Integrating Voluntary Sustainability Standards in Trade Policy: The Case of the European Union’s GSP Scheme

: Trade policy is increasingly being used as a policy instrument to pursue non-trade objectives such as environmental protection or the protection of labour rights. A key example is the European Union’s Generalised Scheme of Preferences (GSP). The current approach is being confronted with signiﬁcant challenges. How these challenges can be addressed is currently subject to debate, and increased attention is turning to the role that private governance mechanisms can play in this context. This paper will look into the potential role that Voluntary Sustainability Standards (VSS) can play. The paper will analyse and assess the complementarity between VSS and EU GSP, and it contributes to the literature on interactions between private and public policy-making for sustainability. The main research question focuses on what role VSS can play in the European Union’s GSP scheme. To answer the question, the paper develops two models by which VSS can be integrated in EU GSP (a mandatory and a voluntary approach). The study is based on interviews with key experts from different stakeholder groups and an analysis of the ITC standards map database. The paper shows that the integration of VSS in EU GSP, as outlined in the two models, is confronted with several challenges. These challenges are discussed, and alternatives are explored.


Introduction
Trade policy is increasingly being used as a policy instrument to pursue non-trade objectives, such as environmental protection or the protection of labour rights. This trend is increasingly being well captured by the term governing through trade, which is actively pursued by the European Union (EU) [1,2]. This governance through trade can be executed by providing additional tariff preferences as an incentive to ratify and effectively implement a series of international human rights, labour, environmental and good governance instruments, such as in the case of the Generalised Scheme of Preferences (GSP) regulation [3]. Whether this governing through trade is effective is still an outstanding question and a report by the European Commission [4] identified several challenges in the case of GSP+. How these challenges can be addressed is currently subject to debate, and increased attention turns to the role that private governance mechanisms can play in this context [5][6][7]. This paper will look into the potential role that Voluntary Sustainability Standards (VSS-also known as private standards, eco-labels, and sustainability certificates) can play in this context.
The focus on VSS emerges in the context of a rapid proliferation of VSS. According to the Ecolabel Index Database, the number of VSS has grown by almost 400% between 1989 and 2016, now reaching 465 VSS operating worldwide. The ITC Standards Map, which is more restrictive in recognizing VSS, counts 210 VSS [8]. These VSS aim to govern through trade, and to pursue similar objectives as the EU concerning sustainable development. Hence, in terms of policy design, they might be

EU's Current Generalised Scheme of Preferences
The European Union's Generalised Scheme of Preferences (GSP) is a preferential trade arrangement by which the EU grants unilateral and non-reciprocal preferential market access to goods originating in developing countries. The preferences are given in the form of the partial or entire suspension of import tariffs. The EU has operated a scheme of generalised tariff preferences since 1971 (Regulation (EU) No 978/2013 of the European Parliament and of the Council, applying a scheme of generalised tariff preferences and repealing Council Regulation (EC) No 732/2008, recital (1)), with the objective of "assist[ing] developing countries in their efforts to reduce poverty, [and to] promote good governance and sustainable development" [4] The scheme consists of three arrangements that distinguish between developing countries on the basis of their development status and needs, thus providing for different levels of preferential market access. The three arrangements are: (1) the general arrangement (the "Standard GSP"); (2) the special incentive arrangement for sustainable development and good governance, known as the "GSP+"; and (3) the special arrangement for the least-developed countries (LDCs), known as "Everything But Arms" (EBA). Standard GSP provides tariff preferences for a number of products. GSP+ expands this to almost 90% of export products, while EBA provides zero-tariff access for all products, except for products related to arms. A country might become a GSP beneficiary if it complies with certain conditions, including the ratification and implementation of 27 international conventions concerning human and labour rights, environmental protection, and good governance (see Table 1). GSP+ aims at spreading and promoting the values and principles of human rights protection, sustainable development, and good governance. The additional preferences are intended as a form of compensation, or reward, for having signed up to, and for implementing the relevant international law. In other words, the GSP+ "fosters the achievement of its goals by offering the 'carrot' of preferences" [4].
Compliance with the abovementioned conditions is continuously monitored by the Commission, primarily by way of examining the conclusions and recommendations of the monitoring bodies established under the relevant conventions, but also on the basis of information supplied by the beneficiary countries themselves, and by reference to other sources of information, including information submitted by third parties, or by the European Parliament or the Council [3]. Where a GSP+ beneficiary is in non-compliance with the regulation, the Commission may remove that country from the list of beneficiaries [13]. To date there was only one instance of temporary withdrawal of GSP+ preferences, namely Sri Lanka [14]. Apart from Sri Lanka, two other countries (El Salvador and Bolivia) have been investigated in the past by the Commission for failure to implement all of the conventions, but neither investigation ended with a suspension of preferences [14]. There are currently only nine GSP+ beneficiaries. Alongside the GSP+ mechanism for temporary withdrawal of preferences, the Regulation also features a common mechanism that is applicable to the other two arrangements Standard GSP and EBA. Similar to the GSP+ mechanism, preferences may be suspended temporarily with regard to Standard GSP and EBA, either for all products, or for certain covered products originating in a GSP beneficiary country. The grounds for suspensions include "serious and systematic violations of principles laid down" in certain of the 27 above-mentioned conventions (i.e., the 15 'core human and labour rights Conventions of the United Nations (UN) and International Labour Organisation (ILO)'). Aside from the withdrawal of Sri Lanka's GSP+ preferences, there have been only two other instances, in the Standard GSP, of temporary withdrawal of preferences (Myanmar and Belarus) [14].

The Compliance Gap
A debate has emerged on the effectiveness of the current GSP approach in fostering compliance with international conventions. The Commission review report on the implementation of GSP+ [15] shows improvement in the implementation of 27 conventions in the GSP+ beneficiaries, but also identifies challenges with the implementation on the ground. Also the recently published mid-term evaluation of the GSP regulation, which analysed the economic, social, human rights, and environmental impact in the GSP beneficiary countries, also finds indications of positive effects of the GSP scheme [16] The report notes that it is difficult, in general terms, to assess the impact on sustainable development. It does observe an economic impact in terms of export diversification, but this differs from country to country. The case studies (both sector case studies and country case studies) in the mid-term evaluation provide an indication that the scheme has had an overall positive impact on social development and human rights measured by different indicators but also found many instances of violations of labour and human rights (in Pakistan, Bangladesh, Bolivia, and Ethiopia). In addition, the mid-term evaluation only finds a limited impact on sustainable development and environmental protection. Several people consulted in the context of this study also indicated the presence of shortcomings in relation to the effective implementation on the ground (especially in the context of GSP+). However, for many countries, one can find several indications of violations in practice, resulting in a compliance gap. The compliance gap refers to the difference in the implementation of relevant international agreements in law, and in practice. While several respondents were positive about how GSP+ is currently operating, several recognized that there is room for improvement in the implementation of the GSP+ conventions. A recent report by the European Parliament showed many instances of non-compliance on the ground with several ILO conventions in export processing zones in selected GSP+ countries [17]. In a recent report by the Pakistan Workers Confederation [18], the compliance gap is also illustrated in reference to many ILO Conventions and their implementation in Pakistan. These observations correspond to a broader body of literature that emphasizes the difference between the formal ratification of international treaties and conventions, and their de facto implementation [19][20][21][22][23]. One key reason to further develop the enforcement design of the GSP-scheme would be to address this compliance gap. This compliance gap might be closed through several actions.
Especially relevant in the context of this paper are the VSS. VSS comprise a collection of organisations that certify producers and production processes, taking sustainability standards into account. The United Nations Forum on Sustainability Standards [24] defines VSS as "standards specifying requirements that producers, traders, manufacturers, retailers or service providers may be asked to meet, relating to a wide range of sustainability metrics, including respect for basic human rights, worker health and safety, the environmental impacts of production, community relations, land use planning and others." These standards are developed on the basis of general principles. These principles refer to a series of public goods, such as the protection of labour rights and biodiversity, and addressing climate change. This collection of VSS comprises many different initiatives. VSS have established themselves as significant governance mechanism to govern transnational economic activity and value chains [25]. The number of VSS has proliferated since 1990 [26]. In addition, the number of certified entities has grown rapidly in the last two decades [27]. Drivers for this growth include consumer demand and access to markets, including the requirement for some economic operators to obtain multiple certificates. This demand is driven by private actors (consumers, business-to-business, retailers) and public actors (through public procurement, integration of VSS in regulatory measures) [9].

Two Models to Integrate VSS in GSP
VSS organizations engage in three interrelated processes that are relevant in the context of GSP: (1) standard setting based on international conventions, (2) assuring compliance with standards, and (3) monitoring and evaluation. VSS can be regarded as instruments that implement and enforce international legal commitments on sustainability. They do this in a three-step approach. First, they explicitly link the sustainability standards that they develop to international legal commitments in their foundational principles. Many of these international legal commitments overlap with the 27 conventions identified in the GSP+ regulation, and the 15 conventions that can trigger a temporarily withdrawal in the case of 'serious and systematic violations of principles laid down" in these 15 conventions for GSP and EBA (These conventions are listed in Annex VIII:A (see Article 19.1(a) of the Regulation)). Second, they operationalise these principles into measurable indicators and actions. In a third step, they use a comprehensive set of procedures to monitor compliance with their sustainability standards, including the use of auditing and the provision of complaint systems. The first two steps relate to standard-setting activities. The third step relates to assessing conformity and monitoring and evaluation.
In this way, VSS might complement the current GSP scheme, and it could be integrated in it. VSS can be integrated in GSP/GSP+/EBA in two ways. The first option is to allow access to the European market, conditional on being certified, i.e., products need certification before they can enter the market and receive the GSP scheme preferential tariff rate. This approach would follow some regulatory approaches that the EU is currently pursuing, which provide market access only to certified products, such as in the case of the Renewable Energy Directive [28,29] and the EU Timber Regulation [30]. In this option, all, or a selected subset of, products entering the EU market will need certification before entering the EU market, in order to obtain the GSP preferential tariff. This could be seen as the 'mandatory' approach (see Figure 1). The second option is to give tariff preferences (lower tariffs or elimination of tariffs) to certified products. In this case, certified products would obtain additional tariff preferences (lower duties) compared to non-certified products. In this option, the EU needs to distinguish tariffs for certified and non-certified products. One tariff would be applied to certified products, while another tariff would be applied to non-certified products. In this scenario, all products would be allowed to enter the EU market under the GSP-scheme tariff, but certified products would get an additional financial advantage. A similar idea was proposed by the European Parliament (EP) Committee on International Trade, proposed in its opinion on the EU flagship initiative on the garment sector. (EP-INTA, 2017, §8). This could be regarded as a 'voluntary approach' (see Figure 2) The intent of these two options is not to replace the current system, but to add additional requirements that correspond closely to the commitments made by states in the GSP scheme, for economic actors exporting to Europe. Such an integration of VSS in GSP would add one additional element in the GSP procedure (the other three conditions are related to the rules of origin requirements.), namely, the verification of a recognized VSS certificate (both site-level certificates, as well as CoC certificates). The introduction of VSS would add one additional element in the GSP procedure, namely the verification of the use of a recognized VSS certificate (both site-level certificates, as well as CoC certificates). Firms exporting to the EU from GSP beneficiary countries will be required to have a recognized VSS certificate.

Assessment of the Integration of VSS in the GSP Scheme
Some authors argue [6] that the integration of VSS in GSP would have several advantages, namely, lower implementation costs, since implementation and enforcement is 'outsourced' to the VSS and the ability to truly govern 'beyond EU borders'. The EU checks that quality standards are being followed in the recognition process, but itself is not required to audit businesses in non-EU countries. By relying on existing VSS, the EU does not need to develop its own certification systems, which would be time-consuming and cumbersome, and can it rely on existing VSS that have been operational for a significant amount of time. These advantages are recognized in a growing body of literature on how VSS are integrated in public policy, and how they can contribute to co-regulation [9,25,31,32]. The fact that sustainability cannot be observed in products when they cross the EU

Assessment of the Integration of VSS in the GSP Scheme
Some authors argue [6] that the integration of VSS in GSP would have several advantages, namely, lower implementation costs, since implementation and enforcement is 'outsourced' to the VSS and the ability to truly govern 'beyond EU borders'. The EU checks that quality standards are being followed in the recognition process, but itself is not required to audit businesses in non-EU countries. By relying on existing VSS, the EU does not need to develop its own certification systems, which would be time-consuming and cumbersome, and can it rely on existing VSS that have been operational for a significant amount of time. These advantages are recognized in a growing body of literature on how VSS are integrated in public policy, and how they can contribute to co-regulation [9,25,31,32]. The fact that sustainability cannot be observed in products when they cross the EU

Assessment of the Integration of VSS in the GSP Scheme
Some authors argue [6] that the integration of VSS in GSP would have several advantages, namely, lower implementation costs, since implementation and enforcement is 'outsourced' to the VSS and the ability to truly govern 'beyond EU borders'. The EU checks that quality standards are being followed in the recognition process, but itself is not required to audit businesses in non-EU countries. By relying on existing VSS, the EU does not need to develop its own certification systems, which would be time-consuming and cumbersome, and can it rely on existing VSS that have been operational for a significant amount of time. These advantages are recognized in a growing body of literature on how VSS are integrated in public policy, and how they can contribute to co-regulation [9,25,31,32]. The fact that sustainability cannot be observed in products when they cross the EU border, but are largely based on production process characteristics, means that sustainability has to be assessed at the place of production [28]. To achieve this, VSS offer some distinct advantages, since they operate in the countries in which products are made. However, there are also several challenges that are related to the idea of integrating VSS in the GSP scheme.
In this part, we summarize the key issues that were raised when discussing the integration of VSS in the GSP scheme with more than 40 experts and stakeholders. Some of the issues apply to the GSP scheme as a whole, while others are specific for Standard GSP or GSP+. The aim of the discussions with the experts were to identify key issues related to the integration of VSS in the GSP scheme. These consultations were conducted under the agreement that we would not attribute statements to specific persons or institutions. On the basis of the identification of issues, we also consulted available sources, databases, and literature to make a further assessment of the issues raised. The main issues and challenges identified are: (1) a substantive overlap of standards, (2) impacts on the nature of the GSP scheme, (3) the dynamic nature of the GSP scheme, (4) VSS as an appropriate tool (effectiveness, governance, and capacity), (5) availability of the VSS, and (6) costs related to certification.
We discuss each of them and identify the relevance for the two possible options of integrating VSS in the GSP scheme. A first option (option 1) involves creating access to the European market, conditional on being certified, i.e., products need certification before they can enter the market. This approach would follow some regulatory approaches that the EU is currently pursuing, which provide market access only to certified products. This could be seen as the 'mandatory' approach. A second option (option 2) involves giving tariff preferences (lower tariffs or elimination of tariffs) to certified products. One tariff line would be applied to certified products, while another tariff line would be applied to non-certified products. In this scenario, all products are allowed to enter the EU market, but certified products obtain a financial advantage. This could be regarded as a 'voluntary approach'. Options 1 and 2 are applicable both to the Standard GSP, as well as GSP+ and EBA. For EBA, option 2 does not apply, since (almost) all tariffs are already eliminated. For each challenge, we also identify whether the concern applies for the overall GSP scheme, or whether they are more specific to Standard GSP, GSP+, and EBA.

Overlap in Substantive Provisions
In the context of integrating VSS in the GSP scheme, an important issue concerns the implementation of and compliance with core conventions in the GSP-scheme. For the entire GSP-scheme, this refers to the first 15 conventions that which can be invoked to trigger a temporary withdrawal of the GSP scheme in the case of violations of the conventions. For GSP+ specifically, it refers to the requirement of ratifying 27 conventions in order to be eligible for GSP+. The connection between VSS and the GSP scheme lies in the fact that both base their approach towards sustainable development on a series of international commitments (see also Table 1). Hence, the reason to consider integrating VSS in the GSP scheme originates from the idea that the effectiveness of the GSP scheme in achieving its objectives could be improved by integrating VSS. However, a first critical issue which was raised concerns the substantive overlap between the conventions in the GSP, and the social and environmental standards in VSS.
Most of the well-known VSS today integrate existing international legal commitments, often developed in a multilateral context, in their standard-setting procedures. Especially relevant in the context of this paper is that the VSS, to a degree, rely on the same conventions as those that are included in the GSP scheme. In order to explore, this we analysed the ITC database, which contains some information on which conventions are covered by which VSS. Fiorini et al. [8] provide a good descriptive overview and main statistics on the VSS in the ITC database. We screened the database for a selection of VSS that could be considered in the context of the GSP scheme. We selected ISEAL-Alliance members (20 in total), since they need to comply with a series of guidelines on how to set and enforce standards. These guidelines correspond to the procedural requirements that the VSS would be expected to meet, in order to receive accreditation under a public regulatory system which would use the VSS. Table 1 reports on which international conventions are referred to by a selected number of VSS in the context of their standard-setting procedures (N = 20 VSS). Table 1 shows that these selected VSS are based on public international law, and derive their standards from international conventions, but also that they do not cover all of the conventions of the GSP scheme. This partially corresponds to the findings of a recent study on VSS that shows that VSS use or incorporate international law to varying degrees [33]. Table 1 shows that several VSS do base their standards on some of these international conventions. However, they do not cover all of the international conventions in their standard-setting process. A few aspects need to be taken into account, in this context. First, some conventions fall substantially out of the scope of issues that sustainability standards typically deal with, especially the UN Conventions on narcotic drugs and psychotropic substances. Hence, it will be difficult for VSS to cover the whole scope of conventions that are included in the GSP regulation. In addition, some of the conventions target more states than organizations or firms, such as the UN Convention against Corruption, or the Convention on the prevention and Punishment of the Crime of Genocide. Hence, these conventions are less suitable for further enforcement through VSS. Second, for those conventions falling within the scope of VSS and not covered yet within a specific VSS, it should be noted that VSS that did not yet include elements of some of the conventions can still do this. Most VSS have open and consensus-based standard-setting procedures that allow them to develop new standards in relation to other conventions. VSS are dynamic systems that align standards with changes in the regulatory environment. Hence, the integration of VSS into GSP might constitute an incentive for VSS to also include references to additional conventions in their standard-setting practices.

Impact on the State-to-State Nature of the GSP Scheme
One of the arguments to include VSS in the GSP-scheme is that it would allow a more differentiated approach towards fostering compliance, since it would allow for a more graduated systems of incentives for firms [34]. As Schukat & Rust [6] made clear, it allows for a targeted approach towards individual firms. Under the terms of the current GSP scheme, states that violate the conventions can be confronted with a suspension of tariff preferences. Such a suspension implies punishing all companies, including those whose production systems already comply with social and environmental criteria, and with some of the relevant conventions. The VSS can directly promote the implementation of sustainability criteria, laid down in the conventions, at the level of economic operators, without either granting or withdrawing tariff preferences in a blunt fashion for all economic operators in that state. Hence, one key argument would be that by integrating VSS in the GSP scheme, not only states, but also economic operators need to comply. This is not to suggest that certified economic actors can continue to export if a beneficiary state is excluded from the GSP-scheme due to non-compliance, but that additional requirements are added, which address economic actors. Currently, the system, in the case of non-compliance, heavily relies on either inter-state dialogue or a hard sanction tariff preferences withdrawal. The latter option is hardly used, and it is not considered to be effective by some. As a result, a recent EP report on labour rights protection in GSP+ countries [17] stresses that "finding a middle ground between 'soft' inter-state dialogue and 'hard' nation-wide sanctions is crucial to enhancing the EU's role in realizing labour rights in GSP+ beneficiaries." VSS could facilitate finding this middle ground through involving economic actors, by rewarding access to the EU market for only those economic operators that comply with the relevant requirements (option 1), or who provide additional tariff preferences to only those economic operators who comply with the relevant requirements (option 2). In this context, the introduction of VSS that which target economic actors more directly could develop a dynamic in which economic actors who need (option 1) or want (option 2) to participate in VSS push governments to pursue compliance with all economic actors. Not only might economic actors influence internal dynamics, but also influence other actors, such as non-governmental organisations. Some respondents noted that the integration of the VSS could create more space for civil society actors to become more active and contribute to creating institutional reforms within countries since several VSS engage local stakeholders in monitoring and enforcement of standards. This might in turn lead to further capacity building of civil society in countries in which VSS are active.
However, according to other respondents, integrating VSS in the GSP scheme could fundamentally change the nature of the system. The EU GSP scheme is designed as a unilateral instrument between the EU and specific developing countries, with the aim of promoting development across a variety of economic, social, and environmental domains, and based on state-to-state dialogue. The idea is to build capacity in those countries through trade. The system is built on state-to-state dialogues and cooperation. The VSS, on the other hand, focus on the level of firms, not states. Some respondents indicated that the two operate "according to different logics" and target different actors. If the VSS are incorporated as an additional compliance mechanism, not only states, but also economic operators would be responsible for achieving compliance with the GSP requirements. This might make it more challenging for the EU to hold the government accountable, since they could shift the burden of responsibility to private actors. Some respondents indicated that the integration of VSS could 'hollow up' the responsibility of the state, since the responsibility will be split between actors. This refers to a dynamic in which private regulatory action replaces public regulatory action (see also reference [9]), and which could create a dynamic in which the state capacity to deal with the issues and conventions is weakened, rather than strengthened. These arguments hold mainly when option 1 is applied (making access to the European market mandatory on being certified), and not for option 2, since in option 2, the adoption of the VSS is voluntary, and certified products get a reduced tariff. In the context of option 2, reduced tariffs for sustainable products would be an additional benefit, and they would not alter the current systems significantly. More generally, this discussion, for both options, underlines the importance that the integration of the VSS into the GSP scheme should be seen as a complementary measure, not one that would in any way substitute the current way of working, based on state-to-state dialogue.

Dynamic Nature of the GSP Scheme
Another identified issue concerns the dynamic nature of the GSP scheme. Countries move from one system (GSP) into another trade regime, such as a bilateral agreement or no special trade regime at all [16]. Table 2 focuses on GSP+, and shows the current and previous GSP+ beneficiary countries since January 2014. Up till now, there were 16 countries that had benefitted from the GSP+ since the January 2014 reform. Seven countries left the scheme because they entered a bilateral trade agreement. If the requirements for VSS was solely required in the context of the GSP+, it might create disproportionate costs to comply, if these requirements no longer hold under other trade regimes of the EU, especially if option 1 was applied. This concern is not only valid in comparison to bilateral agreements, but also in the context of the GSP scheme. Some respondents voiced concerns that the integration of VSS in GSP+ and Standard GSP might create disadvantages for GSP+ or Standard GSP beneficiaries, compared to beneficiaries that fall under the EBA scheme. In this context, reference was made to countries such as Bangladesh, who also compete with GSP+ countries on some export commodities.
It is also important to put the GSP scheme in perspective, in terms of its overall contribution in EU's trade and imports. The importance of GSP for import in the EU market is low and decreasing [16]. In 2016, more than 82% of imports entered under WTO Most Favoured Nations arrangements, and only 4.1% under GSP. Hence, if one aims for ambitious targets to be reached in terms of sustainable products (for example in relation to sustainable textiles), a reform of the GSP scheme will only be able to contribute marginally. However, it should also be noted that for some GSP beneficiaries, their share of exports to the EU is important, and for them, the GSP scheme constitutes an important mechanism. For example, Pakistan (GSP+) and Bangladesh (EBA) gain significantly from GSP+ and EBA, respectively, as both count the EU as one of their most important trading partners. Similarly, in both countries, the textile sector constitutes a major portion of exports. The EU accounts for 43.3% of Bangladesh's exports [16]. Exports to the EU have increased by 50% since 2013, and by almost 80% since 2011. Approximately 91% of Bangladesh's exports to the EU-under the EBA, comprising of textile related products [16]. Furthermore, both countries have a high utilization ratio (see below) for the EBA and GSP+, meaning that a majority of their products that are eligible for preferential treatment utilize these benefits. The objection of the dynamic nature of the GSP scheme only applies for option 1, since economic operators would be required to obtain certification in order to access the European market. This objection does not apply for option 2, since in this case, VSS certification remains voluntary. In option 2, economic operators will make an assessment of whether it is worthwhile for them to obtain VSS certification in the context of their wider export strategy. In the case of option 2, one could contemplate also using additional preferences for sustainable products in FTAs.

Limitations on the Possibilities for Further Tariff Differentiation
For option 2 (tariff preferences for certified products), the issue was also raised for whether there was sufficient scope to lower tariffs, in order to make the system work. Currently, an analysis of the tariff lines used for Standard GSP, GSP+, and EBA shows that a very high percentage of tariffs under both the EBA and GSP+ arrangements have received duty-free treatment in 2014, with over 97 per cent under the GSP+ regime, leaving very few tariffs to be differentiated in terms of different tariffs for certified and non-certified products [16]. This evidence indicates that tariff differentiation (option 2) for certified products in the GSP-scheme has different implications for the different schemes. Concerning GSP+ option 2, including granting zero duty for certified products can only be applied to a small part of the export and products. If one was to consider pursuing option 2 further in GSP+, one would probably need to re-introduce tariffs. One can think of using the tariffs under the general GSP scheme and giving a zero tariff to the certified products. Theoretically, this could be done. However, on the basis of interviews conducted for this study, this option would be politically very sensitive, and some interviewees indicated that this would undo the purpose of GSP+. On the other hand, it could be argued that even if the number of products to which this could be applied is small, it still sends an important political message on the importance of sustainability. Concerning Standard GSP, the application of option 2 would hold more promise, since there is more scope for tariff differentiation, and more products could benefit from additional tariff reductions if they are certified. Option 2 would not be applicable to EBA, since there is no scope for tariff differentiation.

Effects on the Utilisation Rate of GSP Tariffs
The observation was also made that the application of the VSS requirement (option 1) might actually have a negative effect on the GSP scheme utilisation rate. The utilisation rate indicates preferential imports as a percentage of eligible imports under a trade agreement or the degree of usage of existing preferences [35]. Hence, it is the share of imports that are actually imported under GSP scheme preferences over all imports that are eligible in principle. Not all imports that are eligible for a preference will always be imported under a GSP scheme, because of several reasons: not all individual shipments would fulfil the rules of origin, and preference is not claimed or granted for a specific reason. Currently, utilization rates for GSP+ countries are very high. For Pakistan, Paraguay, Cape Verde, Bolivia, and Armenia, the utilization rate is above 90% [16].
High utilization rates indicate that the preferential rates are used significantly by GSP+ beneficiaries, indicating that GSP+ is reaching its objectives in terms of export support. Applying VSS according to option 1 might result in the fact that the utilization rate goes down, since economic operators may not be able to prove that they fulfil the criteria, may choose not to do so because of the costs involved in becoming certified, or they would rather pay the WTO-MFN duty rate. Especially when the costs of certification would outweigh the benefits of tariff reduction, economic operators might enter the EU market without the tariff preference. This might hollow out the purpose of GSP+. This concern applies especially for option 1 (economic operators would choose to export to Europe but not use the GSP-scheme). Similar concerns hold for Standard GSP and EBA. The utilization rates are already significantly lower and decreasing for Standard GSP beneficiaries [14]. For EBA beneficiaries, the utilization rates are currently high and increasing [16]. In both cases, the application of option 1 would in most likelihood result in a decrease of the utilization rate. For Standard GSP, the same argument holds for option 2, as is the case for GSP+, while option 2 cannot be applied to EBA countries.

VSS as an Appropriate Tool
In the interviews, several issues were raised in relation to the VSS and how they operate. Four challenges were identified concerning VSS: (1) the effectiveness of VSS, (2) the governance of VSS, and (3) the capacity of VSS and (4) the availability of VSS for export sectors.

Effectiveness of VSS
An argument for the integration of the VSS in the GSP scheme is that it offers an additional enforcement mechanism. The question then focuses on the effectiveness of the VSS. This issue was raised in several interviews. Some respondents questioned the effectiveness of the VSS and their potential contribution to closing the compliance gap. It is impossible to make a general assessment on the effectiveness of the VSS, since this depends on the VSS that are studied, and which parameters are analysed in the context of impact, i.e., socio-economic parameters such as poverty reduction, compliance with labour rights, etc., or environmental parameters. In this context, it is important to highlight that many studies on the impact of the VSS have been published. They provide a basis for considering the inclusion of the VSS in public policy. This is not to argue that all studies only found positive impacts. There are many critical studies on the effectiveness of the VSS as well. Throughout the literature, one can identify several different types of impacts on different levels for different VSS, and in relation to different products. Concerning agricultural commodities, several meta-studies have been published, which summarize the main results of several other studies. Two leading sources are literature reviews covering the most relevant impact studies for agriculture-related VSS across standards, commodities, and regions [36,37]. The Food Agriculture Organisation literature review [36] covers 101 studies, out of which 30 studies were deemed to be sufficiently rigorous. The ITC literature review includes 47 studies [37]. These studies focus on the impact on the ground for standard-takers on multiple dimensions of sustainability. The first dimension focuses on the level of farms and farmers, and investigates the impact on economic aspects of sustainability, such as profitability for farmers, impact on yield and quality, and market access and technology transfer. The second set of studies focuses on the social dimension of sustainability, and provides an impact assessment on parameters such as wages and assets, gender equity, health and education, and environmental sustainability. The final set of studies provides an assessment on environmental impacts covering topics such as resource management, soil conservation, biodiversity, and water quality. These studies show that on balance, certified farms perform better than non-certified farms, but that it depends on which VSS scheme is analysed, and in which country's context. Other areas in which significant impact studies have been published, also with to a certain degree of mixed results, are in the area of forest governance and forest-based products. Here also, there is evidence that certified forests (forest-management certification) and chain of custody actors perform better than non-certified entities (see [36] for an overview and discussion). The most critical studies on the effect of certification have been published in the context of the textile sector and the enforcement of labour rights; see for example [38][39][40][41][42]. Several respondents in the interviews pointed to the Rana Plaza disaster and the Ali Enterprises fire in Karachi to make the argument that auditing and certification by VSS is not always effective. In both disasters, companies were involved who were certified. The criticism on the effectiveness of VSS in textiles is also documented in academic research. They have been criticized on several grounds concerning their impact [42,43]. However, Locke [42] also shows that some VSS perform quite well in the context of specific countries.
In summary, there is a debate on the effectiveness of the VSS. It is clear that the VSS potentially generate positive impacts and contribute to closing the compliance gap. However, it should also be noted that not all studies univocally indicate a positive impact on the socio-economic and ecological indicators for all VSS, on all products, in all countries. How to assess this in general terms is difficult, and it is to a degree a function of the appreciation for what one can achieve with the VSS. It is safe to assume that VSS with a well-elaborated enforcement structure, and that cover a number of GSP conventions, will perform better in terms of impacts, than VSS that lack this.

Governance of VSS
Several respondents raised the issue and concern of the credibility and governance of the VSS, and indicated that a system should be put in place which checks the robustness and credibility of the VSS, including the process of setting standards, the assessment of the standards, and the governance of the scheme. Given the proliferation of the number of VSS, one needs to distinguish credible from non-credible systems, and assess the diversity among systems. The VSS differ significantly in how they are governed, how accountable they are, how effective they are in enforcing the standards they develop, etc. Many comparative studies have highlighted the difference between VSS [33,40,[44][45][46] and the potential effects on effectiveness and legitimacy.
The diversity in the design of VSS [33,45] indicate that not all VSS are equal, and a recognition system to distinguish credible from non-credible VSS is necessary. This concern should be addressed in a recognition system that needs to be set up to recognize VSS in the context of a regulatory scheme and that distinguishes credible from non-credible VSS. This recognition system should be developed by the EU. Some respondents expressed the preference that the EU should build upon existing benchmarking schemes, as identified above, and support those, instead of duplicating approaches. If the EU were to recognize the VSS, it might create duplication in a context in which several actors are currently certifying or rating the 'certifiers'. On the other hand, it could also be argued that if the European Commission recognizes the VSS, this might have a strong influence and would de-facto regulate the market of VSS, and distinguish credible from non-credible VSS.
When questioning respondents on key elements of such a recognition system, several experts referred to the following requirements: a clear link to the conventions, an open and consensus-based standard-setting process that further operationalizes principles into standards, the presence of third-party auditing, the presence of public disclosure of audits (transparency), and the presence of complaint systems. Some respondents indicated that this recognition system should be very stringent, which might result in the fact that only very few VSS are recognised. For some products, no VSS might be recognised, which might influence the scope of the application of options 1 and 2.
The concern for developing a robust recognition system was also voiced in the context of not creating a race to the bottom in international standard-setting with regard to sustainability. If too many VSS were to be accredited, producers would have a choice between stringent and less stringent standards especially in a context where the business case for becoming certified still needs to be established, or even in a context where the business case for becoming certified disappears, because so many producers are certified. The later can occur when certified products are no longer be sold on the market as being certified, and hence distinctive, since many products are certified. Some noted that it will be difficult to create a level playing field among recognized VSS since there is no consensus on some of the standards.
A final governance issue that emerged in the discussions concerned the traceability through the supply chain. Some VSS have different systems to govern their supply chain [47]. Several respondents raised questions on whether it is possible to apply the VSS to the whole supply chain. Especially in relation to textiles and garments, this might be difficult. Producers in GSP/GSP+ countries source sometimes to thousands of homeworkers. Lund-Thomsen and Nadvi [48], for example, show how in the region of Sialkot in Pakistan, companies in the sportswear and apparel sector use thousands of subcontractors, who in turn outsource to a few thousand homeworkers. In addition, it should be noted that since VSS sometimes only certifies part of the supply chain, i.e., the stage at which it is produced, or converted, it does not guarantee that that the inputs to the production or conversion part of the value chain are necessarily sustainably produced. A t-shirt can be sustainably produced by labour receiving fair wages and working under safe conditions, but there is no guarantee that the cotton used in the process came from farms that similarly respected the labour laws. While these arguments have some merit, it should be noted that the proposal to integrate VSS in GSP could be limited to the trade-link between the EU and the GSP/GSP+ beneficiary country. Hence, only the economic operator in the GSP/GSP+ country should be certified, regardless of the position in the supply chain. In other words, the economic operator in the GSP/GSP+ country who exports to the EU needs to be certified. This could be a cotton farmer exporting cotton to the EU, or a sewing company making T-shirts for the EU market.

Capacity of VSS
An integration of the VSS in the GSP scheme might generate capacity problems with some VSS, since it will spur demand for VSS. This issue is most problematic for option 1. Although the VSS recognize accreditation organisations, which in turn accredit certification bodies to perform the conformity assessment with standards, some VSS remain actively engaged with certified entities, in order to ensure compliance with standards, both in terms of monitoring, as well as in terms of handling complaints and disputes. This implies that some VSS can only certify a limited number of companies, and they are not necessarily interested in certifying as many as possible. An interesting example in this respect is the Dutch Fair Wear Foundation (FWF), which was created in 1999 and is a VSS that is active in the textiles and apparel sectors; for a more elaborate discussion, see [1,43]. In both the substance of its standard and its verification procedures, the FWF aims to be the "gold standard" [49]. FWF focuses on the management practices of firms, and on labour conditions in their first-tier supplier factories [50]. Firms not only sign up to the FWF Code, but they actually become "members" of the FWF, clearly committing to its values. In addition, the FWF requires members to actively engage with their suppliers (other independent firms), while taking into account that members "have influence, but not direct control, over working conditions" in supplier facilities [50]. In terms of monitoring and implementing the Code, FWF has sought to shift from factory auditing alone to a "multilevel verification process" that utilizes inter alia factory audits and complaints procedures. This approach exemplifies a shift from a regulatory process that involves only an independent third-party auditor, to a system involving a more complex governance structure in which several dependent and independent actors provide information, expertise, and capacity in the certification process and fostering compliance with standards [43,51,52]. This potential capacity problem is quite important in the context of integrating VSS in GSP+, since the different governance models of VSS imply the different capabilities and capacities of the VSS.
A VSS governance model that only involves independent certification bodies for granting certificates and dealing with complaints can probably quite easily deal with a significant increase in demand. However, some respondents also identified challenges for independent certification bodies if the uptake of certain VSS increases quickly and strongly. Key challenges identified were that (1) there is probably too little time to conduct a proper and correct audit, (2) the availability of qualified auditors and ensuring that auditors are competent, and (3) they might not be able to cover so many producers. Other governance models employed by the VSS, as illustrated above, will have greater difficulties in dealing with the increased demand for certificates when whole markets within one country are becoming certified. There are two issues to be considered in this context. First, the capacity problem only really emerges in option 1 which would have significant impacts on the adoption required by economic operators. In option 2, capacity issues might be addressed more gradually, following the pace of economic operators who might come to benefit from additional tariff reductions. Second, the capacity problems might, to a certain degree, be addressed using accompanying measures such as capacity-building support for VSS in GSP-beneficiary countries.

Availability of VSS and Applicability to Export Sectors
For both options 1 and 2, there is an issue of which products could be included in such an approach. As noted, theoretically, option 1 and 2 could be applied to one or more products, ideally to as many as possible. In order to obtain a sense of which products could be included, one needs to have a look at the major export products, and whether VSS are available for these products. In order to assess whether VSS are available for key export sectors, an analysis of the ITC Standards Map database was conducted. Table 3 provides an overview of how many VSS are available per country, per export product category. For all GSP+ countries and sectors available in the ITC database (N of VSS = 235), we calculated how many VSS certify products per sector/country that are exported to the EU. In other words, Table 3 provides an overview of how many VSS are active in a GSP+ country, and which certify export products in a specific country. Table 3 summarizes the overview. The table shows that for all sectors in the ITC database, several VSS are operational in the GSP+ countries, i.e., VSS are currently certifying products in GSP+ countries, which are exported to the EU. This indicates (1) that VSS are already operational in the GSP+ countries, and (2) that they are able to cover a wide range of products.  Armenia  9  16  11  10  11  12  9  10  14  11  14  11  9  11  10  9  11  Bolivia  13  27  21  20  20  18  17  14  22  18  22  16  17  17  17  16  17  Cape Verde  8  13  10  10  10  12  10  9  10  10  10  11  9  11  10  9  12  Mongolia  10  16  12  12  12  14  12  11  14  12  14  13  11  14  12  11  14  Pakistan  15  28  18  18  18  20  19  16  20  17  20  18  18  24  18  17  20  Paraguay  14  29  21  21  20  19  18  15  21  18  21  17  18  19  18  17  19  Philippines  14  28  21  21  20  19  18  15  22  18  22  17  18  23  17  9  18  Kyrgyz Rep  8  15  10  10  10  13  10  9  11  10  11  11  9  11  10  18  12  Sri Lanka  14  27  20  20  19  20  19  16  19  16  19  17  18  23  19  19  18 Source: created by authors on the basis of ITC database-note: some VSS are active in multiple sectors.
We cannot distinguish how many of these VSS would be recognized by a stringent recognition system, but there are potentially several VSS per product category that are contained in the ITC Standards Map.
For option 1, this means that there are potentially multiple VSS which could cover the main export sectors. For option 2, there is the additional question of whether for these export sectors there are still tariffs which could be lowered. We already indicated that for GSP+, most tariffs are already eliminated. If one were to consider re-introducing tariffs in order to make a differentiation in tariffs between certified products and non-certified products, one could contemplate using GSP tariff lines for non-certified (also for GSP+ countries), and zero tariffs for certified products. The availability of VSS for export products might not be an issue, and neither the identification of products to which it can applied. One can identify multiple VSS for the major export sectors/products. However, some respondents raised two concerns in this context, (1) broad-sector categories do not always accurately cover what is possible on the product level, and (2) there is the problem of 'over-certification'. Concerning the former, tea is an interesting example. There are many types of tea, and for example, Sri Lanka (a GSP+ country) is a major tea producer. However, they mainly produce specific types of tea for consumers from Russia, Turkey, the Middle East, and United Arab Emirates, which do not really demand certified tea. The type of tea demanded by the European Market is not available in Sri Lanka, due to ecological and geographical factors [53] Concerning the latter ('over-certification'), the VSS do not necessarily have the intention of certifying as many companies or producers as possible, but rather to create 'certified markets' in which consumers demand certified products and consciously buy certified products. Hence, a key challenge for them is to have enough downstream players/buyers in consumer markets (i.e., in Europe). The option 1 proposal of integrating VSS in the GSP-scheme might lead to the situation in which producers need to be certified, and buyers need to buy certified products due to the regulations. However, not all products will be sold or marked as such on the market. It was indicated in the consultations that there is currently already an oversupply of some certified products on the markets, and some certified products are being sold without certificates. A measure that would influence adoption on a large scale might defeat the purpose of some of the VSS. This problem might especially arise in option 1, which would make market access conditional on being certified. Option 2 will also create additional incentives to get certified but this-due to the voluntary nature-might generate far less market demand for certificates.

Cost Issues Related to VSS Certification
A final issue concerned the costs of the proposed change. Integrating the VSS in Standard GSP, GSP+, and EBA could imply additional costs, since the VSS aim to internalize external costs, such as the costs related to environmental and social protection. It is impossible to obtain an estimate of the additional costs involved, which would also depend on several parameters, but we can identify where additional costs might emerge. The identified additional costs are associated with the producers in the GSP/GSP+ countries, customs, and possible the VSS themselves. For economic operators, two additional costs might be involved. One cost relates to becoming certified. The other costs are related to the increased costs for production in compliance with the VSS standards. Concerning the former, becoming certified often implies additional investments, technical changes, and corrective actions, in order to be in conformity with the standards. These costs can vary significantly, depending on the steps that are still necessary in order to obtain the certificate. Some studies indicate that these costs might be very substantial. In addition, there are costs that are related to recertification, which need to be taken into account. Concerning the latter, some studies [54] show that production according to VSS is often more expensive. These additional costs might off-set any benefits such as lower tariffs, price premiums, or market access. The cost issue, especially for small exporters, was flagged by several respondents, who then also expressed the idea that governments and the EU should consider funding/financing certifications (especially for smallholders) and invest in capacity building in the area of sustainability certifications. These additional costs might not be offset by any price that premium producers might receive due to certification. In some cases, the rationale for investing in certificates is that they generate more income due to the higher prices for products. If certification becomes a requirement (option 1), this price premium might not be there. Some already doubt that there is a price premium for certified products [54]. This effect does not affect, or affects option 2 to a lesser degree.
These costs can influence the ability of producers to comply with or adopt VSS, and possibly enter certain (export) markets. This might especially be the case for LDCs. Several VSS have tried to get LDCs on board, but some evidence suggests that this is hard to do because of the costs for certification. There is no general data available on the adoption rates of VSS but Marx and Wouters [55] explored the issue for GlobalG.A.P and FSC. They found that the majority of certified entities are based in high-income economies, while hardly any are based in low-income economies. A total of 79% of GLOBALG.A.P. and 74% of FSC CoC-certified entities are based in high-income countries compared to 8% and 4% in lower-middle and low-income economies combined. For FSC FM certification, the distribution is a slightly more spread out, with the exception of 2% in low income economies. Marx and Cuypers [56] refer in this context to the 'stuck at the bottom' problem, which refers to the fact that some countries, especially LDCs, are excluded from VSS dynamics. The latter is confirmed in a paper by Tayleur et al. [57]. They focused on seven agricultural commodities (coffee, banana, cocoa, oil palm, sugarcane, soy, and tea) and received data from many leading VSS about which farms and plantations were certified, and where they were located. The paper looks into the question of where certified commodity crops are located, and whether they are located in places that are important for conserving the world's most important biodiversity, and benefiting the most vulnerable producers. The paper finds that certification appears to be concentrated in areas that are important for biodiversity conservation, but not in those areas that are most in need of poverty alleviation. They also find that global levels of certification are low, since they were able to plot the certified areas on the total area on which the specific commodities were cultivated. This analysis shows that 9% of coffee production is certified worldwide. For the other six commodities, the proportion of certification is lower (banana: 0.3%, cocoa: 2.2%, oil palm: 2.2%, sugarcane: 0.6%, soy: 0.2%, tea: 2.0%). These results indicate that it is very hard to comply for economic operators in LDCs, and hence the certification option might not be possible for EBA. The low level of participation of entities in developing countries provides support for the literature, which argues that "smallholders need to be organized in commercially viable arrangement to be able to participate in certified value chains" [53]. Even producers in those countries that are producing sustainably might not have the capacity to prove it. This generates several challenges in terms of capacity-building and facilitation of compliance with VSS standards. Especially in relation to agricultural commodities produced by smallholders in GSP+ countries, this might create problems in achieving certification. VSS systems are aware of these problems, and they are taking action that are sometimes supported by external partners, such as international donors [58].

Discussion
The interaction between public and private policy instruments is gaining academic attention [9,11]. Several scholars see quite some potential in these interactions in order to govern transnationally. In this paper, I investigate this potential in the context of EU trade policy. The issues identified above signal quite some challenges for integrating VSS in the GSP scheme in the two options developed. The overall conclusion is that it might be difficult, if not impossible, to integrate the VSS in EU trade policy. It might also create undesirable trade-offs in policy interaction, as identified by Auld [12]. These trade-offs can both effect EU trade policy, as well as the VSS themselves. From the perspective of EU trade policy, the analysis shows that the integration of the VSS in EU trade policy might negatively affect the utilization rate of GSP, which would result in a lesser use of GSP, and hence defeat the objectives on which GSP is built. From the perspective of VSS, the analysis shows that VSS should change in order to better fit within EU trade policy, in order to be recognized as policy instruments. The standards should better align with the conventions that are included in GSP+. Although this is possible, it also implies that VSS have to change in order to align themselves and to accommodate one specific policy instrument, which might make them less suitable for an interaction with other policy instruments. Hence, the paper clearly shows that there are trade-offs in the integration of private and public policy instruments, in the case of trade policy.
This does not imply that integrating private instruments, such as VSS, in trade policy, is not possible. During the consultations, two other suggestions were made to integrate VSS in the GSP scheme, namely (1) by using the VSS as information providers in the monitoring processes, and (2) by including the VSS as targets in roadmaps for the implementation of the conventions.
First, the VSS could act as information providers for monitoring purposes. In this context, VSS can play a meaningful role, since they have extensive monitoring requirements, including audits of firms that provide ground-level information on compliance with standards. Table 3 showed that many VSS are active in GSP+ countries. Some of them will make their audit reports publicly available following their transparency policies. GSP+ reporting could use this information of reporting purposes. Hence, instead of only relying on reports by multilateral organisations and European missions, the GSP+ monitoring team could consider the use of other information sources, including those of the VSS. The VSS could provide at least two types of information; one is based on the audit reports, and the other one is based on possible complaints against companies in the country. The audit reports could provide a good overview on what happens in the factories, or on plantations and farms. They might actually also give some sense of trends within a country, in terms of better or worse compliance. Since the European Commission is interested in these types of trends, VSS might provide longitudinal data on what happens in the factories. The complaints system of VSS might provide insights into non-compliance issues within specific firms, and how they might relate to more systemic weaknesses, for example, the absence of existing legislation, which is relevant in the context of the GSP+ conventions. Finally, actors involved in VSS certification in GSP+ countries might be included as stakeholders in stakeholder dialogues on progress, in compliance with the relevant conventions.
Second, one could further reflect on the idea of linking roadmaps to GSP implementation, and including the VSS in these roadmaps. Some respondents proposed that GSP should be reformed in such a way as to include roadmaps for the implementation of the conventions. For each beneficiary, a roadmap would be developed that would contain specific targets that are related to the conventions. These benchmarks can focus on legal reform, as well as on different mechanisms (public, public-private, private) to implement legislation and to contain specific commitments. Such roadmaps are already used in some cases, such as the EU Bangladesh Sustainability Compact. In this initiative, there is, for example, a commitment by multi-national enterprises and retailers to engage in discussion on responsible business conduct along the supply chain [4]. In such roadmaps, commitments to engage with (a selected/recognized number of) VSS could be included, and they could provide incentives for firms, mainly large firms, to engage with VSS or to have their suppliers engage with VSS. Each year, an assessment can be made on the progress towards certification, as well as an assessment to what degree this approach actually contributes to better compliance. Such an approach should not replace any state commitment to enforcing commitments to conventions, but can complement these commitments and create forms of co-enforcement [59]. This can also strengthen state-civil society-firm ties, which could further facilitate compliance with legislation and voluntary standards [60]. Such an approach would be of a very different nature, but it might support the gradual expansion of VSS in GSP+ beneficiary countries.

Conclusions
This study focused on the question of whether VSS can be integrated in the EU GSP scheme. VSS receive increasing attention as a complementary tool to strengthen policies that pursue sustainable development [10,61]. In this study, we explored two ways of integrating the VSS in the GSP scheme as an additional requirement. In this approach, GSP beneficiary states should still have to comply with part or all of the requirements in the GSP regulation. An additional requirement that can take the form of options 1 or 2, would apply. The first option is to make access to the European market conditional on being certified, i.e., products from GSP beneficiaries will need to be certified before they can enter the market. The second option is to give tariff preferences to certified products. In this case, certified products would receive additional tariff preferences (lower duties) or tariff elimination if they are certified. For both options, one needs an approach to recognize VSS, which would be eligible under Standard GSP, GSP+, and EBA. Such a recognition should be built on three types of requirements: substantive requirements (which make reference to the relevant GSP conventions), procedural requirements (which includes monitoring and complaint provisions) and supply chain requirements.
The assessment of these options is that they are confronted with challenges. Most of these challenges apply more to option 1 then option 2. If an integration of the VSS in the GSP-scheme according to option 1 or 2 might not be feasible, a third option of integrating the VSS in GSP might be considered. This third option, of a very different nature, would be to integrate the VSS in the reporting mechanisms of GSP, either by relying on the relevant information by the VSS, or by making an assessment of the adoption of the VSS in GSP countries part of the periodic evaluation of the GSP via a scorecard or roadmap. Such an approach has some advantages. First, it would not entail a major reform of the GSP scheme. Second, it would remain voluntary, but governments could be encouraged to promote the adoption of VSS and to provide incentives. Third, it would be a far more gradual introduction of the idea of VSS and sustainable production, which might set GSP beneficiaries further on a trajectory towards sustainable development and production.