In recent years, the global shift towards sustainable development and circular economy principles has elevated the importance of remanufacturing in the manufacturing sector. Remanufacturing reduces raw material consumption, minimizes waste generation, and extends product lifespan, offering both economic and environmental benefits. Electric vehicles, as key enablers of low-carbon transformation, are central to this transition. However, consumer skepticism regarding the quality and safety of remanufactured electric vehicles poses significant barriers to market adoption. Cognitive biases and varying willingness to pay (WTP) for new, remanufactured, and used electric vehicles critically influence demand structures and pricing strategies within the supply chain. This study focuses on the electric vehicle industry in China, examining remanufacturing supply chain decisions under differentiated consumer WTP. By constructing demand functions for new products, remanufactured products, and used products, I analyze optimal pricing under centralized decision-making, manufacturer-led, and retailer-led power structures. Through comparative analysis and numerical simulations, I explore the relationships between WTP difference coefficients and profits, highlighting the impact of consumer preferences on supply chain efficiency and coordination mechanisms such as two-part tariff contracts.
The rapid growth of the China EV market underscores the urgency of addressing remanufacturing challenges. Consumers often perceive remanufactured electric vehicles as inferior due to concerns about battery life, performance, and reliability, leading to divergent WTP across product categories. This heterogeneity not only affects demand but also amplifies pricing complexities under different supply chain power configurations. Prior research has largely overlooked the integration of consumer WTP differences into remanufacturing decision frameworks for electric vehicles. This paper aims to fill that gap by developing a model that incorporates WTP variation coefficients, enabling a comprehensive analysis of multi-agent game equilibria. The findings provide theoretical insights and practical guidance for enhancing the sustainability of the electric vehicle remanufacturing industry in China and beyond.
Problem Description and Basic Assumptions
I consider a simplified remanufacturing supply chain model for electric vehicles, involving a single manufacturer, a single retailer, and consumers. The manufacturer produces new electric vehicles while also engaging in the collection and remanufacturing of used vehicles. The retailer purchases these products and sells them at differentiated prices. Consumers, based on their perceptions and information assessments, determine their WTP for new, remanufactured, and used electric vehicles. This leads to distinct demand models for each product category, which are used to derive profit functions for supply chain participants. To streamline the model, I define key symbols and their meanings in Table 1.
| Variable Symbol | Symbol Meaning | Variable Symbol | Symbol Meaning |
|---|---|---|---|
| \(C_{ON}/C_{OR}/C_{OS}\) | Cost of new/remanufactured/used product | \(W_{ON}/W_{OR}/W_{OS}\) | Wholesale price of new/remanufactured/used product |
