International Migration and Stereotype Formation: Indonesian Migrants in Hong Kong

This article analyses the impact of international migration on the formation and maintenance of stereotypes about the ethnic and national identities of migrant workers. It demonstrates that the employment conditions of Indonesian domestic workers in Hong Kong encourage certain types of behaviour, which in turn contribute to the processes of stereotype formation and maintenance about them in the host society. Intergroup contact theory predicts that even superficial interaction should disconfirm existing stereotypes, but this article shows how the ‘context of reception’ in host societies prevents this from happening through its creation of sub-optimal conditions, such as unequal power relationships.

proceedings at the Labour Tribunal. The process brought to the fore evidence of how and why employers, employment agencies and government officials rationalize the behaviour of migrants at work.
Hong Kong is a popular destination in the Asia-Pacific region for Indonesian migrant workers, who mostly migrated for work because of economic reasons, such as lack of opportunities for gainful employment at home (Killias 2018). Hong Kong immigration rules do not allow permanent settlement of foreign domestic workers so this group of migrants typically arrives with the expectation that they will eventually return to their home country (Constable 2007(Constable , 2014Ladegaard 2017). This arrangement is unlikely to change in the future because the Hong Kong Administration continues to see the temporariness of migrant domestic labour as desirable. Foreign consulates are also known to help in the administration and control of migrant workers' employment in Hong Kong (Sim and Wee 2009;Palmer 2013). Notwithstanding, these temporary working sojourns, which sometimes last for years, have a significant impact on society in Hong Kong. It is well known that foreign domestic workers take over household chores as well as child and elderly care responsibilities, which in Hong Kong, as in many other places, are traditionally left to women, thus freeing up local women's time for gainful employment, which in turn contributes to household income (Chan 2005(Chan , 2006. However, there is less of an understanding about how the contact with migrant workers influences the formation and maintenance of stereotypes about their ethnic and national identity. Intergroup contact theory predicts that even minimal interaction will result in employers disconfirming stereotypes about the migrants' ethnic and national identity if it occurs under optimal conditions, including when their power relationship is equal (Pettigrew 1998). Conversely, contact in the workplace meets fewer of these conditions so is expected 'to be less successful' in doing so (Dixon and Rosenbaum 2004). Nonetheless, these interactions still have implications, be they perhaps for the maintenance of existing stereotypes or the formation of new ones. This study makes a contribution here by showing how an employer rationalizes his migrant worker's performance by attributing specific behaviours to her ethnicity and nationality instead of to other identity markers, such as to the status as a vulnerable migrant worker. It also highlights how these 'enforced identities' serve to legitimize otherwise socially unacceptable exploitation and abuse (Ladegaard 2013a, 56). Having said that, the purpose of this study is not to judge the stereotypes or the migrant worker's strategies for resisting the legitimized exploitation and abuse. Rather, it is to draw attention to how the employer and other members of the host society rationalize the migrants' behaviour by way of ethnic and national stereotypes.
I was alerted to this phenomenon while working for a non-government organization hat supported mostly Indonesian migrant workers to claim labour and employment rights that they alleged had not been granted by their employers in Hong Kong. In 2014, I supervised the determination, filing and coaching for over 400 claims to be tabled in voluntary conciliation meetings with employers at the Labour Department. This effort accounted for over 40% of the 912 claims Indonesian domestic workers filed that year, 1 but not all meetings resulted in a 'full and final' settlement so the Labour Department referred unsuccessful cases to the Labour Tribunal, which in that same year adjudicated 250 cases involving foreign domestic workers. 2 By contrast to the conciliation meetings at the Labour Department, the Labour Tribunal's hearings are mandatory, and the court does not break down the caseload statistics by nationality. I did not collect data on how many of our 'unsuccessful' cases ended up in the Labour Tribunal, so it is difficult to know precisely how many cases involved Indonesian migrant workers. But each week, my team of four case managers and I attended multiple hearings at the Labour Tribunal to gather any new case facts, but also to monitor that our clients were treated fairly. Upon returning to the office, we would record our observations in the case management system and submit formal complaints about any non-procedural or otherwise illegal handling in the hearings.
