SOURCE
https://jamestown.org/program/jamestown-early-warning-brief-xinjiangs-system-of-militarized-vocational-training-comes-to-tibet/?mc_cid=596b89b4e2&mc_eid=2aedc83db6
TIBET: DEATH OF A NATION
Xinjiang’s System of Militarized Vocational Training Comes to Tibet
By
September 22, 2020
Image 1: Military-style training of “rural surplus laborers” in the Chamdo region of Tibet, June 2016. (Image source: Tibet’s Chamdo, June 30, 2016).
Introduction and Summary
In 2019 and 2020, the Tibet Autonomous Region (TAR) introduced new policies to promote the systematic, centralized, and large-scale training and transfer of “rural surplus laborers” to other parts of the TAR, as well as to other provinces of the People’s Republic of China (PRC). In the first 7 months of 2020, the region had trained over half a million rural surplus laborers through this policy. This scheme encompasses Tibetans of all ages, covers the entire region, and is distinct from the coercive vocational training of secondary students and young adults reported by exile Tibetans (RFA, October 29, 2019).
The labor transfer policy mandates that pastoralists and farmers are to be subjected to centralized “military-style” (军旅式, junlüshi) vocational training, which aims to reform “backward thinking” and includes training in “work discipline,” law, and the Chinese language. Examples from the TAR’s Chamdo region indicate that the militarized training regimen is supervised by People’s Armed Police drill sergeants, and training photos published by state media show Tibetan trainees dressed in military fatigues (see accompanying images).
Poverty alleviation reports bluntly say that the state must “stop raising up lazy people.” Documents state that the “strict military-style management” of the vocational training process “strengthens [the Tibetans’] weak work discipline” and reforms their “backward thinking.” Tibetans are to be transformed from “[being] unwilling to move” to becoming willing to participate, a process that requires “diluting the negative influence of religion.” This is aided by a worrisome new scheme that “encourages” Tibetans to hand over their land and herds to government-run cooperatives, turning them into wage laborers.
An order-oriented, batch-style matching and training mechanism trains laborers based on company needs. Training, matching and delivery of workers to their work destination takes place in a centralized fashion. Recruitments rely, among other things, on village-based work teams, an intrusive social control mechanism pioneered in the TAR by Chen Quanguo (陈全国), and later used in Xinjiang to identify Uyghurs who should be sent to internment camps (China Brief, September 21, 2017). Key policy documents state that cadres who fail to achieve the mandated quotas are subject to “strict rewards and punishments” (严格奖惩措施, yange jiangcheng cuoshi). The goal of the scheme is to achieve Xi Jinping’s signature goal of eradicating absolute poverty by increasing rural disposable incomes. This means that Tibetan nomads and farmers must change their livelihoods so that they earn a measurable cash income, and can therefore be declared “poverty-free.”
This draconian scheme shows a disturbing number of close similarities to the system of coercive vocational training and labor transfer established in Xinjiang. The fact that Tibet and Xinjiang share many of the same social control and securitization mechanisms—in each case introduced under administrations directed by Chen Quanguo—renders the adaptation of one region’s scheme to the other particularly straightforward.
Historical Context
As early as 2005, the TAR had a small-scale rural surplus labor training and employment initiative for pastoralists and farmers in Lhasa (Sina, May 13, 2005). The 11th Five-Year Plan (2006-2010) then specified that this type of training and labor transfer was to be conducted throughout the TAR (PRC Government, February 8, 2006). From 2012, the Chamdo region initiated a “military-style training for surplus labor force transfer for pastoral and agricultural regions” (农牧区富余劳动力转移就业军旅式培训, nongmuqu fuyu laodongli zhuanyi jiuye junlüshi peixun) (Tibet’s Chamdo, October 8, 2014). Chamdo’s scheme was formally established in the region’s 13th Five-Year Plan (2016-2020), with the goal of training 65,000 laborers (including urban unemployed persons) during that time (Chamdo Government, December 29, 2015).
