Abstract
Introduction
In late 2018, the city of Shanghai began work on the “Overhead-Underground” project (架空线入地工程), a city-wide infrastructure upgrading plan that aims to replace overhead electric and communications cables citywide with underground installations. The project's formal goal is to approximately double the percentage of cable that is run underground in the central city area, from an estimated 29% to 62% (Shanghai Municipal Government, 2018). In the existing network, pylons intended to serve specific purposes such as electrical transmission, street lighting, surveillance camera installation and traffic signaling have been repurposed as nodes on the city's fiber-optic infrastructure. In response, the project aims to dramatically reduce the number of roadside pylons, replacing them with modernized “unified pylons” (合杆) that will direct traffic and position surveillance cameras, while overhead electrical and communications infrastructure is moved underground and out of sight.
The first visible change brought about by the project, however, was a noticeable increase in the number of construction work teams present in the streets of the city. One afternoon in 2019, an eight-man team led by Xiao Li, a migrant worker in his early 40s who has spent twenty years moving between Shanghai and his rural Anhui hometown, completed the final stage of a project that had been in process for months: Wearing hard hats and reflective vests over the worn clothing that is an unofficial uniform of informal construction labor, eight middle-aged migrant workers set to work climbing bamboo ladders to remove a large bundle of fiber-optic cables strung alongside a major downtown thoroughfare. As pedestrians looked on and asked Li about the work, men clambered up ladders and into trees. After several hours, the sidewalk was covered in kilometers of waste cable, and by late afternoon, Li's team retraced their steps, collecting the remains of the obsolete cable before returning to their dormitory. With the completion of the project, both the previous fiber-optic infrastructure and the men who worked on it had disappeared from the roadside.
The ideas driving the Overhead-Underground project are rooted in visions of urban infrastructure development that move the structures operating the city out of sight of urbanites. Like other urban restructuring efforts in China, the overhead-underground project is an example of what Chu has called an “infrastructuralization of state power” heralded by the arrival of anonymous men in hard hats who implement state-driven infrastructure reconstruction programs (Chu, 2014: 351). The official announcement for the project describes these goals as follows: In alignment with the city Party Committee and city government's requirements for the fine-tuning and strengthening of urban management, we are beginning the political work of undergrounding overhead cabling and implementing unified utility poles, in order to progress in strengthening the management of overhead cabling and utility poles, gradually eliminate black pollution, reduce the number of utility poles on public roads, ensure safe circulation around the city, and construct an orderly, safe, clean, and beautiful high-quality urban environment. (Shanghai Municipal Government, 2018, author's translation)
Broadly speaking, the “political work” of the Overhead-Underground project represents an intervention in the fabric of Shanghai's urban space in service of an imagined populace that on one hand does not have an opportunity to participate in the design, planning, or construction of urban communications infrastructure, but on the other hand is expected to enjoy and benefit from the resulting “high-quality urban environment” (Li and Zhong, 2021). In this framework, the production of urban space through infrastructure is one tool among many for shaping cities’ development, and in the process generating appropriate and governable publics (Oakes, 2016).
While the Overhead-Underground project does rely on state-driven priorities to build a more manageable urban infrastructure space, this article approaches those goals not from the perspective of the funders, planners, or engineers who devised them, nor from the perspective of the local public whose everyday life is disrupted by the unexpected appearance of anonymous “men in hard hats.” Instead, by focusing on migrant workers tasked with carrying out the project itself, this article provides a novel perspective on the construction of urban infrastructure. It is widely recognized that migrant labor is crucial to the construction of Chinese cities. However, researchers asking “who builds Chinese cities” often focus on top-level developers rather than the sub-contracted migrant laborers who eventually do the building work (Jiang and Waley, 2020), and research on infrastructure in particular, such as Furlong's (2022) exploration of “infrastructure with Chinese characteristics” does not touch on migrants’ roles in building urban infrastructure. At the same time, while infrastructure scholars have pointed to the roles that low-waged rural-urban migrant workers hold as the embodiment of certain forms of urban infrastructure (e.g. Fredericks, 2018; Gidwani, 2015), infrastructure scholarship is just beginning to consider the labor of building infrastructure in depth. As a result, the Overhead-Underground project is well-positioned as a case study from which to understand how a reliance on migrant labor can impact the development of high-tech network infrastructure reconstruction projects.
