I noticed the previous human rights accusations in Xinjiang, and when I checked the relevant Wikipedia entries, I found that key statements seemed to lack evidence, such as "incarcerated more than an estimated one million Turkic Muslims without any legal process in internment camps." The sources cited seem to be only interviews with some people, telling the facts in a story-like tone, and the authenticity lacks support from other evidence.

Some reports are even suspected of rewriting the facts, such as portraying people arrested for supporting radical Islamic terrorism as being forcibly separated from their children. Source:

On a quiet street in the ancient Silk Road city of Kashgar, a house lies empty, padlocked from the outside, the family who lived there gone.

The father was detained in February; three months later the mother was also taken away by authorities. They had allegedly shared extremist Islamist content on their mobile phones, family friends said.

Despite protests from relatives, two of their children, aged 18 and 15, were then detained and their younger two, aged seven and nine, were sent to a state welfare centre. “The grandfather even wept, but the authorities would not let him keep his grandchildren,” recalled an acquaintance.

The family had fallen foul of an anti-terror drive conducted by Beijing, which has forcibly separated families, sending thousands of children to de facto orphanages, according to Uighurs interviewed in China and abroad by the Financial Times.

Or a large number of growth rates of fertility rates are listed. It seems that the fertility rate has dropped by 60%, but in fact it is just that the fertility rate that was originally only 3% has become 1.2% (for example). Source:

The result of the birth control campaign is a climate of terror around having children, as seen in interview after interview. Birth rates in the mostly Uighur regions of Hotan and Kashgar plunged by more than 60% from 2015 to 2018, the latest year available in government statistics. Across the Xinjiang region, birth rates continue to plummet, falling nearly 24% last year alone — compared to just 4.2% nationwide, statistics show.

This ambiguous evidence seems to have been deliberately arranged to be presented in this narrative format to ordinary people who lack understanding. These manipulations of public opinion make me feel that the entire accusation seems to be based on conjecture and difficult to delve into.

Regarding the human rights accusations in Xinjiang, is there any conclusive and multi-faceted material evidence to support it so that Beijing cannot deny it?

  • 5
    If the fertility rate measured as number of children per year per total population drops from 3% to 1.2% then it did fall by 60%. The number of children born is 60% lower then it used to be. That is just maths and describes the exact same fact in different ways.
    – quarague
    Commented Mar 8 at 8:10
  • 5
    Related question with lots of sources that are relevant to this question as well: Why is China primarily targeting the Uyghurs with their internment camps?
    – Philipp
    Commented Mar 8 at 9:11
  • 11
    @Jack_here Please leave the situation in Gaza out of this. It doesn't have anything to do with the Xinjiang situation in China, doesn't add anything to the question and only provokes yet another Israel/Palestine debate.
    – Philipp
    Commented Mar 8 at 9:21
  • 6
    One needs to be careful not to raise the bar too high. Conclusive evidence is hopefully not meant as in a courtroom, but rather as in "reasonable, neutral humans would have a strong inclination to believe the arguments". Otherwise the problem is that maybe nothing will really convince somebody who doesn't want to be convinced. What is exactly the hurdle to take here? Commented Mar 8 at 11:17
  • 7
    As well as a large amount of personal testimony, there's a lot of other corroborating evidence, such as demographic data mentioned, and satellite images, which can easily be found online. I wouldn't accept deficiencies in a Wikipedia page as evidence for or against: many governments and NGOs have produced reports. Absent a definition of the sort of conclusive evidence that the OP wants, this looks like POV-pushing.
    – Stuart F
    Commented Mar 8 at 12:47

3 Answers 3


incarcerated more than an estimated one million Turkic Muslims

This is estimated based on incarceration rate etc.

Leaked data reviewed by the Associated Press revealed that almost one in 25 people in the Uyghur heartland in western China have been imprisoned on 'terrorism-related charges' - the highest known incarceration rate in the world.

Xinjiang has a population of 25,890,000 based on the last census.

AFAICT the estimates are based on sampling some counties, so it is possible that the extrapolation is not entirely correct (to the rest). YMMV.

enter image description here

I'm a bit but right now, so I didn't read all the details, but this doesn't appear to include those just sent to 're-education'.

The list does not include people with typical criminal charges such as homicide or theft. Rather, it focuses on offenses related to terrorism, religious extremism or vague charges traditionally used against political dissidents, such as “picking quarrels and provoking trouble.” This means the true number of people imprisoned is almost certainly higher.

But even at a conservative estimate, Konasheher county’s imprisonment rate is more than 10 times higher than that of the United States, one of the world’s leading jailers, according to Department of Justice statistics. It’s also more than 30 times higher than for China as a whole, according to state statistics from 2013, the last time such figures were released.

And yeah, China hasn't officially provided any figures for the more informally detained in "vocational facilities", so there's that. But there is some official data on the sudden increase in long-term sentences in 2017, as OHCHR reported

The [Chinese] Government has not released official data about the number of individuals who have undergone re-education in VETCs. In 2018, the UN Committee on the Elimination of Racial Discrimination noted that “estimates of the number of people detained range from tens of thousands to over a million”, and called on the Government to provide statistics for the past five years. In response, the Government asserted that it was not possible to state the number of those taking part in education and training, because it “is dynamic, as people are continuously coming and going,” a position it has maintained since.

