Suppression of birthrates

  • Xinjiang's birthrate has dropped at a historically unprecedented speed over the last few years, and these drops are concentrated among non-Han areas
  • It is highly unlikely that the drop is due to simple enforcement of family planning policy; Xinjiang's birthrates—while higher than average—were already below replacement rates and lower than some Han provinces in 2010, according to official census data
  • The report most frequently cited by Western media by Adrian Zenz on mass sterilization contains noticeable numerical errors and at least one instance of misleading wording, and also partially relies on unverifiable evidence; but even discarding the arguments affected by these issues, the evidence of coercive suppression of birthrates still stands
  • All data I use is contained in this Google Spreadsheet unless otherwise noted

In late February 2021, the newest edition of the Chinese Statistical Yearbook was published. The Yearbook, as the name implies, is the annual product of the Chinese government that catalogs a wide array of statistical data for public use. Data from the 2020 Yearbook (which covers 2019) shows that in the space of two years, the birth rate in Xinjiang dropped approximately 48.7%, from 15.88‰ (per thousand) to 8.14‰. The average for all of China is 10.48‰. Uyghur-majority areas represented the bulk of this decline.

This data is not new; similar arguments had been made by Adrian Zenz in a June 2020 report based on publications from Xinjiang's regional government ("Sterilizations, IUDs, and Mandatory Birth Control: The CCP's Campaign to Suppress Uyghur Birthrates in Xinjiang"). There are various data errors and misused statistics in this paper, detailed below. Some impact specific arguments, but do not negate the paper's overall point.

This post first explains the significance of the birthrate decline itself and then explores hypotheses as to its cause. I will not attempt to determine whether or not the decline triggers the conditions of genocide under international law; that will be the subject of a separate post. Much of this draws on primary sources identified in the Zenz report, some of which are provided in side-by-side translation for you to judge yourself. A few arguments by Zenz are made based on data leaked to the author; because I cannot directly authenticate those myself, I ignore them here.

Section I: These numbers are suspect

Not all births are created equal

First and foremost, it's important to understand that Uyghur-majority areas have seen much larger drops than Han-majority areas of Xinjiang. Examining to official government publications, we can see that Hotan Prefecture, which is roughly 96% Uyghur, plummeted from a birthrate of 20.94‰ in 2016 to 8.58‰ in 2018, resulting in a natural growth rate of 2.96‰. Similarly, the birthrate in Kashgar, which is 92.6% Uyghur, dropped from 18.19‰ in 2016 to 7.94‰ in 2018. (Compare with official concern over birthrates in the overwhelmingly Han northeast.) This means that the birthrate decline among Uyghurs is even higher than the province-wide birthrate decline.

Zenz provides further evidence to suggest the the birth control policy is applied or enforced more strictly in Uyghur-majority areas. For example, Gulbagh District of Hotan City, one of the few majority-Han (54%) areas in the prefecture, saw a growth rate of 15.17‰ and birth rate of 16.35‰ in 2019. The district government explicitly stated that the district's family planning/birth control policy compliance rate was 100% that year. Among the 22,469 Han residents, there were 388 births—a birthrate of 17.3‰, over double that of Uyghurs.

The remainder of this section is only relevant if you believe the 48.7% birthrate drop in Xinjiang over two years is not particularly noteworthy. The following sections take for granted that such a drop is highly unusual; skip ahead if you don't need convincing.

Xinjiang's figures are unusual for China and the world

The below thread helps quickly contextualize the birthrate drop in Xinjiang in terms of all of China's provinces and regions since 2010, and includes a Github link to the underlying data.

I similarly compiled statistical data from 2019 back to 2001 (online copies of all national statistical yearbooks are viewable here) in a Google Spreadsheet here. Based on this data, we can reasonably conclude that Xinjiang's drop in birthrate is notable in two major ways.

First, no region in the last two decades in China has seen anything close to a comparable drop between 2017 and 2019, and only a handful of examples are comparable over the last 20 years:


The pattern holds if we just look at the absolute changes of births per thousand people (remember, these numbers represent year-over-year changes, not the birthrates themselves; thus, +5.00 one year followed by -.10 the next year does not mean the birthrate dropped by 4.9, but rather that it stayed very steady after the initial increase).


Thus, in terms of absolute decline, and decline relative to respective birthrates, Xinjiang stands out among Chinese subnational regions. Other provinces, most visibly in Shandong, do show unusual swings followed by "crashes", but the largest of these are attributed by Chinese demographers to the One-Child Policy being rescinded in 2015-2016. A burst of people who had wanted to have more than one child did, but given how policy still generally only allows two children (and the fact that having two kids in two years is a handful, I'm sure), the rate came back down again quickly. Xinjiang has had no such spike—the crash in birth rates appear out of nowhere.