| \(P_{ON}/P_{OR}/P_{OS}\) | Retail price of new/remanufactured/used product | \(q_{ON}/q_{OR}/q_{OS}\) | Market demand for new/remanufactured/used product |
| \(B\) | Unit recycling cost paid by manufacturer | \(\theta\) | Consumer WTP for new product |
| \(\alpha_1\) | Relative cognitive value coefficient between new and remanufactured products | \(\beta\) | Profit allocation price offered by retailer to manufacturer |
| \(\alpha_2\) | Relative cognitive value coefficient between remanufactured and used products | Superscripts M/R/MR/MT | Denote retailer-led/manufacturer-led/centralized decision/contract coordination models |
| \(\pi_M/\pi_R/\pi_I\) | Manufacturer/retailer/supply chain profit function |
For model simplification, I define the following expressions:
$$ \epsilon_1 = \frac{1}{6} + \frac{37}{36} + \frac{C_{OR}^2}{36\alpha_1^2} – \frac{2C_{OR}}{36\alpha_1} – \frac{C_{OR}}{6\alpha_1}, $$
$$ \epsilon_2 = \frac{-B – C_{OS} + \sqrt{B^2 + 2BC_{OS} + C_{OS}^2 + 24\alpha_1^2}}{4\alpha_1}, $$
$$ \epsilon_3 = \frac{-B + C_{OR} – C_{OS} + 3\alpha_1 + \sqrt{(B – C_{OR})^2 + 2(C_{OS} – 3\alpha_1)(B – C_{OR}) + (3\alpha_1 – C_{OS})^2 + 16\alpha_1^2}}{8\alpha_1}. $$
I base the model on several key assumptions. First, remanufactured products are typically sold at lower prices than new ones; for instance, remanufactured electric vehicles may be priced about 20% lower than new models, while used products have even lower values. Thus, the costs satisfy \(C_{ON} > C_{OR} > C_{OS}\), reflecting the cost advantages of remanufacturing. Second, the manufacturer incurs an additional refurbishment effort and recycling cost \(B\) for producing used products, with \(0 < B < 1\). Third, I focus on a single-period scenario where the market size is normalized to 1, and each consumer purchases at most one electric vehicle. Consumer WTP for a unit product is denoted by \(\theta\), which follows a uniform distribution on \([0,1]\), with distribution function \(F(\theta) = \theta\). Consumers can discern quality differences between new, remanufactured, and used electric vehicles, leading to differentiated WTP: \(\theta_{ON} > \theta_{OR} > \theta_{OS}\), where \(\theta_{OR} = \alpha_1 \theta_{ON}\) and \(\theta_{OS} = \alpha_1 \alpha_2 \theta_{ON}\), with \(\alpha_1, \alpha_2 \in (0,1)\). The utility functions for the three products are defined as:
$$ U_{ON} = \theta_{ON} – P_{ON}, $$
$$ U_{OR} = \theta_{OR} – P_{OR} = \alpha_1 \theta_{ON} – P_{OR}, $$
$$ U_{OS} = \theta_{OS} – P_{OS} = \alpha_1 \alpha_2 \theta_{ON} – P_{OS}. $$
Fourth, the parameter \(\alpha_1\) represents the relative cognitive value between new and remanufactured products, while \(\alpha_2\) represents the relative cognitive value between remanufactured and used products. Higher values of \(\alpha\) indicate stronger WTP and a greater inclination to purchase the respective product. Finally, as refurbished used products are protected by patents, the manufacturer has the autonomy to decide whether to engage in refurbishment activities for used electric vehicles.
Model Construction and Analysis
Centralized Decision Model
In the centralized decision model, the manufacturer and retailer operate as an integrated entity aiming to maximize total supply chain profit. When all three products coexist in the electric vehicle market, consumers choose based on utility comparisons: they purchase a new product if \(U_{ON} \geq 0\), \(U_{ON} \geq U_{OR}\), and \(U_{ON} \geq U_{OS}\); a remanufactured product if \(U_{OR} \geq 0\), \(U_{ON} \leq U_{OR}\), and \(U_{OS} \leq U_{OR}\); and a used product if \(U_{OS} \geq 0\), \(U_{ON} \leq U_{OS}\), and \(U_{OR} \leq U_{OS}\). The demand functions are derived as follows:
