Xinjiang – what do we know, how and why

This is not my Blog. It is put here to provide on online link for wider dissemination. The authors have written one of the calmest, detailed and most dispassionate investigations on Xinjiang that I have read. The information they provide is not covered at all in the press or broadcast media here. This is disturbing, because all the journalists have to do is read it. The complete blackout of information indicates that questioning and debate is not welcome – and nor is anyone who tries to do it. Please read and make up your own mind.

Events in Xinjiang, and how they are interpreted, have become one of the most prominent issues in public discussions in the West about China.

What we know about these events, how we know it, and why, should therefore be key questions for debate and discussion among researchers, policy makers and the media. This paper suggests that, while there have been plenty of publications and media coverage on Xinjiang, there has been little critical examination of many of the claims. Instead, the topic of Xinjiang has become highly emotive and political.

A number of claims about the Chinese state’s policies in Xinjiang have become received wisdom in the West. These were summed up in the US Department of State’s statement of 19 January 2021, issued on Secretary of State Mike Pompeo’s last day in office: ‘The People’s Republic of China (PRC) … has committed crimes against humanity against the predominantly Muslim Uyhgurs and other members of ethnic and
religious minority groups in Xinjiang, [including] the arbitrary imprisonment … of more than one million civilians, forced sterilization, torture of a large number of those arbitrarily detained, forced labor, and the imposition of draconian restrictions on freedom of religion or belief, freedom of expression, and freedom of movement.’ Pompeo stated that this campaign had ‘dramatically escalated’ since 2017 and went on to
say that he concluded the PRC had ‘committed genocide against the predominantly Muslim Uyhgurs and other members of ethnic and religious minority groups in Xinjiang’.

Major stories in leading media outlets have delivered similar conclusions in varying levels of
detail. (4)

This paper looks at the publicly-available evidence for the received wisdom on Xinjiang in the West. It comments on what we know about events in Xinjiang, how we know it, and why this issue has become so prominent. The paper is not based on any new empirical research, but a reading of many of the key publications on which claims about Xinjiang are based. (5)

It concludes that the evidence presented so far is limited and that most of the claims about developments in Xinjiang should therefore be treated with a high degree of caution. The authors’ goal is critical questioning rather than definitive answers.

The paper begins with a brief discussion of the background, based on academic literature
on Xinjiang (more plentiful over recent years than the literature on any other region of
China). It then discusses the three main criticisms which have emerged since 2018, relating
to ‘arbitrary imprisonment’ or ‘internment camps’, ‘forced labor’ and ‘forced sterilization’.
The paper then addresses the question of whether the evidence supports a determination
of ‘genocide’, and how we might understand the Chinese state’s policy in Xinjiang. The final
section of the paper addresses discussions of Xinjiang in the West.

Xinjiang today is formally known as the Xinjiang Uyghur Autonomous Region, established in 1955.

It accounts for around one sixth of the PRC’s territory and borders eight of its neighbors. Today its population of around 25 million is around half Uyghur and 36 percent Han, 6 with a number of other minority groups which – like the Uyghur – are Muslim majority. It has long been a strategically important region to the PRC, and developments there were influenced from the early parts of the 20th century to the 1960s by relations between China and the Soviet Union. Following the latter’s demise in the 1990s Xinjiang saw a new phase of contestation, (7) ranging from resistance to Chinese modernization to
desires for a Uyghur homeland (akin to the new Stans of Central Asia). Some violence has
been part of this since the 1990s.

The historical roots of this are complicated, but the recent phase of events can be traced back to July 2009, when protests broke out in the regional capital, Urumqi, with violence committed by both Uyghur and Han populations, leaving around 200 dead. The following years saw both a growing state security presence in the region, and a series of violent incidents caused by extremists, characterized by the Chinese government as ‘terrorism’. (8)

One of the most serious was an attack in 2014, in which ‘at least 39’ were killed. Violence spread briefly beyond Xinjiang, with a knife attack at Kunming railway station killing 29 and a jeep ramming Tiananmen Square in 2013. (9)

The reasons for these developments are somewhat disputed in the academic literature on Xinjiang, and the purpose of this paper is not to adjudicate between them. Some blame the violence on state policies, arguing that it is understandable resistance to efforts to constrain Uyghur culture or religious freedom. Others highlight a growth in extremism and links to Islamist militants elsewhere. For a long time, Chinese state responses were based mainly on the idea that economic and social development could reduce tensions and lead to more acceptance of the Chinese state, though with limited reflection on the ways that capital-intensive and state-driven development in Xinjiang might also exacerbate discontent. (10) Alongside this, though, a clear state response to the earlier unrest was a rapid investment in security infrastructure in Xinjiang. (11)

Another important response to the events of 2009 was a growing debate in Chinese intellectual and policy circles about whether existing ‘ethnic minority’ policies remained fit for purpose in Xinjiang. A number of influential Chinese academics began to advocate ‘second generation’ ethnic policies, essentially adopting a ‘melting pot’ model instead of the differentiated institutional treatment of ethnic minorities in the form of preferential policies, for example, in admissions to universities or in family planning. These debates continued through the 2010s and have been well covered in the academic literature, most of which concludes that formal change in the policy framework has been limited but
practice may have shifted in the direction of assimilation. (12)

From late 2013, Xinjiang was designated by Beijing as a ‘core area’ on the Silk Road Economic Belt connecting China westwards across the Eurasian continent. This geostrategic designation has added another layer to the multi-faceted context within which Western discussion of Xinjiang began to grow from late 2017 onwards. (13)

