CSR codes and the principal-agent problem in supply chains: four case studies

Journal of Cleaner Production 19 (2011) 885e894 Contents lists available at ScienceDirect Journal of Cleaner Production journal homepage: www.elsevier.com/locate/jclepro CSR codes and the principal-agent problem in supply chains: four case studies Francesco Ciliberti a, *, Job de Haan b, Gerard de Groot c, Pierpaolo Pontrandolfo a a Polytechnic of Bari, Department of Environmental Engineering and Sustainable Development (DIASS), via de Gasperi, 74123 Taranto, Italy Tilburg University, Department of Organization and Strategy, Tilburg, The Netherlands c Tilburg University, Development Research Institute (IVO), Tilburg, The Netherlands b a r t i c l e i n f o a b s t r a c t Article history: Received 18 November 2009 Received in revised form 3 September 2010 Accepted 6 September 2010 Available online 16 September 2010 The benefits of corporate social responsibility (CSR) affect the entire supply chains a firm participates in. However, not every firm is in a position to force the implementation of CSR in its supply chains as some, especially small and medium-sized enterprises (SMEs), lack the necessary power. Chain directors can implement it acting as a principal, whereas the other chain members can act as agents. In the principal-agent framework, two main problems occur due to information asymmetry: adverse selection and moral hazard. This paper examines how a code of conduct (i.e. Social Accountability 8000) can help address the principal-agent problem, for SMEs, between chain directors and partners. The research method involves four case studies on CSR practices as implemented by Italian and Dutch SMEs within their supply chains. Ó 2010 Elsevier Ltd. All rights reserved. Keywords: Corporate social responsibility Codes of conduct Agency theory Supply chain management Small- and medium-sized enterprises 1. Introduction Investigating the socially responsible behaviour of small- and medium-sized enterprises (SMEs) in their supply chains is relevant as SMEs account for the great majority of European Union (EU) enterprises and considerably contribute to employment (Spence, 2007). Most SMEs are directly managed by owners, are closely linked to regional business partners and to the local community, hence they often implement corporate social responsibility (CSR) practices even if they are not aware of that. However, they often lack resources to implement formal CSR systems (Lepoutre and Heene, 2006). In the paper we define CSR as “the voluntary integration, by companies, of social and environmental concerns in their commercial operations and in their relationships with interested parties” (Commission of the European Communities, 2001, p. 7). A socially responsible company can achieve several benefits, among which: reduced operating costs, enhancement of corporate image and reputation, and increased customer loyalty and sales (Blowfield and Murray, 2008). Such benefits affect the entire supply chains the firm participates in (Enderle, 2004). A socially responsible company should thus persuade supply chain partners (e.g. by selecting suppliers on the basis of their fulfilment of CSR requisites) * Corresponding author. Tel.: þ390994733265; fax: þ390994733304. E-mail address: cilibert@poliba.it (F. Ciliberti). 0959-6526/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.jclepro.2010.09.005 to implement CSR (or even force them, should persuasion be not effective), so as to reduce the risk to be blamed if supply chain partners’ actions might damage the company’s reputation. However, not every firm is in a position to do so as some (especially SMEs) lack the power to influence or force partners. Supply chains have to meet final customers’ demands and hence the processes carried out by chain members have to be coordinated. Two extreme coordination mechanisms can be considered: hierarchy, in which all processes are carried out within one vertically integrated firm; and market, wherein each process is carried out by a different firm. The choice among these two (and many other intermediate mechanisms) depends on the risk of opportunistic behaviour. If this risk is low, market would be preferable due to its transparency; otherwise hierarchy should be preferred as it is the strongest safeguard against opportunism (de Haan et al., 2003; Ciliberti et al., 2009). Obviously many companies (especially SMEs) are not able to adopt hierarchy, since they lack the resources and capabilities needed to vertically integrate supply chain activities. In some cases hybrid solutions were developed, integrating aspects of the two extreme coordination mechanisms. One example is the role defined by Gereffi (1999) and labelled as chain director by Sarkis and Talluri (2004). In a chain governed by market relationships one dominant partner can urge the other members to act in a specific way as if he were their manager. This dominant partner is usually better informed about the final customers’ demands (e.g. as a retailer or producer of very popular 886 F. Ciliberti et al. / Journal of Cleaner Production 19 (2011) 885e894 branded products) and uses such an information to persuade chain members to collaborate with him in order to gain benefits, in terms of e.g. higher prices, higher number of sold products or long-term contracts. This collaboration can thus be beneficial for the whole supply chain. However, chain members may not live up to these expectations (e.g. they can declare their commitment to fulfil the final customers’ requirements but do not live up to this promise or the chain director can promise to share benefits with committed chain members but does not live up to this expectation) which may have consequences for both the chain director and the other chain members, in terms of e.g. losing the contract or decreasing the business volumes. Consequently all partners would welcome an adequate safeguard against such opportunistic behaviour by one of them. This position resembles the principaleagent relationship in which one party (the principal) delegates work to another (the agent) who performs that work. The contract between the two parties should resolve both the agency and the risk problems (Eisenhardt, 1989a). Supply chain partners are driven by selfinterest, are prone to bounded rationality, have partially conflicting goals, and information is asymmetric (i.e. one partner is better informed than another). Because of these assumptions, agents can show moral hazard or adverse selection. Within the field of CSR, firms can apply numerous codes and management systems to show their commitment to CSR aspects and their fulfilment of demands coming from stakeholders. Examples of codes are: Business Social Compliance Initiative and Ethical Trading Initiative for labour conditions, the International Chamber of Commerce Guidance on Supply Chain Responsibility, Responsible Care by International Council of Chemical Associations for health, safety and environmental impact of chemical products and processes. Examples of management systems are: ISO14001 for environment, Social Accountability 8000 (SA8000) for working conditions and human rights, Occupational Health and Safety Assessment Series (OHSAS) 18001 for health and safety. Each of these codes and management systems covers one or more aspects of CSR. This paper examines how a specific code of conduct (i.e. SA 8000) can address the principaleagent problem, for SMEs, between chain directors and partners. The research method involves four case studies on CSR practices as implemented by Italian and Dutch SMEs within the framework of their supply chains. In the paper, SA8000 is used as a proxy for CSR as it focuses on invisible aspects of CSR (such as child labour) that have a strong impact on the public. The paper is organized as follows. The next Section reports a literature review on the principal-agent framework in the supply chain, focusing on SMEs and the SA8000 certification. Then the design of the research is shown and the methodology used in the research is described. The four cases are reported and then discussed. Finally, conclusions, limitations of the research, and insights for further research are provided. 