Uncovering Missing Links in Global Value Chain Research – And Implications for Corporate Social Responsibility and International Business

Purpose—Amidst burgeoning attention for global value chains (GVCs) in international business (IB), this paper identifies a clear “missing link” in this literature, and discusses implications for research and corporate social responsibility (CSR) policy-making and implementation.<br>Design/methodology/approaches—The paper combines an overview of relevant literature from different (sub)disciplinary fields with insights from practitioner and expert interviews and secondary data.<br>Findings—Because IB GVC research stems from a focus on lead firms and their producing suppliers, it lacks attention for intermediary actors that may significantly impact the organization of production in general, and firms’ CSR commitments in particular. Import intermediaries are often “hidden” in GVCs. This article indicates the emergence of GVC parallelism with “frontstage” chains managed by lead firms and increasingly exposed to public scrutiny following calls for transparency and CSR, and “backstage” ones in which buyers and intermediaries operate more opaquely.<br>Practical implications—This study points at salient yet little known practices and actors that influence the organization of production and the implementation of CSR policies in various ways, and therefore offers ground for reflection on the design of proper supply chain and CSR policies.<br>Originality/value—This study exposes a hitherto neglected category of actors in GVCs and broader IB research and discusses implications, relevance and areas for further investigation. An illustrative example explicates the importance of carefully considering this “missing link”. The article emphasises the need for further study into ways in which both lead firms and intermediaries deal with contradicting demands of implementing CSR policies and offering competitive prices with short lead times.<br>


Introduction
The organization of global production and its consequences have long been studied in different (sub)disciplines, ranging from geography, political economy, sociology and development studies to business, while also receiving much attention from societal actors, policymakers and managers. This paper uses the term global value chains (GVCs), but we are very aware of related concepts to analyze the phenomenon of our interest, including global productions networks (GPNs), global commodity chains and global supply chains (Gereffi and Lee, 2012;Horner and Nadvi, 2018;Kano et al., 2020;Neilson et al., 2014). In the past decade, international business (IB) scholars have increasingly adopted GVC approaches to study multinational enterprises (MNEs) and better understand their strategies, governance structures and firm interrelationships (for recent work, see Fortanier et al., 2020;Kano, 2018;Sinkovics and Sinkovics, 2019;Strange and Humphrey, 2019). However, the field was relatively late compared to "other" social sciences to delve into GVCs (Kano et al., 2020).
The GVC literature has traditionally sought to shed light on the globalization of production because of an interest in the role of firms and questions of economic development, especially considering best ways for developing countries to insert themselves in the global economy and to realize economic and social upgrading. Since the 2000s, a GVC lens has also been taken to obtain insight into corporate social responsibility (CSR) initiatives by multinational buying firms (global brand retailers) and their effects on factories, plants and farms supplying to these firms Gereffi, 2014) as well as on vulnerable groups outside the formal economy (Narula, 2019). A wave of academic and practitioner publications has explicated the emergence and development of CSR standards (for an overview, see . It should be noted, however, that studies have thus far mostly focused on production facilities in the global south and the buying policies of well-known firms selling to final markets, often considered to be "lead firms" in GVC governance. In particular, when discussing the implementation of CSR policies, observers generally put the relationship between these lead MNEs and their (firsttier) suppliers as central. Figure 1 portrays a simplified overview, which can be linked to the framework of Coe et al. (2008) to indicate both complexities and asymmetries of power in the global organization of production.
Despite its obvious relevance for the study and practice of GVCs, we argue that IB research, because of the predominant attention for lead firms and their producing suppliers, has overlooked crucial intermediaries that may significantly impact the organization of production in general, and firms' CSR commitments in particular. There is emerging interest in other actors, both commercial and non-commercial, in lead firms' networks, as will be explained in the next section, but this work merely recognizes those situated outside the immediate value chains (Kano, 2018;Schrage and Gilbert, 2019; cf. Figure 1). Our article explains the importance of import intermediaries in GVCs and for global production and, accordingly, for the implementation of CSR initiatives, thus exposing a "missing link" in GVC research. Although extant literature provides arguments for why quality and process CPOIB concerns (Fernandez-Stark and Gereffi, 2019;Gayi and Tsowou, 2016), downward pressures on prices (Anner, 2018) or calls for increased supply chain transparency (Gardner et al., 2019) could diminish the role of these "hidden" actors in GVCs, we positbased on practitioner and expert interviews, secondary data and own experiencethat they are still significant in value chains, and that their obscurity is an important feature of their enduring economic success. In our view, scholars and policymakers should thus take into account the possibility that current GVCs might often resemble the structure shown in Figure 2.
