Scholar Adrian Zenz’s Newly Released ‘Xinjiang Police Files’ Affirm Scale of Campaign and Beijing’s Role in Shaping It
Since 2017, China’s government has detained large numbers of Uyghur and other Turkic populations. Outside observers have worked to understand the nature and scale of this interment drive, but they have been hampered by limited physical access to the Xinjiang Uyghur Autonomous Region as well as by strict controls on digital communication. Two questions in particular have cropped up again and again: How many people has the Chinese government sent into prisons or “re-education” camps? And how direct is General Secretary Xi Jinping’s involvement in this effort?
Now, an internal Chinese government document provides new support for the extraordinary scale of internment during what was likely its peak in 2018 and 2019. The document, a transcript of an internal June 15, 2018 speech by Minister of Public Security Zhao Kezhi, reinforces the plausibility of previous detention estimates, adding to the evidence that the People’s Republic of China (PRC) had interned between one and two million Uyghurs and other ethnic minority individuals by the late 2010s. The speech also points to Xi Jinping’s informed and active support for Xinjiang’s “re-education,” “strike hard,” and “de-extremification” campaigns, as well as for continued spending on additional detention facilities and staff to manage the influx of detainees.
Zhao’s speech, along with tens of thousands of Chinese government documents and spreadsheets, which I refer to as the “Xinjiang Police Files,” were obtained by an individual who gained access to the internal police computer networks of two predominantly Uyghur and Kazakh counties in Xinjiang: Konasheher (Shufu) and Tekes (Tekesi). Given the likely repercussions, this individual has asked to remain anonymous. The transcript of Zhao’s speech is an unredacted scan of a paper copy, labeled “classified document” on its first page. The nature of the document makes it difficult to fully corroborate its authenticity through openly available information. But a summary of the speech published on a local government website repeats key phrases from the transcript. Another document from the Xinjiang Police Files, a transcript of a speech by then-Xinjiang Party Secretary Chen Quanguo, delivered on June 18, just three days after Zhao’s, echoes several of Zhao’s phrases and overlaps substantially in its overall content.
I have detailed my further authentication of the full cache of the Xinjiang Police Files in a separate, peer-reviewed article published in the Journal of the European Association for Chinese Studies. The credibility of the documents is bolstered by their sheer volume; the detail, specificity, and metadata in many of the included photos; internal textual consistency; and overlap with previous document leaks.
Research based on both publicly available and internal Chinese government documents is necessary because China’s government has worked so hard to keep international observers from gathering first-hand information. The campaign to intern large numbers of Uyghurs and other predominantly Turkic ethnic groups, holding them in prisons as well as “re-education” (literally “transformation through education”) facilities, began five years ago, in 2017. Scholars and reporters have written extensively about the campaign, which employs not only incarceration but also coercive measures to limit childbearing, restrict cultural and religious practice, and facilitate forced labor.
In the absence of independent data—or any reliable official information from the Chinese government itself—observers have been left to estimate how many people the state has incarcerated. These estimates rely on a range of different methodologies, from phoning local police officers, to extrapolating from leaked government spreadsheets, to analyzing food subsidy payments for “re-education” facilities. The estimates themselves also have varied widely. My own projection from 2018 posited between several hundred thousand and one million interned; in 2019, based in part on leaked government spreadsheets listing detention figures for several Uyghur townships, I increased this estimate to between 900,000 and 1.8 million. Many scholars of the region have accepted these estimates and have cited varying numbers in their own publications, ranging from “up to one million” at the low end to “between one million and three million” at the high end.
For several years, observers were similarly unsure of the precise role of General Secretary Xi Jinping or of the central government in the campaign. While it was obvious that Xi would have given his blessing to any set of policies with such far-reaching consequences, a dearth of information about internal Chinese government processes prevented scholars from concluding with confidence that Xi played any direct role in those policies’ formulation. Though the “Xinjiang Papers,” another series of internal Chinese government documents, contained quotes from Xi that linked him to many aspects of the ongoing crackdown, they did not contain any statements in which he spoke directly about mass internment in “re-education” camps.
