Locked in, logged out: pandemic and ride-hailing in South Africa and Kenya

Abstract

This article examines the impact of the pandemic on ride-hailing drivers and their mitigation strategies during lockdown in Africa. Ride-hailing has emerged as one of the latest paid-work opportunities for the continent's many unemployed. Yet, ride-hailing companies such as Uber and Bolt misclassify drivers to avoid regulation and responsibilities towards workers’ welfare. Drawing on 34 in-depth interviews with ride-hailing drivers, driver representatives and trade unions in South Africa and Kenya, this article makes two arguments. First, the gig economy in Africa provides work opportunities for the unemployed on the continent and simultaneously vitiates the working conditions through the commodification and informalisation of work. Second, the state-directed emergency measures act as a veneer to capital's efforts to commodify labour and the gig economy platforms have emerged as primary tools for it. Our account points to an urgent need for better regulatory systems to hold platform companies accountable and a collective bargaining mechanism in the gig economy.

Uber sees you as a machine that makes money for them. (Uber driver in Cape Town)

INTRODUCTION

From Washington DC to New Delhi and from London to Cape Town, the gig economy platforms (henceforth platforms) such as Uber and Bolt are connecting drivers with customers for taxi rides. This is commonly referred as ‘ride-hailing’ and has emerged as one of the alternative forms of public transport. Ride-hailing provides income-earning opportunities for the unemployed. Yet, it is also fraught with risks as it offers fickle livelihoods and poor working conditions (Anwar and Graham 2022, 2021, 2020a).

The International Labour Organisation (ILO) estimates that over 80% of the employment in Africa is in the informal sector, the highest proportion in the world (ILO 2018). Informal workers (e.g. waste collectors, domestic help, migrants, artisans, street vendors) exist in the most precarious conditions, i.e. low pay, insecure contracts, lack of social protection, and have poor occupational and health standards at workplaces (ILO 2018). The Covid-19 pandemic has made their situation worse. They face an unimaginable choice between hunger and infection during the pandemic. Various African governments extended the cash-transfer programmes and introduced new emergency grants to mitigate the loss of livelihoods (see KPMG 2020 for a global coverage on government responses around the world). However, these have proven inadequate as many, including ride-hailing drivers (henceforth drivers), remain outside the ambits of these measures. With this in mind, the article sets out to examine the implications of the pandemic-induced lockdown on drivers in South Africa and Kenya, two of the biggest markets in Africa in terms of the number of drivers.

This article contributes to the emerging ride-hailing literature in the following ways. There are several country-specific studies on ride-hailing in high-income countries such as the UK (Woodcock 2020), the USA (Ravenelle, 2019) and Australia (Veen et al. 2020). There is also some literature on low-income countries, e.g. Brazil (Amorim & Moda 2020) and Indonesia (Raharjo & Febrianto 2019). But so far, little attention has been paid to African countries (exceptions include Carmody & Fortuin 2019 and Iazzolino, 2021). This article, therefore, not only brings the African perspective to the emerging ride-hailing research but also presents one of the first empirical accounts of the pandemic's impact on ride-hailing in low- and middle-income regions. Also, despite gig economy platforms touting their potentials for flexible work opportunities, empirical research has shown increasing commodification of work and precarious working conditions in the gig economy (e.g. Ravenelle, 2019; Anwar and Graham 2021, 2022). This article adds to this literature by showing that the pandemic has further amplified gig workers’ vulnerabilities, in this case drivers, particularly those with migrant backgrounds (an aspect relatively less-studied so far in the gig economy literature), especially in regions such as Africa where social protection measures are inