Sharing Economy: Challenges for the Labor Market and the Labor Law in China and Globally

The sharing economy has had a substantial impact on the labor market and created challenges for companies in traditional industry and businesses. In conducting micro-comparative research on the existing standards that test the employment relationship between Uber and Uber drivers, this study reveals how different traditional labor law dogmas deal with the conundrum brought about by the sharing economy. The test of personal dependence derived from German labor law has been a model for China, Japan, and other civil law countries as the main standard to identify an employee. This test encounters a failure in the context of the sharing economy. Unlike China, the latest court decisions in the jurisdictions of the UK, the EU, and the US tend to increasingly categorize Uber drivers as employees. This study proposes that a new test or standard in labor law should be developed ‘bottom-up’ through the increasingly consistent development of local court decisions and local regulators’ efforts. A category that lies midway between the employee and the semi-employee is proposed for Uber drivers and other workers intermediated by sharing firms. The sharing firms (platforms), rather than individual workers, would have the ability to undertake the social cost and economic burden incurred when they are categorized as employers in labor law. Only these platforms can invoke their technological and organizational power to reasonably allocate such costs and burdens.


Introduction
The sharing economy, also known as the 'platform economy,' or 'collaborative economy', has emerged as the alternative suppliers of goods and services traditionally provided by long-established industries. 1 In 2017, the number of employees of China's sharing economy platform companies was approximately 7.16 million, an increase of 1.31 million over the previous year. In 2017, around 70 million people participated in providing services in the platform economy, an increase of 10 million over the previous year. 2 Sharing firms such as TaskRabbit, Airbnb, and Uber have appeared in the United States over the past decade and have quickly spread globally, also in China. 3 The ride-sharing company, Uber, has developed a business model that relies on a peer-to-peer platform, which enables individuals to collaboratively make use of an under-utilized inventory via fee-based sharing. This business model is driven by two key factors: technological innovation and supply-side flexibility. 4 With rapid urbanization and a fast-paced population growth in emerging economies such as China, many residents with relatively low incomes have gathered in major cities. The widespread availability of smartphones and technologies like rating systems has facilitated trust among strangers. 5 With this background, large cities have provided a natural hotbed for the development of the sharing economy. 6 In the urban passenger transport service business, Uber and similar internet-driven car-sharing firms have rapidly filled the huge demand gap in the existing taxi services of numerous metropolises.
Although the economic influences and regulatory framework of the platforms have drawn extensive attention from academia and policymakers, 7 until recently their impact on the labor market and labor law has not been sufficiently deliberated. 8 To fill the remaining gap, it is necessary to discuss and solve one of the most essential challenges brought about by the sharing economy, namely the conundrum it has caused for the labor market and labor law. This study takes up that question, adopting the perspective of comparative law to analyze different approaches in Asia, the EU, and the US. Considering that China is one of the most important Asian markets for Uber and similar sharing firms, functioning as a pilot for regulating and policymaking in this regard, this paper specifically concerns China as a typical Asian approach.
This study's contributions are as follow: firstly, via normative and comparative research on the existing standards that test the (employment) relationship between the sharing firms and the workers, it reveals the conundrum that traditional labor law dogma faces when dealing with the challenges brought about by the sharing economy. Secondly, the study proposes that a new test or standard in labor law should be developed 'bottom-up' by the emerging collective initiatives of the workers along with the increasingly consistent development of local court decisions and local regulators' efforts. Part 2 introduces the labor market problems caused by Uber's disruptive business model that challenges the basis of the existing labor law framework. Part 3 adopts a micro-comparative legal analysis of the standards that test the employment relationship by comparing court decisions in different jurisdictions. Part 4 discusses the controversy in different labor law theories as to whether platform-intermediated workers (Uber drivers) should be categorized as employees or independent contractors. Part 5 introduces the most recent collective initiative by the labor organizations of platform-enabled workers (in different jurisdictions) that advocate fairer conditions in the sharing economy, which would reshape the development space left for the sharing firms and other stakeholders. Part 6 contains a conclusion.