2. The principal-agent framework in the supply chain 2.1. Agency theory One of the earliest works in the literature on agency theory is due to Jensen and Meckling (1976), who focused on ownermanaged companies. Other relevant contributions have been made, among others, by Fama (1980), Demsetz (1983), and Demsetz and Lehn (1985). The most significant problems that may arise from agency relationships are moral hazard and adverse selection. Both problems are caused by information asymmetry, i.e. when at least one party to a transaction has more or better information than the other(s). Moral hazard occurs when the ex post behaviour of the agent is not appropriate, i.e. the agent with more information about its actions has an incentive to behave not in line with the principals’ interest. On the contrary, in adverse selection models, ex ante information exchange is not appropriate, i.e. the principal is not informed about a certain characteristic of the agent. Agency theory assumes that individuals are self-interested creatures and addresses the problem of opportunism (Fama, 1980). In order to avoid opportunism, it is necessary to provide the agent with incentives to act in accordance with the principal’s interests. This can be done either by monitoring behaviour or rewarding outcomes (Eisenhardt, 1989a). Which of the two alternatives should be chosen, depends on their effectiveness and related costs. 2.2. Agency theory in the supply chain Applications of the agency theory to supply chains are rather scarce, yet add to the understanding of the many dilemmas that exist in the field. This could depend on the fact that the agency theory is traditionally used in dyads, whereas in a supply chain there are a number of agents serving one principal. Halldorsson et al. (2007) emphasize the potential goal incongruity in supply chains and the potential role of the agency theory in mitigating such a misalignment. When applying the principal-agent framework to the supply chain context, the chain director can be considered as the principal and the other members can be considered as agents. Supply chain partners have partly conflicting goals as they all aim at creating more value collectively, but might want to get a larger part, or want to make less effort in for the same reward as before. The overall results do not depend only on the effort of all partners, as uncertain, external conditions also contribute. The partners can measure the actual outcomes (in terms of e.g. costs and quality) of both the whole supply chain and each of them individually. However the individual effort is only known to the agent involved (i.e. asymmetric information). Some potential partners may promise more than they are able to produce (i.e. adverse selection), and/or may underperform and blame limited output to external conditions once accepted as chain members (i.e. moral hazard). In order to address these problems, the chain director needs a monitoring device to learn about the agents’ efforts. Codes of conduct can be used to this aim. Several papers addressed the use of codes of conduct in supply chains. Emmelhainz and Adams (1999) analyzed 27 US large firms in the apparel industry. They found that monitoring the employee rights performance of suppliers is a major problem. 22 of these codes stated that the actions of suppliers would be monitored, but relatively few specified how monitoring would take place. 16 of these 22 codes indicated that on-site visits could be made to verify compliance, and six of these companies required at least annual reporting by suppliers on their adherence to code provisions. Roberts (2003) analyzed the use of codes of conduct in three sectors: branded clothing and footwear, forest products, and branded confectionary. Comparing these supply chains revealed that four characteristics affect the propensity to introduce a code of conduct in a supply chain: (1) number of links between members demanding code of conduct and stage of supply chain under scrutiny; (2) diffuseness of stage of supply chain under scrutiny; (3) reputational vulnerability of different chain members; (4) power of different chain members. Salomone (2008) investigated the potential for integration between quality management systems (based on ISO9001), environmental management systems (based on ISO14001), occupational health and safety management systems (based on OHSAS18001), and social responsibility management systems (based on SA8000), by surveying a sample of Italian organizations. A significant number F. Ciliberti et al. / Journal of Cleaner Production 19 (2011) 885e894 of organizations stressed the difficulty to constantly monitoring the entire supply chain. Cramer (2008) developed a step-by-step plan to implement CSR in international product chains. Based on this plan, how a company can organize global chain responsibility depends on four main factors: (1) the diversity of the chain in which the company operates; (2) the extent of the ambition to reach social and environmental standards; (3) the complexity of the chain; and (4) the power of the company in that chain. 2.3. SA8000 and supply chain management An example of code of conduct that can applied to the supply chain is SA8000, which is a voluntary accountability standard developed in 1997 by Social Accountability International (SAI), an international non-profit human rights organization, and mainly based on the principles of core International Labour Organization (ILO) conventions, the United Nations (UN) Convention on the Rights of the Child, and the Universal Declaration of Human Rights (Leipziger, 2001). The standard defines eight principles related to working conditions and human rights: (1) child labour; (2) forced labour; (3) health and safety; (4) freedom of association and the right to collective bargaining; (5) discrimination; (6) disciplinary practices; (7) working hours; (8) compensation. The ninth issue of the standard concerns the establishment of a social management system. SA8000 is worldwide diffused. 2,103 facilities were certified at the end of 2009, in 63 different countries (Italy, India and China are the countries with the highest number of SA8000-certified companies) and 66 different industries. All together, these facilities employed over 1.2 million people.1 SA8000 was designed to fit into the ISO9001 and ISO14001 certification and performance audit process, because the management system elements in SA8000 follow the same logic as in ISO standards, and make it easier to implement the code at the operational level (Leipziger, 2001). Similarly to ISO9001 and ISO14001, SA8000 is formulated to allow a third-party certification body to audit and certify on behalf of SAI. A company can have a management system even if it is not externally certified. However thirdparty certification facilitates communication to customers. The third-party certification requires a public description of the behaviour of the audited firm and the management systems implemented. Once certified, firms are monitored to ensure that they live up to SA8000 norms. A SA8000 certification lasts for three years and there are surveillance audits every six months during this period. The company itself should conduct internal audits to ensure that the system is being maintained, at least on an annual basis (‘continuous monitoring’). Four basic types of out-of-pocket costs associated with certification are: 1. Taking corrective and preventive action in order to qualify for compliance. After this, an organization would seek verification of its compliance. 2. Preparing for the audit. 3. Going through an independent audit by a certifying body accredited by SAI. 4. Taking corrective actions to solve problems. The number of required audit days varies, but an average audit would consist of the following steps and related auditor days: 1 See http://www.saasaccreditation.org/certfacilitieslist.htm, last update: December 31st, 2009.     887 Document review: one or two days; Office audit: one or two auditors for two or three days; Witness audit: two auditors for twoethree days; Report writing: oneetwo days. This gives a total of 10e15 days per facility, while the costs typically range between $500 and $1500 per day (plus travel). There is also an application fee (depending on the number of employees) and an annual royalty. Over three years (the period of validity of a certificate), the total cost would range between $20,000 and $40,000 per company. Consequently many SMEs may face difficulties in achieving this standard (Rohitratana, 2002). By integrating SA8000 with other management systems (especially ISO9001 and ISO14001), companies (and especially SMEs) can achieve a cost reduction (Salomone, 2008). The principles of SA8000 tend to create a supply chain effect, being used as a tool to manage suppliers. The organizations may consider the benefit of SA8000 certification in terms of improved public perception of their activities (Rohitratana, 2002). A company seeking certification must implement proper steps to ensure that its suppliers comply with the standard. This may happen through personal visits, documentation review or requiring a second- or third-party audit. Auditors are not obliged to visit these suppliers, but to verify the steps taken by the company. As a first step, first-tier suppliers must provide the company with a written