Choosing Recovery Strategies for Waste Electronics: How Product Modularity Influences Cooperation and Competition

Abstract

Modular design facilitates easy disassembly and reduces the manufacturer’s remanufacturing costs. However, the simplicity and modular structure of products can intensify competition between manufacturers and third-party recyclers. To improve recovery efficiency, this study examines the impact of modular design on the manufacturer’s selection of recovery strategies, including centralized, cooperation, and competition strategies. We examine the optimal recovery strategy for achieving both economic goals, such as supply chain profit, and environmental goals, such as collection quantity. Our results indicate that the manufacturer should adopt cooperation recovery and invest in higher modularity when faced with strong competition from third-party recyclers. Conversely, when the competitiveness of third-party recovery is relatively low, a competition recovery strategy is more advantageous. Contrary to conventional wisdom, which suggests limiting product disassembly to reduce third-party recovery competitiveness, our results indicate that manufacturers should invest in higher modularity and avoid engaging in price wars to prevent third-party entry. Moreover, competition recovery leads to a higher collection quantity, while cooperation recovery is preferred in terms of supply chain profit. This study provides theoretical guidance for manufacturers in selecting optimal recovery strategies and offers recommendations for governments on regulating product disassembly effectively.

1. Introduction

The global surge in e-waste generation is a pressing concern. According to the Global Transboundary E-Waste Flows Monitor [1], projections indicate an annual e-waste total of 74.7 million metric tons by 2030, escalating to 110 million metric tons by 2050. Unfortunately, a substantial portion of e-waste remains uncollected and lacks systematic processing, resulting in low collection and recycling rates. A major factor is the intricate design of electronic products, many of which are not engineered for easy reuse. Due to the difficulty in separating core components, extracting and recovering all valuable materials is challenging [2].

Modular design facilitates the easy disassembly of products into several components, which can then be reused in new customized products. This enables the manufacturer to effectively minimize remanufacturing costs [3,4]. Within the electronics industry, modular design has been adopted by some manufacturers to increase recovery efficiency. For example, all Xerox products are designed to minimize energy consumption and come with energy-saving features. In 2023, over 1.7 million Xerox toner cartridges were produced using recovered cartridges, achieving an average of 90% reuse by weight of reclaimed parts. For instance, Xerox’s adoption of modular design for printers and photocopiers, exemplified by the Xerox DC265 copiers, resulted in a twofold increase in remanufacturing savings compared to a nonmodular design [5]. The modular design approach significantly enhances the efficiency of module remanufacturing and reuse processes. Another example of modular electronics is Fairphone. The modular design of the Fairphone 4 allows for the recovery of 28% of valuable resources, while the Fairphone 5 shows an even more impressive improvement, exceeding 43% [6]. Similarly, the Fairbud XL headphones are constructed from 95% recycled aluminum, 40% recycled tin, a travel pouch made from 100% recycled polyester, and 89% recycled plastics, all contributing to enhanced recovery efficiency through modular design.

However, the manufacturer is concerned about investing in product modularity due to the involvement of a third-party recycler (TPR). Normally, recovery activities are adopted by the manufacturer due to technical barriers, yet modularity lowers these barriers and reduces recovery costs for the TPR by making returned products highly modularized and simplified. This allows TPR to participate in the profitable recovery business, reducing the quantity of used products the manufacturer can collect in the recovery market [7]. Consequently, modularity increases the strength and recovery competition from the TPR even further. This situation presents an interesting trade-off for the manufacturer: balancing the cost reduction benefits in its remanufacturing against the increased competition from TPRs. Alternatively, the manufacturer could also cooperate with a TPR to leverage local advantages and avoid competition. This is particularly beneficial in developing regions, where manufacturers may lack the infrastructure and expertise to profitably collect used products [8,9]. The presence of TPR makes the manufacturer’s recovery strategies more complex and challenging. Therefore, the manufacturer needs guidance on how to strategically invest in modularity and choose between a competition or cooperation recovery strategy to achieve both economic and environmental goals.

Competition recovery: When the manufacturer and the TPR compete to collect and dispose of cores, they often use different recovery methods. Many TPRs, lacking knowledge or techniques for sophisticated remanufacturing, tend to adopt a simplified approach—extracting materials like plastic, batteries, and metals through cores via recycling [8]. Unlike TPRs, manufacturers adopt advanced recovery techniques, specifically remanufacturing. This process reuses the components of cores to produce new products. For instance, Hon Hai Precision Industry, acting as a TPR, initiated the recycling of used Apple devices in specific regions. In addition, Apple conducts official remanufacturing activities and competes with Hon Hai Precision Industry [7]. IBM establishes collection channels and departments to participate in collection activities, competing with local TPRs [10].

Cooperation recovery: Some manufacturers have shifted their strategy from competing to cooperating with the TPR to share recovery benefits and avoid cannibalization. Consequently, manufacturers are increasingly outsourcing the collection business to TPRs strategically located near consumers. For example, Dell and Acer subcontract the collection of used products to third-party firms [11]. Microsoft collaborates with TES for the collection and processing of devices [12]. Xerox has introduced the Green World Alliance (GWA) program, focusing on collecting used spent toner cartridges. Third parties, such as Greiner Associates, in Europe, and Close the Loop, in the U.S., are responsible for collecting, inspecting, disassembling, cleaning, and packaging used toner cartridges for shipment to Xerox for reprocessing [13]. Other TPRs, including Huishoubao, Call2Recycle, and Gadgetgone, have operated similar businesses [14]. However, when outsourcing collection activities, the manufacturer needs to provide sufficient incentives to motivate TPRs to collaborate effectively or achieve better results. Without proper incentives, the TPR may prefer to collect and process cores independently, presenting a trade-off between the benefits of the cooperative and the competitive scenario.

Modular design allows for easier and cheaper repairs and upgrades, influencing customer recovery behavior by providing the flexibility to independently replace specific modules. This enables customers to keep their products longer and update only the necessary parts with newer versions [15], and it promotes a positive self-repair experience. Data from smartphone manufacturer SmartMod reveal that modular devices are three times more likely to be repaired by users themselves (39%) compared to semi-modular devices (13%) [16]. Apple incorporated iPhone and MacBook modularity and repairability scores into its online store. The iPhone 11s is rated between 4.5 and 4.6 out of 10, while the members of the iPhone 12 lineup all have higher scores of 6. Higher modularity scores give customers a sense of how easily a device can be disassembled and strengthen the likelihood of self-repair modular products [16]. Apple’s Self-Service Repair program supports 40 Apple products across 33 countries, providing consumers with access to repair manuals, genuine Apple parts, and tools for independent repairs. Additionally, thanks to the easy disassembly design of its products, Apple’s robot, Daisy, is capable of disassembling 29 iPhone models into individual components. This functionality allows for the recovery of more valuable materials for reuse and enhances the efficiency of remanufacturing [17]. This initiative enables consumers to replace worn-out or broken parts, such as screens, batteries, and cameras, through repair instructions. Consequently, higher modularity leads to a reduced number of used products collected, encouraging customers to keep their products for extended periods.

Overall, modular design poses significant trade-offs between remanufacturing efficiency, recovery competition, and consumer recovery behaviors. Therefore, we compare three recovery models for the manufacturer: (1) Centralized, in which the manufacturer independently collects and disposes of cores; (2) Cooperation, in which the manufacturer outsources core collection to a TPR; and (3) Competition, in which the manufacturer competes with an unauthorized TPR for core collection and disposal. Our analysis considers the influence of modular design on both profit and the quantity of cores collected. In this study, we aim to address the following questions:

• What is the optimal modularity level for each recovery strategy?
• How does the TPR influence product modularity, manufacturer’s profits, and collection quantity?
• Faced with a TPR, which recovery scenario yields greater benefits for both the manufacturer and the environment?

