Beyond the ‘All Seeing Eye’: Filipino Migrant Domestic Workers’ Contestation of Care and Control in Hong Kong

ABSTRACT This paper draws on ethnographic data about Filipino migrant domestic workers’ perceptions of and responses to the use of surveillance cameras in the home to intervene in recent debates about surveillance, care and social control. On the one hand, our participants disclose what we refer to as the gendered ironies of care and control. Digital surveillance practices in the home not only produce tactics for evading control but also reduce the capacity of migrant workers to deliver the best possible care that is ostensibly the basis for the deployment of new forms of watching. On the other hand, the responses we document here speak to critiques of the Foucauldian vision of surveillance derived from the panopticon that are ‘abstract, disembodied and distrustful’. In contrast to the Benthamite reading of God’s all seeing eye, Filipino migrant workers invoke a relational vision which speaks to connectedness, trust and the possibility of mutual concern. While the use of covert surveillance cameras especially was perceived as undermining the trust necessary for care relationships, some respondents used the devices to provoke face to face encounters deemed necessary to re-establish relations of trust.


Introduction
People who travel to live and work as domestic workers in Hong Kong (and elsewhere) are situated at the nexus of different forms and scales of surveillance. They are subject to extensive file based data collection in pre-departure processes of securing employment, bio-metric border controls and digital and face-to-face forms of surveillance mobilized to monitor their intimate labour (Lee, Johnson and McCahill, 2018). In this paper, we focus on the latter and draw on a range of ethnographic data about people's perceptions of and responses to the use of surveillance cameras in the home to intervene in recent debates about surveillance, care and social control.
In Hong Kong, domestic workers are hired by high-and middle-income groups to look after the children and to perform household chores and a multitude of tasks that stretch the boundaries of domestic work. People employed to perform that work are subject to immigration rules that require them to 'live-in' with their employers and that restrict their ability to change employment. They are also subject to a set of normative expectations surrounding the provision of care in a system of privilege where the dependent child is deemed to be of a higher class and more valued ethnicity than the people who provide their care on a daily basis. The 'home' is a key site where these relational systems of difference and power are problematized, constantly negotiated and enacted on an everyday basis. So how do people, mainly women from the Philippines and Indonesia, living and working in Hong Kong make sense of and respond to the different sorts of surveillance that monitor their everyday behavior and being?
Our research confirms the widespread use of surveillance cameras ('nanny cams') in the home and discloses the consequences and effects of that. Domestic workers highlighted employers' concealment of cameras and perceived violations of their privacy by male employers. When employers do reveal their use of cameras, they normally justify that on the grounds of protecting the elderly or children. In thinking about the impacts of surveillance technologies we begin by drawing together feminist insights about the organisation of care and intimate labour with the observations made by Gary Marx (1981) about the ironies of control. That is while cameras are meant to extend care through digitally enabled forms of control, the results may be precisely the opposite of what is intended. We refer to the gendered ironies of care and control, rather than simply the ironies of control, for the following reasons. First, employers monitoring of domestic workers is legitimated in terms of preventing harm and ensuring that people are well cared for.
Second, we refer to the gendered ironies of care and control because of the specific power dynamics at play. The monitoring of domestic workers is gendered, raced and classed not just because the care delivered by female migrant domestic workers is seen as a replacement for the care work of wives, mothers, sisters and daughters. Rather it is gendered in so far as watching extends the male gaze (even if not exercised exclusively by men) and re-enacts the gendered division of the doing of care (gendered feminine) and the control of that care (gendered masculine). Third, as we elucidate below, the ironies of this new nexus of care and control are evident both in people's tactics for evading monitoring and in the reported diminishment of their capacity to deliver care.
Beyond the ironies of care and control that the use of cameras in the home produce, the people's responses we document here also speak to recent critiques of the vision of surveillance derived from the panopticon that are 'abstract, disembodied and distrustful' and that some theorists contend need to be replaced by surveillance strategies 'that involve more human interaction, more attention to embodiment and trust' (Lyon, 2014: 31). Rather than produce compliant and docile bodies, attempts by employers to install a disembodied gaze engender affective and embodied resistance. In contrast to the Benthamite reading of God's all seeing eye, the people we met and talked to invoke a vision of care and control which speaks to connectedness, mutual concern, and 'a relational vision that supports and protects' (Bauman and Lyon, 2013: pp. 138-39).
The paper is based on ethnographic data from a study on migration and surveillance combining conventional ethnographic practice with 'live methods' that enable people to record their experiences in the midst of ongoing daily activities (Back and Puwar 2012). Participants were recruited through the professional networks and personal connections of one of the authors, Lenlen Mesina, an independent researcher and formerly executive director of a local NGO with a remit to enhance the financial literacy and empowerment of migrant domestic workers. The participants were asked to complete a smartphone diary of their daily lives and encounters with surveillance over a one to two week period and participate in a series of interviews, both prior to and after completion of the diaries: 1 15 smartphone diaries were completed. We also conducted two larger focus group discussions with 17 participants, including people who did not complete smartphone diaries, about their perceptions and experiences of surveillance at different stages of migration. All in all 21 women took part in the research. The research, including informal 1 This research was funded by a British Academy grant: BA SG151983. In order to safeguard participants' privacy and protect other people's anonymity, we asked participants never to disclose employers' identities, take or share photos of other people, or their homes, without their consent. Taking photos of other people's children was to be especially avoided. We also asked our participants to only record diary entries when it was appropriate and safe and to ensure phones were password protected.
conversations and ethnographic encounters in a variety of situations, took place intermittently and iteratively over three years and enabled us to follow people's stories and relations with their employers as they unfolded over this period.
