📈 economics

Common Traits, Different Roles: Structural Position and Nonlinear Dynamics in Three-Party Bargaining

This paper demonstrates that in three-party bargaining, a party's institutional role fundamentally shapes negotiation outcomes and the accessibility of stable settlements by altering how fixed actor traits—such as leverage, moderation, and vulnerability—interact through nonlinear dynamics, often creating multiple long-run paths from the same initial conditions.

Original authors: George Lymperis, Lambros B. Drossos

Published 2026-09-19

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: George Lymperis, Lambros B. Drossos

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Negotiation is often treated as a contest of personalities, where the outcome depends on who is stubborn, who is clever, or who holds the most power. In economics and political science, this focus on individual traits is the standard way to understand why some talks end in peace while others spiral into conflict. However, a deeper layer of reality often gets overlooked: the structure of the meeting itself. Just as a game of chess changes if the pieces are rearranged on the board, even if the players remain the same, the rules and positions in a negotiation can fundamentally alter how information and pressure flow between the parties. This structural arrangement determines whether a small disagreement fades away or grows into a chaotic, unresolvable standoff. Understanding this distinction is vital because it suggests that the path to a deal is not just about the people involved, but about the invisible architecture that guides their interaction.

A new study by George Lymperis and Lambros B. Drossos explores this hidden architecture by asking a simple but profound question: can the same three parties, with the exact same personalities and resources, reach a completely different outcome simply by swapping their roles? To find out, the researchers built a mathematical model of a three-party negotiation involving two main opponents and a third party who steps in only when the conflict reaches a certain point. They assigned each party three fixed characteristics: a capacity to apply pressure, a tendency to moderate or back down, and a degree of vulnerability to the other side's actions. In their simulation, these traits remained constant, but the researchers systematically shuffled which party occupied which structural position—initiator, responder, or reactive third party. They then watched how the negotiation evolved over time, treating the process as a continuous flow of adjustments rather than a series of discrete offers.

The results revealed a striking reality: the structural position of a party matters more than the party's own traits in determining the stability of the outcome. In many cases, the model showed that a stable agreement could exist mathematically, yet the negotiation would fail to reach it. Instead of settling, the parties would get trapped in a cycle of constant, irregular repositioning, bouncing between states without ever finding a resolution. This phenomenon, known as multistability, means that a deal is not just about whether one exists, but whether the specific starting conditions of the negotiation allow the parties to find it. The researchers found that in one specific scenario, simply reassigning the same three sets of traits to different roles increased the likelihood of reaching a settlement from roughly 36 percent to more than 99 percent. This dramatic shift occurred without changing a single personality or resource; only the channel through which those traits were transmitted was altered.

To ensure these findings were not just abstract math, the authors tested their model against real historical data from the Phoenix project, a massive archive of international events. They reconstructed actual three-party conflicts involving government actors and the United Nations, identifying who held the initiating role, who was the responding counterpart, and who acted as the reactive third party. By ranking these real-world actors based on their observed leverage, moderation, and vulnerability, they mapped the historical data onto their theoretical model. The data confirmed that the structural positions observed in history align with the model's predictions, showing that the same actors can face vastly different coordination problems depending on how their roles are assigned. The study also demonstrated that these dynamics are robust; even when the researchers slightly tweaked the rules of how the parties influenced one another, the core finding held: changing the structural position changes the entire trajectory of the negotiation.

The most significant takeaway is that the accessibility of a settlement is a distinct economic object from the existence of a settlement. A stable agreement might be sitting right there, theoretically available, but if the negotiation starts from a slightly different opening position or if the roles are arranged differently, the parties may never stumble upon it. Instead, they might wander into a chaotic path where every attempt to compromise triggers a new round of escalation. This suggests that the design of negotiation institutions—deciding who speaks first, who responds, and when a third party intervenes—is not merely a procedural detail. It is a powerful mechanism that shapes the very possibility of peace. By understanding that the same actors can produce chaos or calm depending solely on their structural placement, mediators and policymakers can see that the key to resolving conflict may lie less in changing the people at the table and more in carefully arranging the table itself.

Technical Summary: Common Traits, Different Roles: Structural Position and Nonlinear Dynamics in Three-Party Bargaining

Problem Statement
The paper addresses a persistent puzzle in negotiation theory: why do identical actors, possessing the same resources and operating under the same institutional rules, sometimes reach a stable settlement and other times remain locked in irregular repositioning or conflict? While existing literature attributes bargaining outcomes to actor heterogeneity (beliefs, preferences, capabilities) or specific institutional rules (proposal rights, network position), this paper posits that structural position itself is a distinct margin of heterogeneity. The central question is whether reassigning fixed actor trait bundles (moderation, leverage, vulnerability) to different structural roles within a three-party bargaining framework alters the feedback dynamics and the accessibility of settlement, even when the actors themselves remain unchanged.

Methodology
The study employs a continuous-time, nonlinear dynamic model derived from a local Nash revision game.