By 2016, Chamdo had established 45 related vocational training bases (TAR Government, November 17, 2016). Starting in 2016, the TAR’s Shannan region likewise implemented vocational training with “semi-military-style management” (半军事化管理, ban junshihua guanli) (Tibet Shannan Net, April 5, 2017). Several different sources indicate that Chamdo’s military-style training management was conducted by People’s Armed Police drill sergeants.[1]
Policies of the 2019-2020 Militarized Vocational Training and Labor Transfer Action Plan
In March 2019, the TAR issued the 2019-2020 Farmer and Pastoralist Training and Labor Transfer Action Plan (西藏自治区2019-2020年农牧民培训和转移就业行动方案, Xizang Zizhiqu 2019-2020 Nian Nongmumin Peixun he Zhuanyi Jiuye Xingdong Fang’an) which mandates the “vigorous promotion of military-style…[vocational] training,” adopting the model pioneered in Chamdo and mandating it throughout the region. [2] The vocational training process must include “work discipline, Chinese language and work ethics,” aiming to “enhance laborers’ sense of discipline to comply with national laws and regulations and work unit rules and regulations.”
Surplus labor training is to follow the “order-oriented” (订单定向式, dingdan dingxiangshi) or “need-driven” (以需定培, yi xu dingpei) method, [3] whereby the job is arranged first, and the training is based on the pre-arranged job placement. In 2020, at least 40 percent of job placements were to follow this method, with this share mandated to exceed 60 percent by the year 2024 (see [2], also below). Companies that employ a minimum number of laborers can obtain financial rewards of up to 500,000 renminbi ($73,900 U.S. dollars). Local labor brokers receive 300 ($44) or 500 ($74) renminbi per arranged labor transfer, depending whether it is within the TAR or without. [4] Detailed quotas not only mandate how many surplus laborers each county must train, but also how many are to be trained in each vocational specialty (Ngari Government, July 31, 2019).
The similarities to Xinjiang’s coercive training scheme are abundant: both schemes have the same target group (“rural surplus laborers”—农牧区富余劳动者, nongmuqu fuyu laodongzhe); a high-powered focus on mobilizing a “reticent” minority group to change their traditional livelihood mode; employ military drill and military-style training management to produce discipline and obedience; emphasize the need to “transform” laborers’ thinking and identity, and to reform their “backwardness;” teach law and Chinese; aim to weaken the perceived negative influence of religion; prescribe detailed quotas; and put great pressure on officials to achieve program goals. [5]
Images: Examples of “military-style” vocational training for ethnic Tibetans in the Chamdo region. / Figure 2 (left): Tibetans dressed in military fatigues practice painting. (Image source: Tibet’s Chamdo, June 30, 2016). / Figure 3 (right): Tibetan women in military fatigues are trained how to be restaurant waitresses. (Image source: Sina, July 27, 2020)
Labor Transfers to Other Provinces in 2020
In 2020, the TAR introduced a related region-wide labor transfer policy that established mechanisms and target quotas for the transfer of trained rural surplus laborers both within (55,000) and without (5,000) the TAR (TAR Human Resources Department, July 17). The terminology is akin to that used in relation to Xinjiang’s labor transfers, employing phrases such as: “supra-regional employment transfer” (跨区域转移就业, kuaquyu zhuanyi jiuye) and “labor export” (劳务输出, laowu shuchu). Both the 2019-2020 Training and Labor Transfer Action Plan and the TAR’s 13th Five-Year Plan (2016-2020) only mention transfers outside the TAR in passing, without outlining a detailed related policy or the use of terminology akin to that found in related documents from Xinjiang. [6]
In the first 7 months of 2020, the TAR trained 543,000 rural surplus laborers, accomplishing 90.5% of its annual goal by July. Of these, 49,900 were transferred to other parts of the TAR, and 3,109 to other parts of China (TAR Government, August 12). Each region is assigned a transfer quota. By the end of 2020, this transfer scheme must cover the entire TAR.
Specific examples of such labor transfers identified by the author to other regions within the TAR include job placements in road construction, cleaning, mining, cooking and driving. [7] Transfers to labor placements outside the TAR include employment at the COFCO Group, China’s largest state-owned food-processing company (Hebei News, September 18, 2020).