The participants in our study are fairly typical of migrant domestic workers in Hong Kong. They are women from ordinary backgrounds in the Philippines, range in age from early 20s to their mid-50s, from mainly working class but not impoverished urban and rural locales and are all high school graduates, some with college education. Some of the participants have lived and worked in Hong Kong for considerable periods (the longest, 25 years), while others were relative newcomers. For some, Hong Kong is their first experience of work abroad while others had previous experiences working elsewhere (the Middle East and Singapore). The majority were or had been married -including those who were separated and widowed. Their diaries record routine interactions with partners, children and extended family that shape the rhythm of their days and reflect the gender regimes in migration. As Madianou and Miller (2012) have documented in their study of new media among Filipino migrants, the complex and ambiguous relationships of care and control between women and their left behind children are shaped by their own polymedia practices (for example, leaving the webcam on which enables absent mothers to observe, comment on and participate in aspects of life from afar). The use of new media and mobile networks therefore means that Filipino migrant workers continue to perform transnational parenting and community care work at-a-distance in ways that extend and refigure conventional gender arrangements (see also McKay 2016, Francisco-Menchavez 2018. Confronting the gendered ironies of care and control in the surveillance of migrant domestic workers Surveillance has been defined as 'any systematic focus on personal information in order to influence, manage, entitle, or control those whose information is collected' (Bennett et al., 2014: 6). As this definition illustrates, most work on surveillance focuses on control despite the assertion that surveillance is 'Janus-faced' and always operates along a continuum between 'care' and 'control' (Lyon, 2007: 22). The emphasis on control is also evident in the wider theoretical debates that have dominated surveillance theory. Thus, while some writers argue that the emergence of new surveillance technologies is consistent with the 'disciplinary power' and 'self-governing capabilities' identified by Foucault (Staples and Decker, 2008), for others disciplinary power has been replaced with 'modulation' which works through models, simulation, codes, statistical tracking, and new methods of social sorting (Bogard, 2012: 32-33).
In contrast, our research has attempted to un-pack the 'Janus-faced' nature of surveillance. As Bauman and Lyon (2013: 141) point out, 'the lines separating "care" from 'dependence", and "freedom" from "abandonment", are endemically contentious; each apparent opposition seems more like a couple of inseparable (indeed, complementary) aspects of the same relationship'. This is explicit in Foucault's (2007) later work where care is not counterpoised to control, but rather is deemed to lie at the heart of modern practices of governmentality. The central message is that 'we must train ourselves to recognize that benevolence and coercion can be two sides of the same coin' (Moore, 2011: 257).
At the same time, we have responded to the call from 'feminist surveillance studies' (Dubrofsky and Magnet, 2005) to historicize research in this area by showing how the discipline-control thesis lacks historical specificity and a consideration of how global changes might take on different forms across different geographical locations or different migrant groups in a transnational setting. Ethnographically, it is safe to say that migrant domestic workers are keenly aware of the proximities of care and control, benevolence and coercion, and the paradoxes and ironies of surveillance that brings. Feminist scholars too have repeatedly shown not only how ideologies about and the organisation of care is gendered, but also how practices of control are central to contemporary care regimes that not only make and mark distinctions between women and men but also variously rely on state enforced boundaries between 'foreign' care providers and citizens (Graham, 1991;Bakan and Stasiulis, 1997;Michel and Peng 2017).
People employed to provide care, like the often unpaid female labour they are employed to replace or supplement, are frequently ascribed a kind of innate capacity to care that is associated with specifically gendered, ethnic and/or racialized bodies. The production (and scrutiny) of those caring bodies, however, begins well before arrival. As Guevarra (2006) has shown, while the Philippine state introduced 'gender-sensitive criteria' within its overseas employment policy framework in recognition of Filipinas 'vulnerabilities' in the workplace, this attempt to 'empower' Filipinas was 'not only about producing economically productive workers, but also about generating 'good' wives, mothers, and women' (Guevarra, 2006: 523; see also Lee, Johnson and McCahill, 2018).
In sum, feminist scholarship alerts us to the ways that the organisation of care underpins relations of power. However, it also invites us to consider the ways that an ethics of care runs up against, transgresses and might enable a fundamental reform of gender regimes premised on domination and control (see e.g. Gilligan 2013; Richards 2014). By an ethics of care we do not mean a universally agreed set of rules and definitions about care, but rather forms of moral reasoning arising within practices of caring labour that emphasize relationships, connectedness, responsibilities, and particularity (Walker 2007: 58). These interventions by feminist writers raise the possibility of a broader critique of the disembodied and objectifying surveillance practices that take place in wider public sphere. Noddings (2002) argues that those engaged in public policy and social justice would do well to start in the domestic sphere where the origins of benevolent forms of watching and care have their roots: arguments that are further complicated by the growth of new media and mobile technologies which blur the public-private divide and reveal many layers of complexity in notions of care and coercion which have yet to be adequately addressed in the literature.
Similarly, a number of writers have sought to extend critically the Foucauldian vision of surveillance regimes represented in the 'panoptic' prison, drawing upon the philosophical work of Emmanuel Levinas and the ethics of proximity (Bauman and Lyon, 2013;Introna, 2003;Lyon, 2014;Stoddart, 2011). For Levinas, 'ethics starts with the impossibility of being indifferent to the Other … In being exposed to the other as Other the ego becomes unsettled, shaken, fundamentally and irrevocably interrupted' (Introna, 2003: 2). Lucas Introna, for example, suggests that with the emergence of electronically mediated forms of surveillance and interaction 'the primordial source of our social being is being eroded' (Introna, 2001: p. 12). In the context of global surveillance and national security, 'it is easier to … ban at the border certain categories of ethnic or national origin, when the bodies and especially the faces of the persons are absent' (Lyon, 2007: 193). Nearer to home, it is much easier, Introna contends, to deny assistance to a needy person who presses the button on an intercom system in a gated compound, than it is to ignore the person who knocks on the door because their face disturbs you. And finally within the home, the objectifying gaze has now entered the 'private' or 'domestic' sphere in ways that enable employers to monitor migrant domestic workers at a distance through parent blogs, mobile phones and nanny cams (Brown, 2011).
As Zelizer (2010, 268) contends, paid and unpaid care in intimate settings, such as the home, are dependent on relations of trust; that is, a willingness to share and safeguard intimate knowledge about other people. Evidently, the use of 'nanny cams' significantly challenges those relations of trust. Our participants repeatedly record both the violations of their privacy and the feelings of alienation from employers engendered by the use of webcams. They also reveal the consequences of those forms of watching in terms of forms self-monitoring and evasion that they perceive as detracting from rather than enabling care. The latter is resonant both with forms of disciplinary power, ala Foucault, and the ironies of control identified by Marx (1981). However, as Lyon (2014) suggests and as our participants confirmed, social technologies of control do not entirely remove the relational and embodied aspects of intimate relations but rather makes them even more affectively acute. More specifically, the participants in our study recounted the ways they confronted the perceived breakdown of trust necessary for care relationships that the covert use of digital surveillance especially is seen to register and engender. Renewing trust, they suggested, required face to face confrontation and some participants described the ways that they used the 'nanny cam' as a device to provoke and precipitate those face to face encounters in an attempt to reestablish relations of trust necessary to the work of care.

