Norm contestation and the weakening of migrant fisher protection in Thailand

In 2019, Thailand ratified the International Labour Organization Work in Fishing Convention to protect migrant fishers. The Thai government vigorously promoted the Convention among its constituents. Yet, despite the government ’ s proactivity, such norms have been weakened upon local translation. This phenomenon is not readily explained by existing norms research that pays scant attention to norms in the domestic space. To addressthis oversight,thisarticleunpacksdomesticprocessesofnormvalidationandmakes a two-step argument. First, it argues that norm validation is crucial for instigating domestic implementation by allowing stakeholders to deliberate and contest the ‘ appropriate ’ application of norms. Second, it argues that the locally acquired normative understanding, as influenced by norm antipreneurs, or actors who defend the status quo, weakens norm implementation while the norms ’ discursive strength is retained. In presenting the findings, this article contributes to the norms literature by illuminating a new life given to validated norms. The contribution also interrogates norms ’ global – local dynamics by scrutinizing norm contestation and the extent to which it generates an internalized sense of legal obligation to implement international norms.


I. Introduction
Thailand ratified the International Labour Organization's (ILO) Work in Fishing Convention (hereafter the Convention) in January 2019. In doing so, it became the first Asian country to adopt legal norms to protect migrant fishers. The Thai government actively conveyed its support for the Convention at international fora and promoted it among domestic constituents in order to deliberate norms for domestic implementation. As captured by a coalition of 30 Thai and international non-governmental organizations (NGOs), 'These consultations have paved the way for the successful ratification of this critical convention' (ITF 2018). However, despite the government's serious efforts, such norms have been weakened upon domestic translation. This phenomenon is not readily explained by existing norms approaches in IR, such as in the rhetorical trap perspective (Mumford 2021;Schimmelfennig 2001), de-coupling approach (Hafner-Burton and Tsutsui 2005;Jetschke and Rüland 2009), spiral model , localization (Acharya 2010) and norm implementation (Betts and Orchard 2014). As discussed further below, these perspectives are inadequate for capturing the weakening of migrant fisher protection in Thailand, and therefore warrant greater scrutiny of international norms upon entering a domestic space.
To address the shortcomings in the existing norms literature, this article unpacks the domestic processes of norm validation. To do so, it makes a two-step argument. First, it argues that the domestic validation of normsor what Wiener (2018: 43) calls 'cultural validation'is a crucial process for establishing norm implementation where stakeholders deliberate and contest the 'appropriate' application of norms. Second, it argues that this locally compromised normative understanding weakens domestic implementation despite the wider discursive support of the norm. Such weakening sheds light on the role of norm antipreneurs (Bloomfield and Scott 2017) in actively resisting new implementation while defending the normative status quo through discursive engagement. As a result, validated norms are discursively robust but weak in implementation.
By illuminating the domestic validation of norms, this article contributes to new knowledge by tackling norms scholarship's over-emphasis on norm institutionalization at the international level and shreds light on a new life given to validated norms (for critique, see Betts and Orchard 2014;Wiener 2018: 45). The focus on domestic validation has added values. It contributes to the bourgeoning body of literature on the global-local dynamics of norms, which examines how norms are reinterpreted and translated in a new context (Auethavornpipat 2021;Medie 2020;Zimmermann 2017;Zwingel 2012). It reveals how domestic validation may shape and reshape normative understandings, thus influencing how actors subsequently re-enact global norms. This aspect of norm implementation is important because norms often do not live up to their intended purpose (Krook and True 2012). Domestic validation of norms further assesses a 'stable reciprocity' or congruence between legal norms and actions of norm followers by scrutinizing the sense of legal obligation among norm followers beyond the mere treaty ratification (Brunnée and Toope 2019). The emphasis on domestic validation in this article also incorporates a more expansive analysis of roles played by domestic institutions, international actors and non-state actors in implementing international norms (Betts and Orchard 2014). Such an approach sheds light on the importance of stakeholders' distinctive beliefs and social entities in redefining norms that manifest in practice without romanticizing local agencies, and pre-existing cultures and beliefs.
The argument unfolds over the ensuing sections. Following this introduction, Section II discusses the limits of previous norms literature. Section III conceptually unpacks the processes of norm validation and the robustness of validated norms. Moving to an empirical focus, Section IV outlines normative expectations on fisher protection as outlined in the ILO Work in Fishing Convention. Section V focuses on domestic validation of the Convention by examining the course of contestation among stakeholders in Thailand from 2015 until 2019. Stressing the consequences of these interactions, Section VI analyses the impact of norm validation and illustrates that norm antipreneurs have weakened the practices of migrant fisher protection in Thailand. In the conclusion of the article, I encapsulate key contributions of this article, urging IR scholarship to pursue a more nuanced view of norm diffusion that recognizes the ways in which international norms can be manipulated and weakened without completely losing their validity.

II. Shortcomings of conventional norms approaches
The weakening of domestic norm implementation despite rigorous promotion by state actors cannot easily be explained by existing norms approaches. The norm socialization account in the early constructivist research would expect actors to simply follow norms upon adoption because states either think it is the right thing to do (Finnemore and Sikkink 1998) or are trapped in the rhetorical action in referencing norms (Mumford 2021;Schimmelfennig 2001). According to these accounts, states would be reluctant to diverge from norms due to the risk of being criticized and delegitimized by the international community. Consequently, norms are taken for granted to be stable and strong, and the socialization approach therefore cannot capture the weakening of adopted norms. In an opposite account, the de-coupling literature (Hafner-Burton and Tsutsui 2005; Jetschke and Rüland 2009) follows the acculturation approach (Goodman and Jinks 2013) but diverges from the latter to explain the lack of norm implementation after adoption. Within this literature, authoritarian regimes with limited legitimacy are expected to enhance their standing by mimicking peers but purposely dismissing subsequent implementation of norms (Katsumata 2011; for the Thai authoritarian regime, see Auethavornpipat and Tanyag 2021). In this circumstance, norms are not even translated into domestic implementation but used merely as strategic tools to discursively signal normative commitment so state actors can gain recognition from the international community. The decoupling literature therefore cannot explain the twin outcome of active promotion and weakened implementation, as observed with migrant fisher rights in Thailand.
Alternatively, the seminal 'norm spiral' model  can explain the weakening of norm implementation, but only to a limited extent. The five-stage spiral model (repression, denial, tactical concession, prescriptive status and rule-consistent behaviour) starts with the assumption that domestic activists establish an alliance with transnational advocacy networks (TANs) after facing domestic repression. The TANs then pressure the government to adopt norms and change domestic policy. The spiral model does not assume evolutionary progress in norm adoption and is open to reversal at any stage. However, such reversal is only imagined to be caused by the government's repression of domestic civil society. It treats backlash against norms as an outliner and as a direct result of state actors' reaction (Risse-Kappen, Ropp and Sikkink 1999: 242; for the critique, see Minami 2019). In this circumstance, it expects repressive states to aggressively resist and water down normative standards. Furthermore, the spiral model depicts local supporters to be empowered after a treaty ratification to the extent that they can leverage their normative power and international support to pressure the state into rule-consistent behaviours . On the contrary, as illustrated in this article, the Thai government defended norms and organized extensive consultation with stakeholders rather than using a strategy of repression. The spiral model therefore largely overlooks the complexity of domestic space, where counter-movement against new norms is mobilized by norm antipreneurs.