The central terminology employed for the labor transfer process is identical with language used in Xinjiang: “unified matching, unified organizing, unified management, unified sending off” (统一对接、统一组织、统一管理、统一输送 / tongyi duijie, tongyi zuzhi, tongyi guanli, tongyi shusong). [8] Workers are transferred to their destination in a centralized, “group-style” (组团式, zutuanshi), “point-to-point” (点对点, dianduidian) fashion. The policy document sets group sizes at 30 persons, divided into subgroups of 10, both to be headed by (sub-)group leaders (TAR Human Resources Department, July 17). In one instance, this transport method was described as “nanny-style point-to-point service” (“点对点”“保姆式”服务 / “dianduidian” “baomu shi” fuwu) (Chinatibet.net, June 21). As in Xinjiang, these labor transfers to other provinces are arranged and supported through the Mutual Pairing Assistance [or “assist Tibet” (援藏, Yuan Zang)] mechanism, albeit not exclusively. [9] The transferred laborers’ “left-behind” children, wives and elderly family members are to receive the state’s “loving care.” [10]
Again, the similarities to Xinjiang’s inter-provincial transfer scheme are significant: unified processing, batch-style transfers, strong government involvement, financial incentives for middlemen and for participating companies, and state-mandated quotas. However, for the TAR’s labor transfer scheme, there is so far no evidence of accompanying cadres or security personnel, of cadres stationed in factories, or of workers being kept in closed, securitized environments at their final work destination. It is possible that the transfer of Tibetan laborers is not as securitized as that of Uyghur workers. There is also currently no evidence of TAR labor training and transfer schemes being linked to extrajudicial internment. The full range of TAR vocational training and job assignment mechanisms can take various forms and has a range of focus groups; not all of them involve centralized transfers or the military-style training and transfer of nomads and farmers.
The Coercive Nature of the Labor Training and Transfer System
Even so, there are clear elements of coercion during recruitment, training and job matching, as well as a centralized and strongly state-administered and supervised transfer process. While some documents assert that the scheme is predicated on voluntary participation, the overall evidence indicates the systemic presence of numerous coercive elements.
As in Xinjiang, TAR government documents make it clear that poverty alleviation is a “battlefield,” with such work to be organized under a military-like “command” structure (脱贫攻坚指挥部, tuopin gongjian zhihuibu) (TAR Government, October 29, 2019; Xinhua, October 7, 2018). In mid-2019, the battle against poverty in the TAR was said to have “entered the decisive phase,” given the goal to eradicate absolute poverty by the end of 2020 (Tibet.cn, June 11, 2019). Since poverty is measured by income levels, and labor transfer is the primary means to increase incomes—and hence to “lift” people out of poverty—the pressure for local governments to round up poor populations and feed them into the scheme is extremely high.
The Training and Labor Transfer Action Plan cited above establishes strict administrative procedures, and mandates the establishment of dedicated work groups as well as the involvement of top leadership cadres, to “ensure that the target tasks are completed on schedule” (see [2]). Each administrative level is to pass on the “pressure [to achieve the targets] to the next [lower] level.” Local government units are to “establish a task progress list [and] those who lag behind their work schedule… are to be reported and to be held accountable according to regulations.” The version adopted by the region governed under Shannan City is even more draconian: training and labor transfer achievements are directly weighed in cadres’ annual assessment scores, complemented by a system of “strict rewards and punishments.” [11] Specific threats of “strict rewards and punishments” in relation to achieving labor training and transfer targets are also found elsewhere, such as in official reports from the region governed under Ngari City, which mandate “weekly, monthly and quarterly” reporting mechanisms (TAR Government, December 18, 2018).
As with the Uyghurs in Xinjiang, overcoming Tibetans’ resistance to labor transfer is an integral part of the entire mechanism. Documents state that the “strict military-style management” of the vocational training process causes the “masses to comply with discipline,” “continuously strengthens their patriotic awareness,” and reforms their “backward thinking.” [12] This may also involve the presence of local cadres to “make the training discipline stricter.” [13]
Because the military-style vocational training process produces discipline and transforms “backward employment views,” it is said to “promote labor transfer.” [14] Rural laborers are to be transformed from “[being] unwilling to move” to becoming willing to participate, a process that requires “diluting the negative influence of religion,” which is said to induce passivity (TAR Commerce Department, June 10). The poverty alleviation and training process is therefore coupled with an all-out propaganda effort that aims to use “thought education” to “educate and guide the unemployed to change their closed, conservative and traditional employment mindset” (Tibet’s Chamdo, July 8, 2016). [15] One document notes that the poverty alleviation and labor transfer process is part of an effort to “stop raising up lazy people” (TAR Government, December 18, 2018).