Internment camps or vocational education?
The issue which really began to focus Western attention was the claim that large numbers of Uyghurs were being detained for ‘political re-education’. Many reports and articles on Xinjiang have since referred to ‘reports’ about mass detention, though often with minimal sources. An early report gave some anecdotal claims, (14) while a subsequent report from Radio Free Asia in January 2018 claimed in its headline that 120,000 were being held in Kashgar (Xinjiang’s far west), though it also gave a figure of 32,000 for the whole prefecture. (15) But probably the most influential report on this issue was a paper by Adrian Zenz, published in the academic journal Central Asian Survey in fall 2018. (16) Zenz describes
extra-judicial political ‘re-education’ efforts beginning in 2013, based on a number of examples taken from media and government internet sources in Xinjiang. Zenz used these reports to build a picture of a proactive program of ‘political re-education’ targeted at people in Xinjiang who had been influenced by ‘religious extremism’. The Chinese government initially denied the reports but has since admitted to a ‘deradicalization’ program in Xinjiang. (17)

Given the dominance of this issue in subsequent claims about what is happening in Xinjiang, several points from Zenz’s article deserve further discussion. The first is to note that the length of time he says was spent by individuals in ‘political re-education’ or ‘deradicalization’ seems to vary, but in many cases it was quite short, perhaps a number of days or a matter of weeks in some cases. For example, Zenz reports that in Ghulja (Yining) county, ‘focus persons’ received from 4 days training to 20 days (for ‘the most recalcitrant “strike hard”’ detainees’), and a number of other examples he cites are also short duration.
In a minority of cases, training may have lasted much longer, or led to detention for prolonged periods of time, as has been claimed by a number of Uyghur exiles, though further systematic evidence is needed to corroborate this. A subsequent report of ‘leaked files’ published by the New York Times suggested that some Uyghurs had been taken for ‘training’, but the numbers or length of time were not clear. (18)

A second, related, point is that only part of the (Uyghur or Muslim) population is subject to this ‘re-education’. Here, we get to one of the biggest issues in the Western discussions about Xinjiang, the number of individuals who may have been (or are) in ‘re-education centers’. Zenz’s 2018 article is ambiguous and he himself says there is ‘no certainty’. The abstract simply refers to ‘untold thousands’, while the main text at one point suggests an estimate between several hundred thousand and one million, with another reference to a figure of ‘just over one million’.

The imprecision here is a natural consequence of Zenz’s methodology, which is to look for internet reports giving examples from certain counties and then to extrapolate from these. (19) In one county, Khotan (Hetian), a local official is cited as describing 5 percent of those who had been subject to extremist religious influence as ‘belong[ing] to the hardened faction’. Based on a couple of counties in southern Xinjiang (where the Uyghur population is much higher, and where violence was more prevalent in the period from 2009), Zenz basically assumes that 10 percent of the Muslim population has been through
this ‘re-education’. Based on Xinjiang’s population figures, this gives a figure in the region of one million across Xinjiang.

In spite of this ambiguity and methodological uncertainty, this one million figure has since become a cornerstone of the narrative about what is happening in Xinjiang. Others have defended the figure, such as Jessica Batke, (20) who gives as sources Zenz’s article, a report by China Human Rights Defenders (a US-based NGO) based on interviews with eight ethnic Uyghurs in different villages in southern Xinjiang, a leaked report (which came via a Uyghur exile media organisation, cited by Zenz) giving a figure of 892,000 across 68 counties in Xinjiang in spring 2018, and a report from Radio Free Asia citing four local officials who been told they had detention quotas (10 percent of the population in one village and 40
percent in another). These are hardly conclusive. And while Zenz’s report is the longest, the figure of one million is (somewhat tautologically) based on an assumption that 10 percent of the adult Muslim population is targeted, and the ‘actual’ figure for those who have undergone some re-education could well be anything. (21)

This needs to be considered with the varying lengths of ‘re-education’. The received wisdom has it that ‘Over one million Uyghurs and other Muslim peoples in … Xinjiang are in mass internment camps’ (emphasis added: note the present tense). (22) But – accepting the figures for the moment – the logic of the evidence is captured better by The Economist’s ‘Chaguan’ columnist who wrote that ‘Across Xinjiang, over a million Uyghurs have passed through such camps in recent years’. (23) We should stress that the point here is not about whether such actions are ethically acceptable, but that many of the claims about their scale appear to be seriously exaggerated, and not well based in evidence. In fact, Zenz’s article reveals no direct, verifiable evidence of mass internment (as opposed to evidence for short-term
‘training’ or public tenders for the construction of facilities), only citations of media reports (especially RFA) and anecdotal reports (including from exiles) claiming that this is taking place. Given how much Zenz’s work has been cited by the media, these claims have become circular and self-referential.

A linked question is how many such facilities (‘camps’ or ‘centers’) there are in Xinjiang. Zenz counts the number of bids for re-education related tenders, implying that this tells us something about ‘detention camps’. Other estimates of the number of ‘camps’ have come from analysis of satellite imagery, though the sources of this imagery are usually not clear. For example, Batke refers to a BBC report which cited analysis from ‘analysts from a multinational aerospace company’ judging that there were 44 facilities which ‘had a high or very high likelihood of being a “security facility”’ and data from ASPI which focuses on 28 facilities, while noting that the total number may be over 1,000. (24) It is not clear how these
judgments are made or why ‘security facilities’ or facilities with high border walls (for example) should be assumed to be ‘internment camps’ or ‘re-education camps’ and not factories or state complexes (perhaps even military or paramilitary in nature). Countering these reports, the Chinese government has published pictures of a number of the facilities claimed by ASPI as ‘internment camps’ which show that they are other commercial or public buildings.

