**1. Introduction**

Complex decisions with major social, environmental and economic consequences are often made by groups consisting of multiple organizations rather than by individuals alone (Cur¸seu and Schruijer 2017). In multiparty systems stakeholders explore their interdependencies and use their knowledge and expertise in order to integrate and develop their different perspectives and interests (Vansina et al. 1998; Schruijer and Vansina 2008; Schruijer 2016). The decision making in such systems is characterized by high complexity, given the complex nature of the problem domain, the various stakeholder parties involved and the diversities in interests, identities, perspectives and power positions. Relational dynamics in such multiparty collaborative systems shape the decision outcomes (Cur¸seu and Schruijer 2017; Schruijer 2008). Participants interact within their own stakeholder party to discuss their party's goals, aspirations and interests, while simultaneously they interact with the other stakeholders that have their own, often differing goals, aspirations and interests, so as to arrive at and realize a joint goal. Through this collective goal stakeholders can jointly address a problem which they cannot solve on their own, while through working towards the joint goal, stakeholders can serve their intra-organizational goals (Schruijer 2008).

In such multi-party systems, composed of several stakeholder parties, relational dynamics unfold at two levels: within stakeholder parties as well as between stakeholder parties. These two relational

dimensions are not independent as the relational dynamics that arise from the stakeholder parties may influence the larger system as a whole (bottom-up influences) while system-level relational dynamics may spiral down from the larger system to influence the different stakeholder parties (top-down influences). So far, the literature on multi-party systems lacks systematic investigations of these jointly operating influences and it does not explore how task conflict and relational conflict experienced within the stakeholder parties have an impact on the relational dynamics of the system as a whole. Likewise, there is no direct empirical evidence on how changes in conflict and collaboration in the whole system influence the dynamics of task and relationship conflict within the participating stakeholder parties. We set out to explore the interplay of these bottom-up and top-down influences in multiparty systems, using a behavioural simulation.

In line with interdependence theory (Holmes 2002), when people foresee they will engage in social interactions, they build expectations about: (1) the nature of the situation and (2) the goals, motives and behaviours of the ones they will interact with. The expectations about others and the social situation will eventually shape behaviour in a variety of social contexts (Holmes 2002). In other words, in social situations, expectations and social behaviour are entwined.

In field theory and interdependence terms (Lewin 1936; Holmes 2002) two interdependence fields are generated in multiparty systems that explain the relational dynamics within as well as between parties. In the small parties (first interdependence field) people will build expectations about their own party (social situation) and their teammates (interaction partners), while in the larger multiparty system people will build expectations about the interests of other stakeholder parties and the general climate of the multiparty system as a whole. It is our contention that the expectation—behaviour entwinements at the two levels (party and multiparty system) are interdependent. More specifically our paper sets out to explore the extent to which the association between expectations and experienced social interactions within parties influences the expectation-behaviour entwinement in the larger system (bottom up influence) and the extent to which the entwinement between expectations and behaviours in the larger system is tied to the expectation-behaviour association within parties (top-down influence).

We consider the two forms of social interdependence described by Deutsch (1949), namely positive (as illustrated by collaborative intentions and behaviour) and negative (as illustrated by conflictual intentions and behaviours) interdependence. It follows that the entwinement between expectations for positive interdependence and realized collaboration within the stakeholder parties impacts the entwinement between expected collaborativeness and real collaborativeness in the larger multiparty system. Moreover, we also expect that the entwinement between expectations of negative interdependence and experienced (relationship) conflict within stakeholder parties is dependent on the same type of interdependence entwinement at the multiparty system level. We set out to explore the way in which the entwinement of expectations-experiences for positive and negative interdependence is transferred from the parties to the multiparty level (bottom-up influence) and the way in which the expectations-experiences entwinement for positive and negative interdependence at the multiparty level trickles down to impact the parties in the system (top-down influence). This interdependence dynamics and the bottom up and top down influences is depicted in Figure 1.