statement that expresses the supplier’s commitment to conform to SA8000 requirements. The company can include such a clause in its purchasing contracts and ask suppliers to provide a written commitment that they will require the same from their sub-suppliers. In addition, suppliers and sub-suppliers should identify the root causes of possible non-compliances with SA8000, repair them and implement a plan to avoid them in the future, as well as inform the customer company on the steps taken to put this plan in effect. In this way, the supplier can prove its reasonable effort to adhere to SA8000 and requires its sub-suppliers to do the same. From a buyer’s perspective, previous research shows that SA8000 provides more confidence with the suppliers and leads to a lower risk of being associated with human rights violations (Henkle, 2005). According to findings from Ciliberti et al. (2009), a social management system based on SA8000 can be used to transfer socially responsible behaviour along the supply chain, even when SMEs are involved. SA8000 spreads to supply chain actors, resulting in a domino-like effect (Miles and Munilla, 2004). This counts more in particular for second- and third-tier suppliers, especially in developing countries, where the countervailing powers of governments and civil society are weak and poverty prevails (Wolters, 2003). Monitoring activities on suppliers become more critical in these cases, since opportunism is more likely to occur due to a number of reasons: lack of strict regulation or reduced effort to enforce it by public governments and local public authorities; difficulties in retrieving information on suppliers; differences in culture and language; and lack of adequate tools to communicate with suppliers (Mamic, 2005; Ciliberti et al., 2008). Mueller et al. (2009) observe that only producers are certified, not their first- and second-tier suppliers, which means that certified companies do not admit full responsibility for the working conditions along their supply chain. SA8000 is site-centred, claiming that the suppliers of the location will be SA8000-certified in the future. A certification of suppliers is therefore intended. However, Mueller et al. (2009) highlight that the request on suppliers is not very precise and stipulates no clear period in time. From a supplier’s perspective, Stigzelius and Mark-Herbert (2009) identify the lack of support from buyers as an obstacle to the implementation of SA8000. Buyers do not share the costs incurred by suppliers to become compliant with the standard. In 888 F. Ciliberti et al. / Journal of Cleaner Production 19 (2011) 885e894 addition, there are no contracts to ensure that complying with standards represents a safe investment. Combining the standards’ requirements while also keeping attractive prices for the buyers is therefore difficult (Stigzelius and Mark-Herbert, 2009). This is in line with the theory on bonding costs, i.e. the costs a party bears to show the outside how they behave (Jensen and Meckling, 1976). Bonding costs have the same purpose as monitoring costs, since both costs are incurred to collect information on the behaviour of the agent. The main difference is that bonding costs are carried by the agent and monitoring costs are carried by the principal. In the future it is likely that SA8000 will be replaced by the ISO guidelines for CSR, named ISO26000, which are expected to be released by the end of 2010. ISO26000 offers guidance on socially responsible behaviour and possible actions. It does not contain requirements and therefore, in contrast to ISO management system standards, is not certifiable (ISO, 2009). 2.4. The case of SMEs The implementation of codes can be costly and time-consuming, and the related potential benefits might be not clearly identified or unevenly distributed among supply chain partners (Pedersen and Andersen, 2006). This is even more important when SMEs are involved in the chain. The application of agency theory to business relationships involving SMEs has been considered, among others, by Hand et al. (1982), Ang (1991), McMahon (2004). In smaller businesses, ownership and control are typically merged. Such coincidence should reduce incentives, among others, to make inefficient investments (Easterwood and Singer, 1981). According to Hand et al. (1982), the effects of agency relationships are most significant if the businesses are small. In smaller firms there is a greater level of uncertainty in the estimation of risk (Pettit and Singer, 1985), since for such firms the problem of information asymmetry is greater. This happens because SMEs generally do not have adequate resources, in terms of time, money, and personnel, to monitor the agents (Lepoutre and Heene, 2006). In particular, monitoring and bonding are likely to be more costly (Ang, 1991). The implementation of codes like SA8000 requires major investments that can be difficult to bear for SMEs (Bremer and Udovich, 2001; Welford and Frost, 2006). Reputation can emerge as a relevant way of securing commonality of interests when considering SMEs. However, reputation effects are less visible by end customers when the principal is an SME, compared to larger companies (Jenkins, 2004). Because SMEs rarely have a brand name to protect, they are unlikely to be affected by adverse effects of poor reputation on brand image. 3. Research design The goal of the paper is to show how SMEs can use codes of conduct like e.g. SA8000 to solve the principal-agent problem between chain directors and partners. The research question was articulated in four research propositions. Codes facilitate communication between supply chain partners that have indirect relationships. Manufacturers can thus inform final customers or the general public about certain features of their products. Vermeulen and Ras (2006) see as the challenge of sustainable supply chains that ‘both ends meet’, i.e. that the chain director understands the requirements of the final customer and translate those requirements into demands that have to be fulfilled by primary producers, e.g. farmers in developing countries. The features of the products at stake often reflect aspects of the processes from which they emerge, e.g. no child labour or no polluting emissions during the production process. Although producers can inform the other partners, these might not believe them. If a third independent party monitors whether the producer lives up to these claims and publishes about the results, this will enhance credibility. In third-party certification, an independent party sets the norms, audits suppliers, and informs buyers (Dewally and Ederington, 2006; Terlaak and King, 2006). Such parties might have special competences in evaluating compliance with codes, and can improve their reliability. As an alternative to third-party certification, firstand second-party certifications can be used. In first-party certification, a supplier sets its own norms and communicates them to its customers: brand names and labels are examples. In second-party certification, large buyers audit both potential and actual suppliers on whether the latter live up to norms set by the buyers themselves. Previous research shows that buyers prefer third-party certification, due to additional agency costs (i.e. the costs associated to possible unwanted inspections) in case of second-party certification (Hwang et al., 2006). By using codes of conduct that can rely on third-party certification, more information can be exchanged between the principal and the agents in a supply chain. This is a presumption for reducing the information asymmetry between supply chain partners, which contributes to solve the principal-agent problem between chain directors and partners. This leads to the definition of the first proposition. Proposition 1: Codes of conduct can improve the communication flows on intangible aspects of business between indirect partners within a supply chain and consequently reduce information asymmetry between the principal and the agents, whether in direct or in indirect relationships. The chain director as the principal can choose to use codes and urge the agents to do as well for two reasons. Firstly, codes can be a translation of the requirements by final customers. Secondly, the chain director may consider these codes to be right for their own sake. In the first case it should be easy to communicate this to the agents who can see the application of codes as a relevant investment in supply chain membership to achieve the related benefits. In the second case it may be more difficult to convince the agents as not economic but societal reasons are at stake. In this case the agents may comment that the investment is too high. However, the principal may refer to benefits from the