Impossible children: illegality and excluded belonging among children of migrants in Sabah, East Malaysia

ABSTRACT This article makes a case for attending to the specificities of child illegality in migrant contexts. This is not simply because children have been left out of previous accounts, but also because their status as minors makes both their citizenship and their illegality different to that of adults. The analysis is based on research with children born to migrants in the state of Sabah, East Malaysia. I argue that such children are configured as Sabah’s impossible children, and that this configuration influences their experiences of illegality and exclusion in distinctive ways. From a young age, children are aware of document ‘checking’ raids and, as ‘foreigners’, are unable to attend Malaysian schools. However, informal documents from learning centres, as well as age and contingent circumstances, may give them a temporary, ‘liminal’ legality. Finally, given that irregular migrants experience both exclusion and inclusion in a host nation, the article describes children’s urban forms of belonging. These forms of inclusion demonstrate children’s engagement with Sabah as a home, as against their political construction as an impossible problem.

I am at an alternative learning centre for the children of Filipino migrants, talking to a teenage class about the positive and negative aspects of Sabah, the Malaysian state where they live and where most of them were born. Looking at the worksheet of Herman, a 12-year-old Suluk boy, whose ambition (ironically) is to become a policeman, I see that he has answered the question, 'What is bad about Sabah?' as follows: The first thing, there are lots of beggars in Sabah. Secondly, there are lots of robbers and thieves in Sabah. And thirdly, there are lots of police who arrest people without an IC [identity card].
On another occasion, at a learning centre for the children of Indonesian migrants, I distribute a worksheet encouraging children to compare family life in Indonesia and Malaysia. In response, one child anonymously replies: It is safer in Indonesia than in Malaysia, because in Malaysia there is lots of checking. Malaysia is a modern country. People without a passport will be arrested. Malaysia is a country which is more developed than Indonesia.
On another day, I am sitting underneath a makeshift building in a squatter settlement mostly populated by Filipino migrants. Nasira, a 13-year-old girl, shyly hands me a completed questionnaire. At the end, she has written: I don't have a pass, and I don't have any money to buy a pass. I have money, but just for food … . If you don't have a pass you are scared of the police and scared of going out. I am unhappy when I am just at home. I am bored. I want to go out, but although I like moving around I must be careful, as I don't have a pass.
What does it mean for children, born to migrant workers in a host country, to be treated as if they themselves are migrants? How does illegality influence such children's lives, in a context where the 'checking' of documents is a frequent occurrence? And why does a 'modern country' contain a substantial population of children too scared to leave their homes? This article describes the difficult situation of the children of migrants living in Kota Kinabalu, the capital of the East Malaysian state of Sabah. Such children are mostly undocumented and were mostly born in Sabah to parents who migrated when they themselves were young. As children, they are denied access to Malaysian government-sponsored education and healthcare, and are at risk of detention and deportation to a country they may never have visited before. Yet, as much as they are excluded from Malaysian society and its institutions, they are also informally included: as residents, as speakers of Sabahan Malay, and as present and future workers.
This article makes a case for attending to the specificities of child illegality, not simply because children have been left out of previous accounts, but also because their status as minors makes both their citizenship and their illegality rather different to that of adults. After outlining some of the distinctive issues raised by child illegality, I describe the particular situation of children of migrants in Malaysia, a country both dependent on and antagonistic towards migrant labour. I argue that such children are configured, through immigration regulations and public discourse, as impossible children. Such a distinctive configuration shapes children's particular experiences of illegality and exclusion that, as I go on to describe, are both similar to but also different from the experiences of adults. However, as my fieldwork encounters made clear, these children's lives cannot only be characterised in terms of fear and exclusion, but are also full of dreams, hopes and connections with family, friends and religion. Therefore, in the final part of the article, I describe forms of child belonging, as a way to emphasise such 'impossible' children's engagements with Sabah as a home.

Migrant illegality and children
A key focus of recent ethnographic and theoretical work on migration and its consequences has been the notion of migrant 'illegality'. This refers to the multiple processes by which migrants and their families come to be produced and labelled as 'illegal', and the ways in which such a juridical status shapes individual lives. Illegality means that migrants can be physically present but legally absent in a society (Coutin 2000), and analyses have explored the consequences of this contradiction, demonstrating its impact on embodied experiences of being-in-the-world (Willen 2007, 10). Powerful experiential accounts have helped to draw attention to the multiple exclusions of unauthorised migrants, and their social and physical suffering, whether in the shape of ongoing fear of deportation (Coutin 2000, 33), constant 'embodied tension and anxiety' (Willen 2007, 16) or health deterioration and even death (Khosravi 2010, 106).