Uncovering a hitherto neglected category of GVC intermediaries has implications for IB research and opens several areas for further investigation, as shown through an illustrative exampleclothing importers. We point at salient yet little known practices and actors that influence the organization of production and the implementation of CSR policies in various ways, and in this manner, offer ground for reflection on the design of proper GVC and CSR policies. Arguably, clothing production is used as a key exemplary sector in academic studies on the re-organization of intensive manufacturing production and its developmental consequences (Gereffi, 2014;Khattak et al., 2017;Narula, 2019), as well as on the effectiveness of CSR instruments (Braun-Munzinger, 2019;Kolk and Van Tulder, 2002;Locke, 2013;Schrage and Gilbert, 2019) it also seems to best represent the "global factory" model (Buckley et al., 2020;Johns et al., 2015). Our plea for a reconsideration of the state of the art on GVCs in light of the role of intermediaries in this sector will, therefore, resonate with much broader debates about GVC governance and CSR policy-making and implementation (Kolk, 2016;Rygh, 2020 2. On missing links in our understanding of global value chains and corporate social responsibility Scholars in IB, geography, development studies, sociology and political science have extensively analyzed the geographic dispersion and functional disintegration of production and trade in consumer goods and agricultural commodities Kano et al., 2020). The early 2000s already saw the first studies on a range of sectors, including electronics, footwear, toys and clothing, as well as coffee, cocoa, tea, palm oil, sugar and soy (Gereffi, 2014;Kaplinsky and Morris, 2000;Schmitz, 2006). Increasingly, GVC research also addresses environmental, human rights and labour implications of globally produced goods and discusses the interrelation between the mode of organizing production across borders in GVCs and the mode of governing CSR concerns in specific nodes of GVCs Bair and Palpacuer, 2015;Holzberg, 2019;LeBaron et al., 2017;Ponte, 2019). Despite these insights, however, one "component" has consistently received little attention. Researchers seem very well aware of activities, strategies and circumstances at both the final market side of GVCs (often related to brand/retailer policies and activities) and the "manufacturing end" (most often, farms and factories in low-wage countries), but only few studies consider or analyze what goes on between these nodes in terms of economic actors, activities, transactions, strategies and power. It appears that, cumulatively, the body of literature assumes that these factors are of much less importance in understanding how production is organized and CSR policies are affected.

Cross-country interacƟons
Host-country networks (usually in the 'South') Lower-Ɵer supplier firms Import intermediaries CPOIB Most recently, however, several scholars have argued for a broadening of our understanding of GVCs and a consideration of more actors, economic activities and interactions. Coe and Yeung (2019), for example, call for more attention to the intersections between GVCs (GPN 2.0 in their terminology) and "constituent outsides"broad overlapping domains encompassing the state, labour, finance, development and the environment. Schrage and Gilbert (2019) consider the private, social and public sectors, as well as the gap-creating and gap-addressing governance mechanisms in GVCs. Their analysis, however, remains within the confines of Figure 1, as the companies included are just the buying MNEs and suppliers and covers their interactions as well as those with other governmental and societal actors (non-governmental organizations, trade unions). Alexander (2018Alexander ( , 2019 distinguishes between horizontal and vertical relations among lead firms and lower-tier suppliers, also in value chains emerging from multiple local production networks, while Fontana and Egels-Zandén (2019) point at the relevance of inter-supplier dynamics in shaping GVCs and CSR policies. Linkages to the informal economy are discussed by Narula (2019) in light of the enforcement of higher labour standards by MNEs, and by Haugen (2018) in her analysis of so-called "petty commodities", in which unregistered businesses play a clear role.
