Chinese Journal of Communication
ISSN: 1754-4750 (Print) 1754-4769 (Online) Journal homepage: https://www.tandfonline.com/loi/rcjc20
Digital utility: Datafication, regulation, labor, and
DiDi’s platformization of urban transport in China
Julie Yujie Chen & Jack Linchuan Qiu
To cite this article: Julie Yujie Chen & Jack Linchuan Qiu (2019): Digital utility: Datafication,
regulation, labor, and DiDi’s platformization of urban transport in China, Chinese Journal of
Communication, DOI: 10.1080/17544750.2019.1614964
To link to this article: https://doi.org/10.1080/17544750.2019.1614964
Published online: 28 May 2019.
Submit your article to this journal
View Crossmark data
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=rcjc20
Chinese Journal of Communication, 2019
https://dx.doi.org/10.1080/17544750.2019.1614964
Digital utility: Datafication, regulation, labor, and DiDi’s
platformization of urban transport in China
Julie Yujie Chena
and Jack Linchuan Qiub
a
School of Media, Communication and Sociology, University of Leicester, Leicester, UK;
School of Journalism and Communication, the Chinese University of Hong Kong, Hong
Kong, China
b
This article develops the critical concept of digital utility through studying the
case of DiDi Chuxing and the platformization of transport services in urban
China. By examining DiDi’s business model, its datafication strategies, its
relations with the Chinese government, and its labor management systems, the
article demonstrates how the platformization of transport is emblematic of a
private company becoming a digital utility provider. With technological
imagination and practical inconsistency, this process remediates service delivery
while reworking infrastructures and redefining the access to public and private
services. We argue that platform companies are able to become digital utility
suppliers because of their capacity to straddle the public and the private
sectors, their aspiration to become “ecosystem builders,” and their heavy
reliance on the constant intensive labor of users, particularly drivers, to
produce data. However, these factors also make instability a definitive feature
of digital utility companies in their present condition. Morphing into the
terrain of utilities is a common undertaking by DiDi and similar platform
companies. To problematize the logics of digital utility, especially its laborintensive datafication processes and its complex relations with regulators,
provides a conceptual anchor for further debates on the infrastructuralization
of platforms and the platformization of society.
Keywords: digital utility; ride-hailing platforms; informal economy; digital
labor; algorithms
Introduction
According to official statistics, ride-hailing apps served more than 480 million
Chinese in 2017 (State Information Center & Internet Society of China, 2018,
p. 30), including most of China’s 793 million urban population. The predominant
company is DiDi Chuxing, a private startup that was founded in Beijing in 2012,
subsequently monopolizing 94.6% of China’s ride-hailing market (Xiao, 2017). By
2017, DiDi operated in more than 400 Chinese cities, where the company’s 21 million registered drivers handled more than 30 million rides each day (DiDi, 2018a).
This daily operational volume far surpasses Uber in the US (Crabtree, 2018);
hence, DiDi is the world’s leading ride-hailing platform.
When apps become the mediator of rides, they constitute an underlying digital
condition for transportation services because Chinese urban dwellers depend on
them for taxi services. Analogous to electricity, these apps provide a new type of
utility – a digital utility – in urban transport.
Corresponding author. Email: julieyj.chen@gmail.com
ß 2019 The Centre for Chinese Media and Comparative Communication Research, The Chinese University of Hong Kong
2
Julie Yujie Chen and Jack Linchuan Qiu
The concept of digital utility is not new. Since the early 2000s, digital utility
has been discussed by critical media scholars, who have emphasized the public service functions of the Internet (Murdock, 2005; Moe, 2008; Andrejevic, 2013). The
concept is also in line with proposals by computer scientists to consider computing, especially cloud computing, as the fifth utility after water, electricity, gas, and
telephony (Buyya, Yeo, Venugopal, Broberg, & Brandic, 2009). It has been long
established in economics that utilities markets are prone to monopolization
(Newberry, 2002), so they require state regulation to balance the interests of
investors (i.e. capital) and consumers (i.e. citizens and workers) in pursuit of longterm system stability and the common good (Demsetz, 1968). The recent rise of
digital platforms, such as Uber and DiDi, have made urban mobility both mundanely convenient and seemingly harmless; hence, users tend to forego their inner
logic, while the lack of clear regulations hinders a systematic analysis of this phenomenon. According to van Dijck, Poell, and de Waal (2018, p. 139), “[t]he current platform ecosystem is predicated on an architecture that is primarily staked
in, and driven by, economic values and corporate interests,” which often endangers “public values.” Therefore, it is urgent, both theoretically and practically, to
critically examine the concept of digital utility and apply it to the analysis of realworld cases.
We attempt to fill the gap in the work on this phenomenon by examining the
digital utility provider DiDi. The ascendency of DiDi and the platformization of
transport services in China provide valuable case studies to explore what characterizes a digital utility and a digital utility supplier. The following questions guide
the present study: How does an app-based platform grow into a digital utility provider? What are DiDi’s datafication strategies, its complex relationships with the
Chinese government as an “infrastructural state,” and its labor practices, especially
regarding the intensive labor performed by its driver-workers? How does DiDi
resemble and differ from Uber in the platformization of transport? To what extent
does our analysis of the transport sector shed light on the platformization of
Chinese society?
This study is part of a two-year research project conducted from 2015 to 2017
to examine the digital platforms for on-demand services in China. The data were
collected from primary and secondary materials, including official statistics on
Internet users and the digital economy in China, public data, reports on DiDi,
and news reports, as well as official policies and regulatory documents at national
and subnational levels. We collected information from DiDi’s official websites in
both Chinese and English, as well as research reports published or sponsored by
DiDi. We conducted more than 60 interviews with DiDi drivers in several cities
during the project. The argument in this article, however, is not developed through
examining the primary interview data, but by critically examining the secondary
materials collected from government, corporate, and news documents. Our critical
point of view nevertheless benefited from the driver’s insights and our knowledge
of their perspective.
