The Domestication of Protection


 This article seeks to address some of the issues surrounding Indonesia’s overseas labour protection programme. Despite the fundamental reform of the programme since 2004, observers have assessed that this programme has faced a considerable implementation deficiency. While many studies have identified that this programme was not successful, in this article I aim to explain why. I argue that the government and civil society organizations have helped define, frame, and impose the term ‘protection’ in the process of policy production. In that process, the two stakeholders have demonstrated ‘the domestication of protection’, or an attitude in which the stakeholders of migration view the problems that appear in overseas labour settings as being ones originating in the migrant-sending countries. Therefore, protection is limited to taking action directed at the migration process within the jurisdiction of the sending country. In my research on this topic, I mainly employed open-ended interviews and field observations between 2013 and 2019 in Indonesia and Malaysia.


Introduction
Since the mid 2000s, there has been increasing public interest in the management of the migration of Indonesian labourers overseas, especially with regards to Indonesia's labour protection programme. One partial explanation for the increasing interest in this topic might be that the plight of these labourers gained greater publicity starting from 2002, when the Indonesian government mishandled at least about 150,000 of deportees from Malaysia who were landed almost at the same time on Nunukan Island (Ford 2006). Another reason could be that in the mid 2000s the Indonesian government adopted its first-ever comprehensive regulation, the uu 39/2004, which reformed the management of its temporary labourers overseas. This comprehensive regulation adopted a systematic labour protection programme as the new government took responsibility for migration management. However, after years of implementation, international aid agencies, scholars, and civil society organizations (csos) have assessed that this labour protection programme has failed to deliver its primary mission, that is, to safely place and ensure the rights of Indonesian labourers overseas (Komnas Perempuan 2005;Susilo, Hidayah, and Mulyadi 2013;Farbenblum, Taylor-Nicholson, and Paoletti 2013). One scholar is even more sceptical, arguing that the state, agencies, and employers in Indonesia have used the notion of 'protection' as an excuse for the control and extortion of migrants (Rother 2017:965). While a description of how this programme failed has been overlooked, only a few scholars have attempted to explain why the Indonesian labour protection programme has not been successful (Palmer 2016;Farbenblum, Taylor-Nicholson, and Paoletti 2013). In this article, I will join the latter scholarship in providing an alternative explanation.
In this article, I examine the Indonesian labour protection programme by focusing on the two main stakeholders in the migration of labourers from Indonesia: the state and csos. I argue that these two actors have helped define, frame, and impose the term 'protection' in the process of policy production. In that process, the two stakeholders have demonstrated what I call 'the domestication of protection' . The term 'domestication' often appears in globalization studies and transnational-politics literature, in which it is referred to as the process in which global/international policies are adopted by national authorities (Acharya 2004;Levitt and Merry 2009). However, in this article, what I mean by the domestication of protection is an attitude in which the stakeholders of migration consider the problems that appear in overseas settings to be ones originating in the domestic/internal setting, or a supply-side problem. Therefore, protection is limited to taking actions directed at the migration process within the jurisdiction of the sending country. I will show that a form of this domestication attitude is observable from the late 1970s to the present, which has contributed to the current state of the labour protection programme in Indonesia.
The data presented in this research were collected between 2013 and 2019 in Jakarta, Sukabumi, Mataram, Sampang, and Kendal, all of which are in Indone-dewanto Bijdragen tot de taal-, land-en volkenkunde 176 (2020) 504-531 sia, and in Kuala Lumpur, in Malaysia. I also conducted open-ended interviews with 31 key decision-makers from the government, politicians, private labour agents, and csos in Indonesia. Aside from interviewing them, I also lived with more than 50 migrant workers in Malaysia from various job sectors, with whom I participated in everyday conversation. Finally, I also conducted media archival research.
I have organized this article in the following way. First, I briefly discuss the discourse on the protection of migrant labourers and how the government plays an active role in providing a protection service. In the second and third parts, I discuss the evolution of the overseas labour protection programme that the Indonesian government undertook in two eras: authoritarian and postauthoritarian Indonesia. Next, I observe the role of csos and their contribution to such a programme. Finally, I discuss the impacts of this protection programme.

The Discourse of Overseas Labour Protection and the Government
Although it often appears in discussions related to Indonesian migration management, the term 'protection' is rarely given attention in detail. Very few scholars have used the term 'protection' in contexts other than migration management. More broadly, it is more common for scholars to point to this kind of activity as being the fulfilment of the citizenship rights of temporary labourers from overseas (Huang and Yeoh 1996:488;Hugo 2009). Such rights are often associated with the immigrants' formal incorporation into the welfare systems of the host countries (Sainsbury 2006). Yet, in major guest-worker-sending countries, the issue of rights fulfilment is also often called 'protection' , partly due to the prevailing narrative that a particular group of overseas workers is considered vulnerable to exploitation in the workplace (Chang 2018:698;Kofman and Raghuram 2010:72). It is the csos that initially propose the use of such a term in the public-policy domain (Gurowitz 2000:872). In such settings, protection is often understood as something that the labour-sending government is providing. As a result, the state of being protected for migrants depends on the structural conditions that surround them. In practice, when protecting the rights of their citizens who work outside the national borders, the labour-sending governments often suffer from a 'lack of jurisdiction' Piper 2002:1017). This means that once across the national borders, labourers become partially excluded from attaining their full rights as citizens (Ball 1997). At the same time, the call to address migrant labourers' precarity arose hand in hand with the discourse on clandestine migration that the migrant-receiving governments were attempting to reduce. As a result, the labour-receiving governments, especially in the Global South, became less willing to provide immigrant labourers with access to social justice (Avato, Koettl, and Sabates-Wheeler 2010). As a result of this, overseas labourers became distant from the rights they should have enjoyed. Rather than relying on the United Nations' international conventions that aim to protect labourers' rights (Bosniak 1991:738), some countries have developed their own mechanisms to counter this jurisdiction limitation. For instance, in Indonesia, the government has obliged domestic workers departing overseas to attend 'prolonged mental and bodily discipline' training to achieve skills that the government believes will 'activate' a self-protection response in the labourers if faced with exploitative conditions (Chang 2018:698-9).
Between the 1970s and 1990s, authoritarian Indonesia approached migration management within the pillar of poverty alleviation in rural areas. To those who advocate a rights-based perspective, such as non-governmental organizations, temporary migration overseas was initiated to protect the economic rights of citizens in many underdeveloped areas (Ford 2003:96-7). From a broader, static economic perspective, Palmer (2016) has argued that the development of this migration programme was the result of the extreme drop in oil prices after the oil boom in the early 1970s had significantly reduced Indonesia's foreign exchange reserves. In these circumstances, the government sought different ways to regain its foreign exchange reserves, one of which was to stimulate the remittances from the export of manual labour. During this period, the government even formed a task force coordinated by the Departemen Perdagangan (the Ministry of Trade) to chart a comprehensive roadmap to maintain the development programme through the temporary migration of labour. In promoting this programme in rural areas, the Indonesian government even looked to adopt an Islamic value as a niche that justified women's migration (Robinson 2000).
Yet, since the early 2000s, many of the labour-sending governments have changed their perspective and come to view overseas temporary labourers as vulnerable to exploitation (Firdausy 2006:146;Ball 2006:132). As a result, the government has imposed a protection programme aimed at controlling and supervising the recruitment of overseas labourers. There are variations in the implementation of this protection programme. In the case of the Philippines, the government has deregulated the programme through implementing a 'fulldisclosure policy' , which obliges labour recruiters, during recruitment, to publicize the terms and conditions of the job vacancies they are advertising. By so doing, the government has made the assumption that migrants are making rational decisions on whether to take the jobs, thus rendering the exploita-dewanto Bijdragen tot de taal-, land-en volkenkunde 176 (2020) 504-531 tion of workers an issue that individual migrants could have rationally avoided before migrating (Tyner 2000).
In Indonesia, the labour protection programme was manifested in the legislation of Law 39/2004.1 Through this law, which I later explore in detail, the Indonesian government has adopted a comprehensive policy intended to ensure the rights of labourers in the pre-departure, departure (this includes the period when labourers are residing in the host countries), and post-departure stages. Yet, in practice, such programmes have never been effective in dealing with the 'departure' stage because of the lack of jurisdiction. In implementing the other two stages, government agencies, through the decentralization of authority, are present as a watchdog.2 This policy aimed to safeguard potential migrant labourers from exploitation in the recruitment process.
In 2017, the Indonesian government replaced Law 39/2004 with Law 18/2017 on the protection of overseas labourers, which I will hereafter refer to as the Labour Protection Law.3 In drafting this law, the government involved the csos, which led to a human rights perspective being better incorporated. A consortium of overseas labourers' networks, for instance, was invited to a parliamentary hearing during the bill-drafting process. As a result, unlike the previous law, which received considerable criticism, csos were satisfied with the Labour Protection Law. csos consider it more comprehensive, as it specifically includes ships' crews in the classification of overseas labourers, involves more government subagencies, further limits the role of private recruitment companies, and decentralizes protection.4 Despite the optimism on the part of csos because of the law's sole focus on 'protection' , the new regulation also faces existential challenges. In this law, the Indonesian government has insisted on protecting overseas labourers in the workplace, repeating its previous mistake regarding the lack of jurisdiction. With regard to protection in the workplace, for instance, the government has specified several protective clauses, including that overseas labourers maintain 'the right to obtain wage based on the minimum standard wage that the labour-receiving governments set and/or on the agreements between Indone- sia and the labour-receiving governments and/or on working contract' ,5 and 'the right to hold their travel documents themselves' .6 However, the application of such policies depends ultimately on the end user and the regulations that apply in the country where the migrant resides (Killias 2010). Once labourers reside with foreign employers, any 'protection clauses' that are stipulated in the Labour Protection Law, or even the host country's law concerning the rights of workers, become less useful, as they do not have jurisdiction.7 Also, despite the extended protection responsibility mandated to many Indonesian embassies, there is often a lack of the organizational resources essential for enforcement. All these weaknesses in enforcing the labour protection programme in Indonesia are a legacy of the earlier approach that authoritarian Indonesia employed in its migration management from the 1970s. The section that follows will focus on the development of the protection policy in Indonesia during the authoritarian era.

State-Organized Overseas Labour Protection in Authoritarian Indonesia
As early as the 1970s, incentives for sending Indonesian labourers to work overseas were triggered by the high demand for manual and domestic workers in the Middle East. In response to such demands, in September 1970, for the first time, the government declared that it 'may agree on exporting Indonesian workers' , although it did not yet intend to establish a formal migration channel with partner countries.8 Because such demands could counter poverty problems in rural areas, as mentioned in the previous section, the government promoted this labour export as a national development programme. As a result, the Soeharto administration, through the Departemen Tenaga Kerja dan Transmigrasi (Depnakertrans, Department of Manpower and Transmigration), formulated a national migration management scheme, facilitating overseas labour recruitment. Soon after the labour migration formally began, problems associated with migrant labourers started to appear, which triggered the government to enact a 'protective' measure. However, given the country's young age, the poor con-5 Labour Protection Law, article 6(1f). 6 Labour Protection Law, article 6(1j). 7 The situation is much worse for domestic workers because the 'household' is often untouched by the host country's regulations regarding labour rights; see Huang and Yeoh 1996. 8 'Pemerintah setudju ekspor tenaga kerdja' , Kompas, 12-9-1970, p. 1. dewanto Bijdragen tot de taal-, land-en volkenkunde 176 (2020) 504-531 nectivity between the islands, and suppressed civil society participation, these problems did not receive enough attention and left the government undertaking reactive instead of preventative measures. As the media highlighted, the problems surrounding migration were often those that would be easily identifiable by individuals in their home country, such as scamming and the withholding of labourers by their recruiters. In the first quarter of 1972, the Indonesian government had dealt with one big labour-recruiting company called Lembaga Pembinaan Tenaga Kerdja Indonesia (lptki, Indonesian Overseas Workers' Service Agency), whose director was arrested for allegedly scamming 1,900 recruits in the pre-departure stage.9 While efforts to tackle migratory problems took place at the national level in that period, overseas labourers were already experiencing abuse in the destination countries that was as yet unheard of in Indonesia.
The first report of labour abuse in the destination country that reached the Indonesian government did not trigger any immediate or comprehensive measures. Reportedly seven Indonesian labourers in Saudi Arabia sent a letter to the Indonesian consulate in Jeddah, informing the consul that their employer had severely exploited them. In the letter, the seven workers described their plight: they were assigned to the middle of the desert, paid less than what was agreed, and were not given access to their documents. Not long after that, it turned out that more workers in the same company were experiencing the same deprivations. News about the letter quickly reached Jakarta. However, instead of taking the necessary measures against the employers and the local labour facilitators in Saudi Arabia, the government domesticated the problem by pointing fingers at the Indonesian labourers as the ones responsible for the problems. Along with the government, the labour-recruiting company, pt Dwi Putra Metropolitan, also downplayed the problem further by stating that it had been a 'misunderstanding' among the workers.10 The government's response to this incident shows that the government preferred a domestication approach in the search for a resolution.
Responding to this situation, the government imposed a sanction on the recruitment company pt Dwi Putra Metropolitan in 1977. This sanction meant that the company had to temporarily cease operations. According to Harun Zain, then the 'Menteri Tenaga Kerja' (manpower minister), the temporary ban was more of an attempt to shift overseas migration to other destinations, as cited in Palmer (2016), rather than a move away from encouraging it altogether. 9 '1.900 orang mendjadi korban penipuan lptki ' , Kompas, 24-4-1972. 10 'Dubes Saudi Arabia tentang kasus 7 buruh Indonesia: Hanya pertikaian perburuhan saja ' , Kompas, 24-8-1977. Bijdragen tot de taal-, land-en volkenkunde 176 (2020) 504-531 This effort, however, failed and instead triggered clandestine recruitment to the oil-rich Gulf countries. The domestication approach of the Indonesian government to protect overseas labourers in the 1980s oscillated between three dimensions: safety, economics, and repression. The safety dimension was made clear in September 1980, when 220 Indonesian labourers were stranded following a ship bombing during the Iran-Iraq war. This incident alerted the Indonesian government that overseas labourers were mainly suffering from the dangers of war. In other words, 'protection' was defined as the provision of safety for Indonesian labourers from a security threat.11 However, the idea that overseas labour protection was about safety provisions in times of war did not last long. Over time, stories about the social plight of Indonesian overseas labourers became widely known.
Despite the many problems surrounding Indonesian overseas labourers, the Indonesian government realized the economic potential of the 'labour export' . To boost the economic return from overseas labourers while ensuring their citizenship rights, from 1982 the government focused on penertiban, or a disciplining measure applied via the labour-recruiting companies.12 The Ministry of Trade, rather than the Departemen Tenaga Kerja dan Transmigrasi (the Ministry of Labour and Transmigration), led this penertiban. This disciplining measure, which was aimed at both recruiters and labourers, should be seen as a simultaneous effort to both boost the economic benefit and 'protect' labourers. However, in the lead-up to its implementation, incidents involving the abuse of overseas labourers in Saudi Arabia also became prevalent and attracted media attention.
In May 1984, public awareness of the plight of Indonesian overseas labourers reached another level as the biggest Islamic authority in Indonesia took part in responding the overseas labour issue. The first community organization to express its concern over the plight of overseas labourers was Majelis Ulama Indonesia (mui, Indonesia's Ulema Council, the highest Islamic authority in Indonesia). The mui criticized the government for not taking the protection of Indonesian labourers in the Middle East seriously. The mui stepped up in this case because it felt it was one of the parties representing Muslim overseas labourers in the Middle East. In response to this criticism, President Soeharto expressed his discomfort. Yet, instead of addressing their plight through a bilateral approach, the government considered such a development 11 'Menteri Nakertrans: Para tki di luar negeri akan lebih dilindungi ' , Kompas, 29-10-1980. 12 See a threat to national political 'stability' . To the Soeharto regime, growing public concern over an issue could snowball into an opposition force against the regime. In a public address, he stated that the plight of Indonesian overseas labourers should have been a matter for internal discussion between the mui and Depnakertrans.
As the government realized that the plight of overseas labour could be a threat to the stability of the Soeharto regime, the government embedded in the penertiban measure an attempt to silence potential critics, including the labourers. For instance, in 1985 Sudomo issued a decree obliging overseaslabourers-to-be to sign a declaration consenting to not speak to either the media or csos about their negative experiences in the countries in which they worked. In addition, to prevent the sexual abuse of Indonesian domestic workers, Depnakertrans even came up with a sexist and vague recruitment criterion for labour-recruiting companies, which was that 'pretty and sexy women shall not be qualified to work as domestic workers ' .13 In addition to the penertiban, the government also adopted the concept of pembinaan for the departing overseas labourers. The meaning of the word pembinaan changes according to the context in which it is used. It can mean 'education' , 'training' , 'services' , or 'fostering' . In the context of overseas labour protection, pembinaan was to be understood as 'indoctrination' . Pembinaan was particularly used during the authoritarian regime in Indonesia to 'straighten' individuals whose ideology was considered melenceng or 'off-course' , which could be interpreted as criticism of the Soeharto regime. Therefore, pembinaan was an ideological inculcation that the government believed would prevent overseas labourers from being 'contaminated' by 'unsur-unsur negative/politis ideologis' (negative elements/political ideologies).14 On many occasions, the newly appointed labour minister Admiral Sudomo warned departing Indonesian labourers not to get involved in any political activities in the destination countries.15 Pembinaan manifested itself in many ministerial ordinances and regulations in the 1980s. This inculcation succeeded the government's Bijdragen tot de taal-, land-en volkenkunde 176 (2020) 504-531 programme of 'Bersih Lingkungan' (Clean environment), which according to Bourchier and Hadiz (2003:184) refers to an attempt to ensure 'political hygiene' from both leftist and rightist ideologies. This 'mental and ideological' straightening out took place not only among the recruitment of government officials and the employees of state-owned enterprises but also in overseas labour recruitment. Furthermore, in 1985, Sudomo distributed an ideological questionnaire meant for screening labourers who were considered melenceng in some vital industrial sectors (Bourchier and Hadiz 2003:215).16 The overseas labour protection that the government undertook during the Soeharto administration was framed within an authoritarian approach that aimed to maintain the regime's power. The government during this era domesticated the protection of migrant labourers using a repressive approach, which included the silencing of workers, ideological inculcation, and the disciplining of recruitment agents in Indonesia. This era was followed by a substantial change regarding the government's approach to labour protection in postauthoritarian Indonesia. I will discuss this in the next part.

The Government's Protection Responsibility in the Post-authoritarian Era
Since 2004 the government has formally extended its responsibility to protect overseas labourers through the enactment of the Migrant Labour Placement and Protection Law (Law 39/2004). In the first sentence of its preamble, the law defines the term 'human rights' , which signified its main goal to protect overseas labourers. More specifically, in point c of the preamble, the law also exemplified the dangers that overseas labourers had experienced, such as 'human trafficking, slavery, persecution, mistreatment, exploitation, and other forms of human rights' violation' .17 Although the law attempted to articulate the values that csos had advocated, csos remain critical of it. csos argue that the Migrant Labour Placement and Protection Law prioritized the economic aspect over protection. This is indicated from the fact that out of 108 articles legislated in the law, only eight of them specifically dealt with protection. csos consider the law too lenient towards private recruitment companies. 16 The questionnaire surveyed interviewees regarding their thoughts on communism (with special reference to the 1965 incident that allegedly involved the Partai Komunis Indonesia, or Indonesian Communist Party) as well as their thoughts on radical Islamism. 17 Law 39 Despite the principal idea that the law advocated human rights values, the government has faced more problems in the enforcement of this law. In implementing the law, the government continues to face what Barak Kalir, Malini Sur, and Willem van Schendel (2012:12) call an 'implementation deficiency' , or simply an inability to put policy into practice. Asis and Agunias (2012:7) indicated that there is a disconnection between two government agencies: Kem-nakertrans18 and the Badan Nasional Penempatan dan Perlindungan Tenaga Kerja Indonesia (bnp2tki, National Agency for Placement and Protection of tki), which are the key government agencies responsible for facilitating the placement and protection of overseas labourers. However, the problems do not stop there. Studies also suggest that problems related to the long bureaucratic chain for the protection of overseas labourers are apparent at multiple governmental levels, including the central (national), the local (provincial and municipal), and those in the receiving country (diplomatic representatives) (Farbenblum, Taylor-Nicholson, and Paoletti 2013;Santoso 2017). Furthermore, Palmer (2016) documented that the implementation of the law has triggered a duplication in the government's authority.
One of the highlighted drawbacks of this law was that instead of simplifying the government bureaucracy-a primary goal that was also mentioned in its preamble-the law stimulated the involvement of multiple government institutions in exercising protection measures. The law, as well as the amended version, situates 'protection' in a three-stage process. In this law, protection is interpreted as [A]ll efforts to protect the interests of overseas labourers-to-be/overseas labourers so that they can fulfil their rights as specified in the government's regulations during their preparation for departure, their stay in the receiving countries, and after they return to their place of origin.19 As opposed to the previous system in which Kemnakertrans was the sole authority, the expansion of protection responsibility into three stages has expanded the number of government agencies involved in the protection of labourers. For instance, Kementerian Luar Negeri (Kemlu, the Ministry of For- 18 Since 2001, the name kementerian (ministry) has replaced the word departemen. Therefore, the name Depnakertrans has changed into Kemnakertrans (which stands for Kementerian Tenaga Kerja dan Transmigrasi, Ministry of Manpower and Transmigration). 19 Article 1, number 4. The amended version, the Labour Protection Law, has the innovation of making each stage more specific and operational.
eign Affairs), through the Direktorat Perlindungan Warga Negara Indonesia dan Badan Hukum Indonesia (Direktorat pwni, or Directorate of Indonesian Citizens' and Legal Institutions' Protection), was given responsibility for screening potential employers in the destination countries, as a result of which a 'demand letter' was issued, allowing the employment companies to recruit labourers.20 However, this expansion was also the consequence of pressure from csos that wanted the government to make itself 'present' in protection.21 Hence, the government has empowered itself by expanding its authority to protect overseas labourers by domesticating the issue through dealing with the entire process of labour recruitment, but not the post-recruitment process.
The expansion of bureaucracy, however, created a political moral hazard for the government in that to be seen to be properly protecting labourers, the government opted to add more institutions rather than utilizing existing agencies when incidents occurred. As an example of this moral hazard, in early 2012, instead of relying on the existing government agencies to deal with the execution of Indonesian domestic workers in Saudi Arabia, President Yudhoyono formed a Satuan Tugas Penanganan Kasus Hukum Tenaga Kerja Indonesia (Satgas tki, or Special Task Force for Indonesian Overseas Labourers Legal Matters). This task force consisted of government officials and civil society activists.22 One clear mission of Satgas tki was to serve as a public relations unit, downplaying the significant public pressure on the president to protect the overseas labourers on death row in Saudi Arabia. The establishment of Satgas tki also indicates the inability of existing agencies, including Kemnakertrans, the Direktorat pwni, and the bnp2tki, to function in this area.
Aside from complicating the bureaucracy of the protection service, the expansion of protection responsibilities triggered a conflict between government agencies. According to Palmer (2016), Kemnakertrans and the bnp2tki engaged in more competition and conflict rather than cooperation. The bnp2tki was initially established to take over Kemnakertrans's role in facilitating the migration of labour overseas. Surprisingly, officials at Kemnakertrans thought that the bnp2tki had meddled in its work, especially in authorizing administrative requirements for private recruitment companies. This conflict sparked ego-sectorial, rent-seeking behaviour among the bureaucrats and led to the domination of private recruitment companies over state regulations. Palmer (2016) argued that the cause of the conflict was that there were too many discretions, leading to a contentious relationship between the government agencies. According to Palmer (2016), government agencies that manage the migration of labourers have initiated too many discretions, or the institutions leader's judgement over particular situations, aside from the existing regulations and decrees. Government agencies relied on many discretions to solve overlapping authorities caused by unclear regulations. However, I would argue that these discretions alone are not enough to explain why such conflict arose.
To understand the conflict between these governmental institutions, one must first understand the policy innovation in protection that came with the setting up of the bnp2tki. The establishment of the bnp2tki was followed by several bureaucratic innovations aimed at increasing public participation in the governmental process. An example of this innovation was the establishment of a crisis centre, a unit within the bnp2tki which emerged from a call centre that processed complaints from migrant workers and their families. On the call centre, the former head of bnp2tki, Jumhur Hidayat, stated: In my first few months chairing the agency, I regularly received a monthold list of complaints sent by regents [regency mayors] from across the country. These complaints did not receive prompt follow-ups, as they did not all immediately arrive on my desk. Today, you [complainants] just find a location where your mobile phone can receive a signal [network] from the provider. Within seconds you can reach us [bnp2tki] via a 24/7 freephone call service.23 As more people benefited from the call centre, the bnp2tki eventually upgraded the unit to a 'crisis centre' that operates under the deputy in charge of protection affairs in the bnp2tki. The establishment of this crisis centre not only created the opportunity for the public to have better access to the bnp2tki but also created jobs for them. Unlike other government institutions whose staff members consist strictly of state officials (pegawai negeri sipil), the bnp2tki invited independent employees and individuals to work as clerks in the crisis centre.24 Henry Prayitno, for instance, was a retired public servant from the era of Depnakertrans. His deputy, Yusri Albima, was a former 23 Interview with Jumhur Hidayat, Jakarta, 27-9-2016. 24 Interview with Henry Prayitno, the then head of the Crisis Centre, 23-9-2013.
overseas labourer, as well as an activist for the Gabungan Serikat Buruh Islam Indonesia (Gasbiindo, Indonesian Muslim Labour Union). Despite the slow service, ngos have frequently used the services of the crisis centre, filing reports about cases they are handling. Jejen Nurjanah, an sbmi activist from Sukabumi, for instance, expressed her gratitude that these days she could chat through a mobile application with staff at the crisis centre when, for instance, forwarding questions from people in her village in Sukabumi about their family members who have gone missing while working overseas.25 Another innovation that the bnp2tki undertook to produce a synchronized database between protection agencies was the Sistem Komputerisasi Tenaga Kerja Luar Negeri (siskotkln), an online computerized registration system. Initially, the adoption of the siskotkln by the bnp2tki invited strong criticism from Kemnakertrans, because the latter believed that it should have implemented such a programme. Yet the law did not delineate which government institution should undertake this responsibility, based on which the bnp2tki took the initiative.26 This programme was part of the 'protection measure' indicated in the Migrant Labour Placement and Protection Law, as well as a response to a demand from Indonesian overseas labour ngos (Krisnawaty 2006:15). In principle, this online registration system fulfilled the aspirations of both the labour recruitment companies (Rudnyckyj 2004:428) and the overseas labour csos, which demanded strict monitoring of the recruitment process. With this online platform, recruitment companies were required to enter their profile in detail in an online database. Agencies also had to upload information about their recruits, including family consent letters, health records, training diplomas, approval letters from the municipal labour office, full addresses, and other required information.
The logic behind this online registration system was that the completion of this profiling aimed to ensure that all applicants follow the formal procedures in the correct sequence before being granted a Kartu Tenaga Kerja Luar Negeri (ktkln, Indonesian Overseas Working Card). This card is the end-product that rationalizes the protection of overseas labourers. The government has often equated possession of the card with the guarantee of protection.27 25 Interview with Jejen Nurjanah, Sukabumi, September 2013. 26 Interview with Jumhur Hidayat, Jakarta, 27-9-2016. 27 Mahkamah Konstitusi (mk, or the Indonesian Constitutional Court) has refused the constitutional lawsuit by 29 ships' crews against the mandatory possession of this card and, on the contrary, restated that the possession of this card will ensure the protection of overseas labourers. The adoption of several innovations in the labour protection policy as outlined above deals with the process of recruiting labourers in the pre-departure stage of migration. The implementation of the siskotkln, for instance, has not only improved transparency but also has substantially decentralized the recruitment process, which was previously dominated by Kemnakertrans. During the inputting to the database, all government agencies responsible for channelling overseas labourers-from the municipal and national levels to the labour attachés in many Indonesian embassies-begin to communicate with one another about the process of recruitment. This transparency, resulting from the implementation of the siskotkln, has enabled government agencies to spot errors and those responsible, for instance, for authorizing 'unfit' candidates as overseas labourers. The impact on transparency that this programme has to offer has thus disrupted the relationship between the government and private recruitment companies, who were previously only responsible to Kemnakertrans. This, I argue, is why such conflict arose between the two institutions. The government was not alone in producing such policies. Civil society groups, which I will discuss in the next part, have fundamentally contributed to domesticating overseas labour protection.

Civil Society Groups and the Domestication of Protection
Along with the government, which considered overseas labourers' problems as a supply-side issue, civil society groups have been essential, regardless of their level of involvement, to providing inputs to the design of labour protection policies. As early as the 1980s, the chairman of the Asosiasi Kontraktor Indonesia (Indonesian Association of Building and Infrastructure Contractors), Santosa Sutrisno, mentioned that the problems of Indonesian workers in the Middle East were the result of ketidaksiapan (unpreparedness) in the pre-departure stage.28 This opinion caused the public to echo the same idea, associating the plight of overseas labourers with the fruits of their own mistakes. Previously, media had reported the problems surrounding overseas labour as being limited to problems of scamming that irresponsible labour agencies based in Indonesia had conducted. Once Sutrisno had made his statement, however, the media helped frame the plight of overseas labourers as being a form of unpreparedness. For instance, the media reported that Indone-28 'Masalah Kontraktor Indonesia di Timur Tengah: Kurangnya Produktivitas Tenaga-kerja karena Persiapan Tak Memadai ' , Kompas, 23-07-1980. sian labourers in Saudi Arabia in the early 1980s were slow, lacking in initiative, and too dependent.29 The domestication attitude that prevailed among stakeholders, especially between government officials and the recruiters' association, pushed the government to take a closer look at the process of overseas labour recruitment. However, instead of building 'checks and balances' into its relationships with private recruiting companies, the government colluded with them. This was because labour export was considered a national development priority, whose success strongly depended on the capacity of the government to recruit as many labourers as it could. In the early 1980s, the government even transformed the function of the Antar-Kerja Antar Negara (Pusat akan, or Overseas Work Facilitation Centre), a government subagency under Depnakertrans, from directly facilitating the recruitment and departure of overseas labourers, to facilitating the work of private labour recruitment companies because of the limited capacity of the Pusat akan to meet the demand for labourers. This relationship also raised awareness among government officials of how lucrative the business was for the personal benefit of the individuals involved. Many studies have revealed that corrupt bureaucrats collected illegal levies during the 1980s (Palmer 2016;Tirtosudarmo 2004). Yet, this situation did not go unchallenged, as it received vociferous criticism from the then leader of the Federasi Buruh Seluruh Indonesia (fbsi, or Indonesian Labour Unions' Federation), Agus Sudono, for prioritizing business interests over those of labourers.
In 1984, the close collaboration between the government and the labour recruiters' association, Asosiasi Pengerah Tenaga Kerja Indonesia (imsa, the Indonesian Manpower Suppliers' Association), further influenced protection policy as imsa was given the privilege of participating in policymaking. imsa's proposal, which perpetuated the domestication attitude, was to legalize field recruiters' (calo) ability to impose what Xiang (2013) calls the 'hierarchy of legality' .30 According to Xiang, the 'hierarchy of legality' is a system in labour recruitment in which the private recruiting enterprises could summon and blame the calo for not selecting and preparing the candidates carefully, should the recruits experience a terrible situation in the destination country (usually when absconding or suffering persecution). In meeting the high demand for labourers, private recruiting companies often relied on the calo, who knew the 29 'Tenaga kerja Indonesia di Arab Saudi: Lamban dan banyak tergantung sikap mandor ' , Kompas, 30-10-1983. 30 'Calo tenaga kerja ke ln dilegalkan ' , Kompas, 6-11-1984, p. 8. dewanto Bijdragen tot de taal-, land-en volkenkunde 176 (2020) 504-531 villages and who lived there.31 By formalizing the role of the calo, imsa implied that law enforcement could prosecute the calo in cases where migrants had had a terrible experience. However, this also meant that private recruiting agencies were freed from their responsibility for contributing to these terrible situations.
Despite the vagueness of this policy, the current administration has also adopted it.
In the early 1990s, there was a substantial change in the discourse in Indonesia on migrant workers' protection. One of the triggers for this was the enactment of the United Nations' International Convention on the Protection of the Rights of All Migrant Workers and Members of Their Families on 18 December 1990, referred to by civil society groups as the Migrant Workers Convention. In addition, it was on the nexus of women's issues and domestic work that the narrative of protection came focused from the mid 1990s to the early 2000s in Indonesia (Ford 2003). The publicity of the plight of female overseas labourers, especially through the discourse of female domestic workers being forced into sex work, and the feminization of labour migration were the two primary driving forces behind the uprising of civil society groups (Ford 2004). Also, after the Cold War had ended, Indonesian civil society groups began to engage with international agencies promoting a human rights agenda in addition to development. Since then, advocacy for overseas labourers by civil society groups in Indonesia has successfully been depoliticized, as the issue of women's protection has dominated the narrative of labour protection. This situation has triggered the establishment of many rights-based csos that focus on overseas labourers (Ford and Susilo 2010).
Since 2010, csos have become increasingly influential in shaping the state's overseas labour protection policy. However, this situation has further rationalized protection in favour of domestication. For instance, in 2014 Migrant Care launched a campaign for the development of a migrant village called Desa Peduli Buruh Migran (Desbumi, Village for Migrant Workers' Care). Two years later, inspired by Desbumi, the Ministry of Manpower established an identical programme called Desa Migran Produktif (Desmigratif, Productive Migrant Village). On the one hand, the programme was meant to implement the 'decentralization' of recruitment and education for potential workers and their families and give the village agency in facilitating communication between workers and their families. With this programme, both the government and csos aimed to prevent potential overseas labourers from falling into trafficking networks. 31 See for instance Lindquist (2010Lindquist ( , 2015; Lindquist, Xiang, and Yeoh (2012); Lindquist (2012).
On the other hand, the implementation of this programme implies that it will distract people in the village and discourage them from migrating. However, despite the assumption that production activities (usually small-scale food production) will be successful and attractive to villagers, there is no evidence yet suggesting that this programme has reduced the number of people wishing to migrate.32 Since the mid 2000s, csos have been constantly addressing the social welfare provision for workers in their land of origin, demanding that the government provide more social protection, education, and health services in the migrant-sending villages (Krisnawaty 2006:33), so that, according to one overseas labour activist, 'the next generation does not need to become overseas labourers like us' .33 Overseas labourers' csos have been unconsciously advocating the domestication of protection because of several limitations in their operations. First, while csos perform a minor role in supporting overseas labourers, they have helped to reinstate the power of the government (Rudnyckyj 2004). csos maintain that the state is the ultimate agent of protection, despite their ability to become one by managing significant resources from an international aid agency. Just as in the Weberian notion of a developmental state, in which the state plays an important role in ensuring the wellbeing of its people, csos also expect the government to actively protect its citizens.34 Even today, csos' expectations of stronger government are noted on many occasions. For instance, Migrant Care and the Serikat Buruh Migran Indonesia (sbmi, Indonesian Migrant Workers' Union) expect the government to take over the role of private labour-recruiting companies in facilitating migration. Anis Hidayah, the former executive director of Migrant Care, even suggested that the gov-32 If so, the moratorium in several countries on recruiting Indonesian domestic workers might be the most relevant explanation of what has reduced the number of migrations. See https://ekonomi.bisnis.com/read/20190108/12/876496/penerimaan-remitansi-menu run-akibat-moratorium-tki-ke-timur-tengah (accessed 21-1-2019). 33 Eni Lestari, the coordinator of the International Migrant Alliance, during the 2nd Indonesian Diaspora Network Congress, Jakarta, 12-14 August 2015. Among others, overseas labour ngos stimulate post-migration entrepreneurship in the villages of origin through the creation of a bulletin, Kabar desbumi, which spreads the message of the entrepreneurial success of ex-overseas labourers. 34 The topic of 'developmental states' sparked an extensive debate between economists, political scientists, and development experts as to the extent to which the developmental state was regarded as a new development model, as well as how different it was to a subtle form of authoritarian state. In sum, bureaucracy has been the central agency in rationally deciding in which direction the state goes to achieve development. See, for instance, the edited volume on this topic by Woo-Cumings (1999). dewanto Bijdragen tot de taal-, land-en volkenkunde 176 (2020) 504-531 ernment should completely take over the functions of the private recruitment companies in channelling migrant workers.35 Second, although migrant labour csos network across national boundaries, their intervention remains spatially constrained within national borders, despite the many problems that originate in migrant-receiving countries. For instance, their campaign for migrant labour advocacy has remained unchanged, focusing on scrutinizing the home country's recruitment system and limiting the operation of private recruiting companies. Even when identifying problems originating in an overseas setting, csos tend to mention actors in the original country. csos such as Migrant Care and Human Rights Watch have been particularly concerned about the operation of recruiters in the country of origin who falsely inform potential labourers about working conditions, the detention-like waiting period in training centres before departure, and the excessive recruitment fees. In another instance, when addressing missing Indonesian labourers in the host country, csos blamed the private recruitment companies for failing to monitor their recruits in the overseas settings, as well as the Indonesian representatives for not adequately offering protection services (Migrant Care 2013:62-5). All the subjects targeted in their criticism would require solutions in the sending countries, downplaying many factors in the destination countries.
Despite the networks of csos that span national boundaries, most csos preferred to focus on protection in the home country because of limited jurisdiction. Consequently, regardless of the comprehension of problems that occur in the host countries, csos are incapable of openly addressing them. Based on my fieldwork, the case of the Indonesian labourers who worked at a birds' nest processing factory in Maxim, Malaysia, could help explain this preference of csos. At this company, a hundred Indonesian labourers in Malaysia who had departed through legal channels were falsely accused of arriving via a trafficking network. The Malaysian magistrate formalized this accusation as a verdict and sent the labourers to jail before deporting them. Initially, their employer, Albert Tei, had relocated the labourers from a cosmetics factory to the birds' nest processing factory.36 Despite the fact that both companies belonged to Tei, the difference between the actual work and the work that was written in the workers' documents was considered a violation of immigration law. When these workers were recruited, they were registered with the cosmetic factory. Due to their lack of jurisdiction and weak labour organization in Malaysia, the Indonesia-based overseas labour ngos, as the labourers' advocate, sued the Indonesian recruiter rather than Tei, who is currently free. The prosecution in Indonesia, which was not completed until mid 2018, resulted in the director of the Indonesian side of the recruitment company being charged with six years of imprisonment for human trafficking.37 The third reason is related to the csos' project delivery. In general, csos operate with external financial resources, a situation which demands not only transparent but also well-calculated and concrete project outcomes. To meet this demand, csos ensure that their programme will succeed. Executing programmes that lie outside the jurisdiction of the csos often risks their success. Projects that are implemented in different states often face risks that are not easy to mitigate. For instance, because unionism is only allowed under a restrictive measure in Malaysia, implementing an empowerment or organizational project there has the potential to breach local law. Under these circumstances, programme-delivery becomes an unpredictable process and the risk of failure becomes higher. Therefore, csos tend to stick to projects within their immediate reach.
In their operation, csos also require the work of petugas lapangan (field agents) to disseminate the cso's programme to their target society, a role that is as vital as field recruiters in a labour recruitment chain (Lindquist 2015). Yet, such field agents are difficult to recruit in other countries. Based on the experience of the sbmi and the Gerakan Aliansi Rayat Daerah untuk Buruh Migran Indonesia (Garda bmi, or Regional People's Alliance for Overseas Labourers), for instance, it is difficult to organize labourers in Malaysia and Saudi Arabia. More specifically, the sbmi has struggled to engage workers in Malaysia since 2014, stating that its aspiration to organize workers was dropped by many Indonesian communities due to issues of trust and value differences.38 These three existential challenges-that is, a strong belief in the developmental state, spatial constraints, and nationally bounded project deliveryhave contributed to the maintenance of domestication in protection. In the next section, I discuss how the current shape of the protection regime has affected the pattern of labour migration from Indonesia.

Some Impacts of the Overseas Labour Protection Programme
The domestication of protection has resulted in a heavy, bureaucratic labour regime, which has caused some adjustments to the attitudes of the parties involved in migration. First and foremost, one obvious impact of the bureaucratic expansion of the current protection programme is a longer and more complex chain of labour recruitment. From the perspective of both migrants and recruiters, this long chain of bureaucracy has resulted in soaring migration costs. Consequently, on the one hand, labour-recruiting agents have had to increase their fees for recruitment. On the other hand, this move has also reduced the number of recruits as they cannot afford the costs involved. Every year several companies are forced to shut down, mainly due to the hefty charges for the cost of migration. The current statistics from the bnp2tki indicate that the number of private recruitment companies in Indonesia dropped from 444 in June 2019 to 316 in March 2020.39 Because of this increase in recruitment costs, labour-recruiting companies have started to divert their business to non-migration-related activities. A representative of a labour recruitment company from Condet Subdistrict in East Jakarta, whom I interviewed in 2013, for instance, pointed out that his company is no longer operating mainly as a labour recruiter but as a travel agent dealing with the legalization of visa and guidance documents, including document translation services. According to the director, his company was not the only one that had diverted its business. Many companies that are members of the two largest employers' associations in the labour recruitment industry, the Asosiasi Perusahaan Jasa Tenaga Kerja Indonesia (apjati, or Labour Recruitment Companies' Association) and Himpunan Pengusaha Jasa Penempatan Tenaga Kerja Indonesia (Himsataki, or the Labour Recruitment Companies' Union), have had to opt for diversification; otherwise, they would have gone bankrupt.
chain in recruitment has made migration more expensive.40 Haikal, an Indonesian overseas labourer from Sakra District in Lombok, West Nusa Tenggara, for instance, recalled during his interview that his nephew had waited for nine months after he had paid the recruitment fees to a labour agent. Yet, he had still not successfully departed to Malaysia, while the debt he had taken out from a loan-shark to fund the departure was due that month. Through official channels, labourers never know when they will finally depart to Malaysia, because it all depends on whether the private recruitment companies can secure demand from their counterparts in destination countries. This experience had driven Haikal and hundreds of people in his subdistrict to avoid official migration channels.41 According to Haikal, overseas-labourers-to-be are generally aware of the formal procedure of migration, yet based on a rational calculation they prefer clandestine migration. In general, statistics show that the number of Indonesian overseas labourers who migrate overseas legally has been declining. Within a five-year period, for instance, the number dropped from 586,802 in 2011 to just 275,736 in 2015 (bnp2tki 2015). The fact that there are individuals like Haikal shows that the moratorium on labour placement in the Middle East might not be the only reason why formal migration from Indonesia has been declining. From a bureaucratic perspective, the transparency in the recruitment system that is part of the overseas labour protection programme was not put in place solely to meet the interests of the workers, but to meet those of government officials too. This transparency has created a 'hierarchy of legality' . This means that whenever the recruited labourers encounter problems in the destination countries, both the sending government and the private companies that had facilitated the recruitment could hold those involved in the earlier steps of recruitment responsible for the problems. Therefore, while it is used to rationalize protection, this transparency has exempted the government and private companies from being the only parties responsible for the plight of overseas labourers. In other words, through this system, the government has developed a measure that shields its institutions from being the target of public blame.
Other than affecting national migration management, the current Indonesian overseas labour protection system has also shaped the labour recruitment policy in the receiving countries. For example, since 2010, Malaysian employers have raised concerns about the increasing costs of recruiting Indonesian 40 Migration costs have been an issue that migrant labourers have always faced (Komnas Perempuan 2005). 41 Interview with Haikal, 28-12-2017. dewanto Bijdragen tot de taal-, land-en volkenkunde 176 (2020) 504-531 domestic workers. In 2018, the Malaysian Maid Employers Association's (mama) president, Ahmad Fauzi Engku Mushein, stated that a Malaysian employer needed to provide at least rm 14,000 to rm 18,000 (or usd 3,350 to usd 4,290) to hire one Indonesian domestic worker.42 The cost covers levies, visas, permits, and medical costs, and yet around rm 10,000 alone goes to, he claimed, the recruitment company, covering the middleman's fee and administrative costs in Indonesia.43 In response to this situation, there has been a continuous demand from Malaysian employers that the recruitment cost should be substantially reduced, or otherwise they would start recruiting domestic workers from Thailand and the Philippines. Because of the high cost of recruiting foreign domestic workers from Indonesia, the Malaysian government has attempted to abandon its existing ties with Indonesian recruiters through the launch of 'Maid Online' , a government-sponsored online platform that bypasses the middlemen in reaching out to potential overseas labourers. However, this development has not been positively received in Indonesia, as apjati, the Indonesian recruiters' association, insisted that based on Indonesian law, the formal recruitment of Indonesian labourers can only be carried out through Indonesian private recruiting companies.

Conclusion
One important development in labour-sending countries' migration management is the initiation of overseas labour protection programmes. In Indonesia, the implementation of such a programme has been rather counterproductive. In this article, I have argued that this counterproductivity is due to the domestication of protection, which has prevailed among the stakeholders of migration since the government took control of overseas labour migration in the 1970s. During the Soeharto era, instead of protecting overseas labourers, the government instead protected its regime in response to the increasing public demand to protect overseas workers from exploitative working conditions. As a result, in this authoritarian era, measures taken to deal with overseas labour protection were approached in a rather more repressive way. Also, the csos blamed the 42 My personal encounters with Indonesian domestic workers in Malaysia confirm that by late 2019 employers would need to spend at least rm 20,000 to hire one Indonesian domestic worker. 43 Ahmad Suhael Adnan, 'Employers can save up to rm10k by hiring maids directly from source countries ' , New Straits Time Malaysia, 27-10-2017. labourers for being unprepared and too dependent on help in the workplaces, which triggered their exploitation. As democratization has taken place in post-authoritarian Indonesia, the state's labour protection programme has accommodated more public opinion and, to a limited extent, invited public participation. With the enactment of Law 39/2004, this era has witnessed the expansion of the government's authority in implementing the labour protection programme through the establishment of the bnp2tki, as well as the expanding role of Kemlu. In addition to implementing this law, the government has also paid attention to csos' criticisms and programme, which has inspired further government protection programmes. However, instead of resulting in a safe and sound channel for overseas labour migration, this protection approach has produced heavyweight bureaucratization, triggered interagency conflict, and encouraged migration through unofficial channels. Therefore, protection in the context of migration from Indonesia only means the rational action of developing measures through an overseas labour protection programme.
This study on migration management might open up opportunities for Indonesian studies' scholars to further study policymaking in post-authoritarian Indonesia. On the one hand, in the current era, public participation has become a central means of producing legal and regulatory products. csos have become more powerful than ever before. On the other hand, csos' involvement in public policymaking has yet to deliver the objective of a regulatory product. In the case of overseas labour protection, policies intervening in this regulatory domain have not achieved the initial goals: safe migration and exploitation-free working conditions. Studies of other subjects of regulation might be necessary to better understand the current process of policymaking in Indonesia. Such studies will contribute to explaining why policy fails despite the democratic process through which it is produced.