Storing data on the margins: making state and infrastructure in Southwest China

ABSTRACT

Examining the emerging data center industry in Guizhou, Southwest China, this article investigates the infrastructure-making processes that are initiated to implement cloud infrastructures, and how they are mobilized to reconfigure Guizhou’s nature. It discusses how these processes have come about in tandem with the expansion of China’s cloud geography, and how they are impacting the region. Contextualizing and historicizing these processes, this article argues that the developing data center industry in Guizhou is part of the broader process of state-building. These processes of implementing cloud infrastructure in Guizhou lead to the co-production of further state legitimation and continued marginalization of Guizhou, thus calling into question the common claim that technology bridges economic disparities and enhances connectivity.

KEYWORDS:

• Data center
• cloud
• infrastructure
• state
• remoteness
• marginality

Introduction

Since 2006, China has prioritized innovation and entrepreneurship in the government agenda and focused on technological innovation driven by industrial sectors. Part of the government’s strategy is to set up national-level new areas (guojiaji xinqu 国家级新区) serving as innovation demonstration zones to foster businesses and investments. As of 2019, 19 national-level innovation demonstration zones have been handpicked by China’s State Council and allowed to offer favorable policies to spur innovation and drive regional economic growth based on their respective strengths and geographical advantages. One of them is in Guizhou, a landlocked Southwestern province in China known for its poverty, remoteness and marginality to the rest of the country. Striving to catch up with the country’s industrialized regions, Guizhou has been constructing the eighth national-level new area called the Gui’an New Area (guianxinqu 贵安新区, Gui’an hereafter) and the country’s first national pilot zone for big data industry since 2014. Specifically, the provincial government and industry are making the most of the province’s natural advantages – mild climate, abundant hydraulic power, and lack of earthquakes – to grow a data center industry in Gui’an. With foreign corporations such as Foxconn and Apple as well as domestic tech giants like Tencent and Huawei setting up their data centers in Guizhou, the province is trying to turn its remoteness and marginality into an edge and join the country’s high-tech industries. A similar trend is seen in other countries where big corporations have become increasingly enthusiastic about establishing large-scale data centers in remote and economically disadvantaged areas (Johnson, 2019; Vonderau, 2018) in hopes that technological connection fosters economic growth and overcomes marginality. This practice is obviously gaining momentum and becoming a global trend as industry-scale data centers are one of the fastest-growing industries worldwide (Cook & van Horn, 2011).

Focusing on the construction project of data centers by Tencent in Guizhou, this article aims to both question the techno-utopian claim of technological connection and show the impacts on the local population when these projects are implemented. The long-held belief in the promise of modern science and technology (S&T) has been central to China’s governance since the 1970s where the official narratives regularly link S&T with the pursuit of modernity, ambitions of national rebirth and global ascent (Greenhalgh & Zhang, 2020). More recently, with the heightened focus on innovation in S&T, the official narrative continues to treat S&T with utmost reverence and optimism, and reiterates that technological development is key to addressing the country’s confounding domestic problems and strengthening China’s global standing. As such, Guizhou presents a fertile ground upon which the dynamics between state governance and the promise of modern technology can be studied.

The plan to implement cloud infrastructure in Guizhou provides a window into the processes that are actively linking with the government’s narrative of modern technology and working to reconfigure Guizhou’s nature. By doing so, Guizhou can be incorporated into the expansion of China’s cloud geography, which is increasingly intertwined with geopolitics on regional, national and global levels. These processes create the conditions that facilitate a certain kind of imaginative work that makes the emerging data center industry in Guizhou imaginable and implementable. Such imaginative work plays an important part in the making of cloud infrastructure in Guizhou (cf. Vonderau, 2017; 2018). The construction of cloud infrastructure in Guizhou is essentially a techno-political project (Mitchell, 2002) geared towards the consolidation of state power – it is part of the larger state-making processes by which the Chinese state seeks to increase and strengthen its power and control over data in its quest to become a great cyber power (wangluo qiangguo网络强国). A focus on the practices and processes permits both an ethnographic approach and analytical intervention in thinking about this cloud project in more material and specific ways. It brings into view how the often veiled ‘political work’ of infrastructural forms tends to exacerbate existing social inequalities (Amin, 2014; Anand, 2015; Larkin, 2008; Mitchell, 2002). In Guizhou, the techno-political work of retooling Guizhou’s nature is largely framed in economic and thus apolitical terms that it will invigorate the local economy without jeopardizing the environment.