In this study, we develop single-period recovery models involving a manufacturer and a TPR. We first propose three recovery strategies and determine pricing decisions, product modularity, and collection prices. We then conduct a comparative analysis, focusing on modularity and collection quantity across these different recovery strategies. Finally, we compare the profits of each strategy to identify the optimal recovery approach from both economic and environmental perspectives. The results demonstrate that, when faced with increased recovery competitiveness from the TPR, recovery strategy selection is influenced by factors such as recycling benefits and consumer channel preferences. From the perspective of collection quantity, competition can always lead to optimal collection quantity, sometimes even surpassing the centralized case. The TPR’s involvement in recovery activities stimulates collection quantity, boosting the circulation of resources, similar to the findings of Wei et al. (2021) [10]. However, from the profit perspective, when TPR exhibits strong recovery competitiveness, the manufacturer strategically adopts a cooperation recovery strategy and invests in higher modularity. Otherwise, for less competitive TPRs, such as local informal sectors, the manufacturer prefers a competitive recovery strategy, using high modularity to deter them from entering the recovery business. We find that modular product design is effective for the OEM in competing with the TPR. This contrasts with the study by Wu (2013) [18], who suggested that manufacturers can limit TPR competitiveness by setting a low degree of disassembly and modularity in products.

The main contributions of our research are summarized as follows: First, we embed product modularity into a closed-loop supply chain (CLSC) context. Thus far, much of the research on modularity has focused on its impacts on the forward supply chain, and few studies have shown how modularity also impacts returned product recovery options and strategies in a CLSC. Second, we investigate the trade-offs introduced by modularity, specifically examining its impact on remanufacturing efficiency and recovery competition. Third, most studies investigate the impact of modularity on purchasing decisions for modular upgradable products, and few discuss modularity’s effect on the recovery behavior of consumers. With a better understanding of the effect of modularity on the recovery behaviors of the manufacturer, TPR, and consumer, our research facilitates informed decision-making on product modularity and optimal recovery strategy selection, thereby increasing profits and enhancing the sustainability of CLSCs.

The remainder of this paper is organized as follows: The literature review is presented in Section 2, while the model assumptions are described in Section 3. In Section 4, the optimal decisions under each scenario are obtained, and a comparative analysis is presented in Section 5. In Section 6, the economic and environmental implications of each recovery strategy are analyzed. An extension part and our conclusions are presented in Section 7 and Section 8. Section 9 presents the limitations of the study.

2. Literature Review

Our study follows three different streams in CLSC research: collecting competition and cooperation, as well as the impact of modular design on recovery, and recovery options.

2.1. Collecting Competition and Cooperation

Early research on CLSCs primarily focused on network design [19,20] and inventory control [21,22,23], employing optimization models such as mixed-integer linear programming and genetic algorithms to enhance overall system performance. For example, Mutha and Pokharel (2009) [24] explored modularization as a strategy to avoid the disposal of reusable modules, addressing key decisions regarding the number, location, and capacities of facilities, as well as the allocation of material flows. More recently, research on recovery strategies has shifted toward sustainable management and coordination, with game theory models being increasingly adopted to facilitate more realistic decision-making among supply chain participants [25]. Some studies incorporate empirical approaches, such as using integrative visual tools to represent the benefits and challenges of product modularity for circular business models [26]. There is also a growing body of work applying game theory in CLSC contexts, particularly for modeling channel selection, remanufacturing, and modularity investment decisions [3,4,27,28].

Savaskan et al. (2004) [29] investigated three collecting channels: manufacturer, retailer, and third-party collecting channels. Based on Savaskan et al. (2004) [29], researchers have extended their studies to collecting competition or cooperation. One part of the literature generally analyzes collecting competition between different supply chain members, that is, the supplier and the manufacturer [30], the manufacturer and the retailer [31,32], the manufacturer and the third party [33,34], the formal and the informal collector [35], and hybrid collection [7,36,37]. This stream of research finds the optimal collection rate for each channel, government subsidies, and taxes. Zhou et al. (2023) [37] investigated the competition among an electric vehicle manufacturer, an electric vehicle recycler, and a third-party recycler. When remanufacturing cost savings are moderate, a recycling channel led by the electric vehicle recycler benefits all parties involved. Another part of the literature focuses on the competition between different channel forms, such as bricks and clicks between retailers, traditional channels, and direct channels between manufacturers. These studies investigate whether supply chain members should open a dual channel to collect and sell refurbished products. Feng et al. (2017) [38] compared single and dual collecting channels and found that dual channels always outperform single channels. Yu et al. (2023) [36] developed a two-echelon CLSC model comprising a manufacturer, a platform, and a collector. They found that in a low-price stage with low disassembling costs, a dual-channel collection is the optimal choice. Otherwise, the manufacturer will solely authorize the collector. In the face of competition, supply chain members are motivated to collaborate in order to achieve higher profits. Other studies investigate collecting cooperation and determine the optimal cooperation form and the conditions under which they prefer to cooperate. Normally, the manufacturer authorizes or outsources collecting or recovery business to a third-party or retailer. De Giovanni and Zaccour (2014) [39] investigated manufacturers’ decisions on whether the product collection business should be outsourced. Their findings reveal that the manufacturer prefers to delegate collection activities to the retailer. Q. Wang et al. (2020) [40] studied a CLSC consisting of one manufacturer and two competing retailers. Research indicates that retailer collusion always brings remarkable profit improvement to retailers as a whole.

Several studies have simultaneously considered competition and cooperation. Ghosh et al. (2018) [33] studied competition and collaboration between the OEM and the independent remanufacturer. Under competition, the optimal strategy for the OEM is to provide higher product differentiation by investing in higher-quality product design. Under collaboration, the optimal strategy is to lower the product design quality and set higher prices to maintain margins. Normally, the manufacturer prefers to pursue collaboration rather than fend against a third-party [41,42]. However, they do not consider the impact of the modular product design on the decisions of supply chain members, which is key to our research. We address the problem of how product modularity affects the firm’s optimal choice among a centralized, a cooperation, and a competition model.

2.2. Modularity’s Impact on Recovery

Many studies have examined the direct and indirect impacts of product modularity on the manufacturing process, especially on product development and innovation [43], mass customization capability and product variety [3,4,28], or reducing the risk of failure of a development project [44]. Product modularity can facilitate economies of scale and component commonality, thereby decreasing the inventory quantities of components and lowering assembly costs [18]. Recently, a stream of research has studied the benefits of product modularity for recovery. For example, modular design may increase the salvage value of returned products because products can easily be dismantled and reused [45]. Modularity can also facilitate the inspection, disassembly, repair, and replacement of used product components, which in turn enhances the efficiency of remanufacturing [4,27]. Machado et al. (2024) [26] proposed a comprehensive visual tool that highlights both the benefits and the challenges of product modularity in circular business models, such as facilitating disassembly, maintenance, reuse, recycling, and remanufacturing. Jiang (2023) [27] explored the optimal modularity level for new mass customization products and the most effective recycling channel for used MC products. The study suggests that direct recycling by the manufacturer yields optimal results in terms of modularity, recycling rate, supply chain profits, and overall social welfare. Yang and Jiang (2024) [46] examined the modular design of new products and the remanufacturing strategy for used products in the context of mass customization. Their findings suggest that the manufacturer-led remanufacturing approach can outperform other approaches in terms of both profit and consumer surplus. In addition to traditional selling and recovery modes, a recent stream of research has extended the study of the impact of modularity on circular business models, including leasing, servicing, or renting, where product ownership is maintained by firms [15].

