1.1 Digital labour platforms and their business models

Digital labour platforms are the fastest-growing business in China’s sharing economy. With the scale of financing, the major platforms, including transportation and household services, have been growing rapidly. The market share of platforms keeps expanding while changing the lifestyle of consumers. For example, from 2016 to 2018, the share of ride-hailing platforms in all taxi passenger transportation rose from 9.5 per cent to 36.3 per cent, and the proportion of online food delivery platforms in the catering industry went up from 1.4 per cent to 10.6 per cent. In terms of the number of people employed, digital labour platforms have become an important driver for job growth. In 2016–17 alone, nearly 21.1 million people earned income on Didi (Didi Research 2017). More than 2.7 million workers on the Meituan food-delivery platform in 2018, representing an increase of nearly 500,000 people compared with 2017 (MRI 2019a). Fengniao, another delivery platform, had more than 3 million registered delivery workers in 2018.7

Based on a review of existing literature, China’s digital labour platforms operate in three main business models (table 1). The first is self-operation aided by platform (平台自营): Companies build their own internet platforms to carry out their business and engage customers directly (the so-called B2C). In this model, the platform usually establishes a labour relationship with their workers and controls the recruitment, service standards and working conditions. Some platforms also use a labour dispatch company, through which by law the labour dispatch firm should sign labour contracts with workers. Shouqi Limousine and Chauffeur and Shenzhou Ucar are two examples of such a business model.

Shouqi Limousine and Chauffeur is a traditional taxi company loaded with online-to-offline (known as O2O) ride-hailing services. The relationship between the platform and drivers follow the standard labour relationship, and labour contracts are signed between the company and the workers. Initially, Shenzhou Ucar leased out all its vehicles to its drivers while a labour dispatch firm was responsible for the recruitment and management of drivers. The management duty included overseeing the signing of labour contracts and monitoring performances. Drivers received a salary on a monthly basis and enjoyed participation in the five mandatory social insurance schemes8 and the housing provident fund. However, when other ride-hailing platforms became fierce competition, the two companies changed their business model; box 1 describes the changes that Shouqi Limousine and Chauffeur made. Similar changes occurred to Shenzhou Ucar.

Box1

Note: A labour service contract (劳务合同) refers to all contracts excluding a labour contract and is governed by the Civil Code and the Contract Law. The main costs for labour service contracts are service fees, and termination of the contract is left to the parties’ free will and agreement.

Source: See www.bbtnews.com.cn/2019/0408/293838.shtml.

The second model is called “franchised cooperation through platform (加盟合作)” whereby platforms attract a large number of third-party micro and small enterprises, thus creating an agglomeration effect on the platform (knowns as B2B2C). Many platforms of household service have adopted the model of franchised cooperation with third-party enterprises and operate in an asset-light way to quickly expand their market. The platforms operating in this model serve as information intermediaries and not as employers (work units who meet the legal qualifications under the Labour Contract Law).9 When the franchisees employ workers, they are the employers of these workers.

Yunjiazheng, a platform of domestic home services created in 2012, is an example of such a model. The platform collaborated with the Shanghai Domestic Work Association and used its domestic work database to develop a domestic workers’ management system. Many offline domestic work companies use the management system for obtaining information. The platform is only responsible for identity authentication of domestic workers and provides fee-charging information to domestic work companies. It does not hire any domestic workers nor own offline companies. In this model, the platform is similar to that of an information intermediary. There is no established labour relationship between the platform and the domestic workers, and the franchisees are usually the employers of the workers.

The third model is a sharing model via crowdsourcing (新型共享). This model uses big data, cloud computing and other internet information technologies to integrate and mobilize labour, skills, production materials and other resources to meet the diversified needs of businesses, consumers and the society at large. Unlike the traditional “firm plus employee” model, this one uses the “platform plus individual” (known as C2C). In this model, the platform uses its intermediary function but also regulates the services by setting up recommendations and rules that can be diverse in nature, ranging from security recommendations to the fixing of fares.

For example, crowdsourced fast-delivery workers only need to go through membership registration, name authentication and simple online training and assessment to become a part-time delivery worker with a platform. Delivery workers usually need to have their own vehicle (most of them use of a motorcycle). Drivers working for the DiDi platform typically own or rent their vehicle. Workers join with their own assets and tools, and they can join and withdraw relatively easily. Usually they work with a somewhat high level of autonomy, but they must follow the rules set by the platforms to a large extent. Defining the labour status of workers or service providers operating under such platforms remains a challenge.

Table 1 Main business models of digital labour platforms

China’s digital labour platforms do not stick to one model. They flexibly adjust their business format according to market demands. In particular, the ride-hailing and food-delivery platforms with large market shares tend to adopt a mixed approach that integrates the self-operation, franchised cooperation and sharing models (box 2).

Box 2

Note: *=DCCI 2019.

Source: CALSS 2017.

Location-based platforms

Location-based platforms have become China’s fastest-growing digital labour platforms in recent years. The platforms first appeared in the transportation sector in 2012 and expanded to food delivery, household services and other areas. Through mobile telephone applications (apps), the platforms can realize real-time matching of the demand and supply. The platforms either enable workers or service providers to compete for orders or directly allocate orders to them. The platforms usually set the standards for services and intervene, to a certain degree, in the selection and management of the service providers. Well-known location-based platforms in China include Didi for transportation, Meituan and Eleme for food delivery and 58.com for household services.

Didi Chuxing

Established in June 2012, Didi Chuxing is the largest ride-hailing service platform in the country. Through the acquisition of Kuaidi and Uber China, Didi now occupies more than 90 per cent of the Chinese online car-hailing market. Didi provides users with a variety of online ride-hailing services, including express, premier, special and luxe riding and carpooling. DiDi provided more than 10 billion rides in 2018, with a total mileage of 48.8 billion km (RCSE 2019, 39). By the end of June 2018, Didi had attracted more than 30.6 million drivers and operated in more than 400 cities.

The DiDi app is user-friendly: Passengers place an order, and the platform matches them with drivers. This matching process has undergone a transition from competing for orders to dispatching orders. In the former model, the platform sent orders to drivers in a nearby range who competed with one another to be the first to reply. Now it uses an algorithm to complete the matching by directly assigning an order to a nearby driver. The driver, after receiving the order, goes to the designated place to pick up the passenger and drives the passenger to the destination according to the predetermined route. Once the trip is completed, the platform receives the fee paid electronically by the passenger and deducts a service fee (commonly known as commission). The driver can withdraw the remaining portion from the platform as remuneration. The proportion of the service fee deducted by the platform varies because the fee structure is based on various factors, including the city where the service takes place, distance, length of time used and the availability of carpooling. On average in the fourth quarter of 2018, commissions deducted by DiDi were about 20 per cent of the fee paid by passengers.10

During the service, drivers are required to abide by the code of conduct developed by the platform and accept the supervision of the platform and passengers. Although the platform is not physically present, it can track the driver’s route and behaviour (such as sudden brakes, rapid deceleration and speed) through the app. The platform uses an algorithm to assess the driver’s service after extracting data from four dimensions: the reception of the order, the pick-up, the drop-off and passengers’ post-trip evaluation. The driver can be rewarded or penalized accordingly. Depending on the result of the assessment, rewards and penalties can be reflected in cash, priority of order dispatch and a ranking determined by the algorithm (Chen 2020; Wang, Wu and Yang 2018).

Meituan Food Delivery

Meituan Food Delivery was officially launched in 2013. In 2018, Meituan accounted for 62 per cent of the domestic market share, with more than 24 million orders completed daily, and a total of 300 million people ordering meals from the platform (RCSE 2019). The Meituan platform is user-friendly: After receiving an order, the platform allocates it to a delivery worker called a “rider”. The rider delivers the food to the customer according to the set time and delivery route determined by the algorithm. The total number of active riders on the platform (delivering at least one order per day) is around 670,000, among which there are two main types of riders: full-time and crowdsourcing.

Full-time riders are usually attached to franchisees or agencies that must be officially registered as a business entity. About 1,000 franchisees or agencies have signed civil cooperation agreements with Meituan and have also signed labour service contracts with 250,000–270,000 riders. These riders are subject to the rules set by the franchisees or agencies and receive orders dispatched to them through different management units called “sites” that the franchisees or agencies have established. There is a link of subordination between the franchisees or agencies and the riders; full-time riders should, according to law, sign a labour contract with them but few have actually signed such a contract.

Crowdsourcing riders are usually attached to more than a dozen labour service companies throughout the country via a labour service contract that treats them as self-employed. They compete for orders online and are not restricted to a specific site. These riders often provide added capacity during mealtimes and peak-hour delivery.

Sharing model via crowdsourcing in manufacturing

There is a trend of location-based platforms penetrating into traditional enterprises, such as manufacturing. Traditional enterprises have developed two main approaches to use online crowdsourcing workers – one is internal and the other is external.

The internal approach entails the establishment of a platform within an enterprise as an agent to transform production into “ecosystems” between partners and customers. With this approach, enterprises have often turned their workers into “entrepreneurs” (self-employed service providers and independent contractors) and terminated the labour contract with them. For example, Haier, a Chinese home appliance manufacturing giant, adopted the platform business model and reduced its number of employees from 86,000 in 2012 to 60,000 in 2014 (Peng and Yun 2015, 126–128).

The external approach is to cooperate with platforms through labour services. Enterprises outsource their businesses to selected platforms via a civil cooperation agreement. In these cases, the enterprise pays the platforms an outsourcing cost against a special value-added tax (VAT) invoice issued by the platform to the enterprise. In the next steps, a platform recruits their workforce, guides them to register as an individual business, or “maker” (the Chinese term for self-employed service provider or independent contractor), through the platform app and then crowdsources the contracted business from an enterprise with the registered individual businesses (makers). Upon completion of the assigned tasks, the enterprise pays the due commissions to the platform, and the platform pays the registered individual businesses (makers) and assists them to issue a general VAT invoice as proof of tax paid.

The two approaches have contributed towards reducing the number of workers with labour relations while increasing the number of service contracts. The difference for the individual involved is that they are no longer protected by the labour laws. This raises concerns for these individuals who, according to Chinese law, would qualify to sign a labour contract but whose labour relationship is disguised into a service contract. This is particularly the case when workers work for one enterprise only and are paid, receive orders and follow the rules and instructions of that enterprise.

Web-based platforms

Web-based platforms provide convenience for crowdsourcing work by quickly matching the demand for technology, management support, design, multimedia and other professional services with the supply. Although web-based platforms for low-skill micro tasks exist worldwide, in China, the service providers on web-based platforms usually have a certain level of professional skill and could come from anywhere in the world because the crowdsourcing has no geographic limitation. China’s web-based platforms emerged in the early 2000s and have grown rapidly since 2016. Well-known crowdsourcing platforms, or “witkey” in the Chinese phrasing, include Zhubajie, PWK, sandaha.com and 680.com (witkey China). As of 2012,11 more than 1 million visits to these platforms occur daily, with a total number of tasks issued is more than 1.6 million (SRI 2012). The market for web-based platforms is smaller than for location-based platforms in 2017, with about 24 million service providers registered on the web-based platforms (RCSE 2018, 48).

Zhubajie

Zhubajie is China’s largest crowdsourcing service platform. Founded in 2006, Zhubajie provides outsourcing tasks, such as logo design, website construction, intellectual property management, finance and taxation consultation. It is the largest online labour market of crowdsourcing. By 2017, the number of registered users exceeded 16 million, with the number of contractors at 6 million and the number of service providers at 13 million (Zhang et al. 2019).

Zhubajie offers three service modes for its providers: competition, piece-rate and matching. In the early stages of development, the competition mode was the main service format: A customer offers a reward to all service providers in the form of competition, and the one who offers the best solution receives the bonus. Customers and service providers are in a one-to-many relationship with no strict contractual constraints. Zhubajie receives a 20 per cent of the transaction cost as a platform service fee.

The piece-rate mode is used when the workload is small, the task is simple, the price is low and there are many respondents who can provide a large number of manuscripts in a short time. Zhubajie receives a 5–20 per cent commission in this mode.

The matching mode can effectively protect the privacy of customers. It consists of a direct matching between demand and offer through an algorithm. The benefit of this mode is the elimination of the complicated steps of repeated selection. It is suitable for website construction, architectural design, decoration design, software development and other projects. Zhubajie receives a 15–30 per cent commission in this mode.