$$ q_{ON} = \int_{\frac{P_{ON} – P_{OR}}{1 – \alpha_1}}^1 f(\theta_{ON}) d\theta_{ON} = 1 – \frac{P_{ON} – P_{OR}}{1 – \alpha_1}, $$
$$ q_{OR} = \int_{\frac{P_{OR} – P_{OS}}{\alpha_1(1 – \alpha_2)}}^{\frac{P_{ON} – P_{OR}}{1 – \alpha_1}} f(\theta_{ON}) d\theta_{ON} = \frac{P_{ON} – P_{OR}}{1 – \alpha_1} – \frac{P_{OR} – P_{OS}}{\alpha_1(1 – \alpha_2)}, $$
$$ q_{OS} = \int_{\frac{P_{OS}}{\alpha_1 \alpha_2}}^{\frac{P_{OR} – P_{OS}}{\alpha_1(1 – \alpha_2)}} f(\theta_{ON}) d\theta_{ON} = \frac{P_{OR} – P_{OS}}{\alpha_1(1 – \alpha_2)} – \frac{P_{OS}}{\alpha_1 \alpha_2}. $$
The total supply chain profit, accounting for the manufacturer’s recycling cost \(B\), is:
$$ \pi_I = q_{ON}(P_{ON} – C_{ON}) + q_{OR}(P_{OR} – C_{OR}) + q_{OS}(P_{OS} – C_{OS} – B). $$
Substituting the demand functions:
$$ \pi_I = \left(1 – \frac{P_{ON} – P_{OR}}{1 – \alpha_1}\right)(P_{ON} – C_{ON}) + \left(\frac{P_{ON} – P_{OR}}{1 – \alpha_1} – \frac{P_{OR} – P_{OS}}{\alpha_1(1 – \alpha_2)}\right)(P_{OR} – C_{OR}) + \left(\frac{P_{OR} – P_{OS}}{\alpha_1(1 – \alpha_2)} – \frac{P_{OS}}{\alpha_1 \alpha_2}\right)(P_{OS} – C_{OS} – B). $$
Proposition 1: In the centralized decision model, the optimal prices for new, remanufactured, and used electric vehicles are:
$$ P_{ON}^{MR*} = \frac{(1 + C_{ON})(1 – \alpha_2) + 2\alpha_1 \alpha_2 – \alpha_1 \alpha_2^2}{2 – 2\alpha_2}, $$
$$ P_{OR}^{MR*} = \frac{C_{OR} + \alpha_1 + \alpha_1 \alpha_2 – C_{OR} \alpha_2 – \alpha_1 \alpha_2^2}{2 – 2\alpha_2}, $$
$$ P_{OS}^{MR*} = \frac{B + C_{OS} + \alpha_1 \alpha_2}{2}. $$
Furthermore, the demand sensitivities satisfy \(\frac{dq_{ON}^j}{d\alpha_2} < 0\), \(\frac{dq_{OR}^j}{d\alpha_1} > 0\), and \(\frac{dq_{OS}^j}{d\alpha_1} > 0\) for any model \(j\).
Proof: The Hessian matrix of \(\pi_I\) is:
$$ H = \begin{pmatrix}
\frac{-2}{1 – \alpha_1} & \frac{2}{1 – \alpha_1} & 0 \\
\frac{2}{1 – \alpha_1} & -2\left(\frac{1}{1 – \alpha_1} + \frac{1}{\alpha_1(1 – \alpha_2)}\right) & \frac{2(1 – \alpha_2)}{\alpha_1} \\
0 & \frac{2(1 – \alpha_2)}{\alpha_1} & -2\left(\frac{1}{\alpha_1 \alpha_2} + \frac{1 – \alpha_2}{\alpha_1}\right)
\end{pmatrix}. $$
The sequential principal minors are \(|H_1| = -\frac{2}{1 – \alpha_1} < 0\), \(|H_2| = \frac{4(1 – \alpha_2)}{(1 – \alpha_1)\alpha_1} > 0\), and \(|H_3| = -\frac{8\alpha_2}{(1 – \alpha_1)^2 \alpha_1^2 (1 – \alpha_2)} < 0\), confirming that \(\pi_I\) is strictly concave and has a unique optimal solution. Solving the first-order conditions \(\frac{\partial \pi_I}{\partial P_{ON}} = 0\), \(\frac{\partial \pi_I}{\partial P_{OR}} = 0\), and \(\frac{\partial \pi_I}{\partial P_{OS}} = 0\) yields the optimal prices. Substituting these into the demand functions gives the sensitivity results.
This proposition indicates that in centralized decision-making, higher consumer WTP for remanufactured and used electric vehicles increases their demand, while negatively impacting new product demand. This suggests that firms should adjust pricing strategies dynamically based on WTP trends and emphasize value-added services for new products to stimulate demand.
