The migration ban policy cycle: a comparative analysis of restrictions on the emigration of women domestic workers

Policies banning women domestic workers from migrating overseas have long been imposed by labour-sending states in the Indo-Pacific region. This article presents the complexities surrounding such bans by developing an overarching model of a migration ban policy cycle, which provides a theoretical framework for understanding the circumstances under which migration bans arise and play out. It examines the history of migration bans for four prominent labour-sending states – Indonesia, Nepal, the Philippines and Sri Lanka - to assess the causes, outcomes and extent of regional convergence of these policies. In doing so, we uncover two prominent policy narratives. The first involves labour diplomacy, where countries employ bans to negotiate superior working conditions and rights for migrant workers. The second concerns paternalist states as ‘protector’, where states are primarily motivated to reaffirm traditional gender norms. We conclude that migration bans have been most effective, both in curbing departures and achieving desired outcomes, when they are primarily motivated by labour issues and not gender politics. Nevertheless, even when used as a form of diplomatic negotiation, migration bans heighten the vulnerability of domestic workers to exploitation by pushing them into irregular pathways fraught with risk.

highlights the cyclical nature of migration bans as negotiating instruments. Indeed, both bans are only the most recent examples of a long-standing policy response by the Philippines and other labour-sending states to ongoing rights abuses in West Asia, which remains the largest destination region for domestic workers migrating across the Indo-Pacific (International Labour Organization, 2016).
Bans affecting women migrants stretch back to the early years of domestic worker migration to the Gulf, with the government of Bangladesh repeatedly banning or restricting the overseas migration of certain categories of 'low-skilled' women workers from early 1981 to 1998 (International Labour Organization, 2014). Since then, there have been periodic attempts by numerous labour-sending states to restrict the migration of workers identified as being particularly vulnerable to abuse and mistreatment, including fairly recent bans by the governments of Cambodia (2011), Sri Lanka (2013), Nepal (2014) and Indonesia (2015). These migration bans have been applied, almost exclusively, to women migrant domestic workers [MDWs], who are widely perceived to be at greater risk of harm due to the circumstances of their employment.
However, the bans can also be mired by paternalist political currents emerging from antagonism between traditional gender norms and the encroaching 'modernity' of increasing female labour force participation (Oishi, 2005). Although migration bans often appear to be enforced in a piecemeal and reactive fashion, they have been embraced by an increasing number of countries and used more frequently over time, drawing nascent academic and civil society attention to their efficacy as a potential mechanism for negotiating better wages and conditions for MDWs (Napier-Moore, 2017;Oishi, 2005). Beyond these discussions, though, there has been little comparative analysis of the varying political logics behind migration bans, the extent to which they might reflect extant gender norms or economic constraints, and emergent patterns of policy convergence or divergence within and between sub-regional contexts.
For all the academic attention directed toward the comparative analysis of policy regimes governing international migration, there has been a disproportionate emphasis on the immigration policies of OECD countries with net migration inflows. The characterisation of wealthy countries as principal countries of destination reflects a knowledge production preoccupied with the political and economic apprehensions of the Global North, and belies the demographic reality that most of the world's international migrants now move between countries of the Global South (Hujo & Piper, 2010; United Nations Department of Economic and Social Affairs (UNDESA), 2019). The comparative study of migration regimes within these countries has been limited (Boucher & Gest, 2015) and, where there is existing scholarship, largely confined to analysis of the rigid immigration policies that maintain the supply of low-wage temporary migrant labour to major countries of destination in West, East and Southeast Asia.
However, migration regimes are mutually constituted by immigration and emigration policies, and among countries of origin whose economies have long been inextricably linked to the vicissitudes of temporary labour migration there are diverse examples of policies intended to promote or curtail foreign employment. Of these, migration bans or restrictions for women domestic workers are perhaps the most commonplace and controversial, yet have remained under-researched at a regional level. Identifying a lacuna in the dearth of cross-national comparisons of partial or total bans on domestic worker migration, we examine the history of migration bans in four prominent labour-sending states in the Indo-Pacific region -Indonesia, Nepal, the Philippines and Sri Lankato assess the causes, outcomes and extent of regional convergence of these policies. Our attention is focused on the migration policies of these South and Southeast Asian countries for reasons of similarity and difference. There are obvious parallels, for example, in the sheer prevalence of temporary labour migration as a livelihood strategy: poorer households in all four countries have looked to foreign employment in the absence of substantive local employment opportunities. Commensurately, the collective remittances flowing into each of these countries has been a significant and growing source of developmental capital (World Bank, 2019).
Amid these broad commonalities are important differences. For Nepal and Sri Lanka, labour migration has primarily involved low-wage employment in the oil-producing economies of the Gulf. Yet, while official statistics indicate that Nepal has low rates of female domestic worker migration in keeping with longstanding restrictions, Sri Lanka's migration pattern has been highly feminised, with MDWs accounting for the majority of departures until the introduction of restrictions in 2013 (Weeraratne, 2016). Temporary labour migration from Indonesia and the Philippines, meanwhile, has also been highly feminised but more diverse in terms of destination regions and the occupational profile of workers. Indonesian MDWs had been fairly evenly split between West and East/Southeast Asia until the introduction of a 2011 ban on women working in the Gulf, and continues to occur alongside smaller feminised migration flows into light manufacturing industries. While Filipina migrant workers labour across a range of occupations, notably as nurses and skilled caregivers, the vast majority are domestic workers spread across Asia and Europe with flows fluctuating in accordance with periodic country-specific bans.
What emerges are diverse histories of temporary labour migration flows and accompanying emigration policies that are united by their focus on women MDWs but also straddle a gamut of cultural and political logics. In the first section, we draw out two major crosscutting themesthe feminisation of employment and the politicisation of gender normsto provide contextual background to the emergence of migration bans. We then develop a 'migration ban policy cycle' that provides a theoretical framework for understanding the circumstances under which migration bans arise and play out. In the following section, we demonstrate variations in our model through historical overviews of feminised migration, gender norms and migration restrictions in each of our four country studies. In the final section, we examine these bans from historical and regional perspectives to explore patterns of convergence and divergence in policy thinking and assess the efficacy of recent interventions. We conclude that migration bans have been most effective, both in curbing departures and achieving desired outcomes, when they are primarily motivated by labour issues and not gender politics. Nevertheless, even when used as a form of diplomatic negotiation, evidence suggests that they increase the vulnerability of domestic workers to exploitation by pushing them into irregular pathways fraught with risk.