Despite such common experiences of exclusion, migrant illegalities do not exist in ahistorical form, but are 'constituted and regimented by the law' in specific times and places (De Genova 2002, 424). A fascinating study of such a process is Mae Ngai's Impossible Subjects (2004), a history of the racialised generation and emergence of 'illegal' immigrants as a problematic category in the United States. Ngai argues that national origins quotas and other restrictions on immigration introduced in the mid-twentieth century produced the 'illegal alien' as a new 'legal and political subject' whose 'inclusion within the nation was simultaneously a social reality and a legal impossibility ' (2004, 4). Her account shows how Filipino, Chinese, Japanese and Mexican immigrants and their children were constructed as 'impossible subjects' because they were cast as 'permanently foreign and unassimilable to the nation' (8). In defining such 'aliens' as 'impossible subjects', Ngai means to draw attention to two senses of impossibility. Firstly, such a subject is 'a person who cannot be' (a person whose identity is defined by the very illegality of their existence) and, secondly, they are 'a problem that cannot be solved' (4). As I will show, this notion of the double impossibility of certain people is highly applicable to the situation of children of migrants in Sabah.
Work on the socio-historical construction of migrant illegality shows how changes in immigration laws generate 'the conditions of possibility for a dramatic expansion' of such illegalities (De Genova 2015, 38, my emphasis). Thus, Suárez-Navaz (2004) argues that in Spain, illegality is increasingly produced by new restrictions that make it impossible for poorer immigrants to regularise their status. More widely, Andersson's work (2014) demonstrates how clandestine migration in Europe has been produced in more (and increasingly distressing) forms by the very 'illegality industry' that is meant to control it. In addition to such work on the construction and expansion of illegality, research has explored how unauthorised migrants are included in a nation, even whilst they are rendered 'illegal' by its laws. Ngai sees undocumented immigrants as 'at once welcome and unwelcome' in the U.S.: excluded as citizens, but 'woven into the economic fabric of the nation' as cheap and disposable labour (2004,2). Mezzadra and Neilson have put forward the concept of 'differential inclusion' as a way to account for the elusiveness of distinctions between inclusion and exclusion, insiders and outsiders, in the European migration regime (2012,(67)(68). They compare this concept with McNevin's suggestion that irregular migrants might be described as 'immanent outsiders' who are 'economically and socially integrated' into places dependent on their labour (2006,141). Such theorisations of inclusion are an important corrective to the nation-state's language of illegality and criminality, and can help to draw attention to the moral claims to membership of irregular migrants and their families.
The literature on migrant illegality tends to assume that the subject under discussion is an adult, or focuses explicitly on adult migrant experiences. There has been far less attention to how children and young people are affected by shifting configurations of 'illegal' subjects, or by being born across borders. In part, this reflects wider problems with migration research, much of which has treated children, not as agents and actors, but as 'baggage, to be brought or left behind' (Orellana 2015, 5). However, whilst there is a growing body of work exploring children's involvement in and experience of processes of migration (Parreñas 2005;Gardner 2012;Coe 2014), research on the consequences of unauthorised status for children, or particular configurations of 'illegal' children, is underdeveloped. I will briefly describe some of that research here. But first I want to show why theorists of migrant illegality should be interested in children.
A central argument of the literature on migrant illegality is that the figure of the 'illegal migrant' is constructed, politically, in opposition to the figure of the citizen (De Genova 2002, 422). However, if the citizen under discussion is a child, then that opposition starts to look less clear-cut. Child citizenship is distinctively different to adult citizenship since it is often negated by children's status as dependent minors. For example, even in contexts where birthright citizenship remains possible, children cannot (unlike adult citizens) petition to keep non-citizen family members in their home country (Boehm 2012, 135). Studies have shown that, because children's status is frequently considered to derive from that of their parents, citizen children are also 'not meaningfully protected from deportation like their adult counterparts' (Bhabha 2003, 54). That is, even if they are citizens, children may be 'constructively deportable' (Bhabha 2014, 80) with their undocumented caregivers, on whom their well-being depends. This means that for children, the notion that migrant 'deportability' is a necessary corollary of migrant illegality (De Genova 2002, 438-439) is in fact rather more complex. In particular, if, as has been argued, children are better considered 'semi-citizens' (Meloni et al. 2014, 305), we cannot assume that child illegality takes the same shape as adult illegality, or that children's unauthorised status is produced in the same way as that of adults. Therefore, an examination of child illegality promises to shed important analytical light on our wider understanding of migrant illegality as both phenomenological experience and juridical status.
Existing ethnographic research on configurations of childhood illegality largely concentrates on two issues. Firstly, work has explored the implications of unauthorised migrant pregnancy in a range of contexts, including Israel (Willen 2005), Ireland (Luibhéid 2013), and Hong Kong (Constable 2014). Secondly, a growing body of literature examines the experiences of undocumented '1.5 generation' youth in the United States (Solis 2003;Gonzales and Chávez 2012;Dreby 2015), where children's unauthorised status is said to have negative psychological and developmental implications (Suárez-Orozco et al. 2011). In a powerful article on 1.5 and 2nd generation undocumented Latino adolescents in the U.S., Gonzales and Chávez (2012) have described how such young people 'awaken to a nightmare' as they move into adulthood. During their childhood, they argue, these young people are protected from illegality by their parents, their age and, most crucially, their ability to attend both elementary and high school. By contrast, in their mid-teens, undocumented youth discover that activities they now desiresuch as driving a car, or applying to collegerequire state-issued identification that they lack.