The widening of our view of GVCs involves more attention to what we call "intermediaries" as third-party actors between firms or between firms and other parties (most often governmental bodies) that are significant for GVCs. As noted, in most cases, these involve the parties identified in Figure 1 already. Strambach and Surmeier (2018), for instance, discuss civil society organizations as information hubs and political actors as building standards that structure GVCs, Munir et al. (2018) particularly non-corporate intermediation in the governance of GPNs, and Kano (2018), in a similar vein, expert and non-profit organizations. An emerging strand of literature notes the significance of service providers such as auditors taking compliance and assurance tasks concerning CSR standards (LeBaron et al., 2017;Paiement, 2019) and consultants developing a range of activities in reshaping lead firms' relations to suppliers (Fransen and LeBaron, 2019;Gond and Brès, 2019). Closer to actually intermediating between firms and (lower-tier) suppliers is the work of Barrientos (2013) on labour contractors and their activities in "supplying" labour, including facilitating forced labour, and Soundararajan et al. (2018) on sourcing agents in production countries (Belavina and Girotra, 2012;Fung et al., 2007). Finally, taking a broader geographical perspective, Scholvin et al. (2019) identify a territorial dimension to the gateway role of some (local) economies (gateway cities) and businesses in integrating firms in or excluding them from GVCs in the organization of oil and gas production.
We welcome this broadening of our understanding of GVCs and their governance but would like to add one important link that has so far been missing, even though it has clearly come to the fore in our research and interactions with practitioners. In various supply chains, intermediary traders occupy positions in connecting global demand and supply that allow them to influence the place, price and size of orders and thereby indirectly conditions of production, including those affecting labour, health, safety and the environment. We notice, however, that their roles and activities take place in relative opacity, in two ways. First, these intermediaries mostly stay "under the radar" in analyses and policy discussions on the functioning of GVCs, in part as a result of their relatively quiet, publicly hardly visible business-to-business marketing approach. Second, in general, intermediaries are not keen on publicly sharing their supplier base, as they see it as key to their competitive advantage. Fierce competition at both the intermediary and client levels forces intermediaries to prevent their clients and/or other intermediaries to surpass them and go to their sources directly, as that would destroy their business model.
It should be noted that various studies of GVCs assume that three trends would work against the business of hidden intermediaries. First, some scholars emphasize the need to assure quality and quantity of products (Gayi and Tsowou, 2016;Fernandez-Stark and Gereffi, 2019), which would inform either a process of increasing vertical integration or increasing use of standardization and monitoring, decreasing intermediation and opacity. Second, the phenomenon of disclosure and transparency (Gardner et al., 2019), adopted by growing numbers of MNEs, means that hiding suppliers or intermediaries from view would be increasingly impossibleand even incentivize their eradication from transactions in line with a commitment to full-chain accountability and more effective governance of CSR issues. This argument conforms to observations about a "cascading compliance" approach in which non-compliant actors will be removed from MNEs' networks (and into the informal economy), accompanied by partial re-internalization and great automation (Narula, 2019). Third, and partly related, increasing pricing pressures in particular intensive manufacturing sectors would lead firms to cut out the "middlemen".
Based on own observations, insights from practitioner and expert interviews and secondary data, we instead argue that opacity may persist up until today in GVCs, and that economic actors benefit from their hidden sources. We accordingly propose that what could emerge are parallel value chains in industries: "frontstage" ones managed by lead firms and other actors so as to increasingly conform to societal and regulatory expectations of CSR and transparency, and "backstage" ones where both buyers and intermediaries can continue to operate in a situation of relative opacity. The next section provides a quintessential exampleimport intermediaries in the clothing industryfollowed by a discussion of key implications for research, practice and policy-making.

A quintessential example
As explained in the introduction, the clothing industry has received ample scholarly and public attention for several decades now and can be seen as the case in point of the "global factory": it links production in several regions of the world, and emerging economies have caught up with advanced economies, thus demonstrating convergence. A recent study by Buckley and colleagues (2020) found textiles, broadly defined, to be the only sector in which the number of employees in emerging economies exceeds those in advanced economiesmost notably, not only in manufacturing but also in knowledge-intensive (pre-and postfabrication) activitiesand where the income gap between these two country groupings is lowest. This situation is the outcome of a long process of internationalization over the past half-century, accompanied by the outsourcing of production for cost reduction in a "race to the bottom". Over the years, this has not only been the object of a wave of research but also of worldwide campaigns to combat the "sweatshops" and their detrimental social consequences, in a contestation over global brand retailers' CSR policies and practices (Adams, 2002;Anner et al., 2013;Bartley and Child, 2014;Chan, 2010;Connor, 2001;Gereffi, 1994;Perry et al., 2015;Tokatli, 2008).