Migrant Domestic Workers and Surveillance in Hong Kong
Hong Kong is a pertinent site for exploring these theoretical issues because of its significant Hong Kong is the site of some important grass roots and transnational migrant workers' movements. However, many domestic workers still endure difficult working conditions such as the deprivation of their weekly rest day, the denial of adequate food and accommodation, and threats of or actual physical abuse. These working conditions may be compounded by debt bondage (because they have to pay exorbitant placement fees to agencies) and limited freedom of movement because their identity documents are confiscated by agencies or employers (Constable, 2007(Constable, , 2014.
Surveillance of migrant domestic workers became a subject of public debate when the Office of the Privacy Commissioner for Personal Data (PCPD) (2003, p. 4) proposed to put forward a code of conduct based on principles of proportionality and transparency for employers who monitor and record the activities and behaviour of employees at work in accordance with the existing Personal Data (Privacy) Ordinance. 2 Opinions about employee monitoring were divided in the ensuing public consultation exercise, especially over household employers' use of the camera to monitor the activities of migrant domestic workers in the home. A body representing employers argued that migrant domestic workers might 'abuse their privacy rights', that 'employers need to monitor their helpers by CCTV for evidential purposes in case their young ones are ill-treated' (Office of the Privacy Commissioner for Personal Data, 2003, p. 13), and that 'the employment of domestic helpers is unique' and should be exempted from the proposed code of conduct (p.

