ARTICLES

 

Achieving decent work for digital platform workers in South Africa

 

 

William Manga Mokofe

LLB LLM LLD; Senior Law Lecturer, Eduvos, Independent Researcher, Advocate of the High Court of South Africa

 

 

---

SUMMARY

Across the globe, digital platform workers lack access to basic workplace rights and protections. Lacking the ability to bargain collectively as a result of this novel, digitally managed market for work, many platform workers have jobs characterised by long and irregular hours, lack of representation, low remuneration, high stress levels, and little ability to negotiate wages or working conditions with their employers. This article questions the narrow definition of "employee" that limits the protection afforded to these workers. Workers have been largely incorrectly classified as independent contractors rather than as employees, a classification that has been challenged in courts around the world. This article explores possible strategies and offers suggestions to improve this situation to achieve decent work for such workers. The article concludes by suggesting that the solution to challenges associated with platform workers rests, among others, in social dialogue, legal solutions, new union strategies, and state intervention through the extension of existing legislation to platform workers.

---

 

 

1 INTRODUCTION

Customarily, the legal concepts of "employer", "employee", and "employment relationship" have been used to define the scope of labour law, differentiating between employees (subordinate and dependent workers) and independent contractors. The different forms of platform work and the different legal interpretations of employment relationships across national contexts make it difficult for regulators to determine where platform workers slot into established labour market concepts. In many instances, these workers are categorised as self-employed, which raises concerns that this legal designation would not match the factual reality of workers' relationship with, and dependence on, a given platform or client.¹ These platform workers are at risk of lacking both the flexibility of traditional autonomous work and the security of an employment relationship. In fact, the issue of worker classification is vital because it is a fundamental determinant of a person's access to labour rights, benefits, and entitlements. In particular, classification opens entry to social protection. There is a need to manage the distribution of risks and costs related to the "new markets opened up by digital platforms". However, the challenge for policymakers lies in approaching the platform economy in such a way that the downsides can be tackled while pursuing the opportunities that this branch of the world economy opens up.

Digital platforms are now a crucial part of contemporary law; they permit us to arrange a ride, order food and access myriad other digital services. They achieve this by connecting clients or customers with workers who undertake these tasks or "gigs". The last decade has seen a global rise of "gig" or platform workers; an Uber is an obvious example. Digital work platforms have generated unprecedented opportunities for workers, enterprises, and society by releasing innovation on an enormous global scale. Yet, simultaneously, they have posed serious threats to decent work and fair competition.

The platform economy is growing fast with estimates that digital labour platforms worldwide now earn at least US$50bn per annum.² Examples include platforms operating in ride-hailing, food delivery, personal services, and digital content creation. There are estimated to be up to 40 million platform workers in the global South alone - some 1,5 per cent of the total workforce.³ It is difficult to measure the size of the digital-platform labour force, but some estimates suggest it involves up to 60 million workers in the global South, concentrated in India, Pakistan, Bangladesh, and the Philippines, with a limited presence in sub-Saharan Africa.⁴ In South Africa, it has been estimated that there are some 100 000 workers in web-based crowdwork and 35 000 in location-based platform work.⁵ The current estimate of the size of the digital-platform labour market in South Africa is approximately 135 0000 workers or 1 per cent of those in employment.⁶

While platform work offers income and opportunities to many, there are also numerous instances of unfair work practices. Examples of issues encountered in research are low pay, wage theft, unreasonable working hours, discrimination, precarity, unfair dismissal, lack of agency, and unsafe working conditions.⁷ In most places and sectors, workers cannot bargain collectively, and, because of their employment status, are not protected by relevant employment law. Problems faced by platform workers when it comes to recognition of their rights are broadly similar across the globe. In recent years, legal action regarding the misclassification of platform workers as "independent subcontractors" as opposed to "employees" has also become commonplace the world over.

 

2 WHO IS AN EMPLOYEE IN SOUTH AFRICA?

Ordinarily, the term "employee" means an individual who works part-time or full-time for another individual, organisation, or the State, and is paid for rendering a specific service to his or her employer - whether or not in terms of an oral or written contract of employment. It is worth noting that the Labour Relations Amendment Act (LRAA)⁸ has deleted the term "contract of employment" from the definition of "dismissal" in section 186 of the Labour Relations Act (LRA).⁹ Dismissal therefore currently means that "(a) an employer has terminated employment with or without notice".¹⁰

The employee also has recognised rights and responsibilities. It is important to define the term "employee" because much hinges on the distinction between employees and persons who are not employees. South African labour legislation applies only to employees, and only an employee can claim protection against unfair dismissal.

An employee is defined in South African labour law in the LRA as:

(a) any person, excluding an independent contractor, who works for another person or the State and who receives, or is entitled to receive, any remuneration; and

(b) in any manner assists in carrying on or conducting the business of an employer.¹¹

The term "employee" is defined in similar terms in the Basic Conditions of Employment Act (BCEA),¹² the Skills Development Act (SDA),¹³ and the Employment Equity Act (eeA).¹⁴ This definition is significant as only those who meet the requirements of the definition of an employee are protected by the labour legislation. There are different definitions of "employee" in the Occupational Health and Safety Act (OHSA)¹⁵ and the Unemployment Insurance Act (UIA).¹⁶ It is, therefore, possible for a worker who is not regarded as an employee in terms of the LRA and the BCEA to be entitled to compensation in the event of an injury at work and to be entitled to unemployment insurance.

There are two parts to the definition of an employee in the LRA and the BCEA. Paragraph (b) might be sufficiently broad to include workers who otherwise might not be viewed as employees under paragraph (a).