Another question relates to the nature of the ‘camps’ or ‘centers’ themselves. Zenz cites government procurement notices for a range of items, from sanitary and catering facilities to security and surveillance equipment, which taken together do not seem conclusive. Ultimately there is little – if any – independent evidence of what happens inside them on which to base judgments about whether they are better described as ‘vocational education’, ‘deradicalization’ or ‘re-education’ centers, or something else. The Chinese government claims that they have provided education and training, equipping people with
useful life skills. The small number of exiles who have been interviewed outside China have made claims about torture and other inhumane treatment. At the moment, all the evidence is anecdotal, and all the sources have skin in the game (though the Chinese government has many more sources than those outside the country).

The most recent reports on what may have taken place inside the camps are worth looking at for what they tell us about what we know (or don’t know), how and why. On 2 February 2021, the BBC published a report claiming ‘systematic rape’ was occurring inside the ‘camps’, based on a few individual testimonies, described in the same report by Zenz as ‘authoritative’. (25) These disturbing accounts are difficult to corroborate, but – however harrowing – do not constitute evidence of a systematic problem. Towards the end of the same report, China commentator Charles Parton is quoted as saying ‘It was unlikely that Xi
or other top party officials would have directed or authorised rape or torture … [but they would] certainly be aware of it’; no evidence is offered for this statement, though its implications are incendiary. (26) The Chinese government has rejected these allegations, and sought to discredit the individuals, including by drawing attention to conflicting testimony they had given in previous interviews.

Finally, it is worth noting that the Chinese government has stated that the last ‘graduates’ of vocational training centers’ completed their training at the end of 2019. (27) This, however, does not seem to have had any impact on the Western debate about Xinjiang, and most commentators still state regularly that one million individuals ‘are interned in camps in Xinjiang’, or words to that effect.

The conclusions of this section suggest that reports of a ‘political re-education’ or ‘deradicalisation’ program in Xinjiang are plausible, but that it targets a small proportion of the (Muslim) population there and that it is difficult to confirm how much volunteerism or coercion was involved in participation in training. Vocational training has probably been offered to a greater number of people than political re-education. The number of ‘centers’ for this remains highly uncertain, as does the number of individuals who have passed through them, and the length of time spent there. But claims such as ‘one million Muslims are incarcerated’ in Xinjiang do not appear to be supported by a careful examination of the
evidence, even that presented by strong critics of the Chinese state, such as Adrian Zenz (whose 2018 article contains no direct evidence of mass internment). It is also possible that the re-education program has already been terminated.

‘Forced labor’ or employment programs?
The idea that the Chinese government’s policy has moved on from this ‘re-education’ approach to another phase was explicitly stated in a February 2020 report by ASPI which appears to be the first setting out another major claim about Xinjiang, that of systematic ‘forced labor’. The second paragraph of the executive summary states: (28) ‘The “re-education” campaign appears to be entering a new phase, as government officials now claim that all “trainees” have “graduated”. There is mounting evidence that many Uyghurs are now being forced to work in factories within Xinjiang. This report reveals that Chinese factories outside Xinjiang are also sourcing Uyghur workers under a revived, exploitative government-led labor transfer scheme. Some factories appear to be using Uyghur workers sent directly from “re education

Uyghurs working in other parts of China is not a new phenomenon, dating back almost two decades or longer (as the ASPI report itself notes); the unrest in Xinjiang in 2009 was sparked by a confrontation between Uyghur and Han workers in a factory in Guangdong. (29)

The ASPI report uses Baidu internet engine search results relating to employment or labor transfer outside Xinjiang to argue that there has been a dramatic increase in this practice from 2017. But arguing that this is encouraged or even organised by the government does not on its own constitute evidence of ‘forced labor’ or coercion. The uncertainty of the linkage here is indicated by ASPI’s own language in the statement of ‘the problem’ at the top of the report: ‘The Chinese government has facilitated the mass transfer of Uyghur and other ethnic minority citizens from … Xinjiang to factories across the country. Under conditions that strongly suggest forced labor, Uyghurs are working in factories that are in the supply chains of … well-known global brands…’ (p. 3, emphases added).

This report does not really delve into its other claim about workers ‘being forced to work in factories within Xinjiang’. That appears to be because the main goal of the report is to highlight the risks that ‘global brands’ might be complicit in the use of forced labor, and that issue (which feeds the wider ‘decoupling’ debate in the West) dominates the report. Labor issues go wider than Xinjiang. There have long been criticisms of labor conditions in China, including for many of the country’s over 250 million ‘migrant workers’. (30) The point here is that what the APSI report says about Uyghur ‘labor transfers’ and conditions in factories far from home appears similar to the situation for other migrant laborers. Fundamentally there is no particular Xinjiang angle to this story, other than the elevated attention in the West to anything relating to that region (one could write a report about labor transfer from Sichuan or Henan, for example).

A subsequent report on ‘forced labor’ focused on cotton farming in Xinjiang, one of the major industries in the region. Written by Adrian Zenz, it argues that ‘hundreds of thousands of ethnic minority laborers in Xinjiang are being forced to pick cotton by hand through a coercive state-mandated labor transfer and “poverty alleviation” scheme’. As with the ASPI report, a major aim of this report is to lobby international businesses to disconnect themselves from supply chains which might involve labor from Xinjiang (these
reports seem to have been quite successful in this, furthering the ‘decoupling’ agenda advocated by many in the Trump administration).

The evidence for this as ‘forced labor’ is the interpretation of government statements about the organization of ‘labor transfer’. A BBC package based on Zenz’s report (and visits to Xinjiang where the journalists were not able to film much or talk to many people) highlights the ambiguous nature of the claims, talking for example about ‘conditions that … appear to raise a high risk of coercion’.(31)

An earlier report by CSIS had concluded that it was ‘possible’ that minorities were being sent to pick cotton but that ‘more information [was] needed’. At best, the evidence for coercion remains weak. (Some of the media reporting has gone beyond the notion of ‘forced labor’ to claims about ‘slave labor’; even Zenz’s reports do not support that, noting that workers are paid.)

More recently, the Jamestown Foundation published another report by Zenz on labor. (32)
Zenz takes issue (in constructive terms) with some of ASPI’s conclusions, including the linkage between ‘graduates’ of the ‘camps’ and labor transfer and some of the evidence ASPI offers for coercion, as well as highlighting some problematic conclusions which can be drawn from anecdotal evidence. (33) Zenz instead argues that there are two separate programs (conflated by ASPI): cross-provincial labor transfer, which tends to be focused on rural surplus labor (in two counties he cites figures for unemployment as 43.6 percent and 71.3 percent; see pp. 18-19), claiming some of these programs are increasingly coercive;
and a different program for ‘graduates’ from the centers who are usually placed in factories close by, often specifically constructed for the purpose. A headline claim of the report is that 1.6 million rural surplus laborers across Xinjiang are ‘at risk of forced labor through the labor transfer program’, though ‘at risk’ is hardly a precise term, and the number is tautological, based on assumptions about the proportion of the population which could be categorised in this way (p. 19). Zenz’s own definition of ‘coercion’ extends to the need to be ‘obedient to factory management’ (p. 24). The survey data cited in the report on the willingness of individuals to move to other parts of China is inconclusive.

‘Forced sterilization’ or family planning?
Another major claim is that Uyghur women have been subject to ‘forced sterilization’ on a massive scale. This is based on a report written by Zenz, published by the Jamestown Foundation in June 2020.(34) The conclusions of this report appear to be rather fuzzy, with apparent inconsistencies and contradictory data points. (35) For example, one claim is that the government planned ‘to subject 80 percent of women of childbearing age in the rural southern four minority prefectures to intrusive birth prevention surgeries … with actual shares likely being much higher (sic)’ (Summary, p. 3), while later we read that ‘In 2018,
Xinjiang sterilized 1.1 percent of all married women of childbearing age’ (p. 17), and that itis ‘likely that Xinjiang authorities are engaging in the mass sterilisation of women with three or more children’ (p. 18). Zenz notes that the 2010 national census showed that 19.7% of Uyghur females had three or more children.

This last point gets to the crux of interpreting what could well be a shift in policy approach. The obvious interpretation is that the authorities are finally implementing family planning policies in Xinjiang, thereby limiting family size, an interpretation which is supported by the public statement in 2017 that Xinjiang would implement a ‘uniform family planning policy for all ethnic groups’. (36)

Whether or not the women involved want to have more than two or three children requires further research, and normative views on birth control (whether state-driven or not) also vary. But there is nothing here to suggest that family planning policies implemented in Xinjiang are specifically designed for Xinjiang or Uyghurs, or intended to stop them having any children. Neither does the implementation of this policy support the claim that about promoting Uyghur ‘assimilation into the “Chinese Nation-Race” (Zhonghua minzu)’, as claimed by the editor of Zenz’s report. (37)

As Zenz himself notes, the latest available statistics (up to 2018) show that the Uyghur population has grown, though the rate of growth (the second derivative) may have fallen more recently, while the Han population of Xinjiang declined from 2012 to 2018 (the reports discussed in the previous section suggest there is also some evidence of a reduction in Han migration into Xinjiang in the cotton sector). As has been pointed out by the Chinese government, the Uyghur population of Xinjiang has grown in absolute and relative terms over recent years, and was over half the region’s population in 2018. (38)

Another claim made based on exiles’ testimony is that Uyghur children have been forcibly removed from their families. One investigative journalist who visited Xinjiang pointed out a big increase in the number of kindergartens in Xinjiang from 2017, and claimed that she had waited outside a kindergarten for children to leave, but none had come out at the end of the day. (39)

Again, these data points are open to straightforward interpretations, that there has been a greater investment in education in Xinjiang, and that boarding is part of school life, as it is elsewhere in China (again, normative assessments of the desirability of this will vary).

Are genocide claims credible?
Discussions in policy circles about whether what was happening in Xinjiang constitutes ‘genocide’ were already taking place in early 2019 (before much of the evidence on which subsequent claims have been based emerged). More frequent public use of the term became apparent in the middle of 2020 and there was a significant uptick in discussion of the concept in early 2021, probably prompted by the US State Department’s designation of 19 January 2021 (see the Introduction to this paper).

The first prominent public mention of ‘genocide’ in relation to Xinjiang was in Zenz’s June 2020 report on ‘forced sterilizations’. This report begins with an editor’s note consisting first of a paragraph praising Zenz’s work and his credentials, then a paragraph summarising the report, followed by this: ‘Based on research in original Chinese-language source materials, Dr. Zenz presents a compelling case that the CCP party-state apparatus in Xinjiang is engaged in severe human rights violations that meet the criteria for genocide as defined by the U.N. Convention on the Prevention and Punishment of the Crime of Genocide.’