This article addresses the following questions: How is this techno-political project manifested on the ground? Is this cloud project delivering the promised results? To shed light on these questions, I focus on what practices and processes that are initiated to repurpose Guizhou’s nature to grow a data center industry in Gui’an and their impacts. Against the historical relationship between Guizhou and the Chinese state, my analysis shows that the implementation of cloud infrastructure in Gui’an has come about in a specific time and place. I argue that the construction of data centers in Gui’an runs the risk of prioritizing Guizhou’s natural resources, which are framed as ideal for data centers, over the welfare of the local population and undercuts the government’s promise that technological development bridges economic disparities and enhances integration. My analysis is built in three steps. First, drawing on the work on the materiality of the Internet within science and technology studies (STS), I highlight the politics of the implementation of cloud infrastructure and how this article addresses the existing deficiency in research representing the Chinese experience of S&T. Second, I describe the historical relationship between Guizhou and the Chinese state and discuss how this relationship bears relevance to the current development of the data center industry in Gui’an. Third, I present my ethnographic findings of how government officials and industrial actors reconfigure the cultural history of Guizhou as a remote place and repurpose Guizhou’s nature to suit the data center industry. Lastly, I discuss the tension that arises when the cloud touches the ground concerning the people affected and the socialist legacies of the place.

Notes on method

Materials for this article are based on one year (2018-2019) of research consisting of ethnographic fieldwork in Gui’an and semi-structured interviews with government officials, industry experts and residents living in the near surroundings of these data center projects, and archival research. Studying data centers poses several ethnographic challenges. First, getting a permission to do research inside a data center has proved impossible mainly because of the corporate security strictures. Second, it is questionable how fruitful it is to immerse oneself in a space comprised of racks of servers, power generators and cables, and a handful of security guards (Vonderau, 2017, 2018). Hence, I have conducted fieldwork on what is going on outside of data centers; I study the policy documents, attend industry conferences, and talk with government and industry actors and people affected.

I was also affiliated with the governmental research institute for ecological civilization (guian shengtaiwenming guoji yanjiuyuan 贵安生态文明国际研究院), which provides evaluations of policy implementation and recommendations for the government in Gui’an. The research institute served as a good platform for me to meet government officials from various bureaus and agencies, particularly those involved with data centers. I was able to attend some meetings, business forums and industry conferences where I could observe, for example, the interactions between the government and companies, and talk with some officials and industry actors in the data center industry.

During my fieldwork, the data center projects of Tencent, Huawei, and Apple, which are geographically close to one another, were still under construction and my requests for interviews with these companies were all flatly rejected. Thus, most of my fieldwork took place around these construction sites. I collected my research materials from the everyday conversations and life history interviews with the construction workers coming mainly from different parts of Guizhou and Sichuan, vendors selling food and drinks around the sites, and residents living nearby. These conversations shed great light on the lived experiences of constructing digital infrastructures, which show different ways in which the infrastructure-making processes connect and disconnect as well as reveal and obscures social relationships.

The materiality of the cloud

When we hear about data stored in the cloud, that data is in effect materially stored in a data center. Contrary to the ephemeral-sounding term ‘cloud,’ the ‘cloud’ exists in a very material shape and form. The servers in data centers use huge amounts of electricity and generate enormous amounts of heat, which, in turn, requires large investments in cooling systems to prevent servers from overheating and to continue operating. Furthermore, data centers also attach great importance to the architectural forms and designs to protect the facilities from external threats and attacks. With these specific requirements, the locations of data centers can display how such facilities are intertwined with the political and social life in a specific context.

Attending to the materiality of the Internet, a growing number of scholars are investigating the infrastructures that comprise cloud computing. They have shown that the cloud is in fact very specifically embedded in and interacting with the material world in historical, cultural, social, political, and environmental ways (Gabrys, 2011; Hogan, 2013; Starosielski, 2015; Vonderau, 2018, 2019). This body of literature reveals the spatializing and unevenly distributed impacts of digital capitalism. While the focus on the materiality of the cloud is much warranted, there is also a clear discursive process going on that has not received enough attention. Attention to the discursive work further illuminates the materiality of infrastructure in that it considers how infrastructural and technological promises are formed, articulated, and implemented on the ground (Edwards et al., 2009; Hu, 2015; Mosco, 2015), an analytical focus which this essay pivots on. It discusses how particular imaginaries of and about a place are incorporated into promoting cloud infrastructure projects in a predominantly rural province in Southwest China.