30). Migrant domestic workers took a different view:
Their principal appeal was for their employers to treat them fairly. They are generally opposed to employers "spying" on them and using pinhole cameras to record their movements in the domestic household. If that wish cannot be accommodated then they are of the view that employers should notify them of an intent to use monitoring cameras. They are strongly opposed to employers using monitoring cameras in areas where there is a clear expectation of privacy such as toilets, bathrooms and bedrooms. (ibid p. 31) In the end, the PCPD adopted a 'self regulatory framework', extended 'best practice' guidelines on employee monitoring to migrant domestic workers, 3

General experiences of surveillance in the home
Among the group of 21 participants in our study, the majority have had previous experience of working for employers who used surveillance cameras in the home. Most were not told in advance about the use of cameras. They discover the use of cameras indirectly though they generally remain uncertain about the number and location of cameras. Employers who do disclose the use of cameras are more likely to do so when the domestic worker is given responsibility for looking after the elderly or young children. Whether told about the use of cameras or not, they were generally neither informed about who had access to the images nor whether the images were recorded and stored and if so for how long. In one exceptional case, cited as an example of good practice, one participant reported that her employers had not only disclosed the use and location of cameras in the home, but also downloaded an app on her phone to enable her to view the images from the webcams. She said that on her days off she would sometimes check the app to see what her employers were doing before she came home.
Overall, the women that we spoke to said that they preferred employers not to use cameras.
However, they were more likely to tolerate, if not fully accept, that when employers disclose their use in advance and when justified by on the basis of the care of younger children and infants. Some reported that cameras in certain circumstances could be beneficial to employees, for example, to be used as evidence for work completed and hours laboured. However, most challenged the conditions under which the cameras were used. They talked frequently about the lack of trust that the use of cameras conveyed. More positively, dealing with cameras produces also a set of stories about how that lack of trust is overcome: the presence of the camera becomes a marker of whether or not a relationship of trust has been established with employers. In some cases, cameras were reported to have been physically removed after a period of time and domestic workers told by employers that cameras were no longer needed because they now had confidence in the employee. In other cases, domestic workers report that they are no longer anxious about the camera because their employers are no longer continuously challenging them about every little detail of their observed behavior. That was assumed to mean that they are no longer continuously under the gaze of the camera.
The women we spoke to who were subject to covert surveillance discovered the camera in a number of ways, including being told by children that 'daddy was watching', physically discovering the camera equipment while cleaning, or observing employers comparing images on smartphones and/or adjusting cameras. Most often, they discovered their employer's use of cameras when employers called to ask them about an event or an action they observed at a distance. Alternatively on return from work, an employer might ask or challenge them about a prior event or action that indicated they had either been watching at the time or had subsequently reviewed the days' events which had been recorded. In one case grandparents reportedly called and indicated they had observed something on the camera. Either way it is evident that, among this group of women at least, while most employers did not openly reveal the use of cameras, they did not go out of their way to entirely conceal their use. In that way and in partial conformity with the idea of the panopticon as analysed originally by Foucault, this practice of control operated through the continuous ambiguity of never knowing when one is being watched, by whom and exactly what one might be called to account for.

