Break Their Roots: Evidence for China’s Parent-Child Separation Campaign in Xinjiang

Journal of Political Risk, Vol. 7, No. 7, July 2019

A Uyghur woman holds her son outside her house in the Kashgar old town, northwestern Xinjiang Uyghur Autonomous Region in China. Kashgar is located in the north western part of Xinjiang province where nearly 10 million Muslim Uyghurs are living. It is considered as the crossroads linking China to Asia, the Middle East, and Europe, the city has changed under Chinese rule with government development, unofficial Han Chinese settlement to the western province, and restrictions imposed by the Communist Party. The central government of China says it sees Kashgar’s development as an improvement to the local economy, but many Uyghurs consider it a threat that is eroding their language, traditions, and cultural identity. The discord has created a separatist movement that has sometimes turned violent, starting a crackdown on what Chinese government see as ‘terrorist acts’ by religious extremists. Tension has increased with lot more security in the city including restrictions at mosques, after closing and removing most of them in the Xinjiang province, the Chinese authorities have also restricted to the women to wear veils and the young men to grow beards. Source: Flickr.

Adrian Zenz
Independent Researcher [1]


In spring 2017, the Xinjiang Uyghur Autonomous Region (XUAR) under its new Party secretary Chen Quanguo embarked on an unprecedented internment campaign. Subsequently, up to 1.5 million predominantly Turkic minorities (notably Uyghur and Kazakh) were swept into different types of political re-education, detention and “training” camps.[2]

About half a year after the onset of this horrifying campaign, first reports started to emerge that the children of so-called “double-detained” parents were being placed in state care. Due to a complete lack of official reporting and the state’s denial that this internment campaign is even taking place, it has been extremely difficult to ascertain the scale and exact nature of such intergenerational separation. Some informants claim to know that these children are kept in orphanages. Others, including some of the affected parents, were told that they are being sheltered in regular public schools with boarding facilities. This includes boarding preschools that can care for children who are younger than the regular school age.

The possibility that the Chinese state is implementing a larger-scale or even systematic policy of intergenerational separation of parents and children is a highly emotional topic among the affected exile communities. Few issues have the potential to inspire more concern about cultural or physical state-sponsored genocide than this one. Notably, Xinjiang’s government appears to be as nervous about the children’s situation as they are about the internment camps. When Associated Press (AP) reporters went to Hotan City’s “Kindness Kindergarten”, which reportedly shelters four children of one of their Uyghur informants and photographed the facility, they were immediately surrounded by armed police and ordered to delete their visual evidence.[3]

In the context of this urgent human rights crisis and challenging research context, this article attempts to systematically present and analyze all available evidence regarding state-initiated intergenerational separation in the context of Xinjiang’s political re-education and internment campaign. This evidence consists of government policy and implementation directives, different kinds of official reports and related state or private media articles, educational statistics, public construction and procurement bids, village-based work team reports, and official propaganda pieces that extol the benefits of the so-called “vocational training”.

Government documents provide clear evidence that there are large numbers of children with one or both parents in some form of internment. These documents specifically refer to “couples where both partners are detained in re-education” (夫妻双方被收教), or “couples where both partners are in vocational training center” (夫妻双方在教培中心).[4] They also testify to the fact that this has developed into a concrete and urgent societal issue. From early 2018, the state began to issue urgent directives on how to deal with the virtually orphaned children of single or “double-detained” parents, be it through special care institutions or the regular education system. Local governments began to require schools to provide one-on-one “psychological counseling” and to proactively scan the state of mind of students with parents in detention in order to preempt trouble. Schools must now be prepared to mobilize entire teams of teachers, staff and other students to deal with such students when they are in distress, as well as taking measures for making up for their loss of family ties. Other evidence shows that schools have developed “emergency response plans” that include dealing with students with detained parents in a timely and effective manner in order to prevent violent incidents.

Additionally, the state has issued very detailed forms that are to be used by local authorities to log the situation of children with one or both parents in extrajudicial internment or prison. This data, which is fed into extensive databases, indicates that in some Uyghur majority regions, significant numbers of children are without the care of both parents. Government data shows that just in one particular township in such a region, well over 400 minors have both parents in some form of internment, with many others having one parent interned.[5] Children whose parents are in prison, detention, re-education or “training” are classified into a special needs category that is eligible for state subsidies and for receiving “centralized care”. This “care” can take place in public boarding schools or in special children’s shelters.

This does not mean, however, that these children are well taken care of. The real-life report of a Han Chinese volunteer teacher, posted on the Jiangxi Teacher’s College website, paints a harrowing picture of the consequences of systematic intergenerational separation.[6] The young man taught in an impoverished rural primary school in southern Xinjiang, where pupils were mostly without parents due to seasonal work or internment in vocational training camps. The young teacher wrote that these Uyghur children were in an extremely pitiful state, wearing thin clothes despite freezing December weather. The classroom was filled with an unbearable stench because the children neither washed nor changed their clothes.

In addition, the government has issued propaganda pieces that argue that the children of detained parents derive significant benefits from this separation, that both parents and children need to “study”, or that the “left-behind children” of parents who “work” are “happily growing up under the loving care of the Party and the government”.[7]

Overall, this article presents several key areas of evidence that in combination provide significant and potentially incriminating evidence for a coordinated state campaign to promote different forms of intergenerational separation. Xinjiang has not only created most of the necessary preconditions for systematically creating varying and substantial degrees of intergenerational separation; when placed in the wider context of Xinjiang’s securitization drive, the combined available evidence tells a story of the state’s dramatic race against time to create a vast and multi-layered care system that enables it to provide full-time or near full-time care for all children from a very young age (in several instances for infants that are only a few months old). In some Uyghur majority population regions in southern Xinjiang, preschool enrolment more than quadrupled in recent years, exceeding the average national enrolment growth rate by over 12 times.

In particular, this state care is taking place in highly secured, centralized boarding facilities, independently of any guardianship that these children may or may not have. Driven by multi-billion dollar budgets, tight deadlines, and sophisticated digital database systems, this unprecedented campaign has enabled Xinjiang’s government to assimilate and indoctrinate children in closed environments by separating them from their parents. This separation can take various forms and degrees, including full daycare during work days, entire work weeks, and longer-term full-time separation. When taking into account the threat that Xinjiang’s education system makes children report on their parents, it is safe to assume that parental influence in general, and intergenerational cultural and religious transmission in particular, are being drastically reduced. In some instances, parental influence is quite possibly almost completely eliminated.

The available evidence presented in this article can be viewed from four angles. Firstly, existing witness accounts from former detainees and their relatives indicate quite consistently that children whose parents are in some form of internment are put into either orphanages or boarding schools, with the latter case being more prevalent. Secondly, government plans show that the state is requiring local authorities and schools to comprehensively deal with children whose parents are in some form of internment. Thirdly, official documents testify to an entire set of policies, most of them initiated within the first six months of Chen Quanguo’s deployment to Xinjiang, that are designed to systematically boost the ability of the state to house children of all ages in increasingly centralized and highly securitized educational boarding facilities. Fourthly, government reports and construction bids give evidence of the construction of such highly secured boarding facilities in the public education system and through special child protection centers. All this took place in the second half of 2018, at a time when Xinjiang’s internment campaign was affecting more and more population segments, and when re-education and other detention facilities were expanding significantly in both number and size.[8]

It is only with what I refer to as the weaponization of education and social care systems that the region’s hair-raising political re-education and transformation drive is achieving its terrifying degree of seamless comprehensiveness. Increasing degrees of intergenerational separation are very likely a deliberate strategy and crucial element in the state’s systematic campaign of social re-engineering and cultural genocide in Xinjiang.

1 Evidence of Intergenerational Separation from Anecdotal and Witness Reports

The first indications of a large-scale, systematic separation of parents and children in the context of Xinjiang’s re-education internment campaign came from the accounts of affected Uyghur and Kazakh parents or relatives. Accounts of Xinjiang Turkic Muslims in exile, including former detainees and their relatives, indicated that children as young as 2 years, with both parents in either internment or exile, were put into state welfare institutions or kept full-time in educational boarding facilities. The former were variously described as orphanages or different types of “children’s’ shelters”, “child welfare guidance center” or “protection centers”.[9] Yet other children were taken care of at home by grandparents or relatives.[10]

There is further anecdotal evidence of de-facto separations of parents and children apart from the internment system. Several exiled Uyghurs told the AP that in their home districts in Xinjiang, all children above a certain age are required to attend a boarding school where they only return home during weekends and holidays. [11]

Generally, there is substantial evidence that the majority of children who have become “left-behind” in the wake of the massive internment campaign are being sheltered in regular schools with boarding facilities. Similarly, a Bitter Winter[12] report about Huocheng County (Ili Prefecture) states that a child of detained parents lives full-time at a boarding high school, and that that school has substantial numbers of children whose parents are both in internment. [13] Another report by this outlet suggests that in Lop County (Hotan Prefecture), nearly 2,000 children of “double-detained” parents have been placed in specially established nurseries and preschools, often with euphemistic names such as “Kindness” (pre-)schools.[14] Some, apparently fewer in number, are in orphanages, and “shelter houses”.[15]

A later Bitter Winter report describes a similar situation in Kashgar City. A local Han teacher stated that children of detained parents are confined inside their heavily secured schools, which look like prisons and have practically no holidays. By his account, many of the children at his school have parents in detention. A Radio Free Asia (RFA) report likewise cited a local official from Keriye (Ch. Yutian) County in Hotan Prefecture as saying that over 2,000 children aged 4 to 14, many of whom are separated from their parents, are being held in public education facilities from preschool to middle school levels that are called “little angel schools”.[16] An earlier RFA report, also pertaining to Hotan Prefecture, indicated that while children of detained parents were first placed into increasingly overcrowded orphanages, they were then put into nurseries or schools.[17] Similarly, a Financial Times report notes that while these children are being kept in “child welfare guidance centers”, the rapid increase of the internment campaign forced local governments to put many other children into boarding schools, including secondary vocational institutions.[18] A retired official from Kashgar told the Financial Times that the re-education campaign forced “a big readjustment within the educational system”, stating that many Uyghur children now eat and sleep at their schools.[19]

This is confirmed by government statistics, according to which the number of officially-registered orphans is nearly 240 times lower than that of all public school children (23,000 versus 5.5 million).[20] Generally, orphanages and other forms of shelter for special needs children naturally have a much smaller capacity than the regular education system. If the interment campaign has left large numbers of children without guardianship, one would assume that the state would try to somehow integrate them into public schools, because a ballooning archipelago of orphanages is a much tougher sell to western audiences than an expanding education system. Even so, in the second half of 2018 Xinjiang embarked on a massive campaign to construct so-called “children’s rescue, care and protection centers” (儿童救助关爱保护中心).[21] These are possibly what the Bitter Winter article referred to as “shelter houses”.[22]

Subsequently, this article firstly seeks to corroborate these anecdotal reports based on Chinese state and local media sources that directly link various forms of state care for children with the “vocational training”, re-education and internment campaign. Secondly, it examines how Xinjiang massively expanded its ability to shelter children of all ages, both within the regular education system and in the context of orphan care and general children’s welfare and protection mechanisms.