The second notable aspect of the birthrate drop in Xinjiang is that it is also wildly unusual in a global context. In the spreadsheet linked above, I drew on two sources—data from the World Bank from 1971-2018 and data from International Historical Statistics from 1950-2008—to calculate absolute changes in raw birthrates by country worldwide. In both datasets, the decline of -5.19‰ in Xinjiang from 2017-2018 ranks at least in the 99.99th percentile and the decline of -2.55‰ ranks at least at the 99.8th percentile (World Bank data is more conservative, showing far fewer spikes and crashes) among countries and years for which data is available, over 10,000 combinations in total. Obviously, comparing subnational units to nations might entail some methodological issues, but the core fact remains the same: these numbers are extremely unusual.

Section II: Four potential factors behind the decline

Below I explore four potential causes of the decline in birthrates among non-Han. These are not exhaustive, nor are they mutually exclusive, and their salience may vary from region to region. These causes are:

  1. Stricter enforcement of extant family planning laws (based on: a) economic incentives and penalties in line with other family planning policies in China, b) forced sterilizations and other coercive measures, or c) some combination of the two)
  2. Modernization and the reduction of religious extremism, leading to less conservative views on marriage and procreation
  3. Dispersal of population via reeducation camps and labor transfer schemes
  4. Economic development generally leading to lower average birthrates

The above are drawn primarily in reference to two authors. The first author is Adrian Zenz, whose rather notorious reputation among some on the left precedes him. Zenz is an evangelical and works at the very, very conservative Victims of Communism Memorial Foundation. I strongly disagree with most of Zenz' worldview, but I am an adult capable of judging writing in terms of sourcing and reason. If you're not, then this post is probably not useful to you, and I highly doubt you have read this far anyways (but congrats if you have).

Zenz published a report in June 2020 that attributes the drop to forced sterilizations and other coercive measures ("Sterilizations, IUDs, and Mandatory Birth Control: The CCP's Campaign to Suppress Uyghur Birthrates in Xinjiang"), from which I have pulled various primary source documents for full or partial translation. I do not agree with everything in the report, but I think it is more than adequately sourced to prove that a non-trivial portion of the drop in birthrate is the result of coercive population controls administered against Uyghurs in particular.

The second author is Li Xiaoxia (李晓霞), currently the director of the Ethnic Studies Institute of the Xinjiang Academy of Social Sciences. Li is not a government or Party official, although XJASS is directly administered and subordinate to the Xinjiang Communist Party Central Committee. Her work is widely cited in state media, and in response to Zenz 2020 report, Li published "An Analysis Report on Population Change in Xinjiang" (in English; hereafter Li 2020), which was carried in various official Chinese media outlets. Given my understanding of government policy and the Party line, and her regular citation in authoritative state/Party media, I treat her position as more or less contiguous with the official view of policy in Xinjiang. (It is more articulately rewritten in this piece by Gareth Porter and Max Blumenthal of GrayZone, a small media outlet that functions as a serial apologist for violence committed by basically any country outside of NATO.)

1. Stricter enforcement of family planning policy

Zenz and Li both agree that family planning policy has played a major role in reducing birthrates in Xinjiang. The question is how the policy was enforced. Is the drop in birthrates is due to more uniform enforcement of family planning policy that China has had for decades but under which ethnic minorities were traditionally given much more leeway? Or is it because the government decided it need swift, decisive action to specifically halt Uyghur birthrates, including imposition against the will of patients if necessary?

While many countries have crafted policy to impact birthrates and population growth, China has done so in a way that is arguably without historical parallel via the One-Child Policy, which was functionally abolished in late 2015, replaced with a two-child policy. "One-Child Policy" is something of a misnomer—there were plenty of exceptions and certainly not the official title of any actual policy—but suffice to say, family planning has long been a purview of the state in China. Under One-Child Policy, rural and minority families were generally given higher birth limits than urban and Han families. The OCP was primarily implemented via economic incentives and in particular, fines for violation, but because of pressure for local officials to avoid births "outside of policy" (illegal), forced abortions were sometimes inflicted upon pregnant women.

In 2017, Xinjiang amended its family planning policy to permit people of all ethnic groups to have two children in urban areas and three in rural. This comes in the context of lax enforcement of family planning policies in the region; Li 2020 states 10.5% of births in Kashgar from 1989 to 2014 were "unplanned" (i.e. outside the limits set by family planning policy).

