Abstract
Platform work has attracted global policy attention as part of efforts to extend decent work. Such work is critiqued for fostering insecurity by obscuring employer–employee relationships, but in development contexts it also offers accessible employment, especially for excluded groups. This article examines South African platform workers’ experiences of wages, security, and responsibilities and considers implications for reimagining the social contract. Drawing on 19 interviews with location- and web-based workers, the study highlights the paradox of platform work: it reproduces informality yet engages formal businesses and offers inclusion. Findings reveal how workers view themselves as independent contractors, report generally positive pay, but express concern about risk responsibilities shared between platforms and the state. Their perspectives challenge narrow notions of formalisation as the sole route to decent work. The study prompts thinking about alternative social policy approaches to decent work, especially in Global South contexts with high unemployment and informal employment.
Keywords
Introduction
In the past decade, international development agencies such as the World Bank, the International Labour Organisation, and the United Nations have revived interest in the notion of the social contract, as documented by Plagerson (2023). This renewed engagement reflects recognition of persistent and widening inequalities, which in many Western countries have been exacerbated by the erosion of the post-Fordist social contract, particularly through the rise of “new” forms of work in the platform economy. The platform economy has been widely criticised in the literature for fostering low wages, precarious employment, and inadequate social protection. These dynamics stem from the obfuscation of employer–employee relationships, which allow platforms to “operate outside regulatory and normative frameworks that could provide workers with protections” (Graham et al., 2021: 194). This obfuscation has enabled capital to evade the responsibilities traditionally demanded by the Fordist social contract (Anwar, 2024; MacDonald and Giazitzoglu, 2019; Naidoo, 2020).
In response, many European welfare states have sought to increase formalisation to ensure that platform workers enjoy the same benefits as traditional employees. For instance, the European Union stance on platform work has been to ensure that platform workers receive appropriate recognition to enjoy the benefits that formal workers hold (Behrendt et al., 2019; Chesalina, 2020; Montebovi, 2021). International development agencies have also advocated for the formalisation of the informal sector, though with varied definitions of what this entails (Alfers et al., 2022). Most recently, this notion of formalising informal and non-standard workers was incorporated into the Social Protection outcome document of the Development Working Group of the 2025 G20. The overarching aim of these policy initiatives is to extend protection to informal and non-standard workers and progressively realise the ideals of decent work.
While such responses are underpinned by a positive ideal of expanding work and income security to vulnerable workers, they have gained policy traction in contexts where platform work has expanded against a backdrop of relative labour security and stability. In contrast, these same responses may well be insufficient and possibly inappropriate in development settings where the informal economy predominates and where the standard Fordist social contract, which relies on a formal relationship between employer and employee, has never characterised work for most workers (Alfers et al., 2022; Mosoetsa et al., 2016). In these contexts, the traditional approach to promoting fairer work often fails because it overlooks those pushed to the margins of economic activity (Chen, 2012). Furthermore, the narrow notion of formalisation can undermine the efforts of those working in the informal sector, particularly where the benefits of formalisation do not outweigh the costs of formalising (Plagerson, 2023). Nevertheless, the renewed interest in the social contract provides an opportunity to reimagine its scope. Traditionally, cyclical shifts in power between capital and labour have contributed to ebbs and flows in the advancement of protections for workers. For instance, low wages were a challenge in late 18th-century England, resulting in the payment of wage subsidies to employers to increase wages to workers (Midgley, 2014); and during the early 19th century, the rise of trade unions in many different countries resulted in shifts to pay, working hours, and the introduction of social security (Midgley, 2014) ushering in the Fordist social contract. But as Mosoetsa et al. (2016) argue, many of these gains have largely been accrued in the Global North and to workers who are in traditional, formal employment relationships. There remains a need to promote social policy thinking that accounts for, and learns from, those in precarious and informal forms of labour.
Crucially, the perspectives of diverse informal and non-standard workers – particularly in the Global South where informality is the norm – must be incorporated into this process. Alfers et al. (2022), in their edited volume Social Contracts and Informal Workers in the Global South, assemble a wide range of experiences of informal workers and debate the implications for a reimagined social contract. This article contributes to this growing literature by examining the views of a specific group of informal workers: platform workers in two urban centres in South Africa.
Although platform work is frequently portrayed as a “new” form of employment, scholars have argued that it is characterised by many of the same features as informal work and is, in effect, a new iteration of informality (Hlatshwayo, 2022; Katiyatiya and Lubisi, 2021; Naidoo, 2020). At the same time, platform workers differ in that they engage more directly with the formal sector, working for formal businesses like workers employed through labour intermediaries or as independent contractors. Yet they do also experience work in a new way through algorithmic management, a lack of clarity on who the “employer” is, and pay per “gig” (Katiyatiya and Lubisi, 2021). These complexities complicate how to approach social and labour policy. Furthermore, in contexts with widespread unemployment, platform work may be viewed by policymakers and work seekers as a contributor to much-needed employment growth. It also offers work that is accessible to and flexible for many who are otherwise excluded from the labour market – youth, those with caregiving responsibilities (largely women), people with disabilities, and migrants (Ahmed et al., 2021; Veen et al., 2024). How, then, can the legitimate concerns about how platform work drives work insecurity be balanced with the potential it holds to encourage employment inclusion and growth, and what are the implications for social policy?
