Vicious cycles that hinder value creation in closed loop supply chains: Experiences from the field

Abstract Closed loop supply chains can create value by product acquisition, recovery and reselling returned products and parts. Although advanced supply chain optimization models are available, the literature shows that value creation in closed loop supply chains is limited by too many constraints. Strategic success factors may relax these constraints. This study investigates business practices in the closed loop supply chain of four brand owners in capital goods. We find that strategic success factors may relax constraints but they themselves are also constrained, as multiple stakeholders are involved, each having different interests. In all four cases studies, the interaction between success factors and constraints leads to vicious cycles. Breaking these cycles proves to be difficult and requires integral thinking particularly among internal stakeholders of the brand-owners. Further research is needed to differentiate between different types of cycles, e.g. in a taxonomy, and different stakeholder viewpoints, both quantitative and qualitative.


Introduction
Closed Loop Supply Chains (CLSCs) create value in various ways, e.g. expanding markets by offering recovered products to customers with limited budgets (Heese et al., 2005;Purohit, 1992, Schenkel et al., 2015a or reducing a company's environmental footprint (e.g. Atasu et al., 2010;Krikke 2011;. Three processes are key in value creation, namely the return of used products and parts from the market, the product or part recovery, and the re-integration (sales/reuse) of recovered products and parts back into the market (Geyer & Jackson, 2004). These processes are often hampered by a variety of external and internal constraints such as costs for disassembly, uncertainty of timing, quality and quantity of returns (e.g. González-Torre et al., 2010;Inderfurth, 2005;White et al., 2003). In fact, recent studies show that CLSCs are far from reaching their full potential (Krikke et al., 2013;).
Today, optimization of supply chain processes across different echelons takes center stage, however most research is limited to the domain of forward supply chains, see e.g. Seyed Ashkan Hoseini Shekarabi, Abolfazl Gharaei and Mostafa Karimi (2018). In order to apply these models in closed loop supply chains, an empirical foundation is needed. This calls for explorative research into closed loop supply chains. Only then, objective functions, constraints and solution procedures can be tailored to the specific characteristics of closed loop supply chains The aim of this paper is to map qualitatively which constraints limit value creation in closed loop supply chains, and which success factors may help to relax the constraints. Actors in different echelons make various decisions which interact, unfortunately often ending up in viscous cycles. We feel that if we understand these dynamics, we may be able to intervene effectively and optimize the closed loop supply chain.
Vicious cycles are a concatenation of unwanted or ineffective events that arise from the interaction between CLSC key processes and their constraints .
Strategic factors that serve to relax constraints include product design standards, e.g. to ease disassembly (Krikke & Le Blanc, 2004), customer services, e.g. to enhance return volumes (Östlin et al., 2008), information management, which improves decision making on recovery, and business model changes to strengthen reintegration in the market (Schenkel et al., 2015a;Van Nunen & Zuidwijk, 2004). However, the implementation of these strategic factors themselves may be constrained too as multiple stakeholders are involved whose actions are also inter-related (Corbett & Klassen, 2006;Guide et al., 2003Meixell & Luoma, 2015. For

M A N U S C R I P T
A C C E P T E D ACCEPTED MANUSCRIPT example, product design standards involve many stakeholders whose return on investment is slow at best due to long life cycles (Ghazilla et al., 2015). Salespersons are not motivated to sell recovered products as they typically receive lower commissions for selling recovered products than for new ones (Guide et al., 2003). In sum, current literature provides sufficient evidence that stakeholders influence CLSC activities, constraints and strategic success factors that may alleviate constraints, simultaneously and in an interrelated fashion. The lagging implementation of CLSCs in business is due to the fact that this inter-relatedness leads to the emergence of vicious cycles, a phenomenon that has been under-investigated in CLSC research (Meixell & Luoma, 2015).
This study contributes to the literature by examining how constraints and strategic success factors mutually interact and how stakeholders may hamper CLSC value creation. Twentyseven in-depth interviews and eight site visits have been conducted in four case companies, active in the business to business, high capital electronic goods industry. Better understanding of vicious cycles may transform them into virtuous cycles that lead to improved CLSC value creation. The insights gained can also be used in further developing optimization models.