However, despite discourses against 'anchor babies' (Chavez 2015, 98-100), birthright citizenship remains an important right in the U.S. We know much less about children's own experiences of an 'illegal' status outside of this context. As global migration becomes more restricted and securitised, and as routes to permanent residency are closed, children are increasingly beingin Constable's (2014) evocative phrase -'born out of place'. When such births occur in countries without birthright citizenship, children take on the status of a 'migrant' even in the absence of migration. Born and living in the country where their parents are working, rather than 'left behind' at 'home', they may experience multiple forms of exclusion, even whilst they are tacitly accepted as the next generation of low-paid workers, a reproduced, foreign 'proletariat outside the polity' (Ngai 2004, 13). Considering the situation of such children is both ethnographically and analytically important in contexts where temporary labour migration turns out to be rather less temporary and rather more 'sticky' than either governments or migrants intended. In particular, children born 'out of place' force us to consider what difference it makes to the experience of 'migrant illegality', and of 'differential inclusion', if one is not, in fact, a migrant?

Sabah's impossible children
This article draws on a year of fieldwork (2012-2013) in Sabah's capital city of Kota Kinabalu (henceforth KK), explicitly focused on the children of migrants and refugees. Access to such potentially vulnerable children was initially secured through teachers working at 'alternative learning' centres in the city. In all, I paid 182 visits to such centres: 101 visits to four mostly 'Indonesian' centres, and 81 visits to five mostly 'Filipino' centres. My role at these centres varied, depending on the centre set-up and needs. At some, I worked as a volunteer English teacher, whilst at others I was a visitor and observer. I presented myself to children as someone who was keen to hear their stories about their lives, and to find fun and interesting ways to tell them, but also as someone who was happy to teach them English. Indeed, there were some children I taught English to who took little part in my research. Although I did visit a number of children and their parents at home, ethical considerationsin particular, the need for my presence not to draw undue attention to undocumented childrenmeant that I mostly confined my research to the safety of learning centres. In addition to participant observation, I utilised drawing, map-making, spider diagrams, open-ended worksheets and other methods that have proven to be particularly 'child-friendly' (Punch 2002). Seventy children participated in a photo project, 46 children wrote holiday diaries, and I conducted 29 semi-structured interviews with teenagers I came to know well. I asked for children's verbal consent to the future sharing of writing, stories and images, made it clear that they were free not to take part in research activities, and respected their decisions to destroy photographs, to remain silent about particular issues, or their requests for me not to use interview material. I approached informed consent as an ongoing and evolving issue, and stopped visiting some children when their situation changed and I sensed that they were no longer comfortable with my interest in their life. To protect the identity of my research interlocutors, all names of adults and children are pseudonyms.
Through these visits, I collected information on 338 Filipino-and Indonesian-origin young people, aged between 8 and 18. Whilst my young informants were an extremely heterogeneous group, they all spoke Malay, the language in which fieldwork was conducted. Some of these children's parents or grandparents are from the geographically and culturally close islands of Tawi-Tawi and the Sulu Archipelago in the southern, predominantly Muslim, Philippines. Many such Filipinos first entered Sabah as refugees during the Mindanao insurgency of the 1970s and 1980s, in order to escape fighting between secessionist groups and the Philippines government (Hugo 2004, 43;Kassim 2009, 58). Most other migrants originate from rural eastern Indonesia, including Muslim Bugis and Protestant 'Tator' (Tana Toraja) people from Sulawesi, and Catholic migrants from the small islands of the east Flores region, collectively known as 'Timor'. Before giving more detail on the context, it is important to emphasise that the families of these children were of mixed and sometimes-shifting legal status. Some children had their own passports, with or without a valid visa. Several Filipino-origin children also had IMM13 cards, originally issued to Filipino refugees. However, most of the children I knew were not only 'unauthorized' but also 'undocumented' in the fullest sense of the term, lacking even a simple paper acknowledgement of their birth.
Malaysia has, since its formation in 1963, been a particularly important 'node in regional and global human mobility' (Nah 2007, 39). In the late twentieth century, the country embarked on a series of modernisation projects, and experienced strong economic growth. This led to millions of new jobs, and the necessity to recruit foreign labour, first in rural plantations, and later in construction and domestic service (Wong 2006, 213, 216). However, the scale of this dependence on migrant labour varies across the country. Sabah, an East Malaysian state, has experienced migration on a particularly large scale, and is home to at least 44% of Malaysia's 'foreign workers' (Kassim and Hamid 2004, 81). The state has a complex and contested position within the Malaysian nation, and negative reactions to, and 'demonization' of, migrants have been particularly pronounced (Majid Cooke and Mulia 2012, 152).