Model Architecture:
- Agents and Traits: Three actors possess fixed trait bundles: self-moderation ( $C_i$ ), leverage ( $B_i$ ), and vulnerability ( $Y_i$ ).
- Structural Roles: The model defines three distinct structural positions:
  - Role 1 (Initiator): Responds directly to the counterpart, moderated by its own traits, and exposed to the third party via a vulnerability channel.
  - Role 2 (Coercive Counterpart): Responds directly to Role 1, transmitting leverage through a channel dependent on the joint configuration of Roles 1 and 3.
  - Role 3 (Structurally Reactive Third Party): Activated conditionally by the joint configuration of the principals; once active, it feeds back into both principals.
- Dynamics: Actors choose next stances to maximize a payoff function that includes a cost for abrupt movement. As revision opportunities become frequent, the discrete revision game converges to a continuous-time system of differential equations:
  $\dot{x}_1 = p x_2 - A x_1 - Y_1 x_1 x_3$
  $\dot{x}_2 = q x_1 - D x_2 - B_2 x_1 x_3$
  $\dot{x}_3 = B_3 x_1 x_2 - E x_3$
  Here, $x_i$ represents the stance relative to a reservation benchmark. The coefficients $A, D, E$ represent effective dissipation (moderation + vulnerability), while $p, q$ represent direct response strengths.
Analytical Reduction:
- The authors perform an exact rescaling of state and time variables. This reduces the system's twelve primitive coefficients to four dimensionless groups ( $\delta_1, \delta_2, \rho, \chi$ ) that govern the vector-field geometry.
- A key analytical finding is that Role-3 leverage ( $B_3$ ) does not appear in these four geometry groups; it only rescales the principal coordinates. In contrast, Role-1 and Role-2 leverage directly alter the feedback geometry.
Numerical and Empirical Analysis:
- Simulation: The authors simulate the system across various parameter sets (benchmarks) and role permutations. They classify long-run behaviors (stable settlement, periodic cycles, chaos) and calculate settlement accessibility: the probability that a random opening configuration reaches a stable settlement within a finite horizon.
- Empirical Reconstruction: Using the Cline Center Historical Phoenix Event Data, the authors reconstruct structural roles in historical three-party conflicts (specifically government dyads and a composite UNO third party). They derive ordinal rankings for leverage, moderation, and vulnerability based on event codes (CAMEO) and map these to the model's trait scale. External validation is provided using International Crisis Behavior (ICB) and Near Crisis datasets.

Key Results

Multistability and Accessibility:
The model demonstrates multistability: a stable settlement can coexist with sustained irregular (chaotic) or periodic paths. Crucially, the existence of a stable equilibrium does not guarantee its accessibility. In benchmark scenarios, reassigning the same trait bundles to different roles can change the share of plausible opening positions that lead to settlement from roughly 36% to over 99%. This establishes "settlement accessibility" as a distinct economic object dependent on structural assignment, not just the existence of an equilibrium.
Structural Asymmetry in Leverage:
The dimensionless reduction reveals a sharp asymmetry: Leverage held by the two principal roles (Roles 1 and 2) alters the reduced feedback geometry (affecting $\rho$ and $\chi$ ), whereas leverage held by the third party (Role 3) only rescales the state space without changing the reduced vector field. Thus, structural position determines whether a trait acts as a regime-forming mechanism or a scale-forming mechanism.
Role Permutation Effects:
Holding actor traits fixed and permuting their structural assignments generates diverse long-run architectures (settlement, chaos, periodicity). The "role effect" survives local perturbations to the nonlinear channels, indicating that the institutional assignment of traits is a robust driver of dynamics.
Empirical Findings:
The Phoenix data reconstruction confirms that principal roles (government actors) and the third-party role (UNO) exhibit distinct trait distributions. Principals show higher leverage-to-vulnerability ratios compared to the third party. The mapping of these ordinal signals to the model supports the theoretical mechanism where fixed traits yield different outcomes based on structural transmission.

Significance and Claims
The paper claims to offer a mechanism for coordination through structure. Unlike heterogeneous-agent models where changing agents changes the interaction, this model shows that actors can remain fixed while the positions through which their traits are transmitted change the feedback itself.

Theoretical Contribution: It formalizes how institutional design (proposal rights, network position, third-party activation) alters the "basin geometry" of bargaining. It demonstrates that the same set of actor characteristics can lead to vastly different coordination problems solely based on how those characteristics are structurally assigned.
Economic Implication: The existence of a stable bargain is insufficient to predict negotiation success. The "accessibility" of that bargain—how easily it can be reached from uncertain opening positions—is a function of the structural configuration.
Modesty of Claims: The authors explicitly state that the empirical layer provides ordinal rankings and structural reconstruction rather than a full causal test of specific historical outcomes. The model establishes the structural mechanism and its consequences on the declared state and time scales, leaving the explicit estimation of contextual transmission parameters as a natural next step.

In summary, the paper argues that structural position is not merely a container for actor traits but an active determinant of bargaining dynamics, capable of transforming fixed actor characteristics into either stable settlements or persistent instability.

Technical Summary: Common Traits, Different Roles: Structural Position and Nonlinear Dynamics in Three-Party Bargaining

More like this