Competing Reverse Channels’ Performance with Sustainable Recycle Innovation Input

First, this paper is most closely related to the CLSC system with remanufacturing. In this area, Reference [1] considered a manufacturer-led CLSC in a Stackelberg game, which has three options for product collection activities and found that the retailer is the most effective undertaker. Moreover, Reference [3] extended the two-stage supply chain collection model as explored in many of the above-reviewed studies to a three-tier supply chain coordination problem under different channel leadership. Since then, many scholars have compared these three single channels in pairs; for example, [4,5,6]. However, their works only focus on one period problem. Reference [7] presented a two-period closed-loop green supply chain model with a single manufacturer and a single retailer to investigate the impact of green innovation, marketing effort and collection rate of used products on the supply chain decisions. Some scholars proposed a dual distribution channel consisting of traditional retail channels and e-tail (internet) ([5,6,8]), a dual recovery channel hybrid manufacturing-remanufacturing production model [9], but the above researches focused on the collecting channels of the third-party collectors is less. In recent years, with the rise of online recycling service platforms, third-party collection channels have attracted more and more attention from scholars. For example, the authors of [5] dealt with a closed-loop supply chain with two dual channels—a forward dual-channel where a manufacturer sells a product to customers through traditional retail channel and e-tail (internet) channels, and a reverse dual-channel where the used items are collected for remanufacturing through the traditional third party logistics and e-tail channel. Reference [10] considered consumers might choose the online collection service provided by third-party platforms for convenience and explore the influence of consumer behavior on the competitive dual-collecting supply chain. With the spread of digitalization, a virtual logistics company plays a very important role, not only in a regular supply chain, but also in some emerging supply chain areas. For instance, Reference [11] study an implementation of a virtual logistics network for PCs with remanufacturing and [12] investigate the possibilities of establishing and operating virtual logistics centers in a cross-border logistics area.

Second, the authors of [13] extended the study of [1] to a setting where two retailers compete in the forward supply chain, and studied the coordination problem in the indirect collection model. Subsequently, many scholars ignited fierce discussions about the competition between closed-loop supply chain recycling channels and the choice of recycling channels. On the basis of a single recycling channel, the authors of [14] discussed the recycling model of retailers and third-party companies, and found that the prices of remanufactured products under the recycling channels of third-party companies were higher than the prices of products recovered by retailers. In comparison, the recycling channels of third-party companies are most affected by the marginal effect. Reference [15] considered that the retailer collects used products with a third-party and remanufactures used products at the same time. Reference [16] studied a supply chain in which the collecting of waste products is by the manufacturer and the third-party, and three individuals are leaders of a Stackelberg game. Reference [17] proposed a dual-recycling situation where retailers and third parties compete with each other. Research showed that when recycling competition is fierce, dual-recycling channels are superior to traditional single-recycling channels. On the basis of dual recycling channels, the authors of [18] set up two dual-collecting channel models: one that is the collecting of waste products by the manufacturer and the retailer, and another that is the collecting of waste products by the manufacturer and the third-party. Then, the authors of [19] found dual-collecting competition has significant influence on the supply chain. Reference [20] studied the two chains compete with each other in three competition structures: the Centralized Competition Game, the Hybrid Competition Game (including two cases), and the Decentralized Competition Game. Reference [21] studied the influence of collecting convenience on collecting rate in a dual collecting model. Reference [22] considered two collecting reverse supply chains consisting of a retailer and a manufacturer, who compete together by proposing more rewards to the customers. Competition between two channels of one chain infers to internal competition, and external competition that points out to competition among two supply chains.