The policy origins of migration bans
In this section we highlight the feminisation of employment and the politicisation of gender norms as prevalent themes that inform the policy origins of migration bans applicable to women domestic workers. We situate these two factors as exerting differing degrees of policy influence across labour-sending states, with the relative weighting of either factor altering the shape and course of migration ban policies.

Feminisation of employment
The feminisation of employment is a common thread in the modern political economy of the four countries discussed in this article, as it has been across much of the world. As Standing (1989Standing ( , 1999 has argued, 'feminisation' not only implies increases to female labour force participation rates, but substantive alterations to the gendered division of labour itself. Examples of such shifts can include the concentration of women workers in particular industries and occupations, the feminisation of jobs traditionally performed by men, or a more generalised trend towards a labour market characterised by the precarity that has historically accompanied women's work (Standing, 1999). In each case, feminisation is more a process of labour market segmentation and deregulation than it is a gender rebalancing of participation in paid work. Indeed, Elson and Pearson (1981) long ago argued that the feminisation of employment across the emerging economies of Asia had occurred alongside integration with competitive export markets and predicated on leveraging gender discrimination to cut labour costs. On the factory floor, essentialised gendered traits associated with household workin this case, women's 'nimble fingers'have been refracted through the lens of production to simultaneously devalue the work performed and benchmark higher productivity (Elson & Pearson, 1981). Women have thus been categorically subjected to intensified exploitation at work, yet, with their employment ensconced within prevailing gender norms and characterised as an extension of household labour, seldom experience commensurate readjustments to the unpaid care work they perform, even as they become de facto breadwinners for their families (International Labour Organization, 2018a).
We understand the feminisation of employment through the proliferation of domestic worker migration in a similar vein. International domestic worker migration has been a major avenue of women's employment across the poorer regions of Asia since the 1970s and, by way of remitted income, a vital source of foreign exchange earnings for the countries those workers originate from (Withers, 2019). Domestic worker migration has become one of the largest occupational categories for women in each of the countries we discuss (International Labour Organization, 2013). Yet, like sewing garments, domestic work is framed as an extension of innately 'feminine' work: essentially the commodification of various forms of unpaid care work that women overwhelmingly perform within their own homes. As such, it is work that is typically construed as 'unskilled' and, owing to the oligopolistic characteristics of international demand for domestic work and the absence of substantive labour laws in prominent countries of destination (Wickramasekara, 2016), chronically devalued. Short-term and low-wage domestic worker migration to West Asia has taken place under the limited rights afforded by the kafala system since the 1970s, while similarly rigid migration to Singapore and Hong Kong began en masse in the 1980s before diversifying to other emerging economies in East and Southeast Asia (Oishi, 2005). There has been little alteration to the immigration and employment policies of major countries of destination that have come to rely on temporary migrant workers to perform '3D jobs'the dirty, dangerous and demeaning.
Domestic worker migration across Asia has thus been a form of feminised employment that has some 'typical' features: an increase in female breadwinners, accompanying tensions within established gender norms, and devaluation of women's labour, while also possessing some situationally-dependent issues related to working transnationally (i.e. the implicit reorganisation of unpaid care work in migrant households and difficulties enforcing MDW's human and labour rights in third countries). Of these emergent implications, we identify disruptions to traditional gender norms along with the political and economic implications of exploitation as the most common motivations for the emergence of policies that seek to ban migrant domestic work.