chain membership. Codes are typically used as adds-on to contracts as the issues elaborated in such codes need no re-negotiation or incorporation anymore. In addition to this, a third-party can certify firms who want to use the codes and audit them on a regular basis once certified. Consequently, using codes adds hierarchical aspects to the market-based coordination between firms. These additional aspects can reduce the adverse selection faced by the chain director as the principal, as all chain members are screened by the certifying agency. Once certified, all members know that each of them is capable of living up to standards included in the code. Moral hazard problems can be reduced in a similar vein, as the certifying agency regularly audits all certified chain members, checking whether they still live up to the norms. Hence the transaction costs can be reduced because of simpler contracting, reduced risk opportunism, and related costs to prevent and correct if occurred. This leads to the definition of the following propositions. Proposition 2: Codes of conduct can solve the adverse selection problem both in the process of searching new suppliers (selected from certified firms) as well as when negotiating intangible issues in new contracts with current suppliers, when they accept certification. Proposition 3: Codes of conduct can solve moral hazard problems with respect to intangible aspects of business because monitoring improves communication on these issues, in particular in case of thirdparty monitoring of all parties involved. Bonding practices can be defined as activities that promote commitment in a relationship (Heide and John, 1988). Such activities are mainly trust-based and do not primarily rest on legal 889 F. Ciliberti et al. / Journal of Cleaner Production 19 (2011) 885e894 agreements or market power (Kaufmann and Carter, 2006). Bonding includes a variety of activities that are believed to contribute positively to increased commitment in a relationship: for example, developing personal ties between supply chain partners, developing common company cultures, building incentive systems, time spent together to solve third-party problems (Heide and John, 1988). Previous studies have found that higher degrees of bonding practices correspond to lower degrees of perceived opportunism (Rindfleisch and Heide, 1997). In addition, transparency between supply chain partners positively impacts the formation of bonds between the partners. If the principal can evaluate the agent’s performance precisely, then the chances are high that trust and commitment will gradually increase. Additionally, the agent is responsible, at least partially, for transparency in a supply chain with SA8000 certification, as the third-party requires the agents to carry out self-assessments or provide annual report or other information. When the lines of communication are open and the information flow is timely and correct, bonds as perceived by the principal can grow steadily (Kaufmann and Carter, 2006) and moral hazard can be further reduced. Bonding activities have investment-like characteristics, and at certain points in time extra-costs may be better than lower costs because the principal thereby improves the value of the agent (Jensen and Meckling, 1976). This leads to the definition of the following proposition: Proposition 4: Agents can further reduce moral hazard problems when they apply bonding practices towards their partners, such as the principal, to support monitoring activities by a third-party. 4. Research methodology Since the focus of the paper was to examine how a CSR code can help addressing the principal-agent problem in supply chain relationship, the analysis of the selected companies was aimed at studying how they manage CSR in relationships with suppliers. We found that the most critical aspects of CSR in such relationships deal with intangible aspects such as labour conditions at suppliers’ sites. We thus focused on SA8000 as a proxy for CSR as it deals with such aspects. Cases were selected on the basis of a replication logic (Yin, 2003), aiming for as much diversity as possible among the cases in terms of size, industry, country, and position in the supply chain. Although SA8000 is applied in many countries and industries, it is predominantly present in one country, i.e. Italy, and in one industry, i.e. textile and garment. The selection of the case studies starts from this observation, by choosing a garment producer from Italy. A case study from another country, the Netherlands, where SA8000 is hardly present and consequently far less known, contrasts the Italian experiences. Two Italian cases from other industries complement the selection. One case is also in the consumer product sector, but it expresses far less lifestyle as the fashionable garment sector does. The other one is from a businessto-business market in the construction industry. The richness of such different cases provides both literal and theoretical replication (Yin, 2003). The data collection aimed for triangulation as different methods such as interviews and observation were used for primary data, and internal and external documents and Web sites were used for secondary data. In each of the firms several lengthy semi-structured interviews were conducted with senior staff (Yin, 2003). Wherever possible, plant and shop visits were conducted, both for conducting the interviews and for direct observation purposes. Visits to firm sites were made in teams to have complementary insights and enhance confidence in the findings (Eisenhardt, 1989b). To corroborate information deriving from the interviews and increase construct validity, we also analyzed all the firm’s documents relevant to the research objective (Table 1). The interviews and observations were transcribed and summarized, together with the documents, and the results were fed back to the key persons for their consent on correctness of the data. Then these reports were coded according to the literature review on certification and codes in supply chain management. After this indepth analysis of single cases, the cross-case analysis was carried out. In searching for cross-case patterns we selected categories, then looked for within-group similarities coupled with inter-group differences. To address inter-coder reliability, each of the authors read the transcriptions and the documents separately so as to develop an independent point of view on each case. Then the authors separately coded the data, and identified concepts and categories. Finally they described the path followed when analyzing the data. Concepts and categories were afterwards compared among the authors. When evaluations by the authors were conflicting, a discussion among the discordant authors was conducted until a final agreement was reached. As we followed a theory-guided case study approach, a theoretical framework combining observations on supply chain management, agency theory and CSR was developed and is presented here as a checklist: 1. We consider a supply chain as a set of a chain director, acting as a principal, and the other partners as agents acting on his behalf but also guided by self-interest and with partly conflicting goals. The chain director wants to improve its competitive position by including intangibles in its products/ services. This increases agency problems. Table 1 Sources of evidence. Company Country Product Role of the informant Analyzed documents Monnalisa Italy Clothes for children Cora Kemperman Chicom The Netherlands Italy Women’s designer clothing Consumer products Person in charge of the social management system and the social report Person in charge of product design Person in charge of operations Person in charge of the joint quality and social management system Chimica Edile Italy Pre-mixed building materials Social report SA8000 report Company mission statement Periodic reports on suppliers SA8000 report Periodic report on suppliers’ compliance with SA8000 Suppliers’ self-assessment questionnaire Joint quality and social system management handbook Code of conduct Suppliers’ self-assessment questionnaire Company policy Person in charge of the social management system 890 F. Ciliberti et al. / Journal of Cleaner Production 19 (2011) 885e894 2. Conflicting goals cause two problems: adverse selection and moral hazard. This stands not only for direct but even more for indirect partners in the chain. Rewards are not sufficient to solve these problems as the contracts in which these would materialize are incomplete (in case of direct partners) or absent (in case of indirect partners), in particular for invisible aspects of CSR. Therefore the introduction of extra measures such as codes of conduct is required. Existing suppliers can be stimulated to implement the code by threatening to replace them or by improving financial and/or intangible rewards. 