Zhao’s speech offers additional evidence that between one and two million people have indeed been incarcerated. It also shows that Xi himself has been personally aware of the details of this campaign, and that he issued instructions that enabled its continuation and expansion. Zhao delivered the speech in Urumqi at the end of a five-day “investigative” visit to Xinjiang. Other high-ranking central government leaders were also part of the delegation.
Zhao affirms Beijing’s support for the mass internment drive. In the first section of his speech, he summarizes the region’s praiseworthy achievements and commends regional authorities for having thus far “done transformation through education [work] well.” He also lays out the timeline the regional government had been following in implementing its unprecedented crackdown. After the arrival of Xinjiang Party Secretary Chen Quanguo, the region commenced a five-year work plan: in year one (2017), “stabilize” the region; in year two, “consolidate” those gains; in year three, achieve a “basic normalization” of the new state of affairs; and by the fifth year (2021), reach “comprehensive stability.”
Chen Quanguo reiterates the same five-year scheme in the speech he delivered three days after Zhao’s. Chen says that he was “sent” to Xinjiang by Xi himself, not as a routine assignment but on a special mission for the nation: “The General Secretary sent me to Xinjiang; first, not in order to [merely] be an official; second, not in order to make a fortune; third, not in order to have nothing but an empty title. [Rather,] the General Secretary sent me to Xinjiang in order to make a stable Xinjiang arise . . .” Chen adds that he personally told Xi he would be willing to work in Xinjiang for 10 years. But in December 2021, Beijing replaced Chen with Ma Xingrui, the governor of Guangdong province. This change—swapping out Chen, who was likely deployed to Xinjiang because of his ruthless securitization of Tibet and his reputation for iron-clad Party discipline, with Ma, a scientist-turned-technocrat—suggests Beijing considered its five-year scheme for Xinjiang well in hand.
In the second section of Zhao’s speech, he highlights ongoing challenges and work priorities for the current “consolidation” phase of the plan. He lists the problems still facing the region and asserts that its counterterrorism struggle “absolutely cannot be relaxed in any way.” Here, Zhao begins to hint at the number of people targeted for “re-education” or imprisonment. He refers to two groups, each comprising “two million people,” who pose an internal threat to stability: “Xinjiang has two million people who have been influenced by pro-Xinjiang independence, pan-Islamist, and pan-Turkist thinking. Southern Xinjiang has more than two million people who have been severely influenced by the infiltration of extremist religious thought.”
The first “two million,” those “influenced by pro-Xinjiang independence, pan-Islamist, and pan-Turkist thinking,” alludes to people who identify with a broader Turkic religious, cultural, and linguistic heritage—that is, Uyghurs, Kazakhs, Kyrgyz, and people of other ethnically Turkic groups. Beijing links this sense of non-Han Chinese identity, or pan-Turkic identity, with separatism, frequently describing it as a “poison” of the mind that must be eradicated through “re-education.” Zhao specifies that the second “two million” people reside in southern Xinjiang, where the percentage of Uyghur and other Turkic peoples is much higher than in the north. Zhao’s reference to “extremist religious thought” could describe almost anyone who engages in religious behavior, given Beijing’s exceedingly broad definition of what constitutes “extremism.”
Notably, Zhao does not accuse these two groups of people of violent resistance against the state. Rather, it is primarily their cultural and religious identity that make them targets for “re-education.” This capacious rubric for labeling people as a threat to the state broadly aligns with witness accounts, documentary evidence, and satellite imagery that all point to internment on a very large scale.
Though Zhao enumerates the two groups separately, their membership almost certainly overlaps. Someone “influenced by extremist religious thought” in the south, for example, could also be harboring “pan-Turkist” tendencies. It’s hard to know the size of this overlap, whether Zhao is talking about a total closer to two million people or four million people. But given regional demographics, the overlap could be rather small, the total number closer to three or four million: In 2018, some 3.5 million Uyghurs, Kazakhs, and Kyrgyz lived in northern Xinjiang—all Turkic groups the state deems liable to espouse “pan-Turkist” tendencies, meaning that two million people could potentially be designated among this group alone.