Managed crowd

Most of the traditional web-based platforms outsource work to an undefined crowd, and the platforms do not manage or train crowdsourced service providers. However, in 2017, a new model emerged in China called the “managed crowd”. While this type of platform has many similarities with the other traditional crowdsourcing platforms, the main difference is that the platform releasing the task provides specialized technical training to interested service providers, thereby helping them to improve their ability to complete related tasks. Based on an evaluation of the training results, the platform disqualifies 25 per cent of the service providers who apply. Eventually, the platform keeps the individuals with the best results, assign them tasks and thus applies a certain level of quality control over the tasks completed. This new platform model is used in the transcribing of audio files and software testing, among others. Pactera and Tonelink are the Chinese platforms engaged in such a model. The latter was established in 2015 and currently employs 1,000 workers, with 3,000 registered suppliers on the platform and a level of transactions reaching 1.6 billion yuan in value (Oshri 2017).

1.2 Characteristics of the Chinese digital labour platforms

The emergence of the digital labour platforms has and will continue to have a broad and comprehensive impact on the world of work. Although more research needs to be done to fully understand the nature and scope of the impact, ILO and Eurofound research determined a direct impact on work tasks and the occupational skills needed as well as on working conditions and employment status (de Groen et al. 2018, 7). The following sections analyse these working conditions in relation to the three major platforms in China.

Large number of full-time workers in a situation of economic dependence

Three aspects characterize platform workers in China: their large number in total, their full-time working schedule and their economic dependence on the platform. First, with the rapid development of the sharing economy in China, the number of individuals providing services in one way or another through platforms continues to increase, from 50 million in 2015 to 75 million in 2018; its proportion in total employment also has increased, from 6.5 per cent in 2015 to 9.7 per cent in 2018. To give this data a comparative perspective, platform workers in the United Kingdom account for only about 4 per cent of the employed population, and in the United States, they are less than 1 per cent (table 2). And according to the platforms’ websites, in the ride-hailing industry, DiDi has more than 31 million registered drivers,12 while Uber had only about 3.9 million drivers worldwide as of December 2018.13 In the food-delivery industry, there are more than 2.7 million riders registered on Meituan, of which about 670,000 are active (delivering at least one order a day). Deliveroo, the largest food delivery company in the United Kingdom, has about 60,000 registered riders worldwide.14 China clearly has the largest number of platform workers in absolute and relative terms.

Table 2 Comparison of platform employment, by selected countries

Source: Statistics of European countries are from Eurofound 2018b, 13; the Statistics of the United States are from Krueger and Katz 2016; and China’s statistics are from RCSE 2019 and MOHRSS 2019.

Second, an increasing number of platform drivers and food-delivery riders have become full-time workers. At the beginning of its development, the sharing economy had attracted many part-time workers and groups of people with difficulty in finding jobs15 in urban areas, offering them a “reservoir” of employment. As their market shares expanded, more people joined the platforms on a full-time basis, attracted by the relatively high level of remuneration. According to a 2019 Didi Research report, more than 20 per cent of ride-hailing drivers regarded platform work as a long-term occupation and more than 30 per cent of the drivers from urban poor background16 held the same view. A whopping 74.5 per cent of the food delivery riders worked full-time, among which 42.9 per cent had worked for more than a year (CALSS 2019).

Third, a large number of platform workers is economically dependent upon the remuneration they get from the platform. This is illustrated by the share of income originating from the platform in the total household income. In China, about 16 per cent of all food-delivery riders in 2018 earned more than 90 per cent of their household income from a platform. As for the online ride-hailing drivers, 22.3 per cent of them earned, on average, more than half of their household income. The web-based crowdsourcing platforms were the only source of income for 43 per cent of the workers (CALSS 2019, 22). In contrast, for example, only about 9 per cent of platform workers in the United Kingdom earned more than 90 per cent of the family income in 2018 (table 3).

Table 3 Economic dependency of Chinese platform workers (compared with the United Kingdom), 2018

Source: The statistics of Meituan are from MRI 2019a; the DiDi data are from DiDi Research 2019; and the United Kingdom statistics come from DBEIS 2018, 30.

Platforms attract large numbers of young male workers

While the mobile phone use rate for female and male individuals was 48 per cent and 51.6 per cent in 2018, respectively (UNFPA 2019), men seem to be more attracted to the digital labour platforms. The proportion of male workers on the location-based platform is particularly high, at more than 90 per cent in 2019. This may be related to the traditional gendered occupational nature of being a driver or a rider. Women’s participation in the web-based platforms was about 27 per cent in 2019, still lower than what it was for men. This trend is in line with the participation rate in paid work (table 4).

Table 4 Participation rate in paid work, 2018 (%)

Source: UNFPA 2019.

Regarding age composition, relatively young people dominate digital labour platforms. Most DiDi drivers were born in the 1970s or 1980s, accounting for 71.4 per cent of the total DiDi drivers in 2019. Workers on Meituan are a bit younger, with 82 per cent born in the 1980s or 1990s. This might be explained by the fact that the post-1990 generation may not have as large a proportion of car ownership as the post-1970s and post 1980s generations. In comparison, workers with Zhubajie are relatively older, with 61.9 per cent of them born in the 1970s. Interestingly, the percentage of single persons with Zhubajie is much larger than with the other two platforms, with 56.7 per cent of workers currently single.

There are significant differences between the location-based and web-based platforms in terms of education. Location-based platform workers have a comparatively lower level of education; most of them are graduates of high school or have less education, while college graduates only account for a small proportion. In particular, a large proportion of Meituan riders in 2018 only finished junior high school, which is the compulsory level of education in China. The level of education of workers on the web-based platforms was significantly higher in 2019, with 88.1 per cent having completed college or a higher level of education.

The household registration (hukou) divides the rural and urban population, thereby hindering labour mobility and segregating the labour force.17 Rural migrant workers comprised 58 per cent of the Meituan riders in 2018 (table 5), while Didi’s workforce was dominated by local drivers with urban residence status. This is a direct consequence of China’s recent regulation on the online ride-hailing sector, which encourages local governments to require a local residency status for taxi drivers. Zhubajie has a diverse labour force composition, with a relatively even geographical distribution. This indicates that web-based crowd work has eliminated, to some extent, the discrimination caused by household registration and has broken down geographical boundaries.

Table 5 Demographic characteristics of major platforms in China

Source: The statistics of Zhubajie come from ILO forthcoming(b); those of DiDi are from DiDi Research 2019; and the data for Meituan come from MRI 2019a.

Increased income as the main motivation

Across the three platforms, the main motivation for platform workers to engage in their work is to increase their overall income. Around 40.4 per cent of crowdsourced workers surveyed in 2018 said the platform work is a way to supplement their income from other work, and 63 per cent of the food-delivery riders believe that platform work can help them increase their income. The survey results also found that workers do not choose to work on platforms for their personal interests. Only 24 per cent of riders chose the job out of personal interest. Web-based crowdsourced workers were significantly different from the food-delivery riders, with 52.2 per cent of them choosing personal interest as the reason to do the job (MRI 2019a).

According to the survey data, 20.37 per cent of online ride-hailing workers were re-employed after being laid off from a non-platform job, and 10.5 per cent of food-delivery couriers and 2.2 per cent of crowdsourced workers chose their platform job because they could not find any other job. This seems to indicate that digital labour platforms, especially location-based platforms, have a limited role in finding jobs for disadvantaged people, such as laid-off workers, long-term unemployed persons and veterans.

In short, the reasons for crowdsourced workers to engage in platform work are more diversified, including earning extra income, personal interests and flexibility of work. In contrast, a larger proportion of ride-hailing drivers and food-delivery riders regard the platform work as a long-term occupation.

Table 6 Motivation for platform work

Source: The statistics of Zhubajie come from ILO forthcoming(b); those for DiDi are from Didi Research 2019; and the data for Meituan come from MRI 2019a and ILO forthcoming(a).

Contractual arrangements increase informality in employment

The type of contractual arrangements associated with platform work is rather diversified. Yet, there is a trend for platforms to favour business models that can circumvent labour relations. While other forms of labour contracts, such as labour dispatch and casual contracts are commonly used, the majority of the contractual arrangements take the form of labour service contracts that have different names (such as franchising, agent or labour-outsourcing contracts). As indicated earlier, only 8 per cent of platform workers had labour relations with the platform in 2019. In other words, more than 90 per cent of the platform workers in China in the same year either had signed various forms of labour service contracts falling outside the scope of the labour laws or had not signed any contract at all. Despite the various names the labour service contracts may carry, a majority of workers with the digital labour platforms fall into the following three categories.

1. Workers sign a labour contract establishing a labour relationship with the platform. This concerns mainly those who are part of the core internal team of the platform, such as the managerial or technical staff. These labour contracts are generally full-time with a fixed term. This means that only these workers are granted protection under the labour laws regarding minimum wage, working hours, occupational safety and health, union representation and social security.18

2. Workers sign a labour service contract,19 under which there are three subcategories:

• Workers sign a labour service contract with the client or customer. In this case, the platform is just an intermediary.

• Workers sign a labour service contract with an agency and are dispatched to work for a platform. While the Labour Contract Law requires the agency to sign a labour contract with the dispatch workers, in reality, few do.

• Workers sign a labour service contract with the platform. However, if the characteristics of the work being performed and the way it is performed meet the conditions of a labour relationship under the labour law, then the workers are considered to be in a disguised labour relationship. Of the features, a labour relationship under the labour laws is the concept of subordination.

3. Workers are genuinely self-employed without any contract or with a contract for service that is regulated by the Contract Law.

In reality, there are many instances in which the law lacks clarity in its definition of labour relations, making it difficult to determine the workers’ status as self-employed or with labour relations.

According to scholarly analysis of various survey findings, the workers of the major digital labour platforms in Beijing have signed labour service contracts with their platform (such as cooperation agreements and intermediation agreements). Few platforms have signed labour contracts with their workers (Zhang 2018). For example, DiDi only signs labour contracts with its personnel responsible for backend technology and management, it uses labour dispatching for customer service staff, and for the online ride-hailing drivers, it offers labour service contracts. A clear provision is included in the labour service contracts stipulating that “DiDi only has an affiliated cooperation relationship with all drivers who provide online ride-hailing services, and there are no direct or indirect labour relations. As such, the Labour Law, the Labour Contract Law, the Social Insurance Law, the Regulations on the Management of Housing Provident Fund are not applicable.” Similarly, the 3 million registered food delivery riders with Fengniao Delivery and the more than 400,000 crowd-sourced riders of Meituan Food Delivery are also operating under labour service contracts.

The ILO considers most of work on digital labour platforms as a form of non-standard employment.20 The jobs are casual, with work only available when there is a specific task needed by a client (ILO 2018a, 10). And the work in many instances is informal. According to the ILO (2003), workers are considered to have informal jobs if their employment relationship is, in law or in practice, not subject to national labour legislation, income taxation, social protection or entitlement to employment benefits.

While the Chinese experience shows that internet business models have created and continue to create millions of employment opportunities, there are also clear indications that the size of informal employment is increasing (CEC and CNE 2018, 7). This can be attributed to a number of factors, including the following.

First, location-based platforms have attracted traditional manufacturing workers: 22.13 per cent of online ride-hailing drivers and 23 per cent of full-time food-delivery riders in 2018 used to work in the manufacturing sector (Didi 2018; Meituan 2018). This trend may continue, given how the manufacturing industry is going through a transformation period characterized by industrial upgrading and influenced by such external factors as slower economic growth and COVID-19. Platforms’ low requirement for skills and job access, as well as their comparatively good income are likely to attract more manufacturing workers. This move could reduce the number of formal manufacturing jobs while increasing the labour supply for platforms, with a risk of reducing the incentive for platforms to formalize their jobs.

Second, the majority of platforms have opted for flexible business models that rely on flexible forms of contractual arrangements, such as franchised cooperation and crowdsourcing, which leave workers without a labour contract and thus in a certain level of informality at work.

Third, through the platformization of traditional enterprises, workers’ status has shifted to entrepreneurs or self-employed, and a number of workers’ labour contracts have been terminated. This also contributes to the increased informality in employment in China.

Table 7 Sources of workers for DiDi and Meituan (full-time), 2018

Source: The statistics of DiDi are from DiDi Research 2018; the data of Meituan come from MRI 2019a and ILO forthcoming(a).

A significant feature of web-based platforms is that professional and technical personnel from enterprises and public institutions account for the largest proportion of the workers, up to 23.6 per cent in 2019 (table 8). In addition, the proportion of students is also relatively large, accounting for 15.8 per cent in 2019. This seems to indicate that web-based platforms provide more part-time opportunities for people with a higher level of education and skills.