In this article, we first discuss recent developments in the field of platform
studies and develop the concept of digital utility to examine the public, private,
and digital interfaces of the ride-hailing platforms. We then present our analysis
of ride-hailing platforms as developing category of digital utility through the
example of DiDi. The analysis is organized as follows: (1) the company’s business
model centered on datafication and financialization; (2) the discursive and
Chinese Journal of Communication
3
practical strategies embodied in the company’s relationships with national and
local governments; (3) the obscure labor management system involved in the platformization of transport.
Platformization, infrastructure, and digital utility
Platform research has opened multiple lines of inquiry and invited trans-disciplinary theoretical discussion. Early studies revealed that politics influenced platforms
and that corporate logic shaped both platforms and user experiences (Gillespie,
2010). The development of a platform involves continuous negotiations among the
designers, the company, the user community, the socio-economic context and
larger regulatory frameworks (Gillespie, 2010; Plantin, et al, 2018; van Dijck, et al,
2018). Apart from fulfilling the need for socialization and online expression, digital
platform companies have increased in both number and type across the globe.
Although their business models differ, collectively, digital platforms have surfaced
as a significant force for economic configuration as well as social control (Kenney
& Zysman, 2016; Srnicek, 2016; van Dijck et al., 2018). Accordingly, scholars
have shifted their research foci from technical architecture and affordances to the
organizing and dominating power of platforms in the economy and the society.
For example, after examining four types of platforms in news, urban transport,
health care, and education, van Dijck et al. (2018) found that “public values”
must be considered in assessing the transformation to the platform society. Digital
utility, as a common good that is delivered by monopolistic platforms to society,
is one of the conceptual instruments used to articulate public values outside the
corporate and technological spheres.
According to Plantin et al. (2018), the distinction between infrastructure and
platform has become blurred in today’s digital environment. On one hand, private
digital tech companies fill the vacuum when governments retreat from building
and maintaining universally accessible infrastructures. Such infrastructures are
platformized, which implies the continuous enclosure of the open web and the
lock-ins by monopolistic private gatekeepers (Zittrain, 2009). Simultaneously, giant
platforms, such as Facebook and Google, have achieved a tremendous scale,
reaching far beyond their core businesses. This phenomenon is known as the
“infrastructuralization” of platforms.
An increasing number of studies is concerned with contextualizing digital platforms and/or their effects on life experiences. In his account of “platform capitalism,” Srnicek (2016, p. 48) considered platforms “a new type of firm,” the core
business strategy of which was “extracting and controlling data.” Previous studies
that placed platform work and workers in their historical and socioeconomic contexts contributed to our understanding of the ways in which established social
relationships are embedded in the actual processes of crowd work and how they
contribute to building solidarity among dispersed workers (Gray, Suri, Ali, &
Kulkarni, 2016). Other studies highlighted the importance of geography and the
colonial legacy (e.g. certain language abilities of the residents in former colonies)
for the uneven distribution of online work and the unfair competition among
workers (Graham, Hjorth, & Lehdonvirta, 2017; Irani, 2015b).
It is thus imperative to unravel the complex dynamics that shape the historical
development of platforms in specific sectors. For instance, van Dijck et al. (2018,
4
Julie Yujie Chen and Jack Linchuan Qiu
p. 73) stressed that urban transport “is largely a market sector, but it has a considerable public interest” focusing on “two central public values: the quality of urban
transport and the organization of labor and workers’ rights.” Van Dijck et al.
understood “quality” as not only the narrow consumer viewpoint of using transportation services with efficiency and comfort, but also “consumer protection, passenger safety, inclusiveness … , universal service provision … , and affordability.”
Articulated in European contexts, this wide conceptualization sheds light on our
interrogation of DiDi’s digital utility in Chinese cities.
Two questions remain unaddressed in the field of platform studies. First, the
dual process of “platform-as-infrastructure, and vice versa” (Plantin et al., 2018,
pp. 306–307) deserves more rigorous scholarly interrogation, particularly regarding
the reasons why certain platforms become infrastructure while others do not.
What developmental course does “platform-as-infrastructure” take in specific
national contexts? Second, the “extractivist” nature of platform capitalism has
been widely acknowledged (Scholz, 2016; Srnicek, 2016), but little analysis has
been conducted regarding how companies leverage their financial, technological,
and discursive strategies, among others, to accomplish data extraction.
We develop the concept of digital utility to critically evaluate the characteristics
of “platform-as-infrastructure.” We define digital utility as (1) public services and
values, which was proposed by Murdock (2005), Moe (2008), Andrejevic (2013),
and van Dijck et al (2018) and (2) transformed digital conditions embedded in the
access to public or private services, without which the utilization of such services
would be categorically different (Plantin et al., 2018). Thus, digital utility may be
understood to be a subset of the services delivered by platform infrastructures, the
public value of which is discounted, thereby requiring increased state regulation.
The data-driven nature of the platform economy (Srnicek, 2016) determines the
transmutation to digital utility in the constant flux of data capture. Therefore,
digital utility is intertwined with data production and manipulation in users’ access
to and utilization of the services mediated by infrastructural platforms.
The creation and supply of digital utility distinguishes infrastructural platforms
from non-infrastructural platforms. Google is a digital utility platform for the following reasons: (1) it transforms the digital conditions for Internet users to navigate and utilize the web; (2) it expands to several other domains, including maps
and books (Plantin et al., 2018); (3) Google’s service could not be accomplished
without its streamlined data collection and manipulation mechanism. In China,
ride-hailing apps can also be assessed as digital utility platforms because without
them, it would be substantially difficult to obtain a taxi service nowadays. These
apps have also overlain the urban infrastructure of roads and traffic, catalyzing
the datafied transport system. In contrast, Airbnb currently falls short of being a
digital utility supplier because although it arguably has changed the short-term
rental market, it has not yet been linked to other infrastructures to the extent that
it would fundamentally affect tourists’ access to accommodation.