2 Evidence of Intergenerational Separation from Chinese Government and Media Sources

2.1 Propaganda Accounts of State Childcare for Parents in Vocational Training Internment Camps

Government propaganda pieces published by local Xinjiang authorities or related media outlets confirm such a separation in the context of its “vocational training center” de-extremification initiative, or what is here more aptly referred to as Vocational Training Internment Camps. These pieces also lend support to anecdotal evidence that many children of “double-detained” parents are kept full time in educational or childcare facilities that were specifically established for the children of the “students” or “graduates” of Vocational Training Internment Camps. By the state’s own admission, these camps are dedicated brain-washing institutions. To quote:

Vocational Skills Education Training Centers wash clean the brains of people who became bewitched by the extreme religious ideologies of the ‘three forces’. (职业技能教育培训中心把宗教极端思想从那些受到“三股势力”蛊惑的人的头脑中清除出去).[23]

In this context, some propaganda pieces point out that children are placed into full care facilities so that the parents can focus on their “studies” or their new “work”. For example, the Hotan City Vocational Skills Education Training Center (和田市职业技能教育培训中心) has an immediately adjacent “sunshine” preschool and primary school so that parents can engage in “carefree study”.[24] The children “eat and live” at the schools, and many of them reportedly started to call their teachers “mummy” (妈妈). This statement, meant to indicate the close and caring relationship between teachers and children, could also be a reflection of the fact that they see their real parents very rarely, if at all. According to the report, children are told that they and their parents are both “studying” so that their “family life will be happier in the future”. Another report about another preschool connected to the same “vocational training” camp, the Hotan City Kindness Preschool (和田市爱心幼儿园), says that it admits children who are only a few months old.[25]

Several other pieces state that the state established schools with boarding or childcare facilities right next to “vocational training centers” as a convenient service to the (interned) trainees. Such a report about the Yutian County Vocational Skills Education Training Center (于田县职业技能教育培训中心) states that because the children eat and live for free at these schools, their parents can save up their work incomes for the children’s future education.[26] Another article about the same facility notes that while the children are taken care of at the preschool, the female workers can freely enjoy life by dancing in the public square.[27] This “center” also has its own nursery (托儿所) for very small children below the typical preschool age: an article states that the little ones from the “Yutan County Vocational Education Training Center Nursery” (于田县职业教育培训中心职工托儿所) were singing songs and greeting the reporters in Chinese.[28] Another piece notes that the center’s school and nursery are only two minutes away from one of the associated factories.[29]

Several other reports speak of childcare options for re-education detainees who “graduate”. A Global Times article, also about the Yutian County Vocational Training Center, states that a factory where this center’s “graduates” now work has a nursery (托儿所) nearby, where parents can leave their very little children for the day.[30] An article about an Egyptian media team’s visit to Hotan’s Vocational Training Center also describes a nursery (托儿所) right inside a factory where women can drop off their children during the work day. The workers are said to be “graduates” from the region’s “vocational education and training center” (教培中心) where they “studied” for a period of 1.5 years.[31]

2.2 Government Documents and State Policy on State Care for Children with Parents in Different Forms of Internment

According to the Xinjiang Victims Database, which is a non-representative sample of over 4,000 mostly Uyghur and Kazakh persons in some form of detention or house arrest, 26 percent of the victims of the internment campaign have been women.[32] While most of the samples in this database are from northern or central Xinjiang, the southern regions, which are considered to be the most politically sensitive by the state, may potentially have seen a higher occurrence of “double-detentions” of both parents. Consequently, Xinjiang’s government must have developed some kind of systematic response to this growing phenomenon and related challenges.

Despite apparent efforts by the Xinjiang government to conceal detailed information on this matter, a concentrated search effort yielded significant evidence that related, detailed strategies do in fact exist. Overall, the evidence obtained by the author has incriminating potential, and directly suggests that Xinjiang has developed systematic ways to log and deal with the “left behind” children of single- or double-detained parents.

Perhaps the most blatant of direct evidence comes from forms completed by lower-level government administrations that list the exact detention situation of children with one or both parents in detention or external work, grouped by age, who are “in need of being cared for” (需要托养). An example from a well-known Uyghur majority region is given in Figure 1. [33] This table appears to group children according to ages that correspond to school levels: “six years and under” (preschool), “above six years” (primary school), and “teenagers in the middle school phase”. A related, detailed table with children’s names allowed the author to verify that all affected children have Uyghur names. In this particular case, 35.9 percent of children listed as requiring care had both parents in internment.

Figure 1: Local-level government lists of minors “in need of care”. Source: see footnote 33.

Another type of form that is disseminated to and completed by local government administrations is the “left-behind children’s roster” (留守儿童花名册). The meaning of the term “left-behind children” (留守儿童) and related policies are explained in detail in section 4. Here, it suffices to note that this term evidently includes many children where at least one parent is in some form of prison or extrajudicial internment, while the other parent is unable to care for the children due to abandonment, external work, illness, disability or death.

Some of these rosters contain extremely detailed lists of children with names, locations, ages and the exact fate of their parents. One region’s roster lists 621 children aged 1 to 4 years, all from a particular local Uyghur majority region.[34] The youngest is only 1 year and 3 months old, and along with others shown as being cared for in an “educational institution” (教育机构). Others are either with grandparents, still at home (presumably with older siblings or relatives), at nurseries or “infant care stations”, (托管站), at one of the “Kindness Preschools” (爱心幼儿园), or at other preschools. This document therefore provides direct and official evidence that educational institutions, including preschools, provide full-time care for children where both parents in an internment facility.

In this same roster, the parents are listed as either being in detention or re-education (收押或教), or being away for work. However, even though one key aspect of the nation-wide standard definition of a left-behind child specifically pertains to children whose parents are away for work reasons, and this is also a key reason given for why China has more and more of such children, this name roster only shows instances where both parents are absent for work-related reasons, and just a few instances where there is no information or where parents are absent for other reasons or had passed away. In the vast majority of cases, the parents are in prison, in a detention center (看守所 or 拘留所), a VTIC, a “training class” (培训班), a “re-education center” (教育转化中心) or in an unspecified form of “detention” (收押). The form then lists which of these children are not cared for by either a nursery, an educational institution (preschool or school), grandparents, or village committees, and are therefore in need of “centralized care” (需集中托养).

In yet another document issued by a Uyghur majority region of Xinjiang in late 2017, a local government entity specifically details the situation of children under 7 years, likely ones that are not yet in primary schools, where both parents were absent.[35] In 19.6 percent of the listed entries, both parents were absent for work reasons. In 17.4 percent of cases, both parents were in “re-education” (教育转化中心), and in 60.9 percent of instances both parents were in prison (监狱). Only 2.1 percent of these children were without parental care due to other reasons. Since these figures are only from one small local region, they should not be treated as a representative sample. However, this data strongly suggests that extrajudicial internment and imprisonment, at least in this region, are likely the primary reasons for children’s full loss of parental care.

Overall, these forms show that the state is creating a very detailed record of the precise situation of every left-behind child, and how government institutions are to deal with this. Besides these local government forms, various Xinjiang government agencies and educational institutions have issued a range of policy documents and reports that furnish substantial additional evidence about the state’s provisions and strategies for dealing with the fallout of its internment campaign.

Additional evidence was obtained from the website of Xinjiang’s Economic and Informationization Committee. Its website is blocked to users from outside China and has server-side mechanisms in place that prevent its pages from being saved to a web archive. A January 2018 document, referred to as the “XUAR Youth Development Plan 2017-2020”[36], refers to the need to support minors who are not sufficiently cared for because their parents migrated to urban centers for work.[37] This is a very typical statement that is ubiquitous in similar documents for other provinces. Sichuan’s plan adds the need to offer additional support for minors whose parents have been sentenced to prison, are undergoing coerced isolated detoxification, or who have been otherwise detained.[38]

Xinjiang’s document omits the reference to those who were formally sentenced, but stresses that the list includes the “minors of persons in detention and of trainees” (收押、培训人员未成年子女). The inclusion of such so-called “trainees” is unique to Xinjiang and not found in the Youth Development Plans of other regions, including for example Sichuan, Fujian, Guizhou or Heilongjiang.[39]

The meaning of this term and its connection to re-education is readily established. Numerous other official reports or documents from Xinjiang mention persons in custody or detention (收押) in the exact same vein as these “trainees” (培训人员), typically in the context that these two group’s family members are a special needs group that receive assistance from the government. For example, a September 2017 report by the Xinjiang Science and Technology Department (新疆维吾尔自治区科学技术厅) mentions that the department’s village work team donated rice, noodles, cooking oil and sugar to the “family members of those sentenced and in custody, and trainees” (判刑收押和培训人员家庭).[40] Other, similar reports of village visits even abolish the distinction between those two groups and directly refer to “detained trainees” (收押培训人员 or 收押收教人员).[41]  Another report by visiting officials from the Xinjiang Branch of the Academy of Sciences (中国科学院新疆分院) from October 2017 uses similarly explicit language, stating that the visitors had purchased coal for the relatives of “strike-hard detainees and educational training students” (严打收押及教育培训人员亲属).[42] Consequently, the use of the term “trainees” makes Xinjiang’s statement more incriminating. It indicates that by January 2018, when the re-education campaign was running at full steam, this had apparently started to become a major societal issue.

The term “education training students” derives from the standard expression for Xinjiang’s “vocational training” centers, literally “vocational skills and education training centers” (职业技能教育培训中心). A government notice by Aksu City aptly exemplifies the real nature of these “centers” by grouping the four types of institutions together that are administered by the justice bureau and the public security authorities: coerced isolated detoxification centers, (戒毒所), detention centers (看守所), prisons (监狱) and “vocational education and training centers” (职业教育培训中心).[43] In contrast, the same document places “vocational training schools” (职业技术学校) and other types of schools in a separate group which is administered by the public education bureau. This serves as a clear distinction as to which group is in some form of prison-like internment, and which is in a public education setting (where attendance may or may not be voluntary).

Apart from the XUAR Youth Development Plan, a Kashgar City notice pertaining to the expansion of boarding schools is much more direct about the urgency of the state to provide care and extra attention to the children whose parents have both either been “sentenced” or are in “vocational training”.[44] Issued in February 2018, only one month after the Youth Development Plan, this notice places the “school children where one parent has been sentenced” (父母单方被判刑的困境学生) and “school children where [both] parents [are] in vocational skills training” (父母参加职业技能培训的困境学生) into the same needs category. Both types of school children are referred to as “students in difficult circumstances” (困境学生) who are supposed to receive special attention and care. In Xinjiang and elsewhere in China, “children in difficult circumstances” (困境儿童) are typically defined as children where at least one parent is severely disabled or seriously ill, where the children themselves have a serious disability, where at least one parent is dead or serving a long-term sentence in prison or coerced isolated detoxification, or children with a single parent and a low household income where the remaining parent is either unable to financially support the child or unable to fulfill parental guardianship duties.[45] All of these children are “included in centralized [state] care” (纳入集中供养). Whether such “inclusion” is a choice or not is unclear from the relevant document.

The Kashgar City notice emphasizes the role of public education, especially of school boarding facilities and their personnel and management, in caring for such children. This, again, is consistent with previously cited informant statements. The wording of the notice reflects a distinct urgency, indicating that the care of children whose parents are both in re-education had developed into a significant issue. In particular, the text calls for “implementing a stringent joint responsibility” (严格落实包联责任) to these children’s unique situation. This expression refers to the fact that these children are said to be in need of attention by an entire set of persons with this care responsibility: teachers and student class leaders help the school leadership, class leaders help teachers and teachers in charge of dormitory management (宿管老师), as well as student cadres and dormitory heads (宿舍长). All are called upon to “resolutely put an end to negligence in monitoring students in distress.” (坚决杜绝困境学生漏管).

