How to Make Better Factories

Explaining Variance in Labour Standard Compliance Across Asia's Garment Economies

Master's Thesis 2009 72 Pages

Politics - International Politics - Topic: Development Politics







1.1 Subject and Research Question
1.2 Methodology and Collection of Data
1.3 Research Objectives and Structure
1.4 Review of Literature
1.4.1 New Modes of Governance
1.4.2 Compliance Theory
1.4.3 Labour Standards and Business Regulation in Developing Economies

2.1 Dependent Variable: Compliance with Labour Standards
2.2 Independent Variable: Mode of Governance
2.3 Alternative Explanations and Logic of Comparison

3.1 The Global Garment Industry
3.2 Improved Compliance in Cambodia
3.3 Persistence of Non-Compliance in Bangladesh

4.1 The Enforcement School
4.2 The Management School
4.3 Complementary-Compliance-Hypothesis
4.4 Operationalization

5.1 The US-Cambodian Trade Agreement on Textile and Apparel
5.2 Eliminating Free-Riding: A Necessary Condition
5.3 Better Factories by Enforcement
5.3.1 Monitoring: Data about Non-Compliance
5.3.2 Sanctioning Non-Compliance Phase 1 (2001-2005): Hegemonic Enforcement Phase 2 (2005-2008): Transnational Market-Driven Enforcement and Increased Capability of Local Unions
5.3.3 Reduce of Intentional Non-Compliance
5.4 Better Factories by Management
5.4.1 Monitoring: Data about Capacity Deficits
5.4.2 Capacity Building General Capacity Building Customized Capacity Building
5.4.3 Reducing Non-Compliance Due to Capacity Deficits

6.1 Summary of Empirical Findings.
6.2 Outlook on Future Research..
6.3 BFC - A Model for Regulating For-Profit Actors in Developing Economies beyond Cambodia and the Garment Industry?




Table 1: Logic of Comparison

Table 2: Growth of Cambodia’s Export Processing Garment Industry, 1995-2003

Table 3: % of Factories in Compliance with Freedom of Association and Anti- Union Discrimination, Nov 2001 and Oct

Table 4: Growth of Bangladesh’s Export Processing Garment Industry, 1991-2005

Table 5: Operationalization of the Enforcement Path

Table 6: Operationalization of the Management Path

Table 7: % of Factories in Compliance with Wage Related Standards, Oct 2008


Figure 1: Enforcement Path

Figure 2: Management Path

Figure 3: Complementary Compliance Path

Figure 4: Governing Labour Standard Compliance in Cambodia’s Garment Industry (Causal Path), 2001-1990


Graph 1: % of Factories in Compliance with BFC Standards by Category, Oct 2006-Oct

Graph 2: Average Number of FLA Code Violations per Factor by Category, 2003-2007

Graph 3: % of Factories in Compliance with OSH Related Standards, Oct 2006-Oct


Abbildung in dieser Leseprobe nicht enthalten


1.1 Subject and Research Question

Many scholars claim that the patterns of governance are changing as part of a broader process of globalization or denationalization (Czempiel/Rosenau 1992; Held et al. 1999; Zürn 1998). Often, this change is associated with a retreat of the sate and an increased importance of private and public-private forms of political steering (Cutler et al. 1999; Hall/Bierstecker 2002; Haufler 2001; Reinicke/Deng 2000; Strange 1996). Recently, these new political arrangements have attracted a large amount of scholarly interest. To explore the supposed retreat of the state and the newly emerging institutions, most studies adopt a governance perspective, as this allows an analysis of political steering decoupled from traditional governmental and intergovernmental decision making (cf. Graz/Nölke 2008a: 12). According to Mayntz, governance refers to “all institutions designed for the deliberate solving of collective action problems, irrespective of the private or public character of the actors involved and the hierarchical or horizontal mode of their (purposive) interaction” (Mayntz 2007: 6). Recent studies in this line of research focus on questions related to the new arrangements’ authority, design, legitimacy and effectiveness (Beisheim et al. 2007; Bernstein/Cashore 2007; Fuchs 2007; Hall/Bierstecker 2002; Koening-Archibugi/Zürn 2006; Pattberg 2007).

The above described transformation of governance patterns has been most pronounced in the economic sphere. In the last decades, free trade, technological progress and reduced transportation costs increased the mobility of capital and companies and thereby created “a mismatch between markets and politics in terms of governance” (Haufler 2000: 122). Notably, in the newly industrializing economies fast growth induced demand for regulation, but the lack of institutional capacity and political will resulted in a regulatory void of domestic for-profit actors (cf. Vogel 2008: 275). The consequences were environmental disasters and social hardship. In the 1990s, such substandard conditions became a major issue of political contention in the West, most notably in North America, which constituted the point of departure for transnational business regulation (cf. Bullert 2000; Haufler 2001).