This article uses as a case study based on observations at the Labour Tribunal hearing of 1 client. The proceedings were not exceptional in the sense that ethnic and national stereotypes were used to rationalize the behaviour of Indonesian migrant workers, and I formally complained about all non-procedural treatment. I performed my role as practitioner responsibly, but 'practitioner research' is known to influence the quality of data (Cochran-Smith and Lytle 2009). Even though my purpose for analysing the data has shifted to explain how and why the developments I recorded impact the quality of migrant workers' labour rights in their countries of employment, I need to be aware of my 'research agenda' at the time, which shaped the data's analysis in reporting (Groundwater-Smith and Mockler 2007). My purpose when I collected the data used in this case study was driven by an organisational commitment to maximize the chances of the client to successfully claim employment entitlements. The rationale did not blind us to evidence that weakened those claims, and we often encountered such data in employers' defence statements, but awareness of other perspectives and interpretations of events helped us cater our advice to clients and the overall strategy used to convince the Labour Tribunal of our clients' cases' strengths. Recognizing this, my intention here is to revisit the data as a non-practitioner to identify factors that influenced the realization of migrant labour and employment rights.
This article demonstrates that the employment conditions of foreign domestic workers in Hong Kong encourage certain types of behaviour, which in turn contributes to the process of stereotype maintenance and formation about the migrants' ethnic and national identity. The first section outlines how and why three segments of a host society represented by an employer, employment agent and government official may come to hold thoughts and beliefs about migrant workers before going on to explain how these stereotypes can change over time. Using the example of Indonesian domestic workers in Hong Kong, the second section introduces a typical migrant to draw attention to their class standing at home, which in turn influences their performance at work in Hong Kong. The third section explains how the 'context of reception' in Hong Kong also shapes the migrants' behaviour to foreground a case study about how the aforementioned segments of society instead use ethnic and national stereotypes to rationalize migrants' behaviour. The following section considers how this practice contributes to the maintenance of existing stereotypes and formation of new ones before arguing that international migration has intensified the intergroup contact, increasing the opportunities for disconfirming negative stereotypes about ethnicity and nationality, but that the unequal power relationships, which exist between migrant workers and their employers, prevent it from happening on a much larger scale.

International Labour Migration and Stereotypes
International labour migration increased dramatically in the 20th century, exposing host societies more to migrant workers from different ethnic and national backgrounds. The development has also contributed to the formation and maintenance of stereotypes about members of the host society (in-groups) and the migrants (out-group). There are motivational reasons for people in stronger positions of power to redirect frustration towards and project hostility onto weaker groups of people (Hinton 2000). Furthermore, in the case of labour migration, in-groups in host societies have come to feel increasingly concerned as they are called on to share their privileges with an enlarged pool of residents, including migrant workers. But one of the other major reasons why people stereotype is because stereotyping serves an 'energy-saving' function that 'frees up cognitive resources' to think about other matters (Augoustinos et al. 2014, 18). A consequence of stereotyping is that it generally leads to 'social injustice' (Augoustinos et al. 2014, 234), defined as unfair and unjust relations between individuals and society. Intergroup contact should disconfirm stereotypes, for example, if the power relationships between them are equal (Pettigrew 1998). But power relations are unequal in the workplace (Dixon and Rosenbaum 2004), so intergroup contact between host society employers and foreign migrant workers often fails to do so. The following section considers the origin stereotypes that rationalize migrants' behaviour at work in a single host society.