However, the focus of this research stream is on the impact of modularity on manufacturers’ or other supply chain members’ pricing decisions and recovery costs; therefore, it does not consider the effect of product modularity on consumers, especially their recovery behavior. Several studies focus on product modularity’s effect on consumer purchase behavior; for example, Ülkü et al. (2012) [47] examine the initial choice between modular and integral products and the subsequent upgrade decisions. However, they do not account for product modularity in consumer recovery behavior. We investigate the effect of modularity on consumers’ preference for processing their used devices. This research provides firms with insights into how consumers value modular products. By evaluating three recovery strategies, the proposed approach leverages game theory as an effective tool for modeling the manufacturer’s decisions in terms of modular design, pricing, and recovery strategies.

2.3. Recovery Options

After collecting used products, supply chain members need to choose the optimal recovery method based on their capacity and the condition of the used products. A large portion of the literature on CLSCs assumes that all collected products can be remanufactured and focuses on finding the optimal collection rate [48,49]. The research reviewed above seldom considers the different recovery options for the collected products. As used products have different quality levels and members’ disposing abilities vary, used products may be processed through different recovery options [50], such as recycling, remanufacturing, and refurbishing. Normally, remanufacturing reuses products at the component level, which is better than recycling at the material level [51]. A member can have both remanufacturing and recycling coexisting under the recovery process. Han et al. (2020) [52] investigated the manufacturer’s three recovery strategies, including remanufacturing products, recycling materials, and both. They found that the manufacturer prefers remanufacturing products when the products are of high quality. Firms may also implement a differentiated recycling strategy, in which the high-quality WEEE is processed similarly to standard recycling, while low-quality WEEE is scrapped. This approach results in different cost structures for the firm [53]. Zhang et al. (2024) [54] examined optimal remanufacturing strategies within a closed-loop supply chain, considering government-mandated recycling and remanufacturing targets. Huang et al. (2024) [55] investigated three product categories within a modular architecture: new products, partially remanufactured products, and fully remanufactured products. The manufacturer’s choice of remanufacturing strategy depends on the production costs of the partially and fully remanufactured products.

Several studies investigate various disposing abilities between supply chain members; as the manufacturer has advanced recovery technology, it will choose remanufacturing. However, the TPR lacks adequate recovery technologies and can only extract raw materials through recycling [8,31]. In this study, we assume that the manufacturer and third party can handle collected products separately. Previous research has studied competition between the manufacturer and third party and their disposal ability, but the effect of modularity on the manufacturer and the TPR has not been studied. In contrast to He et al. (2019) [31], we consider the effect of modularity on both the manufacturer and the third party, which decreases their remanufacturing and recycling costs simultaneously. To gain deeper insights into product modularity investment from a game-theoretical perspective, this paper adopts a Stackelberg game [7,8,56,57] to address two key aspects: product pricing and modularity investment, as well as the selection of recovery channels (cooperation and competition) by the TPR or manufacturer in a reverse supply chain. These results also guide environmental regulations regarding when to promote investment in modularity and how to choose recovery strategies to improve environmental outcomes and the sustainability of CLSCs. To better highlight our contribution, we present a comparison between this study and the relevant literature in

Table 1. Comparison of this research with the existing literature.

	Product Architecture		Recovery Strategy		Method	Focus
	Integrated	Modular	Competition	Cooperation
Zhou et al. (2023) [37]			✓		Cournot competition	Remanufacturing decisions of two symmetric manufacturers
Ülkü et al. (2012) [47]	✓	✓			Empirical study	The value of modularity for consumers
He et al. (2019, 2022) [8,31]	✓		✓		Stackelberg game	Consumer inconvenience and competitive collecting channels
Amend et al. (2022) [16]		✓			Empirical study	The relationship between circular economy and modular products
Ghosh et al. (2018) [33]	✓		✓	✓	Game theory	Recovery strategy selection and contract design
Yuchi et al. (2021) [23]	✓				Mixed-integer nonlinear model	Locations of distribution and remanufacturing centers
J. Wang and He (2022, 2023) [3,4]		✓	✓		Stackelberg game	Dual channels and return policy/government subsidy
This paper		✓	✓	✓	Stackelberg game	Modular investments and recovery strategy selection of manufacturer and TPR

.

3. Assumptions

As products become highly modularized, we attempt to provide guidance to the manufacturer for selecting a recovery strategy when facing a TPR. Initially, we establish the benchmark model, wherein the manufacturer autonomously collects and remanufactures all used products in the absence of a TPR. Subsequently, we investigate the manufacturer’s optimal decisions regarding modularity investment and collection in a cooperation case. Next, we explore a competitive scenario in which the manufacturer and the TPR independently collect used devices (see Figure 1). The key notations are summarized in

Table 2. Notations.

Indices
$i$	Index for recovery strategy: $i=1,2,3$
$j$	Index for supply chain member (manufacturer, TPR): $j=M,T$
Parameters
$\alpha$	Intrinsic demand for new products
$\beta$	Price sensitivity of the new products
$d$	Customer demand for new products
$c_n$	Manufacturer’s manufacturing cost for new products
$c_r^M$	Manufacturer’s remanufacturing cost
$c_r^R$	Recycler’s recycling cost
$\Delta$	Unit cost savings through remanufacturing via modular design
$\rho$	Unit cost savings through recycling via modular design
$f$	Fixed benefit of the manufacturer
$s$	Customer’s valuation of old devices
$\theta$	Customer’s preference for the TPR over the manufacturer
$\delta$	Customer’s sensitivity to product modularity
$v$	Recovery benefits of the recycler
$\eta$	Cost index of modularity investment
$q_m$$, q_t$	Collection volume of the manufacturer/the TPR
${\mathrm{\Pi }}_{Mi}$$, {\mathrm{\Pi }}_{Ti}$	The profits of the manufacturer/the TPR
${\mathrm{\Pi }}_{Si}$	The profits of the supply chain
Decision variables
$m$	Product modularity level
$p$	Selling price of new products
$r_m,r_t$	Collection price of the manufacturer/the TPR
$b$	Transfer price for returning cores from the recycler to the manufacturer

.

Assumption 1.

The demand for new products depends on the selling price p.

The market demand is negatively affected by price [31], where $d=\alpha -\beta p$. In the demand function, $\alpha$ describes the basic market size when the price is zero. $\beta$ is the price sensitivity of the product, reflecting that demand is downward-sloping in retail price.

Assumption 2.

Modular design generates cost savings for both the manufacturer and the TPR.

Product modularity refers to the observation that interdependence between subsystems is sufficiently low to permit separability and, thus, that components can be physically separated [3,4,27]. Modular product architecture has several characteristics: several components (standard), interfaces (standardization and specification), degree of coupling, and substitutability [58] (see

Table 3. Characteristics of modular product architecture.

Characteristics	Modular Product Architecture
Interfaces	Decoupled
Outcome	Economies of scale
Component Substitutability	High
Component Recombinability	High
Component Separability	High

). The degree of product modularity $m$ can be improved by redesigning module interfaces that allow more frequent reuse of modular components. Previous research, such as [28], has explored how product modularity can reduce production costs through economies of scale. In this study, the key results and insights remain consistent, even when accounting for the impact of product modularity on production costs (see Section 7.2).