While kept at a distance from state-directed planning on one hand and the imagined urban population on the other, the reconstruction of Shanghai's fiber-optic infrastructure relies on migrant construction workers like Li's team. Through an examination of the role played by migrant labor in infrastructural reconstruction projects as well as the development of infrastructural knowledge through everyday, low-waged labor, this article argues for the centrality of construction as a way to understand infrastructure. In doing so, I extend recent theorizations that have expanded from Abdoumaliq Simone's (2004, 2021) concept of “people as infrastructure” and, more recently, Jean-Paul Addie's related assertion of “social relations as infrastructure” (Addie, 2021: 1352). These formulations emphasize the ways that improvisational and contingent human relationships function outside of conceptions of infrastructure as “engineered solutions,” knitting urban lives and livelihoods together in ways that we should rightly think of as infrastructural (Wilson and Jonas, 2021). Expanding on this concept along the lines of Coss-Corzo's assertion of the essential nature of infrastructural work (Coss-Corzo, 2021b), I argue that the social relations that facilitate infrastructural construction, upgrading, and repair play a crucial role in both the material emplacement of infrastructures in urban space, as well as the creation of formal knowledge that is necessary for infrastructural (re)construction.
The article proceeds as follows: In “How is infrastructure built?,” I situate my argument in relation to infrastructure research and emerging scholarship on construction and repair, before introducing fiber-optic infrastructure construction in Shanghai in “Laying Fiber with True Standard.” The final two sections theorize the importance of migrant workers to the material (re)construction of Shanghai's communications infrastructure, first drawing on empirical research to consider how migrant labor shapes infrastructural construction before focusing on the role of ordinary construction workers in the accumulation of infrastructural knowledge necessary for the Overhead-Underground project. As I conclude, if the production and reproduction of high-tech, critical infrastructure in Shanghai cannot be separated from the labor of migrant workers, our understandings of infrastructural development must also acknowledge the importance of social systems that are built around migrants and migration.
How is infrastructure built?
The relationships between infrastructural knowledge, migrant labor, and state infrastructure planning that come together in the Overhead-Underground project offer an opportunity to better understand and theorize how infrastructures are built, and how construction labor functions as a component of infrastructural systems. This section brings together recent scholarship that has brought questions of labor into analyses of infrastructure, making the case that an understanding of low-waged, informal construction work provides an important lens on the prospects and outcomes of infrastructure construction projects.
Despite recent arguments that the intersections of infrastructure and labor have received minimal attention (see Strauss, 2020), significant bodies of literature have recently begun to examine the various forms of work that are necessary for understanding, maintaining, and repairing urban infrastructures. One key thread of scholarship has arisen to address parallel questions of how human labor expands or supplements nonhuman infrastructural networks. Drawing on a broader understanding of infrastructure “not as an inert, technical support structure, but, rather, as a lively, embodied socio-technical system” (Fredericks, 2021: 4), scholars in this tradition have often utilized Abdoumaliq Simone's concept of “people as infrastructure” as a way to emphasizing how social relationships among people in informal, low-waged, or otherwise marginalized positions are crucial for the functioning of urban infrastructural systems. As Simone (2004, 2021) argues, emphasizing technical or engineered systems as the primary focus of infrastructure research can lead scholars to miss the ways that social relationships perform profoundly infrastructural functions. Drawing on this insight, recent scholarship has emphasized that human labor plays a crucial role
Understandings of infrastructure as constituted through social connections have not led to widespread analysis of the work that goes into building technical infrastructures like the Overhead-Underground project. As Lemanski and Massey have recently pointed out, while urban infrastructure broadly has been recognized as a manifestation of sociopolitical processes, infrastructural networks are still often understood as technical rather than social systems (Lemanski and Massey, 2022). Much of the scholarship that emphasizes labor inputs into infrastructure focuses on the work of urban waste disposal and water provision in the global South. Gidwani, for example, emphasizes the way that rural–urban migrants in Delhi have come to rely on waste picking as a key economic niche, transforming waste as a form of “infrastructural labour” (Gidwani, 2015: 590). At the same time, attempts to formalize waste disposal threaten this niche through privatization (Gidwani and Reddy, 2011). Fredericks (2014, 2018) has also emphasized how rural–urban migrants working in all aspects of garbage disposal in Dakar have been enrolled in the production of the city's clean, modern image. For Fredericks, this form of “vital infrastructure” connects material and social worlds, and urges a “‘fleshing out’ of infrastructures’ literal vitality” (2018: 16). Other work takes up a similar emphasis on the politics of labor in trash collection, emphasizing the role played by “hustling” in structuring informal relationships that make up Nairobi's garbage infrastructure (Thieme, 2013, 2018). This scholarship is valuable in particular for the way that it emphasizes the role that low-waged, informal, and marginalized labor plays in producing infrastructural systems that are often pointed to as either emblematic of urban life, or as failing to live up to the standards of a “modern city.”
