‘Lockdown Work’: Domestic Workers’ Experiences During the Covid-19 Pandemic in South Africa

T his article explores domestic workers’ experiences of ‘lockdown work’, which refers to working conditions during the level 5 to level 3 lockdown period in South Africa during the Covid-19 pandemic. Drawing on in-depth interviews with female black African South African and African migrant domestic workers from Zimbabwe and Malawi, the article provides crucial insights into how the pandemic altered existing working conditions and employment relationships. We use the sociological concept ‘boundary work’ to illustrate the relational dynamic and consequence of social and physical distancing during the pandemic. We argue that social and physical distancing deepened the public-private divide in employers’ private households and domestic workers’ intimate workplaces. The findings show that domestic workers experienced limited or no control over decisions regarding Covid-19-related protocols in their workplace, intensified workloads without additional remuneration, and felt voiceless regarding working conditions because they feared losing their jobs. The experience of lockdown work highlighted domestic workers’ vulnerability because of the asymmetrical and intimate nature of domestic work under new management imperatives that positioned most domestic workers as a high-risk group or perceived carriers of Covid-19. We conclude that the experience of personalism/maternalism and distant hierarchy as forms of boundary work undermined domestic workers’ sense of dignity and employment rights.

findings show that domestic workers experienced limited or no control over decisions regarding Covid-19-related protocols in their workplace, intensified workloads without additional remuneration, and felt voiceless regarding working conditions because they feared losing their jobs. The experience of lockdown work highlighted domestic workers' vulnerability because of the asymmetrical and intimate nature of domestic work under new management imperatives that positioned most domestic workers as a high-risk group or perceived carriers of Covid-19. We conclude that the experience of personalism/maternalism and distant hierarchy as forms of boundary work undermined domestic workers' sense of dignity and employment rights.

Introduction
Domestic employment relationships are fraught with tensions because of the intimate nature of domestic work, which includes cooking, cleaning, and caring for others. Domestic work occurs in the power-laden context of employers' private domains. Yet, the private domain of the household becomes an intimate (public) workplace for both the employer and domestic worker, giving rise to tensions associated with the public-private boundary-making process typical of the employment relationship. Since domestic work remains undervalued, mostly informal, and poorly paid, it is often relegated to black women, highlighting the longstanding power dynamic associated with the raced, classed, and gendered nature of paid reproductive work entrenched in South African society (Ally, 2010;Du Toit, 2013a). While great strides have been made to improve the rights of domestic workers in South Africa, through an array of labour legislation, the sector remains afflicted by poor regulation and non-compliant employers (Mullagee, 2021;Patel, Mthembu and Graham, 2020). These issues are further compounded by the devastating effect of rising unemployment in the sector, with many facing job insecurity through ad hoc, part-time work, or gig work. Statistics prior to the pandemic showed that domestic workers' access to employment was negatively affected by the economic recession in South Africa (SweepSouth, 2022). However, like elsewhere, the pandemic exacerbated the employment crisis and heightened precarity when employers set in motion furloughs, flexible or reduced work hours, retrenchments, or dismissals, plunging households into distress. Aside from economic distress, the fear of contracting the virus and the perception of domestic workers as carriers of Covid-19 presented crucial challenges for domestic workers' working conditions and employment relationships (Chan and Piper, 2022).
This article explores how domestic workers experienced their working conditions and employment relationships during the lockdown period, when employers grappled with their own health needs in their private domains. We argue that social and physical distancing challenged and deepened the public-private divide in employers' private households and domestic workers' intimate workplaces. The power-laden and intimate nature of domestic work exacerbated domestic workers' vulnerability during the pandemic, giving rise to new management imperatives. The article begins by providing an overview of South Africa's domestic work sector that continues to position domestic workers as precarious and disposable, especially during the Covid-19 pandemic. Then, the theme of 'lockdown work,' which we propose to describe working conditions during the lockdown period, is framed within the context of an intimate workplace, highlighting the interplay of 'boundary work' (Lan, 2003) and 'practices of power' (Ally, 2010) that domestic workers experienced. This is followed by an overview of the methodology of the study. The final sections focus on the findings and discussion, emphasising how the pandemic 'locked' domestic workers into working conditions in the absence of protection in a legally transformed but poorly regulated sector.

