In the global pursuit of carbon neutrality and sustainable energy transitions, the new energy vehicle industry has emerged as a pivotal force in reshaping the transportation sector. As a global leader in this domain, China confronts a critical challenge: the large-scale retirement of electric vehicle batteries. The imperative for efficient, transparent, and scalable recycling systems has never been more urgent. My research addresses this nexus by investigating the intricate decision-making processes within the electric vehicle battery recycling ecosystem. I focus on two critical leverage points: the design and efficacy of government reward-punishment policies, and the transformative potential of blockchain technology. This dissertation constructs sophisticated game-theoretic models to analyze strategic interactions, aiming to provide a robust theoretical foundation for policy formulation and corporate strategy. By comprehensively comparing static and dynamic policy mechanisms, and by evaluating the impact of blockchain across four distinct recycling models, I seek to uncover pathways towards a more resilient and efficient recycling infrastructure for electric vehicle batteries.
1. Introduction and Research Significance
The exponential growth of the new energy vehicle market has precipitated a corresponding surge in the volume of end-of-life electric vehicle batteries. Projections indicate that by 2030, the cumulative weight of retired batteries in China alone will surpass 3 million tons. This impending “retirement wave” presents a dual-edged challenge. On one hand, improper disposal of these batteries poses severe environmental risks, as the complex chemical compositions—including heavy metals like cobalt, nickel, and manganese, and electrolytes—can leach into soil and groundwater, causing persistent ecological damage that undermines the very environmental benefits of electric vehicles. On the other hand, these retired batteries represent a treasure trove of critical resources. Lithium, cobalt, and nickel are essential for the global energy transition, and their supply chains are often concentrated and vulnerable. Efficient recycling of electric vehicle batteries is therefore not merely an environmental imperative but also a strategic necessity for resource security and economic resilience, unlocking a projected multi-billion-dollar market.
Recognizing the significance of this issue, the Chinese government has established a robust policy framework centered on the extended producer responsibility principle, mandating that vehicle manufacturers shoulder the primary responsibility for battery recycling. Despite these top-down efforts, the implementation of these policies encounters deep-seated structural obstacles. Informal recycling channels, operating with minimal environmental compliance costs, often outbid formal recyclers, leading to a “race to the bottom.” Simultaneously, a fragmented information infrastructure across the battery lifecycle—from production to recycling—creates significant information asymmetries and high trust costs, hindering efficient valuation, traceability, and collaboration among stakeholders. These challenges underscore a fundamental tension: policy intent has not yet fully translated into effective, market-driven action. The core research question thus centers on how to design more effective policy instruments and leverage technological innovation to overcome these structural barriers.
My research addresses these pressing challenges through a comprehensive analytical framework. It examines two interrelated levers for optimizing the electric vehicle battery recycling system. The first lever is policy design, specifically the comparative efficacy of static versus dynamic government reward-punishment mechanisms in guiding the behavior of manufacturers, consumers, and the government itself. The second lever is technological enablement, specifically investigating the differential impacts of blockchain technology on the performance of four dominant recycling models. By pursuing these twin objectives, I aim to offer actionable insights for both policymakers seeking to construct precise regulatory systems and for enterprises navigating strategic decisions regarding recycling partnerships and technology investments. This research seeks to advance the field toward a more efficient, credible, and sustainable recycling ecosystem for electric vehicle batteries.
2. Theoretical Foundations and Methodological Approach
To systematically investigate the dynamics of electric vehicle battery recycling, my research is anchored in several established theoretical frameworks. The concept of the closed-loop supply chain provides the foundational lens for understanding the system as an integrated network that encompasses both forward logistics (production, distribution, sales) and reverse logistics (collection, inspection, reprocessing). This perspective is crucial for modeling how the flow of materials and value from batteries can be cycled back into the production system, thereby reducing reliance on virgin resources and minimizing environmental impact. The analysis begins with recognizing the core structural elements of a closed-loop supply chain for electric vehicle batteries, connecting the manufacturers, automotive assemblers, and consumers in a circular economic model.