With the rapid development of smart logistics, the recycling model of waste products has gradually shifted to the Internet Plus recycling, and the way of third-party recycling occupies an increasingly important position. Reference [10] considered consumers might choose the online collection service provided by third-party platforms for convenience, which also brings competition to other channels. To explore the influence of consumer behavior on the competitive dual-collecting supply chain. Reference [23] considered a closed-loop supply chain (CLSC) with a competitive recycling-market (competitive collectors) and product-market (new and remanufactured products). According to the above literature, such as [1,3], the third-party recycler is the most unfavorable recycling method, but with the support of the internet, the third-party recycler’s innovative investment in recycling work has greatly overcome the disadvantages of third-party recycling, not only in the external impact. It has increased consumer demand and improved recycling efficiency, and realized the unrealized value of waste products from recycled products [24]. Those problems we proposed above are also the research questions we would like to discuss in this particular study.

Finally, we discussed relevant literature on innovation inputs. Reference [25] mainly studied how the competition between the two reverse channels influence supply chain players’ optimal decisions and channel performance. In their work, however, they ignored the impact of the innovation input. Reference [26] examined the development and current state of supply chain innovation research in management and identified research gaps. Reference [2] emphasized imperfect and asymmetric information and technological and consumption externalities such failures that can be overcome (in part) by technological innovation. For example, the authors of [27] reported a simple, effective and sustainable method for the recovery of precious metals from end-of-life vehicle shredder residue and/or automobile shredder residue based on a hybrid ball-milling and microbubble froth flotation process. The authors of [28] used an innovative disassembly approach to identify the profitability of recycling such electronic components. Reference [29] also showed that the retailer remanufacturing after paying for the fixed technology authorization fees remanufacturing (FR) mode not only promotes the retailer to improve the product service level, but also enables the third-party to improve the recovery rate. Reference [9] first identified the root causes of the low return rate of cell phones and then developed a novel Radio Frequency Identification (RFID) based return channel to increase the recycling rate. A mathematical model is developed considering the cost of implementation and the design of the proposed RFID based recovery channel. Reference [30] contributed on waste product collection and remanufacturing by taking the intellectual property rights of the original manufacturer into consideration. Reference [31] found that during the pretreatment process, the failure to operate on high-value materials caused significant economic losses (from 2006 to 2014, computer and mobile phone losses exceeded 70 million dollars). This is also the necessity of this article to explore innovative investment in recycling channels. We have also found that in many documents, specific innovative technologies and innovative models are introduced into the recycling and remanufacturing process. Reference [32] provided an overview of the status of the waste electrical and electronic equipment (WEEE) collection system in China, combined the brick-and-mortar and the internet, and generated four WEEE collection systems. Moreover, the authors of [33] analyzed the effect of environmental subsidies on the incentives of investing in emission-reducing technologies in manufacturing amid the environmental concerns of consumers. Reference [8] said the performance of the supply chain can be improved by instigating either green innovation or marketing effort or both. For example, the authors of [34] focused on the impact of competition and consumers environmental awareness on key supply chain players. Also, they found that as consumers environmental awareness increases, retailers and manufacturers with superior eco-friendly operations will benefit. In addition, many emerging countries are also involved with their corresponding innovation management in their manufacturing industries. For instance, the authors of [35] use Data Envelopment Analysis to investigate the effectiveness of invested finances in the innovation of small- and medium-sized production enterprises in Slovakia. Furthermore, in [36], the authors developed a bibliometric analysis to investigate how the innovation factors benefit to a strategic manager with their given innovative activities.

In this section, we firstly discuss a basic closed-loop supply chain system that includes an integrated manufacturer and a third-party collector. Secondly, we consider the impact of the innovation input, and we construct the two models. One of the models includes an integrated manufacturer and a third-party collector with innovation investment. The other model is the hybrid model that includes an integrated manufacturer and two third-party collectors with and without innovation input. The integrated manufacturer produces the new product and sells it to the consumer directly. We assume that the manufacturer only recycles from third-party recycling channels to implement remanufacturing. The collector recycles from the market directly and sells the used products to the manufacturer. The two or three parties will all pursue optimal profit by making their decision or effort strategies under Stackelberg games. For example, Ai Hui Shou implement smart recycling bins in residential areas, schools, commercial squares and other places. Due to the popularity of smart phones and the background of the intelligent community, people are more likely to put used products into smart recycling bins and get certain variable remuneration. In the model, the manufacturer allocates the reverse channel responsibility to the collectors, and products are collected indirectly via the collectors. The CLSC structures under different business models are as follows.