From a loosely regulated position in the 1970s, Malaysian laws and regulations concerning the employment of foreign workers have gradually been tightened, leading to an explosion of migrant illegality in the country. Punishments for the unauthorised have also been made more severe, including large fines, imprisonment, deportation, and up to six strokes of the cane (Wong 2006, 217-218, 222;Nah 2007, 40;Hedman 2008, 262-263). In 2002, the government reformed the Migration Act, resulting in a zero-tolerance policy towards undocumented migrants (UN Human Rights Council 2009, 15). This was immediately followed by several, large-scale 'crackdowns' on undocumented workers, often as part of specific, named 'operations'. One such operation in Sabah led to a humanitarian crisis and several deaths on Nunukan island, an Indonesian transit point unable to cope with the sudden influx of almost 40,000 returnees (Ford 2006, 238-240;Nah 2007, 45). Such crackdowns are part of a distinctively Malaysian constructionand performanceof illegality, in which mass deportations are followed by industry-specific labour shortages and recruitment-drives (Nah 2012, 489).
According to Malaysian immigration regulations, 'foreign workers' from countries such as Indonesia must be in the age group of 18-45 years, cannot change jobs or employer without prior permission, and must only work, for a temporary period, in a restricted number of employment sectors. 1 As in many contexts, the state desires migrants as 'detached individuals' (Boehm 2012, 139), and immigration regulations assume that migrant labour is 'fluid', entering temporarily and moving out once conditions have changed (Hugo 2004, 65). Unskilled or semi-skilled foreign workers who hold a temporary workers' pass are (in clear contrast to 'expatriates') not allowed to marry whilst they are in Malaysia, nor are they allowed to bring their families into the country with them. 2 Immigration regulations make no provision for family reunification (Gurowitz 2000, 867), and women applying for a Domestic Helper visa must produce medical evidence that they are not pregnant (Nah 2012, 494). In addition, Malaysian citizenship laws, like those of many states, do not adhere to the principle of jus soli, or citizenship by birth. Thus, a Sabah-born child of migrants is herself considered an immigrant who belongs elsewhere, and may inherit the vagaries of her parents' legal position. 3 In a very real sense, the presence of the children of migrants is 'illegal' since migrants are not supposed to have families in Malaysia. These children are treated as a kind of 'category mistake' (Malkki 2010, 63): the Sabah-born and Sabah-embedded children of adults who are meant to be temporary. Indeed, in everyday discourse, newspaper articles and research reports, these children of migrants are frequently described as 'migrant children', a description ignoring the fact that the vast majority have never in fact migrated. Drawing on Ngai's (2004) approach to the impossibility of certain national subjects, I argue that children of migrants in Sabah are rendered 'impossible' by, on the one hand, immigration and other laws that deny their existence, and, on the other, by a public and political discourse that portrays them as an unsolvable problem. However, these are not generically impossible subjects but specifically impossible children. We see this both in the denial of their status as vulnerable dependents, and in their exclusion from common childhood experiences, most notably education.
Why are these children so impossible as children in Sabah? More specifically, why are they seen as an unsolvable problem? I will briefly focus on two concerns of public and political debate. Firstly, the particular politics of migration and demographic change in East Malaysia. Sabah, as a state on Borneo, has a complicated political relationship with the federal government based on peninsular Malaysia. Sabah (and Sarawak) joined Malaysia in 1963, later than peninsular states, on the basis of a '20-point agreement' designed to ensure autonomy, particularly over immigration and religion (Chin 1999, 21). However, many Sabahans consider this autonomy to have been undermined, and relations with the federal government soured, by what is known as 'Project IC'. This refers to allegations that Muslim immigrants have been irregularly granted citizenship in strongly Christian Sabah as a way to bolster electoral support for the ruling coalition. 4 'Project IC' is interpreted by many Sabahans as a politically motivated scheme to alter the state's demographic balance in order 'to "Malayize" or homogenize Malaysia' (cf. Chin 1999, 26;Sadiq 2005, 105). Certainly, numbers of Muslims in the state are increasing, while non-Muslimsincluding Kadazandusuns and other 'native' groupsare declining (Sadiq 2005, 108). Inevitably, such demographic fears are likely to focus on children, particularly given widespread perceptions that 'foreign' families have higher birth rates than 'locals'. Therefore, despite the undocumented status of many migrant families, it is the fact that some migrants have become possible political subjects that makes the current or future political recognition of the children of migrants impossible. 5 A second explanation for the impossibility of recognising the children of migrants as actually existing children with rights in the present lies in Malaysia's very particular politics of race and citizenship. Much work on migration regimes has explored the racialized dimensions to citizenship, in which, as in Ngai's study, the creation of 'alien citizens'those perpetually doomed to second-class statusis the other side of the coin to the legal production of illegality (De Genova 2002, 433). However, in Malaysia the picture is much more complicated: more recent immigrants find themselves at the bottom of a complex hierarchy of racialised belonging, in which those of Malay ethnicity are at the top, granted special rights as bumiputera or 'sons of the soil' (Lee 2004, 126-127). Citizenship does not, in and of itself, guarantee equal rights, and whereas Malays are considered 'organically Malaysian', those of Chinese and Indian descent find their status tainted by the historical immigrations of their communities (Gabriel 2014(Gabriel , 1215. In Sabah, these racialised dynamics of differentiated citizenship are further complicated by the existence of non-Malay and non-Muslim 'natives' who feel they are 'second-class bumiputera' in comparison with Muslim bumiputera (Chin 1999, 26). Such ongoing struggles over 'ethnic balance' and insider/outsider status do not translate into an atmosphere of generosity towards more recent migrants (Gurowitz 2000, 875). Indeed, Nah has suggested that Malaysians uphold complex, race-and class-based 'hierarchies of deservedness' that relegate the rights claims of irregular migrants to the bottom of the pile. 6 Certainly, when the issue of the rights of the children of migrants is brought up in Sabah, it is often countered by an assertion of the priority of the rights of rural, 'native' children in Sabah's underdeveloped interior, who may also lack documents or access to schooling (see Sadiq 2005, 113-115).

