**5. Discussion: Relations between the Different Dimensions of Power in GISCO**

After describing the respective forms of power at different levels, this subsection focuses on the interdependent relationships between these levels. At the network level, as described by Glückler and Bathelt, Mossig, as well as Bathelt and Taylor, formal and informal institutions are the basis of cooperation [28,29,31,37] (p. 93). These are mutually recognized and reproduced by the members. Relevant in terms of power are the voting relationships that emerge from the bylaws. Thereby, all four stakeholder groups A to D in GISCO have veto rights. Institutionally, this ensures that no group can outvote the others. In other words, a balance has been created and individual stakeholders' power over others has been ruled out. Also formally established is the (so far unused) possibility of excluding members, which, however, as a last resort does not guarantee stable cooperation. The actors' manners of dealing with each other have established themselves as an informal institution in the course of cooperation. Other informal institutions in GISCO include solving conflicts through personal discussions and moderation, as well as the actors' mutual expectations of professional and fair dealings with each other inside and outside GISCO. The order of the process steps in which content is negotiated has also become such an institution in the network that is recognized by all participants without being written into the bylaws. Those values and norms have a disciplining effect on the individual actors and thus reduce uncertainties [29] (p. 93). Figure 3 shows the schematic process in GISCO already known from Figure 2, but supplemented by the dominant levels of power on each step. *Sustainability* **2022**, *14*, x FOR PEER REVIEW 12 of 17 actors' mutual expectations of professional and fair dealings with each other inside and outside GISCO. The order of the process steps in which content is negotiated has also become such an institution in the network that is recognized by all participants without being written into the bylaws. Those values and norms have a disciplining effect on the individual actors and thus reduce uncertainties [29] (p. 93). Figure 3 shows the schematic process in GISCO already known from Figure 2, but supplemented by the dominant levels of power on each step.

As shown in Figure 3, processes in GISCO usually start in the executive board. The members of the executive board decide whether a topic should be discussed further. If so, it is delegated to one of the workings groups (WGs) for more in-depth discussion. Within WG 1: Communication, WG 2: PRO-PLANTEURS and WG 3: Sustainability, the influ-

sonality, perceived expertise and persistence in GISCO can influence the debate, using corresponding strategies of influence. This behavior is consistent with Saffer et al.'s conceptualization of power, in which power in MSIs is associated with influence over communication [27]. This, in turn, is consistent with the "power to" approach of influence through relationships and one's own persuasiveness described in prior theoretical work [29] (pp. 92–93). However, the person-based power decreases in favor of the

In the next step of the processes in GISCO, the four stakeholder groups A to D discuss the decision papers negotiated before in the thematic working groups. Contrary to what might be expected from theory, organization-based power is primarily relevant within these stakeholder groups. In the executive board, the groups act (more or less) cohesively as interest groups. This can be seen, first, in the fact that the interbranch organizations of industry (stakeholder group B) and retail (C) exercise a coordinating function within their stakeholder groups and speak for their groups at the network level. Secondly, it is clear from the fact that stakeholder groups sometimes reject compromises reached in the WGs. Third, in the further steps of the process, negotiation loops take place between the stakeholder groups whose representatives on the executive board conduct negotiations. Consequently, in the context of MSIs we should speak of organization-based power *within* the stakeholder groups and group-based organizational power *between* them, which becomes more relevant than person-based power in the further steps of the negotiation processes. Accordingly, membership of a stakeholder group is important for the assertion of common interests. Although the division of the 75 member organizations (as of 2019) into four stakeholder groups—(A) Federal Govern-

**Figure 3.** Processes and power relations in GISCO. **Figure 3.** Processes and power relations in GISCO.

stakeholder groups with each step of the process.

As shown in Figure 3, processes in GISCO usually start in the executive board. The members of the executive board decide whether a topic should be discussed further. If so, it is delegated to one of the workings groups (WGs) for more in-depth discussion. Within WG 1: Communication, WG 2: PRO-PLANTEURS and WG 3: Sustainability, the influence of individuals is high. The interviews show that individuals with appropriate personality, perceived expertise and persistence in GISCO can influence the debate, using corresponding strategies of influence. This behavior is consistent with Saffer et al.'s conceptualization of power, in which power in MSIs is associated with influence over communication [27]. This, in turn, is consistent with the "power to" approach of influence through relationships and one's own persuasiveness described in prior theoretical work [29] (pp. 92–93). However, the person-based power decreases in favor of the stakeholder groups with each step of the process.

In the next step of the processes in GISCO, the four stakeholder groups A to D discuss the decision papers negotiated before in the thematic working groups. Contrary to what might be expected from theory, organization-based power is primarily relevant within these stakeholder groups. In the executive board, the groups act (more or less) cohesively as interest groups. This can be seen, first, in the fact that the interbranch organizations of industry (stakeholder group B) and retail (C) exercise a coordinating function within their stakeholder groups and speak for their groups at the network level. Secondly, it is clear from the fact that stakeholder groups sometimes reject compromises reached in the WGs. Third, in the further steps of the process, negotiation loops take place between the stakeholder groups whose representatives on the executive board conduct negotiations. Consequently, in the context of MSIs we should speak of organization-based power *within* the stakeholder groups and group-based organizational power *between* them, which becomes more relevant than person-based power in the further steps of the negotiation processes. Accordingly, membership of a stakeholder group is important for the assertion of common interests. Although the division of the 75 member organizations (as of 2019) into four stakeholder groups—(A) Federal Government, (B) Cocoa, Chocolate and Confectionery Industry, (C) Retail Grocery Trade and (D) Civil Society—structurally reduces the heterogeneity in GISCO to a certain extent, the groups' common interests must first be negotiated.