In the absence of officially available data, [...] it has been estimated that around 10-20 per cent of the adult “ethnic population” in these counties and townships were subjected to some form of detention between 2017 and 2018.

Another change in 2017 was the increase in the number of people given sentences of five years or longer. Prior to 2017, approximately 10.8 per cent of the total number of people sentenced in XUAR received sentences of over five years. In 2017, that figure rose to 87 per cent of the sentences. According to official Government statistics, during 2017 alone, XUAR courts sentenced 86,655 defendants, or 10 times more than in the previous year, to prison terms of five years or longer, although again it is not possible to disaggregate the number charged and convicted for terrorism or “extremism”-related offences.

And yeah, even that OHCHR report has been much delayed.

TBH, I'm not sure why OHCHR didn't cite these figures (perhaps the very long delay to publication is at fault), but SCMP reported in 2020 that the Xinjiang regional government released some figures on the 'vocational' centers:

Titled “Employment and Labour Rights in Xinjiang”, the white paper said the regional government had organised “employment-oriented training on standard spoken and written Chinese, legal knowledge, general know-how for urban life and labour skills” to improve the structure of the workforce and combat poverty.

It had provided vocational training to an average of 1.29 million urban and rural workers every year from 2014 to 2019, the white paper said, apparently not using the Chinese government’s five-year planning period as the reporting time frame.

Of those workers, about 451,400 were from southern Xinjiang – an area it said struggled with extreme poverty, poor access to education and a lack of job skills because residents were influenced by “extremist thoughts”. That period was also when regional authorities introduced a “systemic de-extremification” campaign to counter terrorism and extreme religious thoughts, according to mainland media reports.

A mainland-based academic who studies Xinjiang issues said it appeared to be the first time Beijing had “indirectly acknowledged” the number of ethnic Muslim minorities receiving “vocational training” under the “de-extremification” programme.

“If you take into account the timing of China’s de-extremification measures that began in 2014, the ‘1.3 million people being trained per year from 2014 to 2019’ is very close to the number [in the camps] estimated by Western critics,” said the academic, who declined to be named because of the sensitivity of the matter.

“But China does not see these training facilities as internment camps, and what it is really trying to highlight [through the white paper] – to counter Western criticism – is that the ‘vocational training’ they provide is actually a social service to improve people’s livelihoods and alleviate poverty.”

Anyhow, the OHCHR report stated that many of those interviewed said they didn't have choice about attending these, amusingly citing apparently the same Xinjiang State Council Whitepaper:

The 2019 White Paper on “Vocational Education and Training in Xinjiang” states that the centres are “residential”, and that referral follows a decision by the court or public security officials, rather than being voluntary. This is the case even for referrals by the procuratorate, where the concerned individual is given a “choice” between referral to a VETC facility and a prison sentence, implying that placement in a VETC is a form of alternative sanction to a prison sentence.

Individuals interviewed by OHCHR who had been placed in VETC facilities described being taken to such facilities, usually by public security officials. The majority of the interviewees who were apprehended between 2017 and 2019 were held at a police station before referral to a VETC facility. They said that they were told that they had to go to a VETC facility and were not given an alternative option. None of the interviewees felt they could challenge the referral process, and none had access to a lawyer prior to being sent to the VETC facility nor at any point during the time they were present there. Several underwent long interrogations in police stations before their eventual placement.

  • 6
    I mean seriously, in China it's impossible to find out what exactly happened to government ministers that disappear for months. They all got not-further-explained 'health issues'. And you expect precise stats on detainees in some province. Commented Mar 8 at 14:17
  • There is a bug in this estimation. In Xinjiang, Han and Uyghur people are almost 1:1, and Han people rarely have religious beliefs. Therefore, if the probability found in the Uyghur hinterland in western China is used to estimate the number of people detained in Xinjiang as a whole, then the estimated result is at least twice as high as the truth. And this is only based on the belief that the probability of being investigated is not exaggerated.
    – Jack_here
    Commented Mar 8 at 16:06
  • 7
    @Jack_here Even if the numbers were half the size, they should still be extremely alarming, so I'm not sure what your point is except just to argue.
    – Michael W.
    Commented Mar 8 at 16:59
  • 1
    @MichaelW. Since the number can be confirmed to be half the size from the beginning, why do you have to wait until others expose it before you change your mind? I think this kind of conviction without actual evidence should be stopped. The answer I want to see shouldn't be ambiguous or even logically flawed.
    – Jack_here
    Commented Mar 8 at 18:11

Frame challenge: There is virtually no kind of evidence that the country being incriminated cannot deny or just plain ignore. Even if you somehow had a 24/7 footage of every affected Xinjiang resident, how would you present it to China and have it act on it? Who would watch billions of hours of video, on the majority of which nothing incriminating really happens?

Failing that, you can cherry-pick evidence to cater it to a specific audience. That's basically what you have dismissed as "narrative format to ordinary people who lack understanding". For example, Russian emigrant media Meduza used to run stories about ethnic Kazakhs in Xinjiang camps.