The norm localization approach has similar pitfalls. Localization is defined by Acharya (2010: 15) as 'the active construction (through discourse, framing, grafting, and cultural selection of foreign ideas by local actors, which results in the former developing significant congruence with local beliefs and practices'. The successful localization of norms fundamentally depends on local agencies to ensure compatibility between external and local norms. The localization rightly pays attention to local actors in interacting with norms. However, the approach has largely been viewed as an active resistance against external ideas (Capie 2008; for further critique, see Zimmermann 2017). As such, insider proponents such as Thai state actors would be expected to purposely discard undesirable norms. Yet Thailand adopted the Convention without any reservation, which suggests its strong normative support. Therefore, the selective borrowing of external norms by the state as expected did not occur. Moreover, the localization model has not gone far enough to interrogate implementation efforts once norms are localized (for this critique, see Jacob 2018). The shortcomings of existing norms approaches necessitate greater scrutiny of domestic validation of norms to better account for the role of relevant and diverse stakeholders in deliberating international norms for domestic implementation.

III. Domestic validation of international norms
To address the shortcomings of aforementioned norms approaches, this section conceptualizes the domestic validation of norms and the role of norm antipreneurs in weakening the implementation of validated norms without damaging norms' discursive strength. Scholars need to pay more attention to domestic processes that validate and reconstitute international norms. The local-global dynamics of norm diffusion are often overlooked. As asserted by Wiener (2018: 45), norms scholarship tends to focus on examining the validation of norms at the global level or by a collective group of actors. As further articulated by Brunnée and Toope (2019), the formal and social standing of norms as codified in international conventions alone is inadequate for comprehending legal normativity, and scholars should investigate into the continued process of practices and contestation that determine the appropriate application of norms. Krook and True (2012: 105) also dismiss 'the assumption that a norm can be equated with a commitment written into international treaties and instruments' and contend that norms can evolve over time in response to the competing interpretation as well as the environment in which norms operate. Norms may be institutionalized in international law, but do not automatically guarantee subsequent implementation or always retain robustness over time (Percy 2014). The international institutionalization of norms therefore cannot be fully understood without considering the implementation of norms at the domestic level (for further critique, see Betts and Orchard 2014;Zimmermann 2017).
This article overcomes the above limitations of norms research by unpacking the domestic processes of norm validation. The validation of norms at the domestic level, or what Wiener (2018: 43) calls 'cultural validation', introduces new assumptions regarding behavioural and norm change. As opposed to viewing norms as stable objects in the earlier wave of constructivist norms research, cultural validation follows critical norms scholars' understanding that norms have a dual quality of structuring and being (re)constructed through interaction in context (Wiener 2007: 49). Norms can have a structuring power in guiding actors' behaviour and at the same time be reconstituted through interactions among involved actors. The reconstruction of norms further entails norms being expected to change upon re-enactment in a new setting (Lorentzen 2017;Sandholtz 2008). Therefore, domestic validation is crucial for revealing new normative understandings in norm implementation.
This article advances an agentic approach to interrogating the domestic validation of norms. Domestic validation is understood as an expression of individual expectation, as mediated by one's background experiences at the micro-level. These background experiences in turn guide how an actor distinctively interpret and comprehend norms (Gholiagha, Holzcheiter and Liese 2020;Wiener 2014). The agentic emphasis thus highlights the importance of actors' previous experiences as both a sense-making instrument and a reservoir of knowledge 'from which communicating individuals draw interpretations when exchanging knowledge about particular issues in the world' (Wiener 2007: 62). As explained further below, domestic validation can involve a wide range of stakeholders such as individuals, firms and government officials who deliberate norms to establish a shared understanding for subsequent norm implementation (Wiener 2018). In other words, it enables international norms to be appropriated in the domestic setting after various interpretations of norms are negotiated (Lorentzen 2017). Domestic validation is thus an important process for actors to interpret norms prior to domestic implementation.
Domestic validation is distinguished from other dimensions of norm validation, namely formal validation at the macro or global level and social validation at the meso or group level. To elaborate, formal and social validation pertains largely to the institutionalization of norms at the international level (Wiener 2018). Formal validation of norms is achieved through negotiations which involve committee members of international organizations, negotiation delegations or high-level state representatives (Wiener 2019). After such negotiations, norms are said to have acquired their formal validity, referring to norms' explicit standing in international law, conventions and treaties. Formal validity serves as a reference frame for subsequent reinterpretation and implementation. For instance, in this article the ILO Work in Fishing Convention acts as a reference point for reinterpretation and deliberation by stakeholders in Thailand. In another dimension of norm validation, social validation refers to a collective recognition of norms achieved through repeated interactions in institutional settings at the meso level or in a social group. Social validation creates an intersubjectively shared understanding of behaviour and is a source of social obligation commonly held by a group of actors (Wiener 2014).
As a distinct dimension, domestic validation can reinforce norms' formal and social standing by enhancing legal fidelity. In particular, domestic validation deals with the norm implementation stage as triggered by 'a state or organizational commitment to the emerging norm' (Betts and Orchard 2014: 6). As this article shows, the domestic validation process was instigated after the Thai government declared its intention to ratify the Convention in 2015. After signalling such commitments, state officials, civil society representatives and business sectors began interpreting the Convention in order to translate international standards into domestic legislation and implementation. Domestic validation is also integral to the Thai dualist legal system, which requires the Thai government to make necessary amendments to domestic laws before the Convention can have any legal effect. The process therefore bestows life on international norms in a new setting beyond formal and social validity. As argued by Brunnée and Toope (2019), it can generate the internalized sense of legal and normative obligation among norm followers even without explicit enforcement. In other words, the domestic validation of norms can prescribe legal standing to norms.
The agentic view of norm validation, as conceptualized in this article, takes into account a diversity of stakeholders and distinctive beliefs. This follows the quod omnes tangit principle as introduced to norms research by Wiener (2018; see also Cornago 2017). The principle signifies that norms should be approved by the governed. Stakeholder agency is therefore important, but it is often glossed over by the previous norm implementation model (Betts and Orchard 2014). Bett and Orchard's model rightly opens the 'black box' of domestic processes that involve a whole range of new actors and Global Constitutionalism 5 ideational, material and institutional factors that impact the implementation of norms. However, despite recognizing the importance of ideational factors in informing distinctive norm interpretation, this model largely treats pre-existing cultural contexts as already collectively shared, and hence monolithic and overlooking actors' agency in norm interpretation. It strongly emphasizes the 'cultural match' between foreign and local ideas (Betts and Orchard 2014;Checkel 2005). In contrast, domestic validation as advanced in this article takes a more diverse view of actors' belief systems, which does not necessarily assume actors have the same experiences, cultures and normative understandings. The domestic validation of norms further foresees that local normative meanings are amenable as a result of stakeholder engagement. The agentic understanding of norm validation is thus more analytically open and exploratory in terms of allowing stakeholders' normative understandings to come to the fore during their social interaction.