Uber's Disruptive Business Model and Its Influence on the Labor Market
The internet-driven car-sharing market in China is made up of firms utilizing one of two types of business models: (1) the 'tailored taxi' and (2) the 'carpool'. The tailored taxi model comprises two sub-types: platform-enabled car-hailing between drivers and passengers, where vehicles are made available by the taxi companies. The second sub-type requires that firms supply vehicles themselves and employ drivers, which is similar to the traditional taxi business, except that it is internetdriven. The 'tailored taxi' model enables sharing firms to manage a large, disaggregated workforce through their ride-hailing platforms. Leading firms of this subtype in China are Uber, 9 Didi Chuxing, and Yidao. Empirical research suggests that under this model in China, 90% of all vehicles used for sharing are owned by the drivers. 10 The 'carpool' model, sometimes referred to as 'ride-sharing', also consists of two sub-categories. The first is 'not for profit', which mediates ride-sharing among individuals carrying out a longer journey together. This is known as 'Gefaelligkeitsverhaeltnisse' in German civil law literature. 11 A typical platform of this type in 9 Uber is a San Francisco-based company founded in 2009 that owns and operates a smartphone application for 'ride-sharing', connecting drivers of privately-held vehicles with riders who pay a fare set by the company. Uber is reputedly valued at $ 62.5 billion in its latest funding rounds; see Newcomer (2015); Uber is available in 195 cities in North America and 68 countries worldwide; see Kalanick (2016). 10 Xiong (2016) Germany is Blablacar. 12 Previous work on 'carpooling' (or 'ride-sharing') in general has explored its ad hoc, not-for-profit, or cooperative context. 13 The second sub-type is 'for profit', which relies on internet platforms to quickly match vehicle supply and demand based on travel routes and to automatically calculate the costs based on mileage. Internet platforms powered by sharing firms such as Uber and Didi Chuxing have enabled large-scale growth for the for-profit carpool. The platform-enabled peer-to-peer taxi and the for-profit carpooling models have much in common. For instance, both models face the claim that they are 'unfairly' competing with traditional taxi services by circumventing the labor protection enshrined in labor law. They have both been subjected to criticism that their long-term impact on labor standards is unclear with unpredictable and dark implications for the future of low-wage work. Therefore, this article examines these two models by combining them under the term 'Uber', despite Uber China having merged into Didi Chuxing in a $ 35 billion deal in 2016, as representative of a typical business model created and utilized by sharing firms in the urban passenger transport business. 14 Uber's business model is as follows: its smartphone-based app connects drivers offering rides and passengers seeking them. Passengers pay mileage-based fees via electronic transactions (in China mostly by Alipay or Wechat-Pay), which Uber keeps on file, taking a percentage of each fare and giving the rest to the drivers. Uber sees itself as a pure technology company that mediates a peer-to-peer contractual relationship. This being so, the value-adding ingredients of the intermediaries (sharing firms) is argued to consist of only services via the platforms to achieve the matching processes between peers. 15 The socio-political effect brought to the labor market by Uber's new business model becomes obvious when compared with the traditional organizational form of value creation in the same service market. The traditional taxi companies that used to run a great deal of the city passenger transport services have to bear numerous labor and socio-legal responsibilities as employers. These compulsory liabilities, consisting of signing an employment contract, the payment of social insurance, vacation leave, maternity leave, and so on, are provided in labor laws that differ from jurisdiction to jurisdiction. 16 For example, in Germany, the provision of the minimum wage of 8.84 euro per hour was introduced on 1 January 2017, as a compulsory provision that would entail significant consequences for 15 Kollmann et al. (2016). 16 Arts. 40, [70][71][72][73][74][75][76]16  Uber and other sharing firms. 17 The business model of Uber is bemoaned as being unfair as it circumvents these labor and social responsibilities and transfers the risks to the individuals providing the driving services. Uber does this by referring to these individuals as 'independent contractors/entrepreneurs' in its user agreements, typically via the general business condition clauses. Uber declares that its drivers conclude separate individual contracts for every ride they offer to passengers.

The Relationship Between the Platform and the Individuals Intermediated by Them
One of the most controversial issues in lawmaking and policymaking regarding Uber is how to categorize the status of Uber drivers, namely whether they are employees or independent contractors, as well as whether and to what extent the existing labor protection provisions should apply to them. This question is an essential issue that imposes an effect far beyond the discussion of dogma within the framework of labor law per se. To a certain extent, the policymaking in this regard shapes the space left for the present and future development of sharing firms. Although this question lacks a sufficient legal analysis, 18 academia's efforts to address it should not be limited to those within the dogma of labor law. This issue is a lever for the digital industry to shape the sharing economy and should be deliberated with a balanced and proactive stance and a strategic social-political vision.

The EU and China: Rationale of the Policymaking and Controversies
In the European Union, the liberalization of labor through unregulated platformbased intermediation has been perceived positively, at least at the Union level. The European Commission has emphasized that the collaborative economy would create new jobs that even could bring opportunities for higher incomes for the EU citizens. The European Parliament has stressed that 'the collaborative economy offers possibilities for young people, migrants, part-time workers and senior citizens to access the labour market'. 19 Even if there is no further justification or supporting arguments for these views, the attitude of the EU's top-level policymakers is that the unregulated forms of jobs in the sharing economy, even as a business model without sufficient social responsibility that passes the risks onto the worker, could reduce the barriers for entrants to the job markets. 20 The following differentiation is noteworthy: when the platform intermediates jobs that only bring occasional additional income for workers who are already covered by social insurance, there is not much concern about the internal relationship between the workers (or the service providers) and the platform operators. In contrast, when the platform-enabled job opportunity is the main source of their income, the delineation of the employment relationship is a fairly serious question in labor law. In China, this issue gained the attention of policymakers on the national level from the very beginning when they approved the sharing economy's market entrance to mainland China. The following provision was stipulated in Article 18 of the 'Interim Measures regarding Internet-booked Car Service Regulations' (hereinafter the Interim Measures Regulations), 21 the first specific regulatory framework that the Chinese legislature has adopted for platform intermediated peer-to-peer car-sharing services: Network-booked car-sharing companies should ensure that drivers who provide services have the necessary legal qualifications, in accordance with relevant laws and regulations. According to the length of the work, service frequency, and other characteristics, the firms should sign with the drivers a variety of labor contracts or agreements in order to clarify the rights and obligations of both parties. [emphasis added] Similar provisions are found in almost every local regulation on the provincial level in China. 22 The wording of the policymakers seems to intend to prevent Uber from categorizing their drivers as independent entrepreneurs instead of employees.