In Liberty Life Association of Africa v Niselow,¹⁷ the Labour Appeal Court (LAC), referring to part (b) of the definition, held as follows:

"The latter part [of the definition] in particular may seem to extend the concept to employment far beyond what is commonly understood thereby. To adopt a literal interpretation though would clearly result in absurdity. I think that the history of the legislation which has culminated in the present statute, and the subject matter of the statute itself, lends support to construction which confines its operation to those who place their capacity to work at the disposal of others, which is the essence of employment. It is not necessary in this case to decide where the limits of the definition lie. It is sufficient to say that in my view the 'assistance' which is referred to in the definition contemplates that form of assistance, which is rendered by an employee, though the person he assists may not necessarily be his employer. In my view, it does not include the assistance of the kind rendered by independent contractors."

Significant groups of workers are either not viewed as employees, or are incapable of exercising their rights as workers, although they are afforded these rights by the labour legislation. It appears that it was these workers that the South African government had in mind when it proposed the adoption of a new presumption as to who qualifies as an employee. The presumption was adopted in 2002.¹⁸

In terms of this presumption, regardless of the form of the contract, a person is presumed to be an employee until the contrary is proven if any one or more of the following factors is present:

"(a) the manner in which the person works is subject to the control or direction of another person;

(b) the person's hours of work are subject to the control or direction of another person;

(c) in the case of a person who works for an organisation, the person is a part of that organisation;

(d) the person has worked for that other person for an average of at least 40 hours per month over the last three months;

(e) the person is economically dependent on the other person for whom that person works or renders services;

(f) the person is provided with tools of trade or work equipment by the other person;

(g) the person only works for or renders service to one person."¹⁹

However, the fact that this is only a presumption means it can be rebutted. Furthermore, the presumption applies only to persons earning below a prescribed threshold amount determined by the Minister of Labour in terms of section 6(3) of the BCEA.²⁰

More significantly, the presumption is not likely to alter the fact that there are still huge numbers of workers who are not regarded as employees and to whom labour legislation does not apply, and workers who are unable to exercise their rights as employees.²¹

Only natural persons can be employees. Any natural person can be an employee, but there are some statutory constraints.²² Juristic persons, on the other hand, cannot be employees but can be independent contractors. The definition of "employee" expressly excludes "independent contractors" from its scope.²³ This makes it necessary to distinguish between the concept of an "employee" and that of an "independent contractor". This distinction is crucial as it has great significance and poses numerous challenges to South African labour law. The Code of Good Practice supports the following distinction between an employee and an independent contractor: an employee "makes over his or her capacity to produce to another", while an independent contractor is someone "whose commitment is the production of a given result".²⁴ The South African courts have developed several tests to distinguish between these two concepts.

The control test was developed in Colonial Mutual Life Assurance Society v McDonald²⁵ where the former Appellate Division had to consider whether an insurance agent was an employee. The court held:

"[T]he contract between master and servant is one of letting and hiring of services (locatio conductio operarum) whereas the contract between the principal and contractor is the letting and hiring of some definite piece of work (locatio conductio opens)."²⁶

Therefore, in terms of this test, the decisive difference between an employee and an independent contractor is that the principal has no legal right to prescribe how the independent contractor achieves the desired result. However, an employer may prescribe the methods that the employee uses and will have recourse against the employee if that method turns out to be inefficient or ineffective. The "control" test has been criticised by Grogan as inadequate if applied in isolation without consideration of the other tests.²⁷

Disaffection with the control test led the courts to construct another test, known as the organisation test. Determining whether a person is an employee, or an independent contractor depends on whether he or she is part and parcel of the organisation. But this test too was dismissed as inadequate by the Appellate Division in S v AMCA Services²⁸ and Smit v Workmen's Compensation Commissioner²⁹ as being too vague to be of any use.³⁰

The inadequacies of the control and organisation tests led the courts to devise a third test, namely the multiple or dominant impression test. This test, often regarded as the standard test used by South African courts, depends on various factors to determine whether a person is an employee or an independent contractor.³¹ The test requires an appraisal of all the "relevant factors". Although it is impractical to put together a comprehensive list of relevant factors, the most important include the authority to supervise - that is, whether the employer has the authority to supervise the other individual;³² the employer's authority to choose who will perform the work; the employee's responsibility to work for a certain period of hours; whether a salary is paid for time worked or for a specific outcome; whether the employer supplied the employee with the tools or equipment to perform the job; and whether the employer has the power to discipline.³³ This test was criticised from its inception by Mureinik,³⁴ who stated that the lack of a clear definition of an employee revealed that the labour statutes occupied "loose and ill-defined ground". According to Mureinik, the "dominant impression test" fails to say anything about the legal nature of the contract of employment and gives no assistance in difficult cases on the cusp between employment and self-employment (which he calls the "penumbral" cases).³⁵To date, the dominant impression test has been applied mainly in the context of the Workmen's Compensation Act, which restricts compensation for injuries at work to employees.³⁶

The Labour Court, in Rumbles v Kwa-Bat Marketing,³⁷ adopted the approach that a contractual relationship is not definitive as to whether a person is an employee as defined; the court must examine the true nature of the relationship between the parties. The issue of whether a person is an "employee" or an independent contractor was also examined by the Labour Appeal Court in SABC v McKenzie,³⁸ where some of the important attributes of the contract of employment and the contract of work were identified and distinguished. The "independent contractor" is bound to produce what is

required in terms of his contract of work. He or she is not under the supervision or control of the employer and is also under no obligation to obey any orders of the employer about how the work is to be performed -the independent contractor is his or her own master. In addition, he or she is used by the employer to provide services to a person through a contract of service rather than through the common-law contract of employment.³⁹ To circumvent the duties of an employer, employers may transform employees into independent contractors or may contract out work previously done by employees.⁴⁰ In practice, the difference lies in the fact that an employee benefits from the protection of labour law while an independent contractor does not.