Literature review
A CLSC comprises the original forward supply chain and extends it with key reverse supply chain processes: acquisition, recovering and reselling returned products and parts (Guide and Van Wassenhove, 2009;Özkir and Bashgil, 2012). Used, recovered and new products constitute the installed base from which used products and parts are returned. Returned products and parts are recovered and subsequently sold and re-integrated into the installed base. These three processes re-enforce each other in creating value. CLSC management is defined as the "design, control and operations (of a system) to maximize value creation over the entire life cycle of a product with dynamic recovery of value from different types of return over time" (Guide and Van Wassenhove, 2009). Clearly, value creation means more than just optimizing profits, and should be rooted in the triple bottom line. However, this proves to be difficult to achieve in business practice.
Previous work on value creation in CLSCs (Schenkel et al., 2015a,b) MANUSCRIPT 2012;González-Torre et al., 2010;. Preliminary studies using surveys , content analyses and case studies (Lau & Wang, 2008;Shaharudin et al., 2015) found mixed evidence with regard to the relative importance and impact of internal or external constraints. Table 1  In additional studies, cause and effect relationships between constraints and succes factors in CLSC processes were investigated using Interpretative Structural Modelling

and the Grey Decision Making
Trial and Evaluation Laboratory (DEMATEL) approach .
For example, the absence of top-management commitment causes deficient strategic planning of CLSC processes which in turn reduces the financial resources dedicated to CLSC processes . Other studies focused on the optimization of strategies and control regarding business models (Bocken et al., 2016), remanufacturing (Gaur et al., 2017;Zhang et al., 2014) and hybrid manufacturing (Zanoni et al., 2006).  suggest that constraints, which pose specific challenges at each stage of the recovery process, can create inefficiencies for the entire CLSC. Given the interdependencies between CLSC processes, feedback loops can occur that influence the entire CLSC performance (Lehr et al., 2013). Research is needed that addresses the interaction and complexity of constraints in CLSC processes. This can be done by using dynamic models that include feedback loops (Huang et al., 2009;Lehr et al., 2013).
Finally, constraints to CLSC processes can originate from stakeholders and a company's response to the interests of stakeholder groups Govindan et al., 2016;. Managing CLSCs involves the interaction and integration of, both forward and reverse, supply chain stakeholders and stakeholders beyond organizational borders and the traditional value chain (Corbett & Klassen, 2006). Freeman (1984) defines a stakeholder as "any group of individuals that can affect or is affected by the achievement of an organization's objective" (Freeman, 1984, p. 46). To stay competitive in today's markets, companies not only have to respond to stakeholder requests, but also prioritize them based on their relevance for the company (Matos & Hall 2007;Mitchell et al., 1997;Olugu et al., 2010;Sarkis et al., 2010). Stakeholders can be distinguished into different groups such as primary M A N U S C R I P T A C C E P T E D ACCEPTED MANUSCRIPT versus secondary stakeholders (e.g. Álvarez-Gil et al., 2007; Chapter 3), or, as followed in this study, internal versus external stakeholders (e.g. Abdulrahman et al., 2012). Internal stakeholders are for example employees, departments, or top-management and external stakeholders can be customers, suppliers, governmental and non-governmental organizations, or even the natural environment. Both stakeholder groups can "promote or constrain the development of more effective (reverse logistics) processes" (Corbett & Klassen, 2006, p. 14).
Constraints for CLSC processes can be overcome by adapting and changing the design of processes and products in the forward and/or in the reverse supply chain (Geyer & Jackson, 2004). Schenkel et al. (2015a) identifies product design, customer services and business models for recovery as strategic factors. Intra-and inter organizational information sharing and stakeholder relationships also strengthen value creation, provided that the first strategic three factors are already present. Product design principles include design for disassembly, design for the environment, modularity, and upgradability facilitate disassembly, low-level recovery and upgradability of parts (Khor & Udin, 2013;Niinimäki and Hassi, 2011;Rashid et al., 2013, Mollenkopf et al., 2011, Jayraman, 2007. From a general perspective, implementing sustainability in design principles is influenced by different human factors, such as resistance against change and communication (Verhulst & Boks, 2012). Customer services, such as service contracts, after-sale-services, pay-per-use or leasing, enable manufacturers to keep track of their products and make targeted service or trade-in offers to customers with used products (Mont et al., 2006;Östlin et al., 2008). Business model aspects include trade-in activities, choice of recovery activities or decisions on recovering in-house or with a third party (Oezkir & Bashgil, 2012;Subramoniam et al., 2013;Toffel, 2004;Wells & Seitz, 2005). Thereby, CLSC business models should aim at integral value creation that is taking into account multiple stakeholders and the entire process instead of separate business functions (Guide et al., 2003). Information management and IT systems (e.g. product data management systems, installed base monitoring) are also used to obtain information from the installed base. This information can be used for forecasting product returns and timing or choosing recovery options for returned products (e.g. Huscroft et al., 2013;Van Nunen & Zuidwijk, 2004;Zhou et al., 2017). Information sharing is an important factor when dealing with resistance to change (Verhulst & Boks, 2012), hence it is an important enabler for alignment of interests and relationships between stakeholders (Gan, 2017(Gan, , Östlin et al., 2008. Abolfazl Gharaei and Seyed Hamid Reza Pasandideh (2017a,b) model a four level integrated supply chain. The aim of both papers is to optimize lot-sizing in each level such that the total M A N U S C R I P T