Game theory, and its various branches, constitutes the central methodological toolkit for my work. The evolution of complex, multi-stakeholder systems is captured through evolutionary game theory, which relaxes the assumption of perfect rationality and models agents as boundedly rational learners who adjust their strategies based on observed outcomes. The goal is to identify evolutionarily stable strategies—states that are resilient to invasion by alternative strategies. This approach is particularly apt for analyzing how social norms and corporate behaviors regarding recycling can emerge and stabilize over time under different policy interventions. For instance, the decision of a manufacturer to actively recycle or a consumer to use a formal recycling channel can be modeled as a repeated game where payoff structures, influenced by government policies, dictate long-term behavioral equilibria.
Furthermore, the strategic interplay between supply chain partners with differing market powers is analyzed using Stackelberg game theory. This is a leader-follower model where the dominant firm (e.g., the power battery manufacturer) moves first, setting its decisions, and the follower (e.g., the vehicle manufacturer) reacts optimally. This framework is instrumental for modeling the hierarchical decision-making structure in the closed-loop supply chain, allowing me to solve for optimal pricing, recycling rates, and profitability under various recycling models. The objective is to explore how these strategic interactions lead to different economic and environmental outcomes. The models are solved using backward induction, and the resulting analytical solutions are then subjected to extensive numerical simulations to explore the impact of key parameters and validate the theoretical findings.
3. Design and Analysis of Reward-Punishment Mechanisms in Electric Vehicle Battery Recycling
3.1 Model Framework and Hypotheses
I begin my analysis by constructing a tripartite evolutionary game model involving the government, new energy vehicle manufacturers (hereafter “manufacturers”), and consumers to explore the dynamics of electric vehicle battery recycling regulation. The central proposition is to compare and contrast the efficacy of two distinct policy regimes: a static mechanism, where rewards and punishments are fixed, and a dynamic mechanism, where these instruments adjust based on the prevailing behaviors of the actors. The players, their strategies, and the key parameters are defined as follows. The manufacturer can choose to adopt an “active recycling” strategy, investing in infrastructure and technology, or a “passive recycling” strategy, shirking its responsibilities. Consumers, as the holders of retired batteries, choose between “formal recycling” channels and “informal recycling” channels. The government, tasked with safeguarding public welfare, decides between “active supervision” and “passive supervision.” Several key assumptions underpin the model: the cost of active recycling for a manufacturer exceeds the intangible benefits gained, such as enhanced brand image.
The following table summarizes the key parameters used to construct the payoff matrix for the static model. These parameters allow me to quantify the economic and social costs and benefits associated with each strategy combination.
| Parameter Definitions for the Tripartite Game | |
|---|---|
| $C_m$ | Total cost for the manufacturer to actively recycle. |
| $R_m$ | Government reward for active recycling. |
| $F_m$ | Government penalty for passive recycling. |
| $I_m$ | Intangible benefits for active recycling (e.g., brand image). |
| $I_1^c, I_2^c$ | Consumer income from formal and informal recycling channels, respectively. |
| $C_1^c, C_2^c$ | Consumer transaction costs for formal and informal recycling channels, respectively. |
| $\mu$ | Consumer environmental preference coefficient. |
| $\beta$ | The coefficient of consumer profit loss from informal channels due to government supervision. |
| $C_g$ | Government cost of active supervision. |
| $\alpha_1, \alpha_2$ | Recycling technology rates for active and passive recycling, respectively. |
3.2 Static Reward-Punishment Mechanism
Under the static mechanism, the government offers a fixed reward $R_m$ and imposes a fixed penalty $F_m$. Through the construction of the payoff matrix for the game (as detailed in Table 3.2 of the original thesis), I derive the replicator dynamic equations, which describe how the proportion of a population adopting a given strategy evolves over time. For instance, the replicator dynamics for the manufacturer ($x$ being the proportion choosing ‘active recycling’) is given by:
$$ F(x) = x(1-x) [z R_m + z \beta F_m + y(\alpha_1 I_{mc} – \alpha_2 I_{mc}) + C_m – I_m – C_{ti} – C_r – C_i] $$
Replicator dynamics for consumers ($y$) and the government ($z$) are similarly derived. By analyzing the Jacobian matrix of this dynamic system, I identify the evolutionary stable strategies under various parameter conditions. The analysis reveals a critical threshold for the government’s penalty $F_m$.
