I'm Keeping My Baby: Migrant Domestic Worker Rights at The Intersection of Labour and Immigration Laws

Abstract Employment relations systems generally fail to enforce all legal rights of migrant workers. This article illustrates a broader approach to the way labour migration is regulated in practice, using the example of migrant domestic workers in Hong Kong. Political economists have shown that the reality of low-wage migration is either ‘more rights, less access’ or ‘fewer rights, greater access’ in terms of rights enforcement systems. Attention to the effectiveness of such mechanisms and processes reveals another feature of regulation: the divergence of theory from practice. Much scholarly attention has been paid to rights, and this analysis, in which enforcement of those rights is sought, contributes to the literature with a frequently-occurring example of how such regulatory practices effectively restrict migrant rights. The article concludes by arguing that regulation uses employers as a further ‘mechanism of control’ to determine the actual quantity and quality of migrant workers’ employment rights regardless of what is stipulated in the law.


Introduction
Premised on the capitalist logic of cheap and flexible labour, migrant workers in low paid jobs are highly disadvantaged. Guest worker programmes, especially for those from Asia and the Middle East, rarely allow migrants to be accompanied by family members, or allow them to settle even after many years of working in the host country. Their migrant status means their employment is circumscribed by immigration laws and as is well-known, these laws often result in a deficit of both rights and rights enforcement (Ford 2019). If rights enforcement is based on the workers' initiative, migration status and lack of social and economic resources can pose a number of challenges. The extent to which the state is obliged to do more to facilitate the process is a question of considerable complexity, a key challenge being that government agencies fail to provide sufficiently clear answers to common problems faced by migrant workers. The absence of clear answers, as will be discussed, exacerbates the challenges faced by migrant workers. In this article, the rights relevant to female migrant workers who work while pregnant together with their access to maternity leave are discussed as a particularly acute example of how rights are diminished through a combination of their precarious status as a migrant and a lack of clear, reliable advice from government agencies.
In general, employment rights enforcement is 'lopsided', as employers' rights are often protected while their migrant domestic workers' rights are ignored (Tan 2000: 358). There is also a difference between litigation for 'term rights', including recurring rights such as wages or rest days, and 'end rights-only' cases-complaints generated from breach of obligations that arise from the ending of a contract (Tan 2016). Flight or eviction is a very significant factor accounting for the worker becoming a litigant, with the implication that a migrant worker who is not given some of her term rights is unlikely to litigate if she is still employed or if her employment has ended amicably. The former situation is in part a reflection of the broader picture in which employees litigating against an employer while still in the same employment is rare. Further complicating the ability of migrant workers to enforce term rights is that employer sponsorship of their visas serves as a 'mechanism of control', effectively giving employers greater power in employment relationships with migrant workers when compared with others who are citizens or permanent residents (Anderson and Franck 2019: 5). Consequently, migrant workers are more likely to acquiesce to employers' real terms and conditions even though employment law stipulates other, less exploitative, arrangements.
To further elucidate how the 'mechanism of control' works in practice, this article examines a particular aspect-the understudied ability of migrant workers to claim maternity leave-as part of the wider discussion about migrant rights. A term-rights claim based on maternity rights is relatively rare amongst migrant domestic workers in Hong Kong (Tan 2016: 223), and as mentioned, even rarer when the migrant is still employed-as is the situation in our case study. The case study used here is consistent with the paucity of litigation against employers for term rights because it shows how one particular migrant worker was prevented from continuing to push for her rights by the circumstances as they unfolded.
This article complements other studies on awareness and workplace rights of migrant domestic workers in Asia (Constable 2020;Er and Paul 2020;Paul and Neo 2018), as well as the role of diplomatic missions in preventing migrant domestic workers from fully claiming their labour and employment rights (for Singapore, see Marti 2019; for Hong Kong, see Palmer 2013; Sim and Wee 2009). We chose this particular case study because the rights-claiming experience was consistent with the pattern encountered by the first author when he supported over 400 migrant domestic workers to bring a claim for employment rights in his role as senior client advisor and manager of a non-governmental organisation (NGO) programme in Hong Kong. The pattern also reflected the second author's two-decade long scholarly observation of rights enforcement for migrants. This case stood out largely because the migrant worker sought advice much earlier than usual; as such, this reveals a great deal regarding the challenges faced when trying to obtain a helpful response from the authorities.