However, in Niselow v Liberty Life Association of SA Ltd,⁴¹ the Supreme Court of Appeal accepted Brassey's construction⁴² that an employee is a person who makes over his or her capacity to produce for another person, whereas an independent contractor is a person whose commitment is to produce a result.⁴³ The dominant impression test was also found to have shortcomings similar to those we see in other attempts to identify a defining attribute. At present, it is usual practice for employees, especially at the managerial level, to be identified by the outcomes to be achieved.⁴⁴Furthermore, the statutory definition in both the LRA and the BCEA is silent on the point at which a person recruited into employment becomes an "employee".⁴⁵ In Wyeth v SA (Pty) Ltd v Manqele,⁴⁶ it was argued that the term "works for another person" is cast in the present tense in the definition of "employee", and that an applicant, therefore, becomes an employee only when he or she begins working for the employer.⁴⁷ Taking account of section 23 of the Constitution of the Republic of South Africa, 1996 (the Constitution), which affords "everyone" the right to fair labour practices, the Labour Appeal Court adopted a purposive approach and concluded that persons who had signed contracts of employment, but who had not yet commenced work, were "employees" for purposes of the LRA.⁴⁸

In Denel (Pty) Ltd v Gerber,⁴⁹ the Labour Appeal Court considered the following facts. Denel had agreed with Multicare Holdings (Pty) Ltd that Multicare would provide certain human resources consultancy services to Denel. Multicare had one employee, Gerber, and Multicare regularly rendered invoices to Denel. Denel subsequently informed Gerber that her services had been terminated on the ground of redundancy. Gerber argued that she was an employee of Denel, while Denel claimed that it had validly terminated a commercial contract with Multicare, and that Gerber was not employed by Denel.

The court accepted that Gerber was a Denel employee "on the basis of the realities - on the basis of substance and not forms or labels".⁵⁰ Next, the court considered the position of persons who agree to render services through a separate legal entity in order to gain a more favourable tax dispensation. In many earlier decisions, the courts had held that such persons would be precluded from reclaiming employee status for purposes of protection against unfair dismissal.⁵¹ Zondo JP put an end to this line of argument and held that an agreement for purposes of a better tax dispensation does not alter the realities of the relationship.⁵² This notwithstanding, the court also held that, in the absence of reconciliation with the South African Revenue Services, the court had been approached with "dirty hands" and that this would be taken into account when crafting a remedy.⁵³ After defining who constitutes an employee, it is important to describe decent work.

 

3 DESCRIBING DECENT WORK

The International Labour Organisation (ILO) defines decent work as:

"Decent work, the core mandate of the ILO, is defined as productive work for women and men in conditions of freedom, equity, security and human dignity. Decent work involves opportunities for work that is productive and delivers a fair income, provides security in the workplace and social protection for workers and their families, offers prospects for personal development and encourages interaction, gives people the freedom to express their concerns, and organise and participates in decisions affecting their lives and guarantees equal opportunities and equal treatment for all."⁵⁴

Decent work consists of four inseparable, interrelated, and mutually supportive strategic objectives: employment, fundamental principles and rights at work, social protection (social security and occupational safety and health) and social dialogue. Gender equality and non-discrimination are cross-cutting principles of decent work.

In the course of the digitalisation of the work process, the relevance of and need for the Decent Work Agenda have become even more significant amid widespread precarity in the world of work, the circumstances ranging from the financial turmoil and economic downturn of the last two decades to increasing unemployment, casualisation of work, lack of social protection, labour migration, challenges posed by the Fourth Industrial Revolution (4IR) and, most recently, the Covid-19 pandemic. These have challenged countries and social partners - worker and employer organisations - to adopt the Declaration on Social Justice (for fair globalisation) to strengthen the ILO's capacity to promote its decent work agenda and forge an effective response to globalisation and digitalisation of work.

Many aspects need to be addressed if the prevailing standard applied to platform workers in South Africa is to be satisfactory. In February 2015, the ILO held a "Tripartite Meeting of Experts on Non-Standard Forms of Employment", which hosted experts selected after consultation with governments, and the Employers' and the Workers' Groups of the Governing Body, to debate the obstacles to a decent work agenda that precarious forms of employment may present. The meeting mandated member states, employer organisations, and worker organisations to develop policy resolutions to address decent work deficits relating to non-precarious forms of employment so that all workers (regardless of their employment arrangement) could profit from decent work. Specifically, governments and their social partners were entreated to collaborate to: implement measures to address unsatisfactory working conditions; support effective labour market changes; promote equality and non-discrimination: ensure sufficient social security cover for all: promote safe and healthy workplaces; ensure freedom of association and collective bargaining rights; improve labour review; and address highly uncertain forms of employment that do not respect essential rights at work.⁵⁵

In light of the definition of decent work, it is vital to examine the platform economy.

 

4 PLATFORM WORK IN SOUTH AFRICA

The world of work is going digital: an ever-growing number of start-ups are establishing online platforms and mobile "apps" to link consumers, firms, and workers for jobs that often last no longer than a few months - some less than a few days or even hours. What started as a tiny niche for digital "crowdwork" on platforms like UpWork and Amazon's Mechanical Turk has developed into a worldwide phenomenon.