'Lockdown work' during the pandemic
Despite South Africa's progressive labour regulation since 1994 to advance the employment rights of domestic workers, access to these rights has been slow and often hampered by poor regulation, a rise in outsourced cleaning and part-time work arrangements, and employers' indifference to labour legislation (Du Toit, 2013a;Du Toit, 2013b). In Du Toit's (2013a) edited collection, domestic work is aptly described in the book's title as perpetually 'exploited, undervalued, and essential.' The volume highlights the contradictory status of paid reproductive labour that maintains households in different ways for the benefit of employers but at the expense of domestic workers. More recently, there have been notable advances to recognise domestic workers' rights to safe working conditions and their struggle for minimum wage. For example, there has been a shift from Sectoral Determination 7 (SD7), which previously determined minimum wages for domestic workers, to the more recent National Minimum Wage (NMW), which now includes domestic workers (Department of Employment and Labour, 2023). In 2020, the Compensation for Occupational Injuries and Diseases (COIDA) Act 130 of 1993 was extended to domestic workers (SERI, 2021). These represent significant developments in the domestic work sector. However,

SPECIAL EDITION
T H E T H I N K E R | J o u r n a l I S S N : 2 0 7 5 2 4 5 8 domestic workers' opportunities for decent work are, overall, burdened by high unemployment, informality, underemployment, and downsizing and outsourcing as cost-saving strategies. These developments undermine efforts to turn domestic work into better jobs.
There has been an alarming decline in domestic work employment from almost one million in 2019 before the pandemic to 797,000 more recently (Stats SA, 2023). This coincides with the dire economic climate, the Covid-19 pandemic, and state-mandated lockdowns, under which several households cut back on domestic services because of economic and health-related distress (SweepSouth, 2022). Unsurprisingly, unemployment in the sector has had a devastating effect on mostly black African women in South Africa and from neighbouring countries who rely on domestic work for employment. In 2020, the South African state declared that under the strict, hard lockdown level 5, only essential workers could work. Others were subject to mobility restrictions that prioritised collecting social grants, seeking medical attention, and buying groceries (South African Government News Agency, 2020). Level 4 lockdown allowed essential and designated workers to return to work. For both lockdown levels, domestic workers were neither essential nor designated workers, meaning that from March to the end of May 2022, they could not work. Only at the start of level 3 lockdown, announced on 1 June 2020, could the majority of domestic workers return to work, but with health and safety protocols in place that applied to all citizens during levels 5 and 4 (Mullagee, 2021).
We use the phrase 'lockdown work' to describe the working conditions during the state-mandated lockdowns implemented in many countries to curb the spread of Covid-19. During the lockdown period, 'sporadic hyper-precarity' -'the kind of sporadic risks, uncertainty, vulnerabilities, and stigmatisation at times of crisis' (Chan and Piper 2022: 270) -exposed the plight of domestic workers globally. Many migrant domestic workers experienced homelessness and precarity when the lockdown was declared or when they were dismissed by employers after contracting the virus (Chan and Piper, 2022). In South Africa, similar cases were reported (Ndaba, 2021). Domestic workers did not know how their situations would unfold and many received no updates about their employment status from their employers (Zanazo, 2023). Domestic workers were either dismissed when level 5 lockdown was announced, received no pay or support from their employers while at home, or no social assistance such as the Unemployment Insurance Fund (UIF) or Temporary Employer/Employee Relief Scheme (TERS). Initially, domestic workers could only access TERS if their employers had registered with UIF. Aside from implementational challenges, there were issues regarding employers who accessed the funds on behalf of employees but kept the money for themselves (Skinner, Barrett, Alfers, and Rogan, 2021: 11). Further, lack of access for domestic workers to TERS was linked to the large number of non-compliant employers who had not registered their domestic worker for UIF, the system used to disburse TERS (Dawood and Seedat-Khan, 2022;Mullagee, 2021). The Casual Workers Advice Office, the Izwi Domestic Workers Alliance, and the Women on Farms Project lobbied for the TERS system to allow employees who were not registered with UIF to apply for themselves (Skinner et al., 2021: 12). Despite this significant victory, however, Skinner et al. (2021) reported low payment rates because of a lack of awareness.
There is now a growing consensus that the pandemic worsened long-existing inequalities, creating 'new conditions' for an 'ongoing crisis of social reproduction' (Acciari, del Carmen-Britez, and del Carmen Morales Perez, 2021: 15). Acciari et al. (2021: 12) argue that ' [q] uite strikingly, the sectors and people most essential