Remanufacturing can recover the residual value of used products and reuse components in new products, which reduces the consumption of energy and materials [59] and, thus, reduces manufacturing costs by 40–65% [11]. This method, often combined with technological upgrades, can restore end-of-life products to a performance level equal to or better than those of new products [8]. For the manufacturer, only new products are available after remanufacturing. The manufacturer reduces the consumption of resources through investment in modularity. Therefore, let $c_n$ denote the manufacturing cost and $\Delta m+f$ denote cost savings from remanufacturing [27,46] or the salvage value of the cores in total. The salvage value of products rises with higher modularity $m$, where $f$ is the fixed salvage value when no modularity has been invested or the product is totally integrated, and $\Delta$ is the benefit of salvage value with respect to the increase in the product modularity level $m$ [3,4]. For example, Xerox’s adoption of modular design has enabled the efficient replacement of Xerographic components, significantly reducing the costs associated with product disposal [55]. Meanwhile, the manufacturer incurs investment costs at the modularity level, denoted by $\eta m^2$ [27,28,45,46,60,61].

However, the TPR lacks the knowledge and techniques needed for remanufacturing used products. In the recycling process, the TPR extracts useful raw materials, such as metals, glass, and plastics, from cores and receives a fixed residual value $v$, leaving the remaining parts buried [8]. When the manufacturer invests in modularity, the modules are designed to be easily separated under a modular architecture, thereby enabling the TPR to disassemble more modules independently, thus leading to cost savings $c_r^R-\rho m$ in the recycling process. The benefit of the cores from the TPR is $v-(c_r^R-\rho m)$. For example, modular design supports TES, a Singaporean third-party recycler, by simplifying the dismantling of devices and enabling the extraction of valuable materials like gold, silver, and copper. Owing to the recovery technique advantage [8,31], the manufacturer can extract more fixed salvage from cores compared with the TPR, where $f>v-c_r^R$. To model the potential benefits of remanufacturing for the manufacturer, we explicitly consider a higher cost-saving factor for the manufacturer than the TPR, where $\Delta >\rho$.

Assumption 3.

Cores are collected by the manufacturer or the TPR.

In the centralized model, the manufacturer sells new products and collects cores. The consumer is willing to return the used product if and only if the collection price $r_m$ is greater than their unwillingness to return or the resident value of the product $s$. The customer’s utility for returning the old products to the manufacturer is $U_m=r_m-s-\delta m$. $s$ is uniformly distributed within the consumer population from 0 to 1, with a density of 1 [38]. Conversely, a modular architecture provides customers with flexibility in replacing modules of the product independently, thereby enabling customers to hold onto their products and only replace one of the modules with a new version [15]. In practice, Microsoft and Apple offer replaceable components that are used to service or repair Surface devices and can support out-of-warranty scenarios [62,63]. According to Amend et al. [16], consumers are more likely to attempt self-repair on modular devices compared to conventional smartphones. While 55% of conventional smartphones remain unrepaired, only 13% of modular smartphones are not repaired. Product modularity motivates users to repair or upgrade rather than replace the entire device, leading to a decrease in the number of used products collected by manufacturers as the modularity $m$ increases. As a benchmark, the manufacturer derives the collection quantity as $q_m=r_m-\delta m$, where $0\le r_m\le 1$.

In the cooperation model, the manufacturer authorizes the TPR to collect cores. The TPR then sends the cores back to the manufacturer and receives a transfer price $b$. The customer’s utility for returning the old product to the TPR is $U_t=r_t-s-\delta m$. The collection quantity is $q_t=r_t-\delta m$, where $0\le r_t\le 1$.

In the competition model, the manufacturer and the TPR compete to collect cores, incurring acquisition costs, which are denoted as $r_m$ and $r_t$, respectively. The utility functions for consumers are denoted as $U_m=r_m-\theta s-\delta m$ and $U_t=r_t-s-\delta m$. Consumer preference for the third-party collecting channel over the manufacturer’s channel is represented by $\theta$, which satisfies $\theta >1$. This indicates that the third-party collection channel has an advantage over the manufacturer’s channel. In other words, when dealing with homogeneous consumers, the third party incurs lower costs than the manufacturer. Some TPRs, typically located closer to consumers, benefit from lower logistics costs and may offer collection services or quicker cash payments to acquire used products [64]. If $\theta <1$, the manufacturer’s collection channel has strength in collection compared with the TPR channel. This scenario is further explored in the extension (Section 7.1) of this paper. The market segmentation is characterized (see Figure 2), detailing the collection quantity for each member and the quantity retained by customers [38] (see Appendix A): $q_m=\left\{\begin{array}{c}{\displaystyle \frac{r_m-r_t}{\theta -1}},r_m<\theta r_t-(\theta -1)\delta m\\ {\displaystyle \frac{r_m-\delta m}{\theta }},otherwise\end{array}\right.$, $q_t=\left\{\begin{array}{c}{\displaystyle \frac{\theta r_t-(\theta -1)\delta m-r_m}{\theta -1}},r_m<\theta r_t-(\theta -1)\delta m\\ 0,otherwise\end{array}\right.$.

Assumption 4.

The manufacturer is a Stackelberg leader, whereas the TPR is the follower.

In a stable CLSC, the demand for new products and the collection of used products are dynamically balanced. We assume that there is a large potential supply of cores due to technical improvements and shortened product life cycles. The manufacturer can collect the required quantity of cores as different generations of products exist in the recovery market. The manufacturer typically has greater bargaining power over the TPR [65] because it possesses substantial channel power and a strong brand to set the prices of new products [66]. Our model implements a one-period Stackelberg game, where the manufacturer acts as the leader and the TPR as the follower. Since game theory focuses on predicting the outcomes of decisions made by individuals who are aware of others’ actions and choices, it is well-suited for analyzing strategic interactions among independent decision-makers. This approach can assist manufacturers in optimizing recovery strategies and achieving equilibrium points.

4. Models

4.1. Centralized Strategy

In the benchmark model, the manufacturer is responsible for taking all businesses in the forward supply chain, such as those selling new products, and the reverse supply chain, such as those collecting and processing returned products. This first model serves as a benchmark for comparison.

$$\begin{array}{c}{\mathrm{\Pi }}_M(p,r_m,m)=(p-c_n)(\alpha -\beta p)+(f+\Delta m-r_m)q_m-\eta m^2\\ s.t. r_m,m\in (0,1)\end{array}$$

(1)

We solve the Stackelberg game using the backward induction method. We assume $4\eta -{(\Delta -\delta )}^2>0$ to guarantee that the matrix is negative and the concavity of ${\mathrm{\Pi }}_M$ with regard to $p,m,r_m$ (all proofs are included in Appendix B). The condition highlights that the cost associated with enhancing product modularity $\eta$ should not significantly exceed the manufacturer’s remanufacturing benefit minus the TPR’s recycling benefit $\Delta -\delta$—the manufacturer’s advantage of remanufacturing through product modularity. If $4\eta -{(\Delta -\delta )}^2<0$, the profit function becomes non-concave. In such a scenario, the cost of investing in product modularity is negligible, leading the manufacturer to invest in modularity to the maximum extent possible. This outcome contradicts reality [8]. By solving the Lagrangian and KKT conditions, we can derive the equilibrium price of new products $p={\displaystyle \frac{\beta c_n+\alpha }{2\beta }}$, the modularity level $m={\displaystyle \frac{f\left(\Delta -\delta \right)}{4\eta -{\left(\Delta -\delta \right)}^2}}$, the collection price of cores $r_m={\displaystyle \frac{f\left(\delta (\Delta -\delta )+2\eta \right)}{4\eta -{\left(\Delta -\delta \right)}^2}}$, and the collection quantity of cores $q_m={\displaystyle \frac{2f\eta }{4\eta -{\left(\Delta -\delta \right)}^2}}$.

Proposition 1.

(centralized case 1): When there is no TPR, modularity m, collection price r_m, and collection quantity q_m increase with remanufacturing cost-saving Δ (see Appendix B).