As the interplay between collaboration and conflict is essential for decision quality in multiparty systems (Cur¸seu and Schruijer 2017), our study has the potential to make several contributions to the literature. First, we answer the call for dynamic models of the interplay between emergen<sup>t</sup> states in teams and multiparty systems (Costa et al. 2017; Shuffler et al. 2015; Waller et al. 2016) and we use a cross-lagged design to capture changes in collaborativeness and conflictuality. Second, we use a realistic behavioural simulation to tap into the complex dynamics of conflict and collaboration in multiparty systems and to explore the positive role of task conflict for collaboration in multiparty systems. Using a round robin method to evaluate collaborativeness and conflict, we capture both the top-down as well as the bottom-up interplay between collaboration and conflict in multiparty systems.

**Figure 1.** Bottom-up and top-down influences in multiparty systems. Note: T1 = time 1 (expectations before the interaction starts), T2 = time 2 (experienced dynamics at the party and multiparty level).

### **2. Theory and Hypotheses**

Literature on multiparty collaboration is diverse and lacks integration (Vangen 2017). Theories of multiparty collaboration range from practice oriented theories of collaboration that theorize collaborative relations as ways in which stakeholders can achieve *collaborative advantage* (Huxham 2003; Vangen 2017) to more systemic approaches that distinguish between inputs, processes and outputs of collaborative relationships (Thomson and Perry 2006; Wood and Gray 1991). Multiparty systems are social systems composed of interdependent stakeholder parties that interact to discover ways of moving forward while dealing with existing contradictions (Vangen 2017). Collaborative relationships are inherently paradoxical as the stakeholders have to work with their similarities and their differences simultaneously in order to achieve collaborative advantage (Huxham 2003; Vangen 2017). In terms of organization, multiparty systems are formed of individuals, nested in stakeholder or interest parties that are ultimately nested in a larger system that strives for collaboration. Various interdependencies (both positive and negative) that exist among these systemic components are conceptualized as inputs for collaborative relationships (Wood and Gray 1991). These interdependencies are however not necessarily subject to formal or centralized control and members of multiparty systems have autonomy to discover viable ways of dealing with the paradoxes of collaboration (Thomson and Perry 2006). It is our contention that the paradoxes inherent to multiparty collaboration (Huxham 2003; Vangen 2017) reflect two co-existing forms of social interdependence. We further build on insights from Social Interdependence Theory (Deutsch 1949) to understand how interactions within the stakeholder parties (like conflict) influence the functioning of the whole system (bottom-up influences) while the interdependencies among the parties will shape the functioning of the individual stakeholder parties (top-down influences).

Social interdependence theory (Deutsch 1949; Johnson 2003) describes positive and negative forms of interdependence in social systems. Positive interdependence is reflected in so-called promotive or collaborative interactions in which an aggregate within the system achieves its goal only to the extent to which the other aggregates achieve their goals as well. The negative interdependence is reflected in contrient or conflictual interactions in which aggregates only achieve their goals at the expense of other aggregates in the system. In line with social interdependence theory we argue that collaborative and conflictual interactions in the system as a whole have an impact on the interactions that unfold within the aggregates.

### *2.1. Summary and Definitions*

In our analysis, we focus on the entwinement between expectations and experienced interactions and the way in which this entwinement transfers in the system in a bottom-up versus top-down manner. We define *positive interdependence entwinement* as the association between expectations of positive interdependence and experienced collaboration (changes from expectations to what is really experienced after the interactions start) and *negative interdependence entwinement* as the association between expected negative interdependence and experienced conflict (changes from expected to experienced conflict). In what follows we will build on social interdependence theory to further specify the bottom-up and top-down influences in multiparty collaboration. Further on, we define *bottom-up influences* as the ways in which positive and negative entwinement transfers from stakeholder parties to the system as a whole. Finally, we define *top-down influences* as the ways in which the entwinement of positive and negative interdependence transfers from the large multiparty system to the stakeholder parties.
