Digital Enterprise, Digital Labour

Microenterprise in the platform era—identifying new ‘platform practices’

Daniel, a shopkeeper and online freelance writer

The platform era is undoubtedly transforming the way goods, labor, services, and attention are exchanged around the world. But to what extent will platformitized markets change the livelihoods of millions of self-employed workers and small and informal enterprises in the developing world? 

Recently, the Partnership for Finance in a Digital Africa (FiDA) —an initiative of the Mastercard Foundation coordinated by Caribou Digital—conducted a study about how platformization is impacting micro- and small enterprises (MSEs) and the self-employed, offering its findings as a set of interlocking essays and video profiles. The study drew on qualitative interviews with micro-entrepreneurs in Kenya about how they use social media, e-commerce marketplaces, and online freelancing websites. Instead of a narrow focus on adoption, we identified emerging “platform practices”—four distinct ways in which MSEs are changing their workflows to adapt to the platform age:

We also found two overarching themes requiring further discussion:

A brief video of Dorcas, one of five entrepreneurs featured on the report site

This post for DIODE focuses on the implications of our research for the digital development community. It is a shorter form of our research implications essay, and is best conveyed as four imperatives:

Broaden the research focus from mobiles and ICTs to platforms

One line of research nearly 20 years strong looks at microenterprises’ use of ICTs. Yet the bulk of this work, whether focused on mobile phones specifically or on ICTs in general, isn’t calibrated to today’s changing internet landscape. Two forms of platformization are becoming more relevant, and need extensive study:

  • Individual micro-entrepreneurs use social media platforms for informal and real-time coordination with suppliers and customers. We aren’t talking about specialized “WhatsApp for business”, or Facebook for business, or even paid digital advertisements here. Wyche, Forte & Schoenebeck illustrate how these consumer-facing omnibus communication platforms have become the lingua franca of commerce in digital Kenya, our new research further underscores the remarkable ways in which micro-entrepreneurs have appropriated these platforms for their own purposes. MSEs have adapted these social media platforms into toll-free numbers, customer relationship managers, billboards and direct mail, pre-sales advertising, and after-sales support, all rolled into one.
  • The same markets are also being platformitized by the entrance of transaction platforms: websites designed to host multi-sided markets between buyers and sellers. These sites have scrambled the traditional means of finding, transacting with, and supporting customers. However, although they increase MSEs’ exposure, they also increase competition. Our work only scratches the surface of how this new visibility and competition impacts the profitability of small enterprises, and illustrates how more research is needed to assess whether transaction platforms work for small-scale producers, especially in developing countries.

Ask “how”, not just “whether”

The FiDA platform lens—by dividing platform functions into aggregation and discovery, transaction support, credibility, and ancillary data—enables us to see both whether and how platforms are altering prospects for small enterprises. We think this approach has merit for future studies, and we suggest that researchers zero in on specific models of change to ask how the platforms they study, be they Facebook and WhatsApp, or specialized transaction platforms like OLX, are transforming the markets they enter. It’s also important that we get to granularity. Specific affordances matched with specific functions on specific platforms matched by specific platform practices (i.e., user behaviors) is where the rubber meets the road.

Specify and replicate

A close read of our findings should quickly underscore the need for (1) replication beyond Nairobi, (2) specificity of various market segments and types of microenterprise, and (3) complementarity of methods. Our largely qualitative interviews could be expanded via quantitative study of how and whether platform use changes the profitability and sustainability of these businesses. Our study also suffers from survivorship bias: we have only spoken to active microenterprises; we have yet to investigate whether and how platformization is driving some micro-entrepreneurs from their markets.

Avoid detaching studies of digitized, platformized markets from the physical world

Our two blurring essays, on tech and touch and digital side hustles, illustrate the material component of digital pursuits. The realities of doing business in resource-constrained contexts may require more face-to-face time, more trust-building, and more creative ways of bridging the last mile in goods and services than those of less constrained markets. These dynamics are best unpacked with the type of ethnographic or qualitative work we were able to do in this study. At the same time, the very concept of “side hustles” challenges what we mean by a “job” or “a livelihood.” Digital side hustles, such as online essay writing or virtual assistant work, blur the lines between taking charge of one’s own fate and being captive to someone else’s algorithm. The Fair Work Foundation is developing a body of evidence around digitized platform work in the Global South. Our study highlights those individuals who keep one foot in a job or material micro-business and another moonlighting in the digital space.