However, the status of Uber drivers warrants more thoughtful discussion. Critiques of the compulsory employment contract model typically advance the following arguments: First, some argue that the provision of compulsory labor contracts would kill off Uber. As mentioned above, the supply-side flexibility is a hallmark of the sharing economy. Uber drivers can add or remove themselves from the available supply of drivers with a swipe on an app. 23 A compulsory implementation of employment contracts would erode supply-side flexibility, and thus be devastating to Uber and similar sharing firms. 24 Second, some commentators argue that the compulsory employment provision in China's local regulations will create huge costs for Uber because of the need to purchase social insurance for its drivers, and would thus impede the innovation-driven development of the sharing economy in this field. Further, they argue that under such a policy constraint, Uber could eventually be assimilated into the traditional taxi business, and passenger benefits derived from 21  Uber's innovative business model would thus be reduced. 25 Third, some argue that sharing platforms should be considered as neutral arbiters of technological transactions, similar to credit card processors, rather than traditional employers with social obligations towards their employees. Uber drivers should be classified as independent contractors, as their employers often communicate job expectations in the form of suggestions or recommendations, rather than instructions. 26 Fourth, some Chinese academics have referred to the local regulations in California, Colorado, and Washington DC that agree with the solution that rejects categorizing the relationships between Uber and its drivers as traditional employment contracts. They argue that the regulatory approach as such in the US is deliberate and pro-innovation and that it should thus serve as a model for Chinese policymakers to facilitate new business models and related innovations.Some have emphasized that a survey taken in Shanghai shows that most Uber drivers even consider themselves to be contractors rather than employees. 27 On the other side, observers who are in favor of introducing labor protection for Uber drivers have adduced numerous counter-arguments, as follows: Critics stress that Uber circumvents legal requirements through regulatory arbitrage and that its approach to matters regarding labor protection is a typical example of this. 28 This view reveals that Uber manages to avoid labor laws by characterizing all participants in its platform as 'consumers' of its technology. 29 Empirical research suggests that Uber enforces significant indirect control over how drivers do their jobs. The information asymmetries produced by Uber's application are essential to its ability to impose control over individuals. Furthermore, Uber uses digital technology and algorithms to structure asymmetric relationships with individuals in a way that is favorable to itself. 30 The latest study on Uber drivers' experiences points to the need for a greater scrutiny of the role of the platforms in shaping power relationships between employers and workers. 31 A report on Uber drivers in the US shows that when workers switched from 'dependent' to 'independent' employment, this did not necessarily improve their revenue situation. Their average hourly wages in 2015 were $ 13.17 in Denver, $ 10. 75 in Houston,and $ 8.77 in Detroit. 32 Considering that Uber drivers use their own vehicle, pay all travel-related costs themselves, and are not typically covered by social insurance, then the US minimum wage of $ 7.25 has not been attained, at least not fully. If the platform workers who lack bargaining power are, in the long run, under-remunerated, the fact that numerous individuals lack social insurance to take care of their health and retirement is an important factor in estimating the influences of Uber's business model. Considering the lack of social insurance, the reduced barriers 27 Tsinghua University China Case Center for Public Policy and Management (2016). 28 E.g., Cherry (2016), p 577. 29 Ibid. 30 Rosenblat and Stark (2016) Calo andRosenblat (2017), p 1642. to entering the job market cannot necessarily be considered to be a positive step despite the 'flexible' financial opportunity and the 'freedom' to shape working conditions (time, location, service objects) made available by the platform operators. 33 Employed taxi drivers often earn even less than Uber drivers, but they are faced with considerably lower additional costs and risks. When the normative question of to which institutional conditions such services should be subject remains unanswered, it is not plausible to argue that the full use (or 'sharing') of the otherwise 'underutilized' resources is an incentive to facilitate new business-sharing models. For the same reason, it is not plausible to argue that each citizen has an autonomous decision to shape the conditions of his or her employment activity, unless the individual is free to choose between different forms of employment types and has made his or her choice based on an informed deliberation that, inter alia, this 'greater autonomy' in his or her working environment has its 'price'.

How to Delineate the Employment Relationship: Court Decisions and a Comparative Analysis
Not many cases on this issue have so far been adjudicated in China, so more scrutiny is required as to how to interpret the provisions of the compulsory employment contracts enshrined in Article 19 of the Interim Measures Regulations and most local regulations. When enforcing these provisions, it should be taken into account that any labor law arrangement should pay close attention to the structure of employment hierarchies constructed through the sharing platforms. The wording of the provision 'according to the length of work, service frequency, and other characteristics' should be subject to a cautious analysis. Simply labeling someone as an independent contractor will not settle the issue, and neither will any measure that obliges workers to sign documents agreeing to be treated as independent contractors. Considering that the frequency of disputes concerning the employment relationship between Uber (and other similar sharing platforms) and its drivers (and other workers intermediated by them) could rapidly increase in the courts in the near future, a functional test should also be established for the Chinese courts. An approach to balancing the interests of different stakeholders should be found in order to apply and develop existing labor laws in response to the challenges posed by Uber's business model. This study discusses this question by addressing two specific issues. First, what should be the (new) standard or test to categorize the relationship between Uber and its drivers, and second, whether and to what extent employment protection should be granted to Uber drivers. It is noteworthy that this question concerns not only the internal relationship between the platform operators and the respective workers intermediated by them, but it is also shaped by the qualification of the external relationships between the platform operator (Uber) and the service recipients (the passengers/users of Uber): When the platform operator has nothing to do with the customers in this external relationship, then the platform workers can hardly be classified as employees in the internal relationship with the platform operators.