We used three methods to examine the case study. First, we examined the contemporary legal setting in 2014 when the interactions took place. This was crucial to establish the legal parameters that defined the extent of rights and procedures for addressing violations. In particular, this entailed identifying relevant employment and immigration law relating to maternity leave for migrant domestic workers. Second, we examined correspondence between the migrant worker and government authorities to understand the timing and substance of questions and advice relating to maternity leave. This was essential for establishing whether the rights-enforcement approach could have been more effective. Third, the migrant worker herself was interviewed on five occasions, providing qualitative data that extended our analysis of the law and her case documents. Importantly, this final method revealed relevant information about the migrant worker's anxieties, working relationship with her employer and interactions with various government officials that were not included in her case documents. Our case study draws on all three data sources to construct a robust narrative about how this unusual attempt to claim the term right of maternity leave developed.
This article is divided into four major parts. In the first part, it is demonstrated that the enforcement of migrant workers' labour rights is at best patchy. The discussion goes on to show how migrants find themselves powerless to claim and secure certain rights because of the way national migrant labour regimes are designed. The next part outlines how the Hong Kong Administration has positioned migrant workers in the labour market and considers the mechanisms and processes used to enforce labour rights for migrant domestic workers. We focus on domestic workers because of their dominance in Hong Kong's labour migration regime, and on pregnancy and maternity leave issues because the overwhelming majority of domestic workers are women of childbearing age. The third part then discusses the typical experience of a migrant domestic worker as a case study to illustrate common challenges faced by migrants when attempting to claim labour and employment rights related to pregnancy. The fourth part then discusses the processes involved in resolving this particular case and how they impacted the migrant's ability to press for her rights. Drawing upon this case study, this article confirms the view that although migrant workers are theoretically protected, their rights are curtailed in practice. Moreover, governments in effect use employers as a further 'mechanism of control' to determine the quantity and quality of migrant workers' labour and employment rights despite the law.

International Labour Migration Regimes and Migrant Labour's Bargaining Power
It is not uncommon for the rights of migrant workers to be more apparent on paper than in practice. In several of the Asian countries that host migrant workers, it is well-known that laws and government policies limit the ability of low-wage migrant workers to claim labour and employment rights once their employment has ended or is prematurely terminated (Ford 2019). It is also known that the rules affect other factors, such as the ethnic composition of immigrant populations (Portes and Böröcz 1989). Referred to as a "context of reception", the various "economic, political, legal and other aspects…tend to form more or less coherent patterns organising the life chances' of immigrants" (Portes and Böröcz 1989: 618). In Hong Kong, the context has been expanded to include laws that determine migrants' legal status, residency rights, and social mobility (Sim 2003). This highlights some of the ways in which the number of migrant rights is reduced as part of a systemic attempt to ensure that the workers serve the limited purpose of meeting local demand for domestic labour, as discussed below. In turn, we expand the phrase to include the different ways that law and policy in a host country determine the quality-the extent to which rights are or can be successfully enforced-of migrant workers' rights.
For migrants, while labour and employment laws may not formally discriminate on the basis of their migrant status, their immigration status (unlike citizen workers) is an ever present part of the 'context of reception'. Commonly, migrant labour results from labour importation schemes that may confine the imported workers to specific segments of the labour market, set terms and conditions of employment that are below the minimum standards applicable to citizens or residents, and which make changing employers difficult. It is this immigration dimension of the 'context of reception' that is often overlooked by policymakers, who argue that migrants have the same labour and employment rights as citizen workers as a means of rebuffing calls for better legal protection of labour rights. For example, the Hong Kong Administration argues that it is committed to "safeguarding the employment rights" of migrants, which it has done since 1980 when it "accorded equal statutory protection in employment to foreign and local employees" after ratifying the 1949 International Labour Organization Convention on Migration for Employment (Commissioner for Labour 2014: 12). This is a formalist conception of equality that ignores substantive aspects of equality and fairness.