This point is not woven into the report’s analysis in any way and only reappears in its final paragraph, where Zenz raises the issue in less clear-cut terms as follows: ‘These findings raise serious concerns as to whether Beijing’s policies in Xinjiang represent, in fundamental respects, what might be characterized as a demographic campaign of genocide per the text of Section D, Article II of the U.N. Convention on the Prevention and Punishment of the Crime of Genocide.’ The lack of any detail or argumentation for this in the report raises the possibility that this point was added to the report at a late stage, perhaps at editorial suggestion.

Article II of the relevant UN Convention states that ‘genocide means any of the following acts committed with intent to destroy, in whole or in part, a national, ethnical, racial or religious group’. (40) It then lists five acts, starting with ‘killing members of the group’ (Section A). Section D (cited by Zenz) reads ‘Imposing measures intended to prevent births within the group’. It is important to note that the scope of Article II is broader than the common perception of genocide, which is generally taken to mean mass killings of a people. Indeed, there is no evidence of mass deaths of Uyghurs in Xinjiang, or any suggestion from the critics of China’s policies that this has taken place – one commentator who has aligned himself with the genocide claim explicitly states that ‘there is no evidence of a campaign to kill’. (41)

The debate over genocide in Xinjiang seems to relate to Section D and to some extent to Section E (‘Forcibly transferring children of the group to another group’). This debate has been heated but inconclusive. The State Department determination of 19 January 2021 simply says that the determination was made ‘after careful examination of the available facts’ (it also says that genocide is not just targeted at the Uyghurs but ‘other ethnic and religious minority groups in Xinjiang’, though there are no further details or evidence for this addition). According to one report, experts in the State Department were split on the issue, with lawyers arguing against the determination. (42)

In the absence of any public argumentation by the State Department, it is quite possible to interpret Pompeo’s determination as being driven more by politics than a clear-headed assessment of the
facts. Similarly, the statements by Biden administration officials supporting the designation can be explained by their desire to avoid being criticised for being ‘soft’ on China, especially when some – such as Blinken – made their comments on the genocide issue before actually taking up their posts. Other prominent commentators have hedged their bets. For example, Sean Roberts (a long-standing critic of China on Xinjiang) stops short of aligning himself with the genocide claim. (43)

Charles Parton, however, told the BBC he believed genocide was being committed in Xinjiang, though was less forthcoming when asked why that policy was being pursued. (44)

Two analytical questions are at the heart of this claim: whether there is evidence of measures targeted at Uyghurs which might qualify under Article II; and whether there is ‘intent’ on the part of policymakers to ‘destroy [the Uyghurs], in whole or in part’. The authors’ analysis of the reports discussed above indicates that – contrary to Jamestown and Zenz’s claims – there is no evidence of birth control policies specifically targeted at Uyghurs, and that China’s long-standing family planning policy continues to allow Uyghurs
to have children. Section D therefore does not appear to apply. Claims about Section E (‘forcibly transferring children’) are not justified by any systematic research the authors have seen (as noted above), though there have been a small number of anecdotal reports from Uyghur exiles that are used by some to make the argument that Section E does apply.

There is also a problem around ‘intent’. In his piece on ‘the truth’ about genocide, Parton addresses this as follows: ‘The CCP does not spell out its intent, although it comes close when Xi insists on the “Sinicization” of religion or advances his vision of a new “Zhonghua minzu” (usually translated as “Chinese nation”, but “Chinese race” is closer).’

However, there is a huge gap between the use of the term Zhonghua minzu in the PRC’s ethnic policies and in Xinjiang and ‘intent to destroy’ the Uyghurs. Zhonghua minzu is a term which has long been a feature of the PRC’s policy discourse on nationalism and ethnicity and cannot simply be reduced to ‘race’ (it has shaped discussions of Chinese national identity since the late Qing dynasty). In the PRC’s usage it encapsulates the idea of commonality among the citizens of China as a nation-state, while the categorization of 56 minzu (ethnic groups) since the 1950s has institutionalised diversity; there are tensions in this, to be sure, and the relative weight given to these ideas is part of the debate about ‘second generation’ ethnic policies, but this is a very different matter from ‘intent to destroy’ groups. As for the ‘Sinicization’ of religion, official policy statements on Xinjiang suggest that this is about combating extremism and religious-based political opposition, (45) and Xi himself is quoted in ‘leaked documents’ published by the New York Times as rejecting the idea that Islam should be restricted or eradicated (in China), noting that officials should ‘respect their right to worship’. (46) Parton also cites Xi saying ‘We must severely crack down on ethnic separatist activities and persist in the anti-separatism struggle with both cultural and military forces’, but again this is far from an ‘intent to destroy’ a people.

In sum, the conclusion in this paper is that no convincing evidence has been offered of ‘intent to destroy’ the Uyghur people. A determination of genocide is not credible.

State policies in Xinjiang
What, then, is the goal of Beijing’s policy towards Xinjiang? This paper has not paid much attention to Chinese government or media statements, a number of which offer detailed rebuttals of some of the allegations. (47) But based on what the Chinese government has said and on the literature on Xinjiang, we can identify a number of strands to Chinese policy goals in Xinjiang.