Illegality and children's lives in KK
Having analysed how childhood illegality is configured in Malaysian immigration regulations, let me turn to how it is experienced. When I first began fieldwork, mindful of the need to let children emphasise their own concerns, and to avoid potentially sensitive and upsetting issues, I was very hesitant to bring up the topics of legal status, immigration raids or detention. However, children mentioned such issues regularly, and often bluntly, attaching no moral stigma to being undocumented (described as 'not having a pas'). To my surprise, they also joked about illegality, and talked obsessively and humorously about the police. Once, after I had visited a learning centre for Suluk and Bajau children, four girls asked me to drop them off at the prayer hall inside their settlement. Their friend Amal came to say goodbye and, as he shut my car door, told me, 'take them to the police station!' Similarly, when I visited a brick factory, and talked with working teenagers gathered in a tiny shack, I heard one of them say, 'this house has become a police station'. These jokes about, and morbid fascination with, the police are extremely common but go hand-in-hand with explicit criticism of police actions. Malaysia has a system of compulsory ICs, which must be shown on demand to police or immigration officials. Non-citizens without an IC must carry other identification, such as a passport with valid visa. In KK, spot-checks on identity documents are regularly conducted on public transport and in places frequented by migrants. Squatter areas and workers' housing are also targets for such 'checking', often in named 'Ops' coordinated by the immigration department with the involvement of RELA, the notoriously violent People's Volunteer Corps (Hedman 2008, 371). The children I knew in KK had, without exception, all experienced such 'checking', and frequently mentioned being 'scared of the police'. During an interview, Ramlah, a Bugis teenager, recounted childhood memories of almost continual 'checking': They usually came without warning. They would knock on every house and they would ram the door open if the tenants were too slow opening it … . They would ask for documents, and sometimes they would go into the rooms. If people didn't have documents they would arrest them and send them to the Police Station. There were also times during the checking that we were asked to go to the field and were asked to gather and sit in the tent they set up.
Emanuel, a 16-year-old boy with Adonaran parents, spent his childhood living in cramped accommodation next to the chicken farm where his father works. He told me that he and his siblings had experienced document-checking 'plenty of times': We ran and hid ourselves in the chicken coops. It was really horrible. We didn't have any passport at that time. There was one time when I was 9 years old. We had our dinner at seven o'clock and then at eight o'clock we ran and hid ourselves.
Children frequently told me that one of their favourite activities was to jalan-jalan, or to move about, whether to visit a relative's house or take a trip to the market. Yet, children who are undocumented often experience their life as one of immobility, confinement and boredom. They are scared to go out, and spend large amounts of time at home, looking after younger siblings and watching endless television. Nasira, a Suluk teenager said that because she lacked a pas, she was 'scared of going out', but longed to do so, and felt unhappy and bored at home. Another boy told me that being undocumented was like 'living in a cage', while a nine-year-old Suluk girl wrote in a diary that, during a break from informal school: I wanted to go out but we had no money. I grew tired of it. I stayed at home and watched cartoons every day.
As this girl's diary entry shows, undocumented status intersects with poverty in confining children in segregated migrant housing. This aspect of children's lives resonates with the restricted mobility described by ethnographers as a key, subjective aspect of illegality in other contexts (Coutin 2000, 33;Khosravi 2010, 98).
Children are concerned not only with their own immobility and boredom, but also with the safety of adult family members, including siblings. Emanuel confided that the money his family earns was 'a waste', because his undocumented mother was 'not free'. Most children's fathers, and several of their mothers, had experienced arrest. Children are usually very open about describing these experiences, and are often critical of the circumstances they involved. Jony, who is half-Suluk, half-Visayan, recounted: There was this one time when my mother got arrested. That time she did not have a passport but my father had one. She had to hide. My father asked the police to arrest him but the police refused to do so because he had a passport. They threatened my mother that they would take us children to the police station if she didn't come out. She had to come out. I was still very young back then. I just cried. Jony said that his mother used to have an IMM13 card, allowing her to live and work in Sabah, but that she lost this after she was 'scammed'. Looking after his younger siblings, and trading food on a casual basis, she has still not been able to 'legalise' her status, and Jony worries about her and the possibility that she might be deported. Moreover, his own and his mother's experiences make him aware of the vagaries of documents and police power: The police here are bad. My mother was arrested once and they were asking for a bribe. They asked for 100 ringgit. But they lied to her, because after she paid, she was still arrested.