The above inadequacy is inherent in labour migration regimes, given that they are typically constituted by two overlapping legal and policy frameworks: one that regulates labour and employment rights and a second that determines the right of the worker to remain in the host country. The former sets out minimum terms and conditions for the relationships between workers, employers, trade unions and the government. At the same time, the regime determines the terms and conditions under which migrant workers are permitted to enter and remain. Further, unlike the situation for the governance of migrants who are refugees and other asylum seekers, there is no comparable international architecture for labour migration. For example, the 1990 UN Convention on the Protection of the Rights of All Migrant Workers and Members of Their Families seeks to define human rights for people who migrate and then work. However, by September 2021 the Convention had only been ratified by 56 states. Regarding migration policy development, an uneasy global partnership has emerged between the International Labour Organization (ILO) and International Organization for Migration (IOM) to promote international cooperation amongst nation states (Piper and Foley 2021). Of particular relevance to the present discussion is that international law and the activities of organisations do not constitute a labour migration regime capable of constraining law and policy at the national level.
Unsurprisingly, the lack of an international architecture together with the persistence of ideas of state sovereignty mean that labour migration regimes have been more influenced by national interests than by international standards. Charanpal Bal (2016) refers to the role of capitalism when describing how regimes in migrant labour destination countries attempt to manage the contradictions of capital by rendering migrant workers politically powerless. This contradiction-where capital and labour cannot both flourish at the same time-manifests in the tendency towards a "limitless drive to increase the rate of exploitation" of labour, with the result that wealth accumulates in the hands of capital (employers) while "human misery and degradation" is heaped upon the other (migrant labour) (Foster 1992: 78). The rate of exploitation of labour increases the higher the gearing between capital and labour (O'Connor 1991). Since there is a tipping point after which labour exploitation becomes counterproductive, it is also the role of the state to intervene to set minimum standards for the employment of migrant workers. The labour rights dimension of labour migration regimes certainly plays that role, but not in isolation from immigration policies that are designed (for example) to meet society's demand for labour. The overall effect is to produce workers whose bargaining power is limited and substantially weaker than that of their employers.
Similarly, an important perspective reveals that the political economy of legal rights for low-wage migrant workers is primarily concerned with ensuring they do not compete with national low-wage workers for the same jobs. Governments provide access to low-wage workers from poor countries provided there are controls in place to ensure that the workers do not take other low-paid employment or access other benefits that citizens are entitled to (Ruhs 2013). Migrant workers agree to these restrictions as a condition of entry; however, access to justice in relation to their permitted activity in the destination country is in theory rarely circumscribed. Before detailing how the actual opportunity is removed by a 'mechanism of control', the following sections outline how those rights versus national interests are parsed in practice.

Hong Kong's Laws on Labour Migration of Migrant Domestic Workers
Hong Kong's labour migration regime, in keeping with most guest worker programmes in Asia (see Oishi 2005;Surak 2018), is characterised by temporariness and flexibility for the host territory. This was made clear by Hong Kong's highest appeal court in 2012. Asked by domestic workers to rule on whether the refusal of the Commissioner for Registration to grant them residency rights ('right of abode') was lawful, the Court of Final Appeal found for the government, declaring that the importation of domestic workers is to "fulfil the special, limited purpose for which they have been allowed to come [to Hong Kong] in the first place". 1 The limited purpose here is to meet society's demand for affordable domestic labour to look after homes and care for the young, elderly and disabled (Chan 2005(Chan , 2006. The need for domestic helpers has not fluctuated but has actually increased over the past decade and a half, which was ignored as was the fact that many migrant workers clearly worked in the territory on successive contracts for many years. Related to our topic, it has even been possible for former domestic workers to remain in Hong Kong with their babies despite the law (Constable 2014). However, the ruling with its emphasis on 'limited purpose' immediately draws attention to the twin constraints of labour law and immigration policy when migrants attempt to insist on their maternity rights while they are still employed.