In the EU, a recent decision by the European Court of Justice (ECJ) has addressed this issue as follows: Accordingly, the answer to the first and second questions is that Article 56 TFEU, read together with Article 58(1) TFEU, as well as Article 2(2)(d) of Directive 2006/123 and Article 1(2) of Directive 98/34, to which Article 2(a) of Directive 2000/31 refers, must be interpreted as meaning that an intermediation service such as that at issue in the main proceedings, the purpose of which is to connect, by means of a smartphone application and for remuneration, non-professional drivers using their own vehicle with persons who wish to make urban journeys, must be regarded as being inherently linked to a transport service and, accordingly, must be classified as 'a service in the field of transport' within the meaning of Article 58 (1)  Uber had originally denied that it was a transport company, arguing instead that it was a computer services company with operations that should be subject to an EU directive governing e-commerce and prohibiting restrictions on the establishment of such organizations. Now, according to the ruling of the ECJ, Uber's service must be excluded from the scope of the freedom to provide services in general as well as that of the Directive on services in the internal market and the Directive on electronic commerce. 35 The ECJ found that Uber's services were more than an intermediation service. It observed that the Uber app was 'indispensable for both the drivers and the persons who wish to make an urban journey'. 36 As provided for in EU law, the transport sector falls under Article 4, paragraph 2 TFEU, one of the areas in which the European Union has competing powers with the Member States. Presumably, Uber will be subject to stricter regulation in the EU Member States, for example, the rules regarding the minimum wage and holiday leave as enshrined in the labor law of the Member States, for instance, Article 1 of the German Minimum Wage Law and the U. National Minimum Wage Act 1998 ('NMWA') and associated Regulations. 37 The judgment of the ECJ referred to a detailed analysis of the business model by Advocate General Szpunar, who asserted that Uber has generated numerous contractual instructions that shape the offer of passenger transport services for the drivers by using algorithms and information technologies. In other words, Uber defines the essential features of the work offered to its drivers. It regulates and organizes the functionality of these job offers. 38 In some districts in Germany, labor courts have followed a similar rationale categorizing Uber as a service provider in the field of road traffic rather than in the field of internet technology. 39 Some courts have directly avoided addressing this issue, but have stressed that Uber should be held at least jointly liable for the impairment to competition in the urban transport market. 40 Uber's circumvention of labor laws has generated several class-action lawsuits in the US. Lawsuits in California have alleged that drivers for Uber are not 'independent contractors' but 'employees'. 41 In response, Uber has argued that drivers should be classified as 'end-users' or 'consumers' of its software, akin to passengers. 42 Drivers in the UK have also sued Uber, alleging violations of the UK's Employment Rights Act 2008. They have cited Uber's substantial control over their working conditions as the main cause of action. 43 The first decision resolved the issue by granting protection to Uber's drivers under British labor laws. (i.e., by finding that Uber drivers were in fact employees, as opposed to independent contractors). 44 In September 2017, Uber's appeal against this decision was dismissed. 45 The court held that 'when they are in their territory, have the app switched on, and are able and willing to work; or when they have accepted a trip, the drivers "are working for Uber"'.
Although the above indicated jurisdictions tend to classify Uber drivers as employees who rely on the job opportunity offered by Uber as their main income, and to grant them basic labor protection, the judiciary in China has adopted a different stance. Most recently, the Beijing District Courts have addressed this issue with a rather conservative stance. They have found the following: The platform of Didi, as the runner of the App, constitutes an intermediary information service provider. No labor relationship can be found between the individuals and the platform. Because Didi drivers have the right to choose when to take orders, the activities of taking orders and providing the service do not constitute a duty. 46 Such a judicial approach in China has to be understood against its particular background. An investigation of several judges in the Chinese courts shows that they are more or less influenced by the policymakers at the highest level so that they agree that the 'technological innovations and new business models in the internet industry that drive economic growth should be protected and encouraged '. 47 This being so, the ethos underlying labor protection has to give way to the value of innovation and economic growth. On the other hand, labor protection arrangements in this industry would result in huge costs and would therefore be devastating for these sharing firms. The roots of such a judicial interpretation lie in the existing labor law theory and provisions. So far, Chinese academia have followed the 'personal dependence test' when delineating the employment relationships between Uber and its drivers. This test was introduced from German law theory by Chinese labor law academics.