Table 8 Main occupation of workers with Zhubajie

Source: ILO forthcoming(b).

1.3 Working conditions on digital labour platforms

Earnings

Relatively high level of income

Overall, the income of workers on China's digital labour platforms is relatively high. According to a survey conducted by the All-China Federation of Trade Unions (ACFTU) in 2017, the wages of platform workers were higher than the national average for workers of the same age (Li 2017). A particularly noteworthy phenomenon is that platform workers generally earned more than front-line manufacturing workers, and the wages of the platform workers in some provinces was even 30 per cent higher than those in traditional manufacturing industries.

The average monthly income of the drivers with DiDi platform in 2018 was 2,522 yuan. The average monthly income in the first-tier cities was 5,176 yuan, 2,133 yuan in the second-tier cities and 1,130 yuan and 1,033 yuan in the third-tier and the fourth-tier cities, respectively. This indicates a correlation between the level of income for DiDi drivers and the population size of the city where they work and the related cost of living. When comparing the highest average income of DiDi drivers in the top-ten income cities with the local minimum wage and the average income in the local urban private enterprises in the transportation industry, it is clear that the average income of DiDi drivers is generally higher than the local minimum wage and that some drivers earn even more than the local private enterprise employees (figure 1).

Figure 1 Income of online ride-hailing drivers, 2018

Note: The unit is yuan.

Source: DiDi Research 2018; NBS 2018.

The monthly income of the food-delivery riders with Meituan mainly concentrates in the range of 2,000 yuan to 5,000 yuan, accounting for 51 per cent of all riders in 2018, followed by 28 per cent of riders who earned between 5,000 yuan and 10,000 yuan (figure 2). Riders who earned less than 2,000 yuan accounted for 20 per cent, while 1 per cent can even made more than 10,000 yuan per month.

Figure 2 Comparison of income of Meituan riders and DiDi drivers, 2018

Note: The unit is yuan.

Source: MRI 2018; DiDi Research 2018.

In contrast, web-based crowdsourced workers have relatively lower income. Their average hourly income in 2019 was 32.6 yuan (equivalent of US$4.6), while the average monthly income of workers at Zhubajie was 1,777 yuan (US$253.8). Workers at Zhubajie earned the highest income, compared with other crowdwork platforms. There is a clear difference between workers who regard platform work as their major source of income and those who do not. The first group earned an average of 599.42 yuan (US$85.63) per week in 2019, while the second earned only 309.91 yuan (US$44.27) on average.

Income instability as a major concern

There are two underlining trends behind the relatively high income of platform workers. First, the level of income often comes at the cost of long hours of work (see section 3.2). Second, most digital platform workers follow the salary structure of a piece rate system without a base salary. The income level is susceptible to such factors as platform trading rules, seasonal demand fluctuation and commissions deducted by the platform, all of which are to encourage workers to work more. While this leads to relatively higher income on average, it also generates instability and unpredictability in income. According to the ACFTU survey, the instability of income was the most concerning issue for workers in the platform work, accounting for 56.95 per cent of surveyed workers in 2017 (Li 2017).

DiDi is a good illustration of these trends. Back in 2014, when DiDi and Uber fiercely competed for market shares worldwide, both subsidized passengers and drivers heavily to gain market shares. As a result, drivers’ income level rose sharply during this period of subsidy. Following the successful acquisition of Uber China in 2016, DiDi adjusted its business strategy and started to reduce the subsidies. As a result, the income distribution for DiDi drivers has changed from the deduction of commissions to different pricing. Under the commission deduction, the platform deducted about 20 per cent of the fee paid by passengers, and the pricing for passengers was not separated from the pricing at the drivers’ end. Starting in August 2016, DiDi began to implement a two-way pricing rule for drivers and passengers. It now charges 0.5 yuan for an information fee and 1.77 per cent management fee for each order, on top of the commission. Before 2016, the reward rate to DiDi drivers was 2.5 times during the normal time and 5 times during peak hours. This rewarding rate was lowered to 1.2 times and 2 times, respectively, after 2016. Passengers' subsidies also were reduced. For example, in Beijing, the price of DiDi express cars increased from 1.5 yuan per km and 0.25 yuan per minute to 1.8 yuan per km and 0.5 yuan per minute before and after 2016.21

The chart (figure 3) indicates the surge in drivers’ income at the early stage and its gradual lowering. A comparison of the income levels of ride-hailing drivers in 2016 and 2017 indicates that the overall income of ride-hailing drivers in 2017 was lower than that of 2016. The proportion of drivers with an income of less than 4,000 yuan increased (by 14.6 per cent), and the percentage of high-income groups decreased (by 12 per cent).

Figure 3 Changes in the income of online ride-hailing drivers, 2016 and 2017

Note: The currency unit is yuan.

Source: DiDi Research 2018 and 2017.

Similar income instability exists for food-delivery riders. Their income consists mainly of commissions from each order. The food delivery platforms calculate the pay based on each order’s distance, amount paid and time spent, and provide awards for large-amount and on-time deliveries, among other incentives. Some riders in Beijing claimed in a media report that the platform for which they have worked give a 3 yuan commission for each of the first 200 orders per month. The more orders they take, the higher commission they can get, with 6 yuan capping the commission after 500 deliveries.22 However, through the platform algorithms, this commission rate could change any time with market fluctuations. A 2017 survey by the Chinese Academy of Labour and Social Security on food-delivery riders found that 46.8 per cent and 45 per cent of the riders believed that, respectively, the frequency and amount of platform’s rewards were decreasing (CALSS 2017).

In addition, insufficient work is a common problem for web-based crowdsourcing workers, which leads to their income instability. Around 50 per cent of them spent less than ten paid hours a week for Zhubajie. When asked if they wanted to spend more time working on the platform, 83.1 per cent of workers said yes. They wanted to add an average of 10.9 hours per week to the current working hours (figure 4).

Figure 4 Working hours for online platform Zhubajie, 2019 (per week)

Source: ILO forthcoming(b).

Long and fragmented working hours

While their working hours are generally flexible, with online work and offline breaks relatively discretionary, platform workers perform long yet fragmented working hours without access to overtime payment. The statutory working time in China is eight hours a day and 40 hours a week. However, according to the study by Wu and Li (2018, 244–245), only 13 per cent of food-delivery riders worked fewer than eight hours a day, while 49.5 per cent worked more than ten hours, and 30 per cent worked more than ten hours each day in a week. There was a significant difference between full-time and part-time riders: Among the full-time riders, only 8.8 per cent worked fewer than eight hours per day, while the proportion of part-time riders reached 36.6 per cent. Full-time online ride-hailing drivers worked about 0.85 times longer than part-time drivers per week.

A similar situation applies to crowdsourcing platforms. For workers who relied on the platform work as the major source of income, the paid, unpaid and total working hours per week were 24.6, 11 and 35.6, respectively. The working hours of those who did not rely on platform work as their major source of income were almost halved, at 13.2, 7.5 and 20.7, respectively (ILO forthcoming b).

The working time for platform workers has followed a trend of fragmentation (figure 5). This is particularly true for food-delivery riders. Surveys show that the workload of riders is not evenly distributed. At mealtimes when orders flood in, riders must work intensively for hours. In non-mealtime periods, riders must compete for limited orders and the work is not sufficient. Riders often work eight hours a day but stay on call for 12 hours, constantly rotating between intensive driving and order searching (Xiaopeng, Chen and Ruan 2018). Online ride-hailing drivers and crowdsourcing workers face a similar situation. Drivers reported in a 2018 study that they only spent 46 per cent of their online time taking orders, which means more than half of their online working time was spent waiting for the orders by the roadside or running empty on the road. Of the 15.6 hours of paid work for crowdsourced workers, 8.3 hours of their time was unpaid (Zhang 2018, 61–70).

These findings indicate that the actual working hours of workers on the three major platforms have far exceeded their paid hours.

Figure 5 Fragmented working time for food-delivery riders, 2018

Source: WeChat official account, Metro Data Team (ID: metrodatateam), 11 September 2018,

www.huxiu.com/article/262184.html.

High risk of occupational injuries but inadequate social protection

Both the ride-hailing and food-delivery businesses are prone to traffic accidents, thus the risk of occupational injuries is high. For online ride-hailing, occupational injury risks faced by drivers are mainly personal injuries caused by traffic accidents, driving fatigue or by passengers (Ban 2017). According to 2017 data from Shenzhen city, the percentage of traffic accidents for online ride-hailing was nearly 7.2 per cent, compared with nearly 1.8 per cent for traditional taxis.23 A 2017 Chinese Academy of Labour and Social Security survey on food-delivery workers found that 70.1 per cent of them regarded unsafe working environment as a top concern, and 34.2 per cent indicated that they were under intense work and psychological pressure due to the stringent time limit for food delivery and the related penalty rules set by their platform. Another factor is the high demand for food delivery in extreme weather. According to the survey findings, 52.9 per cent of the food delivery workers had suffered a work injury while delivering orders.

Despite the high exposure to work injuries, the majority of platform workers are not covered by a work injury insurance scheme. Most platforms have purposely avoided establishing labour relations with their workers, thus disqualifying them to access the work injury insurance scheme. At present, only a few platforms, including DiDi, provide commercial accidental injury insurance for their workers through deductions from the charges of online orders, mostly with the whole insurance being financed by the workers.

The proportion of platform workers participating in other social insurance schemes (with self-bearing contributions) open to workers who have no labour relationship is also low. Only 37.4 per cent and 12.7 per cent of food-delivery riders in Beijing were participating in medical insurance and old-age insurance schemes, respectively in a 2018 study, while 46.3 per cent had no social insurance coverage at all. Those who participated in the insurance schemes were generally paying the contributions themselves (Zhang 2018). There were even smaller proportions of crowdsourced workers participating in a social insurance scheme, with only 29.7 per cent covered by medical insurance and 5.6 per cent having old-age insurance. And 41 per cent of them did not participate in any social insurance programme (ILO forthcoming b).

Absence of skills training and career development

A large number of platform workers are either unskilled or low-skilled, with poor potential for career development. Platforms generally lack incentive for workers to develop their skills and invest in training. The trainings offered by the platforms are usually very basic and focused on platform rules and the use of their apps. In the 2017 Chinese Academy of Labour and Social Security survey, 67 per cent of the food-delivery riders who had received training reported that the focus was on safety issues (88 per cent), simple customer service (78 per cent) and platform rules and regulations (73 per cent). There was no content related to skills improvement or career development. Food-delivery platforms generally cap the age of riders at 45 years. These workers are likely to encounter great difficulty in finding another job after their time in delivery work.

DiDi drivers experience a similar situation. According to the 2016 Survey on the Status of Online Ride-hailing Drivers, 39 per cent of the drivers did not receive any training at all, and 55.4 per cent only received simple online training with no monitoring of the training results in 2016 (SRRI 2016).

At Zhubajie, 20 per cent of the people operating with the platform were offered a training opportunity in 2019, but only 13.4 per cent participated (table 9). Although Zhubajie provides various business services, such as web design, tax management and management consulting, it may be difficult to offer trainings for diversified service providers. The training mainly focuses on how to use the platform; only 16.7 per cent of the content relates to skills upgrading.

Table 9 Training in Zhubajie, 2019 (%)

Source: ILO forthcoming(b).

Lack of workers’ voice

The platforms studied have full and complete control over matters related to business transactions, such as setting the price for each order, fixing the amount of commission deducted, establishing service standards and the reward and punishment system, despite the workers being classified as self-employed. Platforms can penalize their operators and workers for violating the platform rules based on clients’ evaluation and their own supervision and monitoring system via algorithms. Riders with a food-delivery platform report that if there is a complaint filed against them by a client, their account can be suspended for half a day or a day, during which time they are unable to take or compete for orders and earn income. The rider can appeal to the “site” to which they report, but often the appeal is rejected.24 The lack of voice for platform workers is also reflected in the low degree of organization. The ACFTU survey found that only 21 per cent of full-time platform workers had joined a union in 2017 (Li 2017).

The lack of channels for addressing grievances and safeguarding workers’ rights has led to protests against the platforms. According to the statistics of the Shanghai Platform for the Trade Union, there have been nine collective actions by Shanghai platform workers since 2018, thereby becoming one of the factors affecting harmonious labour relations in Shanghai.25 From January to October 2018, there were monthly strikes of food-delivery riders and ride-hailing drivers (Qiao 2019).

Summary

Like most digital labour platforms around the world, the major platforms in China have chosen to classify their workers as self-employed or persons providing labour services, thus establishing a labour service relationship. This means that these workers are not subject to the legal protections that are granted to workers in a labour relationship on pay, working hours, occupational safety and health, voice and representation and social security.