Drawing on in-depth interviews with 2 groups of platform workers – 10 place-based workers and 9 web-based workers, this article explores how platform workers experience their work, wages, and security, and considers whether their perspectives might offer insights for reimagining the social contract, as called for by Alfers et al. (2022). In line with Au-Yeung et al. (2025), it foregrounds a worker-centred analysis of how social policy needs to respond to platform work. It answers the question how do platform workers in South Africa experience their work and working conditions and how do they view their risk and responsibilities regarding those of platform owners and the state?
The article proceeds by providing a background to platform work, with particular attention paid to Global South contexts. It then outlines the framing of the notion of the social contract and its applicability in such contexts. This is then followed by a discussion of the widespread literature on platform work and working conditions. The methodology follows. The findings are organised by the aspects of work that the social contract traditionally governs – relationship definitions, roles, and responsibilities; pay; benefits; and security. The discussion draws out the nuanced implications these findings have for reimagining the social contract.
Platform work in the Global South
Platform work is often portrayed as a new form of informal work (Naidoo, 2020) in which workers are engaged in work outside of a formal labour contract. It shares many of the same features of other forms of informal work – lack of a secure contract, unpredictable and often low wages, and lack of unemployment insurance and benefits. The new feature of platform work lies in algorithmic management, which distances the contractor and worker even further. Defining platform work and estimating its size is difficult because it takes on so many forms. Naidoo (2020) distinguishes between web-based platform work in which tasks are loaded onto a platform and eligible workers, regardless of where they are based, bid to conduct the work. Location-based work involves workers “grabbing” jobs on a platform that requires them to be in a particular location to conduct the work. The wide variety of delivery and driver or domestic worker platforms are examples of this type of work.
Fairwork (2020) estimates that over 60 million people are engaged in platform work in the Global South, with the highest concentrations in India, Pakistan, Bangladesh, and the Philippines. Asia’s platform economy is large and growing, with India having around 15 million gig workers (NITI Aayog, 2022), and other key growth countries including Indonesia, the Philippines, and Thailand (Tan and Gong, 2024). In Latin America, Brazil and Mexico lead the region in numbers of platform workers (Zanatta et al., 2024). Estimates suggest over 4.8 million people in Africa were engaged in platform work by 2023 with countries like Kenya, Ghana, Rwanda, Egypt, Nigeria, Tanzania, and South Africa showing growth in the platform economy (Elsley and Snyman, 2023). Fairwork estimated that there were over 130,000 platform workers in South Africa (Fairwork, 2020), of which 30,000 were place based and 100,000 were web-based workers.
The nature of platform work
Since the emergence of platform work, there has been considerable attention paid to how it is shaping the nature of work. A substantial proportion of this literature has highlighted the challenges workers face, including low and unstable incomes (Campbell, 2022), lack of labour protections that lead to poor health and well-being outcomes (Behrendt et al., 2019), and how algorithmic management benefits the platform at the cost of the worker (Rani and Furrer, 2021). Studies show how platform work deepens existing vulnerability and abuse, particularly in the domestic work sector (Hunt and Samman, 2020), and that workers face challenges of inconsistent work and unpredictable wages (Anwar and Graham, 2020; Kavese and Mbali, 2022), no pay at times when they are unable to work, lack of non-wage benefits (Aloisi, 2022; Heiland, 2020; Makó et al., 2020), and inability to challenge algorithmic management (OECD et al., 2023). They also experience job insecurity, and lack of social protection (De Ruyter et al., 2019; Hauben et al., 2020), which is compounded by a lack of collective bargaining power (Hlatshwayo, 2022; Sibiya and du Toit, 2022). Much of this vulnerability emerges from how platform workers are conceptualised as independent contractors and, depending on the country context, are not subject to the same benefits that formal employees are (Katiyatiya and Lubisi, 2021; Naidoo, 2020).
Despite these challenges, there is also research showing some benefits of platform work. Veen et al. (2024) demonstrate how flexible platform work alongside access to social protection in Australia offers a sustainable and accessible livelihood pathway for groups typically excluded from the labour market. Sibiya and du Toit (2022) and du Toit and Phumzile (2024) confirm this finding, noting that flexibility and increased earning potential, particularly for historically excluded workers, were benefits. Within the Global South, one of the most salient advantages of the gig economy lies in its contribution to job creation. In Indonesia, the introduction of the online platform GoJek was initially intended to improve local livelihoods. Over time, however, it has also made contributions to the national economy by generating employment opportunities (Azzuhri et al., 2018).