PEER REVIEW
This article explores how domestic workers experienced their working conditions and employment relationships during the lockdown period, when employers grappled with their own health needs in their private domains. We argue that social and physical distancing challenged and deepened the publicprivate divide in employers' private households and domestic workers' intimate workplaces.
to the reproduction of human life are also those most exposed and least protected during the pandemic crisis.' Similarly, Kabeer (2020) argues that the pandemic deepened 'the divide between the "haves" and "have-nots," revealing how those who "have" had homes to lock down in, security of income or savings to fall back on.' Evidence suggests that the pandemic laid bare the extent of vulnerability that domestic workers experienced during the health and economic crisis, including limited or no social protection, joblessness, mental health issues, and increasing inability to support their families (Acciari et al., 2021;Mullagee, 2021). Research in Hong Kong revealed that those who continued to work during the lockdown experienced an intensive workload that necessitated 'sanitized divide' as a new form of social distancing tactic between families and domestic workers (Chan and Piper, 2022: 272). Paradoxically, while domestic workers were at the front lines of sanitising homes with harsh chemicals to keep families safe, they were often perceived as a high-risk group or carriers of Covid-19 (Chan and Piper, 2022). These issues highlighted the undervalued and precarious status of domestic workers, despite their integral role in maintaining households. Further, the pandemic exposed how lockdown work was characterised by employers' cultivation and maintenance of sociospatial boundaries between their domestic workers and families. Next, we discuss the concept of boundary work in relation to the public-private divide in households as workplaces to explain our argument about 'lockdown work'.

'Boundary work': Social and physical distancing
In cultural sociology, 'boundary work' is a relevant concept for understanding relationality within the domestic employment relationship (Lamont and Molnár, 2002). Lan (2003) refers to boundary work as a tactic for social and physical distancing in the employment relationship. For employers, the aim of boundary work is to exercise control over poor performance or maintain a stable relationship. Domestic work scholarship is replete with descriptions of employers' practices of power as a managerial or boundary work strategy in employment relationships (Ally, 2010;Lan, 2003;Villiers and Taylor, 2019). Typical examples of inclusion-based boundary work include maternalism or strategic personalism. Barua, Waldrop, and Haukanes (2017: 482) define maternalism/strategic personalism as a form of boundary work in which employers control workers through a relationship of dependency and patronage.
In this form of a relationship, employers become the custodians of their domestic workers, while domestic workers become proteges of their employers. The use of maternalism or strategic personalism ensures that quality care and work performance are maintained in the household (Näre, 2011).
Alternatively, employers may adopt exclusionary or distancing tactics such as 'distant hierarchy' or business-type relationships to avoid the power dynamic inherent in domestic work (Lan, 2003). Business-type relationships are commonly associated with outsourced cleaning companies (Du Toit, 2013b). In relationships characterised by a distant hierarchy, what Barua et al (2017: 491) refer to as the markerbased approach, employers limit personal bonds with their domestic workers. Lan (2003: 531) argues that in these types of relationships, employers create boundaries that enhance their class or ethnic status while domestic workers are treated as insubordinate.
For domestic workers, boundary work occurs when they avoid certain work-related behaviour or demanding expectations of employers. In addition, while employers are known to use boundary work to encourage a good work ethic, studies have shown that domestic workers tap into the same 'ambiguities of intimacy' (Ally, 2010) to negotiate better working conditions informally (Näre, 2011). Domestic workers' efforts to reclaim control through boundary work tactics are necessitated because their right to fair working conditions is rarely acknowledged. Jinnah (2020) argues that often, everyday resistance includes remaining silent or invisible as a survival strategy. In other words, there is a constant negotiation between the employer and domestic worker in the employment relationship, despite the power imbalance that exists because of race, gender, and class inequalities.
Domestic workers' silence or invisibility underlies the social relations and power imbalance of the home as a place of work (Bonnin and Dawood, 2013;Fernandez and de Regt, 2014). Therefore, boundary work is not only about the nature of the employment relationship but also about how domestic workers are included or excluded in employers' private households. Here, we emphasise the meaning employers attach to PEER REVIEW their 'private' home and domestic workers' 'public' workplace because of the contradictory interplay associated with power differentials and the intimate nature of domestic work in the private household (Bonnin and Dawood, 2013). Further, by recognising and emphasising the household as a workplace, Marchetti (2022: 14) argues that a concerted effort is made to focus on interactions between domestic workers and employers in 'a specific location'.
Research shows that when a perceived form of transgression occurs between both parties, the act of silence or invisibility is likely to occur through physical and social distancing (Ally, 2010;Lan, 2003). We adopt Lan's (2003: 527) notion of boundary work and conceptualisation of 'socio-spatial boundaries' to explore the public-private division of the workplace in a particular context. However, in this article, we focus on domestic workers' experience of employers' boundary work tactics to understand the social and physical distancing practices that employers undertook during the state-mandated lockdowns.