Platform workers in China numbered 78 million in 2019, representing 10.1 per cent of total employment. Many of them worked full-time. For a large number of platform workers, the platform remains an important source of their household income. Economic dependency is therefore a feature of China’s platform work.

The majority of the location-based platform workers are young, male and with little education and low skills levels. In contrast, the level of education of web-based platform workers is high, with 88.1 per cent having completed college or higher education. Increased income is the main motivation for platform workers to choose this employment.

The relatively high level of income comes at the cost of long hours of work: 49.5 per cent of the food-delivery riders worked more than ten hours a day in 2019. The online ride-hailing drivers spent more than half of their working time trying to find an order. And the overall average paid working hours and unpaid working hours of crowdsourced workers were 15.6 hours and 8.3 hours, respectively. Other challenges faced by the platform workers centre around income instability, exposure to occupational injuries, a highly stressful working environment, inadequate social protection, insufficient work and lack of channels to address workers’ complaints and grievances.

A large number of platform workers are under civil law contractual arrangements, often taking the form of labour service contracts. This deprives them of the labour and social protection granted to workers with established labour relations with their employers. According to the Annual Report on the Development of China’s Sharing Economy (RCSE 2020), only 8 per cent of the platform workers in 2019 had established labour relations with their respective platform. While platforms represent an important source of job creation, the lack of adequate labour and social protection has left platform workers in a de facto situation of work informality. With more and more businesses and business models turning to platforms, there is a risk that the trend towards the work informalization might intensify in a context of slower economic growth along with the COVID-19 pandemic.

2.1 Regulation of work relationships in China

There are three types of work relationships in China: labour relationships, in the form of a labour contract and covered by the labour laws; labour service relationships; and employment relationships, both regulated by civil law. The classification of work relationships has important impact for workers in terms of labour and social protections. The following provides the specific definitions to better understand the consequences of each contractual arrangement and applicable laws.

Labour law and labour relations

China’s labour laws encompass the Labour law (1994), the Labour Contract Law (2007), the Trade Union Law (1992, second revision in 2009), the Social Security Law (2011) as well as administrative regulations promulgated by the State Council and administrative rules issued by the labour authorities.

The Government abolished the “employment” terminology along with its primary source, the Republican Civil Code after the founding of the People’s Republic of China in 1949 (Biddulph, Cooney and Zhu 2013). During the planned economy period (from the 1950s until early 1980s), the employment contract did not exist in the State-owned economy, and labour law was nearly non-existent. It was not until the late 1980s that labour laws began to be developed. The Labour Law and the Labour Contract Law are the principle legislation regulating labour relationships in China. As explained further on, the terminology of “employment relationship” differs from the concept used in common law and civil law. It also differs from the definition used in the ILO Employment Relationship Recommendation, 2006 (No. 198).

It is through the concept of “labour relations”, or “labour relationship” that workers and their work units, such as employers, are linked under application of the labour laws. The recognition of labour relations also determines workers’ access to the five mandatory social security schemes to which both workers and employers contribute.

The labour laws apply to workers (劳动者, persons engaged in labour) who form a labour relationship (劳动关系) with a “work unit” (用人单位). According to the labour laws and administrative regulations, there are two qualifiers for a labour relationship.

First, both parties concerned (the work unit and the worker) must meet the principle legal qualifications (主体资格). There is no definition of a work unit in China’s laws, but Article 2 of the Labour Contract Law stipulates: “This Law shall apply to the establishment of a labour relationship between workers and enterprises, individual economic organizations, private non-enterprise entities or other organizations (hereinafter referred to as work units) and to the formation, fulfilment, change, dissolution or termination of labour contracts. The state organs, public institutions, social organizations and workers bound up by labour contracts shall observe this law in the formation, fulfilment, change, dissolution or termination of their labour contracts.” Therefore, such entities as public and private enterprises, registered individual businesses, government agencies, public institutions and social organizations qualify as work units.

There is also no specific definition for “workers”. But Article 3 of the Trade Union Law stipulates that all workers doing manual or mental labour in enterprises, public institutions and government organs within Chinese territory and who earn their living primarily from wages shall have the right to participate in and form trade union organizations pursuant to the law, regardless of their nationality, race, sex, occupation, religious beliefs or level of education. Based on this provision, workers who “earn their living primarily from wages” are legally recognized as workers.

Second, a labour relationship is usually established through a labour contract. Article 7 of the Labour Contract Law provides that a work unit establishes a labour relationship with a worker from the date when the unit puts the worker to work. Article 10 of the law stipulates that a written labour contract shall be concluded in the establishment of a labour relationship. Article 82 of the law requires that a work unit pay a worker two times their salary for each month if the work unit fails to conclude a written labour contract with the worker for more than a month but less than a year from the date the worker began the work.

A “labour relationship” is also not defined in the labour laws. To bridge this gap, the former Ministry of Labour (now the Ministry of Human Resources and Social Security, or MOHRSS) issued the Notice on Matters Concerning the Establishment of Labour Relationship in 2005. The notice indicates that a labour relationship is only established if the following conditions are met at the same time:

1. The work unit and the worker meet the principal legal qualifications.

2. The worker is subject to the work unit’s rules and labour management and engages in paid work arranged by the work unit.

3. The labour provided by the worker is an integral part of the work unit’s business.

Therefore, a labour relationship, according to the labour laws and regulations, is established through two qualifiers and three conditions. This is stated in the Labour Law and confirmed in the MOHRSS Reply to Recommendation No. 6979 of the Second Session of the 13th National People’s Congress, which stipulates that a “labour relationship is the foundation for the protection of workers’ rights and interests”.28

The first qualifier and the first condition are similar: both define the boundary of the labour relationship within which a labour contract can be concluded. This defined boundary can prevent workers who engage in subordinate work from being recognized as party to a labour relationship if the worker and/or the work unit do not meet the principle legal qualifications. This is an important distinction with other jurisdictions as well as the related idea of “indicators” of the employment relationship. In reality, this can be the case for families employing nannies, contractors in the construction industry, re-employed retirees, school interns and some platform workers.

Labour contracts established under the labour law include both open-ended and fixed-term labour contracts, which are considered part of the standard labour relationship. Both types of contract provide the same labour protection, including wages, working hours, paid leave, occupational health and safety protection, termination protection and social security. The Labour Contract Law, which became effective in 2008, extends labour protection to two non-standard forms of a labour relationship: labour dispatch and part-time work.

The labour-dispatching unit is responsible for paying the wages and social security of dispatch workers. Dispatch workers have the right to join a trade union. The receiving unit is held accountable for occupational safety and for some other benefits. This results in joint liability of the dispatching and the receiving units for dispatch workers.29 Part-time work is legally defined as “a form of labour under which the remuneration is mainly calculated on an hourly basis, the average working hours of a worker per day shall not exceed four hours, and the aggregate working hours per week for the same work unit shall not exceed 24 hours”. The work arrangement can be concluded orally and be terminated at any time. Part-time workers may enter into a labour contract with one or more work units. They can join a trade union. Part-time workers are protected by the legislation on minimum wage but not on social security.30 This part-time arrangement is similar to “casual work”, as defined in some jurisdictions.

China’s court system often relies on those three conditions to determine whether a labour relationship exists between platform workers and the platforms with which they have engaged directly or indirectly. As discussed in Chapter 1, Chinese digital labour platforms, while signing labour contracts with their core internal team, often adopt more flexible contractual arrangements with the individuals obtaining task orders from them. Such flexible arrangements include franchising, labour service agencies, labour outsourcing contracts and other kinds of service agreements, all of which are regarded as labour service contracts falling under the scope of civil law and outside the scope of labour law. The same practice of signing labour service contracts, as opposed to labour contracts, applies to labour dispatch firms, through which some platforms engage workers. While the two types of non-standard forms of labour relations under the Labour Contract Law could provide the needed flexibility for platforms in hiring workers, in reality, few platforms use these two forms of non-standard labour relations.

Civil law and labour service relations

“Labour service relations” is the concept used to regulate the economic relationship between two or more entities. There is no unified definition, however. Instead, the concept is used as a “storage box” that covers any economic relationship that is not a labour relationship but involves the provision of and compensation for services and/or labour. A contract for service covered by the Contract Law falls into this storage box too.

The labour service relationship is often defined as a civil right and obligation relationship established by a party that provides labour services to another party who accepts the labour service and pays compensation for it in accordance with the agreement. The parties to a labour service contract can be natural or legal persons and are regarded as equal subjects by the law.

The storage-box function of the labour service relationship is best illustrated by the MOHRSS Reply to Recommendation No. 6979 of the Second Session of the 13th National People’s Congress, which states: “Workers who have reached the legal retirement age are no longer meeting the principle legal qualification in the labour laws, which means they do not meet the preconditions for the establishment of a labour relationship, and their labour contracts will naturally be terminated.” Therefore, “if workers beyond the legal retirement age wish to continue their work, the relationship between the work unit and the worker can be treated as labour service relations. The rights and obligations of both parties can be adjusted according to the civil law.”31

Employment relationship

There is no legal concept of employment relationship in the Chinese labour statutes. The Modern Chinese Dictionary defines “employing” or “hiring” as “purchasing labour with money”, while the definition of “employing or hiring labour” means “labour employed by capitalists”. The term “employment relationship” has been deliberately avoided because it derives from capitalist societies. From the perspective of China’s labour law development, private employment did not exist under the planned economy era. It was not until the promulgation of the Labour Law in 1994 and the Labour Contract Law in 2007 that registered private enterprises and businesses were recognized as “work units”. The infiltration of private rights, such as an employment contract, has gradually transformed the labour relations to respond to the characteristics of a market economy.

The terms “employer” and “employee”, however, are mentioned in the Supreme People’s Court Interpretation on Several Issues Concerning the Application of Laws in the Trial of Compensation for Personal Injury Cases of 4 December 2003. Article 9 of the interpretation provides that “[w]here an employee engaging in an employment activity causes damage to another, the employer shall be liable for compensation; …. Engaging in an employment activity…means engaging in a production or business activity or any other labour service activity within the scope of the employer’s authorization or instructions.” From this interpretation, the employer can be understood as either a work unit as listed in Article 2 of the Labour Contract Law or a natural person who does not meet the principal legal qualifications. Consequently, a person who is hired by an employer does not qualify as a worker under the labour laws, even if there is clear subordination, but is instead an employee. Therefore, the relationship between the employer and the employee forms an employment relationship (Hu 2013). The judicial interpretation stopped short of defining an employment relationship.

The Supreme People's Court went on, in its later documents, to replace the term “employment relationship” with “labour services relationship”. For example, the term “labour service (employment) contract disputes” listed in the 2008 Civil Cases Regulations was revised to “labour service contract disputes” in 2011. Merging the notion of employment relationship with labour service relationship is also evident in judicial practice and in scholars’ research, where it is argued that the issue of labour relations and labour service relations is essentially the issue of differentiating between labour contracts and employment contracts (see Zheng 2016, 57). Due to the lack of legal recognition, the notion of employment relationship has been associated with labour service relationship and is thus regulated by civil law, which represents a significant gap in the regulation of work relationships in China.

Workers whose work relationship is characterized as an employment relationship, such as engaged in subordinate work without meeting the principal legal qualifications for establishing labour relations, include (Xu 2001, 1; Xie 2019, 7–9):

• Workers in the platform economy (those who do not have labour contracts, are not genuinely self-employed and are in a subordination position);

• Workers engaged by natural and/or legal persons in rural areas;

• Domestic workers without a labour contract;

• Workers engaged by unregistered private practitioners, such as doctors and lawyers;

• Many home-based workers or workers in non-traditional workplaces engaged on a flexible-hours basis;

• Workers at the end of subcontracting chains (such as construction workers contracted by unregistered individuals or businesses);

• Many student workers; and

• Retirees.

The fact that those workers are regulated by the civil law and the Contract Law represents an important gap in labour protection. The difference between the occupational injury compensation under civil law and the work injury insurance under labour law is a good example. Under civil law, the level of an injured person’s fault determines the scope of compensation, and the injured person can only receive a lump-sum compensation. In comparison, work injury insurance applies the principle of “liability without fault”, which means an injured worker can enjoy all the benefits of the insurance, no matter if they are at fault or not. This also applies to the situation in which an injury is caused by a third party other than the employer. In addition, an injured worker can continue the labour relationship with the work unit if they meet certain standards of injury or disability (Article 33 of the Regulations on Work Injury Insurance). Beyond the work injury insurance, other protections provided by the labour laws, such as minimum wage, working time, paid leave and access to social security, are not covered by the civil law.