The Standard for Testing the Employment Relationship with Uber:
A Comparison Between Germany, the US, China, and Japan

Civil Law Countries
For decades, German labor law has been a model for China, both in jurisprudence and policymaking. German law has adopted the test of personal dependence as the main standard to identify an employee under labor law. 48 Both academia and most court decisions in Germany have held that the crucial criterion in this regard should be the 'employees' characteristic as to whether they are personally dependent.' 49 'Personal dependence' as such is assumed to exist when the person must follow instructions (given by the entity) and that person has integrated into the organization of the client (the entity). The instructions, commitment, and integration are usually considered to be cumulative characteristics that are required for the existence of an employment relationship. In this sense, the new § 611a (1)  Civil Code, which has been in effect since 1 April 2017, comprehensively defines an employment contract as one that contains an obligation to perform instructionissued, extraneously assigned work with personal dependence. By doing so, the legislator has simply acknowledged the judicial practice established by the Federal Labor Court, thereby abandoning the intention to create a new definition of an 'employee'. 50 Following this decision, Uber drivers can hardly be categorized as employees: they are free (although such freedom is limited) to refuse an assignment on the Uber app on their cellphone. Further, whether the Uber-intermediated virtual car-hailing network could be interpreted as an 'organization' under the § 611a (1) of the German Civil Code remains unclear, and it is unknown whether the drivers are 'integrated' into such an 'organization', since they have the freedom to choose when to turn on the app and whether to accept an assignment by Uber. However, as more and more commentators in Germany tend to argue, the description of a certain contractual relationship (as an employment relationship) is not as important as the identification of the substantial features thereof. Any legal analysis of this new phenomenon should consider which repercussions such a legal analysis could bring to the gig economy, and whether the sharing firms could bear and reduce the resulting socio-economic burden if they were to be categorized as an employer in the field of road traffic. It should also be considered how to balance the interests of the platforms, the workers, the passengers and the markets in a fair and appropriate way. A platform operator like Uber imposes technology-based evaluation and reputation mechanisms on both the immediate driving activity and the other behavior of the driver via sophisticated management disguised as 'contract offers', as well as sanctions all the way up to blocking and deleting accounts. Taking all of this into consideration, an employment relationship should be considered to exist. 51 The academic discussion in the US has suggested that even when finding that an employment relationship exists, the legal consequences thereof, such as the application of the minimum wage, could still be practically manageable. 52 More specifically, it does not matter how Uber drivers are labelled (as employees or independent contractors), any analysis of Uber drivers' legal status should focus on the de facto extent of Uber's control over their behavior through the respective information technology infrastructure. A dogmatic analysis in this regard should not be entangled with conceptual law. The German system of labor law divides labor providers into three categories: employees, semi-employees, and the self-employed. The term 'semi-employees' as defined in the German Federal Labor Law refers to the regulations in accordance with Article 12a, paragraph 1 of the German Collective Contract Law, namely a person who is economically dependent and who needs social protection similar to employees. According to Article 12a of the German Collective Contract Law, individuals who are economically dependent should obtain the same social protection as employees, as long as the following elements are met (1) they should complete the task under the contract by themselves; (2) they mainly work for one (legal) person, or more than half of their income is paid by one (legal) person. 53 According to Article 2, second sentence of the German Federal Annual Leave Act, semi-employed workers should be treated in the same way as employees with minimum annual leave and public holidays. 54 According to Article 6, paragraph 1 of the German General Equality Law, 55 semi-employees should be protected against any discrimination based on race, nationality, sex, and religion.
When trying to appropriately attribute tort liabilities between the internet platforms and the worker intermediated by them, judicial practice in Japan has developed a new concept called a 'contract labor employee.' Whether these workers should be granted labor protection depends on their position in the 'four concentric circles' system. This system consists of four concentric circles, from the outside to the inside: 'general rights' (Circle I), the 'rights of unpaid workers' (Circle II), the 'rights of paid workers' (Circle III), and the 'rights of dependent workers' (Circle IV). The level of likelihood and necessity to grant labor protection increases from Circle I to Circle IV and so does the level of employer liability. Uber drivers would find themselves between Circle III and Circle IV, depending on their individual circumstances. At least the rights of paid workers should be granted. However, there is still some room left for more tailored divisions and identifications. Noteworthy, however, is that under Japanese law, atypical workers, such as those intermediated by internet platforms, should be covered by social insurance, especially health insurance, as long as the term of their contract exceeds two months. The costs of the insurance are split equally between the workers and the sharing firms Depending on the de facto extent of Uber's control over their drivers' behavior (in different jurisdictions), they should be categorized either as employees or semiemployees or midway between these two categories. Either way, social protection, such as social insurance, health care, and compulsory road traffic accident insurance would be granted for Uber drivers. The concrete cost-sharing scheme could be collectively negotiated or suggested by the local regulators from one administrative district to another.