In Hong Kong, there are three major in-groups that are a direct source of stereotypes about Indonesian domestic workers. First, employment agents are tasked with the role of promoting employment of migrant labour to local employers. Generally, they are known to market migrant workers as obedient (Lorente 2018), but they have a reputation for appealing to existing beliefs about ethnicity and nationality as a promotion strategy (Constable 2007). For example, employment agents that broker jobs for foreign domestic workers in Hong Kong emphasize differences in skill sets between nationality groups. They promote Filipinos to employers who prefer communication in English or who want their children to be exposed to the language (Yelland et al. 2013). By contrast, they promote Indonesians when employers want to use Cantonese, the local language (Constable 2007). This difference in skill sets is largely due to the fact that English is an official language in the Philippines, whereas Indonesians learn English as a foreign language at school, acquire it as an extracurricular activity and/ or learn it when their employment requires them to do so (Palmer 2018). Generally, then, Filipinos are expected to speak English, while Indonesians may only be able to do so in exceptional circumstances, for example, if they have previously worked for an English-speaking family. Employment agents also recommend Indonesians when employers express preference for workers who are 'flexible' and so are more likely to comply with changing demands of the job, such as work hours.
Second, employers directly experience Indonesians as migrants who live and work in their households. To begin with, employers might be pleased or disappointed with the migrants' performance based on promises made by employment agents. For instance, a newly arrived migrant worker might be eager to learn the local language and be flexible in response to the employer's demands about the terms and conditions of employment. But these migrants might also need more supervision and instruction about how to work in line with employers' expectations especially if it is the migrants' first job. Before coming to Hong Kong, 59% of Indonesians performed the role in another country (Asian Migrant Centre (AMC), Indonesian Migrant Workers Union (IMWU) and Hong Kong Coalition of Indonesian Migrant Workers Organization (KOTKIHO) 2007, 25) and the remainder reported to the Immigration Department that they had had at least 2 years of relevant work experience in Indonesia. As workers, the migrants' subsequent performance on the job directly influences their employment relationships with employers, who have the power to grant or deny legally protected labour and employment rights, such as the amount and timing of annual leave.
Third, government officials mostly encounter Indonesian domestic workers in relation to labour, immigration and criminal matters, which are often interrelated. In a very common example, employment relationships are known to end abruptly in response to events such as physical abuse or accusation of theft by employers (Tan 2016). At the very least, these employers owe the migrants outstanding wages, air tickets and food allowances for each day of travel to Indonesia. In the case that employers do not make these arrangements before their employment visa expires, the migrants can request the Immigration Department to grant them a visitor visa to remain in Hong Kong for the purpose of claiming the legal entitlements. Acceptable supporting documents include the registration of claims at the Labour Department, Labour Tribunal or Minor Employment Claims Adjudication Board. Visa extensions for participation in criminal proceedings in which migrants are either victims or perpetrators are granted if the migrants can provide evidence of future appointments with police or other law enforcement agencies. As a result, government officials frequently only encounter migrant workers with problems that need to be resolved.
These in-groups may form stereotypes about the migrants based on their individual and collective experiences. But the stereotypes need not be completely new. There may also be modifications of existing stereotypes that intergroup contact failed to disconfirm because one or more of the optimal conditions for it were not met. The Hong Kong press has played an influential role reinforcing the view that the migrants are somehow responsible for their own victimization and that the guilt of in-group perpetrators can be understood and even excused (Ladegaard 2013b). Existing stereotypes such as these continue to influence how in-groups rationalize the behaviour of out-groups (Goff et al. 2008), not least because they have both explanatory and justificatory power that influence attitudes and behaviour of in-groups towards out-groups (McGarty 2002). Use of stereotypes also has an attendant naturalizing effect that ultimately serves to hide other explanations of behaviour from view (Amelina and Faist 2012). These consequences are salient to understand the outcome of seemingly unfair adjudication of intergroup disputes. Before introducing a case study of the ways in which such stereotypes featured in a Hong Kong Labour Tribunal hearing that was tasked with adjudicating a labour dispute between a local employer and Indonesian migrant worker, the following sections discuss two relevant factors that help explain the migrant's behaviour at work.