Methodology
This article is based on qualitative research that was conducted in Cape Town, South Africa in 2022. Interviews were conducted with 12 domestic workers who worked for mostly white middle-class and uppermiddle-class employers. Data was collected using a semi-structured interview design, with follow-up interviews with select participants over WhatsApp to clarify further queries and/or elaborate on their experiences. When the lockdown measures were eased, interviews were conducted face-to-face but with safety protocols. Interviews were conducted in English and isiXhosa. While the sample appears small, the data collected offers crucial insights into the employment experiences of domestic workers during the pandemic. Given that most domestic workers were from townships and part of established social networks, it was not uncommon for participants to say that other domestic workers had similar experiences to them. Data were analysed thematically through a rigorous process of transcribing and translating, re-reading transcripts several times for emerging key themes, and engaging in three stages of coding including peer-review discussions before finalising the key, emergent themes (Campbell et al., 2021;De Wet and Erasmus, 2005: 300). Table 1 presents the demographic profile of domestic workers in terms of age, nationality, marital status, and number of dependents. The sample consisted of twelve domestic workers. Six were African-black South Africans, four African-black Zimbabweans, and two African-black Malawians. All reported one or more dependents. More than half of the participants were employed full-time, while the remainder had three or more employers with a part-time arrangement of one day per week. Those with part-time employment relied solely on their income as domestic workers. Two domestic workers with one employer reported that they were formally employed, and the rest reported informal employment.

Forms of compensation during lockdown levels 4 and 5
Levels 4 and 5 of the lockdowns were challenging times for domestic workers because they were uncertain about their employment status or if they would receive compensation for not working. Table 2 summarises each domestic worker's compensation during the lockdown levels. Overall, all domestic workers received some form of compensation such as money or food during lockdown levels 4 and 5 if they did not work. This finding suggests that most employers were sensitive to the economic situation of their domestic workers, irrespective of their status as compliant or non-compliant employers.   Zoleka's experience illustrates the uncertainty, precarity, opportunities, and losses during the lockdown. Her harsh response, particularly during times of crisis and despair, exposes domestic workers' vulnerability because of informality and employers' indifference to domestic workers' rights. Zoleka explains that her employers never gave her a clear explanation for her dismissal, apart from being told she was replaced by someone who lives closer to them. Their reason for the dismissal made little sense to her because she lived in Khayelitsha throughout her five years of employment with her employers. As Zoleka reasons, it is likely that she was dismissed because her employers saw her as high risk (travelling to work with public transport; attending a funeral during the peak of Covid-19). Instead of reporting the incident Vo l u m e 9 6 / 2 0 2 3 | J o u r n a l I S S N : 2 0 7 5 2 4 5 8