As domestic validation incorporates a diverse range of actors with different normative understandings, norm validation essentially becomes a process of norm contestation. In such circumstances, stakeholders need to sort out their clashing normative interpretations in order to establish a shared understanding of norms (García Iommi 2020; Wiener 2018; Wolff and Zimmermann 2016). Domestic validation of norms operates within the logic of 'contestedness', understood as a moral or cultural objection that international norms are applicable across global society (Wiener 2019: 130-31). Contestation during domestic validation likely involves testing universal validity claims of norms in relation to specific rules and procedures, which are to be formulated for implementation in a specific setting (Wiener 2019). 1 As opposed to broad prescriptive or fundamental norms such as human rights, rules and procedures reconstituted to implement global standards are narrower and specific to the context of re-enactment. 2 Hence, contestation can be expected from relevant stakeholders who have to implement these rules and procedures (Mende 2021).
By focusing on the contestation of specific rules and procedures during domestic validation, this article also sheds light on an objection to norm implementation, which Deitelhoff and Zimmermann (2020: 57) call 'applicatory contestation'. This type of contestation aims to delineate how norms are supposed to be used, which action is deemed appropriate for implementation and which norms should be implemented in a given situation. Such contestation is a normal occurrence with legal or treaty norms, which are often contested through reinterpretation (Brunnée and Toope 2019). Consistent with Wiener's understanding of contestedness, applicatory contestation is instigated when stakeholders seek to enact norms in a new context where norms are situated in the 'zone of ambiguity' (Sandholtz 2008: 106). This zone is where the standard of behaviour reached at international negotiations cannot readily convey precise meanings and applications of norms for every subsequent situation. As a result, applicatory contestation is necessary because 'treaty drafters do not foresee many of the possible applications, let alone their contextual settings' (Chayes and Chayes 1993: 188). Consequently, applicatory contestation will generate new battles over the precise meanings of norms (Van Kersbergen and Verbeek 2007), and the validation of norms through applicatory contestation is significant for setting normative understandings of implementation. 1 Wiener calls these Type 3 norms. They are distinguished from Type 1 norms, which deal with fundamental norms and are most likely to be negotiated at the macro or global level by representatives of the government. Type 2 norms are organizing principles constituted through policy at the meso level. See further Wiener (2019).
2 See further Contestation site 9 in the cycle grid model in Wiener (2018).
The outcome of contestation signifies an agreement and compromise of legal normativity among stakeholders in a given context. It is also important to pay attention to stakeholders' access to contestation as part of norm validation (Wiener 2018). Such access is contingent on existing formal and semiformal regulations and practices, which structure a mode of participation that shapes the inclusion and exclusion of individuals in political processes (Jayasuriya and Rodan 2007). The meaning of norms can be shared widely when access to contestation is wide and inclusive. However, an agreed meaning of norms is often a product of exclusion, which entails the 'silencing' of particular actors (Krook and True 2012;Richmond 2011). By marginalizing some actors, this silencing act defines the limits of both discourse and implementation that can be part of normative understandings. As seen with the contestation of migrant fisher protection in Thailand, the initial access to contestation was established through a broad consultative process, but the final agreement that has weakened norm implementation was a product of restricting NGOs' and migrants' participation in the final rounds of deliberation. Consequently, such normative understanding established through exclusion is not readily shared by NGOs and migrants, who continue to contest locally acquired and yet weakened meanings of norm implementation. In line with the dual quality of norms explained above, exclusionary access to contestation plays a significant role in reshaping and watering down the implementation of migrant fisher protection in Thailand.
To further explain the weakening of validated norms, this article specifically identifies the role of norm antipreneurs, defined as those who safeguard the entrenched normative status quo against challengers (Bloomfield and Scott 2017). They are 'a relatively small group of powerful political operators' who can undermine norms despite the firm commitment to international laws and conventions (Sikkink 2013: 162). Their activities come to the fore when a new norm challenges a more dominant norm that is deeply and institutionally held by more powerful actors. Norm antipreneurs are largely indistinguishable from norm entrepreneurs as both groups of actors can employ similar tactics such as framing and normative arguing in advancing their claims (Heller, Kahl and Pisoiu 2012). However, norm antipreneurs can be revealed through their goals of resisting normative change in ensuring continuity of the status quo (Nesadurai 2017). Furthermore, by associating themselves with the status quo, they are more likely to have an inherent advantage of the incumbent government and existing normative order. As a result, actors who promote new norms that create normative disruption often fight an uphill battle as they are faced with psychological obstacles and uncertainty accompanying normative change (Bloomfield and Scott 2017). As a consequence, norm antipreneurs can be 'highly effective, distorting, blunting or blocking progressive initiatives' (Bob 2013: 71). In the contestation of migrant fisher rights examined below, the role of business sectors and fisheries associations fits with the conceptualization of norm antipreneurs. Given their contribution to Thai seafood exports, any new normative disturbances to both the economic output and supply of migrant labour will be deemed undesirable, and hence subject to resistance. Businesses and industries also traditionally have political clout and can influence Thailand's labour migration policies (see Auethavornpipat 2019a; Brown and Hewison 2005).
Norm antipreneurs can impact the robustness of validated norms in two dimensions: implementation and discourse. The incorporation of both dimensions follows the notion of legal fidelity, signifying that norms and laws can form a sense of legal obligation among actors even though they do not necessarily agree with such norms on moral grounds (Brunnée and Toope 2010). In other words, analysing robustness in both discourse and implementation can illuminate whether norms remain strong beyond formal or social validity as well as beyond rhetoric. Robustness in norm implementation is evaluated by examining the applicatory scope of norms when norms are transferred into a new setting (Deitelhoff and Zimmermann 2019). It is shown by comparing normative obligations between internationally prescribed standards on the one hand and domestic laws and policies on the other. When the scope of norm implementation becomes extremely limited after domestic validation, 'a norm may well degenerate as a result, since its rules can hardly be applied to empirical circumstances' (Panke and Petersohn 2012: 723). The assessment therefore provides a point of comparison between international standards and domestic implementation (Simmons and Jo 2019). As shown below, the involvement of norm antipreneurs in Thailand waters down norm implementation, which has become increasingly restrictive in the protection scope when Convention norms are reinterpreted into domestic laws. Norm antipreneurs can also weaken norms in the discourse dimension. Norms are said to be robust at least in the discursive dimension when stakeholders actively defend norms and prevent contestation from radicalizing to the questioning of the overall validity of norms. Inversely, the lack of discursive response from relevant actors can be viewed as the weakening of norms. As seen in this article, the ability of norm antipreneurs in weakening the discursive acceptance of norms is more challenging. As opposed to domestic implementation, which is largely negotiated by local constituents, the discursive engagement with norms can go beyond the domestic jurisdiction. To broadly capture the discursive support of norms on the ground, the discourse-based dimension of norm robustness is further distinguished into two indicators: concordance with norms and third-party reactions. These two indicators can operate alongside one another. Concordance is defined in terms of 'how widely accepted the rules are' (Legro 1997: 35). This can include the discursive acceptance of norms from state officials, NGOs, international organizations and private citizens (Deitelhoff and Zimmermann 2019). The concordance with migrant fisher protection can be seen in the discourse among Thai government officials and NGO representatives, expressing their defence of migrant fisher rights.