Generally, there are three categories of actor involved in platform work: the client who requests a task; the worker who performs it; and the platform that provides the infrastructure for the transaction. These exchanges can be done locally for tasks that require human contact, or anywhere in the world in the case of online tasks. Moreover, the platform economy can be captured through three main sectors: peer-to-peer transportation (such as Uber); on-demand household services (such as Task Rabbit); and on-demand professional tasks (such as Amazon Mechanical Turk (MTurk)).

Digital platform labour involves the outsourcing of work through internet-based platforms, either to a geographically dispersed crowd (crowdwork) or through location-based apps (location-based).⁵⁶ crowdwork involves advertising specific work tasks that are then performed by eligible workers who are quickest to respond to the advertisement or are contested by eligible workers.⁵⁷ An example of this type of labour platform is MTurk, where a range of work can be advertised - for example, filling out questionnaires, transcribing receipts, or labelling photographs. crowdwork is also associated with online freelancing, where work is given to specified individuals and could involve software development, data analytics, administrative support, or design and marketing.⁵⁸ Location-based apps are related to on-demand work that is geographically specific and are mainly associated with delivery services, transport, and household cleaning services. A common feature of the labour relations governing this type of work is that platform workers are considered independent contractors and are, therefore, not protected by many of the labour regulations covering employees.

Research, in general, underscores that platform workers are rather young (mainly under 30 years of age).⁵⁹ The potential to create job opportunities of high quality for young people in the labour market seems clear; in theory, it is up to the workers to decide the tasks they wish to perform and also when, where, and how to do them. In practice, this sense of flexibility and autonomy is very closely related to the nature of work, a platform's terms and conditions, and whether the work is the worker's main source of income. Reliance on platform earnings varies considerably across states and some workers can work even longer hours than regular employees to fulfil their financial commitments. In certain occupations, the variability and unpredictability of demand can create high levels of insecurity among workers. The European Agency for Safety and Health at Work (EU-OSHA) notes that the similarities between these groups and temporary and agency workers may mean they are exposed to the same psychosocial and physical risks⁶⁰ - and studies have shown that these are significant. Such risks can be even higher for platform workers, in particular, because they are often younger and may have less experience or knowledge of how to manage and counteract them.

This synopsis of digital platform work has provided some insight into the scale and nature of this workforce. Nonetheless, there is still much we do not know. Statistics South Africa's current Quarterly Labour Force Surveys do not accurately represent platform economy workers. Although they are likely to be categorised as self-employed, there is no obvious distinction between platform workers and, for example, professional self-employed individuals or informal workers. Even in tax records, platform workers may be registered as provisional taxpayers, which means that the parent company of the digital platform may not be recorded. Reflecting on worker organisation and the regulation of digital platform work, coupled with concerns about the negative effect of digital and automation technologies on employment, raises important considerations on how new forms of work affect labour organisation and related protection, as well as the capacity to regulate these forms of work. To the extent that the digital platforms raise the demand for the services on offer (such as food delivery), they could be seen to impact job creation. There are, however, early indications that remuneration and the conditions of work may not meet the minimum standards laid out in domestic labour legislation or the decent work guidelines provided by the ILO. Digital platforms have changed the employment relationship and, as a consequence, have undermined worker rights. By categorising workers as independent contractors, enterprises circumvent the rights of workers that are embodied in more conventional employment contracts⁶¹ notwithstanding that the worker is still subject to considerable control by the platform or parent enterprise and his or her performance is central to the enterprise's core business.⁶²

In South Africa, this means that these independent contractors do not benefit from the labour rights in the LRA and BCEA. As a consequence, a major concern is that even if digital platforms become an increasing source of new job creation, they may not comply with the norms of "decent work". As discussed below, early evidence for South Africa suggests an uneven standard of employment across different types of digital platforms. Workers are managed through the online platform (or app), where they clock in by logging into the app and then become subject to an algorithmic assignment of duties to be performed at pre-arranged pay rates.⁶³ Some similarities can be seen in comparison with the regulation of work in the temporary employment sector in South Africa. As noted by Webster and Englert,⁶⁴when the LRA was introduced in 1995 to broaden labour rights to all workers in South Africa, employers took advantage of a legal lacuna to circumvent these new rights.

This took the form of externalising the employment of workers to third-party labour brokers, which makes workers employees of the labour broker and not of the firm at their place of work.⁶⁵ The collective organisation of workers is also challenged by the fragmentation of production, where workers in different parts of the production and distribution chain are divided according to skill levels and type of employment contract and conditions, thus ensuring that these different groups of workers fall under different bargaining councils.⁶⁶ In recent years, the LRAA has strengthened the LRA concerning part-time, temporary, and temporary employment service (TES) workers.

The introduction of section 198A of the LRA has ameliorated the position of TES employees in many ways. For example, should a low-earning employee be placed with a client for a period longer than three months, or should he or she no longer be substituted for an employee of the client who was temporarily absent, the worker will be considered an "employee" of the client. Among other things, this will entitle the employee to refer disputes with the client concerning unfair dismissal and unfair labour practices.⁶⁷

The amendment aimed to extend employees' benefits to temporary and TES workers earning below a threshold amount, while limiting the length of these contracts to three months without the option of continual renewal.⁶⁸There is some evidence to suggest that as a consequence of the amendment, more TES workers lost their jobs with labour brokers and moved to the non-TES sector.⁶⁹ Nonetheless, for the 20 per cent of TES workers who moved to the non-TES sector as a result of the 2014 amendment, it is estimated that they had a greater probability of receiving higher remuneration than workers who were not earmarked by the amendment (that is, those over the earnings threshold).⁷⁰ Accordingly, in this case, early evidence suggests that the regulation had its intended effect on a subset of TES workers, but some workers may have been displaced from formal employment.