Research on how human labor and relationships become infrastructures in and of themselves has also focused on how people and the work they do extend, supplant, or close gaps between infrastructural systems or functions, particularly in places where technical infrastructures are lacking. While Addie (2021: 1352) has pointed out how people-centered infrastructural systems are also at work in the global North, and Thieme (2018: 534) points out cross-regional similarities in the informality that is characteristic of even centrally-planned infrastructures, scholars have emphasized labor's function as a supplement to or replacement of infrastructural systems while the labor that builds technical systems has gone unremarked: For Fredericks, trash collection in Dakar is particularly significant because this infrastructure “is primarily composed of laboring bodies” (2014: 532).
While conceptions of people or social relations as infrastructure have provided ways to understand infrastructural systems through the people that make them work, this manuscript's focus on the (re)construction of technical structures such as Shanghai's fiber-optic network provides a new way to understand what Lemanski and Massey call “relational infrastructure networks” by investigating how manual labor and specific social relations facilitate the development of highly technical, networked infrastructures: as they argue, “people and communities
One stream of literature that has provides a window on these questions traces what Niranjana (2021) has called “engineering epistemologies”. While early research by Star (1999) examined role of engineers and designers producing information systems, scholars have more recently focused on the kinds of engineering knowledge necessary for the construction of physical infrastructural networks such as roads and water pipes. Building on work by Harvey and Knox (2015), whose ethnographic research on road-building emphasizes how engineers translate abstract plans into physical space, scholars have examined numerous contingent ways in which engineers create knowledge that makes infrastructure possible. This process often involves on-the-ground construction and labor: Lisa Björkman, for example, provides a penetrating analysis of how intimate infrastructural knowledge and the long experience of water engineers is integral to keeping water flowing water in Mumbai (Björkman, 2015). Working in the same city, Nikhil Anand theorizes water distribution as a system in which engineers, residents, pipes, and the earth that pipes are buried in work together in contingent ways. Both scholars are sensitive to the hands-on labor that is necessary for the creation of sociotechnical knowledge, describing how experienced engineers create complex maps of underground structures (Anand, 2017: 177–178) and rely on social relationships as well as technical skills to understand the origins of leaks (Björkman, 2015: 131–134). By emphasizing the “craft dimension of engineering” (Niranjana, 2021: 2), such scholarship reveals the role of on-the-ground experience as necessary for the (re)production of infrastructure. Nevertheless, this work has largely avoided the question of how the infrastructural knowledge used by engineers is put into place through construction processes that rely on supposedly “low-skilled” workers, or how information necessary to map structures that are by their very nature kept out of sight is produced.
To obtain an understanding of how technical infrastructures are built, it is necessary to look beyond engineering and towards on-the-ground, low-waged labor. As Michelle Buckley has forcefully argued, ethnography of construction processes can help us understand how “the material production of urban built environments can depend on the parallel production of complex inequalities and intersecting forms of social difference” (Buckley, 2014: 342). The importance of understanding who is involved in construction is also made clear by Datta and Brickell's ethnography of construction labor in London, which demonstrates how material questions of quality in the built environment are intimately tied to the identities of Polish migrant workers who have gained a central position in the industry: For workers and end-users of infrastructural systems, differences in knowledge, skill, and levels of quality mean that it matters who builds a project and how they go about their work (Datta and Brickell, 2009). At the same time, low-waged construction labor can generate epistemologies that differ significantly from those of engineers and project designers. Research by Alejandro de Coss-Corzo in Mexico City describes an improvisational “patchwork” performed by low-waged laborers who lack formal training but nevertheless gain considerable embodied expertise in understanding how to work on and with underground pipes (Coss-Corzo, 2020, 2021a). Still, while research has begun to demonstrate the close relationship between migration into the construction industry and urbanization across the global South (see Adhikari and Deshingkar, 2015; Kumar and Fernandez, 2016), the role of informal, migrant construction labor in infrastructural (re)development has been under-explored. In response, the remainder of this article examines Shanghai's Overhead-Underground project from the perspective of self-described “low-level” construction workers in the Chinese fiber-optic construction industry, first introducing the project itself before discussing the labor relations that facilitate infrastructural construction in Shanghai, and finally analyzing the production of infrastructural knowledge about the fiber-optic network.