Factor 1: Indonesian Domestic Workers' Socio-Economic Status
In 2017, 140,000 Indonesian domestic workers constituted roughly 0.02% of the 7 million large population in Hong Kong. According to Indonesian government data, between 2011 and 2016, 87% of Hong Kong bound migrant workers departed from Java, with 73% coming from the Javanese homeland in the central and eastern parts of the island. This area of Indonesia is densely populated but offers insufficient opportunities for gainful employment. Government programmes attempt to address the problem of 'labour surplus' by facilitating internal migration to other parts of the country where there is a corresponding 'labour shortage' or international migration to countries where there is a job offer (Palmer 2016). Domestic work in Hong Kong is one such opportunity, which attracts migrants with the following socio-economic status.
Most Indonesian domestic workers are female young adults. The Hong Kong Immigration Department does not specify the age or sex of migrants, so the demography is more likely the result of supply and demand factors, such as migrant and employer preferences, which for one reason or another see domestic work as a job more appropriate for young women. In 2007, a survey found that more than 98% of Indonesian domestic workers were between the age of 20 and 39, with three quarters in their twenties (Asian Migrant Centre (AMC), Indonesian Migrant Workers Union (IMWU) and Hong Kong Coalition of Indonesian Migrant Workers Organization (KOTKIHO) 2007, 22). According to Indonesian law, the minimum age is 21, but employment agents in Indonesia are known to obtain 'real but fake' documents in which younger recruits meet the requirement (Palmer 2012;Ford and Lyons 2011). Just over half were married, which is a bit lower than the average marital status of Indonesians who migrated for work in the 6 years between 2011 and 2016 (BNP2TKI 2016). Nonetheless, 96% reported that they supported at least one dependent at home with their remittances. These remittances amounted to around USD 680 million in 2016 alone. As this contribution indicates, these young women play an important role in financially supporting husbands, children and parents, which in turn raises the migrant workers' social-economic status at home (Chan 2018). However, it does not result in a corresponding rise in Hong Kong, where foreign domestic workers occupy one of the lowest-if not the lowest-social position in the places where they live and work.
More than half of the migrants have at least a junior high school education. A third obtained a senior high school qualification or higher, which means that Indonesian domestic workers in Hong Kong are generally more highly educated than the average Indonesian who migrates under the government's overseas labour migration programme (Badan Nasional Penempatan dan Perlindungan Tenaga Kerja Indonesia (BNP2TKI) 2017). As all foreign domestic workers are required to supply evidence that they have had at least 2 years of relevant work experience as part of a successful application for an employment visa, Indonesian migrants also tend have that experience as explained above. But while some migrants may report work experience prior to leaving Indonesia, it is also common for employment agents to fabricate such details in the full range of migrants' application documents, even including their real names and dates of birth (Silvey 2007). Of course, employment agents do so as part of their business model to collect more fees, but there is also evidence that documents are doctored in order to meet the requirements about what migration should take place, rather than reflecting the reality of who actually wants to migrate (Palmer 2012). At the very least, this practice means that government data has a tendency to overestimate the actual education level of Indonesian migrants.
The demographic of Indonesian domestic workers has changed over time, shifting in response to factors outside their control. For example, successive changes to education policy in Indonesia mean that a larger proportion of the population has access to government-subsidized schooling. In 1985, President Suharto made 6 years of basic education free and mandatory, and in 2003, President Megawati expanded the coverage to 9 years. Of course, there have been implementation issues, which mean that schooling is not always entirely free and that some segments of the population do not have access, such as people living in distant and remote areas of the country so cannot attend government schools. Nonetheless, these interventions have decreased pressure on most families to privately finance primary and secondary schooling of their children, thereby removing a significant factor that motivates mothers especially to seek gainful employment abroad. The changes also mean that the average education of migrant workers tends to be higher. A domestic worker born in 1984, for example, was not even supported to attend primary school, whereas a migrant born in 2000 is expected to at least have a junior high school qualification.