Further work is required to understand the extent to which employers can circumvent this amendment and the ramifications for future labour market regulation. There are also challenges to worker organisations in the digitalplatform era and to regulating digital platform labour.⁷¹ It is noteworthy that certain apps are registered internationally and do not regard themselves as directly employing workers in the state in which they operate. For example, Uber South Africa is a technology company whose parent company, Uber, is registered in the Netherlands. In a recent South African case, Uber drivers filed a case against Uber South Africa in an attempt to be recognised as employees. The case was struck down because the court found that the case should have been brought against Uber BV (Netherlands).

In the USA in August 2020, however, a California court ruled in favour of Uber and Lyft drivers by finding that these companies should categorise their drivers as employees with benefits.⁷² In addition, in the recent UK case, Uber BV v Aslam,⁷³ the UK Supreme Court held that Uber drivers are not self-employed or independent contractors as they have the status of "workers". In this light, it can be contended that the South African Labour Court should have adopted a broader approach. It could have explored ways of piercing the legal complexities associated with triangular relationships created by online platforms and placed less emphasis on the existence of a contract of employment by recognising an employment relationship.

While the South African Uber judgment is being contested by the two companies, it represents progress in efforts to have platform workers recognised as essential to the functioning of these technology enterprises.

A further concern relates to worker organisation, which is made difficult by the spatial separation of workers - particularly for those not engaged in place-based forms of digital platform work. Evidence has shown that place-based workers have engaged in a greater variety of strategies to improve their working conditions than is the case with crowdworkers.⁷⁴ Place-based platform workers can engage in strategies that affect their particular jurisdictions and the existing culture of industrial relations.⁷⁵ Webster and Englert⁷⁶ also note how, in the face of traditional trade unions that fail to represent precarious workers in South Africa, these workers are creating new types of organisation in the form of worker committees or councils at the plant level. In South Africa, a legal-aid NGO, the Casual Workers Advice Office, played an important role in the 2014 LRA amendment by offering TES workers assistance in accessing their new rights. There are examples of precarious (location-based) workers changing the nature of worker organisation and relying on different coalitions such as NGOs and broader worker associations.

It is now crucial to explore some attempts, suggestions, and probabilities that may address the precarious situation of platform workers.

 

5 SUGGESTIONS AND PROBABILITIES

Platforms draft their terms of service agreements unilaterally. These terms are largely not limited by labour protection legislation, and may entirely circumvent processes of social dialogue, thus allowing platforms to set all the working conditions for workers. In terms of ILO Conventions and Recommendations,⁷⁷ every worker enjoys universal labour rights, but the question arising is how these rights can be guaranteed to platform workers. Regulating digital labour platforms is complex and involves labour law, other laws, and policies relevant to decent work. Some of the challenges include applying universal labour rights (such as social protection and collective bargaining) to platform workers, ensuring fair competition, fair data use, improved data protection and algorithmic accountability, and reforming the taxation system insofar as it affects platform workers.

There have been attempts in South Africa to address the nature of work in the platform economy. For example, the Fairwork Project has highlighted the precarious nature of work in the gig economy, evidenced by its ratings for digital platforms such as Uber and OrderIn. Fairwork is a collaboration between the University of Cape Town (UCT), the University of the Western Cape (UWC), and the Universities of Oxford and Manchester. The project has evaluated 11 of the country's largest digital labour platforms against five principles of fairness: fair pay; fair conditions; fair contracts; fair management; and fair representation.

Fairwork interacts with workers and worker organisations to develop and continually shape its principles and aims to support workers in collectively asserting their rights. It also seeks to provide consumers with enough information so that they can make informed choices regarding the platforms with which they interact. They hope this will contribute to putting pressure on platforms to improve their working conditions and their scores. Fairwork also engages with policymakers and the government to advocate extending appropriate legal protections to all platform workers, irrespective of their legal classification.

In the words of Howson:⁷⁸

"Decent work and job creation are not mutually exclusive. This is why, by bringing workers and other stakeholders to the table, Fairwork is developing an enforceable code of basic workers' rights that are compatible with sustainable business models."

It is worth noting that Fairwork has been relatively successful in improving standards of living for platform workers. For example, after meeting with Fairwork, getTOD agreed to adopt a living-wage policy that ensures tradespeople do not earn below the living wage of R6 800 per month when working through the platform. It also has clearly defined grievance and disciplinary procedures and guarantees payment to workers.

In addition, NoSweat has created detailed health and safety guidelines to which clients must sign up. It pays workers above minimum wage (after costs) and has committed publicly to recognise and negotiate with a collective body of workers should one be formed. At the same time, SweepSouth scored well in that workers are paid above minimum wage (after costs), and they have introduced initiatives actively to improve conditions for workers, including providing death and disability insurance.

While we applaud the efforts of organisations such as Fairwork, among others, to improve the working circumstances of platform workers, South Africa can also learn from European countries that have introduced other solutions, including new forms of regulation. For example, in 2018, a collective bargaining agreement that allows cleaning workers to be recognised as employees in Denmark was concluded. The agreement, signed between a platform and a union, debunked many myths about platform work, starting from the flawed perception that, by its very nature, it was not compatible with existing forms of labour protection such as employment rights and collective bargaining.⁷⁹

Further, examples of the extension of rights to platform workers are a judicial decision extending safety and health legal standards to platform workers in Brazil, and the "right to disconnect" and to obtain a "decent price" for gig work in France. In 2020, Buenos Aires included digital delivery platforms by modifying the definition of "delivery" in its Transit and Transportation Code, which had regulated the activity of urban couriers or home delivery of food substances since 2016. Also in 2020, the Argentine Ministry of Labour, Employment and Social Security presented a draft bill that places couriers under the protection of labour and social security.⁸⁰Some Chilean platform companies have introduced forms of illness cover for riders unable to continue working. Examples include the creation of an "emergency fund" (equivalent to two weeks' average pay) to support workers infected with Covid-19 and who have been forced to quarantine. However, these positive solutions are fragmentary and not the norm.