In this reverse channel model, collected used products are transferred to the manufacturer by the collectors in return of a fixed per-unit buy-back payment, $b_i$. Subscript i takes the values of 1 and 2, denoting the collector 1 and collector 2, respectively. Given the retail price p and the buy-back payment $b_1$ and $b_2$, each collector $i(i=1,2)$ solves for the optimal innovation input e and the collection effort ${\tau }_i$, taking into consideration the competition from the other collectors. The manufacturer chooses the retail price p and the buy-back price b, taking into consideration the effect of these decisions on the strategic decisions of the collectors. Consistent with the extant literature ([7,8]), we assume that in that model, the manufacturer has sufficient channel power over the collectors to act as a Stackelberg leader, and the collectors compete in a Bertrand pricing game. We characterize the supply chain performance for each model in terms of the pricing decisions of the manufacturer and collectors, the collection effort, total supply chain profits, and the allocation of total profits between the manufacturer and the collectors.

In the rest of the paper, the following notation is used: $c_m$ denotes the unit cost of manufacturing a new product, and $c_r$ the unit cost of remanufacturing a returned product into a new one. ${\Pi }_j$ denotes the profit function for channel member j. Subscript j takes the values of M, C, $C_1$, and $C_2$, denoting the manufacturer, the collector, the collector 1, and collector 2, respectively. While the manufacturer takes back the used product from collector i at a price $b_i(i=1,2)$. $D(p,e)$ denotes the demand at market as a function of the selling price p and collector’s innovation input e. R denotes cost coefficient of recycling innovation input. Due to the vertical externality brought by the recycling innovation input, we assume that e plays a positive role in the demand of consumers. We denote the returned quantity as $\tau D(p,e)$, and $0<\tau <1$, where $\tau$ represents the reverse channel performance and it is a function of the product collection effort. $\tau$ also denotes the fraction of the current generation products remanufactured from the returned units [13]. We assume the market demand function as a deterministic linear function of the retail price p and the innovation input e. where $\alpha >0$, $\beta >0$, the constant $\alpha$ indicates the market potential while the coefficient $\beta$ represents the sensitivity of demand to price changes. $K>0$, and K represents the demand expansion effectiveness coefficient of the innovation input by the collector, also the coefficient K measures the marginal consumer demand with respect to the innovation input e. When $\beta \ge K$, it means that consumers are more sensitive to the sales price of the product than the innovative input coefficient provided by the collector. When $\beta <K$, it is contrary to the above statement. p represents the retail price of the unit determined by the manufacturer, and e represents the recycling innovation investment of the collector. $\alpha >\beta p$, that is, this is a rational market, and reasonable pricing controls the basis of the general market. This form of demand function has been adapted separately by the research community ([34,37,38]).

The average unit cost of manufacturing can hence be written as $c=c_m(1-{\tau }_i)+c_r{\tau }_i$ (c.f., [3,13]). We assume that $\alpha -\beta c_m>0$. The reverse channel performance is characterized by $\tau$, the return rate of used products from the customers. $\tau$ denotes the fraction of the current generation of products supplied from returned and thus remanufactured units, i.e., $0<\tau <1$. e denotes innovation input of collection activities by collectors which can be understood as not only the investment in the recycling technology (disassembly, refining, etc., and the technology of the intelligent recycling machine), but also the innovation in the recycling mode (the release of the intelligent recycling machine). There are now more than a dozen large-scale waste product recycling platforms, such as Ai Hui Shou, Hui Shou Bao (the website of Hui Shou Bao, https://www.huishoubao.com) and other Internet Plus recycling platforms. Compared with the recycling mode of a petty dealer, it breaks the space limitation and improves the enthusiasm of consumers to participate. People are more willing to recycle used products to recyclers in exchange for a certain variable return. For example, Ai Hui Shou gives different rewards to customers regarding the quality rating for their electronic products. The better the product, the higher the reward. Therefore, the investment in recycling innovation will also stimulate consumer demand for new products to a certain extent. By focusing only on promotional and innovation’s expenses to collect used products, we control for the strategic consumer behavior and solely focus on the channel behavior of the manufacturer and the collectors. Hence, consumers in our model react only to new product prices and innovation input.