Common Law Countries
As the counterpart of the civil law countries, the common law jurisdictions have adopted a more flexible test of 'the right to control' to delineate the internal labor relationship. 56 In the US, the well-known Borello Test was established by the California Supreme Court in 1989, whereby a plethora of factors should be considered to determine whether a worker is an employee or an independent contractor in a 53 See Art. 12a Tarifvertragsgesetz  workers' compensation case. 57 These factors include: (1) the right to discharge at will, without cause; (2) whether the one performing the services is engaged in a distinct occupation or business; (3) the kind of occupation, with reference to whether in the locality in question the work is usually carried out under the direction of the principal or by a specialist without supervision; (4) the skill required in the particular occupation; (5) whether the principal or the worker supplies the instruments, tools, and the place of work for the person doing the work; (6) the length of time for which the services are to be performed; (7) the method of payment, whether according to the time taken to do the work or by a fixed sum; (8) whether or not the work is part of the regular business of the principal; and (9) whether or not the parties believe that they are creating an employer-employee relationship. However, the Dynamex decision delivered by the California Supreme Court on 30 April 2018 presumes that all workers are employees. It set out a new 'ABC' test that businesses must satisfy in order to classify workers as independent contractors in a dispute involving wage orders. 58 By adopting this new test, the business in question bears the burden of proving that the worker satisfies all three of the following factors: A. The worker is free from the control and direction of the hiring entity in connection with the performance of the work, both under the contract for the performance of the work and in fact; B. The worker performs work that is outside the course of the hiring entity's business; and C. The worker is customarily engaged in an independently established trade, occupation, or business.
An undecided question is whether the Dynamex 'ABC' standard will replace the Borello Test in the future. The reasoning in the Dynamex judgment appears to limit the application of the 'ABC standard' to situations involving wage orders, 59 so it is likely that the Borello factors will remain a benchmark for this issue for the foreseeable future. Before the Dynamex ABC standard was introduced, the United States District Court had held in the case of O'Conor v. Uber Techs Inc that the drivers were presumptive employees of Uber, not independent contractors, because these drivers performed a service for Uber and Uber depended on its drivers' performance of these services for its revenues. If the Dynamex ABC test is applied in the future, the courts will tend to grant labor protection because the new test assumes that the drivers are employees, and the business bears the burden of proving that the worker satisfies all three of the introduced factors in order to overturn this presumption. The 57 See the decision in S. G. Borello & Sons, Inc. v. Dept. of Industrial Relations 48 Cal.3d 341, 54 Cal. Comp. Cases 80 (1989). 58 Dynamex Operations West, Inc. v. Superior Court 4 Cal.5th 903, 83 Cal. Comp. Cases 817 (2018). 59 Ibid. Especially the wording: 'Here we must decide what standard applies, under California law, in determining whether workers should be classified as employees or as independent contractors for purposes of California wage orders, which impose obligations relating to the minimum wages, maximum hours, and a limited number of very basic working conditions (such as minimally required meal and rest breaks) of California employees.' new test reverses the burden of proof and thus changes the balancing of the interests involved.

Mainland China
German law has influenced policymaking in China to such a great extent that the latest regulatory policy still focuses on the criterion of personal dependence. In China, there is no doubt that labor law applies to the second sub-type of the tailored taxi model that was referred to above as a 'platform-enabled peer-to-peer taxi'. Because of this type of 'sharing,' for example, the 'Shenzhou Tailored Taxi' (Shenzhou Zhuanche), the vehicles are owned by the sharing firms or leased from professional leasing firms while the drivers are provided by a third-party employment service company. Employment contracts are signed between the drivers and the employment company, which constitutes a typical employment relationship. Further, a labor lease relationship is established between the sharing firm and the employment service company.
Secondly, when delineating labor relationships in the case of Uber and similar sharing firms, regulators and courts in China have been resorting to the resources of comparative law. Although no consensus has been found among different local regulators, 60 the courts still tend to protect Uber as an internet service provider under the principle of technology neutrality. The policymaking in this regard should change the status quo in the near future. As illustrated above, civil law countries, including Germany and Japan, often adopt the criterion of dependence to determine the labor relationship, inter alia, the personal dependence, the economic dependence, and the organizational dependence. As discussed above, in the context of the sharing economy, the function of the personal dependence test in determining the labor relationship is weakening. This test is hardly appropriate in the context of the business model adopted by Uber. In the sharing economy, industrial upgrading and technological progress have led to increasingly detailed levels of labor division. In most circumstances, the personal dependence of Uber drivers and similar workers on sharing firms cannot or can only be partly established. This emerging reality results in challenges to the traditional labor law theory. On the one hand, it is difficult for Uber to exert full control over the work of Uber drivers, who are, compared to traditional employees, free to arrange their working time and place. On the other hand, Uber drivers are still dependent on the economic benefits of the platform, sometimes as their primary income. The purpose of work as a means of survival has not changed. Labor providers are increasingly dependent on the operation of the internet platform, while payment for any work done has also become an increasingly important part of the production and operation of the 'sharing' platform. Considering these factors, labor law scholars in China have proposed that the existing standard should be reformed 61 or that a more flexible standard should be adopted that would take different factors that affect the identification of the labor 60 Zhang (2019), p 462. 61 Xie (2018), p 1546. relationship into account. 62 The relationship between Uber and its drivers seems, at first sight, to be an equal relationship. As a matter of fact, however, their relationship is in many circumstances substantially unequal. Therefore, some academics have suggested that the standard for recognizing the labor relationship should break away from the shackles of the personal dependence test in order to respond to the diversified 'employment' model that exists in the case of Uber. Recently, they have tended to suggest that the test of economic dependence should be relied upon instead of personal dependence. 63 As analyzed hereinafter, it can be found that the Chinese courts have not applied and interpreted the standard of personal dependence in a coherent and reasonable manner. This has led to frustration among the academic labor law community in China and they have therefore tended to propose an alternative standard. However, it is noteworthy that the economic dependence test has not yet been able to substitute the standard of personal dependence in order to classify the labor relationship in the case of the platform economy. It is expected that with the passage of time a consensus will emerge among Chinese labor law academia that a 'one size fits all' standard is not feasible in the context of the sharing economy.