Another Suluk boy, 10-year-old Jeri, confided that he is worried about his mother, who has respiratory problems, but is reluctant to seek medical help, since she is undocumented, and doctors in government hospitals are under contractual obligation to report suspected 'illegals' to the authorities.
What is striking is that many children explicitly contrast the humiliations of 'checking' and immobility, as negative aspects of life in Malaysia, with Sabah's advantages as a place where their parents can find work. Thus, although children perceive Sabah as a place chock-full of police, they are aware that it is also a place their families have chosen for survival. An understanding of children's experiences of illegality must therefore involve an appreciation of the ways in which migrant families understand illegality to be a necessary evil, a kind of gamble in which work and money outweigh loss of freedom. Recent work on Indonesian migration to Malaysia has emphasised how migrants may choose to be 'illegal' in order to maximise profit and freedom, and avoid state-sanctioned processes of 'legal' migration that resemble conditions of bonded labour (Killias 2010, 909). In Sabah, many undocumented children learn to cope with the consequences of 'checking' in order that their parents can work, and the family can survive.

Contingencies of childhood illegality
Children's everyday experiences of being unauthorised in KK share many similarities with those of adults. Like adults, children find that their mobility is restricted, that they become paranoid about police and other uniformed officials, and that they often need to run from their home in order to escape 'checking' raids. Children also share with adults a sense of resignation to the reality that the risk of detention and deportation is the price that must be paid for accessing Malaysia's economic development. However, in this section, I will draw out some subtle but illuminating differences between child and adult experiences. For example, the awareness of the vulnerability of a parent who has been 'scammed' out of their identity document, or who is detained in a 'checking' raid, has particular emotional consequences for children, who witness the humiliation of a caregiver, and feel powerless to help.
Accounts of childhood illegality in the U.S. have described how non-citizen children who came to the U.S. when very young feel American due to shared childhood experiences. Such feelings can be compared with older teenagers' 'dawning realization' that their family is different (Suárez-Orozco et al. 2011, 452). In Sabah, children of migrants (whether born in Sabah or not) do not gradually 'awaken' to the living nightmare of illegality as they get older, since regular 'crackdowns' on undocumented migrants mean children learn, from a very young age, the significance of 'not having a pas'. In the U.S., unauthorised children may feel American in large part because, since the 1982 Supreme Court ruling in Plyer v. Doe, they have been entitled to public education until the twelfth grade (453). Likewise, in most European states, unauthorised children have the same legal right/obligation to attend compulsory education as citizen and documented children (Bhabha 2009, 439-440). The situation for children in Malaysia is very different. Before 1997, all children were entitled to attend primary schools, regardless of nationality, provided they held a birth certificate. But from that year onwards, 'foreigner' has been written on birth certificates of children of non-Malaysian parentage, and since 2002, schools have systematically excluded 'foreign' and undocumented children (UN Human Rights Council 2009). Such exclusion makes it hard for the children of migrants to see themselves as similar to their citizen peers, since they tend not to share any socialising spaces or activities. Indeed, most unauthorised children I knew directly connected 'not having a pas' with exclusion from school, and this aspect of their experience is unique to their status as children.
Nevertheless, the argument, from the literature on the U.S., that children are 'shielded' from illegality, does in some respects apply to younger children in KK. Yanti told me how she was caught up in document-checking on a bus but 'the police just passed right by us and didn't ask anything, probably because I was so young'. A number of children said that, in similar situations, they 'just kept quiet' and thereby escaped attention. Such tactics are only available to children because of their status as minors. Some children told me that, if the police ever stop them, they plan to be 'smart', and to say that their IC is 'still being processed' or 'not yet out'. Other older children employ their knowledge that Malaysian citizens do not actually get an IC ('MyKad') until the age of 12, saying that if they are stopped by the police, they will say they are 11. Fourteen-year-old Joel was once caught in a 'checking' operation in Inanam, an area of KK with a high number of migrants, but said: They let me go because I am a kid. I told them I was only 12 and have not got my IC.
Earlier, I suggested that because child citizenship is of a different nature to adult citizenship, the opposition of citizen and 'illegal immigrant' might not be so clear-cut in the case of children. In Malaysia, although children under 12 are entitled to 'MyKid' identification cards, there is no requirement to carry legal proof of identity until the age of 12. This suggests that younger children, even if undocumented, may have a kind of 'liminal legality' (Menjívar 2006) due to their age and the ambiguous nature of child citizenship. However, any such 'legality' is context-dependent: ironically, children may be safer from arrest if caught up in document-checking when alone. When travelling with a parent, their statusas for children in many contexts (Boehm 2012, 130) is assumed to derive from that of their parents, and therefore if their parent is detained, they will be too. For example, when Jony's mother was arrested during a checking operation, she was taken away with his younger sister. The unstable legality of younger children is also shown by the informal identity documents given to students by learning centres. At one centre for Indonesian-origin children, students must wear green, laminated cards round their necks at all times, with the child's name and photograph, and text asking 'concerned authorities' to contact the learning centre or Indonesian consulate. Significantly, learning centres provided many of the children I knew with the only place they regularly went to outside of their home, and such informal documents offer them some protection from 'checking'.