PEER REVIEW
for TERS. The third domestic worker, Nomthandazo, who was informally employed, was registered for UIF by one of her four employers. However, she did not claim TERS. While she did not provide any reason for not claiming TERS, it is likely she did not because all her employers paid full wages while she was at home. When the lockdown changed to level 4, Bukelwa and Tatenda returned to work. Bukelwa travelled to work and Tatenda's employment arrangement shifted from live-out to live-in. Therefore, in relation to changing employment arrangements during lockdown levels 5 and 4, two domestic workers experienced a shift from live-out to live-in employment. In both cases, domestic workers suggested that the change in employment arrangement was a pragmatic decision linked to their employers' safety concerns because they used public transport. When the lockdown was reduced to level 3, all the participants returned to work.
Two domestic workers reported instant dismissal because their employers feared contracting the virus. Sindiswa was dismissed before level 5 lockdown was declared, but later, during level 3, was called back to work, while Zoleka was dismissed after level 3. Before the pandemic, Zoleka, a 49-year-old widow and mother of three, worked as a part-time domestic worker for three different employers that were part of a family network. For one day per week, she earned a set daily rate of R350 or R400 per household. She had no employment contract but sourced regular Zoleka's experience illustrates the uncertainty, precarity, opportunities, and losses during the lockdown. Her harsh response, particularly during times of crisis and despair, exposes domestic workers' vulnerability because of informality and employers' indifference to domestic workers' rights.

Work intensification and remuneration
Most domestic workers in the sample returned to work after three months of lockdown. Upon their return, they reported increased workloads. Although domestic workers with multiple employers usually experience intense workloads (Du Toit, 2013b), those who worked for one employer also reported increased workloads upon returning to work. Given the months spent at home, they were not physically prepared for work, especially in households where employers had not cleaned regularly. Eight participants complained about their workloads. Of these eight, five worked for one employer. The remaining three participants worked for multiple employers. In other words, two participants of the seven participants with one employer did not complain about their workload, while of the five participants with multiple employers, two did not complain. They explained that the workload was not too intense because their employers cleaned for themselves during the lockdown. Mary, who worked for multiple employers, complained about 'catching up' when she returned to work. However, Mary and Sindiswa both recounted their surprise when one of their employers did not pay them for their first month of work after returning during level 3 lockdown. Without informing them, their employers chose not to pay them because they were paid during lockdown levels 5 and/or 4 when they did not work.
Sindiswa's experience throughout the lockdown was that of sporadic hyper-precarity. Before the pandemic, she worked full-time as a live-out domestic worker. However, her employer terminated her employment before level 5 lockdown was announced and she instead received a once-off payment in March, which was later deducted from her wages upon her return during level 3 lockdown. Sindiswa received no financial support thereafter because of her unemployed status. During level 3, her employer asked her to return to work. Like others, she experienced work intensification, but unlike the other domestic workers in the sample who felt powerless and concerned about their job security, she challenged her employer. She reported to the CCMA that she was not paid for the month of work during level 3 even though her employer threatened to reveal that she was an undocumented migrant worker. In the end, Sindiswa did not pursue the case. She explained how she had no money to travel to the CCMA offices and lost her job. Based on Sindiswa and Mary's experience, it is likely that other domestic workers were not compensated for working extra hours because their employers paid them when they could not work. While some employers appeared to show genuine concern for their domestic workers during the lockdown, others failed to communicate the conditions for compensation when they were not working.
Further, domestic workers reported that their employers increased their work hours without discussing changes with them increasing wages. Most of the domestic workers did not challenge their employer's decision because they felt that they had to endure precarious work conditions to avoid losing their jobs. Thandeka made this clear when she said, 'When it comes to , we just couldn't fight about it.' Eunice elaborated, 'It started during the time of Corona; a lot of things changed. Things were better before than now. Now if you lose your job, you might lose it for good and never get another one again.'