Laying fiber with True Standard
This article is based on 18 months of fieldwork in Shanghai, conducted between the spring of 2018 and winter of 2020 and including nine months of participant-observation research on construction teams employed by True Standard Construction, 1 a medium-sized contractor in the fiber optic sector that was one of many at work on the Overhead-Underground project. As a privately run construction contractor that takes contracts from better-connected firms, True Standard construction is a medium-sized construction operator. The company specializes in the installation of underground fiber-optic infrastructure, and employs 15 work-teams of 8–12 workers each, divided according to type of work performed: Trenchless conduit installation via horizontal directed drilling, excavation for the installation of fiber conduit and the construction of manholes, and cable teams that work on the installation and/or removal of fiber cables from existing underground and overhead conduits or utility poles.
During this period, I lived and worked with a series of three work teams in company-provided housing, participating in all stages of the construction process. In addition to participant observation and informal conversation on construction sites and in living spaces, I conducted a series of twenty in-depth semi-structured interviews with temporary and long-term ordinary workers, skilled workers, and machinery operators from each of the teams I worked on, as well as team leaders, labor recruiters, project managers, and company management. While I had initial concerns about working as an un-skilled researcher alongside men who have spent decades working in this and other construction industries, my experience was similar to that of Coss-Corzo, who reports that while working on Mexican infrastructure projects, his willingness to follow commands, match what others were doing, and anticipate the next step in the process led to gradual incorporation as a low-skilled member of work teams (Coss-Corzo, 2020). As has been pointed out in other research (see Swider, 2015: 141–147), it would be impossible to do reliable fieldwork in this industry without first building the kind of real relationships that were only possible through direct participation in construction work.
Rural migrants and urban infrastructure
True Standard, like other contractors in the Chinese construction industry, depends heavily on the labor of rural-to-urban migrant workers, who now make up nearly 98% of China's construction workforce (China Labor Bulletin, 2019). This labor force is highly informal: 90% of construction workers lack formal labor contracts, and are instead employed through an informal subcontracting system that separates “on-the-ground” labor of migrant workers from project planning, engineering, and other “skilled” elements of construction (Pun and Lu, 2009; Ren and Gu, 2012). Workers at True Standard are no exception, relying on oral agreements with team leaders or labor recruiters who bring them into the company. Wages are paid upon the completion of an agreed term of service, most often prior to workers’ annual return to their hometowns for Spring Festival in late January or early February. A smaller number of workers are brought on seasonally, for terms as short as a week or as long as two or three months, and are paid upon completion of their short-term employment. These conditions are consistent with the broader Chinese construction industry (Peng, 1996; Pun et al., 2012; Swider, 2015). Despite reforms, wage theft remains a major risk for workers across the industry, and several informants have since left True Standard to work elsewhere in the construction industry, either because they either feared a rise in unpaid wages in the future, or because the company had amassed debts in back wages that they estimated would never be paid (c.f. Pun and Lu, 2010b).
Reliance on migrant labor is one product of the Chinese construction industry's division of planning and engineering tasks from on-the-ground construction labor. While construction in the pre-reform period prior to 1979 was primarily conducted through state-owned enterprises, the reform area saw the widespread establishment of private construction contractors relying primarily on subcontracted migrant labor (China Construction Culture Center, 2000). As scholarship on the construction industry has pointed out, informal employment structures suppress wages for construction workers, as well as creating the conditions for widespread wage theft across the industry (Pun and Lu, 2010a, 2010b). At the same time, the intersection of labor subcontracting systems with urban citizenship structures implemented through the
Mr Cao, a True Standard team leader in his late fifties, has an unusually long experience in the Shanghai telecommunications industry, and speaks to the creation of this system as he describes the growing prominence of migrant labor in infrastructure construction. Mr Cao first arrived in Shanghai in 1989, and witnessed the transformations of the industry firsthand. As he explains, telecommunications infrastructure work had previously been dominated by local Shanghainese team leaders and bosses, who later built companies and work teams by hiring lower-level staff from his native Anhui or other adjacent provinces. However, the 1990s saw a clear transition across the industry, which he frames not as an increase in exploitative use of migrant labor, but as migrants out-competing local construction workers and contractors: When I first came, we mostly dealt with Shanghainese––that is, they were Shanghainese leading the work teams. At the time, I was with a large installation team. Originally it had been Shanghainese people [doing the work], but when we peasant workers (农民工) came into the city, the Shanghainese took the opportunity to get out. After that, development happened so fast, they couldn't keep up with the times. So they get peasant workers to do the whole thing.