As shown in the results, all stakeholder groups have specific forms of power, which come to play in the executive board. Empirical evidence shows that, due to the structure of GISCO with its four stakeholder groups, these specific forms of power are less likely to be used by individual organizations at the level of the overall network, but rather are first negotiated by the respective stakeholder groups themselves. Industrial and retail companies use group-based organizational power in a form described in the literature on GPNs as corporate power [20,22]. In GISCO, this is reflected in the perception of corporate representatives that it is their responsibility to intensively scrutinize proposals from other groups in terms of their feasibility. Another way of exerting influence is that no measures can be implemented without their ideational and financial support. Therefore, a materially conditional form of influence, coupled with the use of the veto, is characteristic for companies. This effect can be assigned to the "power over" approach, while the explanation of internal processes and practical experiences to other stakeholder groups corresponds to the argumentative "power to" approach. Civil Society has a specific groupbased organizational power that corresponds to collective power [20,22]. It is virtually the "counterpole" (B6 2020: para. 107) of the industrial group. The Civil Society group is described as active and demanding and appears comparatively cohesive, despite its diverse composition. The interviews show that the opportunities for civil societies to exert influence lie primarily in their power of persuasion, their arguments and demands, and the mobilization of others' resources for their concerns. This effect is in line with the "power to" approach. Nevertheless, this stakeholder group also has (immaterial) resources in the sense of "power over": a threatened withdrawal from GISCO would deprive the initiative of legitimacy to a large extent, and some NGOs use their legitimacy as a resource in public campaigns to exert pressure on companies. Furthermore, policymakers have

their own group-based organizational power, equivalent to institutional power [20,22]. The ministries not only assume a kind of patronage for GISCO, but also work on the content of the initiative. Financial support for policy corresponds to the effect of a resource-based "power over" approach to influence. This includes the possibility of opting out of GISCO or becoming legislatively active if necessary. One of the ministries is monitoring the possibility for a legislative regulation. At the time of the interviews, it appears that the ministries want to encourage companies to become more involved without a regulation, which again is in line with the "power to" approach.

Table 3 summarizes the effects of the empirically observed forms of group-based organizational power in GISCO. Here, the representatives are understood as representing their respective organizations, transferring the power of their organizations in global economic contexts in the form of actions in the context of an MSI [34,35].


**Table 3.** Empirically observable effects of group-based organizational power in GISCO.

The diversity of the effects of group-based organizational power in GISCO means that no stakeholder groups are clearly more powerful in relation to others. With reference to Bathelt and Taylor, it can be seen that the power symmetry between the four stakeholder groups does not lead to quick and efficient decisions [31]. This often results in stalemate situations. The following quotation from B5 (2020: para. 35) illustrates a typical agreement process: "Of course, there is a lot of discussion about such small formulations, which sometimes makes it a bit fuzzy. But that's because, based on the positions that the members have, there can't be as much room for maneuver as one might give the other person in normal life". The lack of necessity and willingness to leave the respective positions leads to situations in which compromises are only reached on a voluntary basis. The quotation from interviewee G2 in Figure 3 shows the tough struggle for compromises: "And we often find consensual solutions, and those are always a compromise. It is very rarely the case that a proposal is made, and all parties agree to it" (G2 2020: para. 87).

It can be concluded that the diversity of power forms leads to a dysfunctional power constellation in GISCO in relation to its goals. Moreover, the group-based organizational power of corporations is institutionally advantaged through their veto power. In comparison, the demanding role of civil societies has been elaborated, which are presumably less dependent on veto power than corporations to achieve their goals. This disparity is not overcome in the negotiation processes. The interviews suggest that within the stakeholder groups, precisely because of the use of organization-based power by individual organizations, it is possible, in most cases, to work out a common position. Negotiations at the network level between the stakeholder groups, on the other hand, are often more difficult because there is no clear power hierarchy between the actors. By their very nature, consensus reached between stakeholder groups in this way is often the lowest common denominator. With regard to the MSI's goals regarding the sustainability of the cocoa sector, it should be noted that these have (so far) not been achieved. One interviewee exemplifies: "[ . . . ] the problems in the cocoa sector have been known for over ten years and the problems haven't really changed that much or even improved. So I think the results are still not that earth-shattering" (B2 2020: para. 117). From the interviews it emerges that the opportunities of GISCO lie rather in generating attention, bringing diverse stakeholders into contact with each other and enabling them to learn with and from each other. Improvements regarding the socio-economic living and working conditions of farmers are achieved by GISCO less directly, but much more indirectly or on a smaller scale through the project "PRO-PLANTEURS". In terms of the global cocoa and chocolate sector, interviewees emphasize the steadily increasing amount of certified cocoa processed in Germany by GISCO members. However, many interviewees state that certification as an instrument alone is not sufficient to solve the "wicked problems" [1] in the food supply chain of cocoa in a sufficient way.