Nowhere does Zhao definitively say that either of these “two million” groups of people have been or will be incarcerated. Yet, time and again, high-level Chinese officials have made clear they believe “re-education” is key to containing terrorism and religious extremism. White papers, state media reports, and other official statements describe “re-education” centers as cutting off “terrorism and religious extremism at the source.” In his speech, Zhao too lauds the “re-education” centers for successfully “implementing centralized re-education for those who have been influenced by extremist religious thought,” suggesting that, if nothing else, both of these “two million” groups were the kind of people authorities sought to detain and transform.
In his speech, Chen, like Zhao, refers to those who have “been influenced by pro-Xinjiang independence, pan-Islamist, and pan-Turkist thinking,” calling them a “relatively large group of people.” Chen exhorts cadres to “keep the alarm bells ringing” and continue to execute Xi Jinping’s “strategy for Xinjiang.” This is consistent with statements by Zhao that more directly link Xi to the mass internment.
In the final section of his speech, Zhao discusses a practical implication of the broad rubric for “re-education” he has just outlined: the overcrowded prisons and detention facilities resulting from the mass incarceration campaign.
Zhao depicts a system of prisons and other internment facilities holding detainees in numbers that strain their capacity. The “success” of the “strike hard” and “de-extremification” campaigns had led, by mid-2018, to “a great number of excess detentions [relative to capacity]” (Zhao mentions “excess detentions” four times in the speech). In fact, Zhao makes note of measures taken to relieve the pressure on Xinjiang’s prisons, including bringing police from other parts of China to Xinjiang and transferring Uyghur prisoners from Xinjiang to other provinces, albeit in comparatively small numbers. Zhao assures regional officials that the central government would soon approve funding to construct additional detention centers in southern Xinjiang. He also asserts that Beijing would “increase the strength of its support” for covering the “high costs” of operating and maintaining Xinjiang’s internment facilities.
Though Zhao never says exactly how many people the region has incarcerated, his emphasis on overcrowding attests to very large numbers of detainees, and it aligns with witness testimonies from around that time. It also explains why, as demonstrated by satellite imagery, construction or expansion of internment facilities boomed in the second half of 2018. Evidently the central government had come through with the funding, just as Zhao promised. In 2021, experts estimated that expansion of detention facility space since 2016 would allow the region to hold at least one million new detainees, if they were housed according to state standards for minimum living space. If detention facilities were overcrowded, as Zhao indicates in his speech, then the space built after 2016 could hold well over a million people, and that is without even including the capacity of pre-existing detention facilities.
Further support for the one-to-two million estimate comes from other material in the Xinjiang Police Files. Spreadsheets containing information on more than 280,000 people living in Konasheher county show that in 2018 over 12 percent of all adults were in some form of internment or imprisonment. This accords with my earlier estimates that between 7.7 and 15.4 percent of Xinjiang’s adult Turkic minority population have at some point been interned, and suggests a total internment figure at or above one million people.
Still other documents in the Xinjiang Police Files refer to a category of “500,000 untrustworthy persons.” Though previous caches of leaked data also described detainees as “untrustworthy persons,” they did not mention a specifically defined set of “500,000 untrustworthy persons.” Like the references to the two “two million” groups, this phrase could indicate yet another group of people the government is targeting for “re-education.” (The government’s definition of “untrustworthy” appears extremely broad: A 32-year-old woman labeled as one of the “500,000 untrustworthy persons” was “detained for re-education” in July 2017 because she allegedly engaged in “unusual nightly activities” and her phone was “frequently turned off.”) Of note, the phrase “500,000 untrustworthy persons” appears in the same spreadsheets as the more general term “untrustworthy persons.” In my view, this indicates that “500,000 untrustworthy persons” potentially functions as a sub-category within the wider “untrustworthy” label, meaning that the total combined number of “untrustworthy persons” is in fact substantially larger than 500,000. Local directives within the Xinjiang Police Files make it clear that “untrustworthy” people should be locked up: One township-level notice from April 2018 states that “all those who are untrustworthy must be detained for re-education”; a few months later, a prefectural police chief said that “untrustworthy” people are to be “put into a trustworthy place to slowly transform them.”