Factor 2: Hong Kong's Context of Reception
Hong Kong's experience of Indonesian domestic workers is shaped in large part by the 'context of reception', which has been used to identify the various ways and degree to which migrants integrate into societies in destination countries (Portes and Böröcz 1989). In the specific case of foreign domestic workers in Hong Kong, this context includes law and policy that determines the migrants' legal status, residency rights and social mobility (Sim 2003). The following section outlines the impact that this context has on the ability of foreign domestic workers to enjoy legally protected labour and employment rights.
A primary intention of Hong Kong's migrant labour importation scheme is to provide households with access to affordable, full-time domestic workers. The scheme has mostly helped mothers become economically active and so contribute to household income (Chan 2005(Chan , 2006. By contrast, local domestic workers are not always affordable if hired on a full-time basis. In 2017, the going rate was HKD 8000 (USD 1025) per month, whereas a foreign domestic worker could be hired for around half that amount on the condition that employers also provide the migrants free of charge food and 'suitable accommodation with reasonable privacy' in their homes. Households with a combined income of at least HKD 15,000 (USD 1920) were eligible to sponsor a visa for employment of a foreign domestic worker. Of course, there are households that earn much more and so may pay wages at rates above the minimum allowable wage of HKD 4410 (USD 565) per month. Underpayment of wages is the norm for most first time migrants from Indonesia (Ladegaard 2017), and a 2017 analysis of around 3000 job advertisements by an employment agency indicates that on average, employers pay between 0.1 and 21% more, with lowest wages in Hong Kong's north and higher wages to the west (HelperChoice 2017).
Hong Kong's migrant labour regime also shapes the context of reception for migrants. The regime is understood to be a 'specific mode of macro-level labour control' that enables greater exploitation of migrant workers by their employers for the purpose of delivering economic growth (Bal 2016, 7). Immigration policy is recognized to enable that kind of control, which has the effect of curtailing the ability of migrants to claim legal rights (Ford 2019). In Hong Kong, all foreign domestic workers are entitled to the same labour rights protections as other workers under the Employment Ordinance. The law regulates the payment of wages, rest days, various forms of paid leave from work, protection from being fired for a reason other than serious misconduct and conditions for ending employment. The Standard Employment Contract, which is mandatory for all foreign domestic workers, is another reference for labour standards that are not stipulated in the law. For example, it includes the wage, which is legally below the statutory minimum wage, arguably because the migrants receive the compensatory benefit of free 'room and board' at the expense of employers. These legal standards are intended to protect the migrants' labour rights, but they are not always enforced, which in turn undermines the purpose.
The live-in arrangement is a cornerstone of Hong Kong's migrant labour regime, which enables willing employers to subject foreign domestic workers to 'abuse, exploitation and forced labour' (Justice Centre 2016, 64). In 2003 the Director of Immigration required all new foreign domestic workers to live with their employers, which he justified by referring to anecdotal evidence that by staying out, the migrants directly competed with locals for limited housing and transport when commuting to work. Reportedly too, the policy change curbed illegal employment of foreign domestic workers, which had put downward pressure on casual rates for part-time domestic work earmarked for locals. In theory, the Hong Kong Administration could charge foreign domestic workers and their employers who flout the rule with the offence of false representation. In practice, however, the rule is effectively enforced through the Standard Employment Contract, which requires migrants to 'work and reside' in the employer's home, which in turn encourages foreign domestic workers to compromise on labour rights as part of a strategy to avoid having unfriendly relationships with their bosses outside work hours. The following case study outlines how an employer rationalized the behaviour of an Indonesian domestic worker to justify his decision to summarily dismiss her.