First, the government is clear that it believes there has been an ‘extremist’ or ‘terrorist’ problem in Xinjiang. The evidence for this was previously accepted by major Western governments, and includes a number of attacks in the period up to 2015, as well as reports of links between extremists in Xinjiang and those elsewhere (including the statement by Daesh’s leader that Xinjiang should be seized). (48)

According to the authorities, since 2016 there have been no major terrorist attacks in Xinjiang, and this is surely seen in Beijing as a sign of a successful policy. A major part of that has been a heavy security presence and surveillance in Xinjiang, evidenced by state spending and by anecdotal reports and video
evidence from Xinjiang.

Second, the Chinese state appears to have gone beyond a security response to this threat by seeking to ‘deradicalise’ those with more ‘extremist’ ideas and ideology, including through the use of ‘training centers’. This is inherently controversial. The Chinese government has compared it to deradicalization policies elsewhere, and while (as discussed above) the scale of this program is unclear, it seems to be significantly larger than deradicalization programs in Europe, for example. Whether it is a proportionate response to the threat of extremist violence in Xinjiang is debatable.

Third is the broader policy debate in China about ‘second generation’ ethnic minority policies, which has primarily been driven by developments in Xinjiang and Tibet. As noted at the beginning of this paper, this debate appears to have led to limited changes in the formal policy framework which allows for preferential treatment for ethnic minority individuals (and ethnic minority administrative units) in a number of ways, though in practice the trend is towards a more assimilationist approach. Some scholars who have studied this issue for many years have suggested that ‘second generation’ ethnic minority
policies are ‘already with us’, (49) though in formal terms this seems premature as Uyghurs and others continue to be able to identify as such and to enjoy some more generous policies, such as university entrance criteria. In other areas, such as a greater use of standard Chinese (Putonghua) in primary and secondary education, the trend is towards more assimilationist practice. One example examined in depth in a recent article is how the relationship of ethnic minorities to the state is discussed in textbooks, demonstrating that nuanced analysis is needed to encapsulate what is happening.(50)

Relationships between national and minority identities are controversial, not just in China. But even if there has been some movement towards the assimilationist model in Xinjiang, this is a long way short of ‘eradication’ of Uyghur identity or ‘cultural genocide’ (a term often used, though it does not appear to have any relationship to the UN Convention). Anecdotal evidence suggests that some Uyghurs are not happy with these changes, while others have benefited from the greater access to jobs which learning Chinese (for example) brings.

There is clearly no sympathy in Chinese policy circles for any form of ‘Uyghur homeland’ in Xinjiang. Those in Xinjiang who want this, and their supporters outside China, are always going to differ on this point. A fundamental clash of visions between some in Xinjiang and Chinese policy is likely to continue.

Economic development remains another key feature of government policy. This includes long-standing encouragement of migration to places where there are more jobs, and investment in industry and infrastructure in Xinjiang. There are plenty of analyses in the literature which suggest that this approach to development has been unequal and imbalanced (as it has in other parts of China). The centrality of Xinjiang to the Silk Road Economic Belt has highlighted some of the contradictions and enhanced the government’s desire to deliver stability in the region. There are many challenges to be addressed still. It is
also worth noting that Western boycotts of cotton from Xinjiang could actually create greater push for the government to create job opportunities for Uyghurs outside Xinjiang if the region’s economy is hit.

To sum up, the evidence reviewed for this paper suggests that there is a strong security presence in Xinjiang, primarily prompted by the state’s assessment of an ongoing ‘extremist’ threat following numerous attacks across the region in the first half of the 2010s. There is also evidence of a gradual shift in policy towards a more assimilationist approach to ethnic minorities and a stricter implementation of family planning policies, but not a radical move to one which denies or seeks to destroy their identity – we suggest that this conclusion is consistent with Xi Jinping’s comments on ‘strengthening of ethnic
interaction, exchanges and blending’ in Xinjiang, and that those comments should not be taken as indicating a goal of erasing Uyghur identity.

Meanwhile, the authorities are keen to promote economic development and modernisation in the region, including by encouraging migration to places where more jobs are available, though in ways which may clash with the preferences of some of Xinjiang’s rural populations. None of this supports allegations of genocide, crimes against humanity, or even ‘cultural genocide’, though it is quite plausible that there are some cases of maltreatment, and that more ‘hardline’ resistance is dealt with in brutal fashion. China’s overall approach to political rights, including their relation to economic development and state interests, is different from that in liberal democracies, and this is part of the context for understanding Xinjiang too.

In case it is not already clear, let us state that we do not agree with statements that suggest peace and harmony reign in Xinjiang. The reality appears to be much more complicated.

Public discussions of Xinjiang in the West
However, complexity is not a feature of the picture that is painted in most of the Western media coverage, statements by politicians, or even governments. The State Department under Pompeo accused China of ‘crimes against humanity’ and ‘genocide’. Parliaments in Canada and the Netherlands have voted to call it genocide.

The space in the Western policy world for dissent, or even questioning, is limited, with phrases such as ‘Xinjiang denier’ being used to ostracise anyone who dares to question the received wisdom. Those who do not ‘take a stand’ against Chinese policies towards Xinjiang are outed on social media or become the objects of ‘cancel culture’ campaigns.

Recently, The Economist was heavily criticised for an editorial suggesting that events in Xinjiang did not constitute genocide (even though it did support the designation ‘crimes against humanity’), with some even calling for it to ‘apologise’ for stating this view. There are only a small number of exceptions
to the dominance of this narrative. (51)

Why might the received wisdom on Xinjiang have fallen on such fertile ground? First, there is the simple volume of reports. One report after another is issued which claims to provide ‘further chilling evidence of atrocities’. On careful reading, most of these add little new (many media reports are mainly reworkings of Zenz’s articles, with added individual testimony). By focusing on what the Western ‘response’ should be, many of these reports simply take the ‘evidence’ as read, or cite previous reports as evidence in a
somewhat circular fashion.