The official account: completely voluntary birth control

Li 2020 argues that family planning in Xinjiang was implemented via enforcement mechanisms that included both economic incentives (subsidies, rewards) and penalties (fines) to promote compliance. She explicitly rejects the notion of forced sterilizations or other such measures. State policy efforts that Li mentions include:

  • Rewards for ethnic minorities who voluntarily have fewer children. Li states that "couples of ethnic minorities in Xinjiang rural areas who are willing to have only one or two children (adopted children included) in their lifetime, are eligible for applying for Certificate of Honor in Practicing Family Planning Policy. They could also receive a bonus of no less than 3000 RMB" from the government.
  • Generally enhanced "quality of civic reproductive health", namely, "constant efforts to improve services in birth control and reproductive health and ... popularizing scientific knowledge in contraception and reproductive health." This includes thousands of sites distributing free condoms and other birth control measures.
  • Notably, prefectural governments specifically promote tubal litigation (surgery that permanently disables the fallopian tubes and thus the possibility of natural pregnancy). Li states "fertile women accept tubal ligation and IUD operation spontaneously."

So far, nothing particularly objectionable—these measures seem similar to what might be offered by a Planned Parenthood in the United States. This only holds up if we take at face value the statement that everyone accepts these procedures "spontaneously", of course.

Note on sterilizations

Zenz convincingly shows through a host of government documents in the appendix that government policy prefers sterilizations (overwhelmingly female) over IUDs. He produces the following graph based off of data from various annual prints of the China Health Statistical Yearbook:


I was unable to locate available online copies of the Yearbook for every year cited; but I did find several years' books here, and on that basis, I was able to partially verify some of the underlying data in the above graph:

Expand this tab for the specific numbers taken from the health yearbooks

2012 yearbook Table 7-6-2: 7.27% all of contraceptive operations in China in 2011 were tubal ligations (1,595,105), and .75% in Xinjiang (2990) (Vasectomies: 196,064 in China/71 in Xinjiang) Table 13-3: Population of China was 1.347 billion, and of Xinjiang 22.09 million

2013 yearbook 8-8-2: 7.18% (1,561,809), .65% (2562) (Vasectomies: 173,231/158) 14-3: 1.354 billion, 22.33 million

2016 yearbook 8-8-2: 5.17% (1,230,805), 1.68% (9,818) (Vasectomies: 149,432/216) 14-4: 1.375 billion, 23.60 million

2018 yearbook 8-8-2: 2.13% (405,648), 4.37% (20,242) (Vasectomies: 21,525/125) 14-4: 1.390 billion, 24.45 million

2019 yearbook (PDF) 8-8-2: 2.19% (404,212), 4.78%* (59,499) (Vasectomies: 53,128/941) 14-4: 1.395 billion, 24.87 million

(Note also that based on the 2019 Xinjiang Statistical Yearbook that non-Han areas have seen a disproportionately large number of birth control procedures.)

Per Table 3-7 and Table 3-10 of XSY 2019

The alternative account: not very voluntary birth control

Zenz and others have presented evidence that family planning policy in Xinjiang relied substantially on forced sterilizations and other contraceptive procedures administered under the threat of force or legal penalty. The evidence, to be frank, is compelling. It is entirely plausible, if not almost certain, that the above measures describes by Li did indeed cause people to voluntarily alter their behavior. But it is almost certain that for others, coercion was employed.

First, though, let's think about the numbers for a minute. Li 2020 says in Uyghur-majority Kashgar Prefecture, 10.5% of births from 1989 to 2014 were in violation of family planning policy. For the sake of simplicity, let's assume that around 10% of all Uyghur births in Xinjiang were similarly "excess" births. Li completely fails to explain how simply enforcing existing birth limits—i.e., preventing these 10% excess births—could lead to a nearly 50% drop in the birthrate in two years. The numbers simply do not add up. And as I discuss later, Li's numbers themselves are actually very questionable, contradicting 2010 Census data, among other things.

(In simplified terms: Say 100 births are allowed in a town of 10,000 (a 10‰ birthrate), but 110 occur (11‰). "Strictly enforcing" birth policy would presumably lead to around 100 births the next year—if there were suddenly only 50 or 60, we'd know there was another factor at play. Li's answer, discussed in later sections, is that people all suddenly changed their minds and became more modern about procreation. Amusingly, she avoids any mention of reeducation camps/"vocational training centers".)

In at least some areas, those who had illegal children were legally required to undergo reeducation and contraceptive measures. This is, by definition, forced.

For example, this regulation was promulgated by Qarqan County in Bayingol Prefecture. This is applied retroactively as well to people who had these violations before July 28, 2017 and are discovered, according to article 14 of the document.