Manufacturer-Led Stackelberg Game Model (Model M)
In this model, the manufacturer acts as the leader by setting wholesale prices \(W_{ON}^M, W_{OR}^M, W_{OS}^M\), and the retailer follows by determining retail prices \(P_{ON}^M, P_{OR}^M, P_{OS}^M\). The profit functions are:
$$ \pi_M = q_{ON}(W_{ON} – C_{ON}) + q_{OR}(W_{OR} – C_{OR}) + q_{OS}(W_{OS} – C_{OS} – B), $$
$$ \pi_R = q_{ON}(P_{ON} – W_{ON}) + q_{OR}(P_{OR} – W_{OR}) + q_{OS}(P_{OS} – W_{OS}). $$
Proposition 2: In the manufacturer-led model, the optimal wholesale and retail prices for electric vehicles are:
$$ W_{ON}^{M*} = -\frac{-2\alpha_1 – \alpha_2 – C_{ON} \alpha_2 + 2\alpha_1 \alpha_2^2}{2\alpha_2}, $$
$$ W_{OR}^{M*} = -\frac{-2\alpha_1 – C_{OR} \alpha_2 – \alpha_1 \alpha_2 + 2\alpha_1 \alpha_2^2}{2\alpha_2}, $$
$$ W_{OS}^{M*} = -\frac{-2\alpha_1 – B \alpha_2 – C_{OS} \alpha_2 + \alpha_1 \alpha_2^2}{2\alpha_2}, $$
$$ P_{ON}^{M*} = \frac{(3 + C_{ON}) \alpha_2 – 2\alpha_1 (-1 + \alpha_2^2)}{4\alpha_2}, $$
$$ P_{OR}^{M*} = \frac{C_{OR} \alpha_2 + \alpha_1 (2 + 3\alpha_2 – 2\alpha_2^2)}{4\alpha_2}, $$
$$ P_{OS}^{M*} = \frac{C_{OR} \alpha_2 + \alpha_1 (2 + 3\alpha_2 – 2\alpha_2^2)}{4\alpha_2}. $$
Proof: Using backward induction, the retailer’s profit function \(\pi_R\) is concave, as verified by its Hessian matrix. Solving \(\frac{\partial \pi_R}{\partial P_{ON}} = 0\), \(\frac{\partial \pi_R}{\partial P_{OR}} = 0\), and \(\frac{\partial \pi_R}{\partial P_{OS}} = 0\) gives the retail prices in terms of wholesale prices. Substituting these into the manufacturer’s profit function and solving \(\frac{\partial \pi_M}{\partial W_{ON}} = 0\), \(\frac{\partial \pi_M}{\partial W_{OR}} = 0\), and \(\frac{\partial \pi_M}{\partial W_{OS}} = 0\) yields the optimal wholesale prices. The resulting profits are derived accordingly.
In this model, consumer WTP does not influence wholesale prices, as the manufacturer bases decisions on costs and market strategy. Retailers should adjust retail prices flexibly to balance profit and sales volume, while manufacturers should share production information to enhance supply chain coordination.
Retailer-Led Stackelberg Game Model (Model R)
Here, the retailer leads by setting retail prices \(P_{ON}^R, P_{OR}^R, P_{OS}^R\), and the manufacturer follows with wholesale prices \(W_{ON}^R, W_{OR}^R, W_{OS}^R\). The profit functions remain similar, but the decision sequence changes.