Tactical resistance, the ironies of control and its potential impact on care
Though the use of cameras might be understood as a practice of coercive control, as we suggested in the introduction, it did not simply produce docile and self-disciplining subjectivities. Rather, women respond in a variety of ways that in part undermine those practices of control. The first is through reported forms of tactical resistance to obscure and avoid the camera and/or test and disrupt the employer's gaze. The latter may be done through small and discreet ways, reminiscent of the 'foot dragging' described by Scott (1990), such as 'accidentally' dropping cleaning cloths on top of cameras or moving picture frames while dusting to obscure their view. More generally, women reported making use of the small space available in the homes where, they assumed or had learned, there were no cameras or where they were out of the view of the camera. So, for example, a number of women reported routinely using the toilet as a place to check their phones: evading temporarily their employer's watching so that they could, in turn, browse and check on their children and family member's social media page. Some of the women reported they did this because employers had made comments about their use of phones while at work, others because they were worried that might convey the impression that they were not fully devoted to their duties. The anxieties about using their phone were poignantly recorded by one of the participants in the smartphone diaries who recorded how she was working in the kitchen and had received a text from her family informing her that one of her daughters had been taken seriously ill back in the Philippines. Though she longed to simply call them on the spot, she had to wait until her employers had gone to bed and she knew that they were not watching or listening in order to be able to safely and comfortably do so.
The second response that often goes alongside of more intentional acts of tactical resistance is an increased awareness of how their actions might be perceived by the employers who are observing them remotely on camera. Angie, a recent arrival to Hong Kong, recounted how she had been playing and dancing with her new employer's young daughter when she received a call from the employer asking her why she was shaking her daughter and telling her to move the camera so she could see her. When she saw that her daughter was laughing and smiling, the employer said she simply wanted to see her daughter and make sure she was safe. Angie knew there were cameras in the home, but that was the first time her employer had openly indicated she was watching. The effect of that incident was to make Angie think carefully about how her actions appeared on camera to her employers before engaging in the sort of play that she normally would do otherwise.
The third response that surveillance cameras in the home elicit is more or less overt expressions of disaffection. In order to elucidate this ethnographically we draw here on the smartphone diary of a woman we call Rosa. The diary entries here are from over a three-day period, edited here for sake of brevity.

Rosa, Sunday, July 13 at 9:25PM
Day off but I am tired of listening to the mother shouting at her son, she asked me for help so I had to jump out of the bed because her son is not paying attention and not listening… [A]s long as the child is awake I have to be on guard as well. I feel ashamed to sleep. If it's my day off, I can try to ignore, but usually there are considerations I have to make.
It's like having a 'reserve' favour if I would also need it someday.

Sunday, July 13 at 9:35PM
My ward is at the foot of the ladder, he keeps calling auntie because his mom went out already. I got out of bed, went to the toilet. Staying at the CR [toilet], funny, they have their surveillance camera, I got mine as well [referring to the smartphone].

Sunday, July 13 at 10:04PM
Give a hand, then would want both. Abusive. "Auntie, excuse me help…" she whispered, her son didn't want to brush his teeth…Life of a helper, and it is still my day off. But anyway, the camera can monitor that I am still working late even during day off. I have my card.

Monday, July 14 at 9:49AM
Went to the toilet hoping there is no camera, but it is not only the camera following me 24/7 but my ward as well. He is always around me and in front of my room when I open the stock room. Sir is sleeping.