第十七条 自然居民2017年7月28日后违反规定生育子女的处理: (一)违反相关法律法规规定违法生育一个子女的,按照新修订《新疆维吾尔自治区人口与计划生育条例》相关规定,征收社会抚养费[,]违法生育当事人实际收入低于当地上一年人均收入水平的,社会抚养费按3倍征收;相当于当地上一年人均收入水平的,按5倍征收;高于当地上一年人均收入水平的,按8倍征收。在适用幅度时,对隐瞒、隐藏已有子女而违法多生育的,应在幅度的上限或者在最高限额上征收社会抚养费。即:违法当事人实际收入低于当地上一年人均收入水平的,社会抚养费按4倍征收;相当于当地上一年人均收入水平的,按7倍征收;高于当地上一年人均收入水平的,按8倍征收任何単位和个人不得擅自提高或降低征收社会抚养费标准。 (二)违法多生育两个以上子女的,除按照新修订《新疆维吾尔自治区人口与计划生育条例》相关规定逐胎加倍征收社会抚养费外,同时采取长效节育措施并进行职业技能教育培训。

Article 17 The handling of natural residents who give birth to children in violation of regulations after July 28, 2017: (1) For violation of relevant laws and regulations to illegally give birth to a child, in accordance with the relevant provisions of the newly revised "Xinjiang Uyghur Autonomous Region Population and Family Planning Regulations", social support fees will be levied. If the actual income of the party concerned who gave illegal birth is lower than the previous year's local per capita income, social support fees will be levied three-fold; for those equivalent to the local per capita income, five-fold; and for those higher than above the previous year's local per capita income, eight times. Within the applicable ranges, social maintenance fees shall be levied on the upper limit of the range against those who have more children illegally while concealing or hiding their existing children. That is: if the actual income of the offending party is lower than the local per capita income of the previous year, the social maintenance fee will be levied four-fold; if it is equivalent, seven-fold; and if it is higher, eight-fold; no unit or individual shall arbitrarily raise or lower the standard of collecting social maintenance fees. (2) For those who have more than two children illegally, in addition to doubling the social maintenance fee for each child in accordance with the relevant provisions of the newly revised "Regulations on Population and Family Planning of Xinjiang Uyghur Autonomous Region", long-term birth control measures and vocational skills education and training are also to be enacted.

Nilqa County in Ili Prefecture passed similar regulations in late 2019:

第十九条 户籍地应积极配合现居住地督促流动人口育龄夫妻落实相应的避孕节育措施。本县户籍的流动人口已婚育龄妇女,已生育一个子女的首选上环措施,已生育两个以上子女的首选结扎措施。 第三十一条 对政策外怀孕拒不终止妊娠、违法生育拒不缴纳社会抚养费、外出“躲生”的流出人员及政策外怀孕的流入人员,户籍所在地和现居住地及时将相关人员名单报县政法委,由县政法委将名单推送至公安机关。公安机关对相关人员进行集中教育,促使政策外怀孕人员终止妊娠、违法生育人员缴纳社会抚养费,对拒绝、阻碍计划生育工作人员依法执行公务构成违法行为的,加大打击力度。

Article 19 The place of household registration shall actively cooperate with the current place of residence to supervise and urge floating couples [i.e. internal migrants without official, permanent residence] of childbearing age to implement contraceptive measures. For married women of childbearing age with household registration in this county, the first choice for those who have given birth to one child is an IUD, and ligation [i.e. 'tube tying'] for those who have given birth to two or more children. Article 31 Those who refuse to end a pregnancy that is outside policy [allowances], who refuse to pay social support fees for illegal births, and for those who go outside [their locale] to "hide births" or who have come to the county and have pregnancies outside of policy within, the [authorities of their] place of household registration as well as place of current residence shall promptly report their names to the County Political and Legal Affairs Commission, which shall forward the list to the public security organs. For these persons, the public security organs shall initiate concentrated education, induce* those who are pregnant outside policy to terminate their pregnancy, and induce those who illegally give birth to pay social support fees. [Public security organs shall] intensify the crackdown on those who refuse or hinder family planning staff from performing their official duties in accordance with the law. * 促使 can mean urge, induce, promote, etc.

This undermines that argument put forth by Li and the Chinese government, but they are far from the only evidence that complicates the official narrative. Several individuals have provided testimony describing accounts forced contraception, but by nature of the closed system that is the camps in Xinjiang, absolute verification is impossible. For example:

  • An October 2019 article in the Washington Post details statements by Gulzira Mogdyn, who says she was forced to undergo an abortion while under house arrest, and Rakhima Senbay, who says she was fitted with an IUD against her will before going to an reeducation camp.
  • Gulbahar Jalilova states that she was regularly subject to unidentified injections while in a camp; "We soon realised that after our injections that we didn’t get our periods any more.” (per The Independent) (These injections possibly were CICs.)
  • Zumrat Dawut describes undergoing sterilization for fear that refusal would result in detention in a reeducation camp in a November 2019 Post article.
  • In an on-camera interview with British station ITV News in September 2020, a woman in Turkey claiming to be a gynecologist stated that she performed between 500-600 operations, "including forced contraception, forced abortion and forced sterilization, and forced removal of wombs." She did not reveal her name nor show her full face.