While there is a substantive body of literature on the experiences of platform work and how these workers are vulnerable, relatively less research has focused on what the social policy responses have been, particularly in Global South countries.
Social policy responses to platform work
The challenges that research on platform work has highlighted point to the need for policy responses. Many European countries reacted quickly with policy responses that ranged from banning of certain platforms for failure to comply with employment regulations (Lenaerts et al., 2017) to ensuring that platform workers could be recognised under different categories of workers to ensure that labour and social protections could be extended to this group (OECD et al., 2023). In these latter cases, where only one client is the contractor, this will obligate that contractor to assume the same responsibility for the independent contractor as they would for an employee (Aloisi, 2022). Yet many platform workers continue to be excluded from social protection. Antonucci (2025), focusing on platform work in Italy, Sweden, and the United Kingdom, argues that this has to do with how platform workers are defined and categorised in social and labour policy and trade-offs between taxation and access to benefits. These are factors to consider as countries respond to the needs of platform workers regarding social and labour protections. In Hong Kong, Au-Yeung et al. (2025) argue that state intervention has been weak, leaving platform workers excluded from occupational welfare benefits.
In countries where the need for employment growth is more pressing, there has been far less research attention focused on how governments have responded or could respond to platform work. In India, some states have made strides towards ensuring that platform workers are covered under unorganised worker social schemes (ILO, 2024). Here, there have been steps towards ensuring that platform intermediaries contribute to these schemes (albeit at a low level). In the Philippines, some steps have been taken to regulate platform work. Despite these responses, the literature on policy responses to platform work in developing contexts remains limited. Most of the literature on platform work in Global South countries that does consider how platform work needs to be regulated focuses on how platform workers themselves organise and make demands (Johnston and Land-Kazlauskas, 2018), which has implications for negotiating improved working conditions. Thus, while there is some emerging literature that speaks to the complexities of negotiating improved labour and social protections for platform workers in Global South countries, there remains a need for research that explicitly seeks to assess the potential for shifting social and labour regulations and renegotiating the social contract.
The social contract as foundational to social policy
The notion of the social contract is a way of examining the rights and responsibilities that are implicit in relationships between the state, citizens, and the private sector. It makes explicit what are often implicit expectations about which actors are responsible for various aspects of meeting the social and economic needs of people. Applied to work, the social contract, and particularly the Fordist notion of the social contract, is usually viewed in terms of formal employer–employee relationships and how the state mediates this relationship to ensure workers are treated fairly.
Although the Fordist notion of the social contract is not characteristic of the realities of most of the world’s workers, it nevertheless continues to influence labour-related social policy thinking. Many have critiqued dominant thinking about the social contract, arguing that it does not suit Global South countries where most workers are in the informal sector (Alfers et al., 2022). They argue that social policy needs to be shaped by heterodox approaches that fundamentally question the dominant notion of the social contract, primarily so it can better account for informal work (Moussié and Alfers, 2022). Informal work fundamentally disrupts notions of the standard Fordist social contract demonstrating that in both informal employment in the formal sector and work in the informal sector the worker is left to carry the risks and responsibilities of work and that the state and employers are able to abdicate their responsibilities (Agarwala, 2016; Alfers et al., 2022; Kenny, 2016; Mosoetsa et al., 2016). Applying the notion of the social contract to the realities of informal workers ensures that the implicit relationships and expectations that have unfolded are made explicit and can be examined. To understand the nature of the social contract underpinning labour relations, three questions must be asked. First, are the parties equally aware of the nature of the contract between them? Clarity on the nature of the labour contract and how the worker is positions (either as employee or independent contractor) determines what rights they can claim and what responsibilities they hold. Second, what are the rights and responsibilities of each party and who is responsible for managing risk? Knowledge about the relationship and what this entails about what a worker can claim against an employer is critical to understanding who carries the burden of risk in the relationship. That is, when something goes wrong – either at work or in life, what can the worker claim from the employer/company? Third, is there fair reward for the work? Fairness is subjective but can be assessed against minimum and living wage assessments in each context. The nature of pay and benefits clarifies the level of responsibility that the employer/company carries towards the worker.
Platform work, labour legislation, and social protection in South Africa
The South African labour market is characterised by high rates of unemployment (33% in quarter 1 of 2025) and, relative to similar middle-income country contexts, low rates of informal employment (13%) particularly in the own-account worker category (Asmal et al., 2024). The formal employment rate of 46% (Statistics South Africa, 2025) is consistent with similar economies. This has led some to consider whether growth in the informal sector, including platform work, could be an answer to the staggering unemployment levels (Asmal et al., 2024).