Against this growing research field, the study of Chinese S&T remains underdeveloped and, more particularly, focuses largely on science (Chen, 2003; Palmer, 2007) and very little on technology. In the existing scholarship on China’s S&T, there is also a tendency to prioritize western knowledge and experiences and uses it as a yardstick against which China’s development is measured. While these studies illuminate some parts of China’s S&T development, it has yet to examine important questions such as how members of society negotiate with S&T, and how S&T affects the lives of ordinary people. China’s S&T power is gaining significance globally, but its approaches to develop and apply science and technology are still distinctly missing from the field of STS (Greenhalgh, 2020, p. 6). Recently, some anthropological work has directed attention to the science and society domain and shown how S&T has immensely restructured social, cultural, and political life since the 1970s, but not always in the ways intended by the government (Greenhalgh, 2008; Hathaway, 2013). While much warranted, this emerging literature focuses largely on science-society relationship; the technology-society domain remains under studied. This article tackles this bias by adopting an ethnographic approach to demonstrate the everyday experience with China’s technological development and contributes to the Chinese STS focusing on a specific place and population.

Furthermore, in contrast to the popular claim of digital capitalism and technology as unfettered by geographical borders and national regulations, this article demonstrates the materiality of cloud infrastructure by showing the central role of Chinese state in creating the conditions conducive to the construction of data centers that is deeply affecting the local population. I have observed a kind of imaginative work geared to create a data center industry in Gui’an, which is predominantly dictated by the state (cf. Vonderau, 2017). This imaginative labor is predicated on a re-imagination of Guizhou. As such, the state-initiated infrastructure-making processes and imaginative work of implementing cloud infrastructure in Guizhou make up more than just a space of economic calculation, anticipation, and aspirations. More crucially, the imaginative work in Guizhou is part of the broader processes of state-building to further consolidate state legitimation and governance (Bach, 2016; Cartier, 2018).

Re-imagining Guizhou

Plans to put Guizhou on the fast track to economic growth were already in the making in 2011 followed by the State Council’s decision in 2013 to establish the Gui’an New Area. In the following year, Guizhou was designated as the country’s first pilot zone for big data industry with a specific focus on fostering the data center industry in Gui’an. When traveling from Guizhou’s capital Guiyang to Gui’an, which is about 30 kilometers westward and is paved with the newly laid tarmac roads, observant travelers surely would not miss all the signs and posters attached to the lamp posts, overpasses, and buildings along the roads. These roads are particularly plastered with more of these signs when big events such as the International Big Data Expo takes place in the last week of May every year. These posters and signs publicize the government policy and goals such as ‘Let the world understand Gui’an, let Gui’an go to the world’ (rang shijie liaojie guian让世界了解贵安, 让贵安走向世界); a few signs are commercial advertisements jumping on the big data bandwagon such as ‘China Data Valley. Central Data Valley. Be the Landlord of Fortune 500’ (zhongguo shugu shugu zhongyang zuo shijie wubaiqiang de fangdong中国数谷, 数谷中央, 做世界500强的房东). These slogans represent the central and provincial governments’ big data dreams and visions for Guizhou’s future, which is believed to be mediated and brought about by big data technology and economy. They demonstrate the unwavering technological utopianism – the panacea to isolation and economic ills that have plagued the province for as long as one can remember.

Kept on the edge

Guizhou has been one of the most disadvantaged and poverty-stricken regions in China largely due to its mountainous landscapes, which has made farming challenging as 90% of cultivatable land is located in highly mountainous areas. This is often attributed to the province’s endemic rural poverty and long history of relative isolation from the rest of the country. The common perception of Guizhou as backward, inaccessible and thus immune to governance has been so ingrained in centuries of Chinese history and state-building that scholars characterize the province’s nature-induced ‘remoteness’ (pianpi 偏僻) as iconic (Luo et al., 2019). Nevertheless, Guizhou’s remoteness does not shield it from state intervention throughout history. Its natural resources, geographic proximity to resource-rich Yunnan (p. 120), and culturally diverse inhabitants have made it a frequent and yet often recalcitrant target of cultural and political assimilation during different Chinese political regimes (Weinstein, 2013). Thus, Guizhou has always occupied an ambiguous place in the national imaginaries in which it is constructed as a remote and marginalized area while at the same time it also figures as the crossroads that had to be traversed to reach those strategic borderlands (Luo et al., 2019, p. 272). In other words, there is a malleable and productive (from the state’s perspective) tension or ‘doubleness’ (Luo et al., 2019, p. 272) between its remoteness and geographic connectivity, and distinct cultures from dominant Han culture. This tension and paradoxical co-existence of remoteness and connectivity has served as a powerful trope in state imaginaries and interventions of Guizhou throughout history.