As mentioned above, the primary means the government has enacted family planning policy in Xinjiang is through tubal ligation (permanent sterilization surgery) and IUD placement. (IUDs in China are not easily removable; most require in-office procedures because they lack attached strings used elsewhere to facilitate extraction—see e.g. Cheung 2010, MacDonald 2006.) Unless we are to believe that a region with traditionally low use of contraceptives has magically been altered at such a scale to where everyone is now fine with permanent birth control, we must suspect coercion is at play given the above evidence.

Government documents themselves, in fact, freely admit to problems of noncompliance and resistance to mandatory birth control measures, and the fact that such measures are mandatory. For example, this report from Bayingol Commission of Health in 2018 talks about resistance to IUDs (thanks to D. Zhang for assistance translating):

根据2016年我州长效节育措施落实率调查结果及近期自治区卫生计生委副主任周美林对我州库尔勒市、尉犁县、博湖县基层计划生育与妇幼健康工作调研情况,提出我州基层计划生育技术及妇幼健康服务工作存在以下几点问题: 一是基层工作人员对人口条例不熟悉,对群众宣传不到位。育龄妇女对计划生育各项政策不了解,对“查环、查孕、查病”服务抵触情绪大,存在“怕、厌、躲”的现象。[...]

Based on the 2016 survey of implementation rate of long-term birth control measures in the prefecture and the recent investigation by Zhou Meilin, deputy director of the Health and Family Planning Commission of the Autonomous Region, on the grassroots [i.e. local-level] family planning and maternal and child health work in Korla City, Yuli County, and Bohu County, several issues in our grassroots family planning and maternal health services have been raised: First, grassroots staff are not familiar with population regulations and do not promote them to the masses. Women of childbearing age do not understand various policies on family planning, and are highly resistant to the services of "checking for an IUD, checking for pregnancy, and checking for disease disease", and there is "fear, disgust, and evasion". [...]

(二)提高对婚前医学检查、长效避孕措施落实工作紧迫性的认识 从婚前医学检查源头开始,抓住服务对象,孕、产期全程跟踪管理。对育龄妇女顺产后三个月内,剖腹产后六个月内坚决落实放置宫内节育器,做到“应落尽落”。各县市每半年组织农村已婚育龄妇女进行查环查孕,做到“应查尽查”,杜绝政策外怀孕。[...]

(2) Raise awareness of the urgency of implementing pre-marital medical examinations and long-term contraceptive measures Starting from the source of pre-marital medical examinations, we will take hold of the service targets [i.e. patients] and follow along with and manage the whole process of pregnancy and childbirth. For women of childbearing age, within three months after delivery, and within six months after caesarean section, resolutely implement the placement of IUDs, so as to "place all that should be placed." Every six months, counties and cities shall organize rural married women of childbearing age to carry out [intrauterine] ring-checks and pregnancy checks, so as to "check all that should be checked" and prevent pregnancy outside policy. [...]


One [work task] is to integrate the implementation of long-term contraceptive measures. After an IUD check and disease check, all those who meet the conditions for placing IUDs without contraindications are to have IUDs placed immediately. For circumstances that would disallow an IUD, a diagnosis certificate issued by a second-level or higher medical institution must be provided, and follow-up must be strengthened. Ensure that the implementation rate of long-term contraceptive measures exceeds 75%.

It is perfectly acceptable, of course, to try to explain to patients the benefits of an IUD and even encourage its use, but as we see above, the birth control policy as implemented is highly invasive with regular state-mandated checks. Yet authorities and researchers seem loathe to acknowledge that people might refuse birth control for their own reasons. In early 2017, our friend Li Xiaoxia published "An Analysis of Xinjiang Population Problems and Population Policy" (新疆的人口问题及人口政策分析) in the Journal of the Central Institute of Socialism, a periodical managed by the United Front Work Department of the CCP. On page 25, she wrote of her view on why people refused birth control, all of which stem from extremism and separatism, for some reason:


Under the interfering influence of religious extremism and ethnic separatism, some people have come to abide in certain traditional beliefs regarding marriage and birth. One [such beliefs is], they are not willing to use contraceptives. There are many reasons people would not want to, such as believing that pharmaceutical contraceptives can impact a woman's health or ability to have children later; that using external contraceptives can impact the pleasurability of one's sex life; or that the private parts of a woman's body cannot be exposed to strangers. Others with sinister intent greatly exaggerate the extremely low chances of contraceptive measures causing complications, and some outright deny [as ineffective] technical contraceptive measures, causing some of the masses to fear or despise [the measures]. Two, they deny the legitimacy of the state, do not obey regular social management, and do not acquire national marriage, birth, etc. legal documentation. Three, they illegally have children with the goal of 'promoting the development of the nation [i.e. ethnic group]'. While the author is unable to judge the respective extents of these influences, it is clear that they represent obstacles to the implementation of family planning policy.