The concordance with norms alone is insufficient and should be analysed together with the feedback from other norm addressees in the wider international community, particularly third-party reactions. Third-party reactions are witnessed mainly with discursive responses, but material sanctioning such as international trade embargos is also possible. In preserving norms' strength, other norm addressees can discursively link a particular norm to a larger norm cluster to defend the validity and application of the norm in question. Lantis and Wunderlich (2018) explicate that norms embedded in norm clusters are more resilient when facing challenges, and are therefore more likely to survive contestation without a serious damage to norm validity. This strategy can help ensure that norms remain discursively robust. Third-party reactions to migrant fisher exploitation in Thailand are seen with the discursive response from the US government and the European Union. Both actors condemn Thailand's domestic practices and simultaneously mobilize discursive sanctions threats against Thailand. Both the United States and European Union associate the exploitation of migrant fishers with the larger normative cluster of anti-human trafficking and 'anti-slavery', which is well established and highly robust (Auethavornpipat 2017). By embedding norms in the larger normative cluster, it becomes more difficult for norm antipreneurs to resist new norms and for state officials to make generous concessions to norm antipreneurs. To relate to the implementation dimension introduced above, concordance with norms and third-party reactions can operate alongside one another to prevent norms from losing their core strength despite norm weakening in implementation.
To empirically show norm validation and robustness, I use a reconstructive analysis technique, which enables researchers to recreate and map the activities and discourses of relevant stakeholders (Wiener 2009). This entails analysing discursive interventions of stakeholders in the text corpus. The texts were generated with the collection of ILO and media reports, parliamentary debates, governmental reports and declassified documents. The text corpus covers a period from the start of contestation in 2015 until the Convention's ratification in 2019. These materials were subsequently organized in chronological order to gain insights into the course of contestation. Interviews were also conducted with civil society representatives and government officials to better understand their views of the Convention deliberation. Given that the issue of migrant worker protection and human trafficking is highly sensitive in Thailand, interviewees were anonymized to protect their identity and ability to work in policy-making without backlash. To ensure data validity, I compared versions of the draft legislation against the ILO Convention and adopted law to identify the impact of contestation on norm robustness at different stages in the chronology developed above. This method reveals crucial moments in contestation and the direct impact such moments have on validated norms and their robustness. It particularly illuminates whose applicatory discourses are advanced, rejected and accepted as 'appropriate' among relevant actors in Thailand. The next section discusses normative obligations as outlined in the ILO Work in Fishing Convention before examining the contestation of migrant fisher protection in Thailand.

IV. ILO Work in Fishing Convention
The concerns for fisheries worker protection came to the forefront of international labour agenda in the early 2000s, which eventually led to the adoption of the 2007 Work in Fishing Convention that established the formal validity and social recognition of norms. The ILO began propounding work in fishing to be 'among the most dangerous of all professions' (ILO 1999). Fishing is physically demanding and fishers commonly sustain injuries from working with heavy gear, powerful machinery and slippery decks, especially in adverse and extreme weather and sea conditions (Politakis 2008). Furthermore, the inability to provide immediate medical attention while at sea contributes to higher fatality rate among fishers, making the occupation even more hazardous than timber cutting and construction. The ILO reported that as many as 24,000 fishers were killed each year and the fatality rate for the fishing industry in the United States in 1996 was sixteen times higher than that for fire-fighting and 40 times the national average (ILO 1999). From the 1960s onwards, ILO standards on fishing were adopted in various conventions and only granted partial protection for fishers on industrialized, commercial maritime fishing vessels. These standards therefore left small-scale fishing vesselswhich are the majority of fishing operations in the worldoutside the scope of protection. Given that approximately 44 million people worldwide were estimated to be employed as fishers and fish farmers, and that they often lacked access to protection, the ILO urged its member states to adopt a new and more comprehensive convention (Mathew 2010).
Following such concerns, the ILO Governing Body decided in 2002 to place labour protection for fishers on the agenda of the 92nd (June 2004) Session of the International Labour Conference. The Conference is annual tripartite forum among 187 ILO member states, consisting of governmental officials, employer and employee representatives.
The priority during the 2004 Conference was to discuss 'a comprehensive standard (a Convention supplemented by a Recommendation) on work in the fishing sector' (ILO 2003: 1). To prepare for the deliberation, existing laws, regulations and practices of ILO member states were requested and surveyed by the International Labour Office, the permanent secretariat body, which was put in charge of publishing the survey result in 2003. The survey was compiled on the basis of the replies to the questionnaire from 82 ILO tripartite members in which 64 governments replied in favour of adopting a comprehensive standard in a form of a new Convention with a non-binding Recommendation (Mathew 2010). The preliminary report then served as the foundation for the deliberation and negotiation involved in setting a new international norm for protecting fishers.
To facilitate the establishment of new protection norms, the ILO also examined its previous, older conventions that dealt with the protection of fishers in order to consolidate them into a new and comprehensive convention (Politakis 2008). The ILO reviewed seven existing instruments (five Conventions and two Recommendations) that would later be revised for higher protection in the new Work in Fishing Convention adopted in 2007 (ILO 2003). To elaborate, the new Convention raised the minimum age of fishers to sixteen years, instead of fifteen years under ILO Convention No. 112 (Minimum Age [Fishermen], 1959). The new Convention also set the maximum period of medical certificate validity to two years whereas there was no specific limit outlined in the ILO Convention 113 (Medical Examination [Fishermen], 1959). The new Convention required national laws specify the minimum level of crewing for safe navigation of fishing boats and defined a minimum and specific period of daily and weekly rest for workers remaining at sea for more than three days (Politakis 2008). It also obligated every fishing vessel to carry a crew list on board rather than leaving to local authorities as previously required by the ILO Convention 114 (Fishermen's Article of Agreement, 1959). To prevent occupational accidents and infectious diseases, the measures for onboard medicare were also introduced in the new Convention. Moreover, the Work in Fishing Convention ensured fishers' access to social security and guaranteed that fishers were not discriminated against compared with other types of workers. It further advanced the protection of fishers against abandonment by guaranteeing that repatriation be arranged by the employers or the flag state in case employers failed to do so.
Older ILO conventions did not enjoy popular ratification among ILO members, so it was hoped that the new Convention with higher standards would spark immediate and rapid ratifications. To accelerate the ratification, the Convention innovatively applied the principle of 'no more favourable treatment' to ensure that fishing vessels of any Flag State that had not ratified the Convention would not gain unfair competitive advantages against vessels of member states that were parties to the Convention. In particular, the Convention required that the same standards be applicable to all fishing vessels visiting members' ports, which could be subject to inspection by port state control officers. The new Convention therefore protects ratifying state from unfair competition, and in doing so, extends the internationally prescribed living and working conditions to all fishers by establishing an equal level-playing field (ILO 2012).
The Work in Fishing Convention was eventually adopted in 2007 when 437 constituent members of ILO voted in favour, with only two votes against and 22 abstentions. This suggests that the votes have established the formal and social recognition of a new norm for protecting fishers. Yet the Convention would need to be interpreted further to determine specific measures for implementation in a ratifying country. Various clauses are ambiguous, purposely left flexible in order to accelerate their ratification and implementation by ILO member states. As the ILO governing body stated, the new Convention 'should be sufficiently broad and flexible to address a number of issues and to be effective for the majority of the world's fishermen … It should be based on principles which could be implemented in a manner which would accommodate the diversity of economic and social conditions of countries' (ILO 2004: para. 70). The ILO also stressed that the Convention should not be 'overly prescriptive' (ILO 2004: para. 70). Consequently, due to ambiguity and flexibility, stakeholders in Thailand were left to deliberate, argue and contest their interpretation of the Convention in order to establish a shared understanding of the appropriate protection of migrant fishers. The next section illustrates how norms were domestically validated through applicatory contestation in Thailand.