Given that platforms also operate across multiple jurisdictions, international policy dialogue and coordination are essential. Social dialogue between platforms, governments, platform workers, and their representatives is likewise vital to ensure that the digital economy becomes a powerful driver for fair competition, decent work, and the advancement of social justice.

It should also be noted that the National Minimum Wage Act (NMWA)⁸¹potentially offers some measure of protection to persons rendering on-demand services on online platforms. The NMWA applies to "workers". The definition of this term does not exclude "independent contractors" (unlike the definitions in the BCEA, the LRA and the EEA). Section 1 of the NMWA defines a worker as "any person who works for another and who receives ... any payment for that work whether in money or kind". Added to this, workers and employees who only work limited hours per day are also protected. Section 9A of the NMWA reads:

"Daily wage payment-

(1) An employee or a worker as defined in section 1 of the National Minimum Wage Act, 2018, who works for less than four hours on any day must be paid for four hours of work on that day.

(2) This section applies to employees or workers who earn less than the earnings threshold set by the Minister in terms of section 6(3)."⁸²

 

6 CONCLUSION

This article has examined challenges facing the decent work agenda resulting from the rise of platform work in South Africa. It provides a backdrop to the South African platform economy, underscoring the fundamental issues of unemployment, inequality, and poverty - in particular for those who find themselves working in the platform economy such as young people and women, the majority of whom are Africans. There are some indications of increasing casualisation of work in South Africa, mirrored by the scale of the informal economy and the rising trend of workers being employed in the temporary employment sector and as platform labour.

Jobs are concentrated in the services sector, while the share of employment in the primary and manufacturing sectors has declined. There is still much to learn about the impact on the South African labour market of the 4IR, which is characterised by innovation, automation, robotics, and digitalisation. Early studies suggest that there is a positive connection between firm-level innovation and employment growth. This includes process innovation, which is related to productivity-enhancing production technologies such as automation.

The evidence from other countries suggests that automation technologies will have an uneven effect on labour displacement and earnings across industries and jobs. The COVID-19 pandemic is also likely to accelerate the use of digital technologies while, at the same time, the related economic decline has already had an uneven negative effect on the labour market. The existing labour force survey data in South Africa does not allow for a better understanding of the types of workers included in these jobs, their earnings, and the quality of their employment.

More refined categories for self-employed workers are needed in order to identify platform workers. The capacity to measure the size of this workforce and the nature of its work will offer greater insight into the need for labour regulation in the digital platform economy. There are many routes for future study in the South African context. First, the literature on the vulnerability of jobs to automation needs to address the issue of which types of automation are economically viable, not only which are technically possible.

As noted by Aloisi and De Stefano,⁸³ the scale of recent technological change may empower capital over labour in ways that make the organisation of workers more onerous in general and exert downward pressure on the value of work that remains available. There are also areas of the economy in which innovation can assist to incentivise growth and decent work creation.

More study is required on the firm- and sectoral-level employment effects of an investment in innovation. The existing national innovation surveys need to be scaled up to cover a wider range of sectors and to be conducted more regularly. There is also much we do not yet know about the size and nature of platform work in South Africa owing to a paucity of data. A study on the categories of workers in this sector, their reasons for engaging in platform work, earnings, working conditions, and capacity to organise would provide a valuable understanding of this rising section of the labour market. In addition, another path of study relates to the regulation of platform-based labour markets. As earlier underscored in this article, labour regulation needs to be better adapted to mirror the underlying exchange in the employment relationship between the platform and the worker, which some labour law scholars, jurists, and organisations (such as the ILO) have already started to examine.⁸⁴

It is worth noting that although trade unions have traditionally been the chief defenders of workers' rights, there is scant evidence that trade unions in South Africa are currently succeeding in adapting their approaches to the digital epoch. In the present circumstances, platform workers are unlikely to be successful if seeking security through trade unions. Their best chance of improving working conditions will be through government intervention and law reform. As in the case of TES employees, fixed-term employees and part-time employees, policymakers will at some stage have to address issues that emerge during the 4IR. State power rather than the power of collective bargaining may yet prove to be the best way to improve the conditions of workers involved in new forms of work. Broadening the scope of application of national minimum wage legislation to cover all "workers" may yet open the door of opportunity for digital workers to secure improved conditions of service.

Without strengthening labour regulations and institutions, platforms workers will not benefit from basic labour rights afforded workers. Decent work for platform workers will remain an illusion if these workers are not treated as employees and if the government continues not to use its power to improve the conditions of these workers.