Driven by strong trade unions during the early post-apartheid era, labour reforms favoured formal sector employees. Labour policy includes the Basic Conditions of Employment Act 75 of 1997 (BCEA), which governs minimum leave, unemployment insurance contributions, workmen’s compensation contributions, stipulated hours of work and overtime pay requirements, and employers’ contractual obligations. Protection includes contributory social insurance to public, state-managed mechanisms – the Workmen’s Compensation Fund and the Unemployment Insurance Fund. In 2018, the National Minimum Wage Act 9 was introduced. Labour legislation is robust. Importantly, it is relatively expansive in its definition of employees. The BCEA defines employee as any person, excluding an independent contractor, who works for another person, or for the State, and who receives, or is entitled to receive, any remuneration, and any other person, who in any manner, assists in carrying on or conducting the business of an employer.
Furthermore, it includes a presumption of an employment relationship if the nature of the work and hours of work are under the control of another person, if the worker has worked for that person or organisation for at least 40 hours per month, if the person is economically dependent on the person or organisation, and if the worker is provided with tools of trade by the person or organisation they work for. Despite this fairly expansive definition, and efforts to extend these protections to groups of non-standard workers, there remains a reliance on employers or contractors paying contributions and reporting unemployment, which continues to exclude groups of workers including platform workers (Katiyatiya and Lubisi, 2021). As such, platform workers, like many other informal workers working via labour intermediaries, have not been able to claim benefits they could be entitled to. Furthermore, aside from some emerging innovation, trade unions typically represent formally employed workers, leaving those in precarious work without organisational affiliation.
Outside of labour legislation, social protection policy has focused, until recently, on those outside of the labour market. Social assistance includes cash transfers to children, people with disabilities who are unable to work, and the elderly. During Covid-19, a temporary social relief of distress grant was introduced. This was the first time that working-age adults could access a cash transfer from the state. Although this grant is still in place while policy debates about basic income support are underway, it is temporary, low in value (~$20 per month), and is only available to people who are receiving no other form of income.
The informal sector, within which platform work falls, largely falls outside of any social insurance or social assistance coverage in South Africa, as it does in many other countries.
Methodology
A qualitative, exploratory research design was used to explore the nature and experiences of platform work in South Africa from the perspective of two types of platform workers – web-based and location-based workers. The location-based platform provided services to a national supermarket chain, engaging motorcycle drivers via the platform to deliver groceries to clients. Drivers are matched with a client who has ordered groceries online. They receive the groceries that have been shopped by in-store shoppers and then deliver them to the client’s address. Clients are charged a delivery fee either per delivery or on a subscription basis, and they can reward drivers with a tip on the application. The application tracks drivers’ delivery times and clients can rate drivers. The web-based platform secures large contracts from a range of public and private sector clients that are primarily aimed at gathering data. These include monitoring and evaluation, compliance, or market research data. The company breaks these contracts down into micro-jobs that it then makes available on a platform for individuals all over the country to “grab.” Micro-jobs are tasks that can be completed in a short period of time (usually less than half a day). Affiliated workers then complete the required task. Sometimes this can take place in their homes but often will require travel to a location nearby (e.g. the local clinic to check that certain medications are available, or the local tavern to confirm that an advertising campaign is implemented). Once the job is completed it is submitted on the platform and quality is algorithmically assessed. Payment for the work is only made if the job passes this quality assessment. Commonalities across both platforms included that work was algorithmically assigned and managed, and there was almost no interaction with human beings in the contracting and work process.
Semi-structured interviews were conducted to enable the asking of a structured set of questions, thereby allowing for comparison of experiences between the groups, while maintaining flexibility for participants to elaborate on their experiences. Flexibility in scheduling was also facilitated by the interview format, allowing participants to be interviewed outside of their working hours. All interviews were conducted telephonically in English and lasted approximately 45 minutes. Topics included understanding of employment status, rights and responsibilities, contracting processes, income and benefits, management experiences, worker protections, and recommendations regarding the nature of their work.
Purposive sampling was employed. The two platforms were selected as typical examples of location-based and web-based platforms to facilitate an assessment of whether experiences varied across these types. Participants were eligible to be sampled if they had worked on the platform for a minimum of 6 months, ensuring adequate experience. Although an even gender distribution had been anticipated, it was observed that all location-based workers were male. There was a more even gender split for web-based workers.
Distinct recruitment strategies were implemented for each platform. For the location-based platform, groceries were ordered by research team members located in different parts of the city, and delivery drivers were invited to participate. Additional recruitment was carried out at shopping centres where drivers typically congregate, targeting those awaiting job assignments. Although this strategy concentrated the sample in Johannesburg and surrounds, this is a major centre of platform work (Heeks et al., 2020) and recruitment occurred across different areas of Johannesburg. For the web-based platform, recruitment was conducted via social media groups used by workers for mutual support. A study advertisement was posted, and interested individuals were contacted via direct message to provide details. As a result, web-based workers were geographically dispersed across the country, primarily in major economic centres and peri-urban areas.
A total of 9 location-based workers and 10 web-based workers were interviewed. An overview of the participant demographics is provided in Table 1.
Participant profile.