A C C E P T E D ACCEPTED MANUSCRIPT
cost of the chain is minimized while (stochastic) constraints such as limited procurement cost, limited space, and limited ordering cost are satisfied. Note that these types of optimization models presume perfect information availability.
In conclusion, current literature provides mixed evidence about the relative impact of internal or external constraints to CLSC activities as well as strategic success factors that may alleviate constraints. We argue that in practice vicious cycles emerge from the interrelatedness between constraints and strategic success factors. Our study will investigate this phenomenon that hitherto has only been scarcely addressed in CLSC research (Meixell & Luoma, 2015). Understanding vicious cycles, and how to turn them around, is crucial for further advancement of the field, both in qualitative and quantitative research as well as business practice.

Data collection approach
This study conducts explanatory multiple case study research (Eisenhardt, 1989;Miles & Huberman, 1994) focused on the CLSC of a product group of brand owners, as they are best equipped to take decisions in CLSCs.
The sample consists of four European brand owners of high capital goods who operate in global electronics and baggage handling equipment industries. These cases are suitable for studying constraints to CLSC for several reasons. First, companies in this industry are actively involved in CLSC processes (Talbot et al., 2007) and hence, probably encounter several constraints. Second, high-end capital goods are suitable for our study, because they entail complex, closed loop systems that incorporate many relevant aspects like high financial value, a long life cycle or service level agreements. Third, electric and electronic goods are subject to environmental regulations such as Waste Electrical and Electronic Equipment (WEEE),

Restriction of Hazardous Substances (RoHS), or Registration, Evaluation, Authorization and
restriction of Chemicals (REACH), which both trigger and constrain recovery processes.
Fourth, brand owners of high-capital goods tend to have close relationships with external stakeholders, for example, by offering service level agreements to their customers. These are common characteristics of all four cases.
The four companies differ in size and maturity of recovery activities, their business model, recovery activities, type of industry, and service activities (Table 2 and 3). This variety allows us to study the impact of CLSC-, organizational-and product characteristics on the value creation process, success factors and their constraints (Eisenhardt, 1989). Data were collected via interviews, participative observation in workshops and site visits (Table 4).
Secondary data was obtained from corporate internal research projects, environmental reports and internal documents and presentations obtained from interview respondents. Twenty-seven semi-structured interviews were held with managers in service, reverse logistics, sustainability, sales, R&D and product management. The interview questions were related to product characteristics, reverse supply chain processes, service offers, product recovery options, value creation and constraints to CLSC processes.

Data analysis approach
For each case, data was coded using NVivo 10 to identify relationships between different constructs. While coding, memos were written to capture emerging ideas (Miles & Huberman, 1994). Next, the effect of constraining factors on the three key processes was examined.
Codes were constantly reviewed, and constraints to CLSC processes identified, as well as constraints to implementing the strategic success factors (internal and external). Subsequently, findings on the constraints with four strategic success factors for various stakeholders were derived (Table 5).
Causal loop diagrams were developed to capture interactions between CLSC constraints, strategic success factors and stakeholder interests (Figures 1 and 2). Vicious cycles can be described as a sequence of reciprocal causes and effects leading to a worsening of the situation. Causal loop diagrams generate deeper insights about complex issues by investigating causes and effects in feedback processes (Kim, 1992;Moorecroft, 2007). Due to the explanatory power of interactions among actors and processes, this method is often used by researchers investigating complex issues (e.g. Georgiadis & Besiou, 2008;Lehr et al., 2013;Spengler & Schröter, 2003).
The relationships between the variables in a causal loop diagram can have a positive or negative polarity and encompass delays (Miles & Huberman, 1994;Moorecroft, 2007). A "+" means that if the cause increases, the effect increases too, while "-" means that if the cause increases, the effect decreases below what it would have been otherwise. A "delay" means that a given cause leads to effects after a time delay (Moorecroft, 2007, 39-40). A "-" in the centre of the loop indicates a balancing feedback loop. In a balancing loop, a change in one variable is counteracted in the course of the loop. In a re-enforcing loop, indicated by a "+", a change in one variable is re-enforced in the course of the loop (Moorecroft, 2007). Please note that the purpose of causal loop diagrams is to show interdependencies rather than inflows or outflows in a process.