V. Validation of migrant fisher protection norms in Thailand
This section captures the validation of migrant fisher protection norms through stakeholder contestation in Thailand. The domestic validation of norms created a contextualized understanding of the Convention application. It occurred after an intense international pressure on Thailand regarding the exploitation of migrant workers in fishing and seafood industries. The international pressure spurred a series of domestic reforms, which in turn led to the identification of measures in the Convention to protect migrant fishers. The Convention sparked subsequent applicatory contestation, notably among state officials, NGOs and employers or fisheries associations.
The domestic recognition that the rights of migrant fishers should be protected followed a period of intense international criticism of Thailand. Criticisms emerged in response to the trafficking of migrant workers whose labour was used in export production. International NGOs such as the United Kingdom-based Environmental Justice Foundation (EJF) and media outlets such as the Guardian and Associate Press started propounding labour rights violations and human trafficking in the seafood supply chain from 2013 onwards (Auethavornpipat 2017). Their reports implicated both the US and EU markets in importing commodities made with exploited migrant labour, effectively mobilizing the United States and European Union to use the threat of sanctions against Thailand. The United States downgraded Thailand to the lowest rank in its Trafficking in Persons (TIP) Report for two consecutive years, in 2014 and 2015 respectively. The European Union issued a 'yellow card' warning to Thailand in April 2015 on the issue of illegal, unreported and unregulated (IUU) fishing. In the former, the United States downgrade of ranking threatened to cut non-humanitarian, non-trade related foreign assistance if Thailand continued to ignore the trafficking of migrants. In the latter, the European Union gave Thailand six months to combat IUU fishing and improve working conditions for migrant workers. Otherwise, a trade embargo would be imposed on all Thai fisheries goods. The sustained pressure from both the United States and European Union over the years was necessary to prompt the Thai government's initial reforms to protect migrants but insufficient to generate domestic agreement on the Convention ratification. The government adopted the Ministerial Regulation Concerning Labour Protection in Sea Fishery Work issued by the Ministry of Labour in December 2014. It granted fishers a set of rights such as rest hours, minimum age, repatriation, food and water and medicare on board. During this period, the European Union also brought up the Convention so Thailand could simultaneously solve human trafficking and IUU issues (Prachachat 2017). Subsequently, the debates on the Convention ratification entered the domestic sphere and became contested after the Thai government signalled its intention to ratify the Convention.
The Thai government expressed its intention to ratify the Convention two months after the EU mobilized the threat of sanctions, thus initiating the course of norm contestation. On 10 June 2015, Labour Minister General Surasak Karnjanarat met with Guy Ryder, Director-General of the ILO, in Geneva to convey his decision to adopt the Convention. At the meeting, ILO Director-General iterated the concern on the protection of migrant workers and trafficked victims, and in doing so reinforced the normative understanding that Thailand should improve the working conditions for migrant fishers (Ministry of Labour 2015). During the International Labour Conference in the same year, the Thai Labour Minister explained that the government had already taken initiatives in revising Thai laws with the view of complying with the Convention, further reaffirming the intention to use Convention norms to the international community. Such commitment was reiterated over the years. The official announcement with the 2017 timeframe necessitated identifying the discrepancy between domestic and international standards. The ILO Regional Office in Bangkok provided technical assistance and released a gap analysis report in May 2017. The ILO found that the scope of Convention 188 covers all fishers and all types of fishing vessels engaged in commercial fishing. However, this was lacking in Thai domestic laws, which did not cover fresh water fishing (ILO 2017: 14-15). The ILO also identified several other aspects of protection where Thailand needed reforms in order to support the ratification. Among many discrepancies, the ILO stressed inadequate health and fitness test prior to employment; rest period; annual, sick and paid leave; right to repatriation when fishers cannot perform duties; size and quality of accommodation on board; hygienic and sanitary facilities; and social security. By specifying the gap between domestic and international standards, the ILO in effect urged for the widening of norm's applicatory scope in Thailand before the ratification.
The call for the expanded protection in compliance with the Convention was met with opposition from fisheries associations, particularly the National Fishery Association of Thailand (NFAT). Throughout the domestic deliberation, fisheries associations resisted the Convention that would grant more rights to migrants. They believed that protection provisions in the Convention were developed abroad, and therefore, were not suitable for using in the context of Thai fishing. Fisheries associations also considered that such measures would impose additional burdens on businesses to reform the industry. While the latter concern related to the cost of reforms, it provided employers with a basis for expressing their disagreement and objection to granting more rights protection in implementing Convention measures, which NGOs and migrants themselves asserted should be applied in Thailand. To explain further, NFAT President Mongkol Sukcharoenkana indicated in September 2017 that 'the [Thai] context did not allow' for the implementation of the Convention (Prachachat 2017). He elaborated that the financial burden would be put on employers to remodel their fishing vessels in order to incorporate migrant rights to proper accommodation and sanitation facilities on board. It should be noted these amenities were largely non-existent in the traditional Thai wooden fishing fleet. By November 2017, approximately 1,000 fishing operators in coastal Southern Thailand complained that more than 10,000 boats had to stop operating due to a more stringent law on IUU fishing. They further claimed that the fishing sector had lost approximately 500 billion baht (US$16 billion) in damages and the number of Thai fishing fleet had decreased from 40,000 to 10,600 vessels (The Nation 2017). As a result, employers demanded that all 'affected stakeholders', including fishing business owners, be included in the deliberation on any new laws that would regulate Thai fishing industries (77 Jowo 2017). This particular request was also submitted to the Governor of Pattani Province and then forwarded to the Prime Minister.
To stop employers from opposing the Convention, the government organized public hearings to listen to stakeholders in 2018. From March to October 2018, the Ministry of Labour organized six consultations in Bangkok and coastal provinces (Ministry of Labour 2018f). The purpose, as stated by the Ministry of Labour (2018c), was 'to have a common understanding and lead to the amendment of the Bill in accordance with the context of Thai society and the best practice of all parties'. The government hoped to incorporate the input from fisheries associations, employee organizations and NGOs in order to bridge the protection gap between domestic laws and the Convention. The Ministry of Labour organized the first consultation on 23 March 2018 in Samut Sakhon Province, Thailand's fishing industries hub (Ministry of Labour 2018c). At this consultation, private fishing operators repeatedly urged stakeholders to consider 'the Thai context' in drafting the new law and avoid copying the Convention text verbatim without thinking about 'the reality' in Thailand (Ministry of Labour 2018g). Furthermore, fisheries associations put forward their demand to relax Convention provisions such as minimum age, medical and fitness examination and onboard accommodation. They also opposed the inclusion of migrant fishers in the social security system because, they asserted, migrants were already required to buy health insurance schemes.