Gender norms
Women from key labour-sending states in Asia have undoubtedly pushed the gendered frontiers of the productive sphere forward through their rising participation as independent labour migrants in the global market. However, the feminisation of overseas labour migration conflicts with prevailing gender norms across Asia with respect to women's roles and identities in the reproductive sphere (Asis et al., 2004). Traditional gendered ideals still depict men to be breadwinners and legitimate labour migrants, while women are portrayed as obedient, domesticated citizens located primarily in the private sphere of the home as nurturing mothers and wives (Chan, 2014;Platt, 2018). Therefore, 'women who move to make a living are deemed out of place' (Khoo et al., 2017, 327). Women's economic contributions from paid work -both at home and overseas -remain subordinate to their domestic responsibilities, as it is considered a male responsibility to support families financially.
The emergence of 'female breadwinners' alongside waning male employmentparticularly in rural areas where traditional agricultural livelihoods have falteredhas placed these norms under stress, resulting in commonplace stigmatisation of working women (Lynch, 2007) and stubborn expectations that women should, irrespective of their paid work commitments or their spouses' lack thereof, perform the bulk of unpaid care work and particularly childcare (Gunatilaka, 2013). This patriarchal familial ideology is now explicitly incorporated into state policy in many instances, which demonstrates the underlying unease towards women's migration despite enabling conditions presented by gendered migration systems and recruitment practices (Khoo et al., 2017).
For example, women in Sri Lanka and Indonesia are required to obtain permission from their spouse or male guardian in order to migrate overseas as domestic workers (International Labour Organization, 2018b;Platt, 2018). Women's access to labour migration not only hinges on their ability to convince their husband or father to provide consent as her 'protector', but in the case of Sri Lanka, it also rests on the government and employment agency to endorse, monitor and evaluate a woman's caregiver role (International Labour Organization, 2018b). Despite women's significant contribution to national development, such policies reinforce the traditional ideology of mother as primary caregiver and moral guardian of the family and father as head-of-household (Abeyasekera & Jayasundere, 2015;Withers, 2019). Women's overseas labour migration, therefore, 'remains couched in notions of domesticated dependencyincluding their dependency upon their husbands, and children's dependency upon their mothers' (Platt, 2018, 93). Accordingly, states permit and encourage the migration of women on the condition that they do not neglect their domestic duties (Chan, 2014). The continuing promotion of women's overseas employment, however, suggests that labour-sending states expect women to fulfil a dual role in both the wage-earning sphere and domestic sphere. This vision of idealised femininity translates into a 'migratory income-earning woman for the sake of the "national family's" larger goal of economic development, a process of transnational domestication' (Silvey, 2004, 253). Thus, labour-sending states encourage, and in some instances, coerce migrant women into remitting their wages back home to their families (Guevarra, 2006). In this way, women migrants are represented as 'economic heroes' or 'martyrs' carrying out a feminine familial duty as 'good' mothers, daughters and wives whose remittances 'save' their countries, communities and households (Chan, 2014).

The migration ban policy cycle
The extent to which the feminisation of employment and entrenched gender norms inform migration policies varies substantially among labour-sending states, including the four countries highlighted in this study. We nonetheless suggest that it is possible to identify a broadly generalisable model of a migration ban policy cycle that is catalysed by a 'crisis event', subject to various forms of contestation and negotiation, and ultimately results in a corrective policy change through which migration bans are eased or revoked. In this section, we offer a broad four-stage policy model to illustrate the circumstances under which migration bans typically arise and play out.
As per Fig. 1, below, we postulate that migration bans are likely to follow a predictable policy cycle, progressing through four generalisable stages that may nonetheless contain important differences arising from the ways in which women's work is socially and economically situated within different labour-sending states.

Stage 1: crisis event
Total or conditional bans preventing women domestic workers from migrating to certain countries or regions are typically implemented after a 'crisis event' involving the abuse or death of a woman migrant worker during foreign employment. Such events readily capture public and media attention, leading to public outrage in labour-sending states (Shivakoti, 2020). Some examples of 'crisis events' that have led to the introduction of a migration ban on women domestic workers are listed in the table below for each of our four country studies (Table 1). During a crisis event, sending states must choose quickly from a limited set of policy options. Shivakoti (2020) lists three policy options frequently employed in such cases: i) do nothing; ii) find a diplomatic solution or; iii) introduce a total or partial migration ban. The 'do nothing' approach has frequently worked for governments as they attempt to wait until the crisis blows over and is no longer the focus of public attention. The second, more challenging, option is to work with the destination country to find a diplomatic solution to the crisis. However, a significant power imbalance between labour-sending and receiving countries is usually present, which fosters unwillingness on behalf of the host state to change its domestic laws to appease countries of origin. The third option, and one that has been adopted repeatedly by labour-sending states across the Indo-Pacific region, involves total or partial bans on the migration of women domestic workers with the rationale of 'protecting' them from possible abuse and exploitation. While this seems to be an extreme option in the governments' policy toolkit, it is also the most convenient option. By implementing a migration ban the government appears decisive during a time of crisis, which mollifies public anger.