For the study of the newly emerging institutions, the garment industry is a particularly interesting case. Its relevance is owed to the sectors’ avant-garde position in terms of transnational private and public-private business regulation. A short sketch of the industry’s recent history will shed some light on the matter. Garment manufacturing was one of the forerunners of economic globalization, as its labour intensity made outsourcing production to developing economies highly profitable. Therefore, beginning already in the 1960s, Western garment companies started to relocate production to the low standard and low wage cost economies of the South, most notably Asia. For these countries, export oriented garment production constituted a stepping stone to economic development (Mamic 2004: 149). However, evidence about sweatshop conditions in the transnational companies’ (TNC) subcontracting factories revealed the downside of this process. Outsourcing production often went hand in hand with undermining international labour standards, as in these countries business actors are often poorly regulated. In this situation, a number of larger scandals in factories producing for Nike and other leading brands gave rise to the so called anti- sweatshop movement, a global network comprised of non-governmental organizations (NGO), student groups and trade unions. By means of public campaigns, they exercised pressure on TNCs and Western governments to do something about the issue (Bullert 2000). In the second half of the 1990s, this pressure resulted in number of initiatives, ranging from company self-regulation to private-private and public-private arrangements, seeking to address the problem of poor working conditions at the supply chains’ lower ends (cf. Merk 2008; O'Rourke 2006).1 However, such initiatives became a contentious issue themselves (cf. Vogel 2008: 274). In this respect, critiques argue that companies’ self-regulatory efforts are often weak and aimed more at public relations than substantial improvement in social performance. Multi-stakeholder initiatives, involving NGOs, are generally considered more credible but said to often lack the capacity for scaling up the extremely complex and costly monitoring and verification procedures (Esbenshade 2004; Jenkins 2001; Locke et al. 2006; Utting 2002). Given the complexity of the task and the fact that production is constantly relocated to low cost and low standard environments, poor working conditions remain a problem throughout the industry (cf. Merk 2008: 124; Milberg 2004: 8; Oxfam International 2004, 2006).

Against this background, recent news about improved working conditions in Cambodia’s garment factories constitutes an empirical puzzle, as the country scores rather badly with regard to factors commonly assumed to negatively affect labour standards.2 In this regard, Cambodia is one of the poorest countries in the region which is notorious for its weak institutions and highly corrupt officials (Hall 2000: 126ff.; Kolben 2004: 85f.). Furthermore, with expiration of the Agreement on Textile and Clothing (ATC) at the end of 2004, competition among garment producing countries increased considerably.3 The case of Cambodia is even more puzzling, as evidence points to the persistence of poor working conditions in comparable countries such as Bangladesh (cf. Berik/van der Meulen Rodgers 2007).

To find a solution to this puzzle, the thesis’ research question is: Why and how to explain variance in labour standard compliance across Asia’s garment economies?

The rather praxis-oriented literature on the topic points to an innovative policy project called Better Factories Cambodia4 (BFC), which was designed by the US Department of Labor (DoL) and the International Labour Organization (ILO) (Chiu 2007; Kolben 2004, 2007; Polaski 2004, 2006; Sibbel/Borrmann 2007). Established in 2001, BFC is a public-private partnership (PPP) involving actors from different levels to improve compliance with labour standards in Cambodia’s export processing zones (EPZ). To achieve this, the project runs an extensive monitoring programme and offers training to managers and workers. Thereby, BFC’s activities correspond to recent claims of compliance theorists which stress that compliance governance is most likely to be effective if combining management and enforcement strategies (Tallberg 2002). Drawing on this literature, this study seeks to identify the causal mechanism(s) which led to improved compliance of domestic for-profit actors in one of the world’s least developed countries (LDC).

1.2 Methodology and Collection of Data

The examination of the above stated research question proceeds in two steps. First, a controlled comparison of Cambodia’s and Bangladesh’s garment sector is conducted, in order to verify or falsify the contended correlation between BFC and improved compliance with labour standards. Second, a process analysis is used to identify the causal mechanism(s) in operation.

Why has labour standard compliance improved in Cambodian garment factories whereas non- compliance persists in Bangladesh? Is this due to the work of BFC as claimed in the praxis- oriented literature? To examine the research question’s first part, an inter-sector comparison of Bangladesh’s and Cambodia’s garment sectors for the period of 2001-2008 will be conducted.5 Alternative explanations will be controlled by case selection (most similar system logic).6 Data about labour standard compliance is compiled both from qualitative and quotas on the amount of garments and textiles developing countries could export to developed countries. It expired at the end of 2004. quantitative sources. For the case of Cambodia, the paper primarily draws on BFC monitoring data. On a semi-annual basis, BFC publishes so called synthesis reports on its website.7 In an aggregated form, these reports provide compliance data from approximately 200 garment factories located in Cambodia’s EPZs. BFC monitoring is based on a checklist comprising the international core labour standards as well as wage, working hour and occupational safety and health (OSH) related standards. BFC monitoring data is available for the period 2001-2008. Regarding the case of Bangladesh, data is compiled from the factory tracking charts of the Fair Labor Association (FLA). The FLA is a multi-stakeholder initiative which conducts unannounced supply chain audits of accredited companies. In the period of investigation FLA auditors monitored 20 export processing garment factories in Bangladesh.8 On an annual basis, FLA publishes its monitoring results on its website.9 Comparable to the BFC checklist, the FLA code of conduct comprises international core labour standards as well as wage, working hour and OSH related standards. FLA monitoring data is available for the period 2003-2007. Further, both case studies will be supplemented by data from qualitative sources such as the US Department of State’s Country Reports on Human Rights Practices, reports of the International Confederation of Free Trade Unions (ICFTU) and reports of the UN Commission on Human Rights.

How did BFC succeed in improving labour standard compliance in Cambodia’s garment industry? To identify the governance instrument(s) in operation and to analyse how they improved compliance with labour standards, a within case study of Cambodia’s garment sector will be conducted. Drawing on compliance theory, namely the enforcement and management approach of compliance, causal mechanisms will be deduced and put to the test using the method of process tracing. Process tracing is a deductive method which looks at the observable implications, so called process variables, of a causal mechanism. Taken together, these variables form a causal path. The objective is to verify whether the deduced path forms an uninterrupted chain of evidence linking hypothesised cause to observed effect (Bennett 2004: 22ff.). Therefore, the process variables of the causal path have to be operationalized and confronted with empirical data. To test the causal mechanisms deduced for the examination of BFC, two qualitative research strategies have been applied. First, data has been compiled from primary and secondary literature. In this regard, BFC’s website, which contains detailed information about the project’s structure and work, has served as an especially important source of evidence.10 Further, the existing secondary literature on BFC has been reviewed. Second, semi-structured interviews with experts have been conducted. The expert interview is a qualitative method to generate data by interviewing people who possess a particular knowledge about the subject of research (Froschauer/Lueger 2003). Interviewing experts became necessary to fill remaining information gaps. One interview has been conducted with an official of the US DoL who participated in planning and implementing BFC. Further, a leading staff member of BFC has been interviewed.11

1.3 Research Objectives and Structure

The objectives of this study are threefold: first, to contribute to the study of new modes of governance; second, to collect empirical data about BFC; and third, to make a contribution to practical problem solving.

The study’s primary objective is to identify the causal mechanism(s) in operation and thereby to contribute to the research on new modes of governance in the context of developing economies. New modes of governance such as corporate self-regulation, private-private and public-private partnerships are relatively new phenomena. No established theoretical frameworks exist to analyse how these new governance arrangements can contribute to market regulation in developing economies (cf. Beisheim et al. 2005: 2; Graz/Nölke 2008a: 2; Vogel 2008: 275). To address this deficit, the study draws on compliance theory, namely the management and the enforcement approach of compliance, to generate testable hypotheses. Applying this framework to analyse BFC’s success in regulating Cambodian garment companies extends the current reach of compliance theory, originally developed within the study of inter-state cooperation. Moreover, the theory-guided analysis of BFC contributes to a better understanding of governance instruments best suited to improve social performance of for-profit actors in developing economies. Thereby, it generates theoretical propositions for future comparative research.

The study’s second objective is to collect empirical data about BFC. New modes of governance are an empirically under-researched phenomenon (cf. Beisheim et al. 2005). Although BFC has been discussed extensively in the praxis-oriented literature,12 important questions, especially regarding the exact functioning of its compliance system, remain unanswered. The study seeks to fill the remaining information gaps. Drawing on primary and secondary qualitative sources a comprehensive empirical account of the functioning and the form of BFC’s compliance system is provided. This indirectly contributes to future theory building, as empirical data is a necessary prerequisite for comparative research.

A final important objective of this study is to contribute to practical problem solving. Poor working conditions are a widespread problem in factories located in developing economies. Examining the case of Cambodia may yield some valuable insights about how to design programmes aimed at improving social performance of business actors in such environments. This paper is structured as follows: Chapter 1 provides a state of research on new modes of governance and compliance theory as well as a review of the literature on labour standards and business regulation in developing economies. Chapter 2 introduces the dependent and independent variables and discusses competing explanations. In Chapter 3, a congruence analysis of Cambodia’s and Bangladesh’s garment sector is conducted in order to verify or falsify the contended correlation between BFC and improved social performance in Cambodia’s garment factories. Chapter 4 introduces the theoretical framework from which a working hypothesis for the subsequent process analysis is generated. In Chapter 5, a within case study of Cambodia’s garment sector confronts the hypothesized causal path with empirical data. In conclusion, Chapter 6 summarizes the study’s major findings, provides an outlook on future research and discusses whether BFC could serve as a model for other countries and industries.

1.4 Review of Literature

For this study different strings of literature are relevant. These are: the literature on new modes of governance, the literature on compliance theory and the literature on labour standards and business regulation in developing economies. The following sections will provide a short review of the state of research in these bodies of literature.

1.4.1 New Modes of Governance

In the 1990s, governance became the buzzword in international relations (IR) and related disciplines (Held/McGrew 2002; Héritier 2002; Pierre 2000; Rosenau et al. 1992; Young 1999). The concept gained importance within the broader research on globalization, and many refer to it when analysing changing patterns of political steering. In this respect, governance is commonly understood in opposition to government, which political scientists associate with a top-down, hierarchical form of steering (Benz et al. 2007: 11). By contrast, the governance concept is often used in terms of “governance with government” or “governance without government”, that is, forms of political steering which involve private actors or are conducted exclusively by them. Moreover, these new arrangements are often conceived to rely on non- hierarchical or soft modes of steering (cf. Arts 2006; Risse/Lehmkuhl 2006; Rosenau et al. 1992). This commonly made distinction between governance and government, however, is misleading, as the governance concept encompasses all forms of political steering irrespective of their public or private nature (cf. Mayntz 2007). In this regard, it can best be conceived as a continuum that includes hierarchical state steering, state-centred arrangements with private actor access, public-private partnerships, private governance with public supervision and/or support and purely private forms of governance (Koening-Archibugi 2006: 13). Nonetheless, it is true that in most cases the governance concept is used when referring to constellations involving private actors. Recent research on the subject revolves around questions related to the new arrangement’s authority, design, legitimacy and effectiveness.13 However, as the central theme of this study is compliance, the following review will focus on research concerned with the effectiveness of new modes of governance.14

It is often claimed that the process of globalization undermines the state’s capability to steer effectively, which is said to have resulted in a governance gap or regulatory void (Cutler et al. 1999; Haufler 2000; Strange 1996). Against this background, many policy-makers, practitioners and scholars attribute effectivity and efficiency enhancing effects to forms of governance involving private actors. Wolf, for example, believes that “pooling public and private resources in synergetic relationship could improve the overall problem-solving capacity” (Wolf 2001: 2). However, theory-guided research on new modes of governance in general and their effectiveness in particular is still at a very early stage. Most literature on the subject is descriptive and no established theoretical frameworks exist to study why or how public-private, private-private or purely private arrangements are effective (cf. Beisheim et al. 2005: 2; Graz/Nölke 2008a: 2; Vogel 2008: 275).

More recently, some larger research projects seek to address this deficit. In this respect, a DFG financed project on the effectiveness of transnational PPPs deduces hypotheses from rational institutionalist, realist and constructivist theory to examine the conditions under which these arrangements make a positive contribution to public good provision in developing countries. (Beisheim et al. 2005; Beisheim et al. 2007; Liese/Beisheim forthcoming). Another project within the same research cluster draws on the literature on Europeanization to address the question why and how TNCs, in the form of Corporate Social Responsibility (CSR), implement social and environmental regulation in South Africa (Börzel/Héritier 2005; Hönke et al. 2008). Moreover, some recent studies deal with the effectiveness of new modes of governance. In this respect, Reinhard Biedermann draws on realism and regime theory to examine how a purely private regime improves social regulation in the global toy industry (Biedermann 2006). Further, Doris Fuchs applies a rationalist perspective to analyse the utility function of TNCs to invest in private governance (Fuchs 2007). This short review of recent research makes clear that the study on new modes of governance is still in its infancy. Drawing on compliance theory, this thesis seeks to contribute to this literature by enlarging the theoretical tool box necessary for future comparative research.

1.4.2 Compliance Theory

The study of compliance is transdisciplinary in character and rooted in International Law (IL) and IR theory.15 Its central research questions are: why do states comply with international regimes, and how does their design effect compliance (Raustiala/Slaughter 2002). In the 1990s, the literature on compliance burgeoned. In the research programmes’ early phase, many law students and some political scientists treated the subject of compliance with international court decisions (adjudication) (Garret 1995; Helfer/Slaughter 1997; Weiler 1991). Most influential among IL scholars has been the idea that a rule’s legitimacy, determined through the “right process” of rule making, is central to compliance (Franck 1990). Incorporating some of the insights from IL research, Chayes and Chayes, two political scientists, formulated a general theory of compliance in 1993 (Chayes/Chayes 1993). Their so called management approach assumes that states possess a general propensity to comply with international rules due to considerations of efficiency, norms and interest. In their account, non-compliance is largely due to economic and political capacity deficits as well as rule ambiguity. To address non-compliance, managerialists therefore prescribe measures such as capacity building, creation of transparency and rule interpretation (Chayes/Chayes 1993, 1995; Chayes et al. 1998). The central tenets of the management school provoked harsh critique from rational choice theorists. In what is called the political economy or enforcement approach of compliance, Downs and his colleagues challenge the states’ general propensity to comply with international rules. In their conception, compliance is a function of the incentive structure states face. To prevent non-compliance, they thus propose a sanctioning strategy that punishes violators to an extent which outweighs any potential benefit from non-compliance (Downs et al. 1996). Also inspired by IL theorizing, a third perspective on compliance developed within social constructivism. Scholars writing is this vein identify learning and persuasion as central mechanisms for achieving compliance (Checkel 2000, 2001a, 2001b; Risse 1999). However, the management-enforcement divide is most characteristic for the field, and scholars from these opposing schools have long structured the debate on compliance in IR (Raustiala/Slaughter 2002: 543). More recently, some have attempted to bridge this divide. In this respect, Tallberg argues that management and enforcement are not competing but complementary strategies. Analysing the European Union (EU), he claims that compliance systems are most effective when combining the mechanisms proposed by management and enforcement theorists (Tallberg 2002).

So far, most research in the field has focused on the relationship between the design of international regimes and state compliance. More recently, IL scholars have started to open the black box state and to examine the role of domestic actors, namely national courts, in promoting state compliance with international norms (Alam 2006; Barrett 2009; Rana 2009). However, less attention has been directed to the compliance of sub-state actors themselves, an important issue, notably, with regard to developing economies in which the state often is unable to induce domestic compliance. In this regard, BFC constitutes an interesting case, as it involves public and private actors from different levels to improve compliance of local business actors. Drawing on the approach of complementary compliance, this study seeks to indentify the causal mechanism(s) at work and thereby extends the reach of compliance theory.

1.4.3 Labour Standards and Business Regulation in Developing Economies

In the 1990s, a lively debate evolved around the question whether benefits from international trade should be made conditional on compliance with core labour standards.16 To achieve this, developed countries, Northern unions and activists from the anti-sweatshop movement proposed to link a “social clause” to bilateral and multilateral trade agreements. In favour of such a policy, proponents argued that a social clause could help to protect human rights in the North and South. In this respect, they most prominently claimed that linking trade to labour standards could prevent a so called “race to the bottom” causing a downward harmonization of social standards (Blank 1994; Gunderson 1998; van Liemt 1989). This position was harshly criticized by developing countries, Southern unions and economists as a form of Northern protectionism. They argued that linking trade to labour standards would have distortionary effects and endanger the comparable advantage of developing economies (Brown et al. 1997; Golub 1997; Srinivasan 1996). In fact, until today real world instances of linking trade to labour standards are rare, and more recent voices in the debate claim that the main issue is not whether developing countries should have labour standards, but rather what those standards should be and how they could best be implemented (Singh/Zammit 2004: 3).17 This question is of major concern in the rather praxis-oriented literature on CSR and multi- stakeholder initiatives.18 In the absence of effective public regulation, a variety of private, private-private and public-private regulatory initiatives emerged seeking to improve social performance of business actors in developing economies (O'Rourke 2006; Utting 2002). However, the effectiveness of these new regulatory initiatives raises a contentious issue among practitioners. Proponents believe that such arrangements can have a positive impact on labour regulation in developing countries. For example, in their analysis of corporate self- regulation in the garment industry Kolk and van Tulder find that codes of conduct help to mitigate the problem of child labour (Kolk/van Tulder 2002). Others argue that, based on core labour standards and transparent monitoring schemes, private initiatives can complement international and national labour regulation (Fung et al. 2001; O'Rourke 2003; Webb 2004). In contrast, critiques fear that private sector initiatives constitute a strategy to displace more stricter public regulation (Bartley 2005; Compa 1993; Jenkins 2001). They further doubt their alleged effectiveness and claim that companies’ self-regulatory efforts are often ineffective and aimed more at public relations than substantial improvement in social performance. In this respect, a study of Lock and his colleagues finds that Nike’s CSR efforts had very limited effects on working conditions in its subcontracting factories (Locke et al. 2006). Multi- stakeholder initiatives, involving NGOs, are often seen as more positive. However, many claim that they lack the capacity for scaling up the extremely complex and costly monitoring and verification procedures (Cashore et al. 2004; Esbenshade 2004; Utting 2002). BFC is a hybrid form of regulation, as it involves public and private elements. Examining how this arrangement was successful in improving social performance of business actors in Cambodia will make a substantive contribution to the praxis-oriented discourse on business regulation, as it clearly identifies the causal mechanism(s) in operation.


Why did labour standard compliance improve in Cambodia whereas non-compliance persists in Bangladesh, and how can this be measured? To address these questions, the following sections will introduce and conceptualize the dependent and the independent variables. This is followed by a short discussion of alternative explanations.

2.1 Dependent Variable: Compliance with Labour Standards

In his seminal work on compliance with international public authority, Oran Young defined compliance as:

Compliance can be said to occur when the actual behavior of a given subject conforms to prescribed behavior, and non-compliance or violation occurs when actual behavior departs significantly from prescribed behavior. (Young 1979: 104)

Most studies on compliance therefore deal with the behavioural influence of rules or compliance systems. Although closely related, compliance has to be distinguished from implementation and effectiveness (Mitchell 1996: 24ff.; Raustiala/Slaughter 2002: 539; Young 1999: 110). Implementation describes the process of putting a rule into practice. This encompasses rule creation, the setting up of institutions and rule enforcement. Implementation typically is an important step to compliance. However, compliance can occur in the absence of any effort to implement a rule. For instance, when actual practice already matches the rule, no implementation is needed, as compliance is automatic (Raustiala/Slaughter 2002: 539).

There are also conceptual differences between compliance and effectiveness. Effectiveness has been defined in different ways: for example, as the degree to which a rule solves the underlying problem (ibid.). For effectiveness, understood as problem solving, compliance is neither a necessary nor a sufficient condition. It is not sufficient because high compliance with bad rules does not solve problems. It is not necessary, as problem solving might be due to factors exogenous to the rule (Mitchell 1996: 25; Raustiala/Slaughter 2002: 539). Broader definitions of effectiveness blur the lines between implementation, compliance and effectiveness. A widely used concept identifies different dimensions of effectiveness which refer to David Easton’s distinction between output, outcome and impact (Easton 1965; Young 1999: 111). According to Young, outputs are regulations, programmes, and organizational arrangements established to operationalize the provisions of a rule. Outcome refers to changes of behaviour of actors addressed by the rule. Impact concerns the question to what degree this contributes to actual problem solving (ibid.). Although logically not compelling, in most empirical cases the rule applies: better implementation increases compliance, which in turn yields high effectiveness (Mitchell 1996: 25; cf. Raustiala/Slaughter 2002: 539). However, regarding this interrelationship there might be variation across policy areas. Whether high compliance yields high effectiveness is largely determined by the willingness and capability of actors to create rules suited to solve problems. For example, even high compliance with the provisions of the Kyoto Protocol is unlikely to make any noteworthy contribution to the problem of climate change (cf. Simmons 1998: 78). In contrast, high compliance with core labour standards is likely to mitigate the problem of poor working conditions and therefore to result in effectiveness understood as problem solving.

Having mentioned labour standards, this concept also requires further specification. Labour standards are rules (legal and non-legal) which regulate conditions at work, as they specify rights and duties of employers and their employees. In a negative sense, they prohibit certain behaviour such as child labour or discrimination at the workplace. In a positive sense, they provide workers with rights, for example, to associate freely and to bargain collectively. Labour norms are embodied in national labour law, in the conventions and recommendations of the ILO, in regional arrangements such as the EU and increasingly within private codes of conduct (cf. Block et al. 2001; Hassel 2008). Labour standards vary in their normative significance. A commonly made distinction divides the body of labour norms into core and cash standards (Elliot/Freeman 2003). The former refers to the core labour standards identified by the ILO in 1998. These are: freedom of association and the effective recognition of the right to collective bargaining (ILO convention 87, 98), elimination of all forms of forced or compulsory labour (ILO convention 29, 105), effective abolition of child labour (ILO convention 138, 182) and elimination of discrimination in respect to employment and occupation (ILO convention 100, 111). Beyond the ILO core standards, a large body of other standards entitle workers to further protection and benefits (e.g. minimum wages, restriction of working hours, proper health and safety conditions). These standards are often called cash standards, as they are assumed to directly affect labour costs (ibid.: 13).

To examine the contended correlation between BFC and improved compliance with labour standards in Cambodia, the development of working conditions in factories located in Cambodia’s and Bangladesh’s EPZs in the period of 2001-2008 will be examined. Qualitative and quantitative evidence point to the fact that both sectors were characterized by equally poor working conditions at the beginning of the investigation period. Therefore, the dependent variable “compliance with labour standards” will be operationalized with (1) “persistence of non-compliance” and (2) “improved compliance”. Due to the lack of comparable data for Cambodia and Bangladesh, the value of the dependent variable will be measured for each case intertemporally. For the case of Cambodia, the study draws primarily on monitoring data from the BFC synthesis reports. Measuring compliance in Bangladesh’s garment sector basically relies on the FLA factory tracking charts. Both case studies will be supplemented by data from qualitative sources.19

2.2 Independent Variable: Mode of Governance

In the social sciences, political steering is traditionally associated with government and its “shadow of hierarchy”20 (Benz et al. 2007: 11). In this conception, it is the state, based on its (legitimate) monopoly of physical force, which provides public goods and implements collectively binding rules. As described in the introductory chapter, this notion has given way for a more differentiated view of political steering, involving non-state actors and non- hierarchical mechanisms of steering (e.g. market mechanism). The buzzword new modes of governance is often used when referring to such arrangements (Héritier 2002; Koening- Archibugi/Zürn 2006; Risse/Lehmkuhl 2006). In the last two decades, they have proliferated in numbers and across policy areas. With regard to business, public-private and purely private forms of governance play an important role in the regulation of markets today. Most notably, this applies to developing economies (cf. World Bank 2003a; World Bank 2003b). In these environments, characterized by weak institutions, they increasingly function as providers of social and environmental regulation (cf. Risse/Lehmkuhl 2006).

BFC is such a new governance arrangement. It is a PPP involving actors from different levels in order to induce compliance of for-profit actors in Cambodia. Many professionals attribute better working conditions in the country’s garment factories to its activities (Burrow 2008; Carlborg 2008; Kearney 2008; Muzaffar 2008). For example, Sharan Burrow, president of the International Trade Union Confederation, referring to the work of BFC, stated: “We have seen fantastic outcomes. (…) (T)here is compliance with basic labor standards and rights (…). We understand now that this can work, and so we are looking at expanding this to many other countries” (Burrow 2008). Also the literature dealing with labour standards in Cambodia’s EPZs emphasizes BFC as the determining factor (Chiu 2007; Kolben 2004, 2007; Polaski 2004, 2006; Sibbel/Borrmann 2007). Therefore, BFC will be examined as the primary explanatory variable. The following lines will provide a short overview of the project’s origin, structure and major activities.

In 1999, the US-Cambodian Trade Agreement on Textile and Apparel granted bonus export quotas to Cambodia’s garment manufacturers (U.S.-Cambodian Trade Agreement on Textile and Apparel 1999). This privileged market access, which offered a possible 18% of extra export entitlements, was conditional based on “substantial compliance” with international labour standards and Cambodian national labour law (Office of the U.S. Trade Representative 2002: 7f.). The agreement constituted one of the rare instances of linking a social clause to trade. To address the problem of verifying actual compliance performance on the ground, the US DoL approached the ILO to establish a sector based monitoring programme. After consultations with all parties involved, including the Garment Manufacturers Association in Cambodia (GMAC) and Cambodian trade unions, a technical cooperation project was agreed upon in May 2000. In January 2001, BFC (initially called the ILO Garment Sector Project) became operational (Sibbel/Borrmann 2007: 237). It was the first time that the ILO engaged in direct monitoring of private actors (Kolben 2004). With the phasing out of the ATC at the end of 2004, the US-Cambodian bilateral trade agreement came to an end as well. However, BFC remained operational and continues its work until today.