In the years following this transition, Shanghainese locals have been unwilling to work in construction in any capacity except at the highest levels of company administration or highly skilled technical roles. Because of his early arrival, Mr Cao learned Shanghainese to work effectively with local bosses, and is proud of the cultural cachet and connection to the city that this fluency brings. However, at this point local Shanghainese are now only involved in the construction process at the highest levels: Du Zong, the founder and operator of True Standard, began as a migrant construction worker before founding the company using capital his father accumulated as a migrant construction worker in the early 1980s. Much of Du's work as the head of the company now revolves around the cultivation of relationships and negotiation of contracts with higher-level contractors and local government officials who are almost universally Shanghainese.
As migrant workers took over most aspects of the construction industry, a constellation of legal and social structures ensured that were unable to settle down in Shanghai. These include explicit urban citizenship and bordering policies that exclude migrants from formal urban membership and most city services (Zhang, 2018), as well as deeply rooted Shanghainese localism and exclusion of “outsiders” over multiple generations (Honig, 1992). Meanwhile, state policies that aimed to take advantage of surplus labor across rural China had the effect of dramatically increasing the rural–urban divide in wealth and income, and ensuring that rural livelihoods could not be sustained without urban wage labor (Pun et al., 2012; Swider, 2015). The result is a labor regime built on the implicit assumption that workers will return to their home provinces rather than remaining in the cities they help build.
A strong division of labor between teams that specialize in excavation, horizontal directed drilling, and cable installation also means that individual workers rarely if ever have an understanding of infrastructural networks as a whole. Instead, workers like Xiao Dong, who has nearly 30 years of construction experience and has been working for the same construction team for half a decade, build knowledge about specific components that they work with every day, developing techniques for repairing broken digging tools, un-screwing stuck drill rods, and recognizing differences in the specification of inscrutable rubber-coated fiber-optic cable at a glance. These skillsets are partly aligned with what Coss-Corzo (2021b) describes as a process of “getting a feel for infrastructure,” as construction and repair workers learn a variety of sensorial techniques that facilitate infrastructural construction. However, Dong's position as the self-proclaimed “most skilled of the unskilled workers” does not lead to input in project planning or make him essential to understanding the network as a whole. Instead, due to the nature of rural–urban migrant labor, embodied skills are disconnected from both broader knowledge about the network on one hand, and potential advancement in the industry on the other. Workers also regularly move in and out of the construction industry often, and experience in the fiber-optic construction industry was not generally understood as the endpoint of migrant journies that might take any given worker to a different city or construction sub-sector if offered a better opportunity, or if wages went unpaid.
The crucial role that patterns of migration play in the (re)construction of Shanghai's telecommunications infrastructure is made particularly visible when large portions of the workforce leaves Shanghai and returns home. Beyond the emptying of construction sites during Chinese New Year, the pace of construction at True Standard is structured around the rhythms of agricultural life in northern Anhui, where the majority of the company's staff originates. In late September, a cable-installation team that normally relies on the labor of ten to twelve workers was reduced to four staff, including the team lead, as older team members one by one returned home to harvest corn. Xiao Ke, the youngest member of the team and one of the three who did not return home, complained bitterly that although he wanted a break, management staff did not allow him to leave: “I want to go home during the harvest season too, but they won't let me. If I went, there wouldn't be anyone who knows how to work left! Right now, out of the two cable teams [in our dormitory], I’m the only person doing any work at all!” While Ke's boast may not have been precisely accurate, he illuminates real constraints facing the work team during several-week periods in which a majority of team members had returned to their hometowns: annual harvest times result in two-week rest periods in which many of True Standard's work teams are unable to complete projects as normal.
Reliance on migrant workers for the construction of telecommunications infrastructure directly entangles the physical installation of fiber-optic infrastructure with the various sociopolitical structures governing migration, from the cadence with which projects can be completed to questions of labor supply. This is not to say that a reliance on migrant labor has hindered the Overhead-Underground project or Shanghai's infrastructural development. Rather, as Swider argues, the reliance on migrant work functions to lower the overall cost of labor for the Chinese construction industry as a whole (Swider, 2015). It is clear that reduced labor costs have an important role in facilitating large-scale infrastructural redevelopment like the Overhead-Underground project. However, this reliance on migrant labor also configures the complex work necessary to turn infrastructure projects into material reality. In particular, the labor subcontracting system and urban citizenship structures differentiate the migrant workers who are engaged in building the Overhead-Underground project from well-trained engineers or formal staff of state infrastructure management companies. For example, highly-trained water with decades of experience are key to the operation of the Mumbai hydraulic infrastructure, and are employed by longstanding state institutions (Anand, 2017; Björkman, 2015). No such knowledge is available in the Overhead-Underground project. Instead, low-waged workers are employed in flexible and informal roles that facilitate construction without creating a connection between individual workers and the spaces they build.