Monday, July 14 at 5:10PM
The surveillance camera is on the top of the cabinet, the whole room can be viewed through the monitor. The ipad provides a full view and it is in the master's bedroom. If I will sleep at the foot of my ward's bed, they can see me the whole night. There is always a camera shooting and I am the only character. =)

Monday, July 14 at 10:30PM
My ward keeps on calling me now, he wants me to sleep in his room, but I cannot sleep there with the camera on. My employer can see it from his cell phone and the mother can see it from her ipad in the room… Tuesday, July 15 at 7:19AM …One time the child woke up at 4am and he didn't slept until 7am, then when he slept until 10am, I slept with him. The father didn't sleep and waited for me to wake up. He said, 'it's 10am already…' then I responded, yes I know. And asked if he didn't notice in the camera that we are awake from 4-7am? He didn't respond.
There is so much rich material in Rosa's diary and we can only deal with some of that here.
Rosa was reportedly aware when she signed her contract of employment that there would be cameras in the home. However, she was not told and remained uncertain about: a) the number and location of cameras and of what they are able to record, b) of whether and where the images are being stored, and c) of the different ways that they are being accessed. Although the research study made her even more determined to locate all the cameras and 'blind spots', her evident anxiety about being constantly monitored and having her privacy invaded was exacerbated by the unreasonable and uncompensated demands on her time.
Rosa's diary, more than the others, self-consciously sets out to record forms of surveillance and control that she experiences from her employers. As Rosa indicates, it is not only the camera that is 'following [her] 24/7'. Rather, it is her ward and her employers who especially because of her male employer's shift work means that at times she has someone continuously in the house with and checking up on her. Her diary also discloses the many ways that her actions are guided by what can reasonably be called an ethics of care recorded frequently by the respondents in all the diaries. But it is also evident that ethics of care is constrained and challenged precisely by the system of controls that are ostensibly put in place to ensure care. This is clearly seen in Rosa's expressed dilemmas over whether she should sleep on the floor next to her ward's bed and be exposed to the gaze of the camera in his room.
Rosa also enjoyed the sense of turning the tables on her employers: 'they have their camera, I have mine'. Rosa confirmed in follow-up conversations that she does not view the camera in an entirely negative light. As our other respondents suggest, the images cameras produce may be subject to misinterpretation but they can also document their incessant labour and long hours. As Rosa retorts to her male employer when he wakes her up and asks her if she knows it's 10 o'clock, 'Yes I know' but had he not noticed on the recording that she had been up with his child from 4am -7am while they slept?

Challenging control with and through care
Similarly, Elena recounted her recent experiences of surveillance in the home by a new employer for whom she had only been working for an 8-month period. The employer had not disclosed the use of surveillance cameras when she was hired and she only discovered their use inadvertently, though at the time of research she was still not certain of the extent of their presence in the home, of where the images are being stored and of who has access to them. The first time she became aware of the camera was when her employer called to ask why she had left a window in her ward's bedroom open. It was at that point she realised that her employer must be watching her secretly, and she describes how that realisation made her shiver recalling how she must have been exposed previously changing in front of the camera in the room.
At the time of the interview, she had recently confronted her employer openly about the camera after he had repeatedly 'scolded' her for various things observed on the camera, including not ensuring his son had done his homework.
Elena (animatedly, loudly, remembering the confrontation): 'What, pardon?' I answered back like that…'You saw already on the camera, that the kid is not doing his homework.
How many times did you call him, talk to him, ask him why he didn't do his homework… You know what, no one can help your kid regarding his studies only you as his parent. I applied to work with you as a maid, not as a teacher'. I answered back like that.
'So you know what, this surveillance camera is for security reasons, but watching me from behind, you saw and you talk behind me. Every time you come home, you scold me, why?' At that point I am already crying to release my anger… Suddenly, because I am crying, I'm looking at him straight, head to foot. 'Once you are not comfortable with me, you are free to release me'. Or, I will do it for you. And besides you will never rebuild that [trust].
MJ: What did he say when you told him about the cameras?
Elena: He said, always he said to me 'for security reason', that is the reason. Security reason, but you are always watching behind me. I tackle him, 'You are happy you saw me naked and you never revealed it, you are happy. I thought you are a solicitor? I'm sorry if I'm very rude, but you push me. I want to respect you but you never respect me.