The total cost of collection $C\left(\tau \right)$ can be characterized as a function of returned rate and unit innovation input of used products and is given by where A denotes the variable unit cost of collecting and handling a returned unit (e.g., a fixed payment given to the consumer who returns a used product). For simplicity of calculation, we set A as a constant. H and R are scaling parameters, and $H,R>0$. Notice that in order to characterize the diminishing return of investment with respect to ${\tau }_i$ and e, we use the quadratic cost structure. To ensure that the remanufacturing process is economically viable, we have $A<c_m-c_r$, $A<b$.

Several papers in the operations management literature, such as [22,25], addressed cost structure problems taking into consideration the effects on the self-reward of each channel and cross-rewards that are suggested to the customer by the competitors in the other channels.

In summary, the following assumptions are made for the model in this paper,

Now, we start with the simple supply chain described above, including one collector and one manufacturer. Figure 1 shows this basic supply chain system structure. As the basic model, we first study the vertically integrated manufacturer that directly sells its product to consumers and one collector without innovation input e.

The demand function of the basic system without innovation input (model B) is given as:

Many manufacturers play as leaders in the respective supply chains. They have strong market power to influence the decisions of other supply chain participants. As such, we consider that the integrated manufacturer has the power to determine the retail price and the transfer price. At first, the manufacturer determines the transfer price b and the selling price p. Afterward, the collector determines the return rate of used product $\tau$.

The profit function of the collector of the basic model (model B), ${\Pi }_C^B$ (the superscript B representing basic model without innovation input e and the subscript C representing the collector), is as follows,

Solving the stage two optimization problem first, the collector’s response function can be obtained as,

Further, the profit function of the manufacturer (denoted with a subscript M ) of model B is as follows, where $\Delta =c_m-c_r$.

The following proposition describes the findings in the basic model without innovation input e.

From the above analysis, it can be seen that the cost coefficient of recycling efforts H has a minimum value. Below the minimum value does not meet the actual situation, nor does it conform to the assumptions of this article. In other words, if the recycling cost factor H of recycling efforts is small, that is to say, the difficulty of recycling is small and the recycling rate may exceed 1. It is also reflected from the negative side that the recycling cost in real life is high and the recycling difficulty is great. On the other hand, as the transfer price rises, it will indirectly increase the difficulty of recovery.

Next, in Figure 2, we consider a similar basic supply chain structure consisting of a manufacturer and a collector, in which the collector invests in innovation input in the process of collection. As mentioned in the above assumptions, the demand function of the market can be expressed as, $D^{BE}(p,e)=\alpha -\beta p+Ke$.

The collector’s profit function of the model $BE$ can be written as,

From the concavity of the objective functions, the best response functions are determined by the simultaneous solution of first-order conditions for ${\tau }^{BE}$, $e^{BE}$.

The optimal product return rate is ${\tau }^{BE}(b,p)=\frac{R(b-A)(\alpha -\beta p)}{RH-K^2{(b-A)}^2}$. The optimal innovation input is $e^{BE}(b,p)=\frac{K{\left(b-A\right)}^2(\alpha -\beta p)}{RH-K^2{\left(b-A\right)}^2}$.

The manufacturer’s profit function can be written as,

From the above analysis, after the integration of innovative investment, it can be seen that the threshold of the cost coefficient of recycling efforts H has increased compared with that in the basic model B. In other words, when recyclers invest on the innovation, they also increase the entry barriers to the innovational recycling.