Thinking with the concept of people as infrastructure, it is apparent that the material network of subterranean fiber-optic communications infrastructure is in fact a relational infrastructure network that is made possible by and deeply entangled with seemingly disparate structures of urban (non)citizenship, rural welfare policy, and other migration management policies that ensure the provision of a low-waged migrant workforce for urban construction. This analysis expands on scholarship that has pointed out how peripheral or disadvantaged groups (including rural-urban migrant workers and impoverished youth) actively create urban infrastructures throughout the global south (cf. Fredericks, 2014, Gidwani, 2015), as well as providing a broader view of the kinds of communities that should be considered when researching relational infrastructure networks (Lemanski and Massey, 2022). Finally, a close examination of construction labor in the Overhead-Underground project also demonstrates the importance of understanding how infrastructural knowledge is created: If migrant workers cannot develop broad understandings of networks, what kinds of knowledge about infrastructure are possible?
Infrastructural knowledge from the ground up
Because one of the Overhead-Underground project's goals in reconfiguring Shanghai's fiber-optic network is to render the system more effectively manageable, it is particularly important to consider how knowledge about infrastructure that is necessary for construction and management is built and collated. While infrastructural knowledge is often associated with engineering epistemologies that feed into planning and project management, migrant labor is central to the accumulation of systematized, manageable, and transferable knowledge about Shanghai's fiber-optic network. Despite the lack of deep individual knowledge of broader infrastructural systems or projects described in the previous section, informal migrant workers’ sensory and mapping labor are necessary to accumulate the kinds of infrastructural knowledge necessary for the Overhead-Underground project.
In planning meetings for a proposed project that would re-route overhead cable along a downtown road, True Standard management staff discussed the logistics of bidding on the project. Before settling on a reasonable bid, however, company head Du Zong insisted that it would be impossible to discuss details without sending a work team to confirm the actual status of the existing cable network: How many cables were installed along the stretch of road in question, where they branched off to connect to other areas of the local network, and how many manholes or other access points would be required? In order to put together an accurate and competitive bid that could turn a profit, the company needed a precise understanding how many cables would have to be moved, where they were located, and which techniques could be used to build the new cable-routing infrastructure. This information could not be obtained without sending teams of workers to “feel out” the existing lines.
The difficulty of generating detailed knowledge about infrastructural installations is a partial motivation for the broader Overhead-Underground project: While it is possible to hang fiber-optic cables from existing pylons without formally registering them, creating complex tangles of overhead cabling that the project announcement refers to as “black pollution,” cables routed through underground conduits can be effectively managed, particularly as an increasing number of manholes are being equipped with controlled-access covers. Without this systematic and controlled structure, however, the re-construction of an existing network that has been created through an ad-hoc process of accumulation requires the creation of intimate knowledge about fiber-optic infrastructure that does not currently exist. In order to generate accurate maps of any part of the network, work teams made up of ordinary workers are directly responsible for the bottom-up mapping of existing infrastructure.
In late fall of 2019, Xiao Li's cable-installation team was assigned to map an area that would be affected by the construction of a highway off-ramp for the purpose of bidding on the future project. For two weeks, the team painstakingly followed overhead and underground cables in an area spanning several square kilometers, recording where each cable connected to a trunk line hung along electrical pylons following a main road. The team split up into groups of two, climbing bamboo ladders and tugging on each cable to ensure that each connection did in fact lead back to the original trunk line, and not to some other point located outside the planned construction area. As work progressed, notes about existing cables were translated into increasingly detailed representations: First, hand-drawn paper maps gave schematic overviews of cables traversing a series of poles and manholes in the project area. In the work team's dormitory, Xiao Li re-drew these sectional maps by hand, rendering them into a broader network schematic labeled with locations in relation to roads and intersections. Eventually, engineers at True Standard's headquarters translated this hand-drawn map into a GIS and printed maps displaying accurate technical information about the position and status of existing telecommunications infrastructure across the project area. To create this map required weeks of on-the-ground labor, and similar information for the surrounding area simply does not exist (See Figure 1).
Mapping existing cable installations. Left: Counting and labeling overhead cables in downtown Shanghai. Right: Drafting a network map. Photos by author.