This would look weird to e.g. US audience - why would you focus on a specific minority rather than writing about general Uighur experience. But it hits the target within Russian audience: Kazakhstan borders Russia with close ties, and has significant Russian population, so for a Russian reader, a Kazakh is somebody they might know, somebody who lives in the same town and often speaks the same language and shares part of culture with them.

So a story how a Kazakh went for a business trip on a Kazakhstani passport into Xinjiang and was put into a reeducation camp to study Chinese language would definitely resonate with the audience. Whereas a generic story how an Uighur muslim person's organs were harvested in the 80s, while being much more severe, will just not spark that amount of attention.

At the same time, it's kind of obvious that the recurring attention of Meduza towards Xinjiang is due to their friendship with RFE/RL and the like, and may be biased in more ways.

  • The wording in the answer is more accurate. Yes, stories. The media invites some people from ethnic minorities to tell some stories to the camera. Scholars consider which narrative is more appealing between “1.8% reduction” and “60% reduction” based on fertility data. In front of a screen showing satellite images of some factories, experts thought about how the changes in these factories could be related to the concentration camps. Interesting stories.
    – Jack_here
    Commented Mar 8 at 16:21
  • @Jack_here 1.8% reduction is wrong. If there were k births before, and .4 * k births after, that's a 60% reduction. Expressing those as a percentage of something else, the population of the area, the population of the world, the number of births in China, the number of births in the world, etc., and subtracting the two percentages doesn't give you a percentage reduction.
    – prosfilaes
    Commented Mar 8 at 17:08
  • To be fair to Jack_here I agree that an absolute number should be provided alongside a relative number. We often have this problem in news articles and in business. It should read 60% or 1.8%pts.
    – Lag
    Commented Mar 8 at 17:55
  • @Jack_here I've thought for a long time that absolute and relative percentages should be printed.
    – Lag
    Commented Mar 8 at 18:13
  • @prosfilaes I think this issue should be discussed in the math stack, but I think the contributors there should also agree that 3%-1.2%=1.8%, not 3%-1.2%=60%. Forgive me for generalizing mathematical concepts here, but it is relevant to the question.
    – Jack_here
    Commented Mar 8 at 18:23

How about satellite pictures?

Our findings from satellite images reveal 380 detention camps in Xinjiang, pointing to a campaign of ethnic replacement

In total, we have found 380 separate detention facilities that have either sprung out of the deserts and oases, or expanded from smaller detention facilities since 2017. We don’t believe that we have found them all. The largest is more than 300 acres in size. That is more than three and a half Disneylands. Nearly nine Pentagons.

Using official population figures, the 380 camps equate to at least one new or expanded detention facility for every 37,000 people of non-Han nationality in Xinjiang. This would be the equivalent of New York City building more than 55 prisons only for black Americans in just over three years.

OK, but whose satellite images, you say? US DoD? This isn't detailed very clearly on the Guardian's articles, but Maxar satellites are good enough these days that it would be easily disprovable if it was over-bogus.

This is btw the website the Guardian's article was linking to. Keep in mind tho, this is run by the Aussie gov, so bias is likely.

p.s. Western human rights associations, like Amnesty, do criticize China, but before one pooh-poohs that overmuch as bias, keep in mind they also can be pretty harsh towards Western countries, like Canada and UK so this whole bias claim seems, a tad, ad hominem.

  • Yes, the satellite pictures. You concluded that this is an ethnic replacement movement based on the conditions of 380 internment camps in Xinjiang. Very logical. But I'm curious how the conclusion that these houses were internment camps was derived. These so-called "mystery camps" don't look mysterious at all, as they don't even have high walls to hide the buildings. Even though the building in the picture looks like a school, the "internment camp" result is obtained in a strange way.
    – Jack_here
    Commented Mar 9 at 3:28
  • 1
    Of course, people doing this kind of analysis for human rights groups are just guessing that they look prison-ey as opposed to school-ey. Since you obviously know better, you need to contact them right away and set them straight. We've all been fooled! Commented Mar 9 at 3:47
  • This is unrealistic. Because people who prefer to believe stories, even if they hear the truth later, will not be willing to believe that what they believed before is wrong. Therefore, self-certification is useless, because they will not believe China's self-certification based on their first impression. These reports can only be effective if the reporters voluntarily go to Xinjiang to investigate people's conditions on the spot, and they must avoid the suspicion of being paid. But will this happen?
    – Jack_here
    Commented Mar 9 at 3:56
  • the reporters voluntarily go to Xinjiang to investigate Oh, my, they can't. So unexepected Or maybe HRW is a Western stooge? Commented Mar 9 at 4:01
  • It is worth noting that these actions were not initiated by human rights organizations, but were promoted by the government and the media. Otherwise, it is impossible that the voices of these human rights organizations regarding the Palestinian-Israeli issue would be less than half of the human rights issues in Xinjiang.
    – Jack_here
    Commented Mar 9 at 4:01

Not the answer you're looking for? Browse other questions tagged .