Dual-Reward Mechanism in Cooperative Systems

• Dual-reward mechanism is a system that integrates two interlinked reward processes to simultaneously incentivize giving and receiving behaviors in strategic settings.
• It employs mathematical models and spatial game theory to balance costs and benefits, ensuring optimal cooperation through calibrated reward-to-cost ratios.
• Research shows that moderate rewards, supported by network structures, effectively sustain cyclic dominance and prevent free-rider proliferation in cooperative environments.

A dual-reward mechanism is a structural approach in agent-based systems, game-theoretic environments, and multi-agent learning frameworks that combines two complementary reward-related processes or sources—most often to simultaneously incentivize distinct or even opposing behaviors, balance costs and benefits, or sustain complex strategic dynamics. In the context of research, "dual-reward" usually refers to the interplay between giving and receiving rewards (as in social dilemmas), the aggregation of reward and truth-telling incentives in peer evaluation, or the explicit modeling of multiple reward streams in agent optimization. This entry surveys canonical dual-reward mechanisms through mathematical formalization, key theoretical results, typology of behaviors, and empirical relevance.

1. Mathematical Structure of Dual-Reward Systems

Dual-reward mechanisms implement two interlinked forms of reward assignment within a strategic or learning environment. In the spatial public goods game (1010.5771), agents can adopt one of three strategies:

• $C$: pure cooperator,
• $D$: defector,
• $RC$: rewarding cooperator, who not only cooperates but also pays a cost to reward others.

The formal payoff structure for an agent participating in $k+1$ overlapping groups, with composition given by $N_C$ (pure cooperators), $N_D$ (defectors), $N_{RC}$ (rewarding cooperators), is:

• Cooperator:

$$P_{\rm C} = \frac{r(N_{\rm C} + N_{\rm RC} + 1)}{k+1} - 1 + \beta \frac{N_{\rm RC}}{k}$$

• Defector:

$P_{\rm D} = \frac{r(N_{\rm C} + N_{\rm RC})}{k+1}$

• Rewarding Cooperator:

$$P_{\rm RC} = P_{\rm C} - \gamma \frac{N_{\rm C} + N_{\rm RC}}{k}$$

Here, $r$ is the cooperation synergy factor, $\beta$ the per-reward received, and $\gamma$ the per-reward cost.

This structure directly implements a dual-reward mechanism:

• The rewarding cooperator receives extra utility for each neighbor it rewards ($\beta$, benefit) but pays a cost per recipient ($\gamma$, cost).
• As a result, the net payoff captures both the incentive provided to others (via costly rewarding) and the possibility of being rewarded oneself.

2. Comparison of Rewards and Punishments

Within the same spatial public goods game, a core question is whether reward or punishment is more effective at promoting cooperation. The central outcomes are:

• Rewards foster cooperation more effectively than punishments when the synergy factor $r$ is low. Under such conditions, the marginal effect of additional cooperation is small, and punitive mechanisms may not cover their own cost.
• Rewarding is structurally less self-sustaining because, unlike punishers (who need not pay when defectors are absent), rewarding cooperators incur costs as long as there are any cooperators present—even when defectors have disappeared. This means the parameter $\beta/\gamma$ must be high for rewards to sustain cooperation at the same level as punishments.
• The empirical phase diagrams show regions where reward mechanisms can stabilize cooperation in parameter regimes where punishment or standard cooperation would fail.

3. Non-monotonicity and the Optimality of Moderate Rewards

Contrary to naively increasing rewards for maximizing cooperation, the system displays a non-monotonic relationship:

• Moderate levels of reward ($\beta$) lead to maximal cooperation.
• Very high rewards allow second-order free-riders (ordinary cooperators who benefit from rewards without bearing the cost) to proliferate, which enables defectors to re-enter and persist within the population.

This dynamic emerges via cyclic dominance among the three strategies:

• Defectors exploit cooperators,
• Rewarding cooperators outcompete defectors,
• Cooperators (Cs) outcompete rewarding cooperators (RCs) by avoiding the cost of rewarding while still benefiting, which creates openings for defectors.
• The coexistence and cycling of these strategies is an intrinsic feature of the dual-reward setup.

4. Cyclic Dominance and Strategy Biodiversity

Cyclic dominance structures, akin to rock-paper-scissors, are a haLLMark of systems where multiple mechanisms of reward are operating:

$D \to C \to RC \to D$

Each strategy has a region in state space where it is dominant, yet is itself vulnerable to one of the others. In the spatial public goods game:

• $D$ dominates $C$
• $RC$ dominates $D$
• $C$ dominates $RC$

This property ensures persistent heterogeneity:

• No single strategy eradicates the others.
• The resulting spatial structures show complicated domain interfaces, leading to long-term coexistence and strategic diversity.

Importantly, spatial structure is critical; in well-mixed (non-spatial) populations, these cyclic dynamics quickly collapse.

5. Parameter Regimes for Effective Cooperation

Efficient dual-reward mechanisms require:

• Low synergy factor $r$: When cooperative returns are low, rewards provide the necessary boost.
• High $\beta/\gamma$ ratio: The per capita benefit of rewarding must greatly exceed the cost.
• Moderate rewards: Avoid “overpaying,” which ultimately undermines system-level cooperation.
• Spatially structured networks: Enable clustering and promote local reinforcement, vital for maintaining cyclic dominance and preventing the extinction of rewarding cooperators.

6. Implications for Mechanism Design

The findings highlight several critical considerations for the design and application of dual-reward mechanisms in cooperative systems:

• Overly generous rewards can be counterproductive; calibrating incentive structures for moderate rather than maximal direct payoffs avoids proliferation of free-riders and maintains strategic diversity.
• Policy design should account for spatial or network structures rather than relying solely on mean-field or well-mixed assumptions.
• The dual nature of reward (via both benefit to recipients and cost to givers) is essential for understanding real-world cooperative dilemmas in social, biological, and economic networks.

Summary Table: Payoff Equations and Dual-Reward Structure

Strategy	Payoff Equation	Reward Structure
Pure Cooperator (C)	$$P_C = \dfrac{r(N_C + N_{RC} + 1)}{k+1} - 1 + \beta \dfrac{N_{RC}}{k}$$	Receives rewards
Defector (D)	$P_D = \dfrac{r(N_C + N_{RC})}{k+1}$	Receives no reward, pays no cost
Rewarding Cooperator (RC)	$P_{RC} = P_C - \gamma \frac{N_C + N_{RC}}{k}$	Pays cost to reward, receives benefit

References to Key Results

• Mathematical model: Section "Model—Payoff Functions."
• Phase diagrams and cyclic dominance: Results and Discussion.
• Non-monotonicity and moderate reward effectiveness: Phase diagram figures and associated analysis.
• Synergy factor and $\beta/\gamma$ constraints: Analytical and simulation results sections.

---

In sum, dual-reward mechanisms in the spatial public goods game reveal that the intricate balance between reward giving and receiving, the calibration of incentive cost-benefit ratios, and the emergent cyclic dominance among strategies collectively determine whether cooperation can survive, to what degree, and in what form. The dual nature of these mechanisms shapes boundaries of sustainability and efficiency for cooperation in complex, spatially structured societies.