Results
This section overviews the results regarding constraints in the case companies on the level of key processes (section 4.1) and the success factors (section 4.2).

Constraints to the return process and trade-in of used products and parts
The return process is influenced by several factors. A major internal factor is a company's trade-in rate. Cases 1 and 2 have dedicated business units, recovery facilities, and capacity for recovery. Their product portfolios consist of both new and recovered products, which requires continuous availability of returns inventory and, hence, trade-in of used products and parts.
One manager explained (case 1, company specific information was replaced by neutral terms between square brackets […], editorial amendments for readability are indicated by parentheses ()): In cases 3 and 4 trade-ins were occasionally used for offering customers discounts on new products when returning old ones. Hence, dedicated capacity, and inventory were lower in cases 3 and 4 than in cases 1 and 2. One operations manager (case 3) indicated that product and part recovery required a high inventory of used parts, and hence, high working capital costs. Cases 1 and 3 also faced limited capacity and utilization issues due to the uncertainty of timing, quantity and quality of product returns.
Manufacturers offered trade-in prices for returning their used products and parts. They calculated the trade-in price based on information about a product's residual (book) value, maintenance status and life cycle position. Many manufacturers struggled to determine a realistic residual value. They often calculated the residual value of used products and parts based on linear value depreciation after years of use and estimated recovery costs (cases 1, 4).
However, the market value of the product was sometimes higher (cases 1, 2). As manufacturers were not able to pay the same trade-in or acquisition prices as competitors or brokers, some products were sold elsewhere and not returned (cases 1, 2, 3). Furthermore, case 4 showed a high level of customization that inhibited a proper calculation of the residual value of used products. All four brand owners faced constraints to obtaining information about the maintenance status, the life cycle position and the rest value due to a low field traceability of products and parts. For example, low field traceability occurred in the case of products without service contracts. In case 3 products and customers were also not traceable anymore due to fusions and acquisitions of companies.
Complex CLSC design added to this problem. For example, case 2 outsourced the sales and service of products and parts to third parties, i.e. a broker or service company. As a result, the company had little direct contact with customers and therefore less control over the installed base. One manager (case 2, translated):

"Yes, but if you do a lot of sales via dealers […] as we do, then the barrier for the return flow is simply higher."
Having a complex global CLSC design in addition to worldwide installed bases resulted in high reverse logistic costs. Cases 3 and 4 had a worldwide -although much smaller -installed base with low return rates. Here, reverse logistics costs increased due to small economies of scales.
With regard to external constraints, some customers refused to return used products (cases 1, 2). They were hesitant to return parts and products because these contained confidential information.

Constraints to the recovery of used products and parts
Internal constraints to product and part recovery were mostly related to product characteristics. Managers from all cases admitted to face high costs, as end of life issues had not been considered during product development, especially for older products (cases 1, 2, 3).
Harvesting and recovering parts is only possible when products are easy to disassemble, modularly designed, and recovered parts can be used across product groups. This requires inter-release and inter-product compatibility, which is reduced by customization and limited standardization (case 4). All brand owners produced products with long life cycles, implying that products and parts that were returned from the market could not always be upgraded and were not always compatible with new releases (cases 1, 2, 4). Hence, recovered products may have become unattractive for customers.
On the external side, environmental directives such as RoHS and REACH restricted the use of certain substances and materials (cases 1, 2 Part recovery decreased the procurement of new and recovered parts from the supplier and, hence, affected supplier relationships. In case 3, suppliers reacted by increasing prices for new parts and refusing to collaborate. Additionally, third tier suppliers lowered new material prices when manufacturers recovered old materials and lowered their demand. These market price mechanisms increased recovery costs and made it economically unattractive to undertake recovery activities.