In the hybrid system, there is an integrated manufacturer and two collectors. The structure of the game played between the manufacturer and the collectors in the reverse channels is as follows. In this reverse channel model, collectors engage only in the recovering of the product while the integrated manufacturer participates in both manufacturing and distribution. The integrated manufacturer decides on the retail price of the product p, transfer price $b_1$ (per unit returned) reimbursed to the collector 1 and transfer price $b_2$ (per unit returned) reimbursed to the collector 2. We consider that one of the collectors with innovation input (collector 1) decides on the investment in product collection effort ${\tau }_1$ as well as the innovation input e of the recycling process. While the other one without innovation input (collector 2) determines the product collection effort ${\tau }_2$ by taking into account the retail price charged by the manufacturer and the competition from the other collector 1 with innovation input. Both collectors reimburse each consumer a fixed fee A, per unit returned. We assume that in the model, the manufacturer has sufficient channel power over the retailers to act as a Stackelberg leader, and the collectors compete in a Bertrand pricing game. Figure 3 shows this hybrid supply chain system structure.

When obsoleted products are returned to the collectors directly, the manufacturer has strong market power to influence the decisions of other supply chain members, so he firstly determines the transfer price $b_i(i=1,2)$, the subscript 1, 2 represents collector 1 and collector 2, and the selling price p. Secondly, the collectors determine the optimal return rates of used product ${\tau }_1$, ${\tau }_2$ and the optimal innovation input e. For example, the recycling mode of electronic equipment products such as mobile phones is similar to the recycling mode described above. Mobile phones can be recycled from the recycling channels of professional recyclers, from smart recycling bins, and from petty dealers.

The profit function of the collector 1 with innovation input e in the model H is ${\Pi }_{C_1}^H$(the superscript H representing the hybrid system and the subscript $C_1$ representing the collector 1).

Given p and $b_1$, the collector 1 solves the following profit function to determine the investment in product collection effort and the innovation input.

From the concavity of the objective functions, the best response functions are determined by the simultaneous solution of first-order conditions for ${\tau }_1^H$, $e^H$.

The optimal product return rate is ${\tau }_1^H\left(b_1,p\right)=\frac{R\left(b_1-A\right)(\alpha -\beta p)}{RH-K^2{\left(b_1-A\right)}^2}$. The optimal innovation input is $e^H\left(b_1,p\right)=\frac{K{\left(b_1-A\right)}^2(\alpha -\beta p)}{RH-K^2{\left(b_1-A\right)}^2}$.

Hence the collector 2 without innovation input optimizes the following problem (the subscript $C_2$ representing the collector 2).

From the concavity of the objective function, we can easily show that, ${\tau }_2^H\left(b_1,b_2,p\right)=\frac{R\left(b_2-A\right)(\alpha -\beta p)}{RH-K^2{\left(b_1-A\right)}^2}$.

Furthermore, the manufacturer’s profit function is given below,

$b_2$ is independent of p, $b_1$, while p and $b_1$ affect each other. With the above analytical results, we can derive Theorem 4.

Compared with the previous two models, the cost coefficient of the collection effort H has the highest threshold in model H, that is to say, the system has the highest difficulty. The hybrid system is also much more close to real life, which also reflects the high recycling cost and the difficulty of recycling in real life.

It is apparent that the optimal transfer price is bigger when the collector has innovation input. If the manufacturer assigns a large value of b, it would observe a high level of innovation input e and collection effort $\tau$ by the collector, but at the same time, manufacturer’s net savings from remanufacturing diminish. Hence, the manufacturer faces a trade off on the transfer price. Because of this, the manufacturer is more profitable in the hybrid channel.