Global brands/retailers have traditionally been seen as exerting substantial control in final markets, with a power imbalance resulting from globalization that inspired Gereffi (1994) and others to start analyzing these GVCs (or commodity value chains). Since this early research, in which the clothing industry was recognized as buyer-driven, the focus has shifted from "who" has the power, to "how" this power is exported and "why" it has taken this shape. Gereffi's (2005) theory has been important in this regard, as it introduced five new modes of coordination in GVCs: market transaction, relational value chain, modular value chain, captive network and hierarchy. His work suggests that the clothing industry may be moving away from a captive mode, in which end buyers largely determine CPOIB production and price, to a modular one characterized by relatively low power imbalances. However,  question whether this is the case, arguing that the clothing supply chain is too complex and untransparent for such a transition to be credible. By shifting orders between suppliers and/or locations, global retail chains can get the lowest prices and corner the largest part of the profitsas becomes visible when amounts actually paid to manufacturers are revealed (Anner et al., 2013;Appelbaum et al., 2005;Bair and Palpacuer, 2015;Kolk and Van Tulder, 2002). Moreover, using the case of Chinese fashion jewellery produced for Ghana, Haugen (2018) observes another limitation of Gereffi's (2005) typology, which is that it fails to include the governance of petty commodity chains, which rest upon informal economic networks and many unregistered small firms in production, trade as well as distribution.
Regardless of one's assessment on this point, there is an emerging interest for other business actors and activities in between the "traditional" clothing factories and the brands/ retailers, which delivers insights that help to better assess the relative balance of power in the GVC, also vis-à-vis labour and in terms of CSR. In the past decade, several observers have started to pay attention to one or more of the "in-between components", described with a variety of terms. Nevertheless, research has remained remarkably limited, considering what is, in our view, a persistent presence of intermediaries in the industry. As shown in Figure 2, we use the term "import intermediaries" (or importers) to denote the full array of possible services bridging the two extremes of the clothing GVC. Our characterization comes closest to the one of Vedel and Ellegaard (2013) in their study of 20 Danish clothing firms and the supply chain risk management of what they call "sourcing intermediaries", and of Dyer (2008, 2009) in their early work on domestic US "apparel import intermediaries". However, based on interviews with GVC actors in more countries (Western Europe and South East Asia) as well as own experience in the clothing industry, we perceive the potential range of activities that importers may take as broader than risk management, and as being much more international than just serving their same-country (domestic) retailers. At the same time, what importers actually do is as comprehensive (or restricted) as desired by the global brand retailers for which they work.
Importers can perform just one or many tasks for global retailers (see below) and often also involve other parties such as traders, agents or buying houses in (host) production countries or nearby locations or others in "non-production" (often their home) countries. The degree to which importers themselves engage in "outsourcing" depends on their size and geographical presence/spread and on cost and competitive strategies. The "between market and hierarchy" debate related to GVCs (Strange and Humphrey, 2019) thus has additional layers of complexity in the clothing industry and applies to retailers, manufacturers as well as intermediaries. While well-known retail brands from western countries in particular have reduced the use of "middlemen" as part of a re-internalization process, influenced by CSR concerns and standard-setting after Rana Plaza, our sources indicate that complete "inhouse" management of all collections for the full supply chain up to the factories is rare. Frequently mentioned is a "70-30" approach as a rough indicator of the state of play in the clothing industry as far as western retailers are concerned. This means that the overwhelming majority of the work is managed by the retailer (in what can be labelled as "direct outsourcing"), and the remainder through "indirect outsourcing", with more or less help from import intermediaries, which therefore continue to have a substantial place in the global clothing market. Obviously, the role of importers may be even larger in the GVCs of non-western retailers (e.g. those based in countries such as Russia or Brazil), but also much smaller or non-existent, although the latter is seen as relatively exceptional.
The key point of this article is that import intermediaries deserve more attention, as they form missing links in GVC research and in the public debate, while they are still relatively prominent in practice, with the clothing industry as a case in point. The problem is of course that finding data and obtaining information about "hidden" actors is difficult to impossible. To obtain some more insight into the presence of import intermediaries, we resorted to Orbis, a business data set that contains information on 120 million firms around the world, from which we extracted "real-time" figures on importers based on the NACE industry classification codes of firms. A search on the two most relevant NACE codes (4641, wholesale of textiles, and 4642, wholesale of clothing and footwear) yielded a total of over 140,000 firms worldwide, from 96 different countries.