Next Steps

We’re grateful for the opportunity to share this work with DIODE. Another helpful exchange has been with our co-travelers in the Mastercard Foundation community of practice, specifically BFA’s FiBR project, which has been looking at platforms used by micro-entrepreneurs in Tanzania, and Mercy Corps’s research into agricultural platforms in Africa. As a set, these studies shed light on changing practices and market conditions in East Africa and beyond.

We are particularly excited about the prospects for platforms to build on what they’ve already started, becoming digital spaces in which participants acquire a variety of skills and digital literacies required to thrive in the platform era. We call this “transformational upskilling” and will spend much of 2019 looking at this topic in a more comprehensive way.

Have questions, feedback, or ideas for further research on MSEs and platform practices? Please contact us at ideas@financedigitalafrica.org or @FiDAPartnership on Twitter.

Digital Economy, Digital Enterprise

Data Marketplace as an Enabler for Data-Driven Innovations in Developing Countries

In today’s digital era, data have become a valuable resource for digital enterprise to innovate and stay competitive in the market. This also applies to developing countries, where start-ups offer new digital products and services and transform everyday lives. One way to realize data-driven innovations is by sharing data among parties and even competitors to generate new insights and improve the quality of products or services.

In this context, data marketplaces are emerging, which are described as platforms that bring multiple parties together to exchange data with each other. Data marketplaces could play a key role in enabling data-driven innovations for digital enterprises in developing countries since they face a significantly different challenges in terms of resources and capabilities than digital enterprises in developed countries. More importantly, data marketplaces might generate value for digital enterprises in developing countries, because new insights can be generated by analyzing data from multiple sources.

Various initiatives and pilots on data marketplaces are available (e.g. OpenDataSoft, Azure Data Market, Data Market Austria). However, they often struggle and fail to survive. This is mainly due to lack of trust, fear of losing competitive advantage when sharing data, or the risk of violating strict privacy regulations (e.g. GDPR). To tackle these challenges, in the Safe-DEED project (Safe Data-Enabled Economic Development) we focus on developing new privacy-preserving technologies to enable firms to expose data via data marketplaces without giving access to actual data. We are also conducting multi-actor business model research in order to bring those technologies to the market in a sustainable way so that digital enterprises can generate value from such a platform.

How is this relevant to the digital economy and development? There are several possible use cases within the digital economy and development context. For instance, SMEs could benefit from accessing data and information provided by other SMEs in data marketplaces to support market research, so that they can formulate a better strategy to reach their target market. Digital enterprises providing on-demand transportation services could also expose their data via data marketplaces, and support the government in developing traffic policies without revealing sensitive information. Furthermore, a peer-to-peer lending platform focusing on providing loans in rural areas can also expose data to conventional banks via data marketplaces to get a better picture of unbanked population in a country so that they can collaborate in achieving financial inclusion.

Of course, there are several challenges in realizing data marketplaces that are safe and secure, such as convincing the first couple of firms to join the data marketplace platform, the impact of such an innovation on trust, ethics, security, and finding a sustainable business model for the data marketplace itself. Hopefully, answers to these challenges will be available by the end of this project.

More information:

  1. Safe-DEED website: https://safe-deed.eu
  2. Finding a Business Model for Decentralized Data Marketplaces: https://safe-deed.eu/finding-a-business-model-for-decentralized-data-marketplaces
  3. Safe and secure data marketplaces for innovation: https://www.tudelft.nl/en/tpm/research/projects/safe-and-secure-data-marketplaces-for-innovation
Digital Labour

The Continuing Relevance of the State in the Age of Digital Gig Work

Technological change has always had an impact on the sphere of work. Every new form of work that has emerged has been accompanied by optimism about its potential, and concerns about its damaging effects. The same is true of digital gig work – a form of work that can be performed and delivered digitally via online platforms. A primary concern has been the exploitative nature of digital gig work, especially the micro-distribution models in which individuals, as “contractors”, register with the platform to complete their tasks (Meyers et al, 2017). Although being able to work from anywhere, at one’s convenience, on a digital platform may offer certain flexibility, the absence of physically proximate fellow-workers also leads to the atomization of the workforce (Kuek et al, 2015). With the supply of labour exceeding demand, there are questions about the tendency of workers to undercut one another in their eagerness to obtain work, and whether that will lead to a race to the bottom in terms of wages, reduce the bargaining power of digital gig workers and perpetuate socio-spatial disparities (Graham et al, 2017).