A 2018 account from Chamdo of post-training follow-up shows the tight procedures employed by the authorities:
Strictly follow up and ask for effectiveness. Before the end of each training course, trainees are required to fill in the “Employment Willingness Questionnaire.” Establish a database…to grasp the employment…status of trainees after the training. For those who cannot be employed in time after training, follow up and visit regularly, and actively recommend employment…. [16]
These “strict” follow-up procedures are increasingly unnecessary, because the mandated “order-oriented” process means that locals are matched with future jobs prior to the training.
“Grid Management” and the “Double-Linked Household” System
Coercive elements play an important role during the recruitment process. Village-based work teams, an intrusive social control mechanism pioneered by Chen Quanguo, go from door to door to “help transform the thinking and views of poor households.” [17] The descriptions of these processes, and the extensive government resources invested to ensure their operation, overlap to a high degree with those that are commonly practiced in Xinjiang (The China Quarterly, July 12, 2019). As is the case in Xinjiang, poverty-alleviation work in the TAR is tightly linked to social control mechanisms and key aspects of the security apparatus. To quote one government document, “By combining grid management and the ‘double-linked household’ management model, [we must] organize, educate, and guide the people to participate and to support the fine-grained poverty alleviation … work.” [18]
Grid management (网格化管理, wanggehua guanli) is a highly intrusive social control mechanism, through which neighborhoods and communities are subdivided into smaller units of surveillance and control. Besides dedicated administrative and security staff, this turns substantial numbers of locals into “volunteers,” enhancing the surveillance powers of the state. [19] Grid management later became the backbone of social control and surveillance in Xinjiang. For poverty alleviation, it involves detailed databases that list every single person “in poverty,” along with indicators and countermeasures, and may include a “combat visualization” (图表化作战, tubiaohua zuozhan) feature whereby progress in the “war on poverty” is visualized through maps and charts (TAR Government, November 10, 2016). Purang County in Ngari spent 1.58 million renminbi ($233,588 dollars) on a “Smart Poverty Alleviation Big Data Management Platform,” which can display poverty alleviation progress on a large screen in real time (TAR Government, February 20, 2019).
Similarly, the “double-linked household” (双联户, shuang lian hu) system corrals regular citizens into the state’s extensive surveillance apparatus by making sets of 10 “double-linked” households report on each other. Between 2012 and 2016, the TAR established 81,140 double-linked household entities, covering over three million residents, and therefore virtually the region’s entire population (South China Morning Post, December 12, 2016). An August 2020 article on poverty alleviation in Ngari notes that it was the head of a “double-linked” household unit who led his “entire village” to hand over their grassland and herds to a local husbandry cooperative (Hunan Government, August 20).
Converting Property to Shares Through Government Cooperatives
A particularly troubling aspect of the Training and Labor Transfer Action Plan is the directive to promote a “poverty alleviation industry” (扶贫产业, fupin chanye) scheme by which local nomads and farmers are asked to hand over their land and herds to large-scale, state-run cooperatives (农牧民专业合作社, nongmumin zhuanye hezuoshe). [20] In that way, “nomads become shareholders” as they convert their usage rights into shares. This scheme, which harks back to the forced collectivization era of the 1950s, increases the disposable incomes of nomads and farmers through share dividends and by turning them into wage laborers. They are then either employed by these cooperatives or are now “free” to participate in the wider labor transfer scheme. [21] In Nagqu, this is referred to as the “one township one cooperative, one village one cooperative ” (“一乡一社”“一村一合” / “yixiang yishe” “yicun yihe”) scheme, indicating its universal coverage. [22] One account describes the land transfer as prodding Tibetans to “put down the whip, walk out of the pasture, and enter the [labor] market” (People.cn, July 27, 2020).
Clearly, such a radical transformation of traditional livelihoods is not achieved without overcoming local resistance. A government report from Shuanghu County (Nagqu) in July 2020 notes that:
In the early stages, … most herders were not enthusiastic about participating. [Then], the county government…organized…county-level cadres to deeply penetrate township and village households, convening village meetings to mobilize people, insisted on transforming the [prevailing attitude of] “I am wanted to get rid of poverty” to “I want to get rid of poverty” as the starting point for the formation of a cooperative… [and] comprehensively promoted the policy… Presently… the participation rate of registered poor herders is at 100 percent, [that] of other herders at 97 percent. [23]
Importantly, the phrase “transforming [attitudes of] ‘I am wanted to get rid of poverty’ to ‘I want to get rid of poverty’” is found in this exact form in accounts of poverty alleviation through labor transfer in Xinjiang. [24]
Given that this scheme severs the long-standing connection between Tibetans and their traditional livelihood bases, its explicit inclusion in the militarized vocational training and labor transfer policy context is of great concern.