An increase in remanufacturing cost savings decreases manufacturing costs; thus, the manufacturer benefits from selling new products and collecting old devices, and the manufacturer has an incentive to invest in higher modularity, making used products become more profitable. Meanwhile, the manufacturer sets a higher collection price, $r_m$, and the collection quantity of used products also increases. The result implies that the manufacturer will put more effort into modular product design and set higher collection prices if remanufacturing becomes more profitable.

4.2. Cooperation Strategy

In the cooperation model, the manufacturer outsources the collection business to a local TPR that is closer to the market. The TPR manages the reverse logistics flow, and the manufacturer buys back the used products from the TPR with a transfer price.

$$\begin{array}{c}{\mathrm{\Pi }}_M(p,b,m)=(p-c_n)(\alpha -\beta p)+(f+\Delta m-b)q_t-\eta m^2\\ s.t. b,m\in (0,1)\end{array}$$

(2)

$$\begin{array}{c}{\mathrm{\Pi }}_T(r_t)=b{q}_t-r_t{q}_t\\ s.t. r_t\in (0,1)\end{array}$$

(3)

The concavity of the profit function has been testified. Similarly, we derive the equilibrium new product price $p={\displaystyle \frac{\beta c_n+\alpha }{2\beta }}$, product modularity $m={\displaystyle \frac{f\left(\Delta -\delta \right)}{8\eta -{(\Delta -\delta )}^2}}$, and transfer price $b={\displaystyle \frac{f\left(\delta (\Delta -\delta )+4\eta \right)}{8\eta -{(\Delta -\delta )}^2}}$. Correspondingly, the TPR’s equilibrium collection price can be derived as $r_t={\displaystyle \frac{f\left(\delta (\Delta -\delta )+2\eta \right)}{8\eta -{(\Delta -\delta )}^2}}$, and the collection quantity of the cores as $q_t={\displaystyle \frac{2f\eta }{8\eta -{(\Delta -\delta )}^2}}$. All proofs are presented in Appendix C.

Proposition 2.

(cooperation case 2): When the manufacturer outsources collection to a TPR, modularity m, transfer price b, collection price r_t, and collection quantity q_t increase with remanufacturing cost saving Δ (see Appendix C).

In the cooperation model, the manufacturer outsources the TPR for collection. The manufacturer benefits from the cost reduction through remanufacturing and is motivated to invest in higher modularity and offer a higher transfer price to the TPR. As a result of the manufacturer’s modularity efforts, the TPR collects more cores.

4.3. Competition Strategy

A competitive scenario between the manufacturer and the TPR often appears in the electronics industries of developing countries. Many unauthorized TPRs participate in the recovery process due to a lack of relevant regulations. In these cases, the manufacturer manages all the activities in the forward and reverse supply chains. However, unauthorized recyclers lack the capability and technology for remanufacturing and can only perform basic recycling, which involves roughly extracting raw materials. The profits of the manufacturer and the TPR are as follows:

$${\mathrm{\Pi }}_M(p,r_m,m)=(p-c_n)(\alpha -\beta p)+(f+\Delta m-r_m)q_m-\eta m^2$$

(4)

$$s.t. r_m,m\in (0,1)$$

$$\begin{array}{l}{\mathrm{\Pi }}_T(r_t)=(v-(c_r^R-\rho m))q_t-r_t{q}_t\\ =(v-c_r^R+\rho m-r_t)q_t\end{array}$$

(5)

$$s.t. r_t\in (0,1)$$

4.3.1. The Manufacturer and the TPR Coexist in the Collecting Market

In the scenario $r_m<\theta r_t-(\theta -1)m$, the manufacturer collects $q_m={\displaystyle \frac{r_m-r_t}{\theta -1}}$ cores and the TPR collects $q_t={\displaystyle \frac{\theta r_t-(\theta -1)\delta m-r_m}{\theta -1}}$ cores. We build the Stackelberg game modeled as follows:

$$\begin{array}{c}{\mathrm{\Pi }}_M(p,r_m,m)=(p-c_n)(\alpha -\beta p)+(f+\Delta m-r_m)q_m-\eta m^2\\ s.t. \left\{\begin{array}{c}{r}_m<\theta r_t-(\theta -1)m,r_m,m\in (0,1)\\ \max {\mathrm{\Pi }}_T(r_t)=(v-c_r^R+\rho m-r_t)q_t\\ r_t\in (0,1)\end{array}\right.\end{array}$$

Similarly, concavity is tested to ensure the optimal solution of the model. The determinant of the Hessian matrix is negative definite, indicating the concavity of ${\mathrm{\Pi }}_M$ under the condition $4B\eta t-{(A-B\Delta )}^2>0$, where $A=\rho \theta +(\theta -1)\delta$, $B=2\theta -1$, $C=(v-c_r^R)\theta$, $t=2\theta (\theta -1)$. As in the centralized case, this condition implies that the cost savings of the manufacturer should be limited to a reasonable range. We derive the equilibrium outcomes as follows: $p={\displaystyle \frac{\beta c_n+\alpha }{2\beta }}$, $m={\displaystyle \frac{\left(B\Delta -A\right)\left(Bf-C\right)}{4B\eta t-{(A-B\Delta )}^2}}$, $r_m={\displaystyle \frac{\left((A-B\Delta )\Delta +2\eta t\right)C+(2B\eta t-A(A-B\Delta ))f}{4B\eta t-{(A-B\Delta )}^2}}$, $q_m={\displaystyle \frac{2B\eta \left(Bf-C\right)}{4B\eta t-{\left(A-B\Delta \right)}^2}}$, $r_t={\displaystyle \frac{\left(\left(\Delta (B+1)(A-B\Delta )\right)+2\eta t+4B\eta t\right)C-\left(A\left(B+1\right)\left(A-B\Delta \right)-2B\eta t\right)f}{2\theta \left(4B\eta t-{\left(A-B\Delta \right)}^2\right)}}$, $q_t={\displaystyle \frac{\left(2\left(\Delta -\delta \right)\left(\theta -1\right)\left(A-B\Delta \right)+\left(4\theta -3\right)2\eta t\right)C+\left(2\theta \left(\theta -1\right)\left(\delta -\rho \right)\left(A-B\Delta \right)-2B\eta t\right)f}{2\left(\theta -1\right)\left(4B\eta t-{\left(A-B\Delta \right)}^2\right)}}$.

To make sure the optimal decisions satisfy the constraint $r_m<\theta r_t-(\theta -1)m$, the inequality must hold, i.e., $C>\mathrm{\Gamma }$ (see Appendix D). The TPR can remain in the collection market only if its benefits exceed a certain threshold. This ensures that the optimal collection quantities for each member are greater than zero, allowing the manufacturer and the TPR to coexist in the collection market.

Proposition 3.

(competition case 3.1): When the manufacturer and the TPR coexist, the manufacturer’s modularity m, collection price r_m, and collection quantity q_m increase with remanufacturing cost savings Δ, while the TPR’s collection price r_t and collection quantity q_t also increase (see Appendix D).

When the recycler joins the collection competition ($r_m<\theta r_t-(\theta -1)\delta m$), a higher reduction in remanufacturing costs $\Delta$ will also increase the collection price $r_m$ and quantity of cores $q_m$ for the manufacturer, due to greater cost savings through remanufacturing. A higher level of cost reduction motivates the manufacturer to invest in higher modularity. Regarding the TPR, the collection prices of cores $r_t$ and collecting $q_t$ also increase in $\Delta$, which means that if the manufacturer obtains more cost savings $\Delta$ from recovery, the competition from the manufacturer will push the recycler to offer a high collection price and, thus, collect more cores in response.