Proposition 3: In the retailer-led model, the optimal prices are:
$$ W_{ON}^{R*} = \frac{\alpha_1 + \alpha_2 + 3C_{ON} \alpha_2 – \alpha_1 \alpha_2^2}{4\alpha_2}, $$
$$ W_{OR}^{R*} = \frac{\alpha_1 + 3C_{OR} \alpha_2 + \alpha_1 \alpha_2 – \alpha_1 \alpha_2^2}{4\alpha_2}, $$
$$ W_{OS}^{R*} = \frac{\alpha_1 + 3(B + C_{OS}) \alpha_2}{4\alpha_2}, $$
$$ P_{ON}^{R*} = \frac{(3 + C_{ON}) \alpha_2 – 3\alpha_1 (-1 + \alpha_2^2)}{4\alpha_2}, $$
$$ P_{OR}^{R*} = \frac{C_{OR} \alpha_2 + \alpha_1 (3 + 3\alpha_2 – 3\alpha_2^2)}{4\alpha_2}, $$
$$ P_{OS}^{R*} = \frac{3\alpha_1 + (B + C_{OS}) \alpha_2}{4\alpha_2}. $$
Proof: The manufacturer’s profit function is strictly concave in wholesale prices, as shown by its diagonal Hessian matrix. Solving the first-order conditions provides optimal wholesale prices, which are then used in the retailer’s profit function to determine retail prices through additional derivatives.
Under retailer leadership, consumer WTP positively affects all wholesale prices, reflecting the retailer’s responsiveness to market demand. For new electric vehicles, higher WTP leads to lower retail prices to attract consumers, while for remanufactured ones, it allows for higher pricing due to perceived quality. Used product prices remain less influenced by WTP, depending more on market conditions.
Comparison of Optimal Solutions Under Different Power Structures
Proposition 4: In all models, \(\frac{dq_{ON}^j}{d\alpha_2} = 0\), indicating that changes in \(\alpha_2\) do not affect new product demand. However, pricing and demand for other products vary across models. Specifically, in centralized decision-making, consumer WTP positively influences prices for new and used electric vehicles, whereas in manufacturer- or retailer-led models, the opposite occurs due to strategic pricing power.
Proposition 5: The following inequalities always hold: \(W_{ON}^M > W_{ON}^R\) and \(P_{ON}^M < P_{ON}^R < P_{ON}^{MR}\). This shows that manufacturer leadership results in higher wholesale prices but lower retail prices for new products compared to retailer leadership, with centralized decision-making yielding the lowest retail prices.
Proposition 6: For remanufactured electric vehicles, \(P_{OR}^M < P_{OR}^R < P_{OR}^{MR}\). For wholesale prices, \(W_{OR}^M < W_{OR}^R\) if \(\epsilon_1 < \alpha_2\), and \(W_{OR}^M > W_{OR}^R\) if \(\epsilon_2 > \alpha_2\). This highlights the impact of WTP coefficients on pricing strategies.
Proposition 7: For used electric vehicles, \(W_{OS}^M < W_{OS}^R\) when \(\epsilon_2 < \alpha_2\), and \(W_{OS}^M > W_{OS}^R\) when \(\epsilon_2 > \alpha_2\). Similarly, \(P_{OS}^M < P_{OS}^R\) if \(\alpha_2 < \epsilon_3\), and \(P_{OS}^M > P_{OS}^R\) if \(\alpha_2 > \epsilon_3\). These results emphasize the need for dynamic pricing based on consumer WTP and cost structures.
These propositions underscore that supply chain power structures significantly influence pricing and demand. Centralized decision-making generally leads to lower prices and higher demand, while decentralized models require coordination mechanisms to mitigate inefficiencies.
Two-Part Tariff Coordination Model (Model MT)
To address profit suboptimality in decentralized decision-making, I propose a two-part tariff contract. This involves the manufacturer setting wholesale prices at marginal cost and the retailer paying a fixed fee to compensate the manufacturer.
Proposition 8: In the coordination model:
(1) The retailer’s optimal pricing decisions align with the centralized model: \(P_{ON}^{MT*} = P_{ON}^{MR*}\), \(P_{OR}^{MT*} = P_{OR}^{MR*}\), \(P_{OS}^{MT*} = P_{OS}^{MR*}\).
(2) The manufacturer sets wholesale prices at cost: \(W_{ON}^{MT*} = C_{ON}\), \(W_{OR}^{MT*} = C_{OR}\), \(W_{OS}^{MT*} = C_{OS}\).