On-the-ground construction workers play a key role in collecting and collating knowledge about the infrastructural networks. Maps of existing networks must be developed by teams of informally-employed migrant workers, rather than emerging from previous documentation or the knowledge of local engineers or other experts. While the kinds of research, standards-management, and on-the-ground adaptation that Harvey and Knox describe as being key to the role of civil engineers in road construction are part of the planning and development of the Overhead-Underground project, the analytical-adaptive perspective of trained engineers was rarely visible in the everyday experience of project implementation (cf. Harvey and Knox, 2015: 79–111). Instead, accumulation of basic knowledge about Shanghai's fiber-optic network is facilitated by the direct work “unskilled” migrant workers, who are not positioned to develop precise knowledge of the broad fiber-optic network despite years of experience.
At the same time, migrant workers do develop their own forms of knowledge about the network that can exceed the types of infrastructural knowledge reported through maps of existing structures. Through experience mapping and working with the material components of fiber-optic infrastructures, construction workers develop the kind of embodied “feel” for network infrastructures that Coss-Corzo (2021b) describes as elements of sensorial infrastructure management. However, these forms of infrastructural knowledge are not always brought to bear on the eventual form the network takes. The following ethnographic vignettes describe how construction workers’ feel for infrastructure plays out in the context of corruption and “fake cables” on one hand, and unauthorized infrastructural installations on the other.
During the process of tracing cable routes along a main road, it became clear to members of Xiao Li's team that the thick bundles of fiber-optic cable strung between electric power pylons on either side of the road were not connected to anything at either end, and was instead made up of “fake cables” that were installed without the intention of ever being used. Although these cables appeared real from the ground, they lacked connections on either end. On closer inspection, it became clear the cables themselves were manufactured to different specifications than those required for cables that transmit data through the network, a fact that experienced workers were able to recognize by touch, as the “fake” cables lacked a steel stiffening core. As Wen, a skilled and energetic member of Xiao Li's team observed somewhat bitterly, “we always know when a cable is fake.” Relying on years of experience across the construction industry, Wen and other team members theorized that the cables had been installed to inflate the value of the original road construction project. Following the discovery, Wen and other team members regularly grumbled about the long hours they were spending numbering and labeling cables that would never be used: As Wen put it, the cables (and the effort spent to map them) amounted to a waste of public funds, “stealing money from the nation.”
In another instance, Xiao Li's cable team spent several days mapping a network of cables that spanned several blocks in a downtown district. Most cable runs in the area already complied with mandates to install new fiber-optic lines underground, so rather than climbing pylons to check overhead cables, workers instead climbed down into hip-deep water in manholes and junction points, leapfrogging from manhole to manhole while counting the numbers of cables entering and exiting against official figures. Working with other team members in adjacent access points, workers pulled on individual cables and confirmed which cables moved, building an understanding of the direction each individual cable took as it ran from one junction to another. As the team moved along the branching network of underground cables, it became clear that the number of cables in certain access points was higher than the numbers that Xiao Li, as team leader, had been told to expect. This excess of cable indicated the existence of an unregistered fiber network that had been surreptitiously installed in the area. As some workers had begun to suspect, the potentially illicit cables were a bundle that had been marked as “military use” cables, with a warning attached. Following a brief debate in which ordinary workers refused to cut the offending lines for fear of reprisal, Xiao Li conferred with management over the phone before severing the cables himself. This severing of connectivity precipitated the rapid arrival of a representative from a client of the unauthorized infrastructure––not a military installation, but a nearby, recently constructed office complex. While the representative refused to offer his name or that of his employer, the team's severing of access apparently forced a settlement: Several days later, a few strands of cable were re-spliced to restore connectivity, but more than half of the unauthorized network was slated for removal. In the end, Xiao Li's team was tasked with removing the remains of the illicit network, re-tracing their steps and pulling each length out by hand. Once again, the work truck was filled with mounds of cable as the team returned to their dormitory on the periphery of the city.
Because accurate information about existing infrastructure can only be confirmed through bottom-up mapping performed by construction workers like Wen and Xiao Li, planners and engineers currently lack the ability to create a complete map of Shanghai's fiber-optic network. Instead, the dirty and laborious work of mapping existing infrastructure is a key step in the creation of infrastructural knowledge. This systematization of infrastructural knowledge relies on the sensory experience of workers touching and manipulating cables as a prerequisite to the systematized and ordered “engineering epistemologies” described by other research on infrastructure (Niranjana, 2021). While engineers are certainly present in the planning and execution of network reconstruction, their understandings rely on the physical work of “feeling out” existing cable lines (摸线), harkening back to the sensorial understanding material infrastructural components that Coss-Corzo associates with construction labor (Coss-Corzo, 2021a; Niranjana, 2021).