The notice places similar urgency on “emphasizing and strengthening psychological counseling” (注重加强心理疏导) of all students in difficult circumstances, and to “strengthen students’ thought education” (强化学生思想教育). In particular, all those responsible are to “grasp students’ state of mind in a timely manner, implement one-on-one psychological counseling and psychological correction to the few students in difficult circumstances, … make up for the lack of family ties, and eliminate the negative impact on personality development ” (及时掌握学生思想动态,针对个别困境学生采取“一对一”心理疏导、心理矫正、…弥补亲情缺失,消除人格发展的消极影响). To this end, the affected students are permitted and encouraged to write letters to their detained parents and even to send them short video clips.

Additional and very important evidence of the role of public schools in caring for the children of “double-detained” parents comes from a somewhat clandestine report issued by an unnamed local county’s education bureau.[46] It betrays its location in Xinjiang, and the wording corresponds with many related official documents discussing the situation of poor students or those in difficult circumstances in southern Xinjiang’s minority regions. The document’s date, February 2018, corresponds to that of the Kashgar City notice.

Titled “*** County Report on Special Supervision Work Conditions of the Spring Break And School Safety and Risk Prevention and Control”[47], the report lists several “current problems”. One of them is described as “further strengthening the educational assistance to the children of persons detained in re-education” (被收教人员子女的帮教有待进一步加强). The expression “persons detained in re-education” (收教人员) is a short version of (收押收教人员) mentioned above. It is commensurate with other terms such as 教育转化人员, literally “transformation through education persons”, or the term ” persons in transformation” (转化人员) that Xinjiang’s government employed as recently as August 2017 in connection with the re-education of Falun Gong members.[48]

More specifically, this anonymized report talks about the fact that in order to control and prevent violence, all schools in the county established “emergency response plans”. These include “loving care for the children of persons detained in re-education, … giving them psychological counseling in a timely and effective manner…and regularly educating teachers and students about safety and the rule of law” (关爱被收教人员子女,…及时有效给予心理辅导;…定期对师生进行安全法治教育). Schools are required to proactively “understand and grasp the family situation of the children of persons detained in re-education” (了解掌握被收教人员子女家庭基本情况).

Generally, the state appears to have made detailed and systematic provisions for the situation of such schoolchildren. At least by August 2017, it had issued a “notice on properly implementing statistics on information regarding the children of persons who are receiving re-education or have been lawfully detained” (关于做好受教育培训和依法收押处理人员在校子女信息数据统计的通知).[49] This permits the state to notify the relevant educational authorities so that they can closely monitor these students.

Another source of information about the fate of children where explicitly both parents are interned in re-education comes from the medical insurance system. The exact nature of the social benefits and subsidies that these families are supposed to receive is detailed in a strangely secretive but extremely insightful notice issued by the Kashgar Prefecture government in September 2018.[50] In the document, the commonly used shorthand phrase for “re-education center” (教培中心) is replaced with four diamond shaped ASCII characters, presumably to fool search engines or full text searches. However, the context surrounding each occurrence makes the decoding of this replacement quite straightforward.

A particularly interesting feature of this document and possibly the reason for its concealment is the use of the terms “couples where both partners are detained in re-education” (夫妻双方被收教), and “couples where both partners are in VTICs” (夫妻双方在教培中心).[51] The result of such “double detentions” is that the care of their children is clearly an acute issue that the state must address. Along with “persons in especially difficult circumstances” (特困人员), double-detained couples and their family members are classified as “especially needy groups” (特殊困难群体).  As above, their children are referred to as “children in difficult circumstances”. The document states that when these children’s parents are both in VTICs (夫妻双方在教培中心) and the family cannot pay their medical insurance bills, the insurance for them and for their children who are at school must be paid for by the local financial authorities. Generally, the document notes that the “re-education bureau is responsible for handling the task of paying for the health insurance of each vocational training and education center student” (教培局负责各教培中心学员参保缴费工作).[52]

A specific example of the negative social impact of “double-detention” comes from the “becoming family ” (结亲) team visit report of the Forestry Department to Aotebeixixiang Township in Wushen County, Aksu Prefecture, from November 2017.[53] The team visits 13 poor households to give them gifts of “consolation” (慰问品). One of the households they visit is a “focus household” (重点户), meaning one that is of particular concern in regards to social stability and combatting extremism. There, “both spouses are in detention” (夫妻双方被收押). Consequently, the family’s oldest son Yi Muran Aili, aged 18 years, relies on odd jobs in order to take care of three younger siblings. The team exhorts him to “face up to the problem”, to “not follow the evil, old path of the parents”, and to “eternally follow the Party”, considering that the “lives of peasants are getting better day by day.”

Another reference to the children of parents who are both in detention comes from an Excel spreadsheet from a website linked to the Kuytun government. [54] This spreadsheet contains detailed entries about persons, their government-issued ID numbers, their livelihood situation, information about their parents and siblings, and so on. One of the entries pertains to a person with a female Uyghur name whose parents are both detained in re-education (父母被收教). Her little brother is in primary school. The household is categorized as low-income and receiving minimum living allowance welfare payments (低保户). Another entry in the same spreadsheet mentions a person with a Kazakh name who is the “child of persons detained for re-education” (收教人员子女). This household is likewise classified as being in “difficult living conditions” and “low income” (生活困难, 低收入). Other documents on this website show that it hosts systematic government information about children in special needs circumstances, including orphans.

Similarly, the Kashgar Prefecture Poverty Alleviation Policy Interpretation document, issued in May 2018, states that for “children in difficult circumstances” due to parents in prison, custody or re-education” (三类人员)[55], the prefecture has earmarked annual livelihood subsidies of 60 million RMB. [56] Similarly, Hotan City’s education department was expecting to spend 10.8 million RMB in 2019 on food and livelihood subsidies for “students in difficult circumstances”.[57]

Another piece of evidence of the fate of children with parents in detention comes from the report of a young graduate from Zhejiang Province who volunteered to teach Chinese in a rural primary school in Pishan County.[58] Dated December 2018, the report details how in the middle of winter, with temperatures as low as -20 degrees Celsius, many children were still wearing thin clothes. These clothes also had not been changed or washed in a long time so that the entire classroom stank. The young teacher learned that the pupils’ parents were away picking cotton or “studying at the VTIC” (在jp中心学习). With both parents away, the “students were in an extremely pitiful state” (学生非常可怜).

While the situation of children is especially desperate when both parents are in detention, the internment of one parent also puts families in a dire predicament. On February 19, 2018, a father of two young children from Ili Prefecture was “detained in re-education” (被收教).[59] On April 17, 2019, 14 months later, his wife posted on a legal counsel website that she did not know how to cope with the loss of her husband’s support and labor, given that she has 6.6 acres of land to farm and is raising her children by herself. Because of that, she is unable to go out to work and seeks advice on how to apply for her husband to be released.

Overall, the XUAR government has had to establish a strategy for dealing with thousands or more children whose parents are both in some form of internment. Available documents show that while a general information requirement on such children existed as early as summer 2017, this strategy was disseminated in more detail and with greater urgency from early 2018, a time when the re-education campaign and the expansion of different types of internment facilities was heading towards a peak.[60] In doing so, the state appears to have been bending over backwards to conceal its larger schemes from the outside world. Due to the blocking of websites based on user location, a deliberate masking out of compromising phrases, a constantly changing terminology, apparent limits on the overall number of related documents published on government and other websites, and widespread efforts to avoid stating the obvious directly, this information was considerably more difficult to obtain than evidence about Xinjiang’s re-education campaign in the first half of 2018.

Evidently, the state considers the fate of these children to be a highly sensitive topic. Xinjiang’s government actively conceals its intergenerational separations. Even so, the available evidence from official government documents provides crucial puzzle pieces that help us expose the situation on the ground.

2.3 “You Both Study”: Extolling the “Benefits” of Intergenerational Separation

Besides the state propaganda analyzed in section 2.1, there is another angle to the government’s public-facing positive spin on intergenerational separation in the context of its re-education campaign. Some propaganda pieces specifically extol the supposed “benefits” to Uighur children growing up with full or at least substantial separation from their parents.

Besides improved Chinese language skills, children who live in boarding preschools in or near “vocational training” camps are said to develop better life habits, such as personal hygiene (face washing, teeth brushing etc.), polite manners, or more “open personalities”.[61] Conversely, these pieces argue that parents who have been affected by “extremist thought” fail to raise their children well: they are accused of neglecting them, failing to teach them “good life habits”, instilling negative perceptions about non-Muslim ethnic groups or general life in Han society, and preventing them from getting an education or proper medical care.[62] In contrast, the beautiful and clean preschool and primary school environments with their good slogans “naturally” bring out good habits and manners in these children. A report of an Egyptian media team visiting Kashgar and Hotan notes that the preschools that have been established throughout these regions are a means to inoculate children against “extremist thought influences”.[63]

“Extremism” is depicted as a key cause of ethnic difference and segregation in a region where the government’s core policy is “ethnic unity” (民族团结).  One woman supposedly confessed, “When my children became sick I would not take them to the hospital, [because] at the hospital they are all Han, the medicine they prescribe I cannot eat, they are unbelievers, I won’t go to the hospital.”[64]

Because of the failed parenting methods and negative influences of parents with “extreme” ideologies, not only the parents need to be interned for “training”. The children are likewise considered to be in need of “study”. As one piece cited the Party secretary of the Hotan City Vocational Training Center preschool as saying: “I tell the children: ‘Your parents and you all alike are studying'”.[65] The message is that state institutions and government teachers are better at parenting then the children’s own parents.

These types of accounts indicate that intergenerational separation is very likely not just incidental to Xinjiang’s re-education campaign, but a distinct and strategic component of it.

2.4 Conclusions

 The combination of informant accounts, state propaganda extolling benefits of state-provided childcare for “trainees” or “graduates”, government documents outlining state provisions and mechanisms for managing intergenerational separation, detailed government lists of children with both parents in detention, and other propaganda pieces that highlight the distinct “benefits” of such a separation amounts to an overwhelmingly strong case for the allegation that Xinjiang’s re-education campaign is causing intergenerational separation on a large scale. Its impact is evidently large enough to mandate region-wide regulations and practical guidelines in order to at least contain part of the resulting social and psychological fallout.

The warm, almost emotional language used to describe the “care”, “love” and “nurture” of the Party and the government towards the re-education detainees and their children is of course designed to make intergenerational separation somehow seem more acceptable. At the same time, it divulges the intention of the Party to aspire towards some kind of metaphorical “parenthood”. The message the government seeks to portray is that the Party is the better parent.[66] It provides superior childcare and shelter facilities, better parenting, more attention and care, a better cultural influence, a superior moral value system, the more useful language (Mandarin), and a generally brighter future for children than the minority parents could ever give. One government directive even directly states that Xinjiang’s new centralized boarding school system “reduces household’s economic and mental burden,…solves parents’ worries,…freed them from the troubles of family affairs“. By vastly outspending the actual parents in terms of material provisions for minority children, the state is effectively creating a fully immersive replacement for the traditional minority childhood and socialization experience.