All location-based workers were male and predominantly migrants. Many studies have observed that most location-based platform work in the delivery industries are predominantly occupied by men (Fairwork, 2023). There are no clear statistics on immigrant workers in platform work; however, there are trends in which most migrant workers have been observed taking up work in the delivery sector (Elsley and Snyman, 2023). Six male and four female web-based workers were interviewed. Almost all of the web-based workers were based in South Africa. Across the sample, age ranged from 21 to 34 years, with location-based workers having a slightly higher average age than web-based workers. There is limited information about the age and sex profile of most web-based workers.
Audio recordings of participant interviews were transcribed and quality-checked by the researchers. Thematic analysis was conducted deductively on the transcripts. A preliminary code list was developed, drawing from the literature review and social contract theory. Emphasis was placed on codes that (a) showed how workers understood their relationship to the platform company; (b) perspectives on pay and benefits; and (c) perspectives on who carried risks and held responsibility for managing risk. Four team members coded two transcripts and then compared codes, after which a final code list was developed and transcripts split between team members. Coding was done using the agreed-upon code list with opportunities to add codes if necessary. The coding team met from time to time to compare and consolidate code lists and adjust coding where necessary. Once coding was complete, code groups were developed and assessed against the three conceptual criteria for assessing the nature of the social contract and worker views: clarity of relationship, level of income, and experiences of managing risk. Data analysis and management were carried out using the statistical software Atlas Ti®.
Ethical approval was obtained from the University of Johannesburg Faculty of Humanities Research Ethics Committee. Participants were provided with an information sheet detailing the study and were included only upon provision of informed consent. Voluntary participation was emphasised in the information sheet. Participants were informed of the recording and assured of confidentiality. No names are used in reporting the findings, and platform names were withheld to protect participant confidentiality.
Findings
How workers understand their relationship to the platform
Workers were asked how they understood their relationship to the platform. Most of the participants, regardless of whether they were working on the location-based or web-based platform, viewed themselves to be self-employed independent contractors, with one exception. For location-based workers, this perception was driven primarily by the nature of their contract and the fact that they were bearing their overhead costs: I consider myself an independent contractor, because I got my own bike, I got my own data, I use my own phone, my own airtime. What they give me is just an order. I’m an independent contractor. I need to be well equipped. I’m an independent contractor. (Participant 5, male aged 34, GP, location-based) Basically, as I spoke before I did sign that independent contract which means I work for myself so in this case I am actually . . . working for myself because let’s say you don’t work you won’t get paid its more like I am doing business. (Participant 9, male aged 31, location-based)
For web-based workers, their self-identification as an independent contractor was related to a sense of autonomy over the tasks they take on: I consider myself as an independent contractor. I see myself as self-employed yes. I see myself as self-employed because they don’t tell me that I should take this task I voluntarily take part in these tasks so when I find them (the tasks) I take them. I decide whether I want to take them, or I don’t want to take them. If I want to make some money I take a task and that is it. I don’t consider myself an employee of the platform. (Participant 6, male aged 26, GP, web-based) I believe I am self-employed, because If I don’t give myself time, they not gonna run behind me saying “you haven’t logged in, you haven’t done this” so it’s basically up to me, so I believe that I am self-employed, because If I don’t go to the app I’m not gonna do no job, I’m not gonna get no income. (Participant 5, female aged 27, web-based) There’s no employer contract, nothing to bind me to say I’m required to work such hours, I have to do such days. (Participant 9, male aged 25, GP web-based)
Although most of the workers indicated that they felt they were independent contractors, for the location-based workers in particular there was some lack of clarity when probed about the management relationships. Although they felt they were independent contractors, they also noted that they had prescribed hours of work, structured “off days” (unpaid), and a roster for work schedules. From the perspective of South Africa labour law, this kind of management implies an employer–employee relationship, which the drivers were not aware of: I am an independent driver, but they create a roster that from Monday to Sunday you are supposed to be at work from 8-5, 8-5 or 8-6, then you’ve got 2 days off or 1 day off. (Participant 2, male aged 22, GP, location-based) I would say I see myself as an employee because I don’t choose my own working hours. (Participant 7, male aged 31, GP, location-based)
There is therefore a clear sense among participants that they are independent contractors. For web-based workers, this holds true in labour law as they do their work autonomously. However, for location-based workers, the management of hours of work implies an employer–employee relationship that obfuscates the nature of the relationship.
The classification of platform workers under labour law plays a crucial role in shaping what rights they can claim. The prevailing social contract in South Africa rests largely on agreed-upon relationships between formally employed workers and the employee, creating contestation over the rights that workers under other arrangements can claim. Although independent contractors can claim employee rights, if there is evidence of explicit management of work and working hours by the contractor, the participants in this study were not aware of these rights. These findings highlight the complexity of worker self-identification in the platform economy and the need for platform owners and the government legal framework to provide clear guidelines on classification for workers and how this is reflected in contracts. It also demonstrates the need for better information being provided to such workers.