Digital utility simultaneously constitutes and alters the terms of access to public and private services. Not all interactions on platforms take place on equal settings or generate identical outcomes. On the contrary, “data-based discrimination”
(Gangadharan, 2014, p. 2) or “selection” (van Dijck et al., 2018, p. 40) is the new
norm, which is particularly the case in labor platforms, which systematically reinforces the economic disadvantages of marginalized workers (Chen, 2017; van
Doorn, 2017). Workers and consumers install different versions of the same app,
Chinese Journal of Communication
5
wherein the former face far more intensive data extraction and surveillance, such
as the rating system, tracking, and feedback, than do their client counterparts
(Irani, 2015b; Rosenblat & Stark, 2016). Consequently, on these platforms, the
unequal access determined by digital platforms and automated by algorithms has
the dual function of turning participant workers into datafication workers.
The distinction between digital utility and traditional utilities, such as electricity, is crucial: the source of power for digital platforms – datafication – depends
not only on algorithms, technology, and non-human natural resources, but also,
more crucially and intensively, on humans themselves. Although it could be
argued that telephony also draws on human resources, such as caller networks, in
providing a utility service, the data extraction from human supplementers on
today’s platform infrastructures, which we call “digital utility labor,” is qualitatively different in economic value and cybernetic control. Labor relations scholar
Xiaoyi Wen contended, “the real nature of the sharing economy is the laborintensive economy” (2018).
Platformization of transport and the rise of a digital utility supplier
The rise of DiDi as China’s ride-hailing platform monopoly demonstrates universal patterns of datafication and the intensive extraction of labor from supplementers. However, the specific case of China and its historical context are equally
important in developing the concept of digital utility.
First, China is a socialist country that is governed as an “infrastructural state”
(Bach, 2016). Without the direct investment by government in its infrastructure,
DiDi became an unparalleled behemoth with a 94.6% share of the ride-hailing
market (Xiao, 2017). Beginning in the 1980s, taxi services in China were part of
the urban public transport systems. However, in 2016, the Measures for
Administration of Urban Taxi Business Operations and Services defined the taxi service as a “supplement” to public transport, which means that the business can be
privately owned or operated, but it must serve the public interest. Ride-hailing
platforms such as DiDi offer both traditional taxi services and private car services.
Hence, DiDi insists on its self-branding as a “one-stop” transportation platform,
differing from Uber, which claims to be a “connective platform” in Europe (van
Dijck et al, 2018, p. 73). In China, the state regulation of ride-haling apps met little opposition because of the sectoral history of taxi services. This history has led
China to become one of the first nations to legalize and regulate ride-hailing apps.
An equally important legacy is that China has never had a nationally centralized taxi management system. Instead, each city has rules and regulatory structures, thereby forming a “multi-headed, multi-tiered management” system (Chen,
2017, p. 7) that continues to characterize DiDi’s business today. Wealthy local
governments on the east coast, such as in Shanghai, tend to take more responsibility, sometimes directly owning or subsidizing taxi businesses, than less-developed
cities in the hinterlands, which follow a hands-off approach or prey on taxi drivers
and companies through exploitative rent-seeking.
A common grievance among drivers concerns the ways in which local governments treat unlicensed vehicles, which are known as “black cars” (heiche) persisting from taxi industry to the ride-hailing market. Ride-hailing apps, including
DiDi, transformed urban transport in China after it overtook the US in becoming
6
Julie Yujie Chen and Jack Linchuan Qiu
the world’s largest car market (Branigan, 2012). From 2000 to 2012, private car
sales in China soared from 4 million to about 18 million. The ride-hailing market
expanded phenomenally within a few years. In 2017, 278 million people used apps
for traditional taxi services. and 217 million people used apps to hire a private car
(CNNIC, 2017, p. 26). On the DiDi platform, 30 million rides are delivered every
day (DiDi, 2018a). DiDi had 450 million users and 21 million registered drivers by
the end of 2017, surpassing Uber’s performance worldwide (Crabtree, 2018; DiDi
Institute of Policy Study, 2017). In DiDi’s report, however, the company did not
clarify the number of legally licensed vehicles or the number of “black cars.”
Datafication, financialization, and “DiDi Traffic”
Similar to its Western counterparts, DiDi’s business strategy involves multiple
integrated operations that help achieve the company’s monopoly, maintain its
competitive advantage, and reshape the terms of access to private transport service. At the core of all operations are datafication and financialization.
Founded in the post-2009 context of quantitative easing, which has provided
structural backing and penetration by the financial sector globally, DiDi combines
financialization to pursue market shares with the construction of its data capture
infrastructures. Jia and Winseck (2018) showed that the leading Chinese Internet
companies, Tencent, Alibaba, and Baidu, resemble global giants like Apple in their
integration into the financial market as investors and recipients of transnational
capital. DiDi is similarly financialized in global capital flows. Among the top
investors in DiDi are the tech companies Apple, Alibaba, Booking, Softbank, and
Tencent (Crunchbase, 2018). In addition to private funds, DiDi attracts investment
from state-owned enterprises such as China Life Insurance, which is the largest life
insurer in China, and the China Investment Corporation, which is a sovereign
wealth fund that manages China's foreign exchange reserves.
Internet companies also buy out other companies to make inroads into
unfamiliar markets and increase their competitiveness (Jia & Winseck, 2018).
Mergers and acquisitions are typical financialization maneuvers to achieve this
end. DiDi’s merger with Kuaidi in 2015 and its subsequent purchase of Uber
China in 2016 helped establish its monopoly in China. Globally, DiDi launched
Kuaidi Taxi in Hong Kong, acquired 99, Brazil’s largest rideshare company, and
invested in or established strategic partnerships with transport platforms for both
traditional taxis and private hires in Europe, Africa, Middle East, Southeast and
South Asia, and North Africa (DiDi, 2018b).