This rather narrow-minded paragraph serves as a useful transition to the next possible reason behind the drop in birthrates in Xinjiang.

Data errors in Zenz report

As far as I am aware, there are three main charges made against the Zenz report over statistical accuracy. The first of these three is a nonstarter, but the latter two are valid.

Uyghur population

Grayzone pointed to the discrepancy between Zenz' and Li's Xinjiang population figures as evidence that Zenz is lying. Li 2020 claims that from 2010 to 2018:


[T]he population of ethnic minorities increased from 12.9859 million to 15.8608 million, an increase of 2.8749 million, or 22.14%; from the perspective of ethnic groups, the Uygur population increased from 10.1715 million to 12.7184 million, an increase of 2.5469 million, or 25.04% [i.e. 3.2% annualized]; the population of Han ethnic group increased from 8.8299 million to 9.0068 million, an increase of 176,900 people, or 2.0%.

Zenz' graph on page four of his report shows about 10.1 million and 11.7 million Uyghurs, respectively. You can directly confirm Zenz' 2010-2018 numbers in the leftmost column of 2019 Xinjiang Statistical Yearbook table 3-8. Table 3-8 data is taken from Xinjiang's Public Security Annual Report, and is presumably counting hukou (residential registrations that are generally tied to one's hometown). In contrast, Li's figures come from table 3-1, which counts year-end total population, which would include migrant workers and other "floating" population types. The 2019 Yearbook does not break down year-end total population by ethnicity, however; where Li is getting that information from is still unclear. (Thanks to Nectar Gan for pointing this out!)

IUDs per capita

The title in Figure 6 reads "Net added IUD Placements per Capita." Per capita should be per 100,000.

However, after computing the rates for 2011, 2012, 2015, 2017, and 2018, I noticed significant discrepancies between the figures I calculated for Xinjiang and those charted by Zenz (the numbers for China overall were accurate). See an amended chart below:

Expand for the math

Xinjiang: 2012 Yearbook: (195198 IUD inserts - 90706 IUD removals) / 22090000 people in Xinjiang * 100000 = 473 per 100,000 2013: (196894 - 88888) / 22300000 * 100000 = 484/100k 2016: (338640 - 87473) / 236000010 * 100000 = 1064/100k 2018: (268648 - 67786) / 24450000 * 100000 = 822/100k 2019: (328475 - 89018) / 24870000 * 100000 = 963/100k

China: 2012: 7296642 - 2818858 / 1347000000 * 100000 = 332/100k 2016: (8227879 - 3528728) / 1375000000 * 100000 = 342/100k 2018: (4639490 - 3935241) / 1390000000 * 100000 = 51/100k 2019: (3775318 - 3474467) / 1395000000 * 10000 = 22/100k

I have no idea what the source of this discrepancies is; I didn't try to recreate them. My numbers do not change the argument, but they weaken its magnitude with respect to IUD placements. Nevertheless, given the accurate accounting of national net IUD rates, it is worth mentioning.

80% of all IUDs

Zenz states that in 2018 "80 percent of all new IUD placements in China were performed in Xinjiang, despite the fact that the region only makes up 1.8 percent of the nation's population." This is (functionally) untrue. There were 7,551,436 IUDs new IUDs placed in China in 2019 and 6,948,934 removed. Xinjiang had 328,475 new IUDs place and 89018, a net of 239,457. Xinjiang's new IUDs only represent about 4.3% of all IUDs added in China. This is still relatively high (Xinjiang is less than 2% of China’s population) but definitely not 80% high.

Xinjiang had 239,457 net IUDs placed in 2018; China, on the whole, had 301,651. It is true that Xinjiang's net IUDs are about 80% of the nation's net IUD figure, but this is because there were provinces and regions with negative net IUDs. Guizhou had 109,021 added, or about 36% of the net. See the issue? Xinjiang and Guizhou alone already exceed 100% of China's net IUDs. That's because you can't compare bare net numbers, as the net has no relation on the magnitude of insertions or removals per province.

(Imagine businesses in a shopping center have a total net profit of $100. Businesses A, B, and C all had net profits of $50; Business D lost $50. While it is literally true that Businesses A, B, and C each earned 50% of the center’s $100—$50 is half of $100—that does not mean they contributed 50% to the net profit.)