The disagreement about specific ways to implement the Convention continued in subsequent consultations. At the second consultation on 29 March 2018, fisheries associations raised the same objection, pressing the government to reconsider the requirement on medical and fitness test, rest hours and social security. The meeting minute showed that NGO representatives of the Migrant Working Group and the Stella Maris Seafarers' Center defended the inclusion of social security benefits in the draft law. By the third consultation in April 2018, it became clear that social security for migrant fishers was among the most contentious issues, which remained unresolved until just before the law was approved. In May 2018, migrant fishers themselves also advanced their own rights demands through the newly established Fishers' Rights Network (FRN). In contrast to other NGOs, which are led by Thai nationals, the FRN establishment is meant to empower migrant fishers so they can lead their own activism in eradicating exploitation in the Thai fishing industry. In supporting the Convention, they demanded an improvement in working conditions, issuance of written employment contracts in migrants' own languages, access to basic first aid training, comprehensive medical kits on fishing vessels, emergency medical protocols and monthly wage payment via electric transfer (FRN 2018). Their demands also went beyond the Convention standards in calling for the elimination of immigration and recruitment fees while pressing the government to ratify ILO Convention 87 and 98 to give migrant workers the freedom of association and collective bargaining. While consistently defending this set of rights, the FRN further added statutory social security benefits and minimum wage increase in the period leading up to the Convention ratification in January 2019 (FRN 2019). The FRN's demand therefore reinforced the claims of NGOs and directly opposed employers' requests. It is rare in Thailand for migrants to make such assertions themselves because domestic trade unions are weak and migrants are prohibited from leading a union under Thai law (Auethavornpipat 2019b). As shown further below, the participation of migrants and NGOs would increasingly be curtailed in subsequent deliberation rounds.
The Thai government responded to the disagreement between fisheries associations and NGOs by stating its commitment to adopting the Convention measures. At the trilateral workshop organized by the Ministry of Labour and the ILO on 10 May 2018, Deputy Permanent Secretary of the Labour Ministry Viwat Tanghong defended the ratification: 'It is necessary to adhere to ILO Convention 188 and lead to the development of a country that is not deprived of trade in the world market and can stand with the export goods for sale abroad' (Ministry of Labour 2018e). On 1 June 2018, the Minister of Labour organized the fourth 'stakeholder consultation' in Bangkok to resolve clashing understandings on the specific ways to apply Convention provisions (Thai Tuna Industry Association 2018: 4). At this meeting, representatives of both national and regional fisheries associations united in opposing the extension of social security to migrant fishers. Instead, fisheries associations proposed replacing the social security scheme with the use of private health and accident insurance as well as compensation fund (Thai Tuna Industry Association 2018: 4). In response, an employee organization, the State Enterprises Workers' Relations Confederation, warned against employers' disapproval of the social security scheme as 'reversing' workers' entitlement to welfare benefits. It explained that private insurance schemes would exclude migrants from broader access to pension, pregnancy and unemployment benefits (Ministry of Labour 2018g). Employers also reasserted that the Convention was not suitable to be applied in Thailand because its provisions requiring sleeping rooms, one bathroom per four workers and 190-centimetre headroom were not practical for Thai wooden fishing boats (MC Digital TV 2018). By this point, the employer associations had reaffirmed their interpretation of Convention standards as unsuitable for implementation in Thailand, calling for alternative ways to apply the Convention to protect migrant fishers.
Fisheries associations continued to protest against using the Convention even though two more rounds of consultation were to be organized by the Ministry of Labour. In July 2018, the NFAT planned to lodge a formal complaint with the Prime Minister because of the 'trouble' associated with the Convention implementation (Ministry of Labour 2018a). At the same time, each of the twenty-two provincial fisheries associations submitted the same complaint to their Provincial Governor's Office (Voice Online 2018). After learning of such plans, the Ministry of Labour responded by reaffirming its determination to adopt the Convention. On 25 July 2018, Labour Minister Police General Adul Sangsingkeo dismissed employers' objections by stressing the rights of fishers: The Ministry of Labor requests that fishermen be assured that the government … and related agencies, will continue to work to prevent and address labor issues in the fisheries sector … to enhance the protection of the rights of workers in accordance with human rights and international standards to drive the country to stability, prosperity, [and sustainability] (Ministry of Labour 2018a).
The Ministry further stated that it was 'committed to ensuring work safety on fishing vessels … the substance of the Convention (ILO) No. 188 on Fisheries 2007 will allow fishermen to work in conditions of minimum quality' (Ministry of Labour 2018a). Thus, the Thai government defended the Convention provisions, and in doing so, attempted to dismiss employers' perspectives. In response to the government's refusal to change its position, the fishing businesses called for their greater inclusion in future deliberation in order to influence how the Convention ought to be used in Thailand. This could be observed in the mobilization of employers and fishing operators in Samut Sakhon Province, the hub of Thai seafood industries. In August 2018, approximately 300 employers submitted a letter to the Provincial Governor, seeking 'fairness' in the deliberation of new laws (77 Jowo 2018). The Advisor of Samut Sakhon Fisheries Association justified the demand for greater inclusion by stating that the attempt to bring Thai laws closer to international standards caused several problems for fishing operators.
The government accepted the fishing businesses' request by organizing a more inclusive consultation, hence broadening employers' access to contestation. As part of the fifth stakeholder consultation, the Ministry of Labour organized a series of meetings along coastal provinces from September to October 2018. In total, five meetings were held with fishing industries along 22 coastal provinces. By doing so, the government increased its effort to incorporate inputs, particularly from fishing operators in remote areas such as Thailand's southernmost provinces. This consultative approach was in contrast to the previous four consultations, which were mostly held in Bangkok and surrounding locations. The fifth consultation concluded on 15 October 2018 and was followed by the sixth and final consultation on 19 October 2018 at the Ministry of Labour in Bangkok. As for the content of the last two consultations, similar issues to those unresolved from previous consultations were raised, including fitness and medical examination, and social security benefits. At the final consultation, NFAT President Mongkol urged that the final consideration would be left to the government, which had to explain both the positive and negative impacts that the ratification would have on Thailand (Ministry of Labour 2018g). By the end of the consultation in October 2018, an agreement on the draft legislation had yet to be reached. It was still unclear how the Convention measures would be used in Thailand.
The Thai government integrated stakeholder feedback into revising the Bill and requested further comments from the public (details of amendments discussed in the following section). The Office of Council of State, which was in charge of preparing the draft legislation and providing legal opinion to state agencies, conducted an online consultation from 16-30 November 2018. Interestingly, in line with employers' request, the revised Bill was circulated together with an analysis of the ratification impact conducted by the Ministry of Labour. The report indicated 'positive impacts' across the economic, social and fiscal spectrum, hence directly responding to employers' concerns. The report also identified 'fishing operators and labourers' to be the beneficiary of the new law (Law Amendment 2018). This official view clearly pushed for the adoption of Convention. However, the National Legislative Assembly (NLA), a legislative body hand-picked by coup leaders, went ahead and gave the green light to the Ministry of Labour to ratify the Convention one day before the online consultation closed.
The NLA's approval was met with a much stronger opposition from fisheries industries. As described by fisheries associations, the decision came 'out of nowhere' and 'stabbed' employers in the back (Thairath 2018). On 1 December 2018, fisheries associations from 22 coastal provinces collectively opposed the Convention provisions that protected migrants in the following aspects: onboard accommodation and recreation facilities; sanitary systems; rest hours; repatriation of workers to a country of origin; recruitment fees; social security; compensation for illness, work and death (Prachachat 2018b). Employers reiterated that these measures would require the remodelling of fishing vessels, causing a large financial burden on fishing operators. Then, on 9 December 2018, the NFAT organized an annual general meeting where its members across the country unanimously adopted a resolution objecting the ratification (Prachachat 2018a). In order to pressure the government to abandon the ratification, the members jointly submitted an objection letter to the Ministry of Labour while NFAT President Mongkol directly submitted the letter to Prime Minister General Prayut Chan-ocha (Prachachat 2018a). With the draft legislation pending, the NLA's decision intensified the mobilization from the fishing sector that urged the government to dismiss the Convention ratification altogether.