The conditions that structure migrant construction labor in Shanghai also affect the kinds of infrastructural knowledge that are possible. A combination of low wages for migrant labor and abundant infrastructure investment facilitates the production of accurate maps, but accumulated knowledge about the network does not result in individual workers’ embodied knowledge of network structures becoming essential to the operation of fiber infrastructure. Instead, subcontracting systems and the assumption that workers are “temporary” residents in the cities where they work result in an on-the-ground construction process in which workers who have developed sensorial skills in working with material infrastructure are often kept in the dark about basic details of specific projects or their relation to the network as a whole. However, the understandings that workers build about infrastructure are not limited to the creation of network schematics. In particular, workers’ lengthy experience working with the network means that even short-term staff quickly learn to understand the installation of “fake” cables installed for assumed corrupt interests. Despite grumbles about corruption, the maps Xiao Li's team produced did not indicate the fact that the cables they had mapped were unused, and it is likely that the same number of cables was included on bids to move the entire installation underground: Although the work of the Overhead-Underground project to move communications infrastructure underground may result in a better-mapped and more controllable network, repeated mapping and physical labor will still be required for the management of the underground cable network the project has created, adding ever-growing numbers of connections to the network while removing unauthorized cable installations.
Conclusions
This article has argued that understanding how infrastructure is (re)constructed through on-the-ground labor is crucial to scholarship on infrastructure as a whole. In particular, I have outlined how migrant construction labor plays an integral role in both the material reconstruction of Shanghai's fiber-optic infrastructure, and in the creation of the kinds of infrastructural knowledge that are crucial to this type of reconstruction project. Taking Shanghai's Overhead-Underground infrastructure upgrading project as an example, I have explored how the material structures of seemingly “high-tech” infrastructure can only be managed and understood through the incorporation of low-wage, de-skilled migrant labor. While the Overhead-Underground project may create a more manageable infrastructural system, repair, maintenance, and management of the system itself continue to rely on migrant labor. In turn, the continued availability of nominally “temporary” migrant workers is itself structured by diverse material and social systems, from the exclusion of low-waged migrant workers from urban citizenship under the
Building on both literatures that have emphasized people as infrastructure (Addie, 2021; Simone, 2004, 2021; Wilson and Jonas, 2021) and the forms of knowledge that make infrastructures possible (Coss-Corzo, 2021a; Harvey and Knox, 2015; Niranjana, 2021), I have also argued for the importance of ethnography of construction labor as a platform from which to understand the specificities of infrastructural knowledge that are created and utilized in the construction of infrastructural systems. There are important similarities between the Shanghai construction industry and construction around the world. Although practices of subcontracting in the Chinese construction industry have their own distinct history, the widespread dependence on structurally and spatially devalued migrant labor has been noted in construction research in numerous other contexts (e.g. Adhikari and Deshingkar, 2015; Datta and Brickell, 2009; Kumar and Fernandez, 2016). Despite these similarities, locally specific research on how infrastructural construction plays out in different places is a necessary means to understand how infrastructures are produced and reproduced. After all, the conditions under which Shanghai's fiber-optic infrastructure can be produced and reproduced through the Overhead-Underground project differ significantly from the circumstances documented in other literature that has investigated infrastructural repair and maintenance (Anand, 2017; Björkman, 2015; Coss-Corzo, 2020). To build a more complete account of how infrastructure functions, it is crucial that scholars understand the processes of repair and construction that keep infrastructures running. In Shanghai and elsewhere, this means understanding infrastructural networks in parallel with social relations and forms of complex inequality that structure construction labor (Buckley, 2014, 2019). Doing so will bring construction labor more clearly into the conversation around how urban infrastructure is made possible, and at the same time open up new and important linkages between infrastructure research and migration policy, housing, and the use of urban space. For scholars of urban China in particular, it is important to understand policies such as Beijing's forced demolition of housing as part of a campaign to reduce the so-called “low-end population,” (Li et al., 2018), or Shanghai's more recent targeted campaign against “group rental” (see Harten, 2021) as reconfiguring core functions that include infrastructure construction. Elsewhere, the social relations that facilitate construction labor should be understood as key factors making infrastructure networks work.
Footnotes
Declaration of conflicting interests
Funding
ORCID iD