Besides portraying itself as the better parent, the Party also places itself in the position of being “parent” for both minority children and their parents. One of its endlessly cited slogans in Xinjiang purposely copies the filial piety of traditional society, and is learned and recited by both adults and children throughout Xinjiang:

“Be grateful to the Party! Obey the Party! Follow the Party!” (感党恩听党话跟党走).[67]

The message for minority parents is that they are to learn and imitate the (superior) parenting practices of the Party, which thinks of itself as rescuing them from spiritual “extremism”, cultural “backwardness” and socio-economic irrelevance. However, even the few government documents on the subject offer us a small glimpse of the likely reality on the ground, confirmed and complemented by the harrowing local informant accounts reported by media outlets such as Bitter Winter. The actual psychological impact of the mass detentions on the parents and children must be absolutely devastating.

The subsequent sections examine significant further evidence that the state may in fact be perpetrating intergenerational separation as a calculated, systematic strategy. Not only does state propaganda extol the benefits of such separation. Under Chen Quanguo’s leadership, Xinjiang embarked on a tightly-managed multi-billion dollar campaign to create comprehensive care and boarding facilities across the region, driven by urgent anti-extremism rhetoric, military-style discipline, and extremely ambitious deadlines.

3 Building Intergenerational Separation Infrastructure: Xinjiang’s Drive to Establish Securitized Boarding and Care Facilities for All Ages

Official documents testify to an entire set of policies designed to boost the ability of the state to house children of all age in increasingly centralized and highly securitized educational boarding facilities. Although some of these initiatives are national rather than regional, it is nevertheless notable that their timing closely correlates with the onset of the re-education campaign in spring 2017. Several of them directly coincide with the emergence of Chen Quanguo’s leadership in Xinjiang. Moreover, those that were mandated for all Chinese regions were typically implemented in the XUAR with unprecedented and unparalleled intensity, speed, and scale. Funding apparently represented no serious limitation.

Moreover, the regional focus and target group of these initiatives just so happens to fully overlap with that of the re-education campaign: Xinjiang’s predominantly Muslim minority regions, especially the south; and children from impoverished families, often those with “special needs”.  While the expansion of social care and education is typically a laudable development, in the context of Xinjiang’s all-encompassing securitization and assimilation drive, and in light of government statements about the benefits of intergenerational separation, this expansion does give rise to serious concerns.

3.1 Chen Quanguo’s Military-Style Preschools Construction Drive

In September 2016, merely one month after ascending to power in the region, Chen Quanguo proposed a massively accelerated preschool campaign for the rural southern Uyghur regions, with early 2017 as the start date.[68] The explicit aim was to achieve in advance the goal of the next 5 year plan of three years universal preschool education. This was clearly a new initiative and a diversion from the original plan, which had just been drafted in May 2016 and only called for an 80 percent preschool enrolment share of children aged 4-6 years by 2020.[69] Chen accelerated this drastically, aiming to achieve 100 percent enrolment by autumn 2017.

A specific example is the original 5 year plan for Kashgar Prefecture for 2016-20, which was only to have a three year age group (4-6 years) preschool enrolment rate of 80 percent by 2020.[70] Drafted on January 28, 2016, Kashgar Prefecture’s 5 year plan envisioned the construction and expansion of 490 new and existing preschools by 2020 to achieve this goal. However, after adjustment under Chen Quanguo’s leadership, Kashgar built a total of 1,152 preschools by autumn 2017, over twice the original figure.[71]

By late February 2017, Xinjiang had begun the construction of 4,387 “bilingual” (i.e. Chinese language focused) preschools, with a planned intake of 562,900 new students. The focus of this campaign was on the rural south, home to most of the Uyghur population, and it was done with great urgency. Several news reports state that the construction was done at maximum pace, using a Chinese idiom that refers to “spurring a horse to full speed” with a whip (快马加鞭).[72] The coordination between the involved government departments for the fast and efficient construction of the schools was said to be done in a “military command fashion” (军令状). Preschool construction was supposed to be completed in time for the start of the new school year in September. However, Hotan Prefecture even required preschool construction to be finished before July 25.[73] This massive and accelerated construction drive was financed through state bank loans totaling 8 billion RMB for the southern minority regions alone. Just in Kashgar Prefecture, central government subsidies for preschool construction amounted to 767.6 million RMB.[74]

Both speed and scale of preschool construction and related enrolment increases in Xinjiang’s southern Uyghur majority regions were unprecedented in the entire nation. Between 2015 and 2018 and measured per population aged 0-5 years, preschool enrolment in three southern prefectures in Xinjiang with Uyghur majority populations rose by 148 percent, with Hotan Prefecture showing the highest increase (Figure 2). The enrolment growth for the entire XUAR was still over twice as high as Gansu Province, and far higher than most other provinces or the national average. Because these figures are adjusted per capita of the population aged 0-5 years, they already compensate for the fact that some regions, notably minority regions such as Hotan Prefecture[75], have a much higher birth rate than for example the national average. When using absolute enrolment figures, Hotan’s 2012-18 enrolment rose by 328 percent, compared to a 26 percent increase of the total national enrolment.

Figure 2: Sources are the socio-economic development reports of each region.

Likewise, enrolment growth between 2012 and 2018 was highest in Hotan (199 percent), followed by a XUAR Uyghur majority region sample[76] consisting of three prefectures (157 percent), the entire XUAR and Gansu province (87 percent), and the XUAR non-Uyghur majority region sample with only 38 percent. By comparison, national preschool enrolment growth during that time period stood at only 20 percent. To exemplify the scale of the increases in some regions with a Uyghur population share of over 95 percent: in Lop (Ch. Luopu) County (Hotan Prefecture), the number of preschools rose from 8 in 2009 to 180 in 2017.[77]

Xinjiang’s planned new preschool student intake for 2017 was 562,900, but the actual number was nearly 200,000 higher at 759,900. [78] Planned total enrolment by autumn 2017 was supposed to be 1.04 million, but the actual end-of-year figure stood at 1.45 million.[79] The originally planned enrolment figure would have been equivalent to 433 preschool children per 1,000 children aged 0-5 years.[80] This would have been commensurate with a near full enrolment of children ages 4-6 years into preschool. The actual figure of 1.45 million equates to 602 preschool children per 1,000 children aged 0-5 years, from which we can infer that the entire age group 3.5-6 years was enrolled. In 2018, Xinjiang’s preschool enrolment rose to just below 1.6 million children, over twice as many as in 2014. At that point, nearly all children aged 3-6 years would have been integrated into a state-controlled educational environment (637 preschool children per 1,000 children aged 0-5 years).

Since the original stated goal of the preschool enrolment drive was to enroll all children aged 4-6 years, it appears logical that the higher than expected enrolment was caused by enrolling children below that age. After all, older children are obligated to advance to primary school, and hence mandatory education. Also, the author found no evidence whatsoever that Xinjiang’s preschools keep children older than 6 years. Enrolment figures may be inflated by unreported children who were born in violation of existing family planning regulations.[81] However, China’s growing efforts from the 2000s to comprehensively enforce mandatory education and family planning in minority regions, combined with Xinjiang’s increasingly fine-grained domestic security system and the emphasis placed on registering every citizen in a digital file, makes it unlikely that unreported children alone can explain this discrepancy. Rather, at least one major reason for enrolling younger children is fairly evident: the massive re-education and detention campaign left many young children without adequate parental care.

Figure 3 indicates that the average national preschool enrolment per 1,000 children aged 0-5 years increased steadily, and recently nearly encompassed the full 4-6 year age cohort. Other regions likewise experienced very gradual increments, except Gansu Province, whose enrolment ratio jumped in 2016. Xinjiang’s preschool enrolment ratio was traditionally below the national average and that of the neighboring provinces of Gansu and Qinghai. This changed dramatically in 2017 and 2018. Xinjiang’s enrolment ratio is now by far the highest in the nation.

Figure 3: Sources are regional socio-economic development reports, the 2010 national census and the 2015 census for Xinjiang.

Within Xinjiang, there is a strong discrepancy between regions where Uyghurs or other Turkic Muslim minorities constitute the majority of the population, and those where their share is below 30 percent. These figures not only confirm the known fact that Xinjiang’s preschool enrolment drive specifically targeted the southern regions with high Uyghur population shares. They also demonstrate that these regions have by now enrolled a far higher share than the target population of 4-6 years. Overall, the three Uyghur majority prefectures in the sample have enrolled nearly all children aged 3-6 years, one full year more than the three non-Uyghur regions.

Figure 4: Sources are regional socio-economic development reports, the 2010 national census and the 2015 census for Xinjiang.

In Hotan Prefecture, the region with the highest enrolment ratio, all children aged 2.7 to 6 years would have been in preschool by the end of 2018. This age range is exactly one year wider than that of Guangdong Province, the region with the highest per child preschool enrolment figure outside Xinjiang, which has enrolled all children aged 3.7-6 years, and nearly 1.5 years more than the national average age range of 4.1-6 years. Hotan’s high enrolment figures are consistent with reports that at least some of its preschools admit children from a very young age, and that the region has established an apparently fairly substantial number of nurseries in the context of the re-education and forced labor campaign. In contrast, the average enrolment ratio of the three non-Uyghur majority regions in 2018 was almost exactly the same as the national average.

Likewise, educational data for Xinjiang that is broken down by ethnic minority memberships shows that preschool enrolment growth has been particularly high among minorities. In 2012, minority children made up 67.5 percent of all of Xinjiang’s preschool children. By 2017, this share had risen to 80.5 percent.[82]

This distinct minority emphasis is further underscored by the fact that Xinjiang’s urgent preschool construction drive has had an overwhelmingly rural focus. For example, in Hotan Prefecture, 1,008 out of 1,040 newly built or expanded preschools were in villages, 17 in townships, and only 15 in urban centers.[83]

3.2 Evidence That Xinjiang’s Nurseries and Preschools Admit Very Young Children

Xinjiang’s preschool expansion plans and many related reports state that the enrolment drive targets children aged 4-6 years. In a published email exchange between a parent who seeks preschool admission for his 3 year old child, and the government of Qira County (Hotan Prefecture), the responding official confirms that the child is too young for regular admission.[84] However, the official notes that this region has established preschool classes from age 3 years 6 months, and can therefore accept this child. In fact, the official minimum admission age for preschool in Xinjiang as stated in the region’s preschool evaluation standard is 3 years. [85] This document lists a preschool age category called “small preschool class” (幼小班) for children aged 3-4 years that is also commonly found in other parts of China.

Several preschools in Xinjiang even state that they admit children from age 2 years, for example a preschool in Aksu City.[86] As noted before, Hotan City’s Kindness Preschool admits children who are only a “few months” old. A Uyghur lady interviewed by the AP, was told by a friend that her children, including a 3 year old, were kept in that very institution.[87] Likewise, a Bitter Winter report mentioned nine preschools in Lop County for children aged 3-6 years.[88]

Some preschools throughout China, and others in Xinjiang specifically, have integrated nurseries that admit even younger children.[89] Nurseries in China generally take children starting from age 0-2 years, and can also be standalone facilities. Generally, a rising number of preschools in China have nursery classes for children below age 4. These are referred to as “care-taking classes” (托班), for example “medium care-taking class” (托中班) for ages 1-2 years and “big care-taking class” (托大班) for ages 2-3 years.