Key Finding 1: The penalty has an effective threshold. When the penalty $F_m$ is below this threshold, the system is more likely to converge to an unfavorable equilibrium where the manufacturer is passive, consumers use informal channels, and the government is passive. Only when the penalty is sufficiently high does it effectively deter passive recycling and nudge the system towards the desired equilibrium of (active recycling, formal channel, active supervision). Rewards, in contrast, primarily influence the speed of convergence but do not change the final stable state, and excessive rewards can paradoxically slow down the evolutionary process due to diminishing returns.
3.3 Dynamic Reward-Punishment Mechanism
In contrast to a static policy, I then design a dynamic mechanism where the government adjusts its policy tools in response to the state of the system. The specific formulations are proposed as: the penalty $\bar{F}_m = v/z$ is inversely proportional to the government’s supervision probability z, meaning if the government is less often active, it imposes a higher penalty to maintain deterrence. Conversely, the reward $\bar{R}_m = R_m \cdot x$ is directly proportional to the manufacturer’s recycling probability $x$, meaning the higher the compliance rate, the greater the reward. The replicator dynamics are then recalculated under these dynamic functions, and the stability of the equilibrium points is analyzed.
Key Finding 2: The dynamic mechanism exhibits superior performance. The numerical simulations clearly demonstrate that the system converges to the ideal equilibrium point with significantly faster speed and greater stability under the dynamic mechanism compared to the static one. The convergence time to reach the desired state where both the manufacturer and consumers adopt positive recycling strategies is reduced by approximately 60% under the dynamic regime, and the evolutionary paths are smoother, with reduced oscillations, signifying greater predictability and robustness. This suggests that adaptive policy design is more effective in guiding multi-agent systems toward socially optimal outcomes.
The following comparative table (based on the simulation in Section 3.5) illustrates this core result:
| Mechanism | Convergence Time | Path Stability | Final Equilibrium |
|---|---|---|---|
| Static Reward-Punishment | ~12.5 time units | Moderate, with oscillations | Active, Formal, Passive (Ideal) |
| Dynamic Reward-Punishment | ~5 time units | High, smooth path | Active, Formal, Passive (Ideal) |
This first part of my research concludes that the “dynamic” nature of policy is not merely an administrative detail but a fundamental determinant of its effectiveness in shaping the complex behaviors within the electric vehicle battery recycling ecosystem. The success of the dynamic mechanism lies in its ability to create a self-correcting feedback loop that aligns the incentives of the government, manufacturers, and consumers more efficiently.
4. Benchmark Analysis: Recycling Model Optimization without Blockchain Empowerment
4.1 Description of the Four Recycling Models
Having established the critical role of policy design in motivating stakeholders, I proceed to investigate the operational efficiencies of different recycling models. This chapter establishes a benchmark framework by constructing a three-tier closed-loop supply chain model comprised of a power battery manufacturer, a new energy vehicle manufacturer, and consumers, operating without blockchain technology. The supply chain engages in a Stackelberg game, where the battery manufacturer is the leader, and the vehicle manufacturer is the follower. The demand for new energy vehicles is given by $D = \phi(1+g) – \theta p$, while the volume of recycled batteries depends on the recycling price $r_j$ and is formulated as $Q_r = a + b r_j$. Within this baseline, I model and analyze four prominent recycling structures to understand their inherent differences in performance. The four models are:
- Model $N_B$ (Battery Manufacturer Recycling): The battery manufacturer is solely responsible for establishing the recycling network and collecting used electric vehicle batteries directly from consumers.
- Model $N_V$ (Vehicle Manufacturer Recycling): The vehicle manufacturer leverages its existing dealership network to collect batteries and transfers them to the battery manufacturer.
- Model $N_{BV}$ (Hybrid Recycling): Both the battery manufacturer and the vehicle manufacturer independently operate their own competing recycling channels, simultaneously setting their own recycling prices.
- Model $N_{(B+V)}$ (Alliance Recycling): Both manufacturers form a strategic alliance, integrating their resources to establish a unified collection system. They act as a single entity to set a common recycling price and maximize the combined profit.