In summarizing typical cases adjudicated by the Chinese courts in this regard, it is noticeable that the criteria for determining labor relations are fairly arbitrary in China. As indicated hereinafter, judicial practice does not yet have a unified standard as how to interpret the 'personal dependence' for the identification of labor relations when platform-intermediated workforces are involved. 64 For the above-mentioned reasons (in Part 3.2), some Chinese courts are reluctant to determine labor relations in these cases. In the labor dispute between the driver Zhuang and the 'e-drive' company, the court held that the driver did not have a fixed workplace and he could determine his own working hours; the driver did not receive a steady income from the company every month. Considering the characteristics of this new industry and the fact that the driver and the company had signed a 'driver cooperation agreement' in this case, the court held that no labor relationship could be confirmed. 65 In the decision on the dispute in Shenyang Zhong Ye Technology Co. Ltd. v. Li Hui, the court held that the courier worker in question and the platform company had established a labor relationship. The main reasons for this were that the company would pay for the damage incurred by third persons, in accordance with the original prices, if the delivered items had been broken in transit; the courier used the transportation method provided by the company; the courier's work was a component of 62 Wang and Wang (2020), p 57. 63 Feng and Zhang (2011), p 92. 64 The identification of labor relations is the focus of labor dispute cases involving sharing firms. For example, from 2015 to 2018, the People's Court of Chaoyang District in Beijing accepted a total of 188 labor dispute cases related to the Internet platform of sharing firms, of which 61.2% required the identification of labor relations; among the 171 cases completed, in more than 84% of these cases both parties were in dispute as to whether there was a labor relationship between the parties. No consistent application of the personal dependence test was established, see Chao Yang District Court (2018); see also the comment by Xie (2018) the company's main business; the courier had to comply with the distribution rules established by the company; and that the company paid remuneration to the courier. Therefore, the court held that the two parties had entered into a labor relationship. 66 In 2017, in two disputes concerning sharing firms' liability in road traffic accidents, both of the courts in Chengdu and Beijing held that Uber should bear liability for the damages to the third party. However, the courts did not clarify the legal relationship between the drivers and Uber. 67 For drivers who rely on Uber as the source of their primary income rather than as hobbyists or part-time drivers, the role of platforms as an intermediary in shaping power relations between the firm and the drivers is more dominant. This being so, a labor relationship is more likely to be found to exist. Correspondingly, Uber drivers are more likely to be granted labor protection. But, so far, the standard of economic dependence has not been established in China. For example, no consensus has been reached as how to recognize economic dependence. What percentage of the incomes of Uber drivers should come from Uber to support the finding that economic dependence exists? How can the income of the drivers be investigated when they work for multiple sharing firms or even run their own Wechat online shop at the same time? So far, this recommendation by academia to establish a new standard has not been accepted by the policymakers. The Labor Court of the German Reich used to adopt this standard. However, it later gradually realized that economic independence is neither a requirement nor a sufficient condition to identify labor relationships. The Labor Court then corrected its opinion and stressed that the standard that determined the employment and labor contract should be the personal dependence test. 68 But Chinese academia should bear in mind that the personal dependence test was formed under fairly different social-economic conditions in the first half of the 20th century. When addressing the conundrum of Uber's labor law issues, it should be taken into account that the technological development of the internet has reshaped the roots of many labor law theories.
Indeed, a more flexible standard would be helpful. The factors of this new standard could be identified or developed via a bottom-up approach. The local courts in China decide on this issue on a case-to-case basis, which could form a database for policymaking by the Supreme People's Court of China when drafting a guideline for the interpretation and application of the law in this field. This guideline could include different categories tailored to different forms of service provision by the platform. When drafting this guideline, a comparative analysis of this issue could 66  provide a basis for the Chinese judiciary and policymakers. Over the last three years, Seattle has started to require employers to provide paid 'sick and safe' leave, created an Office of Labor Standards to target wage theft, passed legislation that will raise the minimum wage to $ 15 per hour, and required that certain employers follow fair scheduling practices. 69 In China, the drivers should be granted, as the case may be in different jurisdictions, (at least) limited labor protection, such as paid vacation and inclusion in the technical safety protection system. But the minimum wage or protection from dismissal should not necessarily be granted.
Another mechanism along with labor law for policymakers, regulators, and the courts could be control over the general terms and conditions adopted by the platforms, since these contractual provisions are consistently formulated between the platform operators and the workers intermediated by them.