What sets mergers and acquisitions by platform companies apart is their special concern with data – the essential assets in financialization measures, because
they enable platform companies to possess colossal amounts of data on users and
habituated data capture infrastructures. Data underlie DiDi’s platform operation
and its expansion to a wide range of urban transport from taxi services to diverse
private car services to the recent bike-sharing and smart-traffic systems. In 2018,
DiDi handled 30 million ride requests daily, which generated 106 terabytes of data
per day. Consequently, DiDi processed more than 4,875 terabytes of data daily
(DiDi, 2018a), which were collected to train DiDi’s algorithms, ranging from
pricing and job-allocation to predicting and mitigating traffic congestion and
developing “a safe driving system” (Xiao, 2017).
Chinese Journal of Communication
7
However, unlike patterns in Western countries, DiDi has not disrupted the taxi
industry by replacing taxi drivers with private-gig drivers, which is based on corporate discursive and political strategies at various levels of the Chinese government. DiDi has established partnerships with more than 50 leading taxi companies
in Tier-1 and Tier-2 Chinese cities. Moreover, it has aimed to datafy the urban
transport ecosystem, including taxis, and occupy the center of the converging networks of information, traffic, and transactions involving all kinds of vehicles and
transport services. The company calls this converging datafied system “DiDi
Traffic,” which is in operation in more than 50 cities. The scheme integrates data
captured by DiDi with data sources from local government and related business
partners to manage city traffic.
DiDi Traffic is far more than “a traffic information platform” that offers local
transportation authorities real-time traffic information (DiDi Chuxing
Development Research Institute, 2017, p. 36). It has enabled DiDi to become a
central component in the datafied urban transport infrastructure. Data and data
capture mechanisms have allowed DiDi to shape the digital conditions for urban
transport. Hence, DiDi enters the terrain of supplying digital utility in
urban transport.
Political strategy: align with the central but wrestle with the local
According to Srnicek (2016, p. 92), “far from being owners of information, these
[platform] companies are becoming owners of the infrastructures of the society
[emphasis added].” Infrastructure, however, is always embedded in “other structures, social arrangements and technologies” (Star & Ruhleder, 1996, p. 113).
Infrastructural construction is as material as it is affective because it involves the
imagination of infrastructure as the pathway toward a modern, more advanced,
and better future (Larkin, 2013). DiDi’s rapid expansion and platformization of
transport services in China, therefore, must be examined against the backdrop of
China as a prominent “infrastructural state” (Bach, 2016) and the company’s
infrastructural encounters at the local level. DiDi’s political strategy is characterized by (1) its alignment with the development and technocratic rhetoric of the
Central Government and (2) its contestation of local authorities, which has had
mixed outcomes. The interplay among the technocratic rhetoric adopted by the
platform company, the local state regulations, and existing informal practices has
exacerbated the complexity and contradiction of the governance of transport platforms and digital utility.
China is an “infrastructural state” that channels 43% of its total investment in
infrastructure (Bach, 2016) from building networks of road and railways to constructing urban skyscrapers. Recently, the authorities not only have invested heavily in upgrading telecommunication infrastructures but also have appropriated
Internet and related technologies (e.g. automation and AI) into a discourse of
techno-panacea on modernization, economic restructuring, and social development
(Hong, 2017). DiDi has aligned with this technocratic narrative by positioning
itself as the technological solution provider to problems such as structural
under- or unemployment and traffic congestion. In so doing, the company has
enlisted itself to become part of the driving force to materialize the nation
8
Julie Yujie Chen and Jack Linchuan Qiu
state’s imagination and its desire to become a global ICT leader through an
“infrastructural fix” (Bach, 2016).
Different from the “cultural work” done by crowdwork platforms to dehumanize and invisiblize workers, DiDi’s job-creation rhetoric does the “cultural work”
(Irani, 2015b) of sublimating the platform’s pursuit of its private interest in a discourse of social service provision and contribution to the national strategy for economic restructuring. In addition to promoting the number of jobs created by the
platform, DiDi tends to highlight how flexibility in “gig driving” work helps mitigate the infliction of an economic slowdown on ordinary workers. For example, in
2016, DiDi claimed to have created 17 million flexible jobs, stressing that 2.4 million jobs were held by workers in heavy-industry sectors that had been affected by
national initiatives to reduce excessive industrial capacity. DiDi’s job-creation figures were cited in an official report on China’s sharing economy (DiDi, 2016;
State Information Center & Internet Society of China, 2017). Another report on
DiDi’s gig drivers in 2017 stated that more than 21 million drivers had earned
part or all of their income from DiDi, and the number of drivers from the capacity-cutting sectors had climbed to 3.9 million (DiDi Institute of Policy Study,
2017). The emerging gig driving work is promoted by both the company and governmental agencies as a benign solution to absorb millions of newly laid-off factory workers or underemployed workers. However, empirical studies have shown
contradictory evidence that new gig drivers are concentrated in and are more likely
to come from economically better-off coastal regions than places affected by capacity-cutting policies (DiDi Institute of Policy Study, 2017).
DiDi also has attempted to broker its infrastructural imagination of a datafied
transportation system to local cities. DiDi Traffic is an example. Although cities
allegedly have taken advantage of the talent on DiDi’s engineering team, the platform company has benefited considerably from data-sharing and consolidation
with local authorities because this partnership enables DiDi to design and be part
of the emergent digital transmutation of the urban traffic and transport system.
Moreover, it allows DiDi to further straddle the line between private and public
services. No specific law or policy exists to regulate the sharing and/or commodification of data except the general provision in China Cybersecurity Law to protect
individual’s private data. Because of this regulatory vacuum and DiDi’s dual position, the company benefits from what legal scholar Victor Fleischer called
“regulatory arbitrage” (cited in Calo & Rosenblat, 2017, p. 1627).
When a platform that is designed to harness data at scale sinks in and intersects with local infrastructure and power relationships, incongruities surface. In
contrast to its high-profile market expansion, DiDi and its drivers remain in a
legal “gray zone.” As of June 2018, DiDi has obtained legal operational licenses
in only 51 cities in China while it operated in over 400 cities (Yue, 2018), and
legitimate platform drivers for DiDi accounted for only 0.6% of the entire workforce (China News, 2017). Most DiDi drivers in effect earn money outside regulations or formal institutions. They are unlicensed platform drivers similar to the
“black car” drivers in the days prior to platformization.