Perspectives on pay and benefits
The overarching findings when workers were asked about income and benefits was that income was very variable, that location-based workers earned far more than web-based workers and in fact earned higher incomes than the average wage in South Africa, and that there were no additional benefits earned.
For web-based workers, income was generally low. The platform sets up what they term “micro-jobs” – jobs that should take anywhere from 5 minutes to 1 hour to complete. The pay per job then is also very low ranging from R5 to R100 per task: Okay, like I said they . . . they differ depending on which task that you are doing. For example, taking pictures of the spaza that I told you about it was R5.00 and someone can also go verify that same spaza for R30.00. Then the other ones (other tasks) they pay R100.00 per task then the construction site they will pay you maybe R75.00 depending on which ones you are doing, but they differ. (Participant 2, female aged 22, GP, web-based) Basically, it depends on the specific type of job, because basically the platform has sponsors who gives out the job that the platforms give us. So sometimes there are jobs where they tell you that it’s R75 per job done. Sometimes they (the platform) tell you like you should go to promote somewhere where you set on your location, and they pin you to see if you will be there. For those maybe they said 5 hours, and then when you are there for the 5 hours, they will tell that maybe it is R30 or R40 per hour. If you are there for 5 hours you get your R40 times five or if your there for 3 hours the moment you leave, they tell you they have found that you are no more there, you have worked so and so hours you are gonna get so and so. Basically, as I said it depends on how eager you are, the more jobs you take the more money you make, the less jobs you take less money you make. (Participant 5, female aged 27, Mafikeng, web-based) The most I got (per job) was R40 so it differs. It differs with the task, some are R60. I think sometimes per month there isn’t really good pay per task, but just the tasks that are always there. Sometimes you find there are many tasks, sometimes they are too far away for you to complete. (Participant 7, male aged 28, GP, web-based)
Although the income was low, when worked out on a per hour basis, participants were being paid more than the minimum wage – a finding that is confirmed by Fairwork (2020). However, income was also variable dependent on task availability; whether tasks were accepted or rejected by the workers; and whether the work was deemed of sufficient quality to be paid. Algorithmic management therefore played a strong role in determining pay – both in terms of quality of work submitted and in terms of tracking how long a worker was at a work location. For most of the web-based workers, the pay was insufficient and too variable to rely on and was rather used as a “side hustle” – a source of income to supplement other livelihoods.
For location-based workers, earning was much higher. Responses to the question on income earned per month ranged from R5000 to R19200 (~$300–$1158). These reported earnings are in line with or far higher than median monthly earning in South Africa of R5 200 (~$313) (Statistics South Africa, 2023). This suggests that for location-based workers, there was potential to earn good wages.
However, for both web-based and location-based workers, earnings were undercut by the need to provide their own means of completing the work. For web-based workers, they had to have a smart device and data to access and upload tasks. They also had to travel to locations at their own cost. For location-based workers they had to have their own smart devices and data and either provide or pay off their motor vehicle: OK, first running this business. You need to have a bike. The bike is, it might be like a rented or help you to on your own bike. Those are the costs. You need to buy your own petrol. You need to buy your own data. Yeah, those are the main things. If you are renting a bike, you have to pay each and every week for the bike or on a normal basis. It’s R600 per week. Others they do rented bikes, I heard some of them pay R800 per week and they’ll complete that bike after 18 months. Others have got their own bikes. Others like me have got the platform bike, of which I am paying R273 and some other cents per week, so I am renting this bike to buy. They do deduct that money and then after 14 months the bike will be yours. (Participant 9, male aged 29, GP, location-based) My costs is just only the petrol and the service then that’s it, nothing less.
OK and do you have to buy airtime or data to be on the app or the platform provides?
Yes you have to buy on your own, company doesn’t give you money for . . . you have to take from your own pockets. (Participant 3, male aged 21, GP, location-based)
I think they have a partnership with Cell-C but I am on Vodacom so I have to use my own data. So, for most of these tasks you have to travel to a Spaza shop, or to Pick’ n Pay or Spar. You have to use a taxi, sometimes you can walk to the location so you don’t incur cost but there is a travelling cost either than that I don’t think there is another cost yeah so data and maybe taxi fare. (Participant 8, male aged 27, KZN, web-based) I use my own money to travel and also to buy data. The platform does not provide me with any transport. (Participant 10, female aged 26, GP, web-based)
All of the workers therefore carried the costs of tools they need to do the work. For location-based workers, these costs were higher owing to the need to have a functioning motorcycle and fuel as well as data. For web-based workers, the costs were lower – primarily data and travel costs. Even so, both data and transport in South Africa are among the highest costs that households carry and are particularly high for poorer households when considered as a proportion of household income. Location-based workers did benefit from tips that were added by clients. This is not a cost that the platform company carries, but rather one that is determined by the relationship between the driver and client, often based on a rating of service. Drivers calculate the tips into their overall pay. No other benefits were paid by the platform companies, and it was clear that no leave pay was allowed.
