Guest editorial: Agent-based models of intra-organizational dynamics

Today, the use of triangulation of several methodological approaches is crucial for the understanding of organizational processes. Among these methodologies, computational models offer the opportunity to jointly explore the intricate nuances of individual-level heterogeneity, the communication structures enabling interaction and the dynamics of the overall organizational system. This special issue marks an important step toward studying several processes of internal dynamics in organizations from an agent-based computational perspective. The contributing authors delve into various issues concerning intra-organizational dynamics, offering a comprehensive exploration that promises to enrich the current understanding of organizational dynamics.While computational models have a long history in organizational research (Cyert and March, 1963; Harrison et al., 2007; Davis et al., 2007), these methods (and formal methods generally) have traditionally been secondary methodological alternatives for organizational researchers (Adner et al., 2009). This is despite the organizational research community's consensus to view organizations as complex adaptive systems (Wall, 2016). These systems feature non-trivial arrangements of tasks, resources, knowledge, distributed cognition (Carley, 2002; Secchi and Neumann, 2016; Secchi, 2021), incentives and communication structures (Siggelkow and Rivkin, 2005; Wall, 2017). Such a vision highlights the importance of portraying organizations as adaptive structures of purposeful agents that dynamically shape the collective's fate (Fioretti, 2013; Wall, 2017).Agent-based computational approaches remain a peripheral method for studying organizations. This is partly because many published management models do not sufficiently emphasize a culture of scientific practices like code sharing, replication, calibration and using model-based insights for organizational interventions. This is unfortunate, given the potential of combining rich data and models for better decision-making in today's organizations (Puranam, 2021). Furthermore, a great deal of organizational models has traditionally focused on variance approaches (Estévez-Mujica and García-Díaz, 2021). This is due to an easier alignment with statistical methods, but it comes at the expense of missing opportunities to explore intertemporal, multilevel explanatory mechanisms (Miller, 2015).This special issue, focusing on agent-based modeling and intra-organizational dynamics, provides a forum showcasing novel contributions of agent-based computational modeling for understanding and intervening in organizations' internal processes. We hope this special issue not only encourages the development of computational-based theories using traditional canonical representations (e.g., NK models) but also inspires researchers to extend these and build representations that capture the particularities and contingencies of today's organizations, with an eye toward designing interventions. Ultimately, as a strategy to improve organizations' functioning, we advocate for enriching organizational perspectives by using specific cases to inspire and build computational-based theories and specific interventions to test such theories (Oliva, 2019).This issue features five articles that showcase the richness of studying intra-organizational dynamics. The key contribution of these articles lies in their use of agent-based models to unveil explanatory mechanisms that are either intertemporal, multilevel or structurally related. These articles focus on varied phenomena, including the emergence of discriminatory behavior, the link between modularity and conformity, the implications of hierarchical groups on the advent of cooperation, the association between cultural looseness and organizational innovation and the application of agent-based models for team membership and design.Yasar (2025) explores how discriminatory behavior emerges in agent interactions. This occurs due to different individual-level trait categorization strategies, even when no pre-existing differences are present. The author uses an asymmetric hawk-dove game and an evolutionary algorithm that allows agents to adopt each other's categorization strategies. These strategies vary in their “coarseness,” which reflects how much agents weigh the relevance of others' trait types. The results show that fine-categorization strategies are more advantageous when the cost of fighting is low. Conversely, coarse-grained categorization proves more beneficial when the cost of fighting is considerable.Khodzhimatov et al. (2025) take the perspective of organizations as complex systems to explore the interplay between a system's degree of decomposability and an agent's propensity to conform to others' decisions. The authors find that conformity generally hinders performance across various organizational structures, specifically functional, divisional and geographical structures. The only exception is when departments are engaged in highly similar tasks; in that case, conformity's negative impact is mitigated.Sandretto and Merlone (2025) investigate whether the results of Axelrod's iterative prisoner's dilemma tournament (Axelrod and Hamilton, 1984) apply to situations with varying interaction frequency and structure. These extensions have implications for understanding cooperative behavior in organizations. To explore this, the authors use a triadic supervision structure, aiming to represent hierarchical group interactions in organizations. While they confirm that the number of interactions can affect cooperative behavior, they also conclude that the effectiveness of cooperative strategies can be maintained in hierarchical groups.In the context of collaboration networks, Jung and Gulden (2025) explore the implications of indirect minority influence – the influence of workers with different domain expertise. They examine whether this diverse domain expertise is a crucial factor that potentiates cultural looseness (i.e. cultural tolerance) as a driver of organizational innovations. In their model, cultural tolerance – typically linked to lenient social norms and low sanctions – is represented as an organization's capacity to allow coworkers with diverse expertise to interact. This stands in contrast to what might be perceived as cultural tightness. As agents interact, their degree of agreement (opinion synchronization) is captured by a consistency parameter. Jung and Gulden find that not only is indirect minority influence essential for innovation, but also that, as cultural tightness increases, consistency must likewise increase to sustain organizational adaptation.Lastly, Etemadidavan and Collins (2025) frame their work in the context of the performance implications of different team designs. They simulate design teams to explore overall organizational project wage costs, specifically in relation to organizing schemes like multiple team membership (MTM) and disjoint multi-team systems (MTS). They find that adopting MTM significantly impacts wage costs.The articles in this special issue offer strong evidence for applying agent-based computational modeling to understand internal organizational dynamics. Computational methods are a valuable tool that not only complement traditional statistical methods in organizational studies but also prove highly useful to understand non-trivial aggregations to macro-level properties. These non-trivial aggregations combine member-level heterogeneity, various structural scenarios and an emphasis on evolution.In sum, agent-based modeling offers a powerful way to rethink how we represent the internal evolution of cognitive distinctions and various forms of agent-level interaction within organizations, among other features. This approach allows the study of organizational dynamics to move beyond just the nuances of information exchange and knowledge appropriation between individuals. It lets us use computational thinking to reconsider, for example, the implications of (possibly evolving) asymmetric power in decision-making and the alignment to others' decisions in shaping macro-level properties. We hope this special issue has successfully shown the importance of agent-based modeling for studying these dynamics.