Previous studies showed that digital platforms may formalize and standardize
previously informal service sectors (Ticona & Mateescu, 2018). DiDi has formalized transport services to some extent. By standardizing algorithms that function
as labor control mechanisms disciplining worker-drivers (Rosenblat & Stark,
2016), the platform consolidates DiDi’s control and brings several informal
Chinese Journal of Communication
9
practices to the verge of regulation. Private fare negotiations and cash transactions
have been replaced by pricing algorithms and money transfers through integrated
third-party payment apps, such as WeChat Pay and Alipay. Transaction data on
both ride-hailing platforms and third-party payment apps are traceable.
After the national legalization of ride-hailing platforms in 2016, more than two
thirds of Chinese cities passed and enacted local regulations with varying restrictions on platform drivers and vehicles (Ma & Li, 2017). These restrictions are
focused on qualifications for platform drivers and requirements for vehicles. Three
quarters of the city governments in China, however, have imposed local residency
mandates on drivers (Ma & Li, 2017). Although migrant workers account for the
majority of DiDi drivers, these local regulations preclude them from this employment. Legal barriers of this kind have reduced the number of legitimate platform
drivers, which reflects the local resistance against platformization and the acceptance of the digital utility provided by DiDi. For instance, several cities in Jiangsu
Province issued fines for DiDi’s lack of legal operational licenses in its jurisdiction
as well as the illegal recruitment of unqualified drivers to work on the platform
(Yangtze Evening Paper, 2017).
In practice, DiDi not only acquiesces in the participation of “black car” drivers
to work on its platform (Chen, 2017) but also breaches local regulations brazenly.
In addition to the fines imposed on DiDi, local authorities prohibited unlicensed
platform drivers in the same manner as they used to repress “black cars” in taxi
industry. Hence, the legacy of “multi-headed, multi-tiered management” characterizing the taxi industry has been extended and exacerbated. Local frictions have
caused informal practices to persist during and after platformization, which has
benefited DiDi, which continues to exploit the precarious economic position of
its drivers.
The contradiction between the efforts at formalization by national and local
regulatory bodies, the ride-hailing apps, and the preservation of informal or even
illegitimate practices describes the positions of both DiDi and its drivers. The
incongruities inherent in the ride-hailing platform have allowed its development
into a digital utility, through which DiDi aspires to both scalability and applicability in local settings. The scalability of tech companies is achieved by reducing
workers to “computational power” (Irani, 2015a), whereas the applicability to
local settings demands the agility of drivers who remain in an informal state,
whether it is voluntary or not. DiDi’s complicity in the selective formalization of
its ride service is not coincidental but deliberate.
Platformization and digital utility labor
What is the relationship between the digital utility provider DiDi and its labor
force? In this section, we explain how DiDi’s platformization of transport precipitated a profound shift in platform labor relations toward discriminatory data
extraction. This shift, we argue, lays bare the centrality of digital utility labor,
which means that DiDi’s drivers not only transport people but also feed data into
the system to train algorithms and sustain the platform’s operations and infrastructure. Consequently, Datafied digital utility labor is a distinct feature of
today’s platform infrastructures, which was not observed in traditional
utility markets.
10
Julie Yujie Chen and Jack Linchuan Qiu
In May 2015, the CEO of DiDi, Cheng Wei, announced his Great Tidal
Strategy, or “Tides” (Guo, 2015). In the following three years, Cheng discussed
his vision on multiple occasions, including his keynote speech at China’s Big Data
Industry Summit in 2016 (People.cn, 2016), and expanded it from a strategy to
“integrate professional transport ability with socially scattered transport ability to
satisfy people’s transportation demand in the peak and off peak hours” (Guo,
2015) to a strategy for collaborating with local governments to solve traffic jams
and build a “smart city” (DiDi Chuxing Development Research Institute, 2017).
Since 2015, DiDi has dominated China’s online transport services by offering a
full range of on-demand mobility options, including taxis, private car hailing, ridesharing, chauffeurs (i.e. designated drivers), business limousines, minibuses, and
car rentals. “Tides” was introduced in addition to the surge-pricing algorithm as a
tool for managing and controlling the supply side of transport services. Surge pricing, which incentivizes drivers by cash rewards, can be self-defeating because it
could generate an influx of service providers and then reduce the incentives.
However, “Tides” is more than an algorithm; it is a system that rations, dispatches, and routes all vehicles that are available on the platform to accommodate
transport demands based on the optimal calculations of real-time data (Crouch,
2016). Instead of reacting to the outbreak of an imbalanced supply–demand market, “Tides” is deployed to map the distances and destinations of all requested
rides and to strategize the routes for each driver to travel, thus precluding a surge
price. DiDi aims to “predict traffic hotspots in advance” and “outsmart traffic”
through data analytics (Crouch, 2016). “Tides” is a grand plan for restructuring
urban transport which depends on an enormous amount of data. If it were in full
swing, it would not only help “mitigate traffic jams,” but orchestrate and control
local traffic. It would become the protocol for digital transport infrastructure,
expanding Galloway’s (2001) idea of protocol as the control mechanism for decentralized networks. “Tides” is in effect a cybernetic strategy for the control of platform labor.
Studies on labor platforms have documented the constitutive role played by
algorithms in labor management for quality control, labor performance monitoring and surveillance, and labor valorizations (Chen, 2017; Rosenblat & Stark,
2016). The cybernetic labor control mechanism of “Tides” is concerned with both
labor discipline and data capture. It extracts data through digital utility labor and
disciplines drivers through feeding data into algorithms. As such, “Tides” creates a
two-tiered access and data extraction wherein the drivers produce far more data
than do the passengers, which makes platformization of transport services in
China essentially a “labor-intensive economy” of datafication.