In many ways, and as many of my young informants confirmed, the most vulnerable group of undocumented children is that aged between 12 and 18. This is particularly the case for male teenagers (since girls' mobility is more likely to be curtailed), who are frequently targeted during 'checking' operations in malls. Such teenagers can be said to be triply vulnerable. In the first place, once they are 12 their lack of citizenship becomes clearer to the authorities, since they have not been issued with a Malaysian IC. Secondly, although they often work, they cannot get an official foreign workers' visa, since these are only issued to those 18 or over. Thirdly, the liminal legality afforded by basic learning centres becomes less available for older children. In the very few learning centres sponsored by the Malaysian state, the upper age limit is 12. Thus, it is the teenage children of migrants who are most likely to be constituted by the immigration system as 'migrants' rather than minors, and who face the threat of deportation to the country to which they are assumed to 'belong'.
For many of my teenage informants, illegality, as they got older, led their parents to attempt to acquire documents for them through various means. 14 year-old Yanti told me that her current passport uses her aunt's name as her guardian, since her parents were undocumented at the time it was issued. Other children told me they were in the process of becoming an 'adopted child' (anak angkat) so that their parents could try to get them identity documents. However, the tactical acquisition of borrowed and forged documents can be fraught with problems. Whilst adults tend to make their own decisions about pursuing particular kinds of documents, or choosing to remain undocumented, children tend to have such decisions and choices made for them. One teenager told me that her parents had paid a large sum to get her a Malaysian IC, but this had led to many stressful family arguments about money, and had opened up a source of tension with her siblings, who remained undocumented. Adult tactics to obtain legality can have negative as well as positive consequences, and teenagers' experience of 'solutions' to illegality can be complexly clouded by their lack of agency in pursuing document acquisition.
Children's forms of (excluded) belonging I have shown how child illegality leads to boredom, immobility, and tensions over documents, but how it is also so taken-for-granted that it can sometimes be a focus for laughter. I have also described how children experience illegality in different ways to adults, and how these experiences are influenced by age and circumstance. As the 'impossible children' constructed in discourse and regulations, the undocumented children of migrants should be completely excluded from life in Sabah. And yet, shared humour, trips to malls and the liminal legality that allows a young child to travel on a bus demonstrate certain forms of tenuous and temporary inclusion. Mezzadra and Neilson have argued that irregular migrants occupy an ambiguous subject position that is 'neither fully included nor fully excluded ' (2012, 62). But what forms might such 'differential inclusion' take for children born in a host society that treats them as impossible and refuses to include them in schools? Here, by way of starting a larger conversation about the dynamics and specificities of undocumented children's excluded 'inclusion', I mention four forms of belonging. This is by no means an exhaustive list, and I have chosen to prioritise forms of belonging that are an explicit product of such children's life in the city. I find these forms of belonging particularly revealing, since they suggest the ways in which children's experiences have been shaped by novel living arrangements in KK, arrangements that are markedly different to many of their parents' childhood experiences. Not only is child illegality distinctive, but children are 'insiders' in different ways to adults. The first form of child belonging is nicely illustrated by my encounter with an Indonesian mother one morning at a learning centre. She told me that she and her children had recently returned from a five-month visit to her home village in Sulawesi. However, she said, they were forced to return to her husband in KK because the children did not really like the village. 'It's a real village', she stressed, 'And they are used to the city, not to walking to a village school'. Another mother, a Suluk refugee from the southern Philippines, recounted with laughter her attempt to relocate her children back to her brother's village after their father had died. She described their despondency when confronted by the backbreaking labour associated with a coconut plantation, and their relief when they finally returned to Sabah. Children born in KK, then, are children of the city, profoundly shaped by an urban life far removed from their parents' rural origins. Even as they inhabit many of its hidden areas and backwaters, children would describe to me occasional times when they had visited churches, mosques or shopping malls, asserting their connection to city places and therefore a claim to belong (cf. Coutin 2013, 325-326).
Secondly, these children belong to Sabah in linguistic terms. Many children told me that although their parents would always speak to them in regional languages such as Bugis or Lamaholot, they would always reply to their parents in Malay. Although Malay and Indonesian are mutually intelligible, there are differences of vocabulary and spelling, and Sabahan Malay has distinctive forms of pronounciation. At the official Indonesian school in KK, funded directly from Jakarta, teachers often complained to me about children's use of Malay rather than Indonesian, and posters around the building instructed pupils to speak Indonesian 'correctly and precisely'. The implication of these teachers' complaints was clear: these were Indonesian children, and should therefore speak their national language. However, amongst themselves, and with me, teenage pupils would speak emphatically Sabahan Malay. Some would also privately admit that they found code-switching into Indonesian at school a tedious performance. Since language use, particularly in multi-lingual contexts, can be a critical site for examining children's agency (Paugh 2005), we should see such children's language use as a (both conscious and unconscious) statement of belonging to Sabah.