Orbis is not without limitations, as it covers everything, from large comprehensive intermediaries to "single-task" firms. Moreover, the classification is rough with noisy data (e.g. we found some retail chains amongst the wholesalers, while importers known to us are missing), so these numbers should be treated with caution Dyer (2008, 2009) observed a range of issues already in the mid-2000s when trying to use US census data to identify domestic import intermediaries). Figure 3 thus gives a crude indication and contains a map with the countries that have more than 5,000 importers according to Orbis or that are listed in the top 20 based on turnover. The top countries in terms of number of firms are Brazil, China, Italy, USA, Russia, Japan, Spain and South Korea, followed by a range of further European countries; taken together, the European Union (EU) would surpass all others (also in turnover). Considering turnover, a somewhat different ranking appears with China, Japan, USA, South Korea and Great Britain as leading, and with remarkably high positionsrelatively speaking, compared to numbers of firmsfor Hong Kong, The Netherlands, South Africa and Denmark.
The Orbis data shed light on patterns, despite the limitations, and seem to reveal different roles that came to the fore in our industry contacts as well. For example, China and South Korea may "score" high given the proximity to production locations, with many firms offering specialized services to e.g. western importers that need "local" agents, buying houses or traders. Hong Kong likewise is a relatively easy point of access for several Indicative overview of importers' presence worldwide CPOIB facilities. The prominence of firms in other (non-production) countries may rather be explained from greater nearness to global brands' largest (home) markets. Importers take care of a variety of tasks to "unburden" buyers in their supply-demand management across countries, which might include responsibility for sourcing, quality control, price negotiations, on-time deliveries, monitoring and compliance, but even, for example, coming up with new innovative designs. While it has been assumed that importers would mostly be active in supply chains for discount, cost-competing market segments (Lane and Probert, 2009), they serve the whole range from hypermarkets and discount retail chains to high-end retail chains and clothing brands, especially originating from Western Europe and the USA, but in different ways.
Regarding hypermarkets (or "superstores"), which combine a supermarket and a department store and offer a wide range of products under one roof, clothing is just a small percentage of their total turnover. Consequently, they usually do not have their own design department, but rather outsource the design and production to importers and (directly or indirectly) agents in the Far East. In addition, although (even higherend) retail chains and clothing brands have their own design teams, they often still outsource a part of their collection to importers. In our work thus far, we found examples from all these categories, including high-fashion, high-end retailers, specialized kids and baby retailers, discounters and hypermarkets, located in Western Europe (Belgium, The Netherlands, the UK), Eastern Europe, Russia, and the USA. Usually, the "basic and basic-plus collection" is outsourced directly by the retailers, whereas the more specialized products can go via importers. Basic products are volume articles, and retail chains can buy a large enough volume on their own to achieve economies of scale. For the more specialized products, it is up to the intermediaries to reach economies of scale by combining orders.
Overall, and having asked managers of retail chains and importers themselves about the characteristics of import intermediaries and their function in clothing GVCs, we conclude that the following are most relevant: Serve as production/sourcing agent when brands/retailers develop their own collection and let importers source for the best prices, quality and lead time.
Offer private label and custom-made collections. In these cases, a collection is developed by the importer and sold with the brand/retailer's label in it. Importers have the ability to provide buyers with "fresh" new designs because of their cooperation with many brands/retailers, and they have valuable knowledge about the global industry and trends that their clients would not be able to obtain so easily otherwise. Take care of delivery, packaging and logistics for brands/retailers according to their wishes, with a diverse range of, e.g. packaging and labelling options, delivery in the harbour or at the shop and stock-keeping. Arrange beneficial payment conditions. Although buying via an importer should be more expensive, given that an additional economic actor is to profit, intermediation in this realm is seen as an important aspect of its value added for brands/retailers. When ordering directly in the Far East, a bank guarantee is required before order confirmation, as the factory needs cash flow to buy the fabric and start with the production. Importers usually allow for a payment term of 30 to 120 days after the invoice date. This means that a brand/retailer can sell the garment, earn money and only then pay the invoice.