While no equivalent is found in Xinjiang’s overall development plan, Kashgar’s 13th 5 year plan (2016-20) for socio-economic development states that the establishment of “nurseries” (托儿所) by private enterprises or local community initiatives is to be encouraged. Generally, one certainly gets the impression that nurseries have become far more common in Xinjiang, not just in major cities, but also in rural minority regions.

A report about a nursery in Akelangan village in Hotan Prefecture, which takes care of 20 children aged 2-4 years, shows photos of very young children, including toddlers with pacifiers.[90] An article about a nursery in Pishan County states that the smallest children are only 8 months old.[91] In a village in Hejing County, the local preschool admits children aged 1-3 years.[92] A nursery in Lop County even admits children between 0-4 years.[93] Similarly, a Bitter Winter article notes that that county has built 11 “Loving Heart” nurseries for children aged 1-3 years.[94] Finally, a report about Shache County cites the example of a mother who can leave her 1 year old at the local nursery.[95]

All in all, the available evidence suggests that Xinjiang has been establishing an increasingly comprehensive early care network that is progressively able to admit children even from their infancy.

3.3 Massive expansion of boarding facilities at all levels and evidence for Educational Centralization

A systematic separation of parents and children would require the state to care for these children full time or nearly full time. Since 2017 and especially 2018, Xinjiang rapidly increased its ability to do so.

All of the nurseries mentioned in this article are described as day care facilities. However, some of the reports show facilities with bunk beds, in one case even a room full of small cribs.[96]  Typically, full day care facilities for young children have beds so that children can nap after lunch. There is nothing that would prevent many or most of Xinjiang’s preschools and nurseries to keep at least some of these children overnight if the government told them to do so. Similarly, preschools are typically always equipped with beds, sanitary facilities and kitchens even if they only take care of children until the afternoon. For example, a report about bilingual rural preschools in Lop County (Hotan) from July 2017 confirms that this Uyghur region’s 149 newly constructed facilities all possess such equipment.[97]

At regular educational levels (preschool and up), the state’s ability to shelter children full-time increased greatly throughout 2017 and 2018 as the region accelerated  educational centralization and undertook a truly staggering expansion of educational boarding facilities for all school types. Centralization means that the state can concentrate children in large compounds that provide increasingly comprehensive 24-hour care in a highly controlled and locked down environment.

Xinjiang’s unprecedented preschool construction drive not only dramatically increased the number of enrolled children, but also led to an even more stunning increase in total floor space. This is not only an indication of a larger number of larger and improved facilities, but also evidence of massively expanded boarding capacity. This is especially evident in rural southern minority regions. Between 2016 and 2017, the number of Xinjiang’s preschools rose by 68 percent and enrollment increased by 57 percent, but total floor space grew by 85 percent (Figure 5). Hotan City nearly tripled its total preschool floor space from 87,499sqm in 2016 to 214,347sqm in 2017. Its number of preschools grew by 64 percent, enrollment figures by 90 percent, and floor space by 145 percent. In 2018, its preschool floor space again nearly doubled to 396,235sqm. As a result, the average floor space per preschool more than doubled, from 1,199sqm in 2016 to 2,979sqm in 2018, reflecting a clear centralization trend. The average floor space per enrolled preschool child likewise doubled. Between 2016 and 2018, enrolment figures rose by 126 percent while total floor space grew by a staggering 353 percent. By comparison, China’s national preschool floor space figures are not typically shown, except in the 2014 educational development report. There, a 13.9 percent preschool floor space increase was highlighted as a particularly strong increase that significantly contributed to overall national educational floor space growth for preschools to middle schools.[98] In 2018, Hotan City had on average 426 students per preschools, over twice as much as Xinjiang (203 students per preschool) or the national average (175 students per preschool). For the same year, the average preschool in Hotan City had 2,979sqm of floor space. This is more than twice the average of Xinjiang as a whole (1,452sqm), over four times as much as for Liaoning Province (717sqm) and nearly three times that of Shandong province (1,148sqm).[99]

Figure 5: Sources: Ministry of Education National Education Status Reports, XUAR Department of Education Educational Development Reports, Hotan City Government Social and Economic Development Reports.

This is not just evidence of improved and enlarged educational facilities, but also of the construction and addition of student dormitories. Documentation, including government and media reports, identify several of Xinjiang’s preschools that provide “full care” (全托), as opposed to only “(full) day care” (日托 or 全日托) or “part-care” (半托).[100] In China, “full care” typically refers to a boarding institution where children are dropped off on Monday morning and picked up on Friday, at the end of the work week. In contrast, “(full) day care” means that children are dropped off each morning and picked up at evening after work, without staying at school overnight. “Part-care” means that children must be picked up during the day, for example after lunch. The Hotan City Kindness Preschool is described as a “full care” institution, as is the Yudu Kindness Preschool in Lop County, and there are other examples in Aksu and Kashgar Prefectures.[101] Similarly, a propaganda report describes a preschool in Lop County, where the children are in “full care classes” (全托班) and performed a theater piece about how “left-behind children” are “happily growing up under the loving care of the Party and the government”.[102] As outlined in section 2.2, government rosters show that children with both parents in detention are being taken care of in preschools and other educational facilities.

Several examples of the newly-constructed or commissioned educational facilities are listed in Table 1. Typically, they are much larger and/or more costly than existing ones, with floor space ranging between 4,000-20,000sqm. By comparison, the average floor space per preschool in Xinjiang in 2018 was only 1,452sqm and per primary school 4,859sqm. China’s average preschool floor space in 2016 stood at 1,197sqm, for preschools in villages only about 700sqm.[103] The nation’s average middle school floor space in 2018 amounted to 12,378sqm. Some of the new preschools shown below are therefore larger than an average national middle school, up to 27 times larger than an average national village preschool, and up to 10 times larger than an average national urban preschool. Again, this confirms the educational centralization and full-time care (boarding) strategy of the state in Xinjiang’s ethnic minority regions.

Date School name Location Floor space / Capacity Cost (m. RMB) Source
8/2016 Beijing Xichengqu Hotan City 17,482sqm 80.0
4/2018 Not stated Xinhe County (Aksu Prefecture) 19,000sqm (1,890 preschool & primary students) 45.0


9/2018 Huanhu Preschool Hotan City 16,980sqm (preschool for 1,080 children)

29,368sqm (primary school for 1,728 children)

218.0 and
2/2019 Sunshine Preschool Hotan City, Yulongkashen Township 5,688sqm 16.8
3/2019 East Road Preschool Changji City 5,460sqm 25.0


3/2019 Gelinxiao Township Preschool Changji City 4,000sqm 18.0
4/2019 No.23 District Preschool Changji City 4,800sqm 25.0

Table 1. Preschool and primary school construction bids.

In 2018, Xinjiang embarked on a new campaign to centralize and expand the boarding capacities of schools above the preschool level. In March 2018, the Xinjiang government issued a call to strengthen the construction of boarding schools throughout the region. [104] Related construction and management efforts were a key component of the region’s “rural revitalization” scheme, with a time frame from 2018 to 2022.[105] In June 2018, a government directive specified that all urban and rural regions in Xinjiang were to establish boarding facilities, to promote ethnically mixed classes for Han and minorities combined (推进民汉合校混合编班), and to centralize primary and middle school facilities as combined units (各级教育行政部门统筹推进幼小、小初衔接).[106] The rationale for the latter is to minimize dropouts (控辍保学), ensuring that all students, especially those from poor or rural backgrounds, firmly remain within public school contexts until their mandatory education is completed.

A government directive from Kashgar Prefecture states this even more explicitly as “striving to complete the 2020 goal of completing the construction of dormitory-based schools (寄宿制学校)”, with boarding facilities playing a key role in the retention of students from “poor families” (家庭困难). [107] In particular, the boarding school system is designed to solve the living and study problems of the children of those who “work outside their home regions (外出务工)”. However, from the government roster of left-behind children discussed above it is evident that in Xinjiang such a lack of guardianship is not primarily caused by work-related factors. Rather, it is caused by parents being in prison or extrajudicial internment.

Kashgar’s directive further states that for grades 4 and higher, boarding is essentially mandatory (应寄尽寄). This is consistent with informant statements cited by AP that a 9 year old child was “automatically transferred to a boarding school”.[109] This directive appears to have been implemented in other minority counties. A report from Hejing County in Bayingolin Prefecture confirms that primary school students above grade 3 were ordered to relocate to the local middle boarding school in order to board and study there.[110]

A specific example of what Xinjiang’s mandated educational centralization and boarding scheme looks like can be found in Yecheng County (Kashgar Prefecture). In April 2018, the county mandated the relocation of 2,000 children from the surrounding rural areas to one single boarding middle school in the county seat.[111] This school’s facilities boost a total floor space of 75,000sqm and can hold over 5,000 students. By comparison, China’s average middle school in 2018 had only 12,378sqm of floor space, one-sixth of the Yecheng County school.[112]

As with the preschools, the widespread implementation of an educational expansion and centralization initiative in 2018 can be further confirmed by public bid documents and educational statistics. Hotan City’s primary school floor space rose from 205,569sqm in 2016 to 285,112sqm in 2017. The following year, it jumped to 403,174sqm.[113] The floor space increase was three times higher than enrolment growth. As a result, the floor space per school doubled from 2,538sqm in 2016 to 5,040sqm in 2018, and the average floor space per enrolled child grew by 50 percent.

A comparison of Hotan City with Xinjiang and national figures again shows that the centralization and boarding expansion drive had a clear focus on the rural south. Hotan City’s primary school floor space growth far outpaced both Xinjiang and national growth figures (Figure 6). While this can to an extent be explained by a need for these educationally inferior regions to catch up to regional and national levels, Hotan City’s 2018 average primary school floor space now exceeds the national average by 181sqm (5,040sqm versus 4,859sqm).

Figure 6: Sources: Ministry of Education National Education Status Reports, XUAR Department of Education Educational Development Reports, Hotan City Government Social and Economic Development Reports.

Between July and December 2018, many counties in Xinjiang issued construction bids for large primary and secondary boarding school complexes and numerous individual boarding schools, involving hundreds of millions of RMB. The focus of this drive was, again, in the southern Uyghur-dominated regions. For example, Yecheng County in Kashgar issued a 78 million RMB bid for a 37,300sqm boarding middle school.[114] The dormitory alone was supposed to have a floor space of 13,550sqm, larger than China’s average middle school in 2018. At 4,820sqm, its dining hall alone was planned to be about as large as China’s average primary school (in 2018). Similarly, Bachu County in Kashgar issued a 70 million RMB bid for a 36,400sqm boarding school with a 10,920sqm dormitory.[115] Jiashi County, also in Kashgar, did the same for a 89.4 million RMB boarding middle school with 40,618sqm total floor space and a 16,500sqm dormitory.[116] Large primary boarding schools were also set to become a common sight. Yopurga County, likewise in Kashgar, issued a bid valued at 100 million RMB for the construction of 6 primary schools.[117] The largest of them totaled 10,534sqm of floor space, with its 3,562sqm dormitory as large as many entire regular primary schools. In 2018, Xinjiang’s average middle school had a floor space of 15,388sqm, which is 3,009sqm more than the national average (of 12,378sqm).