One example of how this tension plays out is shown tellingly by the fact that Guizhou had been a penal colony for convicted criminals and disgraced officials since the Han dynasty (Goodman, 1983, p. 115). The entrenched imaginary of Guizhou as an inhospitable and remote place has also made it a container for the undesirable and the marginalized bodies that need to be kept from the political center. Such a cultural logic continues to hold sway during contemporary China and led to the institution of labor reform camps (laogaiying 劳改营) and labor re-education camps (laojiaoying 劳教营). Up until the early 2000s, a total of 58 labor reform camps were recorded to have existed across Guizhou (TLRF, 2008). The construction site of the data centers by Tencent happens to occupy some parts of a former labor reform camp commonly known as the Yang’ai Farm by the locals, a subject to be discussed in detail later.

Guizhou was also a key site of two major state projects: the Third Front Program in the late 1960s and the campaign of the Grand Development of the West, also known as ‘Open Up the West,’ in the late 1990s. The Third Front Program aimed to create a self-sufficient industrial system within the naturally remote and strategically secure region to serve as a Maoist reserve in the event of China being drawn into war. Many industrial factories and workshops assembling sophisticated machinery were scattered across Guizhou’s mountainous terrains. The projects executed in these remote locations could not benefit from the cluster economies present in large urban areas – the availability of a range of supporting services and goods that make any complex project easier to carry out in a city – and so take longer and require expensive imports of locally unavailable items (Naughton, 1988). The original goal for the country’s interior regions to achieve self-sufficiency simply didn’t materialize.

The campaign of ‘Open up the West,’ launched in the 1990s, hoped to reduce the increasing disparity between eastern coastal regions and western inland areas of China and preempt the threat of political instability fueled by ethnic tensions. This campaign had led to many infrastructural projects being carried out in Guizhou such as railways, highways, power plants, airports, and dams. Nevertheless, the long-term trend that China’s west as the primary producer for the industrializing east seemed to only continue if not intensify. This is made most clearly by the west–east electricity transfer project, which basically sells electricity produced in Guizhou at a low price to the east; the east can continue to manufacture goods without worrying about energy pollution. Both state programs capitalized on Guizhou’s natural landscape and resources in ways that have led to Guizhou’s further dependence on state subsidies (Naughton, 1988; Oakes, 1999) and strengthened Guizhou’s role as a cheap energy provider for the industrial east (Oakes, 2004).

These state projects are some recent examples of how state intervention transpired in Guizhou. They pivot on the imaginaries of Guizhou as a remote and marginalized place that can either be utilized to keep unwanted people and used them to cultivate the wasteland, support the greater need for the country in times of war, or supply energy for the nation’s industrialization. It shows the dynamics between the center and the periphery where marginality is not to be defined in binary terms such as exclusion versus inclusion. Here, marginality is a conditioned and measured existence materialized in ways contingent on state calculations. Guizhou has been enrolled in different state projects just as much as it has also been kept at distance. Today the growing data center industry in Guizhou constitutes yet another occasion to mobilize this logic and put Guizhou in a place that is good for the country.

The rise of China Cloud

Since taking power, President Xi Jinping has energetically supported investments in data-driven innovation and advanced technology projects. For more than a decade, China has prioritized the development of cloud computing technology with the goals of both expanding Chinese military and civilian access to cloud computing information technology resources and creating an internationally competitive Chinese cloud computing service industry (Ragland et al., 2014). The 13th Five-Year Plan (2016-2020) highlights the importance of cloud computing technology and its relevant infrastructure following several important policy announcements, such as the Internet Plus initiative in 2015 aiming to encourage Chinese manufacturers to deploy mobile Internet, cloud computing, and to promote Internet banking, mass entrepreneurship and innovation. A significant development is the creation of the Cyberspace Administration of China (CAC) (wangxinban网信办), the new authority regulating and controlling China’s cyberspace. The CAC oversees implementation and enforcement of the Cybersecurity Law, an important legislation that came into effect in June 2017. This law emphasizes the nature and flow of digital information that has been generated in China and places greater restrictions on the transnational movement of data. This law demonstrates that the Chinese government has stepped up efforts to monitor and control the domestic cyberspace while asserting its cyber-sovereignty on the global arena (Yuen, 2015).