Stage 2: policy confusion and resistance
Once a migration ban is in place, it may lead to some policy confusion as the changing nature of its requirements, such as age and country-specific restrictions, can be difficult for migrant workers to understand and adhere to. In practice, the consequences of the migration bans are counter to those intended by labour-sending states. By closing legal avenues for migration, while at the same time failing to provide viable employment and economic opportunities within the home country, the governments' bans push women who hope to migrate as domestic workers for reasons such as poverty and debt to do so irregularly (International Labour Organization, 2015;Weeraratne, 2016). It is well documented that when women domestic workers bypass regular migration processes they are more vulnerable to rights violations and abuse during recruitment and while overseas, as a result of their irregular status (Napier-Moore, 2017). During migration bans, prospective women MDWs do not have access to pre-departure information and training, standard employment contracts, and grievance mechanisms (Napier-Moore, 2017). The bans provide unscrupulous recruitments agents and employers with more power to profit from the worker's undocumented position given the lack of government oversight. Many MDWs deal with local agents engaged in illicit recruitment and transport practices, involving fake itineraries, circuitous travel routes, bribing corrupt officials and paying middlemen for escort services (Hamill, 2010).

Stage 3: labour diplomacy
One possibility, after a migration ban has been enacted, is for the country of origin to work with the country of destination to find a diplomatic solution to the crisis. Laboursending countries with higher bargaining power, such as the Philippines and Indonesia, have occasionally managed to negotiate better working conditions and rights protections for migrant workers through a bilateral labour agreement or MOU before agreeing to lift the migration ban. For example, the Philippines used the threat of a migration ban on domestic workers to negotiate a bilateral agreement with Saudi Arabia in 2012 securing better labour protections and a minimum monthly wage of US$400, despite strong reluctance from the Saudi government. However, it is important to note that where a new agreement is achieved effective monitoring and implementation is often lacking, as most are non-binding and informal MOUs or 'protocols' (Wickramasekara, 2016). Their less stringent nature is preferred by receiving states who do not intend to bind themselves with international obligations.

Stage 4: policy change or reversion
In most other instances, negotiation proves too difficult owing to a power asymmetry between labour-sending and receiving countries. Labour-sending states are at a structural labour market disadvantage and weak bargaining position to demand more rights for their migrant workers. Host states have little incentive to change their labour laws and engage in bilateral negotiations with sending countries because of their relative bargaining power, which hinges on the abundance of low-cost labour available to be sourced from countries of origin across the Indo-Pacific region and, increasingly, Africa. Chi (2008) notes how labour-sending states fear that demanding more protection will lead to the receiving country closing their doors to certain migrant workers and opening up to competing sending states for their labour 'supply'. Such fear is not necessarily misplaced. For instance, Saudi Arabia has on several occasions targeted African countries, including Kenya and Ethiopia, for its labour needs following migration bans on domestic workers imposed by both the Philippine and Sri Lankan governments. As negotiations often fail and rates of irregular migration increase, criticism and political activism from civil society as well as migrant workers themselves can result in a policy change, such as weakening the specific restrictions of the ban, or a policy reversion by lifting it entirely.
Operating with a limited policy toolkit, countries that are heavily reliant on domestic worker migration inevitably confront the social and economic contradictions implicit in the fictitious commodification of migrant labouri.e. simultaneously disembedded as a component of the export economy and enduringly embedded as human lives substantiated by broader social realityand therein struggle to sustainably appease either imperative. Thus, despite differences in labour-sending states' relative bargaining power and the extent to which patriarchal gender norms influence political process, we observe a general adherence to the migration ban policy cycle described above. However, these points of differenceand other contextually specific factorsare important in accounting for contrasting objectives and strategies that play out within the broad arc of this policy cycle. The following section demonstrates variations in our model in more detail through historical overviews of feminised migration, gender norms and migration bans in each of our four country studies.