Images: Different views of the “Chamdo Golden Sunshine Vocational Training School” in the Chamdo region of eastern Tibet. / Figure 4 (top): The facility at ground level. (Image source: https://bit.ly/2Rr6Ekc) / Figures 5 and 6 (below): Satellite views of the facility. (Source: Google Earth, image dates: 11/22/2018)
Militarized Vocational Training: Examining a Training Base in Chamdo
The Chamdo Golden Sunshine Vocational Training School (昌都市金色阳光职业培训学校, Changdushi Jinse Yangguang Zhiye Peixun Xuexiao) operates a vocational training base within Chamdo’s Vocational and Technical School, located in Eluo Town, Karuo District. The facility conducts “military-style training” (军旅式培训, junlüshi peixun) of rural surplus laborers for the purpose of achieving labor transfer; photos of the complex show a rudimentary facility with rural Tibetan trainees of various ages, mostly dressed in military fatigues. [25]
Satellite imagery (see accompanying images) shows that after a smaller initial setup in 2016, [26] the facility was expanded in the year 2018 to its current state. [27] The compound is fully enclosed, surrounded by a tall perimeter wall and fence, and bisected by a tall internal wire mesh fence that separates the three main northern buildings from the three main southern ones (building numbers 4 and 5 and parts of the surrounding wall are shown in the accompanying Figure 4). The internal fence might be used to separate dormitories from teaching and administrative buildings. Independent experts in satellite analysis contacted by the author estimated the height of the internal fence at approximately 3 meters. The neighboring vocational school does not feature any such security measures.
Conclusions
In both Xinjiang and Tibet, state-mandated poverty alleviation consists of a top-down scheme that extends the government’s social control deep into family units. The state’s preferred method to increase the disposable incomes of rural surplus laborers in these restive minority regions is through vocational training and labor transfer. Both regions have by now implemented a comprehensive scheme that relies heavily on centralized administrative mechanisms; quota fulfilment; job matching prior to training; and a militarized training process that involves thought transformation, patriotic and legal education, and Chinese language teaching.
Important differences remain between Beijing’s approaches in Xinjiang and Tibet. Presently, there is no evidence that the TAR’s scheme is linked to extrajudicial internment, and aspects of its labor transfer mechanisms are potentially less coercive. However, in a system where the transition between securitization and poverty alleviation is seamless, there is no telling where coercion stops and where genuinely voluntary local agency begins. While some Tibetans may voluntarily participate in some or all aspects of the scheme, and while their incomes may indeed increase as a result, the systemic presence of clear indicators of coercion and indoctrination, coupled with profound and potentially permanent change in modes of livelihood, is highly problematic. In the context of Beijing’s increasingly assimilatory ethnic minority policy, it is likely that these policies will promote a long-term loss of linguistic, cultural and spiritual heritage.
Adrian Zenz is a Senior Fellow in China Studies at the Victims of Communism Memorial Foundation, Washington, D.C. (non-resident), and supervises PhD students at the European School of Culture and Theology, Korntal, Germany. His research focus is on China’s ethnic policy, public recruitment in Tibet and Xinjiang, Beijing’s internment campaign in Xinjiang, and China’s domestic security budgets. Dr. Zenz is the author of Tibetanness under Threat and co-editor of Mapping Amdo: Dynamics of Change. He has played a leading role in the analysis of leaked Chinese government documents, to include the “China Cables” and the “Karakax List.” Dr. Zenz is an advisor to the Inter-Parliamentary Alliance on China, and a frequent contributor to the international media.
Notes
[1] See for example https://archive.is/wip/4ItV6 or http://archive.is/RVJRK. State media articles from September 2020 indicate that this type of training is ongoing https://archive.is/e1XqL.