Constraints to the sales of recovered products and parts
The sales of recovered products and parts was hampered by internal constraints related to the original and secondary markets. For example, companies were careful with promoting recovered products and parts on the original market in order to reduce the risk of cannibalization of new sales (cases 1, 2). As a result, they focused on cascaded secondary M A N U S C R I P T A C C E P T E D ACCEPTED MANUSCRIPT markets. Another concern was that secondary markets might be too small to make recovery beneficial for the brand owner (cases 3, 4). One manager (case 2) explained how his company manages recovery and sales of used products and parts (translated): Additionally, as products have to be returned from the market, inspected and recovered, lead times for recovered product were longer than for newly built products (case 1).
Furthermore, some countries (e.g. China) impose import and export regulations which, for M A N U S C R I P T A C C E P T E D ACCEPTED MANUSCRIPT example, inhibited the import of recovered products and, hence, reduced the size of the secondary market (cases 1, 2).
<<<Table 5. Constraints to CLSC key processes>>> Figure 1 provides a modelled view of the dynamics of the factors presented in Table 5. For example, the more products and parts are produced, sold and placed in the installed base, the higher is the (potential) trade-in rate and return rate after customers' use (delay). As often old products are traded-in for new ones, a high trade-in rate in turn increases the sales of new products and parts. The trade-in rate is influenced by the installed base visibility, which refers to knowledge about the location, maintenance status, and life cycle of products. For example, a good installed base visibility indicates that trade-in offers can be made to customers with products that become end of life. As a result, the trade-in rate increases. These dynamics reenforce each other negatively, and a vicious circle arises, impeding the three key CLSC processes on an operational level.

<<<Figure 1. Re-enforcing CLSC key processes>>>
Based on prior literature (Östlin et al., 2008, Van Nunen & Zuidwijk, 2004, one may assume that strategic success factors (i.e. customer services, information management, product design and business models) relax operational constraints and ideally, break the vicious circle. For example, customer services affect the return rate and the sales of recovered machines.
Services such as leasing increase sales and oblige customers to return the used product after the leasing period. Information management may increase field traceability, which enables higher trade-in prices. The product design, such as design for disassembly, influences the recovery rate by facilitating the recovery of used products and parts. The business model focuses on internal constraints related to trade-in offer and constraints related to the reintegration. Figure 2 shows the relationships between the operational constraints and strategic success factors.

M A N U S C R I P T
A C C E P T E D ACCEPTED MANUSCRIPT <<<Figure 2. The impact of strategic success factors>>>

Constraints to the strategic success factors
The implementation of success factors is constrained itself in several ways (Table 6). These constraints concern corporate strategic choices on, for example, product design or customer service offerings, and are often rooted in conflicts of interest between stakeholders. Second, integrated data and information management remained challenging, because organizational departments source from different channels and process information in different databases. Complex information systems resulted, involving different databases to which, for example, the recovery department did not have access to.
Third, in our sample, the implementation of product design principles that support recovery was impeded by limited R&D budgets or short times to market that favour design requirements for new products. If part design and manufacturing was outsourced to suppliers, brand owners only specified the required functions of a part or product. Hence, they also had limited influence on low level design that would facilitate recovery activities. Finally, the corporate accounting system was seen as linear rather than circular, aiming at short-term profitability. Every department had to be profitable by itself, which impeded integral thinking among departments. Hence, the value that investments, e.g. in new service models or improved product design, can bring for several departments in the long term were not considered. For example, leasing models could benefit the sales department by increased sales and better customer services and the recovery department by a higher recovery rate.

M A N U S C R I P T A C C E P T E D ACCEPTED MANUSCRIPT
Sustainable and circular business models are difficult to develop, and some managers demanded more top-management commitment to change the business model, as well as stimulations from external parties such as the European Union. <<<Table 6. Constraints for the implementation of the strategic success factors>>>

Theoretical contribution
This paper analysed constraints to value creation in CLSCs at four brand owners in capital goods. Through modelling we mapped (internal and external) operational constraints in a causal loop diagram (Figure 1), which enabled us to identify vicious cycles that hamper value creation in CLSCs. Several insights could be distilled from the causal loop diagrams.
First, based on our results, it can be concluded that internal and not external constraints are most important in CLSCs. The implementation of strategic success factors mostly depends on M A N U S C R I P T