According to Etherington (2016), DiDi collects information about drivers’
locations and speed every three seconds. In a newly developed “safe driving system,” a dashboard camera captures everything from road conditions, following
distances, and reckless driving to potential signs of fatigue (Xiao, 2017). Through
sensors on phones, DiDi gathers data on drivers’ patterns of accelerating, braking,
and steering (Crabtree, 2018). The trip and service histories of drivers are tracked
and categorized by algorithms to predict whether an individual driver has the propensity to inflate passenger fares by taking overly long routes. When these drivers
are detected, “Tides” places these drivers on a dispatch list of drivers that are
eligible for only short-distance trip (e.g. within three miles).
Chinese Journal of Communication
11
“Tides” epitomizes what Zuboff described as the “big other”: operations of
“unexpected and often illegible mechanisms of extraction and control that exile
persons from their own behavior” (2015, p. 85). Zuboff (2015) and others (e.g.
Fuchs, 2010; Turow & Couldry, 2018) acknowledged that all participants in the
“big other” are subject to data extraction, and in the case of DiDi, data are
extracted from both drivers and passengers. Drivers carry out a double task by
offering a transport service and generating data, which constitutes the digital utility labor. The intensification of data capture by drivers could be interpreted as the
multiplication of labor in the work process as a means of generating value to
“secure and obscure surplus labor” (Burawoy, 1979, p. 81) from drivers for
datafication.
The uneven work of data production that pervades labor platforms is justified
by the reasoning that workers need to be controlled and disciplined. A crucial condition for systems such as “Tides” is to have drivers follow instructions given by
algorithms, such as a suggested route. Therefore, the driver must accept the allocated order, instead of going home. Thus, discriminatory data production has
become the default foundation of DiDi and digital utility platforms. Because of
the structural precariousness of the migrant workers and laid-off workers who
comprise the majority of DiDi drivers, these informal drivers are easy prey in
DiDi’s business model, which was first designed to attract taxi drivers and then
was expanded to other transport markets (Chen, 2017). DiDi has become a surveillance-based data extraction company that has enjoyed China’s current lax data
protection laws in its ascendance to a market monopoly. Hence, based on the case
study of DiDi, digital utility labor is both a distinct resource and an inherently
vulnerable component of digital utility platforms.
Discussion
The case of DiDi raises new questions in the current approach to the Chinese
digital economy, which has tended to be top–down in investigations of the power
relationships among capital, private enterprises, and the state (e.g. Hong, 2017; Jia
& Winseck, 2018; Zhao, 2007). No more “dancing” with foreign “wolves,” at least
in the digital platform-based transport sector, domestic companies, such as DiDi,
have become the “wolf” that devours foreign apps, market shares, workers’ labor,
and, most importantly, the data generated by all parties involved. Our study has
examined the historical dimensions of a platform growing into a digital utility
company, thereby transforming the relations between drivers and the platform, as
well as between the platform and governments at various levels. Based on our analysis, the following can be said about the platformization of ride services in China
and about our contribution to the current debate about the platform society
through the conception of digital utility as both an empirical process of pervasive
service delivery and a normative construction of infrastructural platforms.
Based on our analysis of DiDi, platformization includes two processes, both of
which are essential to the provision of digital utilities: (1) the digital mediation of
transport services, including the driving process; (2) the reworking of digital and
social infrastructures for urban transport through the datafication of the first process. A dozen ride-hailing apps, including DiDi, have contributed to the first process. DiDi’s rise to a monopoly set the second process in motion. The case study
12
Julie Yujie Chen and Jack Linchuan Qiu
of DiDi complicates the discourse of innovation, which is often deemed oppositional and “disruptive” of existing models. Undoubtedly, DiDi has transformed
habitual ride services in urban China by becoming an indispensable digital utility
provider. Nevertheless, the company is reluctant to disrupt the existing structures
of the taxi industry because it benefits immensely from its exploitation of informal
labor, which it continues to employ in the datafication of its transport system.
Our analyses about the platformization of urban transport thus offer a glimpse
into the platformization of Chinese society in general. DiDi’s datafication practices
and its infrastructural imagination are shared with the government.
Despite its strategies and its rhetoric of standardization to create a better,
modern transport system, DiDi has not, and probably will not, eradicate its informal practices in the new digital utility system, particularly its reliance on informal
worker-drivers. It is therefore very likely that the relationship between DiDi and
Chinese authorities, especially local state governments, will never be as fully formalized as the traditional utility sectors have been. The new digital utility firms
sometimes may outsmart the authorities by using datafication power and algorithms. At other times, local states, which need to collect taxes to maintain roads
and bridges while facing the consequences of unemployment and inadequate urban
transport may resist the transmutation into digital utilities.
Building on Murdock (2005), Moe (2008), Andrejevic (2013) and van Dijck
et al. (2018), we developed the concept of digital utility to investigate the profound
changes brought about by DiDi and platformization, as well as their consequences
for the articulation of public values and the enhancement of the public welfare.
Our conceptualization of digital utility is based on the socio-techno transformation
and the engineered differential access and hence on the different experiences of privately owned platforms that have become essential in public life. Nonetheless,
DiDi is only one market-based model of the platformization of society. Other
models are possible that could be either state-led or government-owned or based
on platform co-ops and similar civil society initiatives (van Dijck et al., 2018).
The concept of digital utility captures characteristics of platform infrastructure
that have yet to be fully articulated. It shifts scholarly attention away from the technicality of platforms to the outcomes they generate for the larger socioeconomic system and for the public interest. In particular, our analysis of DiDi’s datafication
strategies has determined the constitutive features of digital utility, particularly those
that enable certain platforms to become an infrastructure, which has laid the groundwork for future scholars to investigate the “hidden work” and “invisible workers”
(Crain et al., 2016) in app-based service delivery. This last point is applicable beyond
the platformization of transport in China. For example, the questions of who creates
digital utility and under what conditions should be among the first to be addressed
in examinations of a new platform infrastructure in China or elsewhere.