4.2 Game-Theoretic Solutions and Key Findings
For each model, I employ backward induction to derive the optimal wholesale price $w$, retail price $p$, and recycling price $r_j$. The profit functions for the battery manufacturer ($\pi_B$) and vehicle manufacturer ($\pi_V$) are formulated based on the specific cost and revenue structures of each model. The main analytical results pertaining to optimal pricing and performance are summarized as follows:
Proposition 1: The wholesale price of batteries remains the same across all four models. However, the retail price in the alliance model ($p_{(B+V)}$) is lower than in the other three models, because the alliance internalizes the double marginalization effect, aligning the pricing incentives of both the manufacturer and the retailer.
Proposition 2: The performance of the recycling models is differentiated. The equilibrium recycling price, collection volume, recycling rate, and overall supply chain profits are calculated for each model. The results of the baseline analysis are presented in the table below, using a representative parameter set. The alliance model maximizes total supply chain profit, while the hybrid model is most effective in maximizing the recycling rate of electric vehicle batteries.
| Benchmark Results without Blockchain Empowerment | ||||
|---|---|---|---|---|
| Recycling Model | Supply Chain Profit (CNY) | Sales Volume (units) | Recycling Volume | Recycling Rate (%) |
| $N_B$ (Battery Manuf.) | 4,933,474,747 | 49,700 | 18,228 | 36.676% |
| $N_V$ (Vehicle Manuf.) | 4,859,473,348 | 49,700 | 9,150 | 18.410% |
| $N_{BV}$ (Hybrid) | 5,000,859,104 | 49,700 | 19,225 | 38.682% |
| $N_{(B+V)}$ (Alliance) | 6,540,679,996 | 99,400 | 20,050 | 20.171% |
This baseline analysis confirms that in a conventional, non-digital environment, the choice of recycling model involves a trade-off. The alliance model, by eliminating duplicate investments and channel conflicts, maximizes economic efficiency and overall profitability. In contrast, the hybrid model harnesses the power of market competition between channels to drive a higher collection rate, even though it may result in lower overall profit compared to the alliance due to competitive inefficiencies.
5. Optimization and Simulation of Electric Vehicle Battery Recycling with Blockchain Empowerment
5.1 Model Extension with Blockchain Technology
To overcome the information asymmetries and trust deficits identified in the benchmark models, this chapter introduces blockchain technology as a core component of the recycling decision framework. In the presence of blockchain, the market demand function is expanded to $D = \phi(1+g) – \theta p + k \lambda$. Here, $\lambda$ represents the level of blockchain adoption (or technological effort), and $k$ is a critical parameter capturing the degree of consumer trust in blockchain technology. This parameter is pivotal, as it quantifies how effectively the transparency and security features of blockchain translate into increased consumer willingness to pay for a vehicle equipped with a traceable battery and their willingness to participate in formal recycling. The technology also directly enhances the recycling volume, which is now formulated as $Q_r = a + b r_j + k \lambda$. The total cost of implementing blockchain is modeled as a quadratic function $C(\lambda) = \frac{1}{2}A\lambda^2$, which captures the increasing marginal cost of higher levels of technological sophistication. This cost is shared between the battery manufacturer, with a proportion $t$, and the vehicle manufacturer, with the remainder $(1-t)$. The model is solved for all four recycling models under this new technological context, with the goal of identifying the optimal pricing, technological investment, and profitability.
To facilitate the subsequent analysis, the key new parameters are defined in the following table:
| Additional Parameters for Blockchain Model | |
|---|---|
| $A$ | Blockchain cost investment coefficient. |
| $\lambda$ | Level of blockchain technology investment. |
| $k$ | Consumer trust coefficient for blockchain technology. |
| $t$ | Cost-sharing proportion borne by the battery manufacturer. |
5.2 Comparative Analysis: The Differential Impact of Blockchain
By solving the new models for each of the four recycling modes, I obtain the optimal strategies. I then conduct a comprehensive comparative analysis to evaluate the impact of blockchain on key performance indicators. The central results are summarized below.