Employers failed to entirely dismiss the ratification and returned to the negotiation of specific Convention provisions with the goal of watering down normative obligations. On 13 December 2018, NFAT Advisor Vishal Sirichai-aekwat and President of the Thai Overseas Fisheries Association Aphisit Techanitisawad met Labour Minister Adul to seek a compromise so employers could meet the Convention implementation 'halfway' (Prachachat 2018c). At this meeting, Labour Minister Adul reaffirmed 'the necessity' to adopt the Convention but made concessions to employers in an agreement that any existing national laws and regulations which specified higher standards than those found in the Convention would be brought down in order to 'return rights' to fishing operators (Prachachat 2018c). The Labour Ministry further promised that representatives from the fishing industry would be given access to any future law deliberation as commissioners during parliamentary debates. This compromise clearly favoured a consensus formation between the government and fisheries associations at the expense of NGOs and migrants' voices in the contestation. As illustrated below, it facilitated subsequent negotiations that would result in the weakening of migrant fisher protection in numerous aspects. It should also be noted that this decision to weaken existing domestic practices can be deemed to be in violation of the ILO Constitution Article 19.8, which states that no ratification of ILO conventions can result in lowering national protection of workers. Nevertheless, the compromise between the government and employers was reached and so prevented the latter from opposing the ratification.
Employers seemed unsatisfied despite the above compromise. A declassified document shows that Prime Minister Prayut tabled the Bill to the Cabinet on 20 December 2018 (Secretariat of the Cabinet 2018). The draft legislation was also attached to the aforementioned impact analysis conducted by the Labour Ministry. Subsequently, on 24 December 2018, representatives from the fishing sector organized a mass protest in front of the Ministry of Labour in order to pressure the Ministry to abandon the ratification. In response, Labour Minister Adul reaffirmed Thailand's determination to ratify the Convention. He justified his position by explaining that the Ministry of Labour had already conducted several extensive consultations with relevant state agencies, employers, employees, international organizations and NGOs. He further described the ratification decision as 'righteous', hence defending the Convention implementation in Thailand (Khaosod 2018b). Simultaneously, to disperse the mass gathering, the Ministry filed a legal complaint with the police, which was withdrawn shortly after the fishing sector representatives agreed to end the protest and await a final announcement from the Labour Ministry (Matichon Online 2018).
The final announcement was scheduled for release on 26 December 2018. Prior to the announcement, employers put forward their final demand to relax fourteen protection measures of the Convention, which they insisted had to be met before the ratification (specific measures discussed in the following section). The Ministry of Labour agreed to consider the fourteen demands (Ministry of Labour 2018d). At the same time, Labour Minister Adul pledged 'to end conflicts with commercial trawlers' before the ratification. (Charoensuthipan 2019). His statement thus narrowed stakeholders to fisheries associations and employers to determine the outcome of domestic implementation. Adul therefore saw the government's additional consultations with employers as inevitable for resolving their clashing understanding of the Convention (Charoensuthipan 2019). On the same day that employers presented their final fourteen demands, the NLA held a special parliamentary sitting and voted to consider the revised Bill. Within the same sitting, the NLA set up an ad hoc committee to deliberate the Bill and requested inputs from states agencies, fisheries associations and seafood industries (National Legislative Assembly of Thailand 2019). Particularly in relation to the industry, the request was extended to representatives from the NFAT, Thai Food Processors' Association, Federation of Thai Fisher Folk Association, Thai Frozen Foods Association and Thai Tuna Industry Association. Interestingly, worker organizations, migrants and NGOs were excluded from serving on the committee. This suggests that, by this point, the government was willing to eliminate disagreements by incorporating employers' demands while marginalizing the voice of worker representatives. As an NGO representative recalls, the representation of Thai and migrant workers was noticeably absent (Interview data with an NGO representative, 25 December 2019). The final agreement on appropriate measures of the Convention, it was hoped, would be reached before the ratification at the end of January 2019. In fact, this version of the Bill that was under deliberation by the committee would later be approved into the Labour Protection in Sea Fishery Work Act (details discussed in the following section).
Employers' demands to lower existing protection that advanced beyond the Convention were eventually accepted. On 8 January 2019, Labour Ministry's Permanent Secretary Jarin Chakkaphark chaired a meeting with Fisheries Department, Port Authorities, the Department of Medical Service and the Command Centre for Combating Illegal Fishing to review the employers' requests. The compromise reached between employers and state officials finally led to the ratification of the Convention on 30 January 2019. The Thai government explained that this decision 'reflects its strong political will to ensure that the working conditions in its domestic fishing industry meet ILO standards' (ILO 2019). Despite meeting the international standard, the analysis below shows that employers weakened existing migrant fisher protection that had already gone beyond the Convention through the process of domestic validation.

VI. Weakened protection
The validated norm of migrant fisher protection has been weakened in implementation by norm antipreneursemployers and fisheries associationswhile retaining the discursive robustness due to the wider support at the domestic and international levels. As illustrated above, such norm antipreneurs collectively mobilized and strategically engaged with the ILO Convention by refusing norms that granted higher obligations and advanced those that offered watered-down protection. The weakening of migrant fisher rights was also facilitated by the government, who granted fisheries industries an exclusive access to contestation in the final rounds of domestic deliberation. The following analysis discusses in detail the extent to which the business sector managed to limit the scope of norm implementation in Thailand.
Assessing the overall robustness in norm implementation, progress was made in a sense that the Thai government eventually ratified the Convention after it resolved normative conflicts with employers. The new law, the Labour Protection in Sea Fishery Work Act, was eventually passed in February 2019 in support of the Convention. It is also worth noting that the decision reached between the Thai government and employers extended greater protection to fishers in two areas. First, compensation schemes have been institutionalized and employers are required to contribute 0.2 per cent of the fishers' annual income to such schemes. Fishers can then claim the fund in the case of workrelated accidents, illness and disability. Second, a more comprehensive health examination was introduced to prevent unfit fishers from working (Ministry of Labour 2018b). In this aspect, fishers previously had to only undergo a general health examination that screened for communicable deceases for immigration purposes; however, with the adoption of the Convention, fishers are now required to take eyesight and hearing tests to attest to their fitness before beginning their employment. This measure is important for safeguarding fishers, especially because they have to handle heavy gear in extreme and hazardous conditions at sea.