Because they resemble traditional utilities, digital utility companies should be
obligated to serve the public interest. However, in reality, they often act on selfinterest and ignore workers’ rights; therefore, digital utility labor is a distinct
source of systemic volatility. This revelation has wider analytical applicability to
the contemporary debate on digital platforms and governance, which transcends
the contexts of China and its ride-hailing sector. Digital utility platforms are characterized by algorithmic flexibility and reliance on platform workers to produce
data, which creates the potential for labor resistance. Hence, there is another key
question for future research: Does digital utility labor render the platform
Chinese Journal of Communication
13
infrastructure more fragile than do traditional infrastructure systems? Moreover, it
is important to note that a digital utility is not independent of traditional utilities.
Instead, it extends from and repurposes existing utilities. Thus, the knowledge
about traditional utility infrastructures and their historical transmutations may
lead to innovative analyses of digital utilities and the future of platformization.
Conclusion
This article is among the first to report the development of the concept of digital
utility in a case study of DiDi, including its corporate history, datafication models, relationship with Chinese authorities, and labor management system. Our
findings showed three empirical features that characterized the current digital
utility platform: (1) DiDi’s capacity to straddle formal and informal economies
at the margins of government regulation, especially in relation to local state regulators; (2) its capacity to be an “ecosystem-builder” through financialization,
datafication, and infrastructural projects, such as DiDi Traffic; (3) its constant
and intensive data extraction practices, which are made possible by the exploitation of digital utility labor. These are the three primary reasons that DiDi has
become a digital utility provider of urban transport in China. While the first reason is probably specific to the Chinese context because of the unique development of the taxi industry in urban China, the second and third features are
globally relevant.
In addition to guiding the empirical analysis of datafication, the corporate–government relationship, and labor management systems, the concept of digital utility led us to critically evaluate the platforms’ public values along the lines
suggested by van Dijck et al. (2018). Consequently, digital utilities should be
widely accessible and affordable; they should be regulated at different levels of
government; they should adhere to international and local labor standards. Labor,
including digital utility labor, is the foundation and lifeblood of the sustainability
of digital utility. Utilities must be sustainable; otherwise they risk becoming the
worst urban hazard.
Acknowledgments
The authors would like to thank the anonymous reviewers and the editors of the special
issue – Jeroen de Kloet, Thomas Poell, and Guohua Zeng – for their valuable comments on
the manuscript.
Funding
Funding for this study was obtained from the Seed Money for Project on Digital Labor,
which is supported by the Faculty of Social Science at the Chinese University of Hong
Kong, Hong Kong SAR, and International Development Research Centre (project
#108339-007: Deliver on the promise of the platform economy in China).
ORCID
Julie Yujie Chen
http://orcid.org/0000-0002-0358-0894
14
Julie Yujie Chen and Jack Linchuan Qiu
Notes on contributors
Julie Yujie Chen is a lecturer in the School of Media, Communication, and Sociology at the
University of Leicester. Chen studies how cultural difference, technologies, and existing
economic structure affect the experience and perception of work in the digital age. Her
previous work has been published in New Media & Society among other journals, and she
is the lead author of Super-sticky WeChat and Chinese Society (2018).
Jack Linchuan Qiu is a professor at the School of Journalism and Communication at the
Chinese University of Hong Kong. He is the author of Working-class Network Society:
Communication Technology and the Information Have-less in Urban China (2009) and
Goodbye iSlave: A Manifesto for Digital Abolition (2016).
References
Andrejevic, M. (2013). Public service media utilities. Media International Australia, 146,
123–132. doi:10.1461444817729149/1329878X1314600116
Bach, J. (2016, August 30). China’s infrastructural fix. Retrieved from https://limn.it/
articles/chinas-infrastructural-fix/
Branigan, T. (2012, December 14). China and cars: A love story. The Guardian. Retrieved from
http://www.theguardian.com/world/2012/dec/14/china-worlds-biggest-new-car-market
Burawoy, M. (1979). Manufacturing consent. Chicago, IL: University of Chicago Press.
Buyya, R., Yeo, C. S., Venugopal, S., Broberg, J., & Brandic, I. (2009). Cloud computing and
emerging IT platforms: Vision, hype, and reality for delivering computing as the 5th utility. Future
Generation Computer Systems, 25(6), 599–616. Retrieved from doi:10.1016/j.future.2008.12.001
Calo, R., & Rosenblat, A. (2017). The taking economy: Uber, information, and power.
Columbia Law Review, 117(6), 1623–1690.
Chen, J. Y. (2017). Thrown under the bus and outrunning it! New Media & Society, 1–21.
Advance online publication. https://doi.org/10.1177/1461444817729149
CNNIC (China Internet Network Information Center). (2017). 40th statistical report on internet
development in China. Beijing, China. Retrieved from https://bit.ly/2SlJ5Ic (in Chinese)
Crabtree, J. (2018, April). DiDi Chuxing took on Uber and won. Now it’s taking on the
world. WIRED UK. Retrieved from https://bit.ly/2EW2t8M
Crain M.G., Poster, W., & Cherry, M. A. (Eds.) (2016). Invisible labor: Hidden work in the
contemporary world. Oakland, California: University of California Press.
Crouch, E. (2016, May 27). China’s DiDi wants to make traffic jams a thing of the past.
Retrieved from https://bit.ly/1Rwz2ak
Crunchbase. (2018, July). DiDi Chuxing funding rounds. Retrieved from https://bit.ly/2QO0Lzt
Demsetz, H. (1968). Why regulate utilities? The Journal of Law & Economics, 11(1), 55–65.
doi:10.1086/466643
DiDi. (2016). Job creation - VALUES - DiDichuxing.en. Retrieved from http://www.
DiDichuxing.com/en/values/job-creation
DiDi. (2018a). DiDi brain. Retrieved from https://www.DiDiglobal.com/science/brain (in Chinese)
DiDi. (2018b, April). Milestone. Retrieved from https://www.DiDiglobal.com/about-special/
milestone (in Chinese)
DiDi Chuxing Development Research Institute. (2017). New economy new service. Beijing,
China: DiDi Chuxing Development Research Institute. (in Chinese)
DiDi Institute of Policy Study. (2017). New economy, new jobs. Beijing, China: DiDi
Institute of Policy Study. (in Chinese)
Etherington, D. (2016, December 2). DiDi’s CTO explains why China’s ride-sharing giant
has a data advantage. Retrieved from https://tcrn.ch/2CyIKwt
Fuchs, C. (2010). Labor in informational capitalism on the Internet. The Information
Society, 26, 179–196. doi:10.1080/01972241003712215
Galloway, A. (2001). Protocol, or, how control exists after decentralization. Rethinking
Marxism, 13(3), 81–88. doi:10.1080/089356901101241758
Gangadharan, S. P. (2014). Data and discrimination. Washington, DC: New America.