Thirdly, and highly significantly, some children are intensely connected to Sabah as a place because they are the products of 'mixed' relationships, whether Javanese-Adonaran or Bugis-Bajau. Such 'mixed' kinship would be extremely unusual in most of their parents' home villages, which are largely mono-lingual and mono-ethnic. However, in KK, where migrants work, live and socialise together, even as they are separated from and stigmatised by Sabahans, such relationships are common. Many children told me that it was hard for their parents to live anywhere but Sabah, since they would not be able to choose one home village over another. Significantly, only a small minority of the children I knew had experienced living with grandparents or other kin in their parents' home country. Most of these children's parents had met and formed families in Sabah, and in many cases poverty, lack of kin in places of origin, but also an orientation to Sabah as home made it impossible to think of 'returning' children to Indonesia or the Philippines.
Finally, like adults, children also experience work as a powerful form of belonging. Some teenagers I met had never been to any form of school, and talked instead about the friends they had made by working in a brick factory or a timber yard. Other children combine study at a learning centre with paid work: occasionally helping to unpack onions in a warehouse near their home, or regularly selling iced drinks in a night market. Some work alongside their parents: cleaning a house together with their mother, or helping collect eggs in the chicken farm where the family lives. The latter children are undoubtedly exploited by a system in which whole families contribute labour, even whilst only one parent receives a wage. However, at the same time, many children find work rewarding and fun. Indeed, the extent to which work provides children with a form of productive belonging means that some children decide to leave learning centres, even against the will of a parent, in order to take up work. Wati, a 12-year-old girl, left her Indonesian learning centre to take up work at a carwash, though hid her ethnicity from her Sabahan co-workers.

Conclusion
When confronted by stories of midnight 'checking' raids, or of escape from immigration officials into bushes or chicken coops, it is hard not to feel intense sympathy for our research participants. When those participants are children, the history of humanitarian presentations of children as innocent victims (Malkki 2010) may work to heighten our sympathy. Perhaps this is why, when children occasionally laughed about such encounters, or joked that they were going to round up all their friends and send them to the police station, I felt so unsettled. Throughout my fieldwork in KK, even as I collected depressing stories of immobility and restriction, I was also continually surprised by children's responses. Even quite young children would stoically repeat the mantra that although Malaysia was a place full of police where it was impossible to be free, it was also a place that was 'good' because it had 'lots of work'. Children who felt despised by Malaysians, and resented their lack of time with hard-working parents, would still discuss places in KK that they loved, or would resist attempts by well-meaning Indonesian teachers to encourage them to stop speaking Sabahan Malay.
Such unpredictable, positive responses suggest the considerable resilience of these children in the face of adversity and prejudice (see Lester, Masten, and McEwen 2006). They also demonstrate how important to is to attend to the specificity of how 'migrant illegality' is configured for, and experienced by, children. Children's status as minors means that, in some contexts, their citizenship may not protect them from deportation. In Sabah, it means that, if caught with a mother in a 'checking' raid on migrant housing, a young child might face months of detention. However, if sitting innocuously on a bus, or if wearing an informal learning centre badge, the same child might have an unstable, temporary legality, allowing them to be passed by. Paying attention to child illegality not only puts children's experiences centre-stagean important aim, given the mostly adultfocused migration literaturebut also reveals the contingency that is key to understanding illegality as a frequently liminal condition (cf. Ngai 2004, 6;Menjívar 2006).
Drawing on Ngai's (2004) work, I have suggested that Malaysian immigration regulations, Sabah's history, and Malaysia's racialised politics of 'deservedness' all work to construct the descendants of migrants as 'impossible children'. These are children for whom Malaysia takes no responsibility, and who are considered impossible to educate, impossible to be regularised, impossible to be recognised as 'normal' Sabahan children. This notion of the specific impossibility of certain children is likely to be applicable to many contexts where children of migrants are rendered invisible, stuck in a 'liminal legal space' (Willen 2005, 76). Such children reveal the multiple contradictions in viewing migrant workers as just workers, rather than as people embedded in and constantly forging relationships, people with the same desires for intimacy and family life as those whose homes and shopping malls they build, whose toilets and offices they clean. Notes 4. These allegations were partially confirmed by the 'Royal Commission of Inquiry' into the presence of 'illegal immigrants' in Sabah set up in 2013. Although Sabah-based political parties had long demanded this Inquiry, many felt the Inquiry's eventual conclusions were watered-down for political purposes. 5. More recently, 'Project IC' has also become linked to the activities of armed groups in the Southern Philippines (such as the 2013 'Lahad Datu incident'), and the fear that Filipinoorigin Sabahans have become the 'enemy within.' 6. Alice Nah, 'Hierarchies of Deservedness: Governing Citizens and Migrants in Postcolonial Malaysia.' Presentation at Project Southeast Asia Symposium, University of Oxford, March 2014.