3. Codes of conduct might reduce adverse selection as partners are examined thoroughly. Codes can prescribe either behaviour (e.g. SA8000) or conditions to be met. In addition to this a thirdparty can audit the certified organization following a prescribed procedure capable of providing a convincing ‘snapshot’. All parties involved (i.e. the initiator of the standard, the certifying agent, the principal, the agents, and the public) should be able to rely on this to make the standard effective. Hence, these audits could solve the moral hazard problem by means of additional communication mechanisms provided by the code, in particular when third-party monitoring is involved. 4. In addition to this, the certifying organization may require additional material (e.g. annual reports) from the agents to further reduce agency problems on intangible aspects of CSR. Bonding practices can support monitoring activities, when riskaverse agents provide risk-neutral principals with information to show their commitment to CSR and the fulfilment of requirements posed by chain directors. the basis of several factors: the typology of production; the dimension of supplier companies; the location of these companies; the result of previous inspections. In 2006 the company conducted 12 inspections on suppliers. Three of them were conducted on suppliers in developing countries; two of these inspections were conducted by a third-party and revealed non-compliances. On the basis of the results of the inspections, non-compliant suppliers started to conform to CSR requirements by using a tool similar to SA8000, named WRAP (Worldwide Responsible Apparel Production2). The firm would prefer to conduct direct inspections in order to be more aware of the possible problems that suppliers can experience as to CSR. However the firms’ managers are not able to conduct them very often due to time shortage. The firm states that inspections at suppliers’ have a yearly frequency, on average, and generally last about an hour. This is clearly a point of weakness for the company when monitoring suppliers. Working conditions at suppliers’ sites are good in terms of health and safety issues, at least compared to local standards. Monnalisa acknowledges not knowing all the regulations applied in countries where suppliers are located. The firm tries to stimulate suppliers’ awareness of CSR, for example by inviting suppliers to Italy to observe the way Monnalisa operates. On such occasions several firm documents, such as the social report prepared by Monnalisa, are given to them. As a consequence of the existing cultural differences along the chain, Monnalisa had to modify the self-evaluation checklist prepared for Italian suppliers when monitoring Chinese partners. A few years ago the firm decided to break up the relationship with a supplier that showed critical noncompliances (i.e. child labour). 5. Description of the cases 5.2. Cora Kemperman In this Section we present four cases of firms that apply SA8000 along their supply chain. For each firm, we briefly present the profile, then we characterize the relationships with suppliers (especially in developing countries), and finally we show how SA8000 is used to solve the principal-agent problem in the supply chain. Cora Kemperman is a Dutch firm of women’s designer clothing founded in 1994 by Cora Kemperman and her partner, Gloria Kok. Cora Kemperman designs their collections at the firm’s headquarters in a small village in the countryside in the Netherlands, and Gloria Kok (responsible for operations and commerce) organizes the manufacturing and distribution thereof. The firm owns a retail chain of nine exclusive shops in the major cities of the Netherlands (e.g. Amsterdam and Utrecht) and Belgium (e.g. Antwerp and Brussels). The product designs are exclusive yet bearable in everyday-life. The firm products are in the middle price range. Production is completely outsourced to European and Indian suppliers. One supplier in Mumbai manufactures woven garments, and another one in the Tiripur cluster produces knitted apparel. CSR is an order winner for the firm, because the final customers are willing to pay a price premium for the products and are aware of the firms’ CSR policy, which is reflected in the concern for the wellbeing of workers in the production facilities of their business partners and also in the use of ecologically sound materials. In order to be able to observe labour conditions, Cora Kemperman wants its partners to produce in their own facilities rather than outsource to an army of cottage workers. The firm has signed the Fair Trade Charter. Indian suppliers are SA8000-certified. Cora Kemperman cooperates with local NGOs in India on one hand to ensure that partners live up to the expectations and on the other hand to select social projects to be sponsored. In 1998 Gloria Kok and Cora Kemperman founded The Amma Foundation, to which they donate a part of the profit every year. The foundation supports a project that aims, among others, at: giving medical care for all employees and their families; teaching all employees to read and write, especially women; setting up a fund (managed by the 5.1. Monnalisa Monnalisa was founded in 1968 as a one-person-company, became a limited company in the 1980s and a joint-stock company in 1991. The firm is located in Central Italy. The founder and his wife, who designs the products, own the majority of the company shares. Currently Monnalisa is the core firm of a group of six companies. Products, i.e. clothes for children, are upscale-targeted with a good value-for-money and unique characteristics. Products are sold in 41 countries in the world. Customers include boutiques, shops, and large-scale retail trade. Monnalisa outsources most of the production activities to firms located in China, India, Tunisia, and Turkey. In those countries Monnalisa also sells its products. A license agreement is defined with the Chinese supplier. Suppliers in developing countries are all medium-sized. Monnalisa procures from them a relevant share of its total supplies. The percentage of production sold by suppliers to Monnalisa is low compared to their total production. Monnalisa obtained ISO9001 certification in 1999 and SA8000 certification in 2002. The company defines CSR as “being aware of the role played by the company within the community and being responsible for this role in the community”. CSR is an order qualifier for the company when selecting suppliers, i.e. Monnalisa expects a minimum level of CSR performance from suppliers. Monnalisa plans and conducts inspections at suppliers’ sites both directly or relying on a third-party. External audits are scheduled on 2 URL: http://www.wrapcompliance.org/, last access: June 14th, 2010. F. Ciliberti et al. / Journal of Cleaner Production 19 (2011) 885e894 workers themselves) from which the workers can borrow money inexpensive for large expenditure items such as buying a plot of land or building a small personal home. Gloria Kok visits the suppliers several times each year. Some stipulations in the SA8000 standard clearly reflect Western values, and turned out to be incompatible with the situation ‘on the ground’ in India. As a consequence, the adoption of the standard was not always effective in improving labour conditions and workers’ welfare, and sometimes even had a counterproductive effect. An example of the latter concerns the standard’s requirement that all workers have to be given fixed contracts. In this case, it lowered the well-being of a special group in the textile chain, i.e. tailors. By tradition, the profession of tailor is an independent trade in India and tailors wandered around freely in search of good business. In compliance with SA8000, they are now bound to the Cora Kemperman company by contract, disrupting their social life as they can no longer have long holidays (to participate in religious and social activities) or negotiate better individual contracts. 