BFC reflects the tripartite structure of its host organization. Its steering body, the so called Project Advisory Committee (PAC), comprises three representatives, each from the government of Cambodia, GMAC and Cambodian trade unions. The PAC meets as necessary and provides guidance and advice. BFC’s day-to-day management, however, lies in the hands of the Chief Technical Advisor (CTA), who is appointed by the ILO (Sibbel/Borrmann 2007: 238). Besides this formal involvement of local private actors in the governance structure of BFC, the project keeps close ties with international buyers sourcing from Cambodia. Until now, twenty-four TNCs have subscribed to BFC’s monitoring programme, among them leading companies such as Adidas, H&M and Gap. The annual budget of BFC, which amounts to approximately 1.4 million US$, is jointly financed by public and private actors. In its initial phase, the US government had the largest share followed by the Cambodian government and GMAC (Berik/van der Meulen Rodgers 2007: 15f.). Today, funding is more diverse, including TNCs and international organizations such as the World Bank as well (Better Factories Cambodia 2005a, 2009c).

BFC’s work is based on a code of conduct comprising international labour standards and Cambodian national labour law. Its major activities are monitoring and capacity building. In this regard, a team of specially trained auditors conducts unannounced visits and follow-ups to the countries’ export processing factories. Every single factory is checked at least twice a year. In an aggregated form, monitoring results are made public on BFC’s website. Further, individual factory compliance data is made available to third parties with authorization of factory owners (Better Factories Cambodia 2007c). Besides its monitoring programme, BFC runs a general training programme for managers and workers to raise awareness of labour standard issues and to build capacity in the sector. More recently, it introduced a special remediation procedure aimed at improving individual factories’ capability to better comply with labour standards (Better Factories Cambodia 2009g).


1 For an overview of the new regulatory arrangements in the garment industry see Diller (1999); O'Rourke (2006)

2 For a detailed discussion of these factors see section 2.3

3 Since 1974, global trade in garments and textiles was regulated by the Multi Fibre Arrangement (MFA), which was followed by the ATC in 1994. In order to protect Western garment producers, the MFA regime imposed

4 BFC was formerly known as the ILO Garment Sector Project.

5 The period of investigation results from the fact that BFC was established in 2001 and compliance data is available until 2008.

6 For a detailed account of the logic of most similar system comparisons see Bennett (2004), p. 30ff.

7 See http://www.betterfactories.org/ILO/resources.aspx?z=7&c=1

8 The FLA does not monitor factories in Cambodia.

9 See http://www.fairlabor.org/what_we_do_public_reporting_b2.html

10 See http://www.betterfactories.org/

11 For interview details see Annex 1. Footnotes will indicate when information from the interviews is used.

12 See Chiu (2007); Kolben (2004); Kolben (2007); Polaski (2004); Polaski (2006); Sibbel/Borrmann (2007)

13 For research on authority see: Brühl et al. (2001); Graz/Nölke (2008b); Hall/Bierstecker (2002); Cutler (1999); Cutler (2002); Cutler et al. (1999); Fuchs (2005); For research on institutional design see: Gereffi et al. (2005); Koening-Archibugi/Zürn (2006); Pattberg (2004); Pattberg (2005); Pattberg (2007); For research on legitimacy see: Bernstein/Cashore (2007); Börzel/Risse (2005); Schaller (2007); Steffek et al. (2008); Wolf (2001)

14 See Section 2.1 for a definition of the concept of compliance and the concept of effectiveness and a discussion of their interrelations.

15 For a comprehensive overview of the literature on compliance see Mitchell (1996); Raustiala/Slaughter (2002) 8

16 For a comprehensive overview of the debate see Block et al. (2001); Lee (1997)

17 The US-Cambodian bilateral trade agreement is one of the rare historical cases of linking a social clause to trade.

18 For a comprehensive overview of the literature on global business regulation see Vogel (2008)

19 Deficits with regard to the quality, reliability and comparability of available data might lead to biased results. To minimize this risk, the study draws on different sources of information.

20 The term “shadow of hierarchy” refers to the capability of the state to intervene and to regulate through topdown decision-making. See Scharpf (1993)


ISBN (eBook)
ISBN (Book)
File size
832 KB
Catalog Number
Institution / College
Free University of Berlin – Otto-Suhr Institut für Politikwissenschaft
Public Private Partnership Private Governance Cambodia Apparel Labour Standards Social Standards Private Regulation Compliance Garments Governance Bangladesh ILO Codes of Conduct CSR Verhaltenkodizes Private Regulierung Arbeitsstandards Sozialstandards Kambodscha Bangladesch Bekleidungsindustrie Internationale Arbeitsorganisation Selbstregulierung




Title: How to Make Better Factories