Interestingly, at least part of this construction frenzy is not called for by current educational student flows. Figure 7 shows the share of new students in relation to graduating students for each educational level. A share of 0 percent means that new students and graduating student numbers are the same, while a share above 0 percent indicates that more new students are coming in than existing ones are leaving. The figure clearly reflects the massive preschool intake of 2017, when twice as many new preschool students were admitted than graduated.

Figure 7: Source: Kashgar Prefecture Social and Economic Development Reports.

Between 2016-18, Kashgar Prefecture has been admitting between 56 and 84 percent more new primary school students than those who graduate from the primary level. This explains the need to rapidly expand primary school floor space. However, the same is not true at the middle school level.[118] The new middle school facilities are therefore either designed to centralize middle school education around vast boarding facilities, replacing scattered local middle schools that did not offer dormitories, or, they are meant to incorporate vast numbers of primary school students who are likewise supposed to live and study in vast, centralized boarding facilities. Centralization and boarding in turn facilitate much greater degrees of state control, enable the state to shelter and monitor “children in difficult circumstances”, and could in fact be an essential means for promoting sustained intergenerational separation.

Xinjiang’s educational expansion trends are also reflected in state spending on education. Figure 6 shows annual changes in spending on education in recent years. Xinjiang’s education spending increased significantly in 2017, but this increase was most marked in the southern regions. Spending growth in 2018 was likewise very high, and in Xinjiang, Urumqi and in Kashgar Prefecture even higher than in 2017. This stands in marked contrast to gradually declining educational spending increases for all of China.

Figure 8: Source: Regional finance departments’ final accounts.

A specific example that exemplifies the timing and focus of educational spending in Xinjiang’s Uyghur regions is Shache County in Kashgar Prefecture, where total education spending increases in 2017 were largely related to a doubling in preschool spending. In 2018, total education spending increased by an even higher margin. While preschool spending remained high, spending increases were mainly driven by rising expenditures on primary and especially secondary schooling.

Educational level / year Spending Changes
2016 2017 2018 2017 2018
Education (total) 196,382 222,985 267,377 13.5% 19.9%
Pre-school 28,036 56,075 58,768 100.0% 4.8%
Primary school 83,643 85,918 101,788 2.7% 18.5%
Middle school 55,477 50,240 61,634 -9.4% 22.7%
High school 18,548 19,807 24,278 6.8% 22.6%

Table 3: Shache County (Kashgar Prefecture). Source: Finance department final accounts.

3.4 Public Boarding Education: Full-Immersion Chinese Language Environments

For the state, boarding schools provide the ideal context for a sustained cultural re-engineering of minority societies. There, the next generation is confined into highly controllable environments where the state can determine the rules, and punish those who fail to conform.

In March 2019, an article published by Bitter Winter detailed the government’s indoctrination and assimilation methods in Xinjiang’s public schools. To quote:

“For Uyghur teachers, as soon as they enter a public place, they must speak Mandarin,” Mr. Liu [who is a teacher at that school] said. “They have to teach classes in Mandarin, even if their pronunciation is bad. … All teachers and students must use Mandarin to communicate,” Liu continued. “The school’s leaders conduct inspections of students regularly. If a student uses the Uyghur language to answer a question in class, the teacher’s wages will be deducted.”[119]

Below, the accuracy of these statements will be confirmed by analyzing relevant government documents.

As with the establishment of preschools or the expansion of boarding facilities, the vigorous promotion of Chinese language education in Xinjiang was drastically accelerated during the course of the re-education campaign. The original deadline for the implementation of full Chinese language education, set in May 2017, was by 2020.[120] However, a June 2018 notice issued by the Xinjiang Education Department proclaimed that by the end of that year, the region’s 2.94 million students in mandatory education (grade 1-9) were expected to have a fully Chinese-medium language education.[121] In October 2018, Ili Kazakh Autonomous Prefecture, which is also home to many Uyghurs, proudly announced that it had fully implemented this all-Chinese language education directive, and was busily recruiting more urgently needed Chinese teachers from all provinces in the nation.[122]

Notably, the Chinese language education goal was to be achieved in tandem with the universal establishment of boarding  facilities, of mixed Han-minority classes (民汉合校混合编班), and of combining primary and middle schools in one facility or compound.[123]  The link between mixed classes and the promotion of universal Chinese language education is readily evident. To quote a related news article from late 2017 that praised this approach:

“On school campuses [that follow the] mixed-class education model, there is only one kind of sound, that of the national language; only one type of curriculum, which is the national curriculum…. The campus atmosphere is full of unity, harmony…”[124]

The expansion of preschool enrolment, primary and secondary school integration, mixed classrooms and increasingly universal boarding facilities jointly work in unison to beget immersive and homogenized educational environments where the dominant and only valid language of communication is Mandarin Chinese. As put in a statement by Kashgar University in October 2017:

“Whether in class or outside of class, or in society, in our daily lives, among teachers, between teachers and students, between students or between close friends, all should speak Mandarin.”[125]

Conversely, teachers and others in society who are in any way against the study of Chinese, who do not speak Chinese, or who are not happy about Chinese language education are “double-faced” and “tempted by extremist religious ideology”.[126] These are the “scum of the Chinese nation and the common enemy of all ethnic groups.” Another document notes that cadres and teachers who only support the Chinese language on campus policy “on the surface”, but in reality harbor conflicting views, must be “sternly investigated and dealt with”.[127]

How are schools supposed to promote an immersive Chinese language environment even when most of their students speak a different native language? Two school documents, both issued in October 2017, indicate what has likely become widespread practice by now, and both confirm the statements made by Bitter Winter informants.

One of them is from Tuokezhake Township Central Primary School (托克扎克镇中心小学) in Shufu County (Kashgar), a setting where students would be nearly 100 percent Uyghur and where academic Chinese language skills are typically poor.[128] A separate media article highlighted and praised this institution as the future of education for all of Xinjiang. [129]

The school’s document titled “Implementation Plan for a Completely Chinese-Speaking School Environment” (校园内全部讲国语实施方案) stipulates that all classes except for Uyghur language classes must be entirely taught in Chinese, and within the confines of the entire school, Chinese can be the only language for school management and teaching activities. Even teachers who teach the Uyghur language class are “encouraged” to use Chinese for the preparation and teaching of the class, and for any school management related activities. If at any time a teacher or student uses Uyghur, this must immediately be reported and dealt with as if it were a “serious teaching incident” (严重教学事故). If it happens a second time, the teacher’s annual work appraisal cannot be rated as “excellent”, which typically results in a loss of performance-related benefits or bonuses. In addition, there must be a staff meeting with a written report on the incident. Even worse, if any teaching or management staff at the school do not immediately report any such incidents, they receive the same punishment as the perpetrators.

In order to encourage students to spend more time studying Chinese at home (presumably during weekends or holidays), teachers must ensure that they finish their Uyghur language homework at school. They cannot bring any of it back home. Teachers are encouraged to use any available spare or rest time to conduct additional Chinese language learning activities with students. After school and until 8pm in the evening, the school plays Chinese language recordings. Before the morning reading session, during lunch time and after classes end, the school plays red songs, the Qing Dynasty’s Di Zi Gui (standards for being a good pupil), the Three Character Classics, Tang poetry and other pieces to “create a good atmosphere for learning Mandarin.”

A very similar document titled “Implementation Plan on Fully promoting National Language Education” (全面推行国语教学实施方案) and issued by an unknown primary school in a Uyghur[130] region, likewise requires teachers and students to only use Chinese at any given time while being on the school grounds. It also mandates evening Chinese language self-study times between 9-11pm and Chinese language examinations every Sunday, with the results being submitted to the school’s research office. In addition, dormitory heads are required to regularly test a rolling sample of students for their Chinese skills so that over time the entire student body is evaluated.

This document brims with draconian punishments. For teachers, these include a 100 point deduction for a given month’s performance evaluation for the first violation of language regulations, a notification of criticism broadcast to the entire school for a second violation, and a 30 percent reduction of performance pay (绩效工资) for a period of three months for a third violation.[131] Each time that students fail to use Chinese, they lose performance evaluation points, and the class leader loses 10 points. This use of collective punishment is likely highly efficient in disciplining Uyghur students to only speak Chinese on campus.

Even though there is no direct evidence that the specific policies applied in these two schools are being implemented throughout Xinjiang’s ethnic minority regions, it is a fact that the language of educational policy documents and reports becomes increasingly specific and draconian as they move from regional government to prefecture to county to grassroots (i.e. school) levels. The fact that at least one of these two schools was praised as a model for the entire region makes it likely that at least some, if not most or all of its regulations and punishments, have or are being implemented throughout the XUAR. There certainly is evidence that exclusively Chinese language school campus environments are being promoted in other minority regions in Xinjiang, for example in Makit County (Kashgar Prefecture).[132]

The relentless pressure that is being placed on students and teachers to forgo the use of their native minority languages and to rapidly acquire fluency in Chinese must be devastating. All in all, boarding school campuses appear to be highly coercive and tightly-controlled environments that may well be rendering the daily lives of those within them comparable to some of Xinjiang’s internment camps.

3.5 The Mandatory Securitization of All Educational Facilities

What makes Xinjiang’s boarding school environments even more akin to internment facilities is the large number of security requirements and regulations that apply to all of the region’s public and private education facilities. While some basic security measures would seem appropriate in order to prevent isolated incidents such as knife attacks, the extreme securitization of educational facilities in Xinjiang just in time for the start of the re-education campaign is astounding to say the least.

In March 2017, the  XUAR government issued a “Special Work Plan for Protecting Schools and Securing Campuses” (护校安园专项工作方案), which mandates the installation of extensive security features and the recruitment of related staff for all schools from preschool to high school levels by the end of January 2018.[133] Notably, this regulation was not only issued right before the initiation of the re-education campaign, but also in time for the region’s concerted drive to construct thousands of new preschools.

This document mandates that all educational facilities are to be secured through “hard isolation closed management measures” (硬隔离封闭式管理措施) that prevent unauthorized persons and vehicles from forcefully entering the facilities. Besides high surrounding walls with additional anti-climbing barriers (防攀爬设施) such as barbwire or electric fences, this specifically includes anti-collision barriers (防冲撞设施) and other types of vehicle control barriers such as rising bollards (升降柱). Reports by western media outlets that took photos of educational facilities do show these security features.[134] The installation of security features such as anti-collision barriers even in preschools and primary schools is also reflected in numerous related public procurement bids.[135]

The work plan further mandates that schools must have “no blind spot” (无死角) full coverage video surveillance systems that are linked to the public security organs, and one-button alarm systems that alert the police. In the case of incident, security forces must be capable of reaching the school within one minute. Educational facilities are encouraged to explore the possibility of establishing a convenience police station or campus police office. With the exception of private parking lots, vehicles are barred from stopping within 100 meters of school facilities and within up to 200 meters of school entrance and exit points. Each school campus must have at least four guards, especially for the area where people drop off their children by car. Persons without “firm political views” (政治立场不坚定) or those with “strong religious thoughts” (宗教思想浓厚人员) must “resolutely be prevented” from being employed as security staff.

Besides not having watchtowers, the security measures that Xinjiang mandates for every single educational institution, even preschools, would seem to rival those of its internment camps.