5.3. Chicom Chicom was founded in 1981 as a limited company, and became a joint-stock company in 1994. The firm sells consumer products, e.g. aluminium wraps for food, barbecue products, latex gloves for domestic use, and moth-repellent products (also with a private label). Currently it is part of a group of companies and is located in a village in Central Italy. The main customer is a large-scale retail trader, which is SA8000-certified since 2000. For such a customer, CSR is an order qualifier because all of its major suppliers have to comply with SA8000 requirements. All products are procured from small-sized and exclusive external suppliers. Latex gloves are procured from two companies in Malaysia and coal from three companies in Croatia. The firm has achieved an integrated ISO9001-SA8000 certification in 2001 and has a written code of ethics that mostly deals with SA8000 issues. Chicom asks suppliers to sign its code of ethics and sends them a self-assessment questionnaire to check the respect of SA8000 principles. On the basis of the procedures of the joint quality and social management system, Chicom also writes a periodic report on suppliers’ compliance with SA8000 principles. In such a report, inspections to suppliers are also planned. For example, the latest report showed some non-compliances with SA8000 principles on monitoring second-tier suppliers. To further check non-compliant suppliers, an inspection to their production sites was also planned in the short term. The relationship with one Malaysian supplier is very recent (less than one year), so Chicom is not able to provide information on it (e.g. it does not know if the supplier manufactures in-house, as Croatian suppliers do, or only markets the product). Because of the scarce information, such a supplier is defined as ‘at risk’. The relationships with Croatian suppliers are older and Chicom has carried out direct inspections. The long-standing Malaysian partner has been periodically visited both by a third-party (on behalf of the large-scale retail trader) and by the firm’s top management. Chicom finds it difficult to retrieve information on employees’ wage and working hours in supplier companies. Several years ago, the firm broke up the commercial relationship with a coal supplier in Nigeria, due to the difficulty to check the working conditions and the respect of human rights. In other cases, when non-compliances were lighter, Chicom tried to formally explain to suppliers the relevance of the problem. 5.4. Chimica Edile Chimica Edile was founded in 1968 as a commercial structure (with a different name), and in 1981 created a research division. 891 Currently it is a family-managed limited company with a single partner. It is located in a village in Central Italy. Chimica Edile sells pre-mixed building materials (e.g. plaster, restoration materials and demolition mortars), which are environmentally friendly since they are based on natural elements (e.g. lime and its by-products). Products belong to a niche market and are upscale-targeted. Both suppliers and customers are large companies. Chimica Edile has a stake in an Egyptian firm and in an Argentinean firm; the latter supplies all raw materials. No second-tier suppliers exist. Production started two years ago in Egypt and last year in Argentina. Both the partners produce and sell under license, have the same organizational characteristics as Chimica Edile, and focus on a small market niche. Corporate representatives tied by familiar links to the firm’s top management are present both in Egypt and Argentina. The firm has achieved ISO9001 certification in 1995, SA8000 certification in 2000, and ISO14001 certification in 2006. The firm policy explicitly focuses on the respect of workers’ rights and the appreciation of human resources, and aims at defining fair working conditions. CSR is an order qualifier for the company, since customers expect that the environmental impact of products sold is better than competitors’. Every year Chimica Edile sends its suppliers two self-assessment questionnaires so as to check their compliance with SA8000 principles. Standards related to labour conditions in partner companies are good because Chimica Edile uses the Italian law as a reference. On the initiative of Chimica Edile, several practices (e.g. provision of benefits, such as vaccinations, medical examinations, Christmas bonuses, and respect of local customs, like prayer or tea moments) are adopted in subsidiaries to enhance relationships with employees. Employees are satisfied to work within the partner companies: none of them has ever left the companies. 6. Discussion As indicated before, codes facilitate communication between supply chain partners that have indirect relationships. Although the supply chains analyzed in the cases are rather simple and consist of only a few stages (see Table 2), indirect relationships exist. In most cases the final customers are the first- or second-tier customers of the focal firm. With respect to the supply side, this was more or less the same as in two cases there were no second-tier suppliers, whereas in the other cases information on them was not available. In case of indirect relationships, contracts between two chain members are not sufficient and additional supply chain management tools are needed. In such cases it is important that chain directors exist to ensure that ‘both ends meet’ (Vermeulen and Ras, 2006). In three cases a chain director can be identified, either the focal firm or its main customer (in the case of Chicom). Monnalisa and Cora Kemperman can act as chain directors despite their size, because of their ability to design products that fit perfectly with the demands of the market niches served, thus constructing buyerdriven chains. Although in the case of Chicom the focal firm is not the chain director, it conveys the demands to suppliers on behalf of the chain director, because the latter delegates the monitoring duties to its first-tier suppliers. Such suppliers have thus to be sure that secondtier suppliers live up to the requirements if they want to keep supplying the chain director. In the case of Chimica Edile, the focal firm lacks bargaining power towards its customers, which are larger and can choose from a wide range of suppliers. Yet the firm has high impact on its suppliers because of the family relationships with their top managers. Consequently, in all four chains a principal (the focal firm in the case of Chimica Edile and the chain director in the other three cases) can be distinguished. The other companies are the agents, who have to live up to the principals’ requirements 892 F. Ciliberti et al. / Journal of Cleaner Production 19 (2011) 885e894 Table 2 Supply chain structure in the analyzed companies. Issues Companies Monnalisa Cora Kemperman Chicom Chimica Edile Final customers Parents buying on behalf of children Design Design and production control of garment Medium-sized outsourcees of garment production abroad No information Consumers buying niche products in large retailers Price Trading consumer goods House-owners Order winners Mission Ladies wearing medium-priced design garment Design Vertically integrated retailer of garment Outsourcee of garment production abroad No information Not existent (primary raw materials) Large retail chain and independent retailers (export) Final customers Not existent (not allowed by the focal company) Final customers Mainly one large retail chain Construction firms Not existent Final customers Not existent Focal firm Chain director Focal firm Chain director Large retailer Chain director Not defined Focal firm First-tier suppliers Second-tier suppliers First-tier customers Second-tier customers Chain director Principal because of economic (in case of chain directors) or social reasons (family ties). By using codes of conduct that can rely on third-party certification, more relevant information can be exchanged between the principal and the agents in the supply chain. In all four chains, SA8000 has been implemented by the principal as an add-on to the contracts to reduce the agency problem. Proposition 1. Codes of conduct can improve the communication flows on intangible aspects of business between indirect partners within a supply chain and consequently reduce information asymmetry between the principal and the agents, whether in direct or in indirect relationships. Although all parties have as a common goal to increase the competitiveness of the chain, each of them also has to a certain level a conflicting goal: to increase its benefit from the chain. Reward structures are not sufficient to safeguard against this risk because of incompleteness and/or lack of contracts as well as the intangibility of certain aspects of the relationships, e.g. child labour. This requires additional actions on behalf of the principal to ensure that all agents live up to the norms to reduce information asymmetry. In the analyzed cases, such actions included the use of SA8000 by the principal and the compliance with the principles of such code by all the agents in the supply chain. The intangible aspects of business materialized in the issues addressed by the SA8000 principles (e.g. child labour) and hence communication is facilitated, between both direct or indirect partners. In addition to this, the third-party monitors all parties. As a consequence, the principal is able to have (with more credibility) more information about the agents and information asymmetry between