Distant hierarchy and personalism in the employment relationship
Domestic workers reported different examples of personalism in the employment relationship before the pandemic. The most common examples included receiving gifts and financial support, including loans, from employers, which they sometimes did not have to pay back. Domestic workers also reported that some of their employers provided emotional support to their personal problems and employers also shared their personal issues with them. Overall, domestic workers reported 'good' relationships with employers. Yet, more than half of the sample reported employment relationships characterised by a distant hierarchy before the pandemic. Two domestic workers complained that their employers shouted at them or accused them of theft. Five domestic workers felt that there was a shift from personalism before the lockdown to a distant hierarchy during the lockdown.
Although social and physical distancing was part of precautionary measures to curb the spread of

PEER REVIEW
Covid-19, for domestic workers, social distancing in their workplaces felt discriminatory, signalling a shift from feeling like 'part of the family' to 'outcasts.' They felt discriminated against because employers applied social distancing to them but not to family members, friends, or neighbours. In the excerpt below, Zoleka describes how one of her employers, who previously shared breakfast with her, distanced himself during the pandemic: He just made me feel like I was not a human being or even made me feel like I was the carrier of Covid-19 because he was working with people, he was working for charity organisations under the government. So everywhere he went, he worked with people but when he arrived at his house he would only want to be with his family and no one else. (Interview, 02 June 2022).
Zoleka describes a sudden change in her employer's behaviour towards her. Her employer would request that she should leave the room when he was present. Her movements in the workplace were restricted and differed significantly from the conditions of work before the pandemic. For example, her employer did not allow her to clean the lounge or touch the chairs. Zoleka was not given any explanation for her employer's behavioural change. As such, Zoleka felt as if she was a carrier of the 'disease' and 'not a human being.' Similarly, during the pandemic, Thandeka noticed that her employer and daughter would leave when she entered any room in the house, although she wore a mask. While it is possible that her employer and daughter did so because they were not wearing masks and/or to ensure social distancing, Thandeka shares another significant change that she felt was discriminatory: I remember there is a guy I work with. He is a gardener, and they bought cups only for me and that guy. The strange thing is that I am the one who washes dishes. So, am I not the one who is going to touch the cups? It is going to be touched by me because I am the one who is cleaning it (Interview, 10 June 2022).
Before the Covid-19 pandemic, Thandeka and the gardener could use any of the dishes but during the pandemic, they were given specif ic dishes to use. This sudden boundary work pinpoints the contradictory implications of intimacy, given that Thandeka was employed to clean and keep her employer's family safe. Therefore, as Thandeka astutely notes, the sudden change was nonsensical because cleaning requires touch. At the same time, this change signalled a distant hierarchy because discriminatory social distancing applied to her only and not to the employer's neighbours and family members. Besides buying separate cups, Thobeka's employer instructed her to eat in the laundry room, whereas in the past she ate at the dining table. These examples illustrate a distant hierarchy between domestic workers and employers and the socio-spatial boundaries related to separating the public and private spheres of the intimate workplace.
The Covid-19 pandemic and lockdown measures necessitated new management practices. Wearing masks, social distancing, and the use of sanitisers were precautionary measures for workplaces to curb the spread of the virus. However, domestic workers reported that Covid-19 protocols were partially or not followed. Often both the domestic worker and the employer did not wear masks, nor did they practice social distancing or use hand sanitisers. When protocols were partially followed, domestic workers reported being urged to wear a mask while the employer, family members, and visitors did not. The decision to wear or not wear masks was contingent on the employer, but domestic workers reported instances of asserting their right to safe working conditions. For example: The Covid-19 pandemic and lockdown measures necessitated new management practices. Wearing masks, social distancing, and the use of sanitisers were precautionary measures for workplaces to curb the spread of the virus. However, domestic workers reported that Covid-19 protocols were partially or not followed.