Risk, responsibility, and perspectives on social protection
Even though most workers self-identified as independent contractors and understood that they did not receive any benefits, they did feel the burden of carrying the risks associated with their work. None of the workers reported that unemployment insurance was paid on their behalf by the companies. They did however note that accessing worker protection, particularly income protection and insurance, was very important. Some indicated access to private insurance, but for most this was unaffordable: It (insurance) is very important, because those things they help. For instance, now public hospitals are full of people of COVID, but if you have a medical aid, you just go to private hospital there and just get your thing sorted, and UIF you can just claim it when you are unemployed, so those are some important things somebody must consider in life, even investments. (Participant 7, male aged 28, GP, web-based) No, no work, no pay. That is why I pay for my own policy. Nobody knows tomorrow. I am in this country without any family members. These policies are important because lawyers are expensive and when I start my business, I will be in need of one. I am just getting ready and because I am a husband and a father, if anything happens to me and I die, I don’t want other people to have the burden of taking care of it. (Participant 5, male aged 31, GP, location-based)
Although the web-based platform did not make unemployment insurance contributions, it did offer other forms of insurance through private service providers. For instance, when workers had to travel to construction sites to verify signage, they were covered by a privately purchased insurance that the company had purchased, as these were seen as risky sites to enter: Like we have the . . . what do you call it? The insurance cover I think if you are willing to take it, it is free of charge I think if I remember well it is free of charge to take it so you confirm to take part in construction sites to visit construction sites and then that insurance if it happens that you get injured or get hurt whilst you are there you are covered by that insurance but I have never taken it. (Participant 6, male aged 26, GP, web-based) Yes, there is like accident cover you know if you on a task and you have to travel so they do provide you with accident cover, so you have to fill in your details here, register for the accident cover. It’s free to register.
Oh, okay so you don’t pay anything for that?
No, it’s on behalf of [the platform] and [private insurance company] you don’t pay anything for it but let’s say you were travelling to the job, and the taxi had an accident so they will cover maybe R5000.00 of your medical expenses. (Participant 8, male aged 27, KZN, web-based)
Taken together, the findings above demonstrate that, like other platform workers in previous studies (De Ruyter et al., 2019; Hauben et al., 2020), none of the respondents reported receiving benefits like unemployment insurance contributions or leave pay – benefits typically afforded to formal employees. Yet for the workers, having this kind of insurance or protection was seen as important. Some did purchase private insurance and in the case of the web-based platform there were efforts on the part of the platform itself to offer insurance through private mechanisms.
Discussion and conclusion
The rise of platform work has required policy makers to recognise how the provision of social protection via the labour market, with a primary reliance on formal labour market relationships, continues to place workers at risk. The Fordist model of work has been in decline in Europe and other Global North contexts for some time and has never been a defining feature of work in the Global South. Yet, social protection systems continue to rely predominantly on occupational citizenship as the access point. This is true also in the case of South Africa. Understanding the perspectives of platform workers, as one group of informal workers, offers an opportunity to centre worker perspectives in considering social policy responses, particularly regarding income, how income can be protected, how risk can be distributed, and how access to protections can be mediated.
In the theoretical framing of the study, three aspects of a more just social contract were proposed: clarity of relationships between workers and platforms to ensure all parties are clear on their rights and what they can claim, fair share of rewards, and shared responsibility for risk. In this section, the findings against these three criteria are summarised, before a discussion about the implications for thinking about the social contract and how that might shape global labour and social policy is presented.
The findings demonstrate that all but one of the platform workers interviewed understood clearly that they were independent contractors. This clarity emerged for web-based workers because they managed their own time and tasks. For location-based workers, this was primarily because they bore their own costs and provided their own tools of trade. However, these workers did also mention that they were directly managed with clear rosters and predetermined, unpaid off days that led to some ambiguity. Nevertheless, there was a high level of clarity on the relationship with the platform, meaning that they also clearly understood that they were not entitled to claim leave pay or unemployment insurance. However, the location-based workers, for whom the management relationships might suggest an employer–employee relationship under South African labour law, were unaware of their rights in this regard.
Income was varied, although did seem to be above the minimum wage for web-based workers, and above the median wage for location-based workers. Web-based workers were earning less than location-based workers, primarily because of the nature of the micro-tasks they were taking on and the availability and take up of tasks. Nevertheless, most of the workers felt that they were receiving fair to good income per task (for web-based) and per month (for location-based workers). However, this was undermined by the costs of work and, for the web-based workers, the variability of work. For web-based workers, income from this work was viewed as additional to other livelihoods rather than as something that could be relied on. The findings for location-based workers contradicts much of the literature on income in the platform economy, which shows low and unpredictable incomes (Anwar and Graham, 2020; Campbell, 2022), but does confirm literature on the platform economy in South Africa’s domestic work sector, which shows that income is quite good on many of the platforms (du Toit and Phumzile, 2024; Sibiya and du Toit, 2022). This may in part be explained by conditions of high unemployment and low-waged employment in sections of the formal sector, suggesting that variable income, if high enough, is viewed more positively than lower-wage, regular pay and is certainly preferable to unemployment. It also suggests that some platforms may be taking responsibility for ensuring that income is fair, based on some form of measurement. This is an argument shared by Fairwork (2020), which assesses platforms based on whether they pay minimum or living wages and notes several platforms operating in South Africa that are committed to ensuring that income is at least consistent with living wages.