We can get $p^{BE*}=\frac{\alpha }{\beta }-\frac{2M\left(\alpha -\beta c_m\right)}{4\beta M-4{\beta }^{2}RN}$, $p^{H*}=\frac{\alpha }{\beta }-\frac{2M\left(\alpha -\beta c_m\right)}{4\beta M-4{\beta }^{2}RN-4{\beta }^{2}R{(\Delta -A)}^2}$. Obviously, $p^{B*}>p^{BE*}$. As shown in Figure 4, among the three models, the retail price in model $BE$ is unexpectedly the highest. When the value of $\beta$ is small, $p^{BE*}>p^{H*}>p^{B*}$, which reflects that consumer demand is less sensitive to price, and consumers prefer to purchase the innovation invested products. At this time, both $p^{BE*}$ and $p^{H*}$ have a quicker increase due to the increased cost of innovation input and consumer preference, resulting in an additional increase in retail prices. While the integrated manufacturer considers maximizing consumer utility, it also maximizes his profits. As $\beta$ increases, $p^{BE*}>p^{B*}>p^{H*}$. The retail price of p is relatively small in the hybrid system because the manufacturer can choose to recycle part of the used raw materials from collector 2 with a lower transfer price $b_2$. We assume that the recycled raw materials recovered by the collectors 1, 2 are homogeneous. Therefore, the manufacturer in the hybrid recycling system can save more profit than that in the model $BE$, which is not only the positive impact of recycling innovation investment, but also the social welfare. Another reason is that the manufacturer’s profit in the hybrid model is higher, and some of the profits are derived from the scale effect. In general, the retail price p in the two system models with innovative inputs is not much different from the retail price p in the model without innovative inputs. It shows that innovation investment has not significantly increased retail prices, and can effectively improve channel recovery performance, while increasing consumer demand.

Interestingly, the total recovery rate of the mixed channel is higher than the basic centralized model, and the manufacturer’s recycling compensation has a direct impact on the collection effort $\tau$. When $b_1^{H*}>b_2^{H*}$, for the two recyclers in the hybrid system, collector 1 has a greater recycling effort. When $b_1^{H*}=b^{BE*}$, the collection rate of collector 1 is greater than the collection rate of the collector in the basic model, the main reason is that the retail price is p. It is precisely because of $p_1^{H*}<p^{BE*}$ that consumer demand is greater. For single-cycle issues, the greater the demand, the greater the scale of recycling generated.

Intuitively, manufacturers benefit from increased investment in recycling innovation. Analysis shows that innovation investment not only forces manufacturing to lower retail prices, which leads to higher demand, but also creates more profits for manufacturers by having mixed channels. However, it is counterintuitive that it can be best among integrated manufacturers in the hybrid system.

Such an edge increases when the amount of recyclers in the reverse channel increases, because the scale operated by the integrated manufacturer in model H is higher than that in model $BE$, and the cost savings of recycling and remanufacturing is increased due to collectors competition.

However, the relationship among ${\Pi }_C^{BE*}$, ${\Pi }_{C_2}^{H*}$ and ${\Pi }_C^{B*}$ is not clear. As mentioned in the assumption of [1], when $c_m-c_r>A$ and $b-A>0$ are satisfied, it is economically viable for recyclers. Reference [13] pointed out that for retailers, even $b_1^{H*}<A$, can compensate for losses by expanding demand. However, for the individual collectors in this article, only through cutting the price A and inflating the transfer price b can they obtain the residual value of the waste products, and expand profits by expanding the demand for recycling. The higher the $\Delta =c_m-c_r$, the higher the return rate $\tau$, the more profitable the recycler will be.

In the market competition environment, in order to gain more market share, some recyclers will improve their recycling efficiency by introducing technological innovations, increasing recycling efforts, and expanding recycling scale and other business strategies to gain competitive advantage.

Interestingly, the supply chain profit is the highest in the hybrid system with two collectors in comparison to the basic centralized channel vertical Stackelberg equilibrium. Thus, competition of collectors through the investment of innovation input is advantageous for the supply chain.

In this section, we present numerical simulations to illustrate the analytical results of those two models with the innovation input as shown above. We will illustrate the effects of the marginal demand coefficient K and the cost coefficient R on the analytical results in both closed-loop supply chains with innovation input successively.