 

 

1 DG IPOL "The Social Protection of Workers in the Platform Economy" (November 2017) https://www.europarl.europa.eu/RegData/etudes/STUD/2017/614184/IPOL_STU(2017)614184_EN.pdf (accessed 2021-04-12).
2 Heeks "How Many Platform Workers Are There in the Global South?" (2019) https://ict4dblog.wordpress.com/2019/01/29/how-many-platform-workers-are-there-in-the-global-south/ (accessed 2021-03-30).
3 Heeks https://ict4dblog.wordpress.com/2019/01/29/how-many-platform-workers-are-there-in-the-global-south/.
4 Heeks Decent Work and the Digital Gig Economy: A Developing Country Perspective on Employment Impacts and Standards in Online Outsourcing, Crowdwork, etc. Paper 71 University of Manchester: Centre for Development Informatics (2017).
5 Fairwork "Fairwork South Africa Ratings 2020: Labour Standards in the Gig Economy" (2020) https://fair.work/wp-content/uploads/sites/97/2020/04/Fairwork-South-Africa2020-report.pdf (accessed 2021-04-18).
6 Naidoo Innovation, Digital Platform Technologies and Employment: An Overview of Key Issues and Emerging Trends in South Africa. Future of Work(ers) Working Paper No 9 University of the Witwatersrand: Southern Centre for Inequality Studies (2020) 21.
7 Wood, Graham, Lehdonvirta and Hjorth "Good Gig, Bad Big: Autonomy and Algorithmic Control in the Global Gig Economy" 2019 33 Work, Employment and Society 56-75.         [ Links ]
8 Labour Relations Amendment Act 6 of 2014.
9 66 of 1995.
10 S 18(1)(a) of the LRA.
11 S 213 of the LRA.
12 S 1 of the Basic Conditions of Employment Act 75 of 1997.
13 S 1 of the Skills Development Act 97 of 1998.
14 S 1 of the Employment Equity Act 55 of 1998.
15 85 of 1993. S 1 of OHSA defines an employee as "any person who is employed by or works for an employer and who receives or is entitled to receive any remuneration or who works under the direction or supervision of an employer or any other person".
16 63 of 2001.
17 (1996) 17 ILJ 673 (LAC) 683A-B. See also Borcherds v CV Pearce & Sheward t/a Lubrite Distributors (1991) 12 ILJ 383 (IC) where the Industrial Court held that the "assistance" should be rendered with some degree of regularity and there should be a legal obligation to render such "assistance".
18 See s 83A of the BCEA and s 200A of the LRA.
19 Van Niekerk, Smit, Christianson, McGregor, Van Eck Law@Work (2019) 63.
20 The amount is determined from time to time by the Minister of Labour and is currently fixed at R205 433,30 per annum. The presumption has not been included in other labour legislation such as the EEA, the SDA, the UIA, OHSA or the Compensation for Occupational Injuries and Diseases Act 130 of 1993 (COIDA).
21 Theron, Godfrey and Visser Keywords for a 21st Century Workplace (2011) 21.
22 See s 43 of the BCEA which, e.g., prohibits the employment of children under the age of 15 years.
23 See Benjamin "An Accident of History: Who Is (and Who Should Be) an Employee Under South African Labour Law?" 2004 25 Industrial Law Journal 787, where Benjamin (at 789) states that the "terminology of contract is introduced through the exclusion of 'independent contractor'".
24 Item 34 of the "Code of Good Practice: Who Is an Employee?". This description was used in Niselow v Liberty Life Association of Africa of Africa Ltd (1998) 19 ILJ 752 (SCA).
25 1931 AD 412.
26 Colonial Mutual v McDonald supra 433. See also R v AMCA Services 1959 (4) SA 207 (A).
27 Grogan Workplace Law 11ed (2014) 18.
28 R v AMCA Services Ltd 1962 (4) SA 537 (A).
29 Smit v Workmen's Compensation Commissioner 1979 (1) SA 51 (A) 63.
30 The organisation test was rejected on the basis that it also raised more questions than it answered.
31 Basson, Christianson, Dekker, Garbers, Le Roux, Mischke and Strydom Essential Labour Law (2009) 27.
32 Smit v WCC supra 63.
33 Basson et al Essential Labour Law 27; Grogan Workplace Law 20.
34 Mureinik "The Contract of Service: An Easy Test for Hard Cases" 1980 97 SALJ 246 258.         [ Links ]
35 Ibid.
36 Rycroft and Jordaan A Guide to South African Labour Law 2ed (1992) 41.
37 (2003) 24 ILJ 1587 (LC) par 19.
38 (1999) 1 BLLR 1 (LAC) Myburgh JP 590F-591F.
39 Cheadle, Davis and Haysom South African Constitutional Law: The Bill of Rights 18-7. A person who provides services under a contract of work cannot be regarded as an employee save in special circumstances.
40 Du Toit, Godfrey, Cooper, Giles, Cohen, Conradie and Steenkamp Labour Relations Law: A Comprehensive Guide 77.
41 (1998) 19 ILJ 752 (SCA).
42 Brassey "The Nature of Employment" (1990) 11 ILJ 889.
43 Niselow v Liberty Life supra 753H.
44 Benjamin 2004 ILJ 794.
45 S 9 of the EEA provides that ss 6, 7 and 8 of the EEA (which incorporate the principal protections against unfair discrimination) apply to applicants for employment; see Le Roux "Economic Disputes and the New Unfair Labour Practice" (1997) CLL 91.
46 Wyeth v SA (Pty) Ltd v Manqele (2005) 6 BLLR 523 (LAC).
47 Whitehead v Woolworths (Pty) (1999) 8 BLLR 862 (LC); Herbst v Elmar Motors (1999) 20 ILJ 2465 (CCMA) 2468J-2469C.
48 In Wyeth v Manqele supra par 30, the Labour Appeal Court relied on NEHAWU v University of Cape Town (2003) 24 ILJ 95 (CC) and held that the: "LRA must therefore be purposefully construed in order to give effect to the Constitution".
49 Denel (Pty) Ltd v Gerber (2005) 9 BLLR 849 (LAC).
50 Denel v Gerber supra par 22.