Trends
It is noteworthy that in many jurisdictions so far, judicial practice and local government regulatory policies have played a key role in drawing a distinction between employees and contractors in the context of the sharing economy. 70 As the discussion on the labor policy regarding the sharing economy is still developing at different stages in different jurisdictions, this issue is far from being concluded in China as well as globally. Thus, it is beyond the capacity of this article to define a 'one size fits all' new standard that would clearly categorize internet intermediated workers as employees or as independent contractors. Empirical research suggests that most of Uber's drivers had full-time or part-time employment before joining Uber, 71 and many continued in those positions after starting to drive with the Uber platform. 72 Further, it has been revealed that information and power asymmetries produced by the Uber application are fundamental to its ability to structure control over Uber drivers, and that the rhetorical invocations of digital technology and algorithms are used to structure asymmetric corporate relationships concerning labor that tend to favor the former. 73 As shown above, unlike China, the courts in the jurisdictions of the UK, the EU, and the US tend to increasingly categorize the Uber drivers as employees. Considering all of this, a category somewhere between employees and semi-employees in German labor law should be established for the Uber drivers because the sharing firms, rather than individual workers, have the power to bear the social cost and economic burden incurred when they are categorized as employers under labor law. Only the platforms can invoke the technological and organizational power to reasonably allocate such costs and burdens.
This approach does not necessarily rely on legislation or regulations, but might start from judicial practice or bottom-up collective initiatives. Collective initiatives have emerged in major cities in the US which have enabled drivers to negotiate with the sharing firms, for example, the Seattle Ordinance allowing taxi and for-hire drivers to be classified as independent contractors so that they can unionize and bargain collectively. 74 This law is largely a response to precarious working conditions in the app-based 'gig economy', which depends on an army of workers who are paid on a task by task basis and are unable to enjoy the protection afforded to regular employees. 75 The Seattle Ordinance begins with a statement of purpose, grounded in both commercial stability and workers' rights. In a representative provision, the Ordinance states: Collective negotiation processes in other industries have achieved public health and safety outcomes for the general public and improved the reliability and stability of the industries at issue ... In other parts of the transportation industry, for example, collective negotiation processes have reduced accidents and improved driver and vehicle safety performance. 76 This introductory language may decrease the likelihood that the Ordinance will be struck down by a court, a topic that is currently being discussed in the US. 77 Whether and to what extent the collective initiatives among the Uber drivers in China, comparable to the Seattle Ordinance, would jointly shape policymaking and lawmaking regarding labor law issues remains to be seen. In this regard, more solid empirical research on this issue would be helpful for deliberated policymaking in China.
In order to address the specific issue of identifying the employment relationship between workers and sharing firms, a bottom-up development through court decisions is possible when the local courts have a guideline to facilitate coherent judicial practice. The Supreme People's Court of China could issue such a guideline as a next stage. The preparation for this guideline lies in summarizing existing court decisions and maintaining dialogue among practitioners, policymakers and academia. A (new) test consisting of a set of factors should be established. As provided in the Interim Measure Regulations, 78 these factors include, but are not limited to, the following: the length of the work, the type of service, and other characteristics. One important factor should be added here, namely whether the drivers' income relies mainly on the work intermediated by Uber, or whether they are merely a hobby drivers. The evidence for such an economic dependence could be the number of rides or the time undertaken by the drivers within a certain period of time. For example, 'qualifying drivers' under the Seattle Ordinance were proposed to be anyone who had initiated a contractual relationship with a driver coordinator at least 90 days before the Ordinance's commencement date and who had driven at least 52 trips 'during any three month period in the 12 months preceding the commencement date'. 79

Concluding Remarks
Ideal regulations that establish a reasonable amount of labor protection should have the following functions: (1) to grant bottom-line protection for the workers intermediated by sharing firms in cases of traffic accidents and ride-related injury or rights infringements; (2) to clarify the distribution of liabilities (to external third parties) between the sharing firms and the workers intermediated by them; (3) to encourage the innovation and development of the sharing economy while imposing fair social responsibilities to the sharing firms. The new mechanism must be able to strike a balance between the interests (in conflict) and to find a point of equilibrium between governmental regulation and market freedom. In this regard, the construction of new best labor policies and regulations for the sharing economy should involve interaction between policymakers, regulators, academics, sharing firms, passengers, and algorithm workers. 80 At least one consensus has been reached by the academic community so far: that a new test to determine labor protection in the context of the sharing economy should be established. This test should be more flexible, diversified, taking into account the special features of the work patterns of the algorithm workers. Further, as the comparative analysis above indicates, it is becoming common to tend to grant labor protection for platform-intermediated workers who are economically dependent on sharing firms. These include public holidays, paid sick leave, and especially industrial accident insurance. As to whether there is a need for traditional social insurance for Uber drivers, this could be collectively discussed by and among different interest groups and decided by local policymakers according to the specific economic circumstances of the local communities.
This study is limited in two respects: first, empirical research on the collective initiatives of Uber drivers in different jurisdictions has not been sufficiently and systematically undertaken and reported. Thus, adequate literature and information have not been available for this study to rely upon. Further, the suggested new test of economic dependence to determine an employment relationship needs to be considered by policymakers in different jurisdictions. This process is still developing so the findings of this study need to be tested further down the line and rethought as lawmaking encounters a breakthrough in the future. 79 Seattle Director of Finances and Administrative Services (2017). 80 Rosenblat and Stark (2016), pp 3758, 3766.