[2] Chinese: 大力推广军旅式…培训 (dali tuiguang junlüshi…peixun). See https://bit.ly/3mmiQk7 (pp.12-17). See local implementation documents of this directive from Shannan City (https://bit.ly/32uVlO5, pp.15-24), Xigatse (https://archive.is/7oJ7p) and Ngari (https://archive.is/wip/R3Mpw).
[3] See also https://archive.is/wip/eQMGa.
[4] Provided that the person was employed for at least 6 months in a given year. Source: https://archive.is/KE1Vd.
[5] See the author’s main work on this in section 6 of: “Beyond the Camps: Beijing’s Long-Term Scheme of Coercive Labor, Poverty Alleviation and Social Control in Xinjiang,” Journal of Political Risk (Vol. 7, No. 12), December 2019. https://www.jpolrisk.com/beyond-the-camps-beijings-long-term-scheme-of-coercive-labor-poverty-alleviation-and-social-control-in-xinjiang/.
[6] See https://archive.is/wip/Dyapm.
[7] See https://archive.is/wip/XiZfl, https://archive.is/RdnvS, https://archive.is/w1kfx, https://archive.is/wip/NehA6, https://archive.is/wip/KMaUo, https://archive.is/wip/XiZfl, https://archive.is/RdnvS, https://archive.is/w1kfx.
[8] See https://archive.is/KE1Vd and https://archive.is/wip/8afPF.
[9] See https://archive.is/KE1Vd and https://archive.is/wip/8afPF.
[10] See https://archive.is/KE1Vd.
[11] See https://bit.ly/32uVlO5, p.24.
[12] See https://archive.is/wip/fN9hz and https://archive.is/NYMwi, compare https://archive.is/wip/iiF7h and http://archive.is/Nh7tT.
[13] See https://archive.is/wip/kQVnX. A state media account of Tibetan waiters at a tourism-oriented restaurant in Xiexong Township (Chamdo) notes that these are all from “poverty-alleviation households,” and have all gone through “centralized, military-style training.” Consequently, per this account, they have developed a “service attitude of being willing to suffer [or: work hard]”, as is evident from their “vigorous pace and their [constant] shuttling back and forth” as they serve their customers. https://archive.is/wip/Nfxnx (account from 2016); compare https://archive.is/wip/dTLku.
[14] See https://archive.is/wip/faIeL and https://archive.is/wip/18CXh.
[15] See https://archive.is/iiF7h.
[16] See https://archive.is/wip/ETmNe
[17] See https://archive.is/wip/iEV7P, see also e.g. https://archive.is/wip/1p6lV.
[18] See https://archive.is/e45fJ.
[19] See https://www.cambridge.org/core/journals/china-quarterly/article/securitizing-xinjiang-police-recruitment-informal-policing-and-ethnic-minority-cooptation/FEEC613414AA33A0353949F9B791E733 and https://www.hrw.org/news/2013/03/20/china-alarming-new-surveillance-security-tibet.
[20] E.g. https://archive.is/R3Mpw. This scheme was also mentioned in the TAR’s 13th 5-Year-Plan (2016-2020) (https://archive.is/wip/S3buo). See also similar accounts, e.g. https://archive.is/IJUyl.
[21] Note e.g. the sequence of the description of these cooperatives followed by an account of labor transfer (https://archive.is/gIw3f).
[22] See https://archive.is/wip/gIw3f or https://archive.is/wip/z5Tor or https://archive.is/wip/PR7lh.
[23] See https://archive.is/wip/85zXB.
[24] See the author’s related work on this in section 2.2 of: “Beyond the Camps: Beijing’s Long-Term Scheme of Coercive Labor, Poverty Alleviation and Social Control in Xinjiang,” Journal of Political Risk (Vol. 7, No. 12), December 2019. https://www.jpolrisk.com/beyond-the-camps-beijings-long-term-scheme-of-coercive-labor-poverty-alleviation-and-social-control-in-xinjiang/.
[25] Located as part of the 昌都市卡若区俄洛镇昌都市职业技术学校 campus. See https://bit.ly/2Rr6Ekc; compare https://archive.is/wip/uUTCp and https://archive.is/wip/lKnbe.
[26] See https://archive.is/wip/WZsvQ.
[27] Coordinates: 31.187035, 97.091817. Website: https://bit.ly/2Rr6Ekc. The timeframe for construction is indicated by historical satellite imagery and by the year 2018 featured on a red banner on the bottom-most photo of the website.