2.2 Regulatory gaps for protecting platform workers

Determination of labour relations

The existence of labour relations is the primary point of reference for determining the rights and obligations between work units and workers. As in other countries, the nature of the work relationship between platform workers and platforms is an ongoing debate in China. Most platforms in China regard their workers as self-employed, thus under a contract-for-service relationship. Some scholars hold the same view (Chai 2018, 57–64; Tu 2017, 109–114). In a similar stream, some view the relationship as a general labour service relationship characterized by multiple and flexible relationships, weakened subordination and a phased mode of payment (Wei Tan 2018, 84–90; Peng and Cao 2016, 93–97). Liu and Li (2017) concluded that workers are no longer employed by a platform that earns profit through information service fees, as exemplified by the operation model of crowdsourcing platforms. Some scholars analysed the personal subordination and the economic subordination of ride-hailing drivers and concluded that the relationship between the drivers and the platforms does not fully fit into the subordination nature of a labour relationship (Xiong and He 2018, 101–105).

Others believe labour relations do exist. Although platform workers have a certain degree of autonomy in terms of working hours and where and how to work, they are still controlled by a platform and must accept the platform’s instructions and supervision (Qi, Ma and Bao, 2019, 205–225; Stanford 2017, 382–401; Aloisi 2016, 1–38; Campbell and Price 2016, 314–332; Sprage 2015, 53–76). It even could be regarded as a new type of employment relationship (He, Cheng and Zou, 2015, 91–97; Li 2014). Some scholars even suggested that workers’ subordination in the employment of the internet economy has been enhanced to some extent (Xie 2018), as opposed to being weakened, therefore the nature of a labour relationship between the two parties has not been changed (Chang and Zheng 2018, 78–88).

Some scholars use the Marxist research method to point out that the operation model of platforms merely changes the traditional employment model by breaking down the dependence of traditional production on the production factors, such as factories and equipment. At the same time, platforms wilfully avoid the establishment of labour contracts and social security obligations. By evading the supervision of the Government’s labour authority, they “reasonably” deduct part of the workers’ income, such as overtime pay, social insurance and other benefits, and turn them into the surplus value of their enterprise (Xiao 2018, 185–190; Yuan and Xu 2018, 119–130).

These debates indicate that digital labour platforms pose challenges to the three conditions (previously explained) that define labour relations in China in the following ways.

First, a labour relationship defined by the labour laws provides that a work unit must directly manage workers through workplace rules and procedures and hold disciplinary powers. However, workers on platforms have a certain level of autonomy and flexibility over working time, work schedule and where and how to work, especially for web-based platforms. For example, workers with Zhubajie can choose the work methods they think to be the most effective and finish the tasks based on their own ideas and technical means.

Second, in accordance with the provisions of China’s labour laws, work units must provide the necessary material and technical means, including tools, equipment and facilities. In the platform economy, most workers must provide their own equipment, including transportation vehicle, mobile phone, computer and so on. Although the value of these devices may be of little significance for an enterprise, it can be a valuable asset for the workers.

Third, work units in a labour relationship must assess the performance of their workers and pay them in full and on a fixed pay schedule (for example, weekly or monthly) based on the wage management system of the work unit. However, platforms usually pay their workers on a piece-rate basis, as opposed to an hourly wage. Workers get their pay instantly after they complete each order or task and after the platform has deducted any service or information access fees and/or commissions.

Last but not least, the triangular relationship that characterizes platforms further complicates the traditional labour relationship between work units and workers – for example, workers who provide the work or service, clients who request and receive the service and the platforms that facilitate and organize (and supervise, for some platforms) the process. Clients are involved in task dispatch, result assessment and payment of remuneration. For example, passengers’ ratings of a DiDi driver’s performance will affect the driver’s evaluation by DiDi and consequently the driver’s priority ranking for the next order dispatch. Some critics argue that the clients’ rating system is just a simple transfer of platform supervision to customers (Xiao 2018, 185–190). However, customers have their own subjective views and experiences that are not controlled by the platform. In many cases, some clients’ biased and one-sided rating have undermined workers’ evaluation and reputation and leaving them with nowhere to seek redress.

It is clear that the nature of the platform work has blurred the boundaries between labour relations and labour service relations (Li 2019, 142; Zhou and Li 2017, 159). This calls for a more encompassing definition and clearly defined criteria to determine the existence of labour relations. As noted at the beginning of this chapter, the two qualifiers and three conditions used for determining labour relations are limited in scope, in the sense that they exclude workers under an employment relationship from labour law coverage. In terms of the second and third conditions, there is no specific criteria, factors or elements provided on how to define and assess “subordination”.

The lack of precise direction and guidance on how to determine the existence of labour relations has made it difficult for courts to render decisions over such disputes. In addition, in one of its 2016 decisions,32 the Supreme People’s Court questioned the legal status of the Notice on Matters Concerning the Establishment of Labour Relationship – the only guidance for determining labour relations. The Court pointed out that, given that the Notice does not fall within the scope of the law because it is only a ministerial policy document,33 courts are not obliged to cite it in cases related to the determination of labour relations. In the absence of legal guidance, courts have taken the matter into their own hands and rendered decisions based on their own interpretation, which has resulted in different court decisions for the same case.

For example, on 10 April 2018, the People’s Court of Chaoyang District in Beijing issued the White Paper on the Trial of Labour and Employment Disputes on the Internet Platforms (Shi 2018). The document shows that from 2015 to the first quarter of 2018, the Court accepted a total of 188 cases concerning labour disputes on internet platforms. Of the 171 concluded cases, more than 84 per cent related to the determination of the existence of a labour relationship. In 39 cases (37.1 per cent) of the 105 for which a decision was rendered, the judge found that the two parties had a labour relationship. The judge found that such relationship did not exist in 58 cases (55.2 per cent) (Shi 2018). Across these cases, there appears to be a tendency to decide in favour of the existence of labour relations when the case involved the determination and compensation of work injury (box 3). If the case only focused on the determination of the employment status, the Court decided against the existence of a labour relationship.

Box 3

This court decision has gone beyond the existing criteria – the two qualifiers and three conditions – for determining labour relations to include economic dependence and corporate social responsibility. Another noteworthy jurisprudence is that some court decisions have recognized the existence of an employment relationship despite its absence in the legal documents. And the criteria used to determine an employment relationship is similar to that of a labour relationship, such as subordination, but the legal protections for the workers are not the same.

Individual rights protection

As in many other jurisdictions, a “binary divide” between workers with and without labour relations exists in China, with labour relations serving as the basis for labour regulation as well as labour and social protection. In addition, the fact that an employment relationship is not specified in civil law and that the labour laws exist independently of civil law leaves workers in an employment relationship without a corresponding regulatory mechanism in civil law while unable to access the protection offered by the labour laws. As a result, the level of protection workers in an employment relationship receive is much lower than those in a labour relationship, who are subject to the labour laws. As indicated previously, work-related injury insurance is one important area in which workers in an employment relationship receive much lower protection than workers in a labour relationship. Other areas in which workers have different levels of protection include:

Remuneration. Employers may refuse to pay or reduce the remuneration to workers in a contract for service because remuneration is conditional on finishing the work.

Duty of care. Workers in an employment relationship may be asked to assume an excessive duty of care in a contract of entrustment, while it is assumed that in such a contract, a trustee has the independence and discretion to handle trusted affairs.

Termination of contract. Employers have the right to arbitrarily dismiss and can terminate the contract for service or the contract for entrustment at any time. In addition, the Contract Law does not provide protection in terms of minimum wages, working hours, occupational safety and health protection or discrimination in employment and occupation.

Remuneration

The labour laws in China protect workers’ right to remuneration. The calculation of labour remuneration can be based on either hourly wage or piecework wage. But even for the latter, the work unit is required to reasonably determine the rate based on working time. The laws provide for composition of wage, minimum wage standard, wage payment method and time of payment. However, these provisions are hardly applicable to the digital labour platforms. Minimum wage does not apply. Workers are paid on a piece-rate basis and immediately after the completion of each task, regardless if they are paid above the local minimum wage standard. When workers run into objections of payment, it is difficult for them to communicate with the platform to discuss a settlement. For crowdsourcing platforms, when clients or customers are not satisfied with the work completed and refuse to pay or reduce the amount of payment, platforms usually do not stand up for workers, leaving them with nowhere to seek redress.

Working time

The labour laws stipulate the legal working hours at eight hours a day and 40 hours a week, and extended working hours cannot exceed one hour (three hours in special circumstances) per day and 36 hours per month. The law also covers rest periods and paid annual leave. While workers with the digital labour platforms have the flexibility to choose working time and decide on rest and leave, they must extend their working hours to make a living due to the relatively low level of remuneration per order and their heavy economic dependence on the platforms. In fact, the long hours of work and fatigue from driving are one of the main causes for accidents for DiDi drivers and Meituan riders.

On the other hand, the current legal regulations on working hours, the limits on part-time work and the administrative approval needed for flexible working-time arrangements (such as an integrated working hour system) have created barriers for platforms to rely on the labour laws to regulate working times.

Occupational safety and health

The major work safety laws include the Work Safety Law, the Law on the Prevention and Control of Occupational Diseases, the Mine Safety Law promulgated by the National People’s Congress and its Standing Committees and the Implementation Regulations of the Mine Safety Law, the Implementation Regulations of the Coal Mine Safety Law, the Regulations on the Safety Management of Hazardous Chemicals, the Regulations on Work Injury Insurance and other administrative regulations issued by the State Council and various ministries.

Workers’ rights and obligations concerning occupational safety and health can be divided into two parts. One part entails the rights and obligations in the production process and at the workplace to prevent and reduce safety accidents and occupational injuries. The other part is the protection after accidents and occupational injuries have occurred, including treatment of injuries, compensation and the provision of basic living conditions. The laws largely focus on the first part and provide protection for all workers, not only those having labour relations. For example, the Work Safety Law (2014) uses the concept of “persons engaged in employment” (从业人员) as opposed to “persons engaged in labour” (or workers, 劳动者) and “production and business unit” (生产经营单位) instead of “work unit” (用人单位). Article 6 of the law stipulates that “persons engaged in employment of production and business units shall have the right to work safety protection and in accordance with the law, and they shall, at the same time, perform their duty in work safety in accordance with the law”. Both terms have gone beyond the scope of the Labour Law and the Labour Contract Law by including workers who are not in a labour relationship as well as business operations that may not meet the principle legal qualification.

The National People’s Congress has provided an interpretation on the scope of the Work Safety Law in which the term “production and business unit” is defined as follows: “The law applies (but is not limited to) all state-owned enterprises and institutions, collectively owned enterprises and institutions, joint-stock enterprises, Sino-foreign joint ventures, Sino-foreign cooperative enterprises, foreign enterprises, enterprises of partnerships, sole proprietorships and individual businesses, regardless of its economic nature and scale, as long as it engages in production and business activities, it shall comply with the provisions of this law, and violations of this law will be investigated and pursued in accordance with the law.”34 How this law will be applied to platform workers remains to be seen.

However, the second part of occupational safety and health protection, such as work injury insurance and compensation, only applies to workers in labour relationships. The Regulations on Work Injury Insurance issued by the State Council (2003) provides that workers should submit “documents proving the existence of labour relationship (including de facto labour relationship) with the employing units” when applying for the ascertainment of work injuries (Article 18).

As mentioned earlier, extending occupational safety and health protection to platform workers has become a government priority. This is reflected in the recent State Council Guiding Opinion on Promoting the Standardized and Healthy Development of Platform Economy, the MOHRSS revision of the Regulations on Work Injury Insurance and different trials carried out on work injury insurance for people in flexible employment at the local level.

Social protection

Major laws and regulations related to social security include the Social Insurance Law (2011), the Regulations on Unemployment Insurance, the Regulations on Work Injury Insurance, the Interim Regulations on the Collection and Payment of Social Insurance Contribution, the Decision of the State Council on Establishing the Basic Medical Insurance Scheme for Urban Workers, the Decision of the State Council on Improving the Basic Old-age Insurance for Enterprise Workers and the Guiding Opinions of the State Council on Implementing Pilot Programmes of New-type Social Old-age Insurance in Rural Areas.

The current social security system, as guided by the Social Insurance Law, concerns mainly urban residents because it provides for workers in a labour relationship. Article 2 of the law stipulates that “the State shall establish a social insurance system, including basic old-age insurance, basic medical insurance, work injury insurance, unemployment insurance and maternity insurance, to guarantee the rights of citizens to legally obtain material assistance from the State and society in case of old age, illness, work injuries, unemployment and childbirth”. The law also requires employers to include every worker (including foreign nationals) in a labour relationship in these five social security schemes35 and share the contributions. Part-time workers who do not participate in the basic old-age insurance in the work unit and workers in flexible employment (including self-employed and individual business unit not hiring workers) can also participate in the basic old-age insurance scheme if they pay the insurance contribution themselves. The same rule applies to basic medical insurance. However, in cities such as Beijing, Shanghai and Guangzhou, the requirement of a local hukou status to participate in these two insurance schemes has created difficulty for migrant workers living and working there.

It has become a public concern in recent years that most platform workers do not have social insurance coverage (Xiao and Zhang 2019). The 2016 Survey Report on the Living Conditions of Online Ride-Hailing Drivers revealed that 94.4 per cent of such workers surveyed did not have any social insurance (SRRI 2016). However, research reports released by the platforms, including the 2018 Research Report on the Food Delivery Couriers by Meituan Research Institute (2019b) and the 2017 Research Report on the Employment of Didi Chuxing Platform by Didi Research (2017), did not include data on social insurance. In the absence of labour relations between platforms and most of their workers, platforms are not obliged to pay contributions.

A concern for the platforms is that the payment of social insurance contributions will increase their operating costs. Estimates indicate that if a courier platform pays the social insurance contributions for its couriers, the cost will be at 1,600 yuan to 2,000 yuan per month per person (Wang and Wei 2019). According to China’s current payment structure, the social security contribution, when paid in full as required by law, accounts for around 40 per cent of total business expenditure.36 When the contributions are paid by the workers only, they represent about 37 per cent of their total wage income, which explains why many are reluctant to participate in a scheme.37

Collective rights

Chinese workers’ right to organize and participate in trade unions and their right to collective consultation38 is protected in the Constitution, the Trade Union Law, the Labour Law, the Labour Contract Law and the Government’s administrative rules and regulations. Therefore, these rights are generally exclusive for workers in a labour relationship.

These laws and administrative rules also stipulate the procedures for participating in and organizing trade unions, the organization structure of trade unions, the rights and obligations of union members and the scope and procedures of collective consultation (table 10). Importantly, the laws only allow workers to organize and join unions within the ACFTU system, which is the only authorized federation of trade unions.

Table 10 Legal provisions regulating collective rights in China

In 2010, the ACFTU began “two general promotional campaigns” aimed to “promote the establishment of trade union organizations, and generally promote collective consultation on wages in enterprises in accordance with the laws” (Sui 2011). By the end of September 2017, the number of trade union members reached 300 million, of which 140 million were migrant workers. There were 2.8 million grass-roots trade union organizations, covering nearly 6.6 million work units.39 By the end of September 2017, nearly 2.5 million collective contracts had been signed, covering 280 million employees.40 According to the Ministry of Human Resources and Social Security, “By the end of 2018, a total of more than 1.7 million collective contracts had been submitted to the human resources and social security authorities for review, covering 155 million workers.”41

In the first few years of the platform economy in China, there were no reports of platform workers organizing and participating in trade unions or being engaged in collective consultations. It was not until March 2018 that the ACFTU put forward the Work Plan to Promote the Union Membership of Truck Drivers and Other Groups. The Work Plan specified that the ACFTU should encourage eight types of workers to join a union: truckdrivers, couriers, nursing or care workers, domestic workers, salespersons in shopping malls, online food-delivery riders, real estate agents and security guards. According to the ACFTU, the unionization of couriers has been increasing steadily. In 2018, the membership rate of couriers increased by more than 20 per cent and was expected to reach more than 15 per cent in 2019.42 In some cities, unions for online food-delivery riders have also been set up.43

Despite the efforts of the ACTU to unionize platform workers, difficulties remain for carrying out collective consultations: (i) difficulty in identifying the employers to secure their willingness to engage in collective consultation(CEC 2018, 40); and (ii) the current focus on enterprise-level (or industry-wide below-county level) collective consultation may not work well for platform workers because they move from job to job more frequently and are less likely to bond with their fellow workers.44 A cross-region and sector-wide approach as opposed to an enterprise-focused approach would help facilitate the identification of consultation parties as well as the consultation process.

2.3 Efforts to regulate platform work

Measures and policies at the central government level

The Government has been supportive of platform development since it was introduced in China. In June 2017, Premier Li Keqiang addressed the Summer Davos Forum, stating that “the Chinese Government has adopted an accommodating and prudent regulatory approach toward new industries, new business forms and models…which has enabled their healthy development”.45 In July 2017, the National Development and Reform Commission, together with seven other ministries, issued the Guiding Opinions on Promoting the Development of the Sharing Economy, which includes comprehensive measures for market access, sector supervision and fostering an enabling environment. This landmark document sets an “encouraging innovation and prudent regulation” as the principle and the “participation of multiple parties and collaborative governance” as the general direction for the regulation of the sharing economy. Over the past few years, several administrative documents concerning the sharing economy were successively issued (table 11), both at the central and local levels, most notably the promulgation of the E-Commerce Law by the 13th National People’s Congress in August 2018.

Table 11 Administrative regulations promulgated by the Government

These central-level documents mainly focus on the promotion and supervision of the sharing economy; a few of them address the protection of workers’ rights and interests. While the Government recognizes the need for the protection of platform workers, no indication has been offered so far regarding the employment status of platform workers. The Guiding Opinion on Promoting the Standardized and Healthy Development of Platform Economy, issued by the General Office of the State Council in August 2019, provides that the legitimate rights and interests of platform workers should be protected, mainly in terms of social security, vocational and skills training and occupational injury protection.46 However, the Guiding Opinion avoided hinting at the nature of the relationship between the workers and the platforms. And it uses the term “occupational injury” as opposed to “work injury”, which is used in the labour laws.

The Interim Measures for the Administration of Online Taxi-Hailing Services came close to defining the nature of the relationship between platforms and workers. The 2015 draft provided that “online taxi-hailing service operators shall sign labour contracts with the drivers registered on the platform”. However, the final version, issued in July 2016 by the Ministry of Transport and six other ministries, only noted that “online taxi hailing service operators should…sign labour contracts or agreements with drivers in multiple forms in accordance with relevant laws and regulations and based on the working time, service frequency and other characteristics while clarifying the rights and obligations of both parties” (Article 18).47 Some critics believe that such change failed to respond to the question regarding the legal relationship between the car-hailing platforms and their drivers and thus did not help clarify the unclear employment status. In addition, this “open” provision, which allows various contract forms to be selected, might encourage platforms to circumvent labour contracts by signing contracts for service and cooperation agreements, regardless of the criteria set in the labour laws (Chai 2018, 57–64).

In September 2018, the MOHRSS indicated that it would undertake to draft guiding opinions on the standardization of the administration and services for new forms of work based on classification of work, with a view to promoting and encouraging platforms and industry organizations to strengthen self-discipline while ensuring basic labour rights protection for workers.48 The MOHRSS also indicated it would expedite a study on reform measures for workers in new forms of work to participate in work injury insurance, and introduce an occupational injury protection system (details to be determined) for these workers as soon as possible. Along with the revision of the Regulations on Unemployment Insurance, the MOHRSS would conduct a feasibility study on unemployment insurance coverage for workers in flexible employment and recommend to the Ministry of Justice to proceed at the legislative level. This could provide a legal reference to the study and formulation of specific measures for workers in flexible employment to participate in insurance schemes.49

In February 2019, the MOHRSS stated that it would begin revising the Regulations on Work Injury Insurance at an appropriate time and “strive to resolve the lack of work injury protection in new forms of work and issues that are prone to cause disputes in practice, so as to protect the rights and interests of persons engaged in employment for work injury from the source and minimize the work injury disputes.”50 The use of “persons engaged in employment” (从业者) rather than “persons engaged in labour” (or workers, 劳动者) could potentially indicate a willingness to extend the coverage to workers in an employment relationship, such as employees who are currently regulated by civil law.

Moreover, the MOHRSS indicated in July 2019 that it is “actively pushing for an early promulgation of the Labour Standard Law, which would provide a legal basis for the protection of the basic labour rights and interests of workers who face difficulty in accessing the protection provided by existing labour and social security laws and regulations”.51 However, the MOHRSS statement stops short of the scope and content of the new law as well as its relation to the Labour Law and the Labour Contract Law.

Initiatives and piloting at the local government level

There are generally two streams of measures taken so far by the local governments for platform workers. One is to reform the work injury insurance scheme and extend social insurance coverage; the other is to focus on wider protection. The former can be illustrated by the Interim Implementation Opinions on Promoting Employees of New Economy and New Businesses to Participate in the Social Insurance Schemes, issued by the Chengdu Municipal Government of Sichuan Province in August 2019. The Interim Implementation Opinions require platform enterprises (referred to in the document as “new economic organizations”) to sign labour contracts with full-time workers and written agreements with part-time persons. If the new economic organizations employ full-time workers, the units and individuals must participate in social insurance schemes in the same way as (traditional) enterprises do (Wang 2019).

The latter stream is represented by the policy documents issued by the Beijing municipality and the Zhejiang provincial authority. In August 2019, the Beijing Postal Administration, the Bureau of Human Resources and Social Security and the Bureau of Health Security Administration issued the Notice on Promoting the Standardized Development of the Express Delivery Industry and Strengthening the Protection of Workers’ Rights and Interests in the Industry. The Notice provides that express delivery enterprises shall conclude labour contracts with their workers, pay the social insurance contributions and pay remuneration on time and in full. Express delivery enterprises are to fulfil the responsibility of work safety and labour protection, provide necessary labour protection supplies and facilities for workers and abide by the special regulations to protect female and juvenile workers. These enterprises must conclude collective contracts with trade unions or workers’ representatives or industry-specific collective contracts on overtime pay, non-competition restrictions, labour protection and special protection for female workers.52

In November 2019, the Department of Human Resources and Social Security of Zhejiang Province issued the Guiding Opinions on the Optimization of Labour Employment Services in the New Forms of Work. The Guiding Opinions provides that enterprises of new forms, such as e-commerce, online ride-hailing service, food-delivery service, express courier delivery and logistics, must sign labour contracts with the persons engaged in their employment and ensure those workers’ rights to rest and to social insurance, among other entitlements.53

Yet, these policy documents issued by local governments are not legally binding, and their administrative regulatory effects in effectively protecting the rights and interests of platform workers remains to be observed and assessed.

Even before the emergence of the platform economy, some regions in China had conducted different trials of work injury insurance for persons in flexible employment.54 For example, Nantong city in Jiangsu Province started a pilot scheme as early as 2006. Before the reform of the state-owned enterprises in the 1990s, the light and textile industries dominated the city’s economy. The state-owned enterprises reform resulted in a large number of laid-off workers who could not find opportunities in formal employment due to the restrictions on age and their lack of skills. Among the laid-off workers, some 70,000–80,000 persons turned to flexible employment. To protect their work safety, the Nantong Municipal Government issued the Notice on Flexible Employment Persons’ Participation in Work Injury Insurance in 2006, which took effect in January 2007. Subsequently, Weifang city (2009) in Shangdong Province and Wujiang District of Suzhou city (2018) and Taicang city (2010), both in Jiangsu Province, also launched trials of work or occupational injury insurance schemes for persons in flexible employment (Weng 2019). While Weifang city chose to incorporate people in flexible employment into the existing work injury insurance scheme, Nantong opted for an occupational injury insurance scheme that operates separately from the work injury insurance scheme through an independent fund.

Although these piloting experiences were not directed at platform workers, they could be revisited in light of the growing number of platform workers who do not have labour contracts and could fall into the category of persons in flexible employment.

Measures taken by platforms

In addition to the relevant policies issued by the central and local governments, some platforms in China have taken measures to protect workers in case of accidental injury, serious illness or hospitalization through commercial insurance coverage. For example, DiDi launched its Diandi medical insurance plan in 2017, which covers accident, serious illness and medical insurance. Drivers can opt in voluntarily. DiDi paid the contributions for the insurance plan in the early stage but eventually changed to automatically extracting contributions from drivers’ income for each order or through a monthly payment. Another programme called Care Protection covers accidental injuries, and the cost is entirely borne by DiDi.