Country studies
The Philippines The largest concentration of overseas Filipino workers is in domestic work. After a momentary decline in 2007 and 2008, following the implementation of the Household Service Workers Reform Package, Filipina MDWs have increased and more than doubled between 2009 and 2012 (Scalabrini Migration Center and International Organization for Migration, 2013). Women MDWs provide the Philippines with one of its largest sources of foreign currency; the remittances generated from overseas employment have become the hallmark of Philippine economic development (Zhou, 2017). The Philippine government's extraordinary reliance upon remittances has increased since the 1990s. Remittances from overseas Filipino workers reached a new record high of $3.2 billion in 2018, higher by 3.6% compared to the level recorded in the previous year, accounting for 9.7% of GDP in 2018 (Lucas, 2019).
Bans on the migration of women domestic workers have long been a key feature of overseas employment in the Philippines. It was the first country in the Indo-Pacific region to establish a practice of 'labour diplomacy' through withholding labour force resources to increase its bargaining power with countries of destination for better labour rights protections (Napier-Moore, 2017). The Philippine government first imposed a worldwide ban on the migration of Filipina domestic workers in 1988 amid complaints of abuse. This had mixed results in terms of pressuring destination countries to improve migrants' working conditions. Sixteen states, including Canada, Hong Kong and Singapore, responded soon after and signed agreements with the Philippines detailing labour protections (Napier-Moore, 2017). However, many other governments retaliated by slowing down visa processing for all Filipino nationals, affecting so many workers that the Philippine government reconsidered the ban. In particular, Kuwait reacted by specifically enforcing a ban against hiring overseas Filipino workers that lasted 10 years (Sayres, 2007). Interestingly, countries in West Asia where domestic workers experienced most violations did not agree to any bilateral negotiations (Oishi, 2005).
Despite the mixed success of the 1988 ban, the Philippine government banned the migration of domestic workers to Singapore in March 1995 following the highlypublicised execution of Filipina domestic worker Flor Contemplacion for the murder of a child in her care and another domestic worker (Yeoh et al., 1999). The Filipino public depicted Contemplacion as a martyr and national symbol of their economic plight and participation in overseas employment as a pathway out of poverty (Guevarra, 2006). Various human rights, feminist and migration groups 'condemned the Philippine state's failure to safeguard the rights of workers deemed crucial to the nation's economic development, and questioned the humanity of its state-sponsored policy of exporting labour' (Alipio, 2019, 140).
Since then, the Philippine government has repeatedly enforced country-specific bans on the migration of women domestic workers in response to increasing complaints of exploitation. It imposed a ban on the migration of domestic workers to Lebanon in 2006, to Jordan in 2008 and to Saudi Arabia in 2011. The most recent 'total ban' to Kuwait occurred in February 2018, following the death of seven domestic workers and high-profile murder of Filipina domestic worker Joanna Demafelis (Standing, 1999). The Philippine government lifted the ban on domestic workers to Kuwait in May 2018 having signed an MOU with the Kuwaiti government, in an attempt to provide overseas Filipino workers, particularly women MDWs, with additional rights protection in the Gulf State. However, abuse against domestic workers overseas has persisted, highlighting the ineffectiveness of the MOU. In December 2019, Filipina domestic worker Jeanelyn Villavende was raped and murdered by her Kuwaiti employers. In response, the Philippine government announced in January 2020 that it would once again stop sending domestic workers to Kuwait despite the failure of previous bans to address the mistreatment of workers (Gulf News, 2020).

Indonesia
Overseas migration continues to be seen in Indonesia as a mechanism to alleviate the negative impact of high levels of unemployment especially among women in rural areas (Elias, 2013). The financial crisis in 1997 hit Indonesia particularly hard and, intersecting with increased demand for women migrants in wealthier states to work in carerelated employment, saw huge numbers of rural women leave the country on temporary employment contracts (Elias & Louth, 2016). By 2007, women made up 79% of Indonesian migrant workers with most taking up employment as domestic workers in Southeast and West Asia (Bank of Indonesia, 2009). Wages remitted by women migrants not only play an increasingly important role in Indonesia's attempts to ensure economic stability and development, but to household survival strategies among Indonesia's poor.
Although women now make a significant contribution to Indonesia's economy, ingrained gendered ideals continue to inform state and public discourses regarding women's role in everyday life including their labour mobility (Platt, 2018). This is evident by the stipulation that women must have the express permission of a male guardian (typically a husband or father) in order to migrate overseas (Lindquist, 2010). As more women enter the paid workforce, they are expected to fulfil a dual role: to meet their family and marital obligations on the domestic front, while at the same time contribute to Indonesia's economic development (Austin, 2017).
The Indonesian state has responded to widespread coverage of exploitation and mistreatment of women domestic workers overseas in paternalistic ways, such as implementing laws that further restrict their mobility. In late June 2009, the government placed a ban on its citizens taking up employment in Malaysia as domestic workers following an increase in high profile abuse cases. The government emphasised how this action reflected not only the problem of violent abuse against domestic workers, but also the need for better protections and rates of pay for these workers more generally (Elias, 2013). The ban was lifted in 2011 after the respective labour ministries signed a new MOU, which allowed women domestic workers to retain their passports, entitled them to a rest day and safeguarded their right to communicate with families and the embassy.
However, the execution of Indonesian domestic worker, Ruyati Binti Satubi, in Saudi Arabia in June 2011 prompted the government to once again adopt more drastic restrictions on women's migration. Indonesia imposed a ban on domestic workers from migrating to the country to pressure Saudi Arabia into agreeing to a new bilateral agreement offering workers more protection. The government later announced in 2012 that it would stop sending domestic workers abroad altogether by 2017. In May 2015, the government extended its ban on women domestic workers to 21 countries across West Asia and North and East Africa following a series of abuse cases and subsequent public outcry. President Joko Widodo expressed political humiliation, declaring the deployment of women domestic workers 'a dent in Indonesia's pride and dignity' (Platt, 2018, 89). Such national shame might be thought of as 'reflecting gendered assumptions about the inability of the state to protect 'its' women' (Elias, 2013, 403).