Implications for global value chain governance, corporate social responsibility and international business
This section points to different ways in which importers are able to influence GVC governance and CSR practices and discusses implications for research and policy-making. Building on the insights from the clothing industry, which may be an extreme case but figures most prominently in many academic and practitioner publications, we argue that much more attention should be paid to import intermediaries and their effects. The subsections below suggest three means through which importers could affect their buyers and suppliers, as well as the general character of the GVC, and that have remained largely uncovered thus far. We suggest that importers induce less transparency, more complexity, possibilities for decoupling of standards for production from actual practices, price competition and blame shifting. At the same time, intermediaries also play a key role in unburdening retailers. As such, these intermediaries influence a number of aspects of GVC governance: they can determine from where and how products are sourced, the actual prices paid to suppliers and the degree to which buyers' commitments to CSR policies are implemented and suppliers actively address working conditions in line with these policies. Although we emphasize the role of import intermediaries concerning many crucial aspects of the organization of production and trade and its "CSR regime", at the same time, these firms themselves often seem to strategize and decide from a limited set of options and perceive pressure from various sides in shaping their activities towards suppliers, buyers and each other. Rather than a "free uninhibited agent", the importer is, therefore, better conceived as an actor making choices under significant constraints: it is not a determinant of economic and social conditions of production, but its activities are worth studying because of the potential contribution to and influence on such conditions. The most significant implication of brands' and retailers' sourcing via intermediaries is that it contributes to persistent opacity in GVCs, and, very likely, to a situation where one "frontstage" part of MNEs' supply chains is increasingly exposed to public scrutiny, following calls for transparency and CSR, while a "backstage" part remains hidden. Below, we propose starting premises based on our insights to provide inspiration for a more precise depiction of GVCs and for helping to assess which salient IB, CSR and governance dimensions are missing in the current debate when import intermediaries are left out.

Importers are junctions towards economies of scale
One of the characteristics (and value added) of importers is their network of suppliers (Masson et al., 2007), related to their role of searching and selecting the most appropriate manufacturers, and demand and supply consolidation. Import intermediaries collect and bundle orders from various retail clients to be able to reach economies of scale in production and transportation from Asia to final markets. In addition to "bulk work" for hypermarkets, specialized apparel retail chains usually outsource the more "difficult" products such as dresses, shoes, jackets with complex and detailed designs and that are usually sold in smaller quantities, and/or are not within brands' own specialismhere the "70/30 strategy", already mentioned above, appears common, at least among many European firms. This means that importers have a distinct role in combining orders and increasing quantities. Consequently, they gain buying power in the Far East and can compare quality, time and costs to ensure the best prices and lead times.
However, while this position enables importers to further their economic performance, it goes hand in hand with a fierce competition between suppliers to get the orders. In a situation of imbalance between quantity of orders and available suppliers, the result is a downward pressure on prices, or even a "race to the bottom" across the board. It leads CPOIB suppliers to lower margins, wages and quality; postpone investments; and/or increase overtime within factories . Therefore, we propose that this benchmarking promotes competition between factories to obtain orders. Low wages are, therefore, not only a consequence of competition between brands and clothing factoriesit is enhanced by import intermediation. A factory manager from India whom we interviewed expressed his dissatisfaction about the downward spiral for prices over the past 10-15 years and mentioned to be "always begging for orders". Despite inflation and wage increases, the average amount paid for an apparel product has gone down (Anner, 2018). Making money appears to have become more difficult for all participants in the clothing GVC.

Importers can reduce transparency and increase complexity
The clothing industry is complex, given that it is characterized by sourcing of short lifecycle products from a large number of suppliers across countries, uncertain demand, high product customization with an availability of many alternative designs and/or delivery options, as well as a range of knowledge bases, skills and competences that are required (Masson et al., 2007). On top come cultural, political, legal, administrative and financial (payment) differences. Complexity is obviously further increased through the addition of intermediaries that also engage in the GVC web. And, although a number of initiatives have been taken to map the production facilities of global brand retailers, consumers often do not know where clothes are produced. Adding extra links through import intermediaries tends to reduce transparency, if that existed already.