A Glimpse into How Stereotypes Form
In June 2014 Charles Leung attended a hearing at the Labour Tribunal in which Siti Maesoroh, his former domestic worker from Indonesia, claimed that he had terminated their employment contract without giving her 1 month's notice as required by law. 3 As compensation, Siti asked for the equivalent of 1 month's wages in lieu of that notice. A week before the hearing, Charles submitted a written defence statement to the Labour Tribunal and Siti, arguing that he had given Siti several verbal warnings about serious misconduct so was well within his right to terminate their employment contract on the spot when he found out that she had yet again done something wrong. Ultimately, the Labour Tribunal ruled that Charles ought to have given Siti the legally-required notice largely because he could not provide evidence of the warnings to justify summary dismissal. Charles was ordered to compensate her, but in the hearing, made the following arguments to justify his actions.
Charles claimed that Siti often disobeyed his instructions. On the first day, he expressly forbade her to use a personal mobile phone during work hours (8.00 am until 9.00 pm), but later frequently saw Siti breaking the rule by reading and sending messages when she thought no one was watching. To monitor Siti's behaviour more closely, Charles installed a closed circuit television system in the apartment without her knowledge, which is not illegal but certainly unethical not least because of how it contributes to the domestic workers' humiliation and denigration in Hong Kong. The surveillance allowed him to see what was happening at home anytime through his mobile phone. On one day, Charles calculated that Siti spent two hours and 21 minutes chatting or talking on her mobile phone. He confronted her with the evidence, and she apologized promising not to use her mobile phone during work hours again. But the following week Charles' mother, who also lived in the apartment, overheard Siti talking on the mobile phone in the toilet. Charles was frustrated as the employment agent had assured him that his expectations were reasonable so he asked his sister-in-law for advice. She had previously employed an Indonesian domestic worker and so might know what what had happened. The sister-in-law told him that Indonesians often pretend to follow rules they disagree with rather than directly communicate their disagreement because they are non-confrontational.
With time Charles also grew frustrated that Siti was neglectful of her duties, which primarily included cleaning the apartment. As an example, Charles described how Siti would vacuum the carpet in the living room but hesitated to neaten items on the display cabinet, wall shelves and coffee table. Charles raised this issue with Siti, who then promised to clean the living room more thoroughly in the future. But a few weeks later Charles noticed that the neatened surfaces had become sticky because Siti had only dusted around the display items. This time Charles gave Siti a verbal warning that she did not meet his minimum expectations about how to clean. He called the employment agency to report the incident and asked for a replacement domestic worker. In response, the agent attempted to placate Charles by explaining that it was common for Indonesian domestic workers to receive inadequate skills training before coming to Hong Kong. As a result, Indonesians were sometimes unprepared to live up to Hong Kong employers' high standards for how to perform at work, so the agent offered to educate Siti free of charge in his own facilities on Saturday, Siti's weekly rest day. Indonesians, the agent also argued, have a strong reputation for lacking initiative and so need to be shown everything that needs to be done.
Charles accused Siti of misconduct and argued that she deliberately endangered the safety of members of the household by not maintaining basic hygiene. Very early on in the employment relationship Charles witnessed Siti use cold water to wash used crockery and cutlery. Charles was alarmed because he thought it was common knowledge that hot water kills disease-causing bacteria found on cooking and eating utensils. He promptly forgave Siti for the wrongdoing after she showed him that the dishwashing gloves were too big. The hot water had scalded her hands, but she did not know Charles would buy a smaller pair if she asked. That day Charles bought her a new pair of gloves so that she would use hot water to wash dishes. But soon after he also noticed that Siti did not put soiled towels and other laundry directly in the washing machine. Rather, Siti would hang the laundry out to dry before washing it, which smelled as bacteria proliferated. In the Labour Tribunal hearing, the judge told Charles that she shared his concern, and then asked Siti through an interpreter why she did not maintain good domestic hygiene by allowing germs to flourish. The government official then told Siti that Indonesian domestic workers are known to purposefully behave poorly so that employers end the employment.