4.3.2. The Manufacturer Monopolizes the Recovery Market

In the scenario $r_m\ge \theta r_t-(\theta -1)m$, the customer’s utility is always higher when selling to the manufacturer than when selling to the TPR; in other words, the TPR has no market share ($q_t=0$), and all customers choose the manufacturer channel. The manufacturer collects $q_m={\displaystyle \frac{r_m-\delta m}{\theta }}$ cores. We build the Stackelberg game modeled as follows:

$$\begin{array}{c}{\mathrm{\Pi }}_M(p,r_m,m)=(p-c_n)(\alpha -\beta p)+(f+\Delta m-r_m)q_m-\eta m^2\\ s.t. \left\{\begin{array}{c}{r}_m\ge \theta r_t-(\theta -1)m,r_m,m\in (0,1)\\ \max {\mathrm{\Pi }}_T(r_t)=(v-c_r^R+\rho m-r_t)q_t\\ r_t\in (0,1)\end{array}\right.\end{array}$$

Similar to the competition–coexist case above, the equilibrium results are given as follows, based on the Kuhn–Tucker condition: $p={\displaystyle \frac{\beta c_n+\alpha }{2\beta }}$, $m={\displaystyle \frac{f\left(\Delta -\delta \right)}{4\eta \theta -{\left(\Delta -\delta \right)}^2}}$, $r_m={\displaystyle \frac{f\left(\delta (\Delta -\delta )+2\eta \theta \right)}{4\eta \theta -{\left(\Delta -\delta \right)}^2}}$, $q_m={\displaystyle \frac{2f\eta }{4\eta \theta -{\left(\Delta -\delta \right)}^2}}$ (see Appendix E).

Proposition 4.

(competition case 3.2): Considering the effects of price and modularity, the manufacturer dominates the TPR and collects all the cores. The manufacturer’s modularity m, collection price r_m, and collection quantity q_m are increasing with remanufacturing cost-saving Δ (see Appendix E).

When the manufacturer provides a higher acquisition price and modularity ($r_m\ge \theta r_t-(\theta -1)m$), it can dominate the recovery market, becoming the sole collector and driving the TPR out of the recovery business, thereby forming a competitive monopoly. As the remanufacturing cost savings increase, the manufacturer will set higher collection prices to attract more returns from customers. This leads to greater savings in the production of new products and incentivizes further investment in higher product modularity.

5. Comparative Analysis

By comparing the optimal decisions of Models 1–3, we derive Proposition 5 and Proposition 6. The comparison of the parameters of the competition and cooperation models is given as follows:

Proposition 5.

Competition or cooperation (Case 2 vs. Case 3.1). The modularity and collection quantity are related as follows: when $C\ge C_2,m_2\ge m_{3.1},q_2\ge q_{3.1}$; when $C_1<C<C_2,$ $m_2>m_{3.1},$ $q_2<q_{3.1}$; when $C\le C_1,m_2\le m_{3.1},q_2\le q_{3.1}$ (see Appendix F).

$C=(v-c_r^R)\theta$ represents the combination of the TPR’s benefit from recycling $v-c_r^R$ multiplied by the channel advantage factor $\theta$, which denotes the TPR’s total recovery competitiveness. The modularity level and collection quantity for the manufacturer are higher under cooperation than competition when the TPR’s recovery competitiveness is sufficiently high ($C\ge C_2$). In other words, when the TPR is competitive in recovery, the manufacturer prefers to cooperate with the TPR to offset its competitive advantage. When the TPR’s recovery competitiveness is moderate ($C_1<C<C_2$), cooperation results in a higher level of modularity but fewer collected cores; moderate competition, on the other hand, increases the collection of used products. When the TPR is less competitive in recovery ($C\le C_1$), the manufacturer will rather invest in a higher level of modularity and obtain a higher collection volume in a competitive situation to expand its advantage. The manufacturer uses modularity as a strategic tool to compete with the TPR.

Proposition 6.

Cooperation monopoly or competition–monopoly (Case 2 vs. Case 3.2). The modularity is related as $m_2<m_{3.2}$ when $\theta <2$; the collection quantity is related as $q_2<q_{3.2}$ when $\theta <2$ (see Appendix G).

Under the cooperation strategy, the manufacturer outsources the collection business to the TPR, processes the collected cores, and monopolizes the recovery business through this partnership. Under the competition–monopoly strategy, the manufacturer deters the TPR’s entry by using appropriate pricing and modularity design, thereby monopolizing the recovery market through competition. In both recovery monopoly scenarios, the modularity level and collection quantity depend on the channel factor $\theta$. When $\theta \ge 2$, indicating that consumer preference for the TPR is double that for the manufacturer, cooperation results in higher modularity. When $1<\theta <2$, the manufacturer invests more in modularity and collection quantity under competition if the TPR’s collecting advantage is insufficient. The manufacturer can use product modularity to determine its monopoly position in the recovery market. When the TPR has a significant collecting advantage, the manufacturer can invest in modularity to establish a monopoly through cooperation. However, when the TPR’s collecting advantage is limited, the manufacturer can achieve a competitive monopoly by investing in higher modularity.

6. Numerical Study

6.1. Comparison of the Recovery Strategies

In this section, we explore the manufacturer’s recovery strategy selection, considering the effect of modularity. Drawing on the data from [67,68], we present numerical analyses of how various parameters influence the manufacturer’s optimal modularity investment, collection decisions, and profits. The parameters have been rigorously tested for concavity using the Hessian matrix and meet all the necessary constraints, ensuring that prices and outputs remain positive: $\alpha =5, \beta =2, c_r^R=0.2, c_n=1.5, \Delta =1, \rho =0.9, \theta =1.6, \delta =0.5, v=0.8, \eta =1, f=1$. These parameters serve solely as a foundation for simulation, aimed at developing an intuitive understanding of our model. A robustness check of the data concerning the effects of these parameters on supply chain members is provided in Appendix H.

Observation 1.

Table 4. Numerical results of three recovery strategies.

Recovery Strategies	${\mathit{p}}_{\mathit{i}}$	$\mathit{m}$	${\mathit{r}}_{\mathit{m}\mathit{i}}/\mathit{b}$	${\mathit{r}}_{\mathit{t}\mathit{i}}$	${\mathit{q}}_{\mathit{m}\mathit{i}}$	${\mathit{q}}_{\mathit{r}\mathit{i}}$	${\mathbf{\Pi }}_{\mathit{M}\mathit{i}}$	${\mathbf{\Pi }}_{\mathit{T}\mathit{i}}$	${\mathbf{\Pi }}_{\mathit{S}\mathit{i}}$
1. Centralized model	2	0.13	0.6	\	0.53	\	0.77	\	0.77
2. Cooperation	2	0.06	0.55	0.29	\	0.26	0.63	0.07	0.70
3.1 Competition–coexist	2	0.03	0.75	0.55	0.32	0.21	0.59	0.02	0.61
3.2 Competition–monopoly	2	0.08	0.56	\	0.33	\	0.66	\	0.66

compares price decisions, collection quantity, modularity investment, manufacturer, TPR, and supply chain profits under each recovery strategy.

(i) For the manufacturer: Our results indicate that the choice of recovery strategy should depend on the competitiveness of the TPR. When the TPR’s collecting advantage (i.e., consumer preference for the TPR) is weak (see Figure 3a), the manufacturer should invest in a higher level of modularity and choose to compete with the TPR to prevent its entry. Comparing the two competition models—Model 3.1 (in which the manufacturer and TPR coexist) and Model 3.2 (in which the manufacturer has a monopoly)—the manufacturer sets a higher modularity level (0.08) and a lower acquisition price (0.56) in Model 3.2, compared to a modularity level of 0.03 and an acquisition price of 0.75 in Model 3.1. In other words, when facing competition from the TPR, the manufacturer might strategically use higher modularity and avoid price wars to deter the TPR’s entry. As consumer preference for the TPR increases, the manufacturer’s collection quantity decreases, leading to reduced market share and lower profits for the manufacturer. Consequently, when the TPR’s competitiveness is strong—meaning that consumer preference for the TPR exceeds 2 (see Figure 3a)—the manufacturer chooses to cooperate with it to maximize profits.