So [he said], 'I'm sorry'. [I said], 'Sorry is not enough, damage is already done…
There are a range of issues involved that are not just limited to significant violations of privacy but also the fundamental breach of trust between employer and employee that as Zelizer (2010) contends, and is also recognised by Office of the Privacy Commissioner for Personal Data (2015), is vital to intimate labour in domestic settings. Both are evident in Elena's account and central to her outrage about the covert forms of surveillance she has experienced. It is instructive that in challenging her employer's abusive behaviour, Elena also challenges her employer on the basis of care and, in particular, his perceived lack of care for his son evident in his unwillingness to call or talk to him about his homework. As she puts it, 'what's the use of your surveillance then?' The final story we recount here is from Susan. She had worked for her current employers for six years. Initially she came into conflict with her employers over their use of the camera to monitor her activity. As Susan recounts it, her male employer, like Elena's, justified the presence of cameras in terms of security and because they were new parents and wanted to be sure their baby was safe. When she started work, her ward was only one month old. Susan emphasises that she was not so concerned about the cameras watching her -she had previously worked for employers in Singapore who had used surveillance cameras in the home. Rather, what bothered her was the way that her employers and her female employer in particular, used the camera to constantly interfere in what she was doing. Her female employer repeatedly called whenever she was not visible on camera. Although she challenged them about the continual harassment that interrupted her other domestic work, Susan altered her routine to suit their wish for her to spend most of the time on the sofa in the living room with the baby. That appears to have been the only location where there was a camera and, from what Susan could surmise, the female employer who worked as a school teacher was continuously monitoring 'the video'. However, she found also that constantly being watched by her employer meant that all of her actions became the subject of interrogation. She describes her employer asking what she was eating at meal times and harbouring suspicions that she was taking food. The situation came to a head after a heated argument between Susan and her female employer through the intervention of her male employer's father, i.e. her ward's paternal grandfather. The latter appears to have been a regular visitor to the house and frequently brought Susan food that she liked to eat. Susan overheard him berating his son, her male employer, and telling him that he should tell his wife to stop scolding and bothering her since she, Susan, was a better mother than his wife. Subsequently, Susan reports, her female employer's attitude towards her changed, and she became both a gracious and generous employer, and the camera though still apparently there, no longer an issue.
Susan's story reveals some of the wider and more complex layers of surveillance within the home including both face-to-face and digitally mediated monitoring conducted by members of the extended family.Rosa, Susan and Elena all reported, for example, that they had heard of situations where grandparents had been given access to the surveillance cameras recordings; in one case, it was grandparents in Canada who had been asked by parents to monitor remotely the maid working in their home in Hong Kong. But they also affirmed that long before cameras, it was often grandparents or other relatives whose ostensible social visits to children were perceived by domestic workers to be monitoring on behalf of their employers. In a further twist to this story of care/control, it appears that the grandfather was not only monitoring the 'maid' but also his son's wife, and intervenes both to protect her and to ensure the care of his grandchild. While his intervention secures and enhances significantly Susan's position in the home, it also illustrates that the grandfather's intervention is shaped by wider kinship and gender regimes that often places women, whether as wives or as domestic workers, in a subordinate position in the home and family. Indeed, in another exceptional instance, one of our participants reported that her male employer's family had attempted to enlist her help in watching his wife who was suspected of having an affair.