Finally, Charles found out that Siti had been dishonest about her reason to take annual leave. Prior to Siti's arrival, the employment agency told Charles that he should give Siti the equivalent of seven days wages as a final pay-out for annual leave at the end of their two-year employment contract. Allegedly, Siti agreed to the arrangement shortly after starting work. But a few months later, and not too long after Siti began attending the free skills training at the employment agency on her only rest day, Siti requested annual leave for the purpose of visiting her friend in hospital. Charles agreed, but then saw Siti shopping with friends in Causeway Bay, a popular hang-out for Indonesian domestic workers on Hong Kong Island, when she should have been seeing her friend at Queen Elizabeth Hospital in Kowloon. He called the hospital and found that Siti's friend had never been an in-patient, and when he confronted Siti about the lie, at first she claimed that her friend was registered under a different name. Eventually, Siti admitted that she had indeed lied about the reason for taking annual leave, explaining that she did so because she thought Charles would refuse her request if he knew the real purpose was to celebrate the birth of her friend's first grandchild. It was then that Charles summarily dismissed Siti because he deemed her dishonest behaviour to be very serious misconduct.

Stereotyping Indonesian Domestic Workers
That day Siti lost both her job and home in Hong Kong as Charles told her to immediately pack her things and then leave the apartment. Both parties can terminate the employment relationship at any time, but employers do not lose their place to live as a result. By contrast, the migrant workers do. Such outcomes are not uncommon especially when employment relationships end on bad terms (Tan 2016). The live-in arrangement certainly exacerbated the already unequal power relationship (Justice Centre 2016), which existed between Charles and Siti, as it tends to favour employers over employees at work, insiders over outsiders in family life at home, and those with higher socio-economic status (Cheng 1996). Viewed from this vantage point, the case study about Charles' perception of how and why Siti performed poorly at work also becomes in part a story about a migrant worker's everyday resistance to labour exploitation, such as Siti's deception about her real reason for annual leave when Charles had been requiring her to undergo unpaid skills training on rest days for the purpose of learning how to meet his expectations at work. Siti's behaviour then is recast as something that is not particularly Indonesian, as other ethnic and national groups in similar situations are expected to behave in a comparable fashion.
The above case study of Charles' perception of Siti's behaviour reveals at least three instances where their work related interactions resonated with existing stereotypes. First, Charles' sister-in-law rationalized the fact that Siti flouted his rule banning mobile phones during work hours using the stereotype that Indonesians are nonconfrontational so will pretend to follow rules as part of a strategy to avoid direct conflict. She may have formed this stereotype through her former experience employing an Indonesian domestic worker, but it also reflects an existing stereotype about Indonesians generally, whereby Indonesians are expected to submit or compromise in an interpersonal conflict rather than seek resolution through confrontational means (Haar and Krahé 1999). Lying is understood to be another effective strategy for avoiding interpersonal conflict (Grover 2005). Siti may have lied about compliance because it benefited her to do so, but it is also a possibility that she thought the ban was unreasonable because using the mobile phone at work did not prevent her from getting the job done. Either way, Siti chose to lie instead of challenge Charles about the rule, for whom, the behaviour confirmed and ultimately served to maintain the existing stereotype that 'bad' foreign domestic workers, like Siti, were neither subservient nor did they 'obey their masters without question' (Ladegaard 2013a, 49).
Charles' interaction with the employment agent who recommended Siti for a job with him had a similar effect. The agent appealed to a widely-accepted belief that Indonesians do not assess and initiate work independently so as to dissuade him from demanding a replacement, which would impose an extra cost on business. In this way, the agent attempted to rationalize Siti's inaction as the product of her need for direction about how to complete the many tasks that constitute sundry responsibilities like cleaning, cleaning, cooking and doing the laundry. Indeed, cultural models for excellent leadership styles in Indonesia have an 'authoritative and paternalistic' dimension (Selvarajah et al. 2017, 4). In the case of Charles, the model explains Siti's lacklustre performance at work as a result of his failure to provide the expected guidance and pastoral care necessary for her to meet his expectations. But as admitted by the employment agent, there are other practical reasons that help to explain why Indonesian domestic workers fall short of Hong Kong employers' expectations. Before leaving Indonesia, migrant workers frequently receive inadequate skills training partly because they do not undergo training at all, or because the home appliances used for training are much older or completely different models to the ones they ultimately end up using (Rother 2017). As such, the inadequacy serves to maintain the stereotype that Indonesians need to be micro-managed in order to work effectively.