It is noteworthy that many of the original reports of so-called ‘atrocities’ in Xinjiang emanated from the pen of Adrian Zenz, who has also played a prominent role in spreading the allegations through media appearances and congressional testimony. Rarely has one researcher had such a powerful impact on public policy debates. The consequence of this is that judgments on Xinjiang turn to a great degree on the credibility of Zenz’s work. It is hardly surprising, therefore, that the Chinese propaganda machine has targeted Zenz by name on numerous occasions.

Second, over the last few years, there has been a clear shift in Western public debates on China. This has been driven mainly in the US, but also in Europe, with the Trump administration’s turn to a hostile China policy at the heart of it. The roots of this are complex: China as a possible peer competitor to the US, American policy elites’ insecurity and need for rivals, a desire to pull China away from its relations with Muslim states, the role of interest groups, and so on. However, this hostile policy has taken an ideological
turn, emphasising the difference and alleged inferiority of China’s system. Reports such as Zenz’s on Xinjiang are grist to the mill.

There are probably some among the critics who are genuine in their ideology, believing that pressure could turn China into a liberal democracy. But deeper stereotypes of ‘oriental despotism’ also feed these tendencies (and have come to the surface too in the way Covid-19 has been covered).

In short, current Western discussions about Xinjiang look as much like a political project as an analytical endeavour. With China now such an emotive topic for many in the West, the chances of a more rational and objective discussion of Xinjiang look slim. But the authors hope that this paper may at least encourage the resumption of normal critical discussion about this most important of topics. (52)
52 What is missed in all of this is a serious debate about what Western policy goals in relation to Xinjiang
should be and what can be achieved. Sanctioning individuals doesn’t seem to work. Trying to ‘shame’
Beijing into changing course appears only to reinforce perceptions that the West is out to get China, and that some of its media and think tanks are part of an organised disinformation campaign.

1 Authors’ note: After much thought, the authors of this paper have decided to remain anonymous. They do not want to receive hate mail, letters sent to their employers, or additional risks to securing tenure.

2 Here the West is taken to include the Anglosphere (‘five eyes’ countries) and much of Europe. There
has been much less public interest in Xinjiang outside of the West, with some exceptions such as
Singapore, Malaysia and Pakistan.

3 State Department, Determination of the Secretary of State on Atrocities in Xinjiang, 19 January 2021.

4 For a recent long piece which sums up parts of the conventional wisdom well, see

5 The focus is mainly on academic articles and long research reports, though not all of these are covered
in depth and there may be more written which the authors are not aware of. There has been so much
media reporting of this topic, that it is not feasible to review all of that coverage as well.

6 Based on official statistics for 2018. Global Times (2021). An Analysis Report on Population Change in
Xinjiang, 7 Jan.

7 Hassan H. Karrar (2018). Resistance to state-orchestrated modernization in Xinjiang: The genesis of
unrest in the multiethnic frontier. China Information 32(2), 183-202.

8 Sheena Chestnut Greitens, Myunghee Lee and Emir Yazici (2019/2020). Counterterrorism and
Preventive Repression: China’s Changing Strategy in Xinjiang. International Security 44(3), 9-47.

9 Austin Ramzy and Chris Buckley (2019). ‘Absolutely no mercy’: Leaked files expose how China
organized mass detentions of Muslims. New York Times, 16 Nov.

nsion or Part of the Problem. In China’s Frontier Regions: Ethnicity, Economic Integration and Foreign
Relations, edited by Michael E. Clarke and Douglas Smith. London and New York: I.B. Tauris. Thomas
Cliff (2016). Lucrative Chaos: Interethnic Conflict as a Function of the Economic ‘Normalization’ of
Southern Xinjiang. In Ethnic Conflict and Protest in Tibet and Xinjiang: Unrest in China’s West, edited by
Ben Hillman and Gray Tuttle. New York: Columbia University Press.

11 Adrian Zenz and James Leibold (2020). Securitizing Xinjiang: Police Recruitment, Informal Policing and
Ethnic Minority Co-optation. The China Quarterly 242, 324-348.

12 118-132. James Leibold (2013). Ethnic Policy in China: Is Reform Inevitable? Honolulu: East-West

13 Michael Clarke (2020). Beijing’s Pivot West: The Convergence of Innenpolitik and Aussenpolitik on
China’s ‘Belt and Road’?. Journal of Contemporary China 29(123), 336-353.

14 Megha Rajagopalan (2017). This is what a 21st-century police state really looks like. BuzzFeed News,
17 Oct.

15 Radio Free Asia (2018). Around 120,000 Uyghurs Detained For Political Re-Education in Xinjiang’s
Kashgar Prefecture. 22 Jan.
The 32,000 figure was cited by Zenz, see next footnote.

16 Adrian Zenz (2019). ‘Thoroughly reforming them towards a healthy heart attitude’: China’s political
re-education campaign in Xinjiang. Central Asian Survey 38(1), 102-128. A pre-print posted online is
dated 6 Sept. 2018.

17 Normative judgments about this sort of program will differ, and deradicalization programs elsewhere have proved controversial (this is not intended to draw a direct comparison between the Chinese
approach and that elsewhere, something that Chinese officials have done, simply to note that this sort of
program is likely to be contested).