With data localization, foreign firms will have to either invest in new data servers in China which would be subject to government spot-checks or incur new costs to hire a local server provider, such as Huawei, Tencent, or Alibaba. Some critics point out that the substantial investment by these Chinese technology firms in recent years is partly designed to bolster the domestic Chinese data management and telecommunications industry against global competitors (Wagner, 2018). China’s cloud industry is benefiting from limited competition with major US providers, which has left room for the development of indigenous cloud services. As of 2019, the top two major cloud service providers headquartered in China are Alibaba Cloud and Tencent Cloud (‘Cloud infrastructure services,’ 2019, june 26).

These initiatives, policies and regulations are part of the state-making process to create new modes of conduct, organize and regulate ways and possibilities of government and corporate operations. China’s ambition of becoming a technologically innovative nation presents a unique case in comparison to other countries in a similar race in that, for example, the sophisticated Internet content-filtering and censorship program, commonly referred to as the Great Firewall is still in place, and that the Ministry of State Security has taken an oversight role in projects aimed at bringing foreign cloud computing investment to China. Such restrictions distinguish the Chinese government’s cloud ambition from other nations. China’s cloud strategy indicates its intent in not only protecting Chinese enterprises from foreign competition but also enhancing its influence over the evolving global cloud computing standards. The current development of the data center industry in Guizhou should be considered in this context.

Technological (dis)connection

In this section, I will discuss two ethnographic examples to illustrate the discursive work of the government and industry that tries to retool Guizhou’s nature to attract investments. Both examples indicate how the government’s vision and infrastructural promise are articulated and conveyed to the industrial actors in a way that resonates with the history of Guizhou as a place.

The infrastructural promise

On 12 December 2018, I attend the 13th Internet Data Center Conference in Beijing. It is the biggest annual event for the data center industry in China and invites policymakers, key industrial stakeholders, industry analysts, investors, and the media to convene in Beijing for three days. I am at a session called ‘The Summit of Data Center Industrial Parks’ (shujuzhongxin chanyeyuan gaofeng luntan 数据中心产业园高峰论坛), which brings together various stakeholders to meet and explore opportunities of collaboration. Jin Heping, deputy party secretary of Gui’an is giving a presentation of the ongoing development of the data center industry in Gui’an. The audience consists of tech company CEOs, investors, bankers, market analysts, and government representatives from regions such as Ningxia and Inner Mongolia. Jin has been the deputy party secretary of Gui’an since 2015 and the main spokesperson of the data center industry in the region. Leading a delegation of five people, Jin has come to the conference to promote Gui’an as a good place for setting up data centers.

Jin starts his presentation with a news video from China Central Television (CCTV) broadcast on 29 April 2018 (‘Live at the Yangtz River: Guizhou,’ 2018, april 29). The video shows the construction site of Tencent Gui’an Seven Stars Data Center Complex (tengxun guianchixing shujuzhongxin 腾讯贵安七星数据中心) with an aerial view showing that Tencent has dug six tunnels in a mountain in Gui’an. The reporter in the video describes that each tunnel is 15 meters high and divided into two floors; the huge caves created inside will house tens of thousands of servers for Tencent’s data. The reporter emphasizes that these tunnels are built to the highest standards of civil air defence, which makes Tencent’s data center complex a giant bombproof shelter or a data bunker (Hu, 2015, p. 81). In the event of impending danger or threat, these gates will be automatically locked, according to the report. Showcasing Tencent’s construction project, this news report highly praises the rapid development of Gui’an since its official creation in 2014. It commends the local government’s success to have attracted not only Tencent, but also, among others, Foxconn, Huawei, and Apple to set up their data centers in this largely rural area.

Jin goes further to highlight that Guizhou’s moderate climate, abundant rainfall (cheap hydropower), lack of earthquakes, and mountainous landscape make it an ideal location for data centers. Jin emphasizes that the current construction of data centers by Tencent as well as Huawei and Apple shows the government’s unwavering determination and efforts to develop Guizhou, and praises these projects as the true feats that ‘have been accomplished from nothing’ (wuzhong shengyou无中生有), reminding the audience of the long history of endemic poverty and barrenness of Guizhou. Jin stresses that the cloud infrastructure will allow Guizhou to quickly catch up with the developed regions (houfa ganchao后发赶超) and can take the most appropriate and fastest routes to economic prosperity (wandao quzhi弯道取直). Lastly, Jin invites the audience to come to Guizhou to fulfill the dreams of big data (dashujumeng大数据梦), evoking President Xi Jinping’s famous idiom the China Dream (zhongguomeng 中国梦), a phrase that has become one of the guiding ideological principles promising to transform China into a global economic, military and innovation superpower while trying to connect with the people through a common vision by addressing social inequalities.