Historically, including in the 20th century, states have taken an active developmental role to ensure improvements in labour conditions and in aggregate standards of living. Is it possible for the state to address similar concerns that arise in the context of digital gig work, when technology gives platforms a global reach? This is an urgent question as a 2016 study by the Oxford Internet Institute reveals that a significant share of the gig workforce is in the less affluent regions of the world, primarily undertaking menial and repetitive tasks, while two-thirds of job vacancies are posted by employers from the developed world, including the United States, the United Kingdom, Australia and Canada (Lehdonvirta, 2017).

Such demarcation between the source and destination of work, and the attendant concerns about the digital gig economy, are reminiscent of arguments about the New International Division of Labour (NIDL). NIDL theorists, such as Frobel et al (1980), were concerned with “bloody-Taylorization” following the growth in the offshoring of low-skill, low value-added manufacturing from the 1960s. Yet, for at least four Newly Industrializing Countries (NICs) in East Asia – Singapore, South Korea, Taiwan, and Hong Kong – becoming a part of the NIDL offered a path to technological ‘catch-up’ and ‘late-industrialization’. Those NICs showed how, despite starting off with less-than-desirable work, it is possible to develop indigenous capability to move toward more high-skilled, higher value-added activity. In all these NICs, the dominant institutional force behind the socio-economic transformation was the state.

Amsden and Chu (2003), for instance, point to how Taiwan entered the world semiconductor market by offering low-skill assembly work. Over time, state investments in the education system and technological acquisition, its provision of financial incentives including tariffs and subsidies, and the building of physical infrastructure, created an industry which gave Taiwan a “second-mover advantage” and made the country the world’s leading semiconductor supplier. Specifically, firms pioneered the pure-play foundry, an organizational model in which they became adept at original design and manufacture without necessarily marketing their own brands. Thus, contrary to the fears of the NIDL theorists, the East Asian NICs showed that developing countries have the agency to move beyond exploitative niches in the international division of labour with technological upgrading. More broadly, this highlights that, while concerns about the impact of type and conditions of work on socio-spatial disparities are far from misplaced, there are also institutional means of overcoming these concerns.

The concentration of digital gig work in less affluent countries is largely due to a lack of better local job opportunities. Concerns about the precarious nature of digital gig work are equally true for locally available work in the vast informal sector, with gig work at least offering better monetary returns. But this is not an argument for the status quo. Rather, it is a call to draw from lessons about how institutional means, especially the state, can be deployed to ensure that menial and repetitive digital gig work becomes a stepping stone to more rewarding work and not the only choice arising out of a lack of other options.

One possible initiative by the state is to build, or assist private players in building platforms for local needs which can find global use. With such platforms within its jurisdiction, the state can regulate them to ensure adherence to socially acceptable norms of pay and work conditions. Local language platforms could be encouraged to ensure that digital gig work opportunities are available to a larger section of the population. To undertake such initiatives, the state must minimally promote what the East Asian NICs did so effectively to overcome the debilitating aspects of 20th century manufacturing – investment in education (to ensure employability) and physical infrastructure, financial incentives, and opportunities for ‘learning by doing’. Since current technology and global conditions differ from what they were 50 years ago, the specifics of state interventions will vary. But history suggests the state can play a useful role in harnessing the potential of digital gig work for social transformation. These trajectories and trends are explored at length by the author in a paper being written for the DIODE working paper series.

References
Amsden, A. and Chu, W-w. 2003. Beyond Late Industrialization: Taiwan’s Upgrading Policies. Cambridge, MA: MIT Press

Frobel, F, J. Heinrichs, and O. Kreye. 1980. The New International Division of Labor: Structural Unemployment in Industrialised Countries and Industrialisation in Developing Countries. Cambridge, UK: Cambridge University Press.

Graham, M., V. Lehdonvirta, A. Wood, H. Barnard, I. Hjorth, and D. P. Simon. 2017. The Risks and Rewards of Online Gig Work At The Global Margins. https://www.oii.ox.ac.uk/publications/ gigwork.pdf (accessed 24th June 2018).

Kuek, Siou Chew et. al. 2015. The Global Opportunity in Online Outsourcing. Washington D.C.: World Bank Group.