18 Ramzy and Buckley, Absolutely no mercy.

19 The authors find the logic and structure of these articles hard to follow. They often string together
anecdotal and partial information and present it confidently as established fact without considering
alternative interpretations for the evidence presented, or demonstrating awareness of the limited nature of the evidence on which their conclusions are based. These are basic weaknesses in research methodology.

20 Jessica Batke (2019). Where Did the One Million Figure for Detentions in Xinjiang’s Camps Come
From? ChinaFile, 8 Jan.

21 There have been some media reports that the one million is a UN figure. The only reference we could
find to this is discussions of the claim at a Committee on the Elimination of Racial Discrimination hearing
in August 2018, but these are based on the claims made by others, not any independent UN work. See

22 Sean R. Roberts (2021). The Roots of Cultural Genocide in Xinjiang. Foreign Affairs.

23 Chaguan. China doubles down in Xinjiang. The Economist. 12 Dec. 2020 (emphasis added).

24 See Batke, The One Million Figure.

25 BBC (2021). ‘Their goal is to destroy everyone’: Uighur camp detainees allege systematic rape. 2 Feb.

26 Parton published a comment piece on the same issue in The Spectator the following day; the timing
suggests that he was aware of the content of the BBC report before it was published.

27 Eva Dou and Philip Wen (2020). ‘Admit Your Mistakes, Repent’: China Shifts Campaign to Control
Xinjiang’s Muslims. Wall Street Journal. 6 Feb.

28 The first paragraph summarises existing claims about ‘re-education camps’. See Australian Strategic
Policy Institute (ASPI) (2020). Uyghurs for sale: ‘Re-education’, forced labour and surveillance beyond
Xinjiang. Policy Brief No. 26, 1 Mar.

29 For details, see David Tobin (2020). A ‘Struggle of Life or Death’: Han and Uyghur Insecurities on
China’s North-West Frontier. The China Quarterly 242, 301-323.

30 For an excellent book on this topic, see Pun Ngai (2016). Migrant labor in China. Polity Press.

31 BBC (2021). China’s ‘tainted’ cotton.

32 Adrian Zenz (2021). Coercive Labor and Forced Displacement in Xinjiang’s Cross-Regional Labor
Transfer Program: A Process-Oriented Evaluation. The Jamestown Foundation. One of his sources is a research report from Nankai University (which he translates in an Appendix), and while this may be a good source, a lot is hung on this one report, which is not by an official body. The BBC carried a story on the same issue on the same day, also citing the ‘Nankai report’. See

33 If nothing else, this suggests that normal critical faculties should be applied to all the reports on this

34 Adrian Zenz (2020). Sterilisations, IUDs, and Mandatory Birth Control: The CCP’s Campaign to Suppress Uyghur Birthrates in Xinjiang. Jamestown Foundation.

35 Zenz’s methodology is again to look for official documents or media reports on the Chinese internet,
and to draw wider conclusions assuming that the small number of such reports he finds are
representative. This may explain the apparent inconsistencies in figures cited throughout the report.

36 Cao Siqi (2017). Xinjiang implements new uniform ethnic family planning policy. Global Times 31 July.

37 See below for discussion of the concept of Zhonghua minzu.

38 Global Times (2021). An Analysis Report on Population Change in Xinjiang, 7 Jan.

39 Vice News (2021). China’s Vanishing Muslims: Undercover in the Most Dystopian Place in the World.
Video available at

40 The text of the Convention can be accessed at 20Prevention%20and%20Punishment%20of%20the%20Crime%20of%20Genocide.pdf

41 Charles Parton (2021). The truth about China’s genocide against the Uyghurs. The Spectator. 3 Feb.

42 See Colum Lynch (2021). State Department Lawyers Concluded Insufficient Evidence to Prove
Genocide in China. Foreign Policy. 19 Feb.

43 Roberts, The roots of cultural genocide.

44 BBC (2021). Uyghurs: ‘Credible case’ China carrying out genocide. 8 Feb.

45 陈全国:深入学习贯彻习近平总书记重要讲话精神 以新疆工作的优异成绩庆祝建党100周年, Xinjiang
Daily, 5 Jan. 2021.

46 Ramzy and Buckley, Absolutely no mercy.

47 For example, a press conference held in Beijing in February 2021. See

48 Anna Hayes (2020). Interwoven “Destinies”: The SIgnificance of Xinjiang to the China Dream, the Belt
and Road Initiative, and the Xi Jinping Legacy. Journal of Contemporary China 29 (121), 31-45, pp.

49 Gerald Roche and James Leibold (2020). China’s Second-generation Ethnic Policies Are Already Here.
Made in China Journal. 7 Sept.

50 Taotao Zhao and Sow Keat Tok (2020). “From Academic Discourse to Political Decisions? The Case of
the Xinjiang Ethnic Unity Education Textbook Reform”. The China Quarterly.

51 The most prominent exceptions have been a series of reports from Grayzone, which have attacked
Zenz personally and outlined numerous inconsistencies in his reports. An example of ‘cancel culture’ in
this context was a letter written by a group of NGOs (signed ‘Hong Kong Global Connect’) in early
February to the president of Columbia University, complaining about comments on Xinjiang and other
China-related issues by Prof Jeffrey Sachs.

52 What is missed in all of this is a serious debate about what Western policy goals in relation to Xinjiang
should be and what can be achieved. Sanctioning individuals doesn’t seem to work. Trying to ‘shame’
Beijing into changing course appears only to reinforce perceptions that the West is out to get China, and that some of its media and think tanks are part of an organised disinformation campaign.

One thought on “Xinjiang – what do we know, how and why

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s