Labour law
In terms of the law, the Employment Ordinance (Cap. 57) is the primary source of labour rights. The Ordinance deals with aspects of employment such as the payment of wages, the entitlement to weekly rest days and various forms of paid leave from work. Relevant to the following discussion, a pregnant worker is entitled to at least ten weeks of maternity leave with pay at 80 per cent of her monthly wage if she has worked under a continuous contract for 40 weeks. With her employer's consent, she can commence the leave period on any date between two and four weeks before the expected due date. If there is no agreement, the pregnant worker should start her leave at four weeks before the due date. Employment protection begins after four weeks of employment. Employers are prohibited from terminating employment contracts, except where the pregnant worker has engaged in serious misconduct or where the employee is still on probation and is dismissed for reasons other than her pregnancy. Migrant domestic workers work under contracts that do not include probation. Breach of employment protection law is an offence punishable with a fine of up to HKD 100,000 (approximately USD 12,900), and upon conviction, the employer should also pay the worker an amount equal to two months wages (and if she was entitled to maternity leave, the equivalent of pay that she would otherwise have received).
A second source of workers' rights lies in their employment contract. Such contracts are subject to the provisions of the Employment Ordinance and must be read together with that statute. The parties to the contract may agree to better terms and conditions but are not allowed to agree to terms and conditions that are below those stipulated in the Employment Ordinance. These principles apply to the government-authored Standard Employment Contract (ID 407) that must be used between employer and domestic worker. This standard form contract is used to specify the particulars that vary between employers (e.g. the address of the employer), employees (place of origin) and the list of duties. Most provisions of the Employment Ordinance relevant to the employment relationship are not repeated in the standard form contract. It does, however, contain two central aspects of Hong Kong's labour migration regime for domestic workers. These are the requirements that domestic workers live at and work only at their employer's address as identified by the contract; they may not live at another address nor can they work at any other address. As we shall see, the live-in requirement posed a particular difficulty for the migrant in our case study, whose experience reflects that of thousands of other migrant domestic workers.

Immigration policy
Foreign domestic workers require an employment visa to work. Moreover, there are three other immigration policies relevant to migrant domestic workers whose employment has ended. First, the New Conditions of Stay (implemented in April 1987) allows migrant domestic workers a maximum of two weeks to remain in Hong Kong from the time the immigration authorities are notified of the termination of their contract. The New Conditions of Stay were introduced to discourage migrant domestic workers from leaving their employment and then remaining in Hong Kong until the employment visa expired. Reportedly, migrants ended their visa-related employment to take up illegal jobs (Labour and Welfare Bureau et al. 2014: 4). Employers complained that this made sponsoring such visas a costly and risky investment. Under the new system, at the end of the two weeks, the migrant worker can apply to the Immigration Department for permission to remain in Hong Kong temporarily. Popularly called a visa 'extension', this takes the form of a visitor's pass and thus forbids employment; the extension cannot later be used for a permanent residency application, and importantly, it removes access to governmentsubsidised public services that migrants previously had as domestic workers.
Second, the Immigration Department requires foreign consulates to endorse employment contracts as part of the visa application process through what is known as the notarisation policy. This policy was introduced in the 1990s because the government of the Philippines insisted on having greater control over the migration of its citizens to Hong Kong as domestic workers (Palmer 2013). Consequently, all consulates are required to endorse domestic worker contracts. Consulates with large numbers of migrant domestic workers (such as the Philippines and Indonesia) impose an additional requirement that only approved recruitment agencies must be used. Unfortunately, recruitment agencies are often 'significant obstacles' to migrant workers claiming their rights (Tan 2016: 224). More often than not, they refuse to find employment for migrants who have a pending labour dispute with a previous employer. In part, agencies react in this way because some potential employers believe that migrant workers with cases are not ideal employees (Palmer 2016). Migrant workers who do not secure the support of an approved recruitment agency to obtain a new employment visa is forced to return to their country of origin, or move to the neighbouring Special Administrative Region of Macau for work (Sim and Wee 2009). This places the migrants in a double bind, as pursuing a claim against a former employer often means that recruitment agencies will not help them find a new one (Ladegaard, 2017). Furthermore, for migrants who choose to pursue their claims, this also means not actively seeking a new employer.