Lehdonvirta, V. 2017.Where are online workers located? The international division of digital gig work. http://ilabour.oii.ox.ac.uk/where-are-online-workers-located-the-international-division-of-digital-gig-work/ (accessed 24th June 2018)

Meyers, L., B. Minic, L. Raftree and T. Hurst. 2017. The Nexus of Microwork and Impact Sourcing: Implications for Youth Employment. http://gcyerti.com/wp-content/uploads/2017/02/The-Nexus-of-Microwork-and-Impact-Sourcing_Final_ONLINE_02.28.17_v2.pdf (accessed 25 May 2018)

Digital Labour

Development Implications of Digital Platform Labour

A new paper – “Understanding the Development Implications of Online Outsourcing: A Study of Digital Labour Platforms in Pakistan” – analyses the experiences of some of the millions of gig workers who undertake digital labour in developing countries via platforms such as Upwork and Freelancer.

Using the sustainable livelihoods framework as the basis for analysis, it identifies four things from interviews with workers and other stakeholders in remote areas of Northern Pakistan:

a) Employment Push: The context of politico-economic vulnerability that pushes unemployed individuals into digital work including lack of alternative employment, political instability and concerns about Islamic extremism.

b) Barriers to Gig Work: The typical barriers to digital gig work for those in more remote areas of developing countries. These include poor quality of technical infrastructure such as power and broadband connectivity; a lack of relevant knowledge and skills or the means to obtain them; limitations of current financial payment systems; and cultural norms that do not see online freelancing as constituting a “job”.

c) Worker Trajectories: The four trajectories of digital gig workers who go through training schemes: sinkers (the majority who never undertake digital platform work), strugglers (who try but appear largely unable to make a living), survivors (who can earn small amounts from digital gig work), and swimmers (who flourish and are able to build a career path via digital platforms).

d) Role of Institutions: The “re-institutionalisation” of digital labour. Notwithstanding narratives of the de-institutionalisation of digital gig work, experience in Pakistan shows three institutional forces impinging on online outsourcing to marginalised groups. There are the digital platforms themselves; often seen as improving the context for outsourcing work.  There are interventions of formal organisations – development, government and NGO agencies – who help overcome asset deficits that would otherwise exclude these groups from online outsourcing.  And there are informal linkages between freelancers themselves which provide assistance and work sub-contracts.

Digital Labour

Mobile Microwork in South Africa

When Richard launched the DIODE network, I took this as a cue to initiate a mini-empirical research project and got Zaakirah Roomaney, one of my honours research students, to look at mobile microwork in South Africa. We just wrote up the findings as a conference paper which my colleague/co-researcher Pitso Tsibolane will present at AMCIS, New Orleans, in August. However, I thought it would be useful to mention some of our key findings here as well. (We’ll upload the paper after the conference.)

We sent out a survey request and link via Facebook, soliciting people with mobile micro-work experience or interested in participating. The sample consisted of 125 valid responses of which 70% (n=87) of respondents were female and 30% (n=38) were male. The age distribution was positively skewed with the largest response from the 20-29 years age group which represented 45% (n=17) of the male respondents being male but a fully 71% (n=62) of the female respondents. The education level of the respondents was quite high with 62% of the respondents either have a Diploma or Bachelor degree or above and 37% of the respondents having completed secondary school. 74% of respondents had not participated in any form of microwork while 26% had participated in some form of microwork.

Figure 1 below illustrates the types of microwork that respondents indicated that they would most likely (want to) partake in (single response item). The majority of respondents (20%) indicated that would most likely complete surveys if they were to participate in microwork. The second most popular task amongst the respondents was rating services (14%, n=31) followed by mystery shopping (11%, n=33).

Microwork_SA_Fig_1

Figure 1: Tasks Respondents are “most likely to complete

Potential Contribution of Microwork

When we asked what they think the major contribution of microwork could be to a nation, 69 of respondents think that microwork has the potential to decrease unemployment while 35 indicated that microwork has the potential to help a country develop. Interestingly, 11 of respondents are of the opinion that participation in microwork is a new form of exploiting cheap labour.

Microwork_SA_Fig_2

Figure 2: Potential Contribution of Microwork

Key motivators for participating in Mobile Microwork

We also asked them about a number of potential motivating factor as well as barriers. Figure 3 shows to what extent they rated particular motivators/barriers as being important.

Microwork_SA_Fig_3

Figure 3: Motivators for (intended) participation into mobile microwork.