Conclusion
Participants' smart phone dairies and the conversations about those diaries confirm that domestic workers are subject routinely to a range of different sorts of surveillance and monitoring in the home. The use of surveillance cameras may be seen as an extension of other sorts of monitoring but in doing so potentially alters and intensifies aspects of those relations. On the one hand it creates a situation in which increasingly the labour of care needs to be continuously visible and visualized in order to exist and be recognized. In other words, care becomes more than simply embodied and affective labour in the service of social reproduction (Borris and Parrenas, 2011:7) but rather, following Finch (2007) something that must also be seen and continuously displayed.
On the other hand, precisely because of the uncertainty about the number and location of cameras , as well as who might be watching and when, this form of watching and visibility is regarded generally by participants as more invasive and threatening to their bodily and social integrity than other forms of face to face surveillance. It enables employers not just to monitor at a distance but to interrupt, direct and interfere in the work that the women are doing in a continuous and unpredictable manner. This is gendered, as well as raced and classed, moreover, not just because the domestic workers are women and at times evidently subject to a voyeuristic male gaze but also because it re-enacts and extends the divide between the doing and display of care and the organization, direction, reception and recognition of that care.
Of course, it is not just domestic employees who experience surveillance at work. Nonetheless, the specific characteristics of domestic labour in this situation make the system of surveillance distinctive if not entirely exceptional. There are three interrelated aspects to this. The first has to do with the organisation of domestic and care work in terms of its temporality -people working as domestic employees in the situation described here feel themselves to be continuously on call and subject to round the clock monitoring. The second has to do with its spatiality -people working as domestic employees frequently do not have a secure or adequate private space to retreat to. The third is its primary focus on the body -people's bodies and their bodily performances are continuously subjected to their employers and other people's gaze.
The specific forms of monitoring and surveillance of people in the homes in which they labour is also made possible by the general systems of non-surveillance that governs their position as migrant domestic workers (Hintjens 2013, Johnson 2015. On the one hand, as we indicated above, prior to departure people who move to Hong Kong are subjected to a range of surveillance practices that ideologically at least blur the lines between care and control. On arrival they are compelled to 'live-in' and hence live under the forms of surveillance they encounter in their employers' homes and are subject to other forms of monitoring and controlby government and by employment agencies -that enforce and blur the distinctions between their status as resident foreigners and care and domestic workers. On the other hand, while migrant domestic workers are the focus of one sort of monitoring and control, they are also excluded from many of the legal protections and care that are at least nominally afforded other workers. That includes the systems of redress that are put in place to protect people's rights in the work place, including rights to privacy and data protection that govern the sorts of personal data about employees that can be gathered, stored and shared by employers. Migrant domestic employees, by contrast, are afforded very minimal protections from the sorts of digital surveillance practices disclosed by our participants. Significantly the growth of new media and mobile technologies has blurred the public-private divide and further revealed many layers of complexity in notions of care and control. It has allowed the disembodied and potentially objectifying gaze of surveillance to enter the 'private' or 'domestic' sphere through the proliferation of computers and smart phones. In this context, the voyeuristic male gaze, so often found in the context of public-space CCTV surveillance, has entered the domestic sphere where some of the male employers in our research used smart phones linked to cameras in the home to watch migrant workers in intimate settings. Our findings chime with emerging research on the 'smart home' which has revealed how 'tech savvy' (often male) users are able to monitor and control the activities of 'incidental users' (Zeng, Mare and Roesner, 2017). However, as our research has shown, female employers also use smart phones linked to cameras to monitor domestic workers. In some respects this monitoring reinforces the female employer's roles as primary carers, as they use the devices to monitor interactions between domestic workers and their children and to monitor the amount of food they suspect is consumed by their employees. In this context, surveillance far from being 'intimate, pastoral and productive' (Moore, 2011: p. 257), is disembodied, impersonal and based on a mistrust of those who have been employed to care for their children. The irony, of course, is that while employers may care for their children, their attempts to demonstrate this care through the introduction of the disembodied and objectifying gaze of surveillance cameras, leads to strategies of evasion and avoidance amongst domestic workers which in turn can undermine the 'care' received by the employers' children.
While the use of cameras is often legitimated in terms of preventing harm and ensuring care, monitoring may have precisely the opposite effects in so far as they undermine the trust necessary for care relationships, elicit strategies of evasion, and constrain rather than ensure or enable the attentiveness necessary for good care. First, domestic workers develop bodily and spatial tactics for avoiding the gaze of the camera that, as a consequence, may mean they are not as often physically present with those they are responsible for. Second, domestic workers report that they do not necessarily become more conscientious about what they are doing, but rather more self-conscious about what it looks like they are doing on camera. Third, their apprehension about the potential invasions of privacy, their anxiety about perceived misinterpretation of what is seen on camera, and resentment about the perceived lack of trust that the presence of camera conveys, means that they are likely to experience or at least express disaffection towards employers, if not the people that they care for. This is especially so in the case where cameras are being used covertly, or when they are persistently being questioned by employers about actions either observed on camera at a distance or recorded and reviewed subsequently.