The third immigration policy that negatively impacts migrant domestic workers' ability to pursue claims against their employers is the earlier mentioned rule that requires the migrant domestic helper to reside with their employer: the so-called 'live-in rule'. Since 2003, with the exception of domestic workers already on contracts allowing them to live-out, migrant domestic workers on new employment visas have been required to live at their employer's residence and breach of this obligation is an offence. In recent years, there have been growing calls to abolish the live-in rule. 2 Reasons for these calls include the inadequate accommodation offered to many migrant domestic workers by their employers; the isolation of migrant domestic workers, who are often single employees in the household; the lack of privacy; and the lack of separation between workplace and home, which employers abuse when they impose 2 See HK Helpers Campaign (http://www.hkhelperscampaign.com). Retrieved September 17, 2017. unreasonable demands on their workers day and night. Conversely, the rule has been defended because the live-in option is more affordable for employers, a justification that is part and parcel of the demand for cheap domestic labour. The live-in rule-which overlaps with the similar contractual term to 'reside and work'-complicates migrant domestic workers' ability to claim maternity leave, as the following section demonstrates.
Our Case: Titin and Her Attempt to Claim Maternity Leave In early July 2014, Titin (a pseudonym) walked into an NGO office to seek advice about her employment rights as a pregnant migrant domestic worker. At the time, she was 32 years old and needed to find a solution on her own because the father of her child, a businessman from 'somewhere in Africa', had left Hong Kong and could no longer be contacted. She had worked for her employer for over two years and was thus entitled to employment protection (Employment Ordinance, s 15). Titin was also entitled to paid maternity leave. As specified previously, Hong Kong's employment law also provides criminal penalties for an employer who fails to grant maternity leave or maternity leave pay (Employment Ordinance, s 15A). One of the first hurdles Titin had to overcome was her own erroneous belief that her pregnancy was a ground for dismissal and that she could not give birth to a baby in Hong Kong. Her employment agency had told her-and many agencies tell their worker clients the same-that it is illegal for 'foreign domestic helpers' to be pregnant or to give birth in Hong Kong. However, Titin knew of migrant domestic workers who had given birth in Hong Kong and wanted to do the same. But unlike them, she did not want to choose between caring for her baby and her job. Ideally, she wanted both. 3 It was this desire and her wariness of the employment agent that prompted Titin to approach the migrant support organisation, which then began collecting the documentation necessary for her to claim paid maternity leave. As a first step, Titin obtained a medical certificate confirming her pregnancy. She was already 30 weeks pregnant, and the doctor made arrangements for her to have the recommended pre-natal check-up as soon as possible. That morning, Titin made a copy of the certificate for herself and handed the original to the employer as notification of the pregnancy as required by the Employment Ordinance. The employer's reaction was somewhat unexpected. Instead of congratulating Titin for the good news (a more normal reaction), the employer handed the certificate back to Titin without saying anything. That afternoon Titin consulted the migrant support organisation about the interaction and was advised to send the notification to the employer by registered mail. She did this before returning home in time for dinner with the employer and her family. Everyone in the household behaved as if nothing unusual had occurred, including small talk about the day. No one mentioned Titin's pregnancy then or the following day.
Two days later, Titin's recruitment agency telephoned her to talk about an 'incident' with the employer. The agency explained that the employer had complained that Titin was pregnant and expected to give birth in Hong Kong. The agency told Titin that she could only stay in the job if she terminated the pregnancy, and that if she wanted to keep the baby, she had to resign and return to Indonesia to give birth. Titin informed the agency that she had already consulted a migrant support organisation about her rights and she knew that she could continue working while pregnant, and that it was unlawful for the employer to dismiss her because she was pregnant. The agency then changed tack, reminding Titin that her family in Indonesia was poor and depended on the money she remitted at the end of each month to cover daily expenses like food. Adding an extra mouth to feed would only make their lives harder. This had the intended effect of making Titin feel guilty because she knew her family needed her income; she also wanted to have the baby and was not prepared to consider abortion or adoption. It was at this point that the agency lost patience with Titin and told her they could not offer her any assistance. As Titin had by this time already repaid her loan for the recruitment fee, the agency had fewer reasons to maintain contact with her.