PEER REVIEW
You know what they said, they just said 'We trust you and we know you will not let us down and we want you to also trust us because we know you've got a small baby at home.' So, I was taking care of myself, I would sanitise and everything because…not because they were asking me to, but I had to do it mos (after all), everyone for himself (Chiedza, Interview, 04 July 2022).
Chiedza's experience differed from other domestic workers because her employer did not enforce protocols that made her feel like an outsider by, for example, making her wear a mask while the family and employer did not. Chiedza's employer rationalises the decision to not use masks because of mutual trust and the need for responsibility towards each other's families. However, Chiedza did not feel safe without a mask or sanitising. She explained that her choice to wear masks was her attempt to protect herself because her employer decided to not wear masks or social distance without consulting her. While Chiedza did not challenge her employer's decision to not wear masks, she continued to use sanitisers and to wear masks. One explanation for Chiedza's response is that although she experienced personalism in the employment relationship, the unequal power in the employment prevails. Chiedza also raises an important point about 'everyone for himself' -suggesting that there was no guarantee of sick pay or a job to return to if she was ill.
Only two employers changed their employment arrangement with domestic workers, shifting from live-out to live-in domestic work. Wages did not change when they moved in with their employers. In both instances, domestic workers accepted the request despite not being entirely happy with the new arrangement. Their employer's request was based on fear of contracting the virus because their domestic workers use public transport. Nomawethu moved in with her employer during level 5 lockdown because she did not have young dependents. Tatenda was allowed to move in with her children and did so despite disapproval from her husband: I was not comfortable but they were happy. You know, even the kids were like 'she is here, she is going to sleep here.' So, they were like happier but as for me, you know, as a domestic worker, like as a helper, you don't feel free like my house (Interview, 5 July 2022).
Tatenda reconciles her discomfort with moving in by noting that the employer and children were 'happier' with the arrangement. Tatenda describes the live-in employment arrangement as 'unfree' compared to a live-out arrangement under which she can retreat to her own space and family setting without restriction. Tatenda revealed that besides not wanting to lose her job, she became a live-in domestic worker despite her husband's response because she felt indebted to her employer, who often gave her gifts. Like Tatenda, Bukelwa describes how personalism in the employment relationship serves as a tool to control domestic workers. Rather than question her employer, Bukelwa remained silent. Her concern that making an inquiry would 'backfire' on her also relates to her fear of losing her job. However, she raises an important issue noted in many studies that illustrate the insubordinate role of domestic workers when employers decide what is in their best interest and act as their custodians (Ally, 2010). Bukelwa experienced the classed power asymmetry in her employment relationship when her employer saved her TERS money to purchase a washing machine without consulting her. Bukelwa felt bound to her PEER REVIEW employers, making yet another concession to keep her job with them when they revealed she would not get a wage increase for the next two years because they were facing financial challenges. The findings show how personalism/maternalism diminishes domestic workers' bargaining power. In these specific examples, domestic workers had no voice regarding their wages.