The 2018 work plan published by a township primary school in Kashgar Prefecture, but apparently issued by the township government for all local schools, reflects securitization, surveillance and disciplinary priorities.[136] As the very first item on the plan, the school states that it will “resolutely and thoroughly implement the overall goal, continuing to strengthen the establishment of the three-tiered defense lines”. The plan calls for the installation of a full-coverage surveillance system in the school’s preschool section and reminds school leaders that the township inspection team will visit each school once a week on a random day in order to inspect the guard duty schedule and the security installations. Any failures to meet the required security standards will attract “severe punishment”.

The phrase “three-tiered defense lines” (三防 or 三道防线) is further explained in a remarkable article that details the multi-tiered security systems that Karamay City installed in all of its 41 preschools.[137] Based on a 360-degree intelligent security network, the system combines intrusion barriers, perimeter alarms, interior and exterior surveillance systems, and a police alert system. The first defense line consists of entrance and exit controls, a perimeter alarm system linked to the police, a 100 percent coverage perimeter video surveillance system, and doors with secure access control systems. The second defense line consists of a full coverage surveillance system for the courtyard and interiors of the preschool. The third defense line is represented by a computerized security patrol management system equipped with a one-button alarm that supervises and enforces the seamless and perfect operation of the campus security patrols.

The preschools’ intrusion detection system can distinguish human intruders from intruding animals based on electromagnetic wave technology. All of these 41 preschools have 4-layered 10,000 Volt electric fences atop their high walls, totaling 26,600 meters in length, and segmented into 176 “defense zones”. Via 485 network hubs, these zones are connected to a fully automated central monitoring and management platform that can interact with other security systems in real time.

Even if not all of Xinjiang’s educational facilities have implemented the same level of security as has Karamay, numerous public bid documents confirm that public schools around the region have procured similar security features. For example, a middle school with attached preschool in Alashankou City issued a 579,000 RMB bid for an electric fence with automatic police alarm feature, an access control system, and a video surveillance system.[138] Bügür (Luntai) County, a Uyghur majority area in Bayingol Mongol Autonomous Prefecture, likewise ordered an electric fence for its no.2 preschool.[139] The Chinese company Safeway System (安天下) shows numerous Xinjiang preschools and primary schools with airport-style X-ray scanning units at their entrances.[140]

Large schools spend more money on security. Xinjiang’s Ruizhiheng Electronics and Technology Company (新疆睿智恒电子科技有限公司) sells highly sophisticated electronic surveillance systems, especially developed for entire school campuses, that monitor teaching buildings, dormitories and administrative complexes.[141] A middle school in Xinjiang’s XPCC spent 2.1 million RMB on a video surveillance system, more than several re-education camps previously examined by the author have spent on such systems.[142] In December 2017, Xinjiang placed an “urgent” order for 19 “peaceful campus” (平安校园) systems, totaling 104.1 million RMB or 5.5 million RMB per system. “Peaceful campus” systems are the educational equivalent of the Orwellian “peaceful city” systems that provide comprehensive monitoring and surveillance capabilities for public security purposes.[143] The “peaceful campus intelligent campus” system purchased by Xinjiang’s Teacher College (新疆师范高等专科学校) cost a stunning 16.3 million RMB – roughly equivalent to the price tag of an entire medium-sized re-education camp.[144]

There is no public information on any past Uyghur militant attacks or kidnapping attempts in regards to Xinjiang’s public schools, and no general awareness of such an event in some of the Uyghur exile circles that the author interacted with. Also, even though violence can take on a highly irrational dynamic of its own, schools full of young Uyghur children do not seem to be obvious targets for violent attacks perpetrated by other Uyghurs for the sake of harming the Chinese state. Rather, such attacks have either targeted groups of Han people, or government facilities such as police stations. Even though the state might be seeking to protect the thousands of Han teachers who have been recruited to teach Chinese in Xinjiang’s Uyghur regions, the cost and effort put into securing these schools seems quite extreme.

Why is Xinjiang putting such incredible emphasis, pressure and funding into securing its educational facilities? Why was the plan that mandated this securitization issued in the exact same month that saw the issuance of Xinjiang’s de-extremification ordinance and the initiation of the re-education campaign? Did the state foresee that large numbers of children would become separated from their parents, and seek to preempt any possibility on the part of Uyghur parents, relatives or community members to forcefully recover their children? Even worse, has intergenerational separation been a premeditated government strategy, perhaps as a deliberate component of the re-education campaign? Is it being deployed as the ultimate weapon of the state to “break their lineage, break their roots, break their connections, and break their origins”?[145]

3.6 Conclusions

Within the past two years, Xinjiang’s public education infrastructure has undergone a remarkable and rapid transformation. The massive expansion of educational floor space, boarding capacities and the mandatory securitization of school campuses must be understood in light of the region’s entire securitization and cultural assimilation strategy. Between preschool and high school, students face 15 or more years of intensive, state-controlled and highly coercive Chinese language education and immersion, along with political indoctrination and psychological correction, depending on the age at which they are integrated into the education system. In this regard, Xinjiang’s schools have become like the colonial boarding schools used by the United States, Canada or Australia, to assimilate native ethnic populations.

More specifically, Xinjiang’s education system is being weaponized in the state’s so-called “war on terror” against any form of Uyghur resistance.[146] Figure 9 shows that between 2015 and 2017, its spending on domestic security rapidly increased and nearly caught up with education expenditures. However, in 2018, the situation changed. Domestic security spending decreased by 2.2 percent, while education spending soared by 14 percent.

Figure 9: Source: XUAR finance department final accounts.

It is important to understand that as the focus of Xinjiang’s security crackdown has shifted from public policing to internment, social control and the systematic assimilation of the next generation, its spending priorities are tilting more towards educational as opposed to public security infrastructure.

4 Xinjiang’s Drive to Establish Comprehensive and Centralized Care Facilities Outside the Education System

Reports by AP, Bitter Winter and RFA all indicate that the state considers children of detained parents to be in some kind of “special needs” situation. According to reports, some of them are treated like orphans and apparently even placed in orphanages or other types of shelters or centers for special needs children. The advantage of the state doing so is that the practice of sheltering minors without adequate guardianship or care easily fits into pre-existing categories, defined by the central government.

China, and specifically Xinjiang, have different categories of needy children. Besides outright orphans (孤儿), there are:

  • Left-behind children (留守儿童), typically defined in state documentation as caused by either both parents being away for work reasons, or one parent being away and the other unable to take care of the children.
  • Street children (流浪少年儿), who are without proper guardianship.
  • Children from “poor families” (贫困家庭), whose situation is impacted by difficult economic circumstances.

All of the above types of children can be referred to as “children in difficult circumstances” (困境儿童) or “children in especially difficult circumstances” (特殊困境儿童). By the end of 2017, Xinjiang officially had 80,933 orphaned, disabled or other children in difficult circumstances in centralized state care facilities.[147]

In October 2016, the XUAR government set the goal to centralize all elderly and orphan care in central state institutions by 2020.[148] A related implementation directive, published in February 2017, noted that the region had 23,000 orphans, only 4,900 or 20 percent of them in “centralized adoptive care” (集中收养).[149] The centralization process was scheduled to start in January 2018 and to be completed by December 2020. Just as with the educational centralization and boarding policy, this will make it easier for the state to control and influence the orphans’ lives, to keep them locked into secured compounds, and to easily increase orphanage capacity if required. Already in September 2016, one month after Chen Quanguo assumed power, Urumqi commissioned a bid for a massive children’s welfare center (儿童福利院), 28,000sqm in size, and designed to shelter orphans, street children and other children from special needs backgrounds.[150] The cost of the facility was 134 million RMB, over one third of Xinjiang’s entire spending on children’s welfare between 2011 and 2015.

In November 2016, the XUAR government issued a document outlining the region’s implementation of the central government mandate to organize systematic, centralized care for all rural “left-behind children” (农村留守儿童) by 2020.[151] This included the establishment of a computerized left-behind children information system in order to achieve complete reporting. However, as with the preschool construction drive, the state decided to drastically accelerate this policy. In June 2017, shortly after the begin of the internment campaign, the State Council Information Office (SCIO) in Beijing published a notice that required all left-behind children in Xinjiang to be in care by the end of that year.[152] The notice further called for a comprehensive information system and an early intervention mechanism. Children with significant behavioral issues were to be subjected to “behavioral correction”. As discussed in section 2.2, a sample of local government rosters pertaining to left-behind children indicates that many or perhaps even the vast majority of them are in a left-behind state because both parents are in some form of internment.

In December 2017, Korla City commissioned a “rural left-behind children information database” (农村留守儿童信息数据库), which, together with several other investments, was budgeted at 625 million RMB.[153] The announcement proclaimed that:

“We will comprehensively establish a care and protection system for rural left-behind children, and effectively operate the rural left-behind children’s rescue and protection mechanisms, such as mandatory reporting, emergency response mechanisms, assessment assistance, and guardianship intervention.”

Also in late 2017, Kashgar City published an implementation directive of the regional 2016 mandate, pointing out that the left-behind children phenomenon was becoming a threat to social stability.[154] To quote:

“Some rural left-behind children have psychological barriers due to a lack of spiritual care, are dropping out of school, … are even becoming problem children … due to a lack of guardianship. To this end, all towns, villages and relevant departments must stand at the height of maintaining social stability and long-term stability, fully understanding the importance and urgency of strengthening the care and protection of left-behind children in rural areas.”

As solutions to the situation, the document mandates the promotion of boarding schools and the construction of “rural left-behind children’s rescue and protection mechanisms” (农村留守儿童救助保护机制). Specifically, it calls for the construction of “street children’s rescue and protection centers” (流浪未成年人救助保护中心).

From mid-2018, nearly every county and administrative unit in Xinjiang commissioned “rescue, care and protection centers for children in especially difficult circumstances” (困境儿童救助关爱保护中心). Typically, they cost around 8-9 million RMB and come with 4,000-5,000sqm of floor space.[155] In some of the southern Uyghur majority regions, they are larger. For example, Bachu County’s rescue center’s dormitory alone has a floor space of 3,000sqm.[156] Kashgar Prefecture overall commissioned the construction of 30 such centers even though it only has 12 counties.[157]

Many regions in China saw the establishment of such centers throughout 2018 and the first half of 2019, including neighboring regions such as Inner Mongolia or Gansu. However, the intensity, scale and expenditures of Xinjiang’s so-called “rescue, care and protection centers for children in especially difficult circumstances” (困境儿童救助关爱保护中心) has been absolutely unprecedented. Notably, Xinjiang’s rescue centers are larger and have a higher boarding capacity than those commissioned in other regions. The combined capacity of all these centers commissioned by different counties in Xinjiang appears to exceed that of the region’s orphanage system. This may well relate to the fact that the region’s internment campaign continued apace throughout 2018. Detention facilities expanded massively that year.[158] Xinjiang’s rescue centers are often established in the same context as its welfare centers. The Financial Times article cited in section 1.1 cited informants as saying that three children in Kashgar City aged 2, 7 and 9 years, with both parents in detention, were housed in the local welfare center.[159]

However, children in difficult circumstances and even orphans are also regularly housed in public boarding schools. For example, the Hotan County Kindness preschool also shelters children from poor families and orphans. The facility has 137 students, eight teachers and 28 caretakers (保育员) who look after them 24 hours a day.[160] This means that the facility has nearly four times more caretakers than teachers, and only 4.9 children per caretaker. The Hotan City Kindness preschool and other “kindness” preschools in Hotan’s rural regions also house children from such backgrounds, as well as “left-behind children”.[161] They are referred to as “full care” (全托) preschools where children receive room and board. Kashgar City, Korla City, Aksu City, Changji City and Hoboksar County have “little angel” preschools that likewise house children from “poor families”, “left-behind children”, and orphans.[162]

While many uncared for children are more than likely sheltered within the education system, Xinjiang’s expanding network of children’s rescue centers gives the state a growing range of options for sheltering “special needs” children of different types. For example, those with both parents in detention who develop particularly strong psychological issues could be transferred from regular public schools to one of the rescue centers. The latter facilities could become places that shelter particularly challenging children, keeping them away from the worrying eyes of their communities and removing them from the tightly-run public education system. This would make it easier for the state to hide “problem children” and to mask the consequences of the internment campaign and the general security crackdown.