That night, Titin broached with her employer the matter of a start date for her maternity leave. Her employer offered Titin the equivalent sum of maternity leave pay (ten weeks of salary at 80 per cent) in return for resigning. Titin told the employer that she wanted to keep working while she was pregnant primarily because she needed the regular income. Titin realised that she would be unlikely to find new employment because recruitment agencies tended not to recommend pregnant jobseekers to potential employers. Arguably, agencies are merely reflecting employers' preferences; employers see pregnancy as a form of incapacity that reduces the employee's ability to carry out their duties. Even if there were employers who would employ a pregnant domestic worker, Titin was not able to by-pass an agency because her government's representative in Hong Kong required her to find employment only through an agency. Titin was by this stage also aware that the employment visa gave her access to governmentsubsidised healthcare, which meant the cost of pre-natal and childbirth care would be more affordable. The employer told Titin to think about the offer and get back to her in the following couple of days. However, Titin was already 32 weeks pregnant and entitled to commence maternity leave the following week. After consulting the migrant support organisation for further advice, Titin sent her employer a letter confirming the start and end dates of maternity leave.
The night before commencing maternity leave, Titin's employer informed her that she should leave the employer's residence at 9 AM and return at 9 PM each day during the period of maternity leave. This was apparently because the employer believed an employee who was not working was not entitled to be at her place of work during what would have approximated her working hours. Apart from being allowed to return to sleep at night, it appeared that it had been overlooked that the workplace was also Titin's residence. The employer took Titin's house key and told her that the arrangement would continue until the leave ended, including the time after Titin gave birth. The employer also told Titin that she could not bring the baby back to the house, an attitude that is partly encouraged by the immigration policy that prevents foreign domestic workers from bringing children or other dependants to Hong Kong, even if her salary is enough to support them. Clearly, the employer's position was that she engaged an individual to whom the employer was meant to provide accommodation, and this did not include any children of the employee.
Titin was surprised and became anxious about how she was going to care for her baby if they were separated. She visited the migrant support organisation again the next day. As a practical solution, the organisation arranged for Titin to move into a shelter that caters for mothers after giving birth. There was obviously a risk that Titin's departure to a shelter might be interpreted as termination of her employment without notice, and a risk that Titin would be in breach of the conditions of her stay in Hong Kong. There was also a question of whether the employer's action in imposing restrictions on Titin's access to the home and taking away her key were lawful. As these were matters that were not the subject of explicit provisions in either the law or any guidance or advice provided by government departments, the organisation wrote to the Labour Department to seek advice. The Labour Department referred Titin to the Immigration Department (whose live-in rule requires migrant domestic workers to reside with their employers during the entire employment period when in Hong Kong, including during maternity leave taken there). In response, the Immigration Department advised that the proposed arrangement was not an immigration matter but one of employment relations. They referred her back to the Labour Department to temporarily change the address in the employment contract.
Ultimately, the Labour Department agreed they had authority to consider the alteration of the employee's place of residence, and its Policy Section prioritised the case and later permitted the variation of contract allowing Titin to reside at a different address. The employer was generally supportive of Titin living elsewhere with her newborn, largely because the apartment was not big enough for her own family. Titin was pleased with the response. However, the next day the recruitment agency informed Titin that her employer had rejected the variation of contract because she would still be liable to pay for any medical expenses that would arise during the time Titin lived in the shelter. This was because of the scope of the employer's insurance policy. Titin proposed purchasing another insurance policy at her own expense, but the recruitment agency replied that the employer was not interested. Titin was disappointed and informed the Labour Department of the employer's concern. The officer from the Policy Section replied by first empathising with Titin's situation: the officer had herself recently returned from maternity leave and could understand the dilemma from the perspective of a working mother. Next, she apologised that the Labour Department's policy documents did not offer a solution to the problem of an employer who refused to consent to a change of address, and they recommended that Titin get legal advice about how to proceed.