However, more interesting was when we ran a multiple linear regression test to see which of these actually were a statistical predictor for their intended future participation in mobile microwork. It turns out that only payment/remuneration is significant in the model (p = 0.006)> However, the overall multiple regression model has an R² of only 0.13 with an adjusted R² of a measly 0.03 i.e. we could only ‘explain’ 3% of the variance in intended future microwork participation from the usually cited motivators and inhibitors !

More interpretations and full details in the forthcoming publication of the AMCIS paper. Please feel free to comment or contact me directly.

Jean-Paul Van Belle; Centre for IT and National Development in Africa (CITANDA), University of Cape Town (UCT).

 

Digital Economy, Digital Labour

The Role of Digital Jobs in Solving Youth Unemployment in Kenya

iHub Research in 2014 published a report on Digital Jobs in Kenya. Fast forward to 2018 how far are we? A key insight presented in the report was that there existed a digital skills gap between theoretical skills, attained by youth through various programs, and practical skills, sought after by employers despite the existence of the key trends of: online work, big data analytics, and the mobile applications sector which present great potential for large-scale digital job creation in the future.

February 2018, unemployment is hitting record highs of 39.1% in Kenya, based on a report by the United Nations Human Development Index (HDI) 2017 and in its midst there’s been concerted efforts towards encouraging entrepreneurship, it’s clear that we need to innovate around solving the problem of unemployment.

In December 2016 the government of Kenya launched the online jobs portal, Ajira in a bid to take advantage of ICTs in eradicating unemployment; specifically targeting the youth, with the promise that it would equip 1 million Kenyans with digital skills so that they can secure employment. Ajira’s tag line which states “Online WORK is WORK” aims to raise the profile of online work, promote a mentorship and collaborative learning approach to finding online work, provide Kenyans with access to online work and finally to promote Kenya as a destination for online work.

Since its launch we’ve barely heard of progress, statistics or testimonies of the users of the platform. We do know that together with Kenya Private Sector Alliance, the Ministry of ICT as a result of funding from the Rockefeller Foundation are implementing the first phase of the Ajira mentorship program, to train and mentor future young online workers. Will this be the reason Ajira and the concept of digital work will successfully scale in Kenya?

The concept of digital work is definitely not one that is new in Kenya and definitely not across the world, from where we can learn great lessons. Already existing in Kenya is the platform KuHustle that has 32,000 plus online workers, with over 1,000 jobs posted worth over US$920,000.

In order to adequately take advantage of this opportunity it is paramount to address the barriers affecting the job market as a whole in Kenya and creating mechanisms to overcome these barriers. Based on estimates from the government before the launch of Ajira in 2016, it was assumed that there were already 40,000 Kenyans working online and as adoption of technology and the Internet is gradually increasing in the country, this number has most definitely increased and has the potential to continue to do so, with time.

Digital job platforms serve the purpose of easing the process of connecting employers to a competitive selection of employees from different locations, background and privilege so long as they have an internet connection. Meanwhile it seems in this central narrative that technology in the form of digital jobs will be the salve of solving youth unemployment. How true is this assessment?

There is no doubt that digital jobs will definitely enable and increase the possibility of a greater percentage of the young population in Kenya (who are possibly marginalised due to issues surrounding lack of access) to acquire formal employment at higher wages than they would have previously probably acquired. Beyond the basic digital skill gap that is being addressed through training and mentorship by government initiatives, there exists huge demand for specialised skills, such as developers, data scientists, which the current supply levels fail to meet and this is an example of some of the fundamental barriers affecting the job market in Kenya, that need to be addressed while also focusing on digital jobs.

By utilising this multi-stakeholder and multi-dimensional perspective in analysing the challenges currently being faced in the job market in Kenya today, this approach will propel us closer to solving the issue of youth unemployment.

 

Digital Labour

Be Mindful of the Short-Circuiting of Platform Work Narratives

The fight for fair and decent digital work involves multiple fronts. These include advocacy for accountable applications of technology, an inclusive policy-making agenda, and the construction of alternative narratives to those favored by companies or neoliberal governments or both, to list a few. The fight for fair and decent digital work is also deeply affected by local and national contextual factors such as the existing labor market and local income level. All seem to be uphill battles calling for joint efforts from concerned scholars, activists, and workers.