Jin’s presentation resembles typically an infrastructural promise (Anand et al., 2018) displaying a common narrative about infrastructure as projects of modernization aspiring and anticipating for a better future (Cross, 2014). It paints a vision of Guizhou as an important base for cloud infrastructure in the country, which would result in economic growth and improvements in the well-being of the local population. The presentation is a discursive practice that adheres to the government’s firm belief in technological development and works actively to reconfigure Guizhou’s nature so that it can be incorporated into China’s growing cyber power.

The value of Guizhou’s location

With the ongoing construction projects of data centers by Tencent, Huawei and Apple in Gui’an, Guizhou may very well become an important base for China’s cloud infrastructure. But it remains to be seen whether these projects will ultimately lead to better economic conditions for the local population. In a business promotion meeting organized by Gui’an officials during the Big Data Expo in 2019, a presentation by Tencent’s representative, referred as Mr. Zhang here, seems to raise more doubts than reassurance about Guizhou’s future.

Tencent’s data center project has attracted national and international attention partly because of the spectacle of storing servers in six tunnels. Many wonder why Tencent chose to build a data center complex in Gui’an, a rural and sparsely populated area. Standing in front of a room with approximately 20 investors from across the country, Mr. Zhang explains that the decision to build this cloud infrastructure in Guizhou is highly strategic in terms of data security and energy demand. He notes that 70% of data centers in China is concentrated in the southeastern region; such a high concentration exposes data centers to greater security and safety risks. For Tencent, it is important that data centers operate without natural or man-made disruptions so, according to Mr. Zhang, Guizhou is chosen because it is far away from the coastal region and lack of natural disasters such earthquakes and floods. Referring to the Third Front Program, Mr. Zhang notes that Guizhou was selected for its naturally remote and strategically secure location; similarly, Tencent sees the value of Guizhou’s location in providing data security and ensuring stable operations. He points out that the security design of Tencent’s data center complex in Gui’an is constructed in such a way that it can continue operating during emergency such as natural disasters or war. It also means that the data center complex needs to be energy independent. By putting servers inside a mountain, according to Mr. Zhang, the data center complex can take full advantage of the area’s chilly weather and natural airflow to create a natural and effective cooling system, which further reduces the data centers’ energy-dependency in situations of emergency.

Storing servers in caves is nothing new. Located in central Stockholm, Sweden, Pionen, a former civil defense center built in 1943, was converted into a data center by the Swedish Internet service provider Bahnhof and opened in 2008. The former nuclear bunker was built below 30 kilometers of granite with 1,100 square meters of space and is repurposed to preserve data or data centers (Fish & Garrett, 2019) because of its geographic isolation and securitization of space. Nevertheless, placing data centers in militarized buildings goes beyond the scope of ensuring data security. There is no technical basis for using buildings built to defend chemical weapons or invasions by Soviet tanks to combat cyberattacks (Hu, 2015, p. 92). In the name of data security, what is really at work is the reanimation of sovereign power (ibid.). The development of data centers in Guizhou presents a particularly telling example considering China’s recent introduction of the Cybersecurity Law.

The materiality of data centers is not only embodied by the physical structure of the bunkers but also by the temporality of data centers (Fish & Garrett, 2019, p. 1) which echoes the promises of round the clock operativity and protection from natural calamities as seen in Mr. Zhang’s repeated emphasis that data centers operate without interruptions. A popular notion about digital capitalism is that it is unrestrained by geological, geographical, ecological and geopolitical conditions. Data centers are the modern archives that sustain our digital lives and propagate permanent data storage and capability to weather crises of natural and human nature (Fish & Garrett, 2019, p. 8).

When it comes to the maintenance of data centers in the caves, Mr. Zhang is eager to say that Tencent’s current policy is to reduce its dependence on manpower to manage and maintain data centers. There is a growing trend in implementing automation in data centers so that routines and workflows of a data center – scheduling, monitoring, maintenance and application delivery – are managed and executed without human administration. ‘We are striving to achieve autonomous maintenance,’ said Mr. Zhang. Tencent has already widely employed automation in its data centers; in fact, according to Mr. Zhang, the data center complex in Gui’an will be predominantly monitored, controlled, and managed remotely from its urban headquarters in Shenzhen.