51 CMS Support Services (Pty) Ltd v Briggs (1997) 5 BLLR 533 (LAC); Bezer v Cruizer International CC (2003) 24 ILJ 1372 (LAC). In Callanan v Tee-Kee Borehole Castings (Pty) Ltd (1992) 13 ILJ 279 (IC) 1550D-E, the former Industrial Court held that the courts will be unwilling to assist employees who want to "have their cake and eat it". See also Apsey v Babcock Engineering Contractors (Pty) Ltd (1995) 16 ILJ (IC) 924D-F.
52 Van Niekerk ("Personal Service Companies and the Definition of 'Employee'" 2005 26 Industrial Law Journal 1904-1908) argues that parties should be entitled, for whatever perceived advantage, to decide and agree on their own status and designation even if this does exclude the employment relationship.
53 Denel v Gerber supra par 204-205.
54 See Report of the Director General: Decent Work 87 International Labour Conference 1999 https://www.ilo.org/public/english/standards/relm/ilc/ilc87/rep-i.htm (accessed 202104-23).
55 For further details, see ILO "Conclusions of the Meeting of Experts on Non-Standard Forms of Employment" (17 March 2015) http://www.ilo.org/gb/GBSessions/GB323/pol/WCMS_354090/lang-en/index.htm (accessed 2020-12-06).
56 Berg, Furrer, Harmon, Rani and Silberman "Digital Labour Platforms and the Future of Work: Towards Decent Work in the Online World" (2018-09-20) https://www.ilo.org/global/publications/books/WCMS_645337/lang--en/index.htm (accessed 2021-05-16).
57 Heeks Digital Economy and Digital Labour Terminology: Making Sense of the "Gig Economy", "Online Labour", "Crowd Work", "Microwork", "Platform Labour", etc Development Informatics Working Paper Series Paper 70, The University of Manchester Global Development Institute (2017) http://www.gdi.manchester.ac.uk/research/publications/di/ (accessed 2021-04-18).
58 Ibid.
59 See Eurofound Platform Work: Types and Implications for Work and Employment -Literature Review https://www.eurofound.europa.eu/publications/report/2018/employment-and-working-conditions-of-selected-types-of-platform-work (accessed 2021-06-17).
60 See EU-OSHA (European Agency for Safety and Health at Work) (2017), "Protecting Workers in the Online Platform Economy: An Overview of Regulatory and Policy Developments in the EU" (2017) Publications Office of the European Union, Luxembourg.
61 Webster "The Uberisation of Work: The Challenge of Regulating Platform Capitalism: A Commentary" 2020 34(4) International Review of Applied Economics 512-521.
62 Aloisi and De Stefano "Regulation and the Future of Work: The Employment Relationship as an Innovation Facilitator" 2020 159(1) International Labour Review 47-69.
63 Webster 2020 International Review of Applied Economics 512-521.
64 Webster and Englert (2019) "New Dawn or End of Labour? From South Africa's East Rand to Ekurhuleni" 2020 17(2) Globalizations 279-293.
65 Ibid.
66 Ibid.
67 Ss 198A-D and s 187 of the LRA.
68 Cassim The Impact of Employment Protection on the Temporary Employment Services Sector: Evidence from South Africa Using Data from Tax Records UNU-WIDER Working Paper 79 (2020).
69 Ibid.
70 Ibid.
71 Webster and Forrest "Precarious Work: Experimenting with New Forms of Representation: South African Labour Responds" 2020 19(1/2) International Journal of Labour Research 5.
72 Allyn "California Judge Orders Uber and Lyft to Consider All Drivers Employees" NPR (10 August 2020) https://www.npr.org/2020/08/10/901099643/californiajudge-orders-uber-and-lyft-to-consider-all-drivers-employees (accessed 2021-04).
73 [2021] UKSC 5.
74 Johnston "Labour Geographies of the Platform Economy: Understanding Collective Organizing Strategies in the Context of Digitally Mediated Work" 2020 159(1) International Labour Review 25-45.
75 Naidoo "Innovation, Digital Platform Technologies and Employment: An Overview of Key Issues and Emerging Trends in South Africa" Future of Work(ers) SCIS Working Paper Number 9 Southern Centre for Inequality Studies, Wits University (2020).
76 Ibid.
77 The ILO Governing Body has identified eight Conventions covering subjects that are considered to be "fundamental" principles and rights at work: the Forced Labour Convention, the Abolition of Forced Labour Convention, the Freedom of Association and Protection of the Right to Organise Convention, the Right to Organise and Collective Bargaining Convention, the Equal Remuneration Convention, the Discrimination (Employment and Occupation) Convention, the Minimum Age Convention, and the Worst Forms of Child Labour Convention.
78 Howson "Calls for Fair Pay and Fair Conditions in the Gig Economy" (2020-03-20) https://www.news.uct.ac.za/article/-2020-03-20-calls-for-fair-pay-and-fair-conditions-in-the-gig-economy (accessed 2021-04-22).
79 Countouris and De Stefano "Collective-Bargaining Rights for Platform Workers" (undated) https://www.socialeurope.eu/collective-bargaining-rights-for-platform-workers (accessed 2021-04-21).
80 Fierro "Time to Deliver: Decent Work in Digital Platforms" (undated) https://voices.ilo.org/stories/time-to-deliver-decent-work-in-digital-labour-platforms (accessed 2021-04-21).
81 9 of 2018.
82 See s 5(2) of the NMWA and s 9A of the BCEA.
83 Aloisi and De Stefano 2020 International Labour Review 47-69.
84 ILO Work for a Brighter Future - Global Commission on the Future of Work (Geneva 2019). See Aloisi and De Stefano 2020 International Labour Review 47-69.