Retrieved from https://bit.ly/2LD0oSI
Gillespie, T. (2010). The politics of “platforms.” New Media & Society, 12(3), 347–364. doi:
10.1177/1461444809342738
Chinese Journal of Communication
15
Graham, M., Hjorth, I., & Lehdonvirta, V. (2017). Digital labor and development.
Transfer: European Review of Labour and Research, 23(2), 135–162. doi:10.1177/
1024258916687250
Gray, M. L., Suri, S., Ali, S. S., & Kulkarni, D. (2016). The crowd is a collaborative
network. In CSCW ’16 Proceedings of the 19th ACM (pp. 134–147). ACM Press.
Retrieved from https://doi.org/10.1145/2818048.2819942
Guo, G. (2015, May 22). DiDi Kuaidi releases tides strategy. Retrieved from https://www.
iyiou.com/p/17838.html (in Chinese).
Hong, Y. (2017). Networking China. Urbana, IL: University of Illinois Press.
Irani, L. (2015a). Difference and dependence among digital workers. South Atlantic
Quarterly, 114(1), 225–234. doi:10.1215/00382876-2831665
Irani, L. (2015b). The cultural work of microwork. New Media & Society, 17(5), 720–739.
doi:10.1177/1461444813511926
Jia, L., & Winseck, D. (2018). The political economy of Chinese internet companies:
Financialization, concentration, and capitalization. International Communication
Gazette, 80(1), 30–59. doi:10.1177/1748048517742783
Kenney, M., & Zysman, J. (2016, Spring). The rise of the platform economy. Issues in
Science and Technology, 32(3). http://issues.org/32-3/the-rise-of-the-platform-economy/
Larkin, B. (2013). The politics and poetics of infrastructure. Annual Review of
Anthropology, 42(1), 327–343. doi:10.1146/annurev-anthro-092412-155522
Ma, L., & Li, Y. (2017, November 28). China’s ride-hailing platform policies review and
suggestions. Retrieved from http://www.thepaper.cn/newsDetail_forward_1882926 (in
Chinese)
Moe, H. (2008). Dissemination and dialogue in the public sphere. Media, Culture &
Society, 30, 319–336. doi:10.1177/0163443708088790
Murdock, G. (2005). Building the digital commons. In G.F. Lowe & P. Jauert (Eds.),
Cultural dilemmas in public service broadcasting (pp. 213–231). Goteborg, Sweden:
Nordicom.
Newberry, D. M. (2002). Privatization, restructuring, and regulation of network utilities.
Cambridge, MA: MIT Press.
Plantin, J.-C., Lagoze, C., Edwards, P. N., & Sandvig, C. (2018). Infrastructure studies
meet platform studies in the age of Google and Facebook. New Media & Society, 20(1),
293–310. doi:10.1177/1461444816661553
People.cn. (2016, May 25). Speech by Cheng Wei. Retrieved from https://bit.ly/2CxgtWW
(in Chinese)
Rosenblat, A., & Stark, L. (2016). Algorithmic labor and information asymmetries.
International Journal of Communication, 10, 3758–3784.
Srnicek, N. (2016). Platform capitalism. Cambridge, UK: Polity Press.
Star, S. L., & Ruhleder, K. (1996). Steps toward an ecology of infrastructure. Information
Systems Research: ISR: A Journal of the Institute of Management Sciences., 7(1),
111–134. doi:10.1287/isre.7.1.111
State Information Center & Internet Society of China. (2018). China’s sharing economy
development report 2018. Beijing. Retrieved from https://bit.ly/2QTlmmd (in Chinese)
Ticona, J., & Mateescu, A. (2018). Trusted strangers: Carework platforms’ cultural
entrepreneurship in the on-demand economy. New Media & Society, 20(11), 4384–4404.
doi:10.1177/1461444818773727
Turow, J., & Couldry, N. (2018). Media as data extraction. Journal of Communication,
68(2), 415–423. doi:10.1093/joc/jqx011
Van Dijck, J., Poell, T., & de Waal, M. (2018). The platform society. New York: Oxford
University Press.
Van Doorn, N. (2017). Platform labor. Information, Communication & Society, 20(6),
898–914. Retrieved from https://doi.org/10.1080/1369118X.2017.1294194
Wen, X. (2018, December 10). The real nature of sharing economy is a labor-intensive
economy. The Paper. Retrieved from https://www.thepaper.cn/newsDetail_forward_
2716522 (in Chinese)
Xiao, E. (2017, April 29). DiDi’s quest for China’s transportation data. Retrieved from
https://bit.ly/2pNeSHM
16
Julie Yujie Chen and Jack Linchuan Qiu
Yangtze Evening Paper. (2017, November 15). Online and offline vehicles not match,
DiDi fined by four cities. Retrieved from http://www.xinhuanet.com/2017-11/15/c_
1121957526.htm (in Chinese)
Yue, Shaoxue. (2018, July 2). Rank of business licenses for close to a hundred ride-haling
platforms and their nationwide distribution. Retrieved from http://m.evpartner.com/
news/detail-37434.html (in Chinese)
Zhao, Y. (2007). After mobile phones, what? International Journal of Communication, 1(1),
92–120.
Zittrain, J. (2009). The future of the Internet. New Haven, CT: Yale University Press.
Zuboff, S. (2015). Big other: Surveillance capitalism and the prospects of an information
civilization. Rochester, NY: Social Science Research Network. Retrieved from https://
papers.ssrn.com/abstract¼2594754