Employment opportunities that are available at data centers on the site are generally rather limited. Aside from a handful of security guards, engineers and programmers are usually located in more urban locations. The case of Tencent’s project in Gui’an shows that Guizhou’s location and resources are considered as more useful than its people. Without the prospect of creating employment opportunities for the local population, Tencent’s project raises questions about the prosperity promised by the data center industry, the prospect of labor absorption, and the emergence of surplus populations. It also exemplifies a process of dematerialization characteristic of the data-centric economy and evokes the ‘doubleness’ of Guizhou being remote and yet essential to connectivity elsewhere. Similar processes where already marginalized places are utilized to further connectivity in other more urbanized regions have been documented (Hogan, 2013; Johnson, 2019). The case of Guizhou further highlights the need for a critical view on technological connection as an analytical question rather than an empirical outcome.

Bodies outside the cloud

Like in many other places in China undergoing urbanization, land disputes are not uncommon in Gui’an. Nevertheless, since most of the land in Gui’an is already acquired by the administrative committee of Gui’an, the disputes are usually a matter between the administrative committee and the local villagers. At the time of my fieldwork, there was a dispute over some land that overlaps with Tencent’s construction site. This site happens to occupy an area that used to belong to the Yang’ai Farm, a former labor camp.

Created in 1952, the Yang’ai Farm consisted of a prison, and seven brigades that include laogai camps, laojiao camps and forced job placement camps. To this day, the Yang’ai Farm consists of tea plantations, fruit gardens, forests, paddy fields and some arable land. The farm had stopped letting its prisoners out and work on the farm in 2006. In 2010, this labor reform camp was renamed as Guizhou No. 1 Women’s Prison, and a separate prison for male prisoners was also set up nearby. At the time, there was a system called liuchang jiuye (留厂就业), which means ‘take employment’ (Seymour & Anderson, 2015/1998, pp. 189–197). It is essentially a forced job placement program, which consisted of freed workers, that is, ex-prisoners who had served their sentences and were kept on in internal exile (Mosher, 1991, p. 50). Under this program, released prisoners were given local residence permits (hukou) and jobs on the farm. While the practice of jiuye was abolished in the late 1980s or early 1990s, many ex-prisoners and their families still live in the area, and some live near the construction site of Tencent.

When the plans of constructing Gui’an were made, all land of the Yang’ai Farm was given to Gui’an to plan and manage. The farm has been a state-owned enterprise, so this gives Gui’an a lot of leeway to plan the sales of the land. What was once a socialist landscape of revolution and production is now reconfigured into a landscape of digital capitalism by providing the storage and processing power for data extractivism (cf. Lam, 2020). This reconfiguration has resulted in some material consequences particularly for those who have co-existed with the farm for decades, such as the ex-prisoners and their families. For these people, the most pressing issue concerns the ownership of the land and the houses they live in. For decades, they have been living in the houses provided by the farm and cultivating the land owned by the farm. Now that all the land belongs to Gui’an, these families were asked to move. But they were not financially compensated for the land and the house like the local villagers. They were only compensated for what was grown on the land. These families call themselves ‘dry residents’ (ganjumin干居民). Here, ‘dry’ means being deprived of the land and houses they had lived in and cultivated but never owned.

Mrs. Liu, in her 60s, is a dry resident. When I met her for the first time in early November 2018, she and her son were staying in an abandoned workers’ dormitory perched on a hill adjacent to Tencent’s construction site; the dormitory was owned by the Yang’ai Farm. They had been forcefully removed from their house a week before. In the dormitory, a few wooden stools were scattered around an impromptu stove made of bricks and stones, two mattresses and some bags of personal belongings were stacked up in the back of the room. Showing me the bruises on both of her arms and shoulders, she told how she had fought to try to keep her house, but the police and security guards pushed her aside, threw out her furniture and belongings and let the excavator tear down the house. The Liu family had lived in that house, property of the Yang’ai Farm, for more than three decades. Mr. and Mrs. Liu are originally from Zunyi, a city in the northern part of Guizhou. They moved here in the late 1970s when Mr. Liu was sent here to serve his sentence. Their son was born and raised on the farm. After having served his sentence, Mr. Liu was asked to stay and continue working on the farm. Since then, he and his son had been working on the farm until Mr. Liu passed away in 2010. Her son has been doing some odd jobs around Gui’an.