Still, a strong argument can be made looking at net IUDs per 100,000 in 2018, as is charted below (spreadsheet here):


China as a whole averaged a net of +21.6 IUDs per 100,000 people; in Xinjiang, it was +963, over three times greater than the runner-up province, Guizhou, at +303. As detailed in subsection 4 below, Guizhou had a higher birthrate than Xinjiang as of the 2010 census.

The 80% claim is based on shaky math and unclear wording. Nevertheless, it is still true that Xinjiang's IUD insertion rates are significantly greater than China as a whole.

2. A decline in extremism and changes in mindset

Li 2020 says that another reason for the decline in birth rates is a shift in "the mindset on marriage and procreation." Specifically, "In [recent] years, young people in Xinjiang have already discarded the backward and outdated thoughts on mate selection and procreation. An increasing number of ethnic minority youths began to put more time and energy on personal development, [and] thus postpone the time of first marriage."

Relatedly, she gives the third reason for the decline as the effective "war against extremism" that has facilitated the "idea of civilized modern life [winning] popular support from the public." Li states that women previously oppressed by extremist thought have become "emancipated" by "the scientific consciousness of gender equality and reproductive health," thus ensuring that "women [are] not being reproduction machines any more." These two factors are functionally identical for the purposes of argument.

This argument is actually partially valid, but not in the way Li tries to present it. As discussed previously, the decline in birthrates in Xinjiang is highly unusual; unless we are to believe everyone spontaneously became much more liberal in their view of birth control, this explanation makes little sense. The "change in mindset" did not appear out of nowhere, of course. The war on extremism Li mentions refers to none other than...

3. Reeducation and mass incarceration

Li 2020 avoids any mention of "vocational training centers," i.e., the mass internment/reeducation facilities, nor of (likely forced) labor transfer schemes or imprisonment in Xinjiang. All of these could plausibly impact birthrates; it may simply be that there are so many people in the camps, prisons, and/or who were transferred out of province on labor assignments that it has "naturally" slowed birthrates. Admitting as much would represent a possible contradiction, however, of the official Chinese narrative: for drops in birth rates that large, it would mean there were huge numbers of people in the camps, certainly over the one million figure derided by the government.

Additionally, as has been established in the previous sections of this blog, mass internment/reeducation is an extensive phenomenon, and it is involuntary. A decline in birthrates, by extension, would be involuntary.

One bit of evidence in support of this possibility came from my examination of the various editions of the China Health Statistical Yearbook to corroborate Zenz' graph in the previous section about tubal litigation and sterilization in Xinjiang. The Yearbook records five types of contraceptive procedures: IUD implant, IUD removal, vasectomy, tubal litigation, and abortion. It breaks down the composition of all procedures (what percent are IUD implants, removals, vasectomies, etc.) for each province, which usually add up to 98% or greater (I don't know what other type of procedure they are electing to leave out).

The 2019 Yearbook contraceptive data for Xinjiang is unusual. While it records 1,244,886 instances of contraceptive operations (including 89,018 IUD removals—on a side note, at 7.15%, this is a proportion far lower than pre-2017 averages, wherein removals made up 20-30% of all contraceptive operations), it is only able to account for 45.09% of them in its composition data (cf. 95.05% for the 2018 Yearbook, 92.04% for 2016, 88.26% for 2013, 87.6% for 2011). The only area with comparably low composition data is Tibet, which had less than 3% of the total operations Xinjiang did in 2019, and unlike Xinjiang, historically has had incomplete composition data. This is a pattern we might expect to see if large numbers of procedures suddenly were performed outside of normal clinical contexts or where data is otherwise suppressed. This is not definitive evidence by any means, but it is something we might expect to see if sterilizations were suddenly occurring outside of normal clinical contexts.

4. Economics (and the real birthrates)

The final section of Li's paper notes the well established fact that, in general, as areas develop, their fertility rates drop. She concludes that "the transition of Uygur population is achieved more by their own choice and autonomy than the guidance of policies and external forces. The transition is not only a matter of population size, but also a subject involving the overall improvement of population quality, which is the result of voluntary choice of people of all ethnic minorities."

Of course, as the data in Section I from the World Bank and Mitchell clearly demonstrates, for the last half century or so declines in birth rates among developing countries almost never reach what has been seen in Xinjiang since 2017. We can also turn to Li's 2017 paper, where she notes that birthrates in Xinjiang have indeed declined significantly—over the space of several decades.