A C C E P T E D ACCEPTED MANUSCRIPT
internal rather than external stakeholders' interests. For example, projects for changing product design towards improved recovery opportunities require close contact between the R&D, recovery and financial department. External constraints affect manufacturers' CLSC processes by reducing the return rate and increasing costs. For example, a high market value of products and a global installed base cause high trade-in and reverse logistics costs.
Environmental regulations and customers' quality standards increase recovery costs as a smaller share of returned products can be reused. Hence, external constraints, which arise from demands of external stakeholders, will affect the internal stakeholders, and are internalized as costs and lower return rates in CLSCs.
In the literature there is no consensus about the relative importance of internal and external constraints to CLSC activities.  and Post and Altman (1994) find that internal issues need to be solved first, while González-Torre et al. (2010), Govindan et al. (2016) and Shaharudin et al. (2015) find that external constraints are of major importance.  suggests that inconsistencies in findings may relate to the fact that studies differ with respect to the size of the firms that are in the data set of each study. Our dataset contains both large and small firms, and by taking into account the emergence of vicious cycles, we can conclude that brand-owners should integrate internal departments first. Nevertheless, the effects of the firm size or business specific aspects on the relative importance of internal and external constraints could be further investigated in future research.
Second, the constraints themselves are inter-related. The three key processes discussed in section 4.1 (return, recovery, and re-integration of used products and parts) are interdependent and form a re-enforcing loop with the installed base and rate of products that are traded-in for recovery in a negative way. Constraints that affect one key process indirectly affect others.
For example, due to market uncertainty or customers' demand for service and warranty for older generations of products, manufacturers were hesitant to recover and re-sell used products and parts. Moreover, some products are not easy to disassemble or recover, which results in lower recovery rates and higher unit recovery costs. Manufacturers anticipate and weigh these costs against the benefits that they can yield from recovery. Depending on the expected benefits, they will offer a high or low trade-in value, which steers the return rate.
The constraints found in the four case studies and how they influence other factors ( Table 6) confirm some of the model variables and related interactions recognized and modelled in earlier research, e.g. phenomena such as uncontrollable disposal, product return rate, new product demand, new product sales (Georgiadis & Besiou, 2008;Lehr et al., 2013).

M A N U S C R I P T A C C E P T E D ACCEPTED MANUSCRIPT
Third, constraints can be relaxed by strategic success factors, but in turn conflicting interests -especially among internal stakeholders -limit their implementation. Relaxing constraints requires integral thinking among internal stakeholders, otherwise less value is created than is potentially possible.
Contrary to past research, we did not find lack of awareness , lack of know-how, or lack of training to be relevant for internal constraints González-Torre et al., 2010;Kapetanoupoulo & Tagaras, 2011;Post & Altman, 1994).
This may be due to the fact that our sample companies were already actively involved in

Limitations and further research
Limitations of this study give rise to recommendations for further research. We studied constraints to CLSC key processes and strategic success factors from the perspective of brand M A N U S C R I P T

A C C E P T E D ACCEPTED MANUSCRIPT
owners, while incorporating other stakeholders' views. A multi-perspective view of CLSC actors may lead to a supply chain view on adjusting CLSC processes. Moreover, it would be worthwhile to compare the results of this study with results from studies in other industries, outside the business to business context and electronic industry. The variables in our causal loop diagram are based on constraints deduced from all case studies. The focus of the analysis was not specifically to differentiate between constraints that were mentioned only once, or were mentioned several times. Hence, the diagrams do not provide information on the relative importance of each CLSC constraint on each CLSC key process. More research is needed on ways for breaking the vicious cycle and implementing the strategic success factors. This could be done by studying incentive management and functional integration between stakeholder groups in CLSCs (Guide et al., 2003). Thereby, the benefits could be investigated of a business model that focuses on integral value creation among departments versus individual value creation per department.
Furthermore, this study confirms the framing of ecological benefits and values within an economic perspective, in which long term profits are predominant, both in CLSC context (Schenkel et al., 2015a, b) as well as in broader sustainable supply chain context (Carter & Rogers, 2008). Further research on integrated value creation according to the triple bottom line (Elkington, 2018), specifically in CLSC context is recommended. When discussing investments for implementing the strategic success factors, the question of revenue sharing among CLSC stakeholder groups calls for more attention. For example, the party that recovers returned products and, hence, benefits from product design improvements might not be the one investing in it (Schenkel et al., 2015b). Quantitative studies using system dynamics or

M A N U S C R I P T
A C C E P T E D ACCEPTED MANUSCRIPT

Managerial implications
Our study provides several managerial recommendations for organizations and managers involved in CLSC processes.