Mrs. Liu explained that most of the farm’s employees have moved and lived in the resettlement housing complex in Huaxi, an urban district in Guiyang. Mrs. Liu attributed their poor treatment to the fact that her husband never became a worker (zhuangong 转工). Zhuangong literally means ‘to become a worker.’ It was a practice on the labor reform farm where ex-prisoners became a worker of a state-owned enterprise, or commonly referred to as zhigong (职工). Once an ex-prisoner became a zhigong, he and his family were then entitled to the benefits of a state-employed worker. Mr. Liu never became a zhigong. The distinction between an ex-prisoner and a zhigong was most stark in terms of their monthly wage. Mrs. Liu said, ‘an ex-prisoner earned one RMB while a zhigong earned 24 RMB per month.’ In the end, the Liu family received a total of 280,000RMB for the jasmine trees they had grown on the land, which is barely enough to purchase a small apartment in Guiyang. Mrs. Liu decided that she was going to go back to Zunyi to stay with her relatives while her son continues to stay in Gui’an.

The dry residents embody the lived experiences when the cloud touches the ground. The life history and trajectory of these dry residents are so distinctly embedded in the sticky history of Guizhou as a marginalized place that their lives have been made abject by different infrastructure projects over the years. As such, the dry residents also represent a kind of infrastructural deficiencies that both index pre-existing inequalities and deepen them at the same time. Compounded by the ‘sticky grip’ (Tsing, 2005) of the history of the farm, not only have these dry residents lived through their social marginality but also endured the place-based marginality that has sustained the circumstances of their lives. Their lives demonstrate how the seemingly virtual cloud are entangled with the social environment at this particular historical and geographical juncture.

Conclusion

The emerging data center industry in Guizhou triggers a range of social, economic, environmental and spatial changes that have profoundly affected the people in the region. The implementation of the cloud infrastructure in Gui’an is both a process of regional identity-making – Big Data hub in South China – and a process of consolidating state governance. I have shown that the growing data center industry in Gui’an puts to use not only its natural conditions such as the climate and energy recourses, but also particular imaginaries of Guizhou’s remoteness, isolation, and marginality. Government officials, industry experts and stakeholders re-use these images to reconfigure Guizhou as a naturally suitable place for data centers for foreign and domestic investors. I foreground the history of the province and discuss how it is persistently imagined in certain ways, and how the production and circulation of these imaginaries have rendered it available for certain kinds of state intervention. I also show how the siting of data centers lays bare some of the socialist legacies while overwriting others (cf. Lam, 2020).

I contend that without the state-led imaginative work, it would be difficult to put these (re)imaginaries into work and develop a data center industry in Guizhou. The implementation of the cloud infrastructure is a techno-political enterprise because it shows not only part of the regional processes in which Guizhou is trying to rebrand and reposition itself by actively participating in the national agenda, but also part of the continued making of the Chinese state. Viewed as such, the emerging data center industry in Gui’an can be seen as part of the broader process of consolidating state power. By revealing the techno-politics of cloud infrastructure in Gui’an, this article illustrates how infrastructures forge connections while disrupting others and foregrounds the tension between center and periphery and urban and rural. As such, the common techno-utopian claim that technology connects and overcomes marginality becomes problematic.

Infrastructures continues to provoke questions among populations as to who benefits and who is made abject (Anand, 2012); one person’s benevolent infrastructure can be another person’s burdensome barrier. I have shown some of the ways in which the materiality of the cloud is entangled and ‘embodied within particular communities, shaping livelihoods and possible futures in particular ways’ (Pickren, 2018, p. 237). The current making of the cloud industry in Guizhou should be situated in the historical and geographical context of capitalism’s uneven development in order to recast the emerging socio-technological entities and realities as political rather than merely technological and thus inevitable (Pickren, 2017, p. 23). As part of the Chinese state’s efforts to reconfigure national and global cloud geographies, it does seem that the growing data center industry allows Guizhou to be included in the state project of uplifting the China Cloud. But it remains to be seen how much the project is truly benefiting Guizhou considering the environmental impact and limited employment prospects from the data center industry. While the cloud might potentially further entrench Guizhou’s marginality creating forms of precariousness because of land expropriation and the dematerialization of local economy, the productive tension long characterized Guizhou’s relationship with the state might also be leveraged by the locals to gain more power in the decision-making process.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by The Swedish Research Council.

Notes on contributors

Darcy Pan

Darcy Pan holds a PhD in anthropology from Stockholm University (2017). Her research interests focus on the state, labor, civil society, infrastructure, regimes of governance, big data, and global technology and development. She has done fieldwork in Ireland, China, and Hong Kong. [email: [email protected]].