Proposition 3: The introduction of blockchain increases both wholesale and retail prices of electric vehicles. Due to enhanced consumer trust, the willingness to pay for a vehicle with a verifiable battery lifecycle increases, allowing manufacturers to set higher prices without negatively impacting demand. Concurrently, blockchain reduces the optimal recycling prices. This counterintuitive outcome arises because the information transparency provided by blockchain alleviates consumers’ risk perceptions about unfair valuation, making them more willing to recycle at lower price points.
Proposition 4: There is a strong positive correlation between the level of consumer trust ($k$) and the optimal level of blockchain investment ($\lambda$). When consumers place higher trust in the technology, the economic returns on technology investment are amplified through higher sales and improved recycling efficiency. This enhanced profitability incentivizes firms to invest more heavily in blockchain infrastructure, creating a virtuous cycle.
The following tables present the comparative results for the four models with and without blockchain. The first table shows the performance metrics for the “with blockchain” scenarios.
| Performance Results with Blockchain Empowerment | ||||
|---|---|---|---|---|
| Recycling Model | Supply Chain Profit (CNY) | Sales Volume (units) | Recycling Volume | Recycling Rate (%) |
| $Y_B$ (Battery Manuf.) | 4,998,974,780 | 50,260 | 19,367 | 38.533% |
| $Y_V$ (Vehicle Manuf.) | 4,916,297,388 | 50,158 | 9,608 | 19.155% |
| $Y_{BV}$ (Hybrid) | 5,108,024,171 | 50,456 | 21,906 | 43.415% |
| $Y_{(B+V)}$ (Alliance) | 6,684,186,668 | 100,068 | 23,037 | 23.022% |
The relative changes in profit and recycling rate are visually apparent in the comparison tables from the original thesis (see Table 5.3). It is evident that blockchain technology drives improvements in both economic and environmental performance across all models. However, the magnitude of improvement varies significantly. The alliance model $Y_{(B+V)}$ realizes the largest absolute increase in profit. The hybrid model $Y_{BV}$, which already had the highest recycling rate, sees the most substantial percentage improvement in recycling rate. This divergent impact can be explained by the technology’s effect on the underlying inefficiency each model addresses. For the single-channel models, blockchain’s impact is limited as they still suffer from limited channel reach. For the hybrid model, blockchain enhances performance by mitigating information-driven channel conflicts and reducing valuation disputes, allowing the competitive dynamics to more effectively drive up volume. For the alliance model, which is already coordinated, blockchain primarily reduces internal coordination and verification costs, further unlocking its already considerable economic potential.
5.3 Sensitivity Analysis and the Pivotal Role of Consumer Trust
To provide more robust managerial insights, I conduct extensive sensitivity analyses on the equilibrium solutions. I systematically vary key parameters including the blockchain cost-sharing ratio $t$, the consumer retail price sensitivity $\theta$, the consumer recycling price sensitivity $b$, and, most importantly, the consumer trust coefficient $k$. The findings are summarized as follows:
Finding 1: The Profit-Enhancing Effect of Blockchain is Universal. The analysis reveals that for any feasible cost-sharing ratio $t$ between the battery and vehicle manufacturers, the profits of both parties are always higher when blockchain technology is adopted. This implies a strong economic rationale for firms to voluntarily cooperate in blockchain investments, as the resulting efficiency gains outweigh the implementation costs for all participants.
Finding 2: The Criticality of Consumer Trust. The sensitivity analysis on $k$, the consumer trust in blockchain, highlights it as the key mediating variable that unlocks the full value of the technology. As $k$ increases from a low to a high level, a range of positive outcomes are observed. The optimal blockchain investment $\lambda$ increases, indicating a greater willingness of firms to invest when they anticipate higher consumer acceptance. The retail price p increases, demonstrating the technology’s ability to command a premium. Crucially, the recycling rate $R$ also improves, even when the recycling price remains unchanged. This last point is particularly insightful: it reveals that consumer trust creates an intrinsic motivation to recycle through formal channels, decoupling participation from pure price incentives. The following table illustrates these dynamics:
| Impact of Consumer Trust Level ($k$) on Key Metrics | ||||
|---|---|---|---|---|
| Consumer Trust (k) | Blockchain Investment ($\lambda^*$) | Retail Price ($p^*$) for Y_B | Recycling Rate (%) for Y_B | Profit for Y_B (CNY) |
| k = 2 | 255.9 | 175,444 | 37.135% | 4,947,733,297 |
| k = 4 | 522.6 | 176,237 | 38.533% | 4,988,277,985 |
| k = 6 | 812.7 | 177,640 | 40.953% | 5,083,716,123 |
This analysis crystallizes the understanding that blockchain technology itself is not a panacea. Its economic and environmental benefits are contingent upon the existence of a trusting market environment. This underscores that technological deployment must be accompanied by consumer education initiatives and transparency campaigns to effectively cultivate the trust necessary to reap the full rewards of blockchain in the electric vehicle battery recycling ecosystem.