据 1988 年全国生育节育抽样调查(2‰抽样调查),1970-1974 年维吾尔族妇女平均预期终身生育 6.13 个孩子,1985~1987 年则为 5.4 个孩子。同类数据同地汉族妇女分别为 5.2 个孩子和 2.5 个孩子。1981 年我国分民族人口总和生育率最高的是藏族(5.84)和维吾尔族(5.59);1989 年总和生育率最高的是维吾尔族(4.65)和藏族(3.80);2000 年最高的分别是哈萨克族(3.195)、维吾尔族(3.156)、藏族(2.755)。

According to the National Survey on Birth and Birth Control in 1988 (2‰ sampling), Uyghur women were expected to give birth to 6.13 children on average from 1970 to 1974, and 5.4 children from 1985 to 1987. Han women in the same locations had 5.2 children and 2.5 children, respectively. In 1981, Tibetans (5.84) and Uyghurs (5.59) had the highest total fertility rates of any ethnic group in China; Uyghurs (4.65) and Tibetans (3.80) had the highest total fertility rates* in 1989; and the highest in 2000 were Kazakh (3.195), Uygur (3.156), and Tibetan (2.755).

* Total fertility rate refers to how many children a woman would have on average over her lifetime in a population. Birthrate refer to births in relation to the entire population.

Li never directly answers the question in her response to Zenz as to why, all of a sudden, the government decided that it needed Uyghur birthrates to come down immediately. In fact, Li selectively leaves out newer data when she reiterates the above statistics in her 2020 essay. Government statistics show the Uyghur birthrates have been below replacement level since the mid-2000s. (Populations can keep growing for a generation or so after this occurs; see population momentum.)

Xinjiang's 2010 census shows a fertility rate of 1.53 in 2010 (chart 6-4 on page 1823), far too low to justify such stringent birth control policy. I am unsure where Li's 2000 figure of 3.151 average births per Uyghur woman comes from. Wang 2018 ("Analysis of Changes and Influencing Factors in Xinjiang Fertility Rate Levels" (新疆人口生育水平的变化及影响因素分析)) similarly states Xinjiang overall had a fertility rate of 2.56 in 2000 and 1.53 in 2010, and notes that two areas, Guangxi and Guizhou (both where Han represent about two-thirds of the population), had higher rates in 2010.* With the sole exception of Turpan, all prefectures and prefecture-level cities in Xinjiang had fertility rates below 2.0. This means that birth control measures had already been implemented and "mindsets" had already "changed" in Xinjiang well before 2017. There is little reason to believe birth control policy in Xinjiang is suddenly being implemented more stringently just as a matter of "reasonable" family planning concerns. The evident goal is the prevention of births among Uyghurs.


We have established that Xinjiang has undergone an unprecedentedly large drop in birthrates concentrated among non-Han, and in particular, Uyghurs. This drop is not merely the result of bringing birthrates in line with formal policy; Xinjiang has for over a decade had below-replacement fertility rates, and mathematically, preventing excess births still cannot account for the 48.7% decline from 2017-2019. Some testimony from previous detainees and Xinjiang residents, plus significant direct and circumstantial evidence, further demonstrates that the drop in birthrate is very likely the result of deliberate and coercive policy implemented by the government, which does not reflect the general will of the Uyghur population. Moreover, it does not appear that the nominal justification put forth by the government—to bring down unsustainably high birthrates—can sufficiently explain this decline. The most immediate explanation is that the government seeks to arrest growth of Uyghurs, but at the same time, the disparate rates of implementation across provinces, including Uyghur-majority ones, suggests that this might not be centrally planned policy per se, but rather varied reactions to a general central mandate.

The decline in birthrates must be understood in the context of concurrent policy in Xinjiang. For the rest of the country, restrictions on birthrates have been relaxed, but as authors like Li note, too many Uyghurs represents a threat to stability. (See also the post on colonialism for more on this, including the role of the Xinjiang Production and Construction Corps in maintaining a Han bulwark in the region).

Accusations of deliberate birthrate suppression are severe. This would directly implicate one of the conditions of the Convention on the Prevention and Punishment of the Crime of Genocide, a treaty passed by the United Nations in 1948 and ratified by almost all countries, including China. Article II of the treaty:

In the present Convention, genocide means any of the following acts committed with intent to destroy, in whole or in part, a national, ethnical, racial or religious group, as such: (a) Killing members of the group; (b) Causing serious bodily or mental harm to members of the group; (c) Deliberately inflicting on the group conditions of life calculated to bring about its physical destruction in whole or in part; (d) Imposing measures intended to prevent births within the group; (e) Forcibly transferring children of the group to another group.

Whether or not genocide is a useful or accurate term for what is occurring in Xinjiang is not the subject of this post, but if we are to go by the Convention, it is very much a possibility that the term applies. Even so, we should also note that the Chinese government in the 20th century implemented highly repressive birth control policies against the Han majority, even exceeding the rate of sterilization per 100,000 that we see in Xinjiang today. Regardless, it certainly qualifies as a crime against humanity, particularly in the context of widespread the mass internment/reeducation campaign.