Hence, import intermediaries offer services that may increase the complexity and opacity of the clothing supply chain (Popp, 2000). Conversations with suppliers in India revealed that import intermediaries may contribute to opacity in at least two ways. First, as actors in between final buyers and producing suppliers, they can obscure or even manipulate the depiction of what buyers want from suppliers in terms of price, time and volume, so as to capture extra added value for themselves. Factory managers indicated the lack of transparency when they receive a claim, which an importer can send in case of, e.g. bad quality or late delivery. As the producer is not in direct contact with the retailer, they do not know if the claim is real -Neu et al. (2014) refer to "clawing back" payments in this regard. Therefore, trust is of utmost importance in the collaboration between suppliers and intermediaries.
Second, intermediaries can obscure efforts at CSR management and compliance with voluntary standards focused on labour and environmental conditions at producer facilities. CSR managers and sourcing managers in retail and at import intermediaries in The Netherlands indicated to us that importers shy away from sharing information about their supplier base and keep details about locations and conditions confidential (unless their MNE clients require otherwise). According to them, such information is of competitive value, and its disclosure will, therefore, endanger the bottom line. For this same reason, many intermediaries do not cooperate with a Dutch industrial commitment to enhance transparency in clothing supply chains in which brands/retailers do participate.

Importers and buying firms' decoupling
Academic literature has highlighted a tension between, on the one hand, increasing brand and retail commitments to CSR instruments in various industries, such as the monitoring of labour standards in clothing, and their drive to offer low-priced goods, on the other . In this regard, Xiao et al. (2019) have proposed that it is often left to the buying department staff to resolve this tension between various lead firm demands towards suppliers. Khan et al. (2019) argue that suppliers also need to deal with this tension, often to the detriment of their profits. Similarly, Sinkovics et al. (2016) describe how CSR commitments and price pressures lead suppliers into a double bind . We hold that lead firms, by outsourcing via importers while having CSR commitments, leave it up to the intermediaries to "fix" things. Import intermediaries then function as "decouplers" of these counteracting commitments (Bartley and Egels-Zandén, 2015).
As mentioned above, importers "fly under the radar" from the public and source for competitive prices with high volumes. The question is what they do once clients demand both competitive prices and a degree of compliance with CSR standards as intermediaries are brought into a challenging position. Our conversations with European import intermediary employees show that they engage in different kinds of activities to reconcile CSR and price demands. One of the respondents described that a certain retail chain always demanded that all products were produced in certified factories. As the size of its orders were rather small, and below suppliers' minimum order quantities, they needed to be combined with orders from other retailers that did not require that same certification. As a result, not all products were always produced in a certified factory, even though the importer did communicate that with the client. Afraid of losing the order, certification was promised, yet not lived up to. Interestingly, this could, in theory, also work the other way around in case many brands with sufficiently large orders would require certification, thus tilting the balance even for those not requiring or interested in certification. However, it should be noted that intermediaries also face the issue of multiple CSR standards  to which their global brand clients may have enrolled, with different types of requirements and associated costs and efforts.
Finally, many retailer and brand representatives who we spoke to still source a part of their orders via import intermediaries; next to this, some retailers exclusively rely on importers for their products. We note that this can also be a strategy to shift responsibility for wrongdoings elsewhere. When the orders of a European retailer or brand are placed in South or Southeast Asia via an importer, a part of the GVC is in its hands. If ethical transgressions are discovered, some of our respondents note, retailers do not consider it "their" responsibility what has occurred, but the importer's. The retailer or brand's CSR policy, thereby, in practice, only relates to what is directly sourced, which is most notable if a substantial part or all of the production is left to intermediaries to coordinate. In the latter case, importers are usually supposed to arrange enrolment, membership and registration, instead of the leading firm taking care of this themselves. This practice seems to mirror claims that buying firms tend to disregard certain suppliers when auditing their supply chains (Egels-Zandén, 2017).

Conclusions
As part of a range of initiatives seeking to deepen our insights of GVCs in the context of discussions of CSR, we have argued that research should focus on intermediary actors, which perform a range of tasks to unburden lead firms so that they can concentrate on their retail business. We made the following observations concerning the role of these "hidden" importers: first, importers and retailers are able to influence factory conditions by benchmarking for the lowest prices; second, importers reduce transparency and increase complexity; and third, global brand retailers outsource the tension between CSR commitments and buying prices by sourcing via importers, using them as "decouplers" and as actors to which they can shift their responsibilities. The mutual cohesion between these propositions lies in the "invisibility" of importers. This makes them interesting parties for shifting responsibilities, decoupling business and CSR commitments and benchmarking for lower prices. In our view, it is worth investigating further whether intermediaries such as CPOIB importers influence CSR governance and the governance of GVCs more broadly and how they affect IB.