6. Conclusion and Managerial Implications
My research provides a comprehensive and nuanced analysis of the pathways to optimize electric vehicle battery recycling. The findings offer substantial theoretical and practical insights, which are synthesized into several key conclusions and actionable recommendations for both policymakers and business leaders.
6.1 Research Conclusions and Main Findings
The study yields two prominent sets of conclusions. First, concerning policy design, the findings underscore the superiority of dynamic reward-punishment mechanisms over static ones. I found that while government penalty strictness is a critical determinant of manufacturer compliance, the dynamic policy’s ability to create a self-correcting feedback loop leads to faster, more stable convergence to desirable system states. Furthermore, structural factors, including the manufacturer’s recycling technology level and consumer environmental preferences, serve as vital intrinsic drivers of the system’s positive evolution. Second, concerning technology empowerment, my analysis confirms that blockchain technology has a systemic and positive impact on the entire closed-loop supply chain. It enables a mutually beneficial scenario where retail prices and overall profits can increase while simultaneously lowering the marginal recycling price, demonstrating a solution to the information asymmetry problem that has long plagued the industry. Across the recycling models, the research indicates that while the alliance recycling model is best positioned to maximize total supply chain profit, the hybrid recycling model is more effective in maximizing the overall collection rate. Importantly, the degree of consumer trust in blockchain is shown to be a critical and potent moderator of the technology’s economic and environmental benefits.
6.2 Managerial and Policy Implications
For Policymakers: The implications are clear. Governments should transition away from static, one-size-fits-all regulatory frameworks and toward smart, adaptive regulatory systems that adjust reward and penalty intensities in real-time based on compliance data. Policy should also focus on setting effective deterrence thresholds and promoting measures that enhance the technological capabilities of manufacturers and the environmental awareness of consumers. Moreover, governments have a critical role to play in fostering the digital trust infrastructure by promoting industry standards, encouraging the development of shared blockchain platforms, and supporting consumer education about the benefits of formal recycling. The combined application of dynamic policy and technical infrastructure is essential for creating a self-sustaining and efficient recycling ecosystem.
For Enterprises: The research offers clear strategic guidance. Companies should make their choice of recycling model a function of their strategic priorities. If the primary goal is to maximize overall profitability, the alliance model is a superior strategy. If the goal is to dominate in resource recovery and achieve high collection rates, then a hybrid competitive model may be more appropriate. For companies, blockchain technology should be viewed not just as a cost center, but as a critical strategic asset. Firms should actively invest in blockchain and establish fair cost-sharing partnerships to build a foundation of transparency and trust that will ultimately strengthen their supply chains, enhance their brand reputation with increasingly environmentally conscious consumers, and secure access to valuable recycled materials.
6.3 Avenues for Future Research
While this dissertation offers significant contributions, it is not without limitations. Future research could extend the current framework by incorporating the constraints and opportunities presented by environmental regulations such as carbon taxes and carbon trading, and their interaction with the economic incentives modeled here. Additionally, integrating behavioral factors into the game models, such as fairness concerns or bounded rationality, could improve the realism of the simulation results. From a supply chain design perspective, the inclusion of additional actors, such as third-party recyclers and echelon utilization companies, would lead to a more representative model of the increasingly complex industrial ecosystem. By addressing these areas, future research can continue to build upon the foundations established here, guiding the electric vehicle battery recycling industry toward a future that is both environmentally sustainable and economically prosperous.