Amidst such uncertainty and without the alteration to her contract, Titin did not move into the shelter but spent the hours between 9 AM and 9 PM in public parks and libraries while she waited to give birth. Titin reported the practice to the Labour Department, but they avoided commenting on whether it was legal for the employer to deny her access to the place where she was legally obliged to live, and again recommended that she obtain legal advice about her rights. Before Titin could do so, she went into labour and gave birth to a baby boy. Titin informed the employer that she was in hospital and so would not be returning to the house that night. The employer congratulated Titin but then reminded her that the baby was not welcome in the house. Titin was due to be released the next day, but hospital staff empathised with the situation and so permitted her to stay for a couple of extra days while they also monitored the baby's health. Titin found herself in a difficult situation; she wanted the job and the maternity leave pay it offered but understood that the employer would only let her keep it if she lived separately to her newborn during maternity leave in Hong Kong. The migrant support organisation informed the Labour Relations Division about the dilemma.
An emergency conciliation meeting between Titin and her employer was arranged for the day after she was due to be discharged from the hospital. However, as the employer did not attend, the conciliation officer conducted the meeting over the phone. Both parties were entrenched in their respective positions. The employer was adamant that Titin should return to the residence if she wanted to keep her job. But Titin would only do so if she could bring the baby, which the employer would not allow. At the crux of the dispute were accusations and counter accusations regarding termination of the contract. The employer deemed that Titin terminated the employment contract by refusing to return home. Titin also accused the employer of terminating the employment contract by not allowing her to return with her baby. The issue of determining which party terminated an employment contract is often a difficult question for employment tribunals. In this instance, this is indeed what the conciliation officer suggested Titin do: litigate to find the answer. The immediate consequence was that the employer informed the Immigration Department that Titin was no longer in her employment. Subsequently, Titin's change in status meant that she had not only lost her job but no longer had access to government-subsidised healthcare or to the right to remain in Hong Kong. Titin could have sued her employer for constructive dismissal, but she chose instead to file a torture claim with the Immigration Department, which prevented her deportation to Indonesia, where she claimed she would face physical violence because she was not married to the baby's father.

Between Labour Rights and Limited Purpose for Migrant Domestic Workers
Titin's experience is much like that of labour migrants in general, as governments tend to treat migrant workers differently; they uphold the rights of citizens but not those of immigrants. Had Titin not run out of all options, she would unlikely have invoked torture-something many other migrant domestic workers in her position have done (Constable 2016(Constable , 2020. Until the final stages, she had engaged with the law; she had not rejected it. She had not set out to claim torture. It is difficult to know whether Titin's claim of torture could have been avoided, but the practice certainly overloads an already burdened system for processing other claims made by asylum seekers and refugees. In any case, the government's handling of Titin's labour and employment rights is a textbook example of how the context of reception in Hong Kong curtails the legal rights of migrant domestic workers. Titin's legal rights are protected in theory, but attention to the workings of systems, processes and actors involved in claiming and enforcing those rights reveal much about the actual quality of rights as they are experienced, which is generally quite low.
In Titin's case, she did not have a stable relationship with the baby's father, who, like her, had no residency rights; thus, she felt she had very limited options to care for the baby in Hong Kong. Titin knew of other Indonesians who left their babies in the care of the fathers when employers insisted that the migrants 'reside and work' at the employer's residence as required in the employment contract. In the evenings and on their weekly rest day, employers would allow them to spend time with their babies.
But this kind of arrangement requires migrants to have strong personal relationships with their employers, who are by default allowed to make rules concerning access to the migrant mother's residence-as in Titin's case. While these migrants may live separately to their infants, Titin did not want to do so, at least during maternity leave following her baby's birth. Rather, she wanted to sleep at the same address as her baby, and after returning from maternity leave, continue working as a 'foreign domestic helper'. With the support of a civil society organisation, Titin proposed solutions that were approved by the authorities, but these still required the employer's consent, which was withheld. Had the officials been firmer with the employer, the employer may not have withheld consent. Approving a variation of a contract to allow 'living out' as it is often called, would have created another way for migrant domestic workers to keep their Hong Kong-born babies in Hong Kong even though they cannot rely on the father's support.