Meituan designed an insurance service using artificial intelligence for its partners and riders that is directly connected to the system of selected insurance companies. The system provides convenient channels for each rider to participate in the commercial insurance plan. Franchisees or agencies can purchase commercial insurance for full-time riders directly in the system, while accidental injury insurance is automatically purchased for crowdsourcing riders after their first order. At present, the insurance scheme for Meituan riders covers death by accident, injury and disability, medical care, work delay, third-party personal injury and property loss.

Due to the limited coverage of the insurance scheme, the low level of protection, the high contribution rate and the difficulties in claim settlement, the participation rate in the commercial insurance scheme has remained low.

Summary

The development of the Chinese labour laws occurred in an environment of a centrally planned and state-owned economy, with a sharp distinction between the agricultural and industrial workforces based on the hukou system and a small service sector. The Labour Law of 1994 and the Labour Contract Law of 2007 comprise the legal framework for regulating work relationships. While the labour laws reflect the evolution of the economic transformation and the changes in the labour market in China, the influence of the planned economy is still evident. For example, a natural person does not qualify as a work unit as they did under the planned economy; only a work unit, rather than an individual, is able to provide the necessary financial support to workers, such as social protection. In the face of the ongoing changes in the labour market, in particular the surge of the digital labour platforms, the labour laws may need to be adjusted again to better serve their purpose.

In this legal framework, labour relations or labour relationships, as opposed to employment relationships, is a precondition for the application of the labour and social security statutes. The concept of labour relationship has some similarities with the concept of employment relationships used in the common law and civil law systems, and in the ILO standards, in the sense that “subordination” is a criterion for determination of the existence of such relationship. Yet in practice, one of the difficulties that has emerged relates to the lack of clear criteria to determine the subordinated position for platform workers who usually enjoy a certain degree of autonomy in their work, particularly in terms of working time.

An important characteristic of the country’s labour law is that both workers and work units must meet the principle legal qualifications before labour relations can be established. The boundary of the labour relationship is restrictive. In a legal sense, there are two types of contracts that regulate work relationships, the labour contract covered by the labour laws and the labour service contract regulated by civil law. However, in practice, labour service contracts are divided into (i) contracts for services regulated by the Contract Law and (ii) employment contracts, also regulated by civil law, including the Contract Law, which is arguably not appropriate due to its evident subordination and/or economic dependence.

The increasing labour disputes over the determination of labour relations are calling for precise criteria to guide court decisions, especially in light of the blurred boundaries between labour relations and labour service relations brought about by the digital labour platforms. While some court decisions have gone beyond the two qualifiers and three conditions for determining labour relations by citing, for example, the factors of economic dependence and corporate social responsibility, a legislative change and/or policy direction can serve to provide an authoritative and sustainable solution. A reflection of the legislative change is the use of the terms “persons engaged in employment” and “production and business unit” in the Work Safety Law (2014). These two terms go beyond the scope of labour relations defined by the Labour Law and the Labour Contract Law and extend occupational safety and health protection to workers who are not in a labour relationship as well as business operations that may not meet the principle legal qualification.

The government at different levels has embarked on various initiatives to regulate the platforms. In the State Council’s Guiding Opinions on Promoting the Healthy Development of the Platform Economy, the MOHRSS is mandated to: “Promptly study to improve the social security policies for persons engaged in employment in the platform enterprises and people with flexible employment, carry out pilot programmes for occupational injury support, actively promote the plan for the full coverage of insurance, and guide more persons engaged in employment of the platforms to get insured, strengthen vocational skills training for platform employees and incorporate them into the vocational skills improvement initiatives”. Local governments, such as in Chengdu, Beijing and Zhejiang, have issued policy documents and urged platforms to sign labour contracts with their workers. And the ACFTU is working to organize platform workers and explore innovative ways for collective consultation (ACFTU et al. 2019, 4).

3.1 Existing approaches to labour protection on digital labour platforms

An ILO study on digital labour platforms in 16 Asian and Pacific countries55 found that national reactions can be categorized in two ways: first, if and how countries allow digital labour platforms to operate and how revenues are traced and taxed; and second, if and how countries attempt to reform their labour laws to either situate platform workers within the existing categories of the defined employment relationships or create new categories of workers, all with the aim to determine the obligations derived from the national labour protection framework for new forms of work. Many countries have responded to the first category of issues, but no country in the world has yet taken legislative action regarding the full coverage of platform workers in existing labour legislation (ILO 2019c, 47–48).

One approach so far has been to partially include location-based platform workers into existing employment laws for certain liabilities and protections but without yet fully defining responsibility of the platform-owning enterprises. In the case of app-based drivers, the current regulatory emphasis is on regulating driver behaviour first to ensure rider safety, with secondary consideration – if any at all – paid to the welfare of drivers (with a focus primarily on occupational safety and health). Regarding the legal status of web-based platforms and the nature of the relationship between the crowdworker, or service provider, the digital platform that mediates the transaction of their work and the crowdsourcer, or client, no country has yet fully sorted out how this phenomenon fits within current labour legislation (ILO 2019c, 47–48).

A recent European Union study found that the most discussed and addressed challenges related to platform work among member countries include employment status, representation and social protection. National responses have somewhat addressed the challenges for location-based platforms but have done little for online platform workers (CEPS, EFTHEIA and HIVA-KU Leuven 2019, 226). As Aloisi and De Stefano (Daugareilh, Degryse and Pochet 2019, 67) noted, “Very little systematic effort has been made to establish how existing regulations should apply.”

Some examples regarding the regulation of platform work include: In France, the August 2016 El Khomri law provides that where the platform determines the characteristics of the service, it must also take responsibility for occupational accident or illness liability (OECD 2019, 31). Since October 2017, the Government of New Zealand has applied stringent rules for licensing ride-hailing drivers and ensuring their compliance with working-time laws, and drivers have been prosecuted for violating labour standards (Plumb 2016). In the Republic of Korea, revisions to the Act on Industrial Safety and Health were announced in January 2019. One of the planned amendments is to include platform workers under the protection of the Act.56 In Indonesia, as of May 2019, a minimum tariff for fares was implemented in five cities as a preliminary trial to be monitored closely by the Ministry of Transportation (ILO 2019c, 48).

3.2 Guiding international labour standards

Employment Relationship Recommendation, 2006 (No. 198)

In most national jurisdictions, the employment relationship is the main vehicle through which workers gain access to the rights and benefits associated with employment, and it is the point of reference for determining the nature and extent of rights and obligations of workers and employers. In some countries, an increasing number of workers lack labour and social protection. This can be attributed to many factors, but from a legal perspective, the causes might be: (i) the scope of the law is too narrow and restrictive; (ii) the law is too general and ambiguously formulated; and (iii) the employment relationship is disguised.

ILO Employment Relationship Recommendation, 2006 (No. 198)57 provides the following tools and guidance to help determine the existence of an employment relationship (ILO 2016):

• The principle of the “primacy of facts” – the determination of the existence of an employment relationship should be guided by the facts relating to the actual performance of work and not on the basis of how the parties describe the relationship.

• The means for determining the existence of an employment relationship can be broadened and so can the factors and indicators that may be used to assess the criteria established by law, such as subordination or economic dependence.

• The legal presumption of the existence of an employment relationship whereby one or more relevant indicators are present – indicators that may prompt such a presumption could vary between countries and include the fixing of working hours or workers’ economic dependence.

These valuable tools are often used in both developed and developing countries to ensure that all workers in need of protection are adequately protected. In the case of China, the principles, means and tools offered in the Recommendation can shed light on defining the scope and criteria of a labour relationship and fill the gap in protection for workers in an employment relationship (recognizing the differences in the definition of an employment relationship in China).

Transition from the Informal to the Formal Economy Recommendation, 2015 (No. 204)

The ILO Transition from the Informal to the Formal Economy Recommendation, 2015 (No. 204)58 provides guidance for the development of an integrated policy framework to progressively facilitate an economic transition while preventing informalization. Specifically, it provides guidance to Member States to pursue three objectives: (i) facilitate the transition from the informal to the formal economy; (ii) promote the creation, preservation and sustainability of enterprises and decent jobs in the formal economy and the coherence of macroeconomic, employment, social protection and other social policies; and (iii) prevent the informalization of formal economy jobs.

The Recommendation calls on Member States to progressively (and urgently) extend, in law and practice, occupational safety and health rights to all workers in the informal economy as well as access to social security, decent working conditions and a minimum wage. Tools to improve compliance and facilitate the transition include the reduction of compliance costs, the provision of incentives (such as access to credit and information), the strengthening and extension of inspection, efficient complaint and appeal procedures for workers, effective sanctions and the facilitation of the access to social protection.

With the increased informality in employment resulting from various contractual arrangements on the digital labour platforms in China, the Recommendation offers useful guidance in facilitating the transition from the informal to the formal economy while preventing informalization.

Universal Protection and Recommendation on Social Protection Floors, 2012 (No. 202)

The increased diversity in forms of employment has prompted the ILO to call for the establishment of universal minimal protection regardless of the forms of contractual arrangements. China has progressively established a comprehensive social security system with significant coverage of health insurance and old age insurance. However, challenges remain to be tackled, such as closing the social security gaps for a growing number of platform workers.

The Universal Protection and Recommendation on Social Protection Floors, 2012 (No. 202)59 defines a social protection floor as a set of basic social security guarantees aimed at preventing or alleviating poverty, vulnerability and social exclusion. Countries are recommended, in accordance with their national circumstances, to establish as quickly as possible and then maintain their social protection floor comprising basic social security guarantees. The guarantees should ensure at a minimum that, over the life cycle, all people in need have access to essential healthcare and to basic income security, which together thus secure effective access to goods and services defined as necessary at the national level. Reaffirming that social security is a human right and a social and economic necessity, the Recommendation provides guidance to Member States to establish and maintain a national social protection floor as a fundamental element of their national social security system and to implement their floor with strategies for the extension of social security that progressively ensures higher levels of protection to as many people as possible and guided by ILO social security standards.

The Global Commission on the Future of Work60 regards social protection as “a human right”. The report of the Commission calls for governments to guarantee universal social protection that includes a social protection floor that affords a basic level of protection to all in need, complemented by contributory social insurance schemes of increased levels of protection. It recommends that societies fill the gaps and adapt systems to the evolving world of work by extending adequate social protection coverage to workers in all forms of work, including self-employment, and extend coverage to vulnerable workers in the informal economy. “As the organization of work continues to change, social protection systems will need to evolve to deliver continued protection for workers who move between wage employment and self-employment, between different enterprises and sectors of the economy or between countries, ensuring that rights and benefits are accessible and portable, including for those working on digital platforms” (ILO 2019e, 35–36).

ILO Centenary Declaration

The ILO Centenary Declaration for the Future of Work, adopted in June 2019, calls on Member States to strengthen the capacities of all people to benefit from the opportunities of a changing world of work and to strengthen the institutions of work to ensure adequate protection of all workers. Reaffirming the continued relevance of the employment relationship as a means of providing certainty and legal protection to workers, the Declaration provides that “All workers should enjoy adequate protection in accordance with the Decent Work Agenda, taking into account:

• respect for their fundamental rights;

• an adequate minimum wage, statutory or negotiated;

• maximum limits on working time; and

• safety and health at work” (ILO 2019a, 5).

The Global Commission’s report, which served as a basis and an inspiration for the Centenary Declaration, calls for the establishment of a Universal Labour Guarantee that includes fundamental workers’ rights, an “adequate living wage”, limits on hours of work and ensuring safe and healthy workplaces. It recognizes the need to “review and where necessary clarify responsibilities and adapt the scope of laws and regulations to ensure effective protection for workers in an employment relationship. At the same time, all workers, regardless of their contractual arrangement or employment status, must equally enjoy adequate labour protection to ensure humane working conditions for everyone” (ILO 2019e, 38).

With respect to the fundamental principles and rights at work, the ILO Declaration on Fundamental Principles and Rights at Work61 “[d]eclares that all Members, even if they have not ratified the Conventions in question, have an obligation arising from the very fact of membership in the Organization to respect, to promote and to realize, in good faith and in accordance with the Constitution, the principles concerning the fundamental rights, which are the subject of those Conventions:

• freedom of association and the effective recognition of the right to collective bargaining;

• the elimination of all forms of forced or compulsory labour;

• the effective abolition of child labour; and

• the elimination of discrimination in respect of employment and occupation.”