The latest globally uneven proliferations of platform-dependent jobs (e.g., gig driving and micro-work) raise serious challenges in terms of how to localize the multi-front battle for fair and decent digital work. For developing countries in particular, which often concern themselves with integrating into the global (digital) economy and development agenda, the discursive front of the fight for fair and decent digital work—that is, mainstreaming the worker-centered narratives about platform work—turns out to be tougher than other fronts.

Scholars like to use the trope of the “black box” (Pasquale, 2015) to describe the secretive and opaque way in which private IT companies utilize proprietary algorithms to manipulate consumers, users, and workers. Though scholars may perform reverse engineering to gather information on the inner workings of certain algorithms, systematic knowledge is hard to acquire. For example, the secrets of Uber’s surge pricing algorithm were revealed in 2015. However, the surge pricing algorithm does not function in isolation in shaping drivers’ working experience with the ride-hailing apps, while Uber may change how the algorithm works at its will. Thanks to the work of many scholars (see, for example, Rosenblat & Stark, 2016), the informational asymmetry between companies on the one hand and consumers, users, and workers on the other has been recognized as a structural feature of the power dynamics among workers, the platform company, and the regulatory authority.

Besides opaque algorithms operating in the dark, I have become aware of the back-channel circulation of narratives about digital platforms and jobs (workers) directly from private companies to governmental official reports without engaging with third-party institutions or scholars’ voices, not to mention the inclusion of workers’ perspectives. I call this phenomenon the “short-circuiting of platform (work) narratives.”

Platform companies have the tendency to brand themselves as pioneers in offering technological solutions to social problems, and the leading Chinese ride-hailing company Didi Chuxing is no exception. It builds its platform work narratives around creating jobs and increasing income. Didi claims that the company created 17 million flexible jobs in 2016 and more than 21 million drivers earned their income on the platform from June 2016 to June 2017 (Didi, 2016; Didi Institute of Policy Study, 2017). Taxi drivers do not belong to the category of “new jobs,” despite the fact that a great majority of taxi drivers in China also work on the Didi platform (Chen, 2017a). The two reports highlight the company’s contributions to jobs-creation particularly in regions and sectors where China’s national development strategy for economic restructuring has brought about staggering figures of unemployment. These regions include the heavy-industrial sectors of the rust-belt provinces in northeastern China Take the 2016 report as an example; other than the vague number of 17 million and 5 million job opportunities for laid-off factory workers, the report revealed few details on specific work conditions or if the income earned through the platform accounted for living wages. The China Internet Network Information Center (CNNIC)—China’s authoritative agency on statistics about Internet use—cited Didi’s jobs-creation number (17 million) word for word in its annual report on China’s state of platform economy (CNNIC & Internet Society of China Sharing Economy Committee, 2017). In the absence of other figures or reports on drivers’ work conditions, CNNIC echoed the rhetoric of jobs creation and did not delve into the challenges facing drivers in the ride-hailing platforms.

China is hardly alone in the short-circuiting of platform narratives. This process is dangerous because it allows the language and figures of jobs creation, and more importantly, the framing of the platform’s contributing role in the economy, to travel in a narrative package from the private company to the governmental report. But the governmental report also serves as the reference for scholars and policy-makers. Without incorporating alternative perspectives and narratives on the relationship between platform technologies and workers and development, the governmental report may be at risk of losing its public accountability, which further prevents counter-narratives from surfacing into the mainstream conversations on how accomplish the goal of fair and decent digital work.

Works Cited

Chen, J. Y. (2017a). Technologies of control, communication, and calculation: taxi driver’s labor in the platform economy. In P. Moore, M. Upchurch, & X. Whittaker (Eds.), Humans and machines at work: monitoring, surveillance and automation in contemporary capitalism (pp. 231–255). London: Palgrave Macmillan.

CNNIC & Internet Society of China Sharing Economy Committee. (2017). Report on the Development of the Sharing Economy in China. Beijing.

Didi. (2016). Job Creation – VALUES. Retrieved September 5, 2017, from http://www.didichuxing.com/en/values/job-creation

Didi Institute of Policy Study. (2017). New Economy, New Jobs: 2017 Research Report on jobs on Didi Chuxing. Beijing, China: Didi Institute of Policy Study.

Pasquale, F. A. (2015). The Black Box Society: The Secret Algorithms That Control Money and Information. Cambridge: Harvard University Press.

Rosenblat, A., & Stark, L. (2016). Algorithmic Labor and Information Asymmetries: A Case Study of Uber’s Drivers. International Journal of Communication, 10, 3758–3784.