only way to go forward’ (VW_02)
Supply chain actors
Promises of the RSPO	Associated solutions/ practices	Discursive evidence
Traceability	E-trace tools	‘The RSPO certification is a way to ensure traceability because, otherwise, you have no control over the source’ – interview with a manufacturer
	Four types de cer tification	‘Now we have RSPO mass balance, RSPO Segregated, RSPO Identity preserved, ISCC, so you are almost decommoditising the market, by creating actually different products, we are almost splitting the number of products’ – interview with a trader
Reputation	Addressing risks Alignment with the identity of the company	‘I think many people join RSPO to manage their risks’ – interview with a grower ‘The intentions of the RSPO were aligned with what the company believes’ – interview with a trader

Promises and solutions of the RSPO in relation to other subject positions

RSPO, roundtable on sustainable palm oil; SPO, sustainable palm oil; PPE, Personal Protective Equipment; P&C, principles and criteria; ISCC, International Sustainability and Carbon Certification; NGOs, non-governmental organisations.

Table 9

Main representations of the RSPO subject position

	Visual representation	Usual context	Illustrative quotes
The only practical solution	VC_03 UNEP and RSPO partner to raise profile of Sustainable Palm Oil	VC_03: This video is made to promote the ‘strategic partnership’ between UNEP and RSPO. The video starts by showing a set of images of palm oil trees, products, forests and orang-utans. At one point, the UNEP Executive Director, seen in a suit and tie standing in front of the UNEP flag, explains that the UNEP has decided to sign a ‘memorandum of understanding’ with the RSPO due to its ‘great success stories’. After another sequence of images related to deforestation and climate change, the RSPO Chief Executive Officier (CEO) appears, wearing a suit with an RSPO pin, and affirms that the RSPO has significant experience that can help the UNEP achieve its Sustainable Development Goals (SDGs).	VC_03: ‘One of the great success stories of trying to bring together... It is indeed one of the great success stories with today over 1,000 members’
All-powerful	VC_04 RSPO Smallholder Suppor t Fund VS 02 Smallholders in Thailand	VC_04: This video is a whiteboard animation explaining how smallholders can access RSPO funds. Whilst the narrator explains that the fund is destined to help smallholders obtain the certification (by covering the costs of audits, training and other assessments), the fund is represented as a chest full of gold bars or as a happy smallholder holding gold bars in his hands. This representation conveys the idea that the fund may enrich smallholders.	VC_04: ‘The RSSF can provide support to cover your audit costs. Not only that, RSSF can also be applied for capacity building programs, ..., training on agricultural best practices, improving your market access, and so much more’

Main representations of the RSPO subject position

Revealing the dark side of MOs

50Our research acknowledges the role of discourse in constructing responsibility relationships, and therefore in (re)shaping the meaning and practices of sustainability in global supply chains. We show how the MS-MO is engaging in a particular responsibility attribution discourse (Black, 2008) when constructing the different positions of its members. This is particularly relevant in the context of a voluntary multistakeholder body where there is a ‘democratic’ decision-making process and where responsibility attribution is, at least in theory, collectively negotiated (Banerjee, 2018; Carlos & Lewis, 2018). In contrast with such assumptions and nuancing previous studies portraying MS-MOs as the optimal solution in transitioning towards sustainability (Berkowitz et al., 2017, 2020; Chaudhury et al., 2016), our findings reveal a possible dark side. MS-MOs may be a ‘strategic device used by organisations to manipulate perceptions of their activities and performance’ (Black, 2008, p. 151). Our results suggest that attributing responsibility to specific stakeholders may easily turn into a blame game that could effectively shield the MS-MO from assuming its collective responsibility (Messner, 2009). In our case, the MS-MO’s discursive strategies shift the responsibility for addressing sustainability issues to the actors at the two ends of the chain. Several studies have adopted a discursive approach to investigate blaming strategies and responsibility (Coombs, 2007) and highlight that shifting responsibility onto other actors mainly happens in ambiguous situations where audiences struggle to clearly attribute responsibility due to complexity or uncertainty (Roulet & Pichler, 2020). This is also the case here, where the MS-MO is engaged in coping with several ‘grand challenges’ such as deforestation, climate change and poverty, characterised by their complexity, uncertainty and essentially contested nature (Dentoni et al., 2018; Ferraro et al., 2015). Further research should be done to deepen our understanding of the enablers and determinants of similar discursive strategies aimed at selectively attributing responsibility at the intra-organisational level of MS-MOs. Another distinctive feature of our study is that the palm oil unsustainability problem is constructed from a supply chain perspective. All the actors convening at the ‘round table’ are linked by transactional relationships across the supply chain, which may exacerbate conflicts, tensions and power dynamics amongst them. It would be interesting to see to what extent our results can be confirmed in other settings. Specifically, it would be interesting to discover whether local and regional MS-MOs, characterised by better balanced, established and trust-based relationships amongst actors (Berkowitz et al., 2020), exhibit weaker power dynamics in their discursive strategies.

51We also contribute to the debates around the symbolic rather than substantive nature of corporate sustainability claims (Carlos & Lewis, 2018; Christensen et al., 2013) by showing how the presence of an MS-MO may allow focal companies to further dilute their responsibility for taking action without losing their legitimacy. In fact, their legitimacy seems even stronger and better protected, thanks to their participation in the MS-MO. Shifting responsibility through the MS-MO’s discursive strategies means that large corporations effectively leave the stage, and voiceless actors are placed in the limelight. As the notion of (un)sustainable palm oil is mainly crafted with reference to smallholders and consumers, they suddenly become the ‘new villains’ and are exhorted to become the ‘new heroes’ of sustainability, without being given any actual agency, whilst veiling the role of focal companies (McCar thy et al., 2018). We contend that this shift in the responsibility discourse instrumentally enables focal corporations to maintain their authority and control over the meanings and practices of sustainability in their supply chains and thus nurture their legitimacy. Accordingly, the MS-MO’s actorhood (Grothe-Hammer, 2019) and its presumed collective responsibility for action and commitment (Berkowitz et al., 2020) suffer from the same power-laden dynamics characterising linear and traditional sustainability governance, paving the way for discrepancies between claims and responsibility. Shifting the responsibility onto weaker actors is a comfortable defence strategy for powerful actors whose legitimacy may be threatened by the very enactment of their boundaryless responsibility (Glover & Touboulic, 2020). Our results suggest that when focal firms join an MS-MO, they may have instrumental and opportunistic motivations. In supporting and adhering to a collective entity, focal firms implicitly contribute to recognising the entity’s independent decision-making capabilities and collective ‘actorhood’. This also implies allocating responsibility to that entity because ‘if decisions can be attributed to an organisation as an actor, it can be held responsible for these actions’ (Grothe-Hammer, 2019, p. 326). Powerful actors can then hide behind the responsible collective body, thus reducing their own exposure to blame or contestation.