3.3 Policy options for better labour protection on digital labour platforms

China’s economic growth has been slowing since 2016, with a registered gross domestic product growth of 6.9 per cent in 2017, then 6.6 per cent in 2018 and 6.2 per cent in 2019 (Zou 2019). The platform economy represents a booming sector against the backdrop of decreased international trade and stagnant domestic demand, which has been further complicated by the COVID-19 pandemic. Digital labour platforms have become an important channel to create new jobs and have indeed improved the overall income level of many workers, including the low-skilled ones. The Government focuses on “encouraging innovation” in the development of the platform economy and takes an “inclusive and prudent” approach towards its regulation. At its executive meeting in December 2019, the State Council reiterated the pledge to support job creation while rolling out measures to cancel unreasonable restrictions on non-standard employment. Premier Li Keqiang stated at that meeting that a stable job market is a key indicator of economic performance, and it is important to stabilize existing jobs while creating new ones.62

Caution has been raised regarding the overly heavy costs of social insurance to be borne by platform enterprises if platform workers are to be classified as workers with labour relationships. According to some scholars, this could lead the end of the platform business model and the loss of an important channel of employment. On the other hand, the misclassification of many platform workers and the deliberate circumvention of the labour laws have resulted in the denial of access to basic labour and social protection for those platform workers. Facing this dilemma, the Government has been taking a cautious approach in finding a middle ground between the current divide of “labour relationship for everything” and “no labour relationship for nothing”.

Based on the analysis in the previous parts of this paper, the following policy options are offered for consideration.

Expand the scope and establish clear criteria of labour relationships

As indicated earlier, similar to many other jurisdictions, a binary divide between workers with and without labour relations exists in China, with labour relations the reference point for labour regulation as well as labour and social protection. This makes the definition of labour relations and the classification of the work relationship as a labour relationship a central element in the provision of labour and social protection. For this reason, it is fundamental to clarify the scope of the labour relations, ensuring that the criteria used in the determination are sufficiently clear, comprehensive and up to date to keep pace with the changing labour market and business models.

Chapter 2 specifies that labour relations, as defined by the current labour laws, are bound by two qualifiers and three conditions. Therefore, even if the two parties in a work relationship meet the second and third criterion (engaged in subordinate work), their relationship cannot be considered as a labour relationship if both parties or either of them do not meet the principal legal qualifications. When not meeting this criterion, workers will be regarded as in an employment relationship. In addition, the fact that the concept of employment relationship is not recognized in the Chinese legal framework leaves workers in this relationship without a corresponding regulatory mechanism in civil law while not able to access the protection offered by the labour laws (Xie 2019, 9). As a result, the level of protection that workers in an employment relationship receive is much lower than in a labour relationship because they are subject to civil law, including the Contract Law.

Both the Labour Law and the Labour Contract Law could be reviewed to make sure that workers who are in need of protection are duly protected. Specifically, the following aspects could be considered:

1. Define carefully the principal legal qualifications so as to be more inclusive. Particularly, the term “work unit” could be defined to extend its narrow scope (Dong and Qiu 2005, 57). In this way, parties can at least proceed their cases to the assessment of subordination stage in courts rather than being dismissed at the beginning due to not having met the principal legal qualifications.

2. Apply the principle of the “primacy of facts” as suggested by the ILO Recommendation No. 198. The determination of the existence of a labour relationship should be guided by the facts relating to the actual performance of work and the control of work process, not on the basis of a written labour contract or how the parties describe the relationship.

3. Emphasize a worker’s economic dependence on the entity to better reflect the changing organization of work.

4. Establish a set of unified factors, indicators and standards for assessing a subordinate position. This is particularly important for platform workers who have a certain degree of autonomy, especially in the arrangement of their working time.

To extend the scope of the labour laws, the Work Safety Law could serve as an example. Its intended wide coverage of both workers and workplaces through the use of the terms “persons engaged in employment” and “production and business units” reflects the need to protect workers beyond the scope of labour relations. In terms of the development of specific factors and indicators concerning the existence of labour relations, court decisions at different levels across the country have offered some useful criteria to consider. Likewise, the international labour standards and other countries’ experiences can be a source of inspiration too.

The White Paper on the Trial of Labour and Employment Disputes on the Internet Platforms issued by the People’s Court of Chaoyang District of Beijing (Shi 2018) identifies that the first challenge in dealing with labour disputes at the court level is the lack of policy guidance and standards for the determination of labour relations. China’s judicial community has been calling for clear and up-to-date criteria to determine labour relations, especially in light of the emerging platform economy and new forms of work. The China Enterprise Confederation recognized that “[t]he boundary and scope of labour relationship is expanding and extending, and the interest appeals of both labour and management are constantly updating and changing, which brings challenges to the labour relations coordination mechanism”, and called for the government to formulate corresponding laws, regulations and policies to secure the flexibility of the labour market (CEC and CNE 2018, 40–41).

In the context of the Government’s employment-first policy and support for innovative technologies and business models as new engines for employment generation, any regulation concerning the digital labour platforms must be prudent, practical and based on solid research and data. The rapidly evolving business models in the platform economy and their impact on working conditions certainly merit ongoing research and analysis. In the meantime, developing clear and comprehensive criteria for the determination of labour relations would fill a longstanding regulatory gap, especially given that the only guiding policy document, the Notice on Matters Concerning the Establishment of Labour Relationship, has not been revised since 2005. Clear criteria would contribute towards achieving the policy goals of building sustainable enterprises and protecting workers’ rights and interests by promoting fair market competition, addressing disguised labour relationships, fostering decent working conditions and facilitating effective labour dispute resolutions.

Due to the complicated and protracted process in amending laws, it may be more practical to tackle this issue through a judicial interpretation at the current time, with a view to testing and evaluating the criteria that lead to legislative change at an appropriate time in the near future.

Establish a floor of labour standards by decoupling rights

In many jurisdictions, the binary divide has served as a workable simplification of complex economic and social realities that have enabled an effective policy response to regulating work relationships. However, this simplified fashion may not work anymore because increasingly larger segments of the workforce in many countries no longer fit well within this binary classification. As Countouris (2019) pointed out, the origins of employment law in feudal master-and-servant relationships have left emphases on subordination, continuity and bilaterality, which are increasingly irrelevant to the purposes of workplace regulation today.

Yet the continuing emphasis on these factors in the definition of employment allows those who hire workers to manipulate the legal relationship so as to avoid forms of legal responsibility required by the labour laws. Subordination, as historically defined and understood in most legal systems, has been replaced by other forms of control, which have perhaps brought about problematic outcomes for workers. Meanwhile, the absence of continuity and bilaterality continues to produce the forms of vulnerability found in traditional employment relationships.

It is thus necessary to respond to new forms of control and dependence in a more disaggregated way. The starting point is to rethink what types of control should attract legal responsibility. To this end, there is a need to consider which economic actors control outcomes that affect fundamental interests of workers and to recognize that the answers will vary according to the interest. Rather than having a single boundary, the scope of legal responsibilities will probably vary. This could mean that the identity of the employer or the liable entity will differ or could be multiple, depending on the type of work regulation. It could also mean that for some kinds of regulation, employment should not be the boundary anymore. For example, regarding occupational safety and health in Canada and Australia, the boundary lies well outside traditional employment and is based on which enterprise controls conditions affecting health and safety for workers.

Along this line of thinking and taking into consideration the vulnerabilities that many platform workers face, as described in Chapter 1, a set of core labour standards should be established and applied regardless of the underlying contractual arrangements or employment status for the purpose of ensuring humane working conditions for everyone. Those rights could start with the following:

Fundamental principles and rights at work enshrined in the 1998 ILO Declaration – The Declaration makes it clear that these rights are universal and that they apply to all people in all Member States, regardless of the level of economic development, and therefore to the platform economy. As a founding member of the ILO, China has been working towards the enhancement of these principles and rights for all Chinese workers. Realization of the fundamental principles and rights at work has been a priority area of China’s Decent Work Country Programme 2016–2020.

Occupational safety and health – The Work Safety Law goes beyond the boundary of labour relations, and the liability appears to be based on an enterprise controlling the conditions affecting safety and health. This seems to respond to the call of the ILO Centenary Declaration to incorporate occupational safety and health into the fundamental principles and rights at work. It also responds to the requirements in ILO Recommendation No. 204 regarding the urgent extension of occupational safety and health rights, in law and practice, to all workers in the informal economy.

Minimum wage – The rationale is that workers should not be paid less than a certain amount per hour unless they have a meaningful opportunity for profit and therefore can be expected to bear a corresponding risk of loss. Following this logic, it would make sense to ask whether the platform workers accept the work on terms that offer a meaningful opportunity for profit. Because this is not the case, as highlighted in Chapter 1, a minimum wage serves to protect platform workers against unfair low pay and helps ensure a just and equitable share of the economic gains. It would respond to the call of the ILO Centenary Declaration as well as to the Universal Labour Guarantee recommendation of the ILO Global Commission’s report.

Working time and weekly rest – From a safety and health perspective, a maximum limit on working time and compulsory weekly rest must be respected. The regulation on the hours of work, including the establishment of a maximum working day and week, is considered an element of humane working conditions in the 1919 ILO Constitution.

Extend social protection coverage

The lack of or inadequate social insurance is the most urgent issue facing workers in the platform economy. In the short term, a compulsory and workable work injury insurance scheme can be used as a breakthrough point and can lead to the design of other social insurance schemes. Although platforms are required to contribute to the premiums, the Government should consider subsidies. The work injury insurance piloting experience for flexible employment at the local level should be assessed as to whether it can apply to platform workers. The social insurance system could be improved in the following aspects: (i) extend the workplace-based social insurance system to people who are not in a labour relationship and detach social insurance schemes from the labour relationship; (ii) reform the hukou system to remove barriers for internal migrant workers to participate in the basic old-age and medical insurance schemes in the place of their work; (iii) change the contribution model from wage-based to income-based; and (iv) improve methods of collection, mobility between jobs and national coordination.

Strengthen collective bargaining

The collective bargaining right is especially important to bridge gaps in the legislation and to provide context-specific responses to the needs of workers and employers. While platform workers can be hard to organize, given their high mobility and turnover, network technology provides convenient conditions for their communication. In China, local unions can take responsibility for organizing platform workers and even act as negotiators when workers are not able to elect a representative.

According to the ACFTU, the Beijing Express Delivery Association and the Beijing Express Delivery Workers’ Federation organized enterprises and workers’ representatives in early 2019 to sign China's first Special Collective Contract for Labour Protection in the express delivery industry and agreed on setting up labour protection inspectors and purchasing accidental injury insurance for workers.

Elsewhere in the world, there has been an emergence of rejuvenated or even completely new collective organizations, such as the Spanish workers’ collective and informal associations, solidarity movements like the broodfonds in the Netherlands, the Independent Uber Guild in New York and the Independent Workers Union in Great Britain and other initiatives aimed at helping or supporting the collective organization of platform workers (Daugareilh, Degryse and Pochet 2019, 140–141).

In December 2018, the European Committee of Social Rights ruled on the possibility for self-employed workers to conclude a collective agreement. The Committee made an important statement that “the decisive criterion is rather whether there is an imbalance of power between the providers and engagers of labour. Where providers of labour have no substantial influence on the content of contractual conditions, they must be given the possibility of improving the power imbalance through collective bargaining”. This could serve as food for thought in the search for a solution to platform workers who find themselves confronted with such a power imbalance.63

Encourage social responsibilities of platforms

Corporate codes of conduct, while being typically voluntary, can be a source of self-regulation and also a way for industries to develop best practices. The best practices of an industry often form the basis for the starting point of regulation. Such efforts have already begun in the crowdwork sector (Cherry 2019, 36). There are good examples of platform enterprises taking on social responsibility. For example, in 2017, eight Germany-based platforms signed the Crowdsourcing Code of Conduct in which they agreed to include local wage standards as a factor in setting prices on their platforms and to set up an Ombudsperson’s Office to deal with complaints from platform workers. The code of conduct was first initiated by the Munich-based software testing platform Testbirds and is officially supported by the German Crowdsourcing Association (Deutscher Crowdsourcing Verband). The German Metalworkers’ Union, IG Metall, helped with the code’s development. IG Metal has called for government regulation to ensure that workers who are employees are not misclassified as self-employed and to ensure a “level playing field” among workers.64

In January 2020, the China Internet Association organized 14 platforms to sign and launch the Platform Enterprises Caring for Workers Initiative, calling for them to take on social responsibility and exercise a duty to improve the image of the platforms and enhance competitiveness. A number of measures were suggested in the initiative, including fixing reasonable remuneration standards, protecting the safety and health of workers, reducing the accident rates, participation in social insurance schemes in accordance with the law and actively participating in commercial insurance schemes. The initiative stops short of concrete actions other than presenting general principles. Nevertheless, it is a step forward for the Chinese platforms in recognizing their social responsibility.