Nepal
Temporary labour migration from Nepal is a comparatively recent phenomenon, growing in volume since the 1990s after Nepal liberalised its migrant labour system. A defining characteristic since the 2000s is that the destinations for Nepalese migrants are heavily concentrated in the Gulf States and Malaysia. With increasing numbers of labour migrants working abroad, Nepal's dependence on remittances has also steadily grown over the years, with remittances accounting for 27.3% of its GDP in 2019, aggregating at US$8.1 billion. (World Bank, 2020).
The government has issued over 4 million labour permits to migrant workers since 2008/09, excluding those going to India, Korea and irregular migrants. The Department of Foreign Employment only began to segregate data for employment permits by sex from 2005 (Sijapati & Limbu, 2012). The official data shows that 211,893 labour approvals were issued for women in the last decade, which only accounts for a little above 5% of the total share (Ministry of Labour, Employment and Social Security, 2020). However, this official figure is believed to be inaccurate owing to the recurring partial or total bans preventing women from pursuing foreign employment, which pushes them into using irregular channels. The discrepancy in data can be observed by the fact that Nepali embassies continue to receive requests for assistance from women migrant workers in larger numbers than the official data suggests. For example, the former Nepali Ambassador to Saudi Arabia estimated that 60,000-70,000 Nepali women were working as domestic workers in the country despite the ban (Pandey, 2013).
Gendered and paternalistic ideals have long influenced policies governing Nepali women migrant workers. Until 1998 women were unable to migrate without the written approval of a male guardian, such as a father or husband (International Labour Organization, 2015). A total ban on the migration of women to the Gulf was initially adopted in 1998 following the death of Nepali domestic worker Kani Sherpa in Kuwait. It was claimed that she committed suicide after continuous physical and sexual violence from her employer (International Labour Organization, 2015). Since then, the policy has gone through several changes, including the introduction of country-specific bans and migration bans for women under a certain age or working in certain sectors (Shivakoti, 2020). In it latest iteration, the total ban was lifted in 2015 with a new age restriction of 24 years for 8 destination countries. However, in 2017 the Nepal Parliament's International Relations and Labour Committee instructed the government to reimplement the ban on women migrating for domestic work to Gulf States after its field investigation uncovered widespread abuse and exploitation in the industry (MyRepublica, 2017;Rai, 2020). These bans, which have been lifted periodically and later imposed again as policy directives, lack clear information and consistency, making it difficult for women migrant workers to comprehend and abide by.
By closing legal channels for women to work abroad, without providing alternative local income opportunities, Nepal's policy stance has major implications for women migrant workers. The ban has resulted in women migrants not being able to obtain the official labour permit approval to travel from the Department of Foreign Employment, which also excludes them from the benefits of financial compensation through the welfare fund in case of injury or death abroad (International Labour Organization, 2015). Pande (2014) claims that since women are excluded from government protection, they provide a pool of disposable labour for undesirable jobs in host states while sending remittances back home without imposing any concomitant obligations on their home country. The policy ban, in its different forms, has not prevented women from migrating abroad. Rather, they have found ways to circumvent these bans and continue to migrate in search of work through more risky and circuitous routes, usually through the open border with India.

Sri Lanka
Sri Lanka was another relatively late entrant among South Asian countries participating in domestic worker migration to West Asia, with the first departures commencing in 1977 after restrictions on foreign employment were lifted amid broader reforms to the economy. Sri Lanka had abruptly transitioned from a 'closed economy' predicated on import-substitution industrialisation to an 'open economy' seeking closer integration with global markets (Kelegama, 2006). Gendered labour market segmentation was a key part of this shift. Garment factories established in export processing zones were reliant on women's low-waged employment, while domestic worker migration to the Gulfwhere heightened demand had arisen from contemporaneous migration bans implemented in Bangladesh, India and Pakistan (Eelens et al., 1992) was promoted as a means of securing additional foreign exchange earnings via remittances. Sri Lanka has since had one of the highest concentrations of migrant domestic worker departures in the region, with 'housemaids' accounting for as much as 75% of all foreign employment contracts by 1993 (Ministry of Foreign Employment, 2017). Feminised domestic worker migration has underscored a deepening reliance on remittances, now 8% of GDP and equivalent to 63% of goods exports (Central Bank of Sri Lanka (CBSL), 2018), and has become a vital component of the country's external sector.
However, increasing domestic worker migration has also fomented social tension by disrupting the gendered division of work and care activities. The prevalence of domestic worker migration alongside the waning of traditional masculine livelihoods has undermined the male breadwinner model (Gamburd, 2000) without a commensurate shift in gender norms relating to the performance of unpaid care work and particularly childcare (Gunatilaka, 2013). The 2013 execution of Rizana Nafeek, a young domestic worker charged with murder after a child died in her care in Saudi Arabia, served as a flashpoint for these concerns (International Labour Organization, 2018b). Though initially provoking outrage about the abuse of MDWs, public debate quickly segued into discussion of the social cost of women's migration and the families they 'leave behind' (Withers, 2019). These twin pressures catalysed the implementation of a conditional ban on migrant domestic work, known as the Family Background Report [FBR], later that year. The FBR imposed minimum age restrictions on MDWs bound for various destinations -21 for Singapore, 23 for most Gulf States and 25 for Saudi Arabiaand placed outright bans on the migration of women with children under the age of five (International Labour Organization, 2018b). Women with children older than five are required to complete a FBR, which involves identifying a 'proper' female guardian and attaining spousal consent, which is then subject to government inspections and approval (International Labour Organization, 2018b).
The FBR is the first instance in which Sri Lanka has sought to restrict women's migration, reflecting a hesitancy to disrupt the country's economic reliance on domestic worker migration. Indeed, the introduction of the FBR has overseen a sharp decline in official domestic worker departuresfrom 138,000 per year in 2013 to 72,000 in 2017 (Ministry of Foreign Employment, 2017)and a corresponding dip in remittance inflows, which have effectively flatlined since 2014 (Central Bank of Sri Lanka (CBSL), 2018). This slowdown has prompted questions about the economic viability of the migration ban at a time when Sri Lanka's external debt has reached record heights and additional foreign exchange receipts are sorely needed to offset trade deficits and stabilise a depreciating currency. At the same time, the FBR has done little to protect the welfare of MDWs: there is evidence of an increasing number of women circumventing the ban by migrating to the Gulf through irregular channels that are both riskier and more costly (International Labour Organization, 2018b;Weeraratne, 2016). Given these shortcomings, the FBR appears to have succeeded only in mollifying patriarchal unease about 'female breadwinners' and bringing domestic worker migration into alignment with gender norms embedded in Sri Lankan political discourse (Abeyasekera & Jayasundere, 2015).