Discussion and conclusion
The Covid-19 pandemic and lockdown transformed the nature of domestic work in unsettling ways. First, domestic workplaces became fragile sites of sporadic hyper-precarity. The findings show how domestic workers grappled with employers' uncertainty and anxiety as everyone retreated to their homes, uncertain of what would happen during the lockdown period, especially regarding their employment status. The lockdown rules and health-related protocols called for social and physical distancing to curb the spread of the virus. Yet, social and physical distancing challenged and deepened the public-private divide in employers' private households and domestic workers' intimate workplaces. Domestic workers experienced social and physical distancing as exclusion. The findings show how socio-spatial boundaries adopted by employers were mostly perceived as employers' response to domestic workers as a high-risk group or carriers of Covid-19. In addition, domestic workers were rendered invisible because of the contextspecific nature of the pandemic. The example of the politics of food consumption -where one eats separately or with separate crockery -demonstrated othering and the redefinition of the public-private divide because of the enactment of socio-spatial boundaries. Unsurprisingly, some domestic workers felt restricted to certain spaces or back rooms instead of former public spaces such as dining rooms.
Second, while the need for practicing social and physical distancing was necessary, an underlying assumption of shared understanding between employers, families, and domestic workers often led to poor communication. The lack of communication suggests a distant hierarchy in the employment relationship, including among domestic workers who experienced personalism before and during the lockdown period. What is noteworthy is how boundary work occurs along a continuum where 'individual employers lean toward one or more approaches of boundary work in shifting contexts' (Lan, 2003: 530). Further, poor or no communication revealed the inherent power asymmetry in the employment relationship. The data showed that almost all employers did not communicate with their domestic workers about Covid-19 protocols. Adherence to Covid-19 protocols was partial, ignored, and based on employers' needs. According to de Villiers and Taylor (2019: 7), at times, domestic workers are unable to negotiate employment conditions because of the unequal power between domestic workers and employers. Therefore, the top-down approach of managing domestic workers in an unequal employment relationship marginalised domestic workers' employment rights despite their essential status for keeping families safe and sanitising homes.
Third, the threat of unemployment heightened domestic workers' anxiety during the pandemic. Surveys tracking middle-class responses to South Africa's economic climate revealed that the leading cause of domestic work job losses occurred because employers immigrated or could not afford domestic services as households grappled with job losses, reduced work hours and pay, and the impact of the war in Ukraine on the economy (SweepSouth, 2022). However, SweepSouth (2022) reports that two-thirds of their respondents (from Kenya and South Africa) thought that they were dismissed for invalid reasons.
Although the sample size of this study is small, it suggests, like other studies, that domestic workers were likely dismissed because of employers' fear of domestic workers as a high-risk group or perceived carriers of Covid-19 (Mullagee, 2021;Sumalatha et al., 2021).
In the context of underemployment and mass job losses, the findings show domestic workers were mostly silent. While some employers showed sensitivity to domestic workers' economic situation during lockdown levels 5 and 4, others unfairly deducted what they had paid them previously when they returned to work. Studies by Ally (2011) and Dawood and Seedat-Khan (2022: 7) discuss how the asymmetrical employment relationship and risk of unemployment underlie a 'silent paradox' that is deeply rooted in a culture of 'servitude and institutionalised fear' because of 'a lack of trust in government and low bargaining power with employers due to the shortage of employment opportunities.' The shift from live-out to live-in employment during levels 5 and 4, working during the hard lockdown, work intensification without remuneration, unfair deductions, and silence towards domestic workers' mental and physical health cemented the conditions for what Dawood and Seedat-Khan (2022) refer to as bonded labour. Therefore, most domestic workers felt they had no voice in a society burdened by high unemployment. Their silence, however, must be contextualised considering employers' partial or non-compliant status with labour legislation, which adds to domestic workers' job insecurity and lack of access to social assistance during the pandemic. Like the findings in this study, Patel et al. (2020) found that most domestic workers did not benefit from TERS during the Covid-19 pandemic because of poor monitoring and enforcement of domestic workers' employment rights. A further constraint was a lack of resources or awareness of support mechanisms such as the CCMA or TERS (Skinner et al., 2021).
In conclusion, lockdown work deepened the publicprivate divide in employers' private households and domestic workers' intimate workplaces, thus 'locking' workers into conditions of servitude. The experience of personalism/maternalism and distant hierarchy was entangled with the social and physical distancing under Covid-19. These socio-spatial boundaries undermined domestic workers' sense of dignity and employment rights. The pandemic not only revealed domestic workers' longstanding socio-economic vulnerability but also exposed the dire situation and implications of non-compliant employers during economic uncertainty. The findings suggest a need for proactive government intervention to ensure that labour legislation is enforced to protect domestic workers. Campaigning to foster a culture of compliance among employers and to educate domestic workers about their employment rights is one step towards ensuring that domestic workers have access to relevant information, representation, and support services. Using online and/or WhatsApp support groups can further the reach and accessibility of support services under such campaigns. While this study focused on workers' experiences of employer's boundary work, future research can examine how domestic workers offset employers' control during lockdown work and explore employers' perspectives regarding the complexities of the public-private divide they must navigate. Such studies will likely generate robust debate regarding the future of domestic work in the post-pandemic period.