Shared responsibility for risk is where there are gaps in ensuring a socially just social contract. Most of the workers raised concerns about the risks they carried, which extended to losing the means to earn their livelihood, through to being unable to ensure that their families could manage risks if they passed away or were injured. This confirms much of the research on vulnerability of platform workers (Aloisi, 2022; Behrendt et al., 2019). Some workers had responded by purchasing their own insurance. For web-based workers, the platform had responded by purchasing insurance that would cover workers when they were engaged in more risky work. In the former example, the workers have taken on the individualisation of risk and accepted to manage it themselves (Hamilton, 2016), freeing platform companies from the costs of these risks. In the latter, the platform does demonstrate some commitment to sharing the responsibility via private means, and as such demonstrates that there is willingness to explore how to share responsibility for risk more equitably, confirming work conducted by Fairwork (2020) on how platforms are responding to calls for more equitable share of responsibility for risk. However, this is the area requiring most policy work, and the evidence suggests that workers and platforms are trying to address this issue in the absence of guidance or involvement by the state.
The findings above demonstrate that there is certainly room for platforms to be operating in developing country contexts, contributing to employment growth and opportunities, particularly for typically excluded workers. However, there is also a need for states in these contexts to be more actively driving a socially just social contract that responds to the concerns and needs of platform and other kinds of informal workers. There is also room for regional groupings, such as the Southern African Development Community or the African Union to be collaborating on policy responses given similar economic contexts. Such voices are critical in the policy debates about how to achieve decent work for all, particularly in the face of calls for formalisation of informal work, which, while well-meaning, could undermine the growth of much-needed work in the informal sector.
The current dominant form of the social contract, which continues to emphasise the standard employment relationship, is one option for including platform workers. This is the predominant approach that has been followed in European countries and is one that has strong influence in global social policy thinking, particularly because it does align with ILO decent work approaches to formalising the informal economy. However, platform work demonstrates once again that the idea of the standard employment relationship has never and still does not fit most working arrangements in developing contexts and that the idea of formalising informal work may be, at best idealistic and at worst, undermine employment growth in such contexts.
As such, there are opportunities within global social policy debates to learn from the experiences of developing countries and from platform workers themselves, and to think more expansively and creatively about a more just and inclusive social contract that reaches more non-standard workers. This article has demonstrated that there is room for non-standard work where a living income is paid. Mechanisms for setting and assessing whether incomes meet such standards will be important and there is a need to learn from platforms that are seeking to ensure a fair or living wage is paid (Fairwork, 2020). There is certainly room for states to be involved in such debates and informing such thinking. Second, ensuring that there is relationship clarity and that workers are aware of the rights they can claim as independent contractors is critical. Information campaigns by states or worker organisations are important in this regard. Finally, while there may be a role for private insurance for platform workers, this should not be at the expense of platforms taking responsibility for the costs of risks, and states and worker collectives negotiating on how to include more non-standard workers into state-managed, contributory social insurance mechanisms. Many developing countries are experimenting with how to do this and there is room for peer-learning and triangular cooperation in promoting more expansive social insurance systems that meet the realities of work in developing countries. In so doing, there is potential for developing countries to inform more expansive thinking about how to promote decent work beyond the formal labour contract. Ultimately, the aim must be for all workers to access work with greater security and stability through, at a minimum, clarity of relationships, fair share of reward, and shared responsibility for risk.
Footnotes
Acknowledgements
The authors would like to acknowledge critical review feedback received from the anonymous reviewers, the Just Society group of researchers as well as the CSDA Writing Group members. Eleanor Ross copy edited the document. Grateful thanks to the participants in the study. Hannah Dawson and David Du Toit critically reviewed research instruments and analysis processes. In addition to the authors, Demarco Pillay, Divine Katay, and Lindokuhle Maswana assisted with data collection.
Ethical considerations
The study received ethical approval from the University of Johannesburg, Faculty of Humanities Research Ethics Committee. All participants were guaranteed confidentiality, and their names were changed in all reports and articles emanating from the study.
Informed consent statements
All participants received full information about the study and the voluntary nature of their participation. This was provided in writing and discussed with potential participants. All participants consented in writing to participate.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The study was funded by the Department of Science, Technology and Innovation and the National Research Foundation via a grant made by the Centre of Excellence in Human Development located at the University of the Witwatersrand under grant number ACC2020-4IRHD-2.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data availability statement
Deidentified transcriptions are available upon request.