Nevertheless, the comparison of the draft legislation and adopted law demonstrates clearly that employers managed to bring down protection standards, and in doing so, decreased the robustness in norm implementation by limiting the scope of migrant fisher protection. Despite the aforementioned advancement, employers successfully used the Convention to water down existing protection standards that had already advanced beyond the measures specified in the Convention. This can be directly observed in the draft legislation circulated at the first public consultation in March 2018. The draft was shortened from 68 articles to just 22 articles in the adopted law. An NGO representative indicated that in the early consultation rounds, the law drafting committee was writing the Bill 'freely' (Interview data with an NGO representative, 25 December 2019). The 68-article version of the Bill also similarly replicated the Convention in terms of both structure and protection provisions. For instance, the accommodation provision in the draft legislation used the exact wording as found in the Convention when translated from English into Thai. Moreover, the 68-article draft legislation largely followed the structure of the Convention, whose substantial protection provisions were organized into four separate parts (Parts III-VI on minimum requirements for work on board fishing vessels; conditions of service; accommodation and food; and medical care, health protection and social security). But when the law was adopted, the protection measures were truncated and lumped together into only eight articles under 'Part II: The protection of sea fishery labour'.
To be more specific, employers' engagement in the applicatory contestation weakened the protection of fishers in numerous aspects: social security, minimum age, recruitment process, repatriation, rest period and wage payment (Khaosod 2018a). In the adopted law, migrant fishers are excluded from gaining all benefits in the social security system, but instead provided with separate health insurance, which employers consider similar to social security benefits. In contrast, Article 34 of the Convention specifies that fishers are guaranteed 'social security protection under conditions no less favourable than those applicable to other workers', and such benefits should be portable to fishers' country of origin (ILO 2007). Consistent with employers' demand, the new law also decreases the minimum age of fishers from eighteen to sixteen years. While the age is consistent with the Work in Fishing Convention, it contradicts other ILO conventions which specify child labour to be under the age of eighteen. Furthermore, it was agreed that the rest period does not to have to be consecutive hours of rest as long as the total hours add up to ten hours a day or 77 hours a week (Ministry of Labour 2018b). Without adequate and uninterrupted rest or sleep, migrants are highly vulnerable to fatigue and accidents on fishing vessels. In addition, while employers were previously required to pay wage via electronic transfer, the agreement now allows employers to pay fishers' wages with cash. The ILO asserts that cash payment could subject migrant fishers to many potential labour rights violations, such as wage theft, wage withholding, illegal wage deductions due to the lack of proper payment verifications (Manomaiphibul, Phochanakij and Judd 2018: 9). With the weakened protection, migrants were arguably better off in many protection aspects before the Convention ratification. As succinctly captured by FRN President and former migrant fisher from Myanmar Ye Thwe, protections are 'as thin as the paper they're written on' (FRN 2021).
While employers have watered down the implementation of migrant fisher protection norms, they are unable to weaken the discursive strength of such norms. This is observable in the concordance with norms and third-party reactions. In terms of the concordance with norms, it is clear that throughout the contestation discussed above, the Thai government was firm in defending both the validity and application of the Convention. This is further evidenced in the ratification proceeding in Geneva, where Labour Minister Adul stated the ratification 'underlines Thailand's full commitment to raising the standards of labour protection for both Thai and migrant workers and eliminating forced labour' (ILO 2019). The official stance of the government is translated into the Labour Protection in Sea Fishery Work Act. The law inscribes the purpose of the Act is to 'set the duty of vessel owners and performance of fisheries labour in accordance with international standards … which promotes and increases the capacity of fishing operations of the country'. Based on this remark, the government defended norms by framing them as acceptable to employers while reiterating the positive impact of the Convention for the employers (e.g. increasing their operation capacity). The Thai government's concordance with norms thus acts to defend the protection of fishers' rights. It is also reinforced by worker organizations and NGOs who had disagreed with employers' demands from the beginning. The official and NGO support therefore helps to ensure that migrant fisher protection norms remain discursively robust.
The concordance with norms closely operates with third-party reactions. As noted above, since 2014, other norm addressees such as the European Union and United States responded strongly to the exploitation and trafficking of migrant fishers. In January 2019, the same month that Thailand ratified the Convention, the EU reiterated that it 'will continue to work closely with Thailand to fight [IUU] fishing and to promote decent work conditions in the fishing industry' (European Commission 2019). The US government took a harsher action against Thailand in October 2019 despite the upgrade of Thailand's ranking in the TIP report. Citing worker rights abuses in seafood industries, the United States revoked Thailand's trade privileges of all Thai seafood products under the Generalized System of Preferences. This decision is expected to cost Thailand US$1.3 billion each year (Auethavornpipat 2019b). This is in addition to the continual threat of sanctions associated with the annual assessment in the TIP Report, which sparked Thailand's domestic reforms in the first place. Moreover, being among Thailand's top seafood importers, the European Union and United States are in the strong position of being able to mobilize their purchasing power and sanctions threats as additional pressure to induce Thailand's compliance with international standards. In sum, through the domestic validation of the Work in Fishing Convention, employers managed to reduce the robustness of migrant fisher protection in implementation. But with both the Thai government and NGOs' support of the Convention, together with a close watch by the wider international community, the norm retains its strength at least in the discourse dimension.

VII. Conclusion
The findings of this article show the limits of existing norms approaches, which pay insufficient attention to international norms when they enter the domestic space. To address such shortcomings, this article has unpacked the processes of domestic validation of norms. In doing so, it tackled the puzzle of why domestic implementation of norms has weakened despite serious efforts and commitment by state officials who earnestly promote norms. This article also presented an agentic approach to comprehending domestic norm validation and moved beyond the domestic norm implementation model (Betts and Orchard 2014) that assumes a static and monolithic domestic culture with which foreign norms have to contend. In contrast, the domestic validation approach advanced in this article reinforces the bourgeoning critical norms literature by revealing agentic and diverse interpretations of norms among stakeholders (Wiener 2018;Zimmermann 2017). It further shows that diverging interpretation instigated applicatory contestation that was resolved through domestic deliberation. However, while the contestation paved the way for a domestic agreement in norm implementation, norms antipreneurs' critical engagement with norms managed to reverse existing progressive reforms by manipulating norms during stakeholder interaction, thereby weakening norm implementation without damaging the discursive strength of the norms. With such findings, a new area of research is open to investigation in order to reveal the conditions under which the domestically validated and yet weakened norm implementation can be reversed and strengthened despite the locally acquired meaning and hence entrenchment of norms.
This article has provided another valuable contribution by capturing the robustness of validated norms as a consequence of applicatory contestation. It does not take for granted that a robust normative understanding would be readily established through norm validationthe notion that entails validated norms would remain unaffected. The findings of this article revealed the nuance and interplay between different robustness dimensions in which some aspect of validated norms lose strength while other aspects simultaneously retain salience. In the case of migrant fisher protection, norm antipreneurs have weakened norms in the implementation dimension by restricting the applicatory scope. However, the discursive support of norms has been reinforced by the concordance with norms and third-party reactions. Such support safeguarded norms from further weakening and prevented norm antipreneurs from radicalizing the contestation to the questioning and dismissal of norms' overall validity. This nuanced finding showed that validated norms could be weakened without completely losing robustness. The mixed picture of 'norm health' sheds a new light on the impact of contestatory practices and points to the importance of norm flexibility, not just in meaning reconstruction but also in robustness. Nevertheless, it also raises a question about Thailand's internalized sense of migrant fisher protection. If the country was to demonstrate its sincere commitment to protecting migrant workers beyond discourses, the Thai government would need to seriously reconsider its decision to please employers and businesses that rolled back migrant worker rights. organized by the Collaborative Research Centre 'Global Dynamics of Social Policy', University of Bremen. Any mistake is solely the author's.