Zhao’s speech also offers evidence for General Secretary Xi Jinping’s informed and active role in directing policy in Xinjiang. Zhao says Xi’s “important instructions on governing Xinjiang according to the law, unifying and stabilizing Xinjiang, and building Xinjiang over the long term” provide the basis for numerous policy priorities outlined in the speech, including “bringing the Vocational Skills Education and Training Center management work into the orbit of legalization” (i.e., establishing them as legally operating facilities). Zhao describes Xi as directing Xinjiang authorities to “conduct de-extremification work,” which according to Zhao includes the “transformation through education” work conducted in the “re-education” facilities. According to Zhao, after central Party leader Guo Shengkun reported on the prison capacity challenges he witnessed during his visit to Xinjiang in April 2017, Xi himself ordered regional authorities to, “implement practical measures such as expanding the number of employed [staff in detention facilities], enlarging the capacity [of these facilities], and increasing investment [in these facilities] within the set time frame.”
Reached for comment on Xi’s role in directing the expansion of detentions in Xinjiang, the Chinese Embassy in Washington referred ChinaFile to a statement by spokesperson Liu Pengyu describing Xinjiang as enjoying “social stability and harmony” following “a host of decisive, robust and effective deradicalization measures.”
Unsurprisingly, Zhao’s speech abounds in references to General Secretary Xi, to his leadership and vision, and to his close interest and guiding role in the governance of Xinjiang in the “new era.” Zhao credits Xi with having devised the plan by which the region is now enjoying “stability” and celebrates that plan as “entirely sagacious and correct.” Even if this kind of fawning language in part reflects the political necessity of ceaselessly praising the nation’s leader, Zhao’s speech suggests that Xi’s personal involvement in Xinjiang policy has determined actual developments on the ground. At the very least, Xi has had a detailed understanding of the mass internment campaign and played a role in directing it.
By early 2022, Xinjiang had reached the end of Beijing’s five-year strategy to pacify and subdue the region—meaning that, according to the timeline, Xinjiang is now in a state of “comprehensive stability.” Indeed, the visible police presence and onerous security checks of 2017 and 2018 have largely given way to today’s more imperceptible but still pervasive surveillance regime. In his June 2018 speech, Chen Quanguo explained that the region would shift toward a much more “invisible” security state, where the police force is “not reduced” but simply less evident, and where it can be “immediately” mobilized if required. Measures such as mass internment in “re-education” camps constituted what a group of Chinese scholars in an academic report referred to as “drastic short-term measures” that were “absolutely necessary and effective.”
And yet, even in 2017, in the transcript of another speech from the Xinjiang Police Files, Chen had already predicted that some of those detained in “re-education” camps “may not necessarily have been transformed well even after three or five years.” Indeed, while some detainees have been released, witness accounts and documentary evidence published through the Xinjiang Victims Database indicate that many others have since been sentenced to long prison terms. Zhao’s speech shows that Beijing prioritized not only the construction and expansion of “re-education” facilities, but also the prisons that would confine many individuals after their initial “re-education” period. The “re-education” work that Zhao says “must persist for a long time” is likely now being carried out in prisons instead of camps, which effectively serve as a legal mechanism for arbitrary incarceration.
Beijing’s campaign of repression in Xinjiang, which had relied on middle-of-the-night house raids and large-scale internment, has now entered a new phase. Under Xinjiang’s current Party secretary, Ma Xingrui, the region is now embarking on a multi-billion-dollar initiative to boost infrastructure construction and “high-quality economic development.”
However, in 2022, the regional government’s work plan described the struggle to enforce social stability as “still grim and complex.” In April, Ma gave the strongest indication yet that some forms of “re-education” will continue when he stated that Xinjiang will, “to the greatest possible extent, educate and save people who have been bewitched by extremist ideas”—a commonly used euphemism for “re-education” in camps. Many continue to languish in prisons and detention facilities. For the elderly, including important cultural and religious figures, their terms are tantamount to life sentences. Meanwhile, the state continues to destroy traditional community ties by means of forced labor transfers, separation of parents from children, eradication of the Uyghur language through boarding schools, and “optimization” of ethnic “population structures” through birth prevention policies and a planned migration of 300,000 Han Chinese settlers into Uyghur areas. Along with the muting of the region’s police presence, these measures and their support from the very highest reaches of China’s leadership amount to a largely invisible, slow form of genocide.