(ii) For the TPR: The TPR achieves optimal profit under a cooperative recovery strategy (see Figure 3b). While the TPR might set a higher collection price in a competitive scenario compared to cooperation, the transfer price paid by the manufacturer under cooperation is higher than the TPR’s profits from recycling alone in competition. Additionally, the TPR may face intense competition from the manufacturer, which could prevent its entry into the recovery market (as seen in Model 3.2), resulting in a loss of recovery market share. For a TPR that lacks remanufacturing capabilities, cooperation with the manufacturer is always advantageous.

(iii) For the profit of the supply chain: As the leader of the supply chain, the manufacturer’s profits align closely with those of the entire supply chain, and a competitive monopoly generally results in higher supply chain profits compared to other strategies (see Figure 3c). However, as the TPR’s competitiveness increases, the manufacturer faces intense competition, leading it to opt for cooperation with the TPR. In this case, cooperation yields optimal supply chain profits. Encouraging cooperation in recovery between firms can enhance overall supply chain profitability in the recycling industry.

(iv) Total collection quantity: The collection quantity under competition is consistently higher than under cooperation (see Figure 3d). Our results indicate that the presence of the TPR increases the volume of used products collected, thereby enhancing the circulation of materials and resources, which has a positive environmental impact. When the manufacturer and the TPR coexist in the collection market, the total collection quantity can even exceed that of the centralized case. Therefore, to improve recovery efficiency, government legislation should encourage both the manufacturer and the TPR to engage in recovery activities. However, collection quantity decreases as consumer preference for the TPR increases. Different levels of channel preference $\theta$ can lead to varying preferred recovery strategies (see

Table 5. Recovery strategy selection when θ varies.

Recovery Strategies	Manufacturer’s Profit	TPR’s Profit	Environmental Impact: Collection Quantity	Economic Impact: Supply Chain Profits
Low $\theta$	3.2 Competition (manufacturer monopoly)	2 Cooperation	3.1 Competition (coexist)	3.2 Competition (manufacturer monopoly)
Moderate $\theta$	3.2 Competition (manufacturer monopoly)	2 Cooperation	3.1 Competition (coexist)	2 Cooperation
$\mathrm{High} \theta >2$	2 Cooperation	2 Cooperation	3.1 Competition (coexist)	2 Cooperation

).

6.2. Modularity’s Effect on Members’ Profits and Collection Quantity

Observation 2.

Modularity’s impact on the cost saving of remanufacturing $\Delta$ of the manufacturer is beneficial for the economic–supply chain profit and environmental–collection quantity factors. The situations are opposite to modularity’s impact on cost saving of the TPR $\rho$ and consumers’ sensitivity to product modularity $\delta$.

Product modularity is not always beneficial to the manufacturer and environment, although it allows more components to be reused. When the cost saving of remanufacturing $\Delta$ increases, the manufacturer yields a higher marginal benefit from recovery and is thus willing to invest in higher modularity (see Figure 4a and Figure 5a). Both the manufacturer and the TPR will set a higher collection price to collect more used products. By contrast, the effect of modularity on the cost savings of the TPR $\rho$ is harmful to supply chain profits and collection quantity (see Figure 4b and Figure 5b). However, increasing $\delta$ has limited influence on collection quantity and supply chain profits under all recovery strategies (see Figure 4c and Figure 5c). Considering modularity’s effect on supply chain members, the manufacturer and the supply chain obtain more profits and collection quantities only when the cost savings from the remanufacturing of cores are higher than the effects of the TPR and consumers.

7. Extensions

7.1. Extension 1: Consumers Prefer the Manufacturer Collecting Channel ($\theta <1$)

In the study above, it is assumed that consumers prefer the TPR channel because it is closer and more convenient for them. In this section, we explore consumer preferences for the manufacturer’s collection channel and examine how these preferences influence the manufacturer’s choice of recovery strategy. In practice, the TPR often operates as a non-certified collector, leading many consumers to prefer returning used products to the manufacturer’s official channel. Consumers are more likely to return their used products to the manufacturer due to a preference for offline stores [8,31,38]. As a result, consumer preference for the TPR over the manufacturer is likely to be less than 1. Similarly, we obtain the collection quantity for each channel (see Appendix I) as follows: $q_m=\left\{\begin{array}{c}{\displaystyle \frac{r_m-\theta r_t-(1-\theta )\delta m}{\theta (1-\theta )}},r_m>\theta r_t+(1-\theta )\delta m\\ 0,otherwise\end{array}\right.$, $q_t=\left\{\begin{array}{c}{\displaystyle \frac{r_t-r_m}{1-\theta }},r_m>\theta r_t+(1-\theta )\delta m\\ r_t-\delta m,otherwise\end{array}\right.$.

Based on the profit functions of each member (4) and (5), we derive the following equilibrium results: $p={\displaystyle \frac{\beta c_n+\alpha }{2\beta }}$, $m={\displaystyle \frac{\left(C-fh\right)\left(L-\Delta h\right)}{4\eta gh-{\left(L-\Delta h\right)}^2}}$, $r_m={\displaystyle \frac{\left(L\Delta +2\eta g\right)\left(C+fh\right)-L^{2}f-C{\Delta }^{2}h}{4\eta gh-{\left(L-\Delta h\right)}^2}}$, $r_t={\displaystyle \frac{\rho \left(C-fh\right)\left(L-\Delta h\right)+\left(4\eta gh-{\left(L-\Delta h\right)}^2\right)\left(v-c_r^R\right)+\left(L\Delta +2\eta g\right)\left(C+fh\right)-L^{2}f-C{\Delta }^{2}h}{2\left(4\eta gh-{\left(L-\Delta h\right)}^2\right)}}$, where $C=(v-c_r^R)\theta$, $L=\rho \theta +2(1-\theta )\delta$, $g=2\theta (1-\theta )$, $h=2-\theta$.

We set parameters, $\alpha =5, \beta =2, c_r^R=0.2, c_n=1.5, \Delta =1, \rho =0.9, \delta =0.5, v=1.1, \eta =1, f=0.9$, which are verified for concavity and satisfy our constraints. We exclude the case in which the manufacturer’s collection quantity is zero (competition–TPR monopoly, in Figure 6a–d), as it indicates that the manufacturer has no recovery market share and no remanufacturing cost savings from investing in modularity. Consequently, the manufacturer is unwilling to invest in modularity and only derives profits from selling new products.

When $\theta <1$, consumers have a higher preference for the manufacturer over the TPR. Increasing the channel factor $\theta$ narrows the preference gap between the TPR and the manufacturer. This reduction decreases the manufacturer’s collection strength, leading to lower investment in product modularity and decreased profits in competitive scenarios. From an academic perspective, our findings show that cooperation leads to higher manufacturer profits when the manufacturer has strong collection capabilities and a significant channel advantage (see Figure 6a). On the other hand, competition consistently drives higher collection quantities (see Figure 6d), aligning with earlier theoretical conclusions. From a business standpoint, these results suggest that when a manufacturer has a competitive edge in both remanufacturing and distribution channels, it should prioritize cooperation with TPRs to maximize profits. However, to optimize resource collection, the manufacturer may benefit from maintaining competitive pressure on TPRs. This dual approach balances economic and environmental goals.