5 Conclusions

This article has presented comprehensive evidence in at least four key areas pertaining to the systematic, state-initiated separation of parents and children in Xinjiang. When taken together and analyzed in light of the region’s other policies and practices, these areas provide strong evidence for a coordinated state campaign to promote different forms of intergenerational separation.

  1. Just in time for the re-education campaign, Xinjiang urgently and systematically establishment a vast network of care and boarding infrastructure capable of offering full-time care for children of nearly all ages. Both the construction of preschools and the provision of a comprehensive care network for left-behind children were originally scheduled to be completed by 2020. However, after Chen Quanguo became Party secretary, this deadline was moved forward to late 2017.
  2. These care facilities are heavily secured, tightly controlled, and promote systematic and intensive linguistic and cultural assimilation.
  3. State propaganda articles extol intergenerational separation as being highly beneficial for the children.
  4. The government has put detailed plans in place for handling the socio-economic fallout and the negative psychological impact of intergenerational separation. The public education system is required to closely monitor and care for children of interned parents in order to preempt potential incidents. Relatives and family members of detainees are likewise closely monitored. Additionally, the state set up a database system with mandatory reporting requirements that covers all children with various special needs. Related documents indicate that many of the left-behind or children in special circumstances have one or both parents in detention.

Finally, available informant accounts and western media reports consistently corroborate the findings listed above inasmuch as these accounts pertain to them.

By May 2019, the Xinjiang government is literally able to “parent” at least tens of thousands, if not a hundred thousand or more children. Even after the state releases parents, children can remain in full-time care or boarding facilities at least during the work week, meaning that the state gets more time to influence the next generation than the parents do. This arguably began on a small scale with the establishment of the Xinjiang boarding classes in eastern Chinese cities in 2000, but is now being expanded to a truly unprecedented scale.[163] Generally, the all-encompassing nature of Xinjiang’s intergenerational separation practices affect not only those in internment, but any parent. With the expansion of sophisticated care and boarding facilities, students can remain separated from parents for entire work weeks and possibly longer. This is almost certainly not coincidental, but a deliberate part of “breaking roots” and changing Turkic minority societies through coercive social re-engineering.

The evidence suggests that Xinjiang’s securitization strategy is gradually moving from mere policing towards the development of long-term social control mechanisms. Related research will need to adjust to this shifting focus. While the police force and its infrastructure will remain, the real battle over China’s grasp over this restive region is set to be fought in the hearts and minds of its minors. Police officers can arrest or intimidate people and surveillance systems can track their movements, keeping a recalcitrant population in oppression. A weaponized education system, however, enables the state to control an ethnic group from within their core and their roots rather than merely from the top down.

China’s state-driven campaign of intergenerational separation in Xinjiang is a clear indication that its long-term goal in Xinjiang is a targeted cultural genocide, designed to completely alter and align the hearts and minds of the next generation with the Communist Party ideology. This appears to be part of Beijing’s long-term strategy for permanently securing its hold over this core region of the Belt and Road Initiative (BRI).

Already, China is using Xinjiang as a laboratory for advanced surveillance technologies and predictive policing methods. If the state considers intergenerational separation to be a successful method for suppressing the transmission of religious and cultural identity, it is possible that it might adapt and apply this approach in other parts of China. The coercive social re-engineering methods that are currently being trialed in Xinjiang may then become a model for other Chinese regions, or even for other authoritarian states along the BRI.

Appendix A

The per capita calculations for preschool enrolment were performed as follows. Each region’s population aged 0-5 years was calculated by subtracting the population aged 6 and up from the total population count. This was done for the 2010 census and the 2015 mini-census (which measured 1 percent of the population). The estimated population aged 0-5 years for each region and year between 2012 and 2018 was then calculated based on the two census figures and each year’s total population figure as published in regional socio-economic development reports. For the years 2012 to 2014, this was done by weighting and combining the 2010 and 2015 population shares. Each region’s 2012 total population figure was multiplied by the percentage of the 0-5 year cohort of the entire population, both for the 2010 census and for the 2015 census. Each annual figure is then a combination of weighing each of these two figures. For example, for the 2012 estimate for Hotan Prefecture, Hotan’s total population for 2012 (of 2,123,400) was multiplied by 11.0 percent (2010 census share of 0-5 year olds among the total population), which equals 223,621, and by 16.1 percent (the 2015 census share), which equals 375,162. Then, each of these figures was weighted. For 2012, which is closer in time to 2010 than to 2015, the 2010 figure of 223,621 was multiplied by 60 percent, and the 2015 figure of 375,162 was multiplied by 40 percent. For 2014, it was the reverse (2010 figure weighted at 40 percent, 2015 figure weighted at 60 percent). For 2015 and later, each year’s total population figure was multiplied with the 2015 census percentage of the 0-5 year cohort of the entire population. While this is necessarily an approximation of each year’s actual size of the 0-5 year cohort for which there is no census data, it is far more accurate than to calculate enrolment figures per capita of the entire population.

Dr. Adrian Zenz is an independent researcher and Ph.D. supervisor at the European School of Culture and Theology, Korntal, Germany. He specializes in China’s ethnic minority policy, minority education systems, public recruitment (especially teacher and police/security-related recruitment), public bid documentation, domestic security budgets and securitization practices in China’s Tibetan regions and Xinjiang. He has authored Tibetanness under Threat (Global Oriental, 2013) and co-edited Mapping Amdo: Dynamics of Change (Prague: Oriental Institute, 2017). JPR Status: Working Paper.

[1] The author gratefully acknowledges the helpful comments made by Darren Byler and Timothy Grose on earlier drafts of this article, the help of an anonymous academic with checking the Chinese-English translations, and the amazing flexibility and helpful editing done by JPR publisher Anders Corr. Part of this research was funded by the British Broadcasting Corporation.



[4] or

[5] Document from late 2018, discovered from publicly accessible information by a contact working outside of China. The original document and additional details of this material can be verified by third sources at the sole discretion of the author on a confidential, case-by-case basis. The author chose not to make this information public at this point in time in order to protect the integrity of ongoing research.

[6],, or

[7] or

[8] See e.g. and




[12] Bitter Winter typically publishes first-hand accounts of single, anonymous sources. While this makes their claims harder to verify and means that they often do not meet typical media reporting standards, it is also true that their approach frequently yields insightful information that is hard to obtain in any other way. This article draws on Bitter Winter reporting in order to contrast the author’s findings with the statements of Bitter Winter informants. In many instances, they confirm each other.








[20] or,

[21] See section 4.


[23] or, or or, or or, or or, or or

[24], see also

[25] or

[26] or


[28] or Compare or

[29] or

[30] or

[31] or

[32] Source: Date accessed: May 31, 2019.

[33] The full heading of the form is “需要托养人员调查汇总表”. Data discovered from publicly accessible information by a contact working outside of China. The original document and additional details of this material can be verified by third sources at the sole discretion of the author on a confidential, case-by-case basis. The author chose not to make this information public at this point in time in order to protect the integrity of ongoing research.

[34] Source: see comments under footnote 33.

[35] Source: see comments under footnote 33.

[36] Ch. 新疆维吾尔自治区青年发展规划(2017—2020年).

[37] An alternative link that is accessible outside of China is or


[39] or,, or

[40] or

[41] or, and or

[42] or, or


[43] or

[44] or or

[45] or, see also or

[46] This document can be found on numerous document storage websites, for example: See also

[47] Ch. ***县2018年春季开学暨学校安全风险防控专项督查工作情况汇报.

[48] or

[49] and

[50] or For evidence of the use and meaning of the term 教培中心, see for example or

[51] The original phrase is concealed in the document as “夫妻双方在◇◇◇◇”.

[52] The original phrase is concealed in the document as “教培局:负责各◇◇◇◇学员参保缴费工作”. Compare also this document, where this fact is confirmed, and the phrase “教培中心” is not concealed but spelled out: or

[53] or, or or

[54] Raw data: Compare main page at

[55] For a definition of三类人员, see for example or; also or

[56] or


[58] or, or or

[59] or

[60] See e.g.,,

[61] or



[64] or


[66] See related discourses in regards to the Xinjiang class. See

[67] or

[68] or



[71] or

[72] For example or


[74] or

[75] Hotan Prefecture’s population is over 90 percent Uyghur.

[76] The “XUAR Uyghur prefectures” shown in figures 2 and 3 are Hotan, Kashgar and Aksu, while the “XUAR non-Uyghur prefectures” are Urumqi, Changji and Bortala. Both samples only include regions where preschool enrolment data for 2018 had been published at the time of this publication.




[79] or

[80] The calculation method for the per capita figures is discussed in the appendix.

[81] For a discussion of this phenomenon, especially in the mid-1990s, see Joanne Smith Finley, 2013. The Art of Symbolic Resistance. Brill: Leiden.

[82] and


[84] or

[85] or

[86] See or Other examples include or



[89] For example or

[90] or

[91] or

[92] or

[93] or


[95] or

[96] or

[97] or

[98] or

[99] Sources: and Figures for Liaoning and Shandong are from 2017.

[100] Examples include:,,,

[101] See,,

[102] or

[103] or


[105] or

[106] or

[107] or



[111] or



[114] or

[115] or

[116] or

[117] or

[118] This may also be part be explained by the boarding schools for Uyghur and other minority students in inland China (内地班)  and within Xinjiang (内初班). Compare (notes on Chapter 9).


[120] or

[121] or

[122] or

[123] or

[124] or

[125] or

[126] or

[127] or


[129] or

[130] While no location is given, the names of the schoolmaster, school Party secretary and others who signed the document are all in Uyghur.


[132] or



[135] For example or


[137]  and

[138] or

[139] or

[140] or

[141] or

[142] Compare the author’s previous publication on re-education camps and their security features at

[143] or

[144] or The cost of some re-education facility construction bids can be found in the author’s publication on re-education camps at


[146] See for example


[147] or

[148] or

[149] or

[150] or

[151] or

[152] or

[153] or

[154] or

[155] Examples include: or, or, or, or, or, or, or,, or

[156] or

[157] or

[158] See e.g.,,


[160] or

Also or


[162] or, or

[163] See e.g. Grose, Timothy, 2015. ” (Re)Embracing Islam in Neidi: the ‘Xinjiang Class’ and the dynamics of Uyghur ethno-national identity”, Journal of Contemporary China, Vol. 24, No. 91, 101–118,