In court, the Labour Tribunal judge's comment that Indonesians have a reputation for not directly communicating their wishes draws attention to another stereotype. Put another way, this government official announced that Indonesians are known to prefer using indirect means of communication to show rather than say what they want. In the case of Charles and Siti's labour dispute, the official attempted to rationalize the migrant's poor performance at work as a particularly Indonesian way of asking the employer to terminate their employment. It is unclear if this observation is based on the government official's direct experience adjudicating labour disputes involving Indonesians or on the anecdotes of her colleagues. Having said that, the practice does have a cultural reference in Indonesia, where direct communication is deprioritized in favour of 'sensing the feelings of others' to not just avoid conflict but also get what is wanted without asking (Mulder 2005, 102). However, it is more likely that another non-cultural factor was at play when Siti displeased Charles with what he deemed to be her practice of poor domestic hygiene, remembering that economic factors motivate most Indonesian domestic workers to migrate, and that they pay relatively high fees to recruiters in return for the opportunity to do so. Conversely but logically, Siti would be expected to retain the employment even when Charles subjects her to exploitation and abuse.
These different in-groups' rationalization of Indonesian migrant workers' behaviour draw on existing stereotypes, some of which were formed and maintained in Hong Kong. This fact begs the question: Where, then, do these stereotypes come from? To a certain extent, the Hong Kong press, as explained above, is responsible, especially for the continuing 'us' and 'them' portrayal (Ladegaard 2013a). This treatment serves to legitimize abuse and exploitation of out-groups by in-groups, and resonates with the popular belief that the migrants 'have no right to their own space' in Hong Kong (Ladegaard 2018). Despite this, and perhaps further exacerbating the in-groups' misgivings, these migrant workers have legally used limited public space in Hong Kong for a range of purposes, including public protest (Constable 2009(Constable , 2010. In part, this explains the all too frequent exclusion that is practised in Hong Kong society, such as in lifts and pool areas that are designated as migrant worker-free zones in priorititization of the the right of in-groups to use the limited facilities. An historical cultural experience also helps explain the stereotyping, which relates more to the migrants' job category rather than their ethnicity or nationality. Few amah or paid Chinese women domestic workers were employed by the early 1990s, 'yet powerful ideas about them persist', including that amah were more 'hard-working' and 'devoted' than their foreign replacements (Constable 1996, 448). This combination of factors complicated, but did not ultimately thwart Siti's assisted attempt to clim her legal employment rights.

Conclusion
Increased temporary labour migration can provide society with greater opportunities to disconfirm stereotypes. In practice, however, outcomes have sometimes been quite the opposite as Hong Kong's experience with foreign domestic workers shows. The unequal power relationship that exists between employers and migrant workers is one of the more significant factors that undermines the capacity of their contact to do so. Intergroup contact under these conditions can instead maintain stereotypes by vindicating existing thoughts and beliefs that employers and society more generally hold about migrants' ethnic and national identity. Such outcomes may have negative implications for the migrants' overall well-being in the host society especially if the stereotype is somehow negative, which often predicts worrying behaviour that is 'aggressive, untrustworthy, or unintelligent' (Stephan et al. 2005, 3). Negative stereotypes such as these are known to lead to prejudice and other negative attitudes towards outgroups such as the migrants discussed here. To this end, this article draws attention to how stereotypes can negatively affect employment relationships, for example, when employers come to misunderstand the cause and motivation for their migrant employee's behaviour at work. Such counter-productive outcomes are assumingly an unintended consequence of migrant labour import programmes, which are arguably designed to benefit society.