partners, both in direct and in indirect relationships, is reduced. Proposition 1 is thus supported. Proposition 2. Codes of conduct can solve the adverse selection problem both in the process of searching new suppliers (selected from certified firms) as well as when negotiating intangible issues in new contracts with current suppliers, when they accept certification. As none of the suppliers had been SA8000-certified before the contract was negotiated, the code did not help to reduce adverse selection in the search process. It could in the future as some of the agents are SA8000-certified or are in the process of obtaining the certification as a result of their membership to the chain (e.g. in the Cora Kemperman case). Chicom obtained SA8000 certification because of the pressure exerted by the chain director. In the other cases, the principal, although it is an SME, was able to exert pressure on the agents to fulfil SA8000 requirements. During the first Small producers from abroad (Chicom is their major customer) Quality Pre-mixing of building materials in niche market Small joint-ventures abroad selection of suppliers, however, the principals had to rely on other tools to reduce adverse selection. After that, they had to convince their partners during the negotiation process that SA8000 would facilitate their cooperation. The principals could only be sure that the agents agreed to put in writing the fulfilment of SA8000 principles. These partners (i.e. the agents who are assumed to be riskaverse) would only agree on this if they intended to live up to the behaviour prescribed by SA8000. If they would not live up to these norms, this would be discovered during the monitoring sessions. Proposition 2 is partially supported. More in detail, suppliers could not be selected from a pool of SA8000-certified firms and consequently that part of the proposition is not supported. However, during the negotiation process the suppliers accepted the certification as well as the related procedures which would show non-compliance; hence the latter part of the proposition is supported. The principal now has a tool to be used in case the agents would act opportunistically in advance by providing the wrong information about their intentions. This solves the adverse selection problem to a large extent as intentions of each of the partners as well as their competences are checked in a formalized way and approved whether to fulfil requirements needed. Once certified firms are invited, this facilitates the selection process; once firms are certified during the negotiation process or communicate then their intention, negotiations are facilitated. Proposition 3. Codes of conduct can solve moral hazard problems with respect to intangible aspects of business because monitoring improves communication on these issues, in particular in case of thirdparty monitoring of all parties involved. Moral hazard problems are reduced as the certifying agency audits all certified chain members on a regular basis checking whether they still live up to the norms and act accordingly. Hence the transaction costs are reduced because of simpler contracting, reduced risk opportunism, and related costs to prevent and correct if occurred. All parties (i.e. the principal as well) in the analyzed cases are monitored according to SA8000 procedures by a third-party, which increases transparency for all chain partners. Third-party monitoring by local SA8000 auditors facilitates this in particular for SMEs, due to less expensive visits, familiarity with local circumstances, and more perceived objectivity. On the other hand, such inspections have several limitations, since they are generally agreed in advance, have a limited frequency and time length. Monnalisa F. Ciliberti et al. / Journal of Cleaner Production 19 (2011) 885e894 thinks that these limitations make it difficult to carefully check suppliers’ respect for working conditions and human rights. In some cases (e.g. Monnalisa), in addition to the SA8000 procedures, the agents are invited to visit Italian plants so as they can learn best CSR practices. Agents can thus have a better understanding of what their final customers expect from them and consequently information asymmetry can be reduced. In order to reduce information asymmetry, the analyzed firms may also cooperate with local NGOs in developing countries. Hence, proposition 3 is partially supported as monitoring reduces the moral hazard problem faced by SMEs. Once again, communication between parties in the chain, in particular on intangible aspects of the business, is facilitated and risk-averse agents may fear that not living up to the norms will be detected. Even in case they would not and tried to hide this, they may fear that this might be discovered by the monitoring party. All parties involved (i.e. the initiator of the standard, the certifying agent, the principal, the agent, and the public) should be able to rely on this to make the standard effective. Hence, these audits should solve the moral hazard problem by means of additional communication mechanisms provided by the code, e.g. third-party monitoring. This happens in particular once suppliers understand that noncompliance might cause breaking up the relationship, as happened in the cases of Monnalisa and Chicom. Proposition 4. Agents can further reduce moral hazard problems when they apply bonding practices towards their partners, such as the principal, to support monitoring activities by a third-party. Agents could meet bonding costs to show that they are living up to expectations as explicated in the certification process by providing additional information to the principal. In addition to this, the certifying organization may require additional material (e.g. annual reports) from the agents to further reduce agency problems: risk-averse agents provide risk-neutral principals with information to show their performance. Suppliers of Monnalisa, Chicom and Chimica Edile prepare selfevaluations to facilitate easy monitoring in a way that suits suppliers in various countries. Indian suppliers convinced Cora Kemperman of the Western bias in SA8000 which resulted in alterations in the application of some of its policies. Proposition 4 is supported. In addition to the visits of the certifying agency, the suppliers provide evidence to the principal showing to what extent they live up to norms, as such contributing to the communication on intangible aspects of CSR. These bonding activities facilitate the work by the monitoring agencies and ensure even more security on whether the chain as a whole lives up to the promises made to the public about intangible aspects of the products, such as child labour. 893 principal as an add-on to contracts to reduce information asymmetry and increase transparency. Codes as a hybrid coordination mechanism introduce aspects of hierarchy in market relationships because third parties will monitor agents’ behaviour. Principals will benefit from this in particular when they are SMEs, with limited resources. In addition, transparency is increased as norms are known in advance and all parties, principal included, are monitored along the same lines. In the analyzed cases the agents were not yet certified when selected, but negotiations were facilitated as the suppliers agreed to live up to the SA8000 norms or even started the certification process. Hence the adverse selection problem was reduced. Moral hazard problems were reduced in a similar vein as the certifying agency audits all certified chain members on a regular basis. However, these monitoring results provide only a snapshot that might not fully reflect required behaviour, although it should be in the common interest of all parties involved to show their commitment. Riskaverse agents should be expected to live up to the norms, yet in some cases not all suppliers lived up to the norms and ultimately in two cases the contracts with such suppliers were terminated. Hence doubts may exist on whether the agents do indeed respect the codes’ requirements. They can reduce these doubts by meeting bonding costs to show the principal and the public that they do live up to the requirements. Agents in the examined cases provided annual reports and self-evaluations to underline their positive attitude to the code and to prove that they behave in line with its requirements. One limitation of the implementation of codes like SA8000 along the supply chain is that the risk of opportunistic behaviour is not completely eliminated, as the chain director may not be able to identify all violations of the standard. To some extent the focal firms need to rely on other parties, such as local NGOs for additional safeguarding and interpretation of behaviour. This is even more important if the geographic, economic, and cultural distance between partners increases. The Western origin of many codes, SA8000 included, makes this even more complicated. Applying codes should therefore not be studied as an isolated topic but rather as part of a learning network: partners with different backgrounds should learn from each other to understand the codes’ meaning and their application in different situations. This could be one line for further research. Not in all cases customers were willing to pay a premium price for the application of CSR practices and codes such as SA8000. Consequently, the additional certification, monitoring and bonding costs are not covered by higher incomes for the chain as a whole. Further research could aim at investigating how to increase the awareness of the customers, as part of the public, and make them willing to reward socially responsible companies as well as identifying instruments to redistribute supply chain profits in line with their overall contribution. 