(3) The retailer pays a fixed fee \(\beta_2\) to the manufacturer, ensuring that both parties’ profits are at least as high as in the decentralized model. The profit functions become:
$$ \pi_R^{MT} = \text{Expression based on optimal prices and costs}, $$
$$ \pi_M^{MT} = \beta_2, $$
with \(\beta_2\) satisfying \(\pi_M^{MT} \geq \pi_M^{M*}\) and \(\pi_R^{MT} \geq \pi_R^{M*}\). The total supply chain profit matches the centralized model: \(\pi_I^{MT} = \pi_I^{MR}\).
This contract facilitates Pareto improvements, enhancing overall supply chain performance while ensuring fair profit distribution. It encourages long-term collaboration between manufacturers and retailers in the electric vehicle industry, particularly in China, where market volatility and consumer preferences require adaptive strategies.
Numerical Simulation
I conduct numerical simulations to validate the analytical results and examine the impact of consumer WTP differences on profits. The parameters are set as follows: \(C_{ON} = 0.8\), \(C_{OR} = 0.6\), \(C_{OS} = 0.3\), \(B = 0.2\), and \(\alpha_2 = 0.8\). I analyze variations in \(\alpha_1\) and \(\alpha_2\) on manufacturer profit, retailer profit, and total supply chain profit under different models.
The results show that \(\alpha_1\) (representing the relative WTP between remanufactured and used electric vehicles) non-linearly affects profits. At low \(\alpha_1\) values, remanufactured products dominate, yielding higher manufacturer profits. As \(\alpha_1\) increases, used products gain substitutability, reducing profits until a critical point where remanufactured demand rebounds. Retailer profits mirror this trend, indicating interconnectedness in the supply chain. For example, in manufacturer-led models, profit fluctuations are more pronounced, whereas retailer-led models show greater stability due to better market alignment.
Similarly, \(\alpha_2\) (representing the relative WTP between new and remanufactured electric vehicles) influences profits differentially. Centralized decision-making consistently achieves higher total profits than decentralized models, highlighting the benefits of coordination. When \(\alpha_2\) exceeds a threshold, remanufactured product demand surges, driving profit growth and underscoring the economic value of remanufacturing. In decentralized settings, retailer leadership often outperforms manufacturer leadership, leveraging closer consumer proximity.
These findings suggest that firms should prioritize supply chain collaboration, especially in the China EV market, where consumer WTP is highly variable. Strategies include investing in remanufacturing technology to narrow quality gaps with new products and implementing data-sharing platforms for real-time demand insights. Policy incentives, such as carbon credit systems, can further encourage remanufacturing activities.
Conclusion and Managerial Implications
This study explores remanufacturing supply chain decisions for electric vehicles under consumer WTP differences. By comparing centralized and decentralized decision-making models, I demonstrate that centralized pricing leads to lower prices and higher demand, while decentralized models require coordination mechanisms like two-part tariff contracts to enhance efficiency and profit distribution. Key conclusions include:
– Consumer WTP differences critically influence product pricing and demand, with specific thresholds dictating market outcomes.
– Manufacturers can achieve higher returns through refurbishment of used electric vehicles, but must balance this against substitution effects.
– Supply chain coordination is essential for maximizing overall profits, particularly in dynamic markets like China’s EV sector.
Managerially, firms should segment markets based on WTP: target high-WTP consumers with advanced new electric vehicles, mid-WTP segments with cost-effective remanufactured options, and low-WTP groups with affordable used vehicles. Enhancing consumer trust through quality certifications, extended warranties, and transparent communication can alleviate skepticism and boost demand for remanufactured products. Additionally, adopting two-part tariff contracts can align incentives between manufacturers and retailers, fostering long-term partnerships.
Future research could extend to multi-period models, incorporating product lifecycle considerations and competitive dynamics among multiple electric vehicle brands. This would provide deeper insights into sustainable supply chain management in the evolving automotive industry.