With regard to the period of maternity leave before the baby was born, no official insisted that Titin could not be prevented from being at her employer's during the day. Titin would have experienced less stress in the final stages of her pregnancy if she had a home during the day. They missed multiple opportunities to uphold the law and resolve this issue before Titin went into labour by making it clear to both the employer and worker where they stood in relation to the law and clarifying the law where it was unclear. The final attempt to reconcile the employer and worker the day after Titin left hospital was never going to yield the desired result for the migrant worker. By contrast, Titin's employer got what she wanted. She had already made her intentions clear, well before the baby was born, and reminded Titin of it when she was still in hospital. The baby was not welcome. The authorities do not seem to have emphasised to the employer the concept of constructive dismissal and employment protection, which allowed the employer to remove an inconvenient situation without any penalty. Instead, they enabled an outcome in which the employer's preferences were prioritised at the migrant's expense. It could be argued that Titin's case is consistent with the logic of the low-cost migrant worker scheme that supports the reproductive rights of Hong Kong women by ignoring the reproductive rights of another group of women; namely, migrant domestic workers.
This case enables discussion of another aspect of non-enforcement that takes the form of a lack of definitive answers from government agencies relevant to a migrant domestic worker who is pregnant. As this article has shown, the laws of Hong Kong offer to protect the rights of migrant workers in much the same way as they do for other resident workers, including citizens. In practice, however, the institutional set-up for regulating the importation and stay of migrants for the purpose of working determines the quality of those rights as they are experienced, and even the migrants' ability to have them enforced. This outcome is consistent with the prevailing trend of making a trade-off between the human rights of migrants and access to jobs in the territory of receiving countries (Ruhs 2013). In the minds of policymakers, the choice is a rational one: 'more rights, less access' or 'fewer rights, greater access'. In a way, diminishing the quality of rights enjoyed by migrants serves much the same purpose by allowing policymakers to maintain a relatively liberal scheme for importing migrant labour without upsetting segments of society that are concerned that this openness could result in them sharing their privileges with an enlarged number of residents.
Against this backdrop, it is understandable that governments are reluctant to provide too much definition to the rights and obligations in relation to migrant workers' employment. Doing so would empower migrants who are currently disadvantaged by a status quo that effectively gives priority to the preferences of employers when the law does not provide clear-cut answers to some labour disputes that arise-for example, as a result of pregnancy and childbirth. These answers could serve to counteract false beliefs amongst the migrant workers themselves that their legal rights are somehow conditional on extraneous factors, including "workplace morality and responsibility, and their employer's generosity" (Paul and Neo 2018, 14). Such misconceptions can support employers, who either knowingly deny migrant worker rights or do so out of ignorance of the law. As shown in the case study about Titin, one way of intervening in such situations is to treat the issue as a case that is the result of 'personal motivations', which the government can then address on a case-by-case basis instead of addressing what should have been seen as a systemic failure (Vanace 2011: 935). However, this 'failure' may in fact be an unwritten but intended consequence of a system that enables the continued exploitation of migrant labour.

Conclusion
In migration studies, scholarly attention often focusses on rights in theory rather than rights in practice. This important focus means it is possible to observe how policy regimes impact the power of migrants to demand employment rights (e.g. Wright and Clibborn 2020) and target different groups of people, including migrant workers and refugees (e.g. Schultz et al. 2020). Such analyses enable a better understanding of both the intention and consequences of our immigration systems; however, analysis about how those systems operate to achieve their objectives is missing. As the interactions of multiple government organisations, a migrant support group, a recruiter, employer and migrant worker herself in the case study demonstrate, the state did not explicitly restrict the right. Rather, the employer, already in a favourable position to be an effective 'mechanism of control', was again allowed to control the outcome for the migrant worker. This is not to say that labour officials intended the particular position Titin eventually found herself in. Many government officials would be horrified at the notion, but the unintended-but-systemic consequence is no less disturbing for observers concerned about human rights. This case demonstrates a need to establish an interagency structure with the purpose of better protecting migrant rights and preventing official treatment of the kind Titin experienced. It also highlights the importance of legal certainty for low-waged migrant workers. No law will provide clear answers to all issues; however, the need for certainty has a privileged position in commercial contracts; here, we see the personal consequences of a lack of certainty for already marginalised migrant workers.