7.2. Extension 2: Modularity’s Impact on Production Cost

In the previous section, we focused on the impact of modularity on remanufacturing, particularly in simplifying the replacement of used product components. In this section, we examine the impact of modularity on both manufacturing and remanufacturing operations, and whether it influences the choice of recovery strategy. In practice, product modularity enhances assembly efficiency for new products and offers significant cost-saving benefits, such as economies of scale and scope. By redesigning and producing specific components rather than entire products, the manufacturer can reduce development and production costs. This creates a link between production costs and modularity $c_n^{\prime }=c_n-\tau m$ [4], where $\tau$ denotes manufacturing cost savings through investing in product modularity. We derive the optimal decisions for each strategy (see Appendix J). The parameters we set, $\alpha =5, \beta =2, c_r^R=0.2, c_n=1.5, \Delta =1, \rho =0.9, \delta =0.5, v=0.8, \eta =1, f=1, \tau =0.5$, were verified for concavity and meet all constraints.

The results consistently showed no difference regarding the consequences presented earlier (see Figure 7a–d). Specifically, product modularity positively impacts the manufacturer through economies of scale under both recovery strategies. From an academic perspective, our results indicate that when consumer preference for TPR is low, a competitive monopoly leads to higher supply chain profits. However, when consumer preference is moderate to extremely high, cooperation becomes the more profitable strategy (see Figure 7c). Additionally, competition consistently maximizes collection quantities, regardless of consumer preference levels (see Figure 7d). From a business standpoint, these findings suggest that the manufacturer should align its recovery strategy with its primary objectives. If maximizing profits is the priority, cooperation with TPRs is optimal in cases of high consumer preference. However, if the focus is on maximizing collection quantities, a competitive approach may be more effective. The impact of product modularity on manufacturing costs does not change the manufacturer’s strategic decision-making. Notably, the impact of product modularity on manufacturing costs does not affect the manufacturer’s recovery decision-making, reinforcing the robustness of the main results.

8. Conclusions

In contrast to previous studies that focused on modularity’s impact on manufacturers, particularly in mass production and product variety, the trade-offs involving remanufacturing cost reduction, recovery competition, and consumer return utility among different supply chain members have been overlooked. This study provides insights for manufacturers facing competition from a third party, showing how they can strategically use modular product design policies to establish efficient recovery strategies. Considering product modularity’s effect on CLSC members, whether the manufacturer chooses to work or compete with the TPR and how to strategically invest in product modularity for each recovery strategy remain unclear. Our results offer several insights:

Question 1. What is the optimal modularity level for each recovery strategy? It is not always beneficial for the manufacturer to invest in higher product modularity under different recovery strategies. The impact of modularity on recovery strategy is influenced by consumers’ sensitivity to product modularity and the resulting cost savings in remanufacturing and recycling for both the manufacturer and the TPR. The manufacturer should consider the TPR’s capability and competitiveness when deciding on modular investments. The manufacturer may respond to external threats from the TPR by adjusting the degree of product modularity, as modular design impacts the third party’s recycling costs. Specifically, cooperation recovery shows greater modularity effectiveness when the TPR is highly competitive. Conversely, in scenarios in which the TPR’s competitiveness is low, competition recovery leads to higher modularity. For example, Fairphone’s modular design enables easier and more cost-effective repairs, upgrades, and reuse of components as spare parts. In its recovery process, Fairphone has partnered with the prominent third-party recycler Closing the Loop to address e-waste in Ghana by investing in highly modularized phones and maximizing the value of its modular design. As a result, 36% of the Fairphone 5’s total weight now consists of remanufactured materials, resulting in high recovery efficiency and increased profits.

Question 2. How does the TPR influence product modularity, manufacturer profits, and collection quantity? When the TPR actively participates in the recovery process, modularity reduces recovery costs for both the manufacturer and the TPR. This cost reduction enables the TPR to achieve a higher profit margin, encouraging increased collection activity through higher acquisition prices, and leading to a greater recovery market share for the TPR. As a result, the manufacturer’s profits may be negatively impacted, and modularity levels, along with overall supply chain profit, may not increase compared to a centralized scenario. The findings suggest that the manufacturer should dominate the recovery market by adopting either a cooperation or a competition monopoly strategy, enabling it to manage all core remanufacturing activities. However, the TPR’s involvement in recovery activities stimulates a collection quantity that even surpasses those of centralized cases, boosting the circulation of resources. The presence of the TPR also helps supply chains achieve better environmental regulation goals. In practice, many governments provide subsidies to TPRs and assist the formal sector in disassembling functional parts and valuable materials in an environmentally responsible manner. For instance, Germany mandates that used or end-of-life batteries should be collected by designated third-party recycling agencies, backed by government subsidies.

Question 3. Faced with a TPR, which recovery scenario yields greater benefits for both the manufacturer and the environment? When faced with a TPR, the manufacturer should select recovery strategies based on the TPR’s competitiveness, including recycling benefits and consumer channel preferences. For less competitive TPRs, the manufacturer should lean towards a competition–monopoly recovery strategy by investing heavily in high modularity, deterring TPR entry. Contrary to conventional wisdom, the manufacturer limits TPR recovery competitiveness by setting low product disassembly and modularity. Instead of engaging in a price war, the manufacturer can strategically use high modularity to leverage cost-saving benefits in remanufacturing and preempt TPRs. For instance, manufacturers can invest in modular electronic device design, provide repairability scores or instructions, and influence customer recovery behavior to enhance competitiveness against local informal sectors’ TPRs. Otherwise, severe competition may drive the manufacturer to cooperate with the TPR, outsourcing the collection business and simultaneously investing in high modularity for remanufacturing all used products. For powerful TPRs like TES, iFixit, and Caterpillar, the manufacturer may choose cooperation to achieve optimal profits and invest in high modularity. From an environmental perspective, competition between the manufacturer and the TPR for collection consistently leads to a higher overall collection quantity. The manufacturer must balance profits and collection targets when dealing with varying TPR capacities.

Table 6. Conclusions and managerial insights based on real-world practices.

Decisions	Conclusions	Managerial Insights
Modularity level	Cooperation recovery shows greater modularity effectiveness when the TPR is highly competitive.	The manufacturer can counter external threats from the TPR by modifying the level of product modularity to discourage the TPR’s entry. It can also collaborate with a powerful TPR by increasing its investment in product modularity, like Fairphone.
TPR’s impact	Competition between the manufacturer and TPR leads to lower profits, reduced modularity levels, and diminished overall supply chain profitability compared to a centralized scenario. However, it consistently results in higher collection quantities.	The presence of TPRs can enhance resource circulation. Governments can offer subsidies to TPRs and assist the formal sector in disassembling functional components.
Recovery strategy selection	For less competitive TPRs, the manufacturer should adopt a competitive monopoly recovery strategy by heavily investing in high modularity to deter TPR entry. Otherwise, intense competition may force the manufacturer to collaborate with the TPR.	Outsourcing the collection process and collaborating with established TPRs such as TES, iFixit, and Caterpillar. The manufacturer must strike a balance between profitability and collection targets when managing different TPR capacities.

presents a summary of the conclusions and insights.

9. Limitations of This Study

Our study invites future investigation in several directions. First, we consider a CLSC model composed of a manufacturer and a TPR to investigate their recovery strategy selection. It would be advisable to consider a more complicated supply chain, such as under the cooperation model, whereby the manufacturer authorizes the TPR to collect used products, while also opening a direct collecting channel. Second, we assume that product ownership is maintained by customers. Our paper highlights the need for further research considering different circular business models, in which product ownership is maintained by firms, including leasing and servicing. Third, future studies could examine the impact of government subsidies on the selection of recovery strategies.