Comparative policy analysis
The analysis of restrictions imposed on MDWs from the four country studies presented in the previous section showcases complexity within our overarching model of a migration ban policy cycle. The repeated and enduring use of migration bans by these and other countries shows that, regardless of efficacy, restrictions remain the main policy lever for many labour-sending states and suggests the possibility of policy learning or policy convergence through emulation. As countries of origin have to balance appeasing patriarchal norms at home while considering the economic consequences of labour migration for the country, this policy allows for a performative change to satisfy the public without addressing the root causes of the problem. Policy convergence and divergence can be seen across the four countries of origin, which together reveal two prominent policy narratives: of labour diplomacy and of patriarchal states as 'protector'.
The Philippines presents the clearest example of the former narrative. In recent years, the state has used bans as a form of labour diplomacy and signal to Filipino citizens that it is committed to taking strong action. It has negotiated and signed the most bilateral agreements and MOUs after exerting diplomatic pressure on host countries. The Philippines' strong bargaining power largely stems from the global labour market's demand for well-educated and English-speaking Filipina domestic workers, noted for their 'modern' appearance and professionalism (Ireland, 2018). Therefore, the superior market position of Filipina MDWs has helped to increase the state's influence and emboldened the Philippines to mount a more forceful defence of their interests. However, while negotiations are successful on some occasions, such pressure does not always result in the implementation of systematic changes to prevent future exploitation and deaths. The country-specific bans have often been lifted after the receiving country responds by recruiting domestic workers of other nationalities from competing countries of origin. Accordingly, the Philippine government's structural dependence upon an overseas labour force, along with the short-term benefits of remittances, constrains the extent to which it is able to take meaningful action and strengthen the rights of workers abroad.
The latter narrative of 'patriarchal state' can be seen in the cases of Nepal and Sri Lanka. In contrast to the Philippines, both countries have a fragile market position and consequently less bargaining power with receiving states. Ireland (2018) points to the global labour market's more modest valuation of Sri Lankan domestic workers due to their lower educational qualifications, limited proficiency in foreign languages like English, skin colour and religious makeup, rendering them less desirable to foreign employers. Nepal and Sri Lanka's restrictions, therefore, reflect a more protectionist style of policymaking, grounded in patriarchal norms that construct migrant women as individuals who need protection and surveillance.
In Nepal, women's migration is highly stigmatised, fuelled by media reports on migrant women facing sexual violence in destination countries that equate their migration with trafficking/prostitution (Gioli et al., 2017). While traditional family structures have modernised in recent years, particularly in urban areas, gender stereotypes and patriarchal practices discouraging women's mobility and education persist in rural settings. For instance, a woman's role in more traditional Hindu communities in the Terai tends to be confined to the domestic sphere, in contrast to the Newari and Tibeto-Burman communities where women are more economically active (Gioli et al., 2017). Nepal's migration policies and bans are rooted in concern for women's ijaat (social honour), a concept that is closely tied to their perceived sexual purity within the context of historically dominant high-caste Hinduism. Grossman-Thompson (2016) explains how a woman's honour can be assured in the domestic sphere through social surveillance by her parents and relatives, while being outside of the home is grounds for social censure and an indication that a woman lacks ijaat. Consequently, Nepal's restrictions on women's labour migration uphold gender norms and cultural narratives of social honour.
Sri Lanka's restrictions are also couched in terms of protecting the welfare of domestic workers abroad, but this appears secondary to the preservation of traditional family structures and gender norms that bind them. By banning the migration of women with young children, the FBR 'postulates a scenario in which women working less and caring more comprises an ideal solution to the mistreatment of MDWs and the care needs of their children' (Withers, 2019, 331). Prohibiting women's foreign employment where migration would impair the performance of care activities, the FBR is logically consistent with a deeper patriarchal metanarrative entwined in Sri Lankan nationalism, public discourse and state policy. More fundamentally, by regulating women, and not men, the policy enforces traditional gender norms by placing the provision of childcare and the wellbeing of the family squarely on women's shoulders.
Indonesia appears to adopt an approach that intersects both narratives of paternalism and diplomatic negotiation. The state's bargaining power is gradually increasing with its dominant share of the Malaysian labour market for domestic work and growing preference for 'low cost' and 'obedient' Indonesian MDWs in West Asia, amid complaints that Filipina MDWs are 'too assertive' (Ireland, 2018). For instance, Indonesia successfully negotiated revised MOUs with Malaysia and Saudi Arabia after imposing bans on the migration of domestic workers, allowing them to keep their passports, communicate freely with their families, and a weekly rest day (Elias, 2013). The Indonesian government's promotion of domestic worker migration and reliance on their remittances, however, sits uncomfortably with its depiction of women as dependent upon their husbands and belonging in the domestic sphere despite their increasing economic contributions to the household (Platt, 2018). This is evident by the stipulation that women must have the express permission of their husband or father in order to migrate overseas, and women with children younger than 6 months old are forbidden from migrating in some areas of Indonesia. Therefore, akin to the restrictions in Sri Lanka, women's mobility in Indonesia is seen as incompatible with their maternal role. It could be argued that while the stated intent of Indonesia's migration restrictions is to 'protect' women MDWs following a 'crisis event', the ideological intent seems to be aligned with reinforcing traditional gender norms.
The two overarching narratives present across our country studiesof labour sending states as 'negotiator for' or 'protector of' women MDWsare each informed by differing political objectives that, in turn, determine the relative efficacy of migration bans as a policy measure. For states primarily motivated by a perceived need to reaffirm patriarchal gender norms, migration bans may be ineffective in restricting women's mobility (i.e. due to the availability of irregular migration pathways) but politically instrumental in policing gender norms. A cynical observer might suspect that, in these contexts, the very appeal of migration bans is the opportunity to balance the pretence of intervention against the preservation of remittance receipts. Meanwhile, for states like the Philippines that have used migration bans to negotiate for better labour outcomes, progress has been modest and episodicundermined in large part by economic constraints on the maintenance of bans. Moreover, by trading on the cultural capital of Filipina domestic workers, this strategy hinges upon (and reinforces) a racialised hierarchy of MDWs in the Gulf, to the effect of impeding other labour-sending states from emulating these strategies effectively. Both approaches are therefore limited in their ability to realise sustainable or lasting change to the social or economic regulation of domestic worker migration, a reality evidenced by the frequency and repetition of migration ban policy cycles across the Indo-Pacific region. We determine that the fundamental challenge is that labour-sending states continue to act alone, or in competition with one another, in a situation of pronounced power imbalance that mirrors capital-labour relations in the Gulf. Our analysis of the migration ban policy cycle and its ineffectiveness as a form of advocacy therefore strengthens longstanding calls for laboursending states to collaboratively engage in multilateral bargaining as a 'high road' to achieving rights and protections for migrant workers.