7. Conclusions References The aim of this paper was to examine how a code of conduct (i.e. SA8000) can help to address the principal-agent problem, for SMEs, between chain directors and partners in supply chains. A literature review brought to the definition of four propositions, which were tested by means of four case studies. The cases showed that two propositions (1 and 4) were supported and the other two (2 and 3) were partially supported. Codes of conduct, in particular when involve third-party certification, enable the exchange of more relevant and focused information in a supply chain between both direct and indirect partners. For direct partners, information on invisible aspects is involved; whereas for indirect partners the chain director, as the principal, can communicate the norms the chain members, as agents, have to live up to. In all four cases SA8000 has been implemented by the Ang, J.S., 1991. Small business uniqueness and the theory of financial management. J. Small Bus. Finance 1 (1), 1e13. Blowfield, M., Murray, A., 2008. Corporate Responsibility. A Critical Introduction. Oxford University Press, Oxford. Bremer, J., Udovich, J., 2001. Alternative approaches to supply chain compliance monitoring. J. Fash. Mark. Manag. 5 (4), 333e352. Ciliberti, F., de Groot, G., de Haan, J., Pontrandolfo, P., 2009. Codes to coordinate supply chains: SMEs’ experiences with SA8000. Supply Chain Manag. Int. J. 14 (2), 117e127. Ciliberti, F., Pontrandolfo, P., Scozzi, B., 2008. Investigating corporate social responsibility in supply chains: a SME perspective. J. Clean. Prod. 16 (15), 1579e1588. Commission of the European Communities, 2001. Green Paper. Promoting a European Framework for Corporate Social Responsibility, COM. 366 Final. Commission of the European Communities, Brussels. Cramer, J.M., 2008. Organising corporate social responsibility in international product chains. J. Clean. Prod. 16 (3), 395e400. 894 F. Ciliberti et al. / Journal of Cleaner Production 19 (2011) 885e894 de Haan, J., de Groot, G., Loo, E., Ypenburg, M., 2003. Flows of goods or supply chains: lessons from the natural rubber industry in Kerala, India. Int. J. Prod. Econ. 81e82, 185e194. Demsetz, H., 1983. The structure of ownership and the theory of the firm. J. Law Econ. 26 (2), 375e390. Demsetz, H., Lehn, K., 1985. The structure of corporate ownership: causes and consequences. J. Political Econ. 93 (6), 1155e1177. Dewally, M., Ederington, L., 2006. Reputation, certification, warranties, and information as remedies for seller-buyer information asymmetries: lessons from the online comic book market. J. Bus 79 (2), 693e729. Easterwood, J.C., Singer, R.F., 1981. Are the motivations for leveraged buyouts the same for large and small firms? In: Yazdipour, R. (Ed.), Advances in Small Business Finance. Kluwer Academic Publishers, Boston, pp. 79e92. Eisenhardt, K., 1989a. Agency theory: an assessment and review. Acad. Manag. Rev. 14 (1), 57e74. Eisenhardt, K., 1989b. Building theories from case study research. Acad. Manag. Rev. 14 (4), 532e550. Emmelhainz, M.A., Adams, R.J., 1999. The apparel industry response to “sweatshop” concerns: a review and analysis of codes of conduct. J. Supply Chain Manag. 35 (3), 51e57. Enderle, G., 2004. Global competition and corporate responsibilities of small and medium-sized enterprises. Bus. Ethics Eur. Rev. 13 (1), 51e63. Fama, E.F., 1980. Agency problems and the theory of the firm. J. Political Econ. 88 (2), 288e307. Gereffi, G., 1999. International trade and industrial upgrading in the apparel commodity chain. J. Int. Econ. 48 (1), 37e70. Halldorsson, A., Kotzab, H., Mikkola, J.H., Skjoett-Larsen, T., 2007. Complementary theories to supply chain management. Supply Chain Manag. Int. J. 12 (4), 284e296. Hand, J.H., Lloyd, W.P., Rogow, R.B., 1982. Agency relationships in the close corporation. Financial Manag. 11 (1), 25e30. Heide, J.B., John, G., 1988. The role of dependence balancing in safeguarding transaction-specific assets in conventional channels. J. Mark. 52 (1), 20e35. Henkle, D., 2005. Cap Inc. sees supplier ownership of compliance with workplace standards as an essential element of socially responsible sourcing. J. Organ. Excell. Winter, 17e25. Hwang, I., Radhakrishnan, S., Su, L., 2006. Vendor certification and appraisal: Implications for supplier quality. Manag. Sci. 52 (10), 1472e1482. ISO, 2009. Draft International Standard ISO/DIS 26000. Guidance on Social Responsibility. International Organization for Standardization. Jenkins, H., 2004. A critique of conventional CSR theory: an SME perspective. J. Gen. Manag. 29 (4), 37e57. Jensen, M., Meckling, W., 1976. Theory of the firm: Managerial behaviour, agency costs and owner structure. J. Financial Econ. 3 (4), 305e360. Kaufmann, L., Carter, C.R., 2006. International supply relationships and non-financial performance e A comparison of U.S. and German practices. J. Oper. Manag. 24 (5), 653e675. Leipziger, D., 2001. SA8000: The Definitive Guide to the New Social Standard. Financial Times e Prentice Hall, London. Lepoutre, J., Heene, A., 2006. Investigating the impact of firm size on small business social responsibility: a critical review. J. Bus. Ethics 67 (3), 257e273. Mamic, I., 2005. Managing global supply chain: the sports footwear, apparel and retail sectors. J. Bus. Ethics 59 (1), 81e100. McMahon, R.G.P., 2004. Equity agency costs amongst manufacturing SMEs. Small Bus. Econ. 22 (2), 121e140. Miles, M.P., Munilla, L.S., 2004. The potential impact of social accountability certification on marketing: a short note. J. Bus. Ethics 50 (1), 1e11. Mueller, M., Gomes dos Santos, V., Seuring, S., 2009. The contribution of environmental and social standards towards ensuring legitimacy in supply chain governance. J. Bus. Ethics 89 (4), 509e523. Pedersen, E.R., Andersen, M., 2006. Safeguarding corporate social responsibility (CSR) in global supply chains: how codes of conduct are managed in buyersupplier relationships. J. Public Aff. 6 (3e4), 228e240. Pettit, R.R., Singer, R.F., 1985. Small business finance: a research agenda. Financial Manag. 14 (3), 47e60. Rindfleisch, A., Heide, J.B., 1997. Transaction cost analysis: past, present, and future applications. J. Mark. 61 (4), 30e54. Roberts, S., 2003. Supply chain specific? Understanding the patchy success of ethical sourcing initiatives. J. Bus. Ethics 44 (2/3), 159e170. Rohitratana, K., 2002. SA 8000: a tool to improve quality of life. Manag. Audit. J. 17 (1/2), 60e64. Salomone, R., 2008. Integrated management systems: experiences in Italian organizations. J. Clean. Prod. 16 (16), 1786e1806. Sarkis, J., Talluri, S., 2004. Evaluating and selecting e-commerce software and communication systems for a supply chain. Eur. J. Oper. Res. 159 (2), 318e329. Spence, L.J., 2007. CSR and small business in a European policy context: the five “C”s of CSR and small business research agenda. Bus. Soc. Rev. 112 (4), 533e552. Stigzelius, I., Mark-Herbert, C., 2009. Tailoring corporate responsibility to suppliers: managing SA8000 in Indian garment manufacturing. Scand. J. Manag. 25 (1), 46e56. Terlaak, A., King, A., 2006. The effect of certification with the ISO 9001 quality management standard: a signalling approach. J. Econ. Behav. Organ. 60 (4), 579e602. Vermeulen, W.J.V., Ras, P.J., 2006. The challenge of greening global product chains: meeting both ends. Sustain. Dev. 14 (4), 245e256. Welford, R., Frost, S., 2006. Corporate social responsibility in Asian supply chains. Corp. Soc. Responsib. Environ. Manag. 13 (3), 166e176. Wolters, T., 2003. Transforming international product chains into channels of sustainable production. The imperative of sustainable chain management. Greener Manag. Int. 43, 6e13. Yin, R.K., 2003. Case Study Research. Design and Methods. Sage Publications, Newbury Park.