Based on our insights thus far, we are assuming that intermediaries predominantly react to pressures from external actors rather than actively developing their own assessment of what is responsible behaviour through formalized CSR policies. At the same time, there is at least some decision-making power among these intermediaries, with their relative hiddenness in GVCs as key feature of their relevance. It is especially important to improve the understanding of how they deal with contradicting demands of implementing CSR policies, on the one hand, and aiming for competitive prices with short lead times, on the other. For the clothing industry as the totemic sector studied in CSR and GVC literature, our propositions point to understudied actors and mechanisms that seem to affect outcomes in this sector significantly. Of course, empirical research should further uncover whether these starting statements hold, within this sector and beyond. Such studies should bear in mind that the clothing industry combines a few distinctive features that, in our view, could make further investigation of intermediaries in this sector particularly salient.
First, clothing production does not require particular climate conditions or rare resources and is, therefore, not confined to specific geographic conditions like many other globally traded commodities. Many countries can, therefore, seek to insert themselves in the clothing industry by encouraging manufacturing, which implies that clothing production is more globalized, and global competition for manufacturing orders more intense than in other industries. Second, clothing production for most products and stages of production does not require skilled labour, further lowering barriers to entry for entrepreneurs. Third, clothing production has a high turnover of products due to fashion cycles. These factors, on the one hand, signal particularly intense price pressures and unpredictability in clothing production, and on the other hand, increase the likelihood of supply chain opacity as mediated by importers. As such, it would make sense to think of the role of intermediaries in clothing production as serving as an extreme case in methodological terms.
Having said that, we do observe parallels in other industries that are worth further investigation. In the palm oil industry, for example, there has been discussion of how trade intermediaries increase opacity in terms of the origin of palm oil plants, particularly when it concerns questions of the sustainability of farming conditions at these plants, or to what degree they apply proper labour standards (Gardner et al., 2019). In the coffee sector, complexities have been noted for, e.g. multinational retailers that need large quantities from a variety of (intermediary) sources, both conventional and certified sustainable coffee types, which must be kept separate (Kolk, 2013). Opacity due to intermediaries has also been a long-standing issue in smallholder cocoa production for global exports and therefore excludes the possibility of full traceability of cocoa origins for some chocolate products by brands and retailers that endorse labour and environmental standards; according to Fairtrade (2020), this applies to tea, sugar and fruit juices as well. Our conversations with experts of the Indonesian footwear industry furthermore imply that intermediary sourcing is a significant phenomenon in this sector as well. Nevertheless, as noted, within clothing, there is a wide variety already in the types of importers and the intermediary tasks they fulfil themselves or outsource to others, let alone when including other sectors, and considering the wide variety of countries, both home and host, of all the GVC actors involved.
To conclude, this article suggests further avenues for research as many questions are still unanswered, considering both firm and societal perspectives. For example, how widespread is the role of intermediaries across industries and countries, and how does using them fit leading firms' strategies, within specific GVCs and in terms of their CSR? How does digitalization affect the dynamics of market actors' involvement in GVCs? We suggest that digitalization may serve all of them, but to different degrees, especially for those consumer goods and commodities that continue to require substantive manual labour or are tied to specific locations, and that this deserves further study. In addition, and more specifically, how do global brand retailers monitor importers for compliance with CSR standards? What are the main differences in expectations in CSR between importers and global retailers? Do both types of actors experience the same pressure to invest in CSR? To what degree do retail buyers distinguish between direct sourcing relations and sourcing through intermediaries when they establish priorities for their CSR commitments in their GVCs? What is the experience of producing suppliers concerning working with importers versus working with global retail brands as buyers? And, to what extent does working with intermediaries in GVCs help or hinder efforts to move from mere symptom control to the alleviation of root causes in designing CSR? . There are many more questions that can be asked based on the ideas and insights included in this article, and that we hope will inspire further research to help shed further light on these issues and in this way uncover missing links in our current understanding of GVCs and hence of IB and CSR.