Conclusion: from competition to collaboration
The comparative analysis of the four country case studies illustrates the realities labour-sending states face while making policy decisions related to banning women domestic workers from migrating overseas. Both the domestic politics and patriarchal norms surrounding the role of women in society, as well as a power asymmetry between labour-sending and receiving countries, are key factors driving the decisions of countries of origin and shaping their ability to negotiate superior rights protection. The migration bans in each of the four labour-sending states exemplify their contradictory stance on women's role in national development as remittance generators, and the moral discourses that underscore women's labour overseas as incompatible with their roles of wife, mother and nurturer. While origin states can and do intervene in the form of migration bans as a mechanism to 'protect' women domestic workers, their interest in sustaining remittances limits how forcefully they push for better rights protection in destination states. Labour-sending states are well aware that migration bans are grossly ineffectual at preventing women domestic workers from migrating overseas, but instead drive the process underground and push more workers to pursue unauthorised, and riskier, migration routes. Yet, such reactionary policies are a more convenient option than making the necessary structural changes within the home country and reforming current migration policy and practice.
When states implement bans as a form of labour diplomacy, the receiving state often responds by simply increasing its recruitment of domestic workers from other laboursending countries prepared to accept lower wages and inferior working conditions, rather than agreeing to improve protection. When receiving states are able to choose between a range of countries to satisfy their labour needs, this diminishes the bargaining power of sending states to establish minimum terms and conditions of employment to protect migrant workers overseas. Labour-sending states have so far worked in conflict and engaged in 'underselling' the rights of migrant workers to receiving countries through separate bilateral labour agreements, in order to ensure their share of an increasingly crowded labour market.
Negotiating collectively, rather than competing for market share, would provide sending states with more influence and allow them to implement the same model employment contracts, including reference wages and superior working conditions. This would prevent host states from undercutting such standards and recruiting from other sending countries. Regional Consultative Processes, such as the Colombo Process and Abu Dhabi Dialogue, are forums that afford labour-sending states the opportunity to come together and identify common ground on issues related to migrant workers. Ultimately, adopting harmonised positions through regional agreements that challenge the ever-shifting hierarchy of remuneration rates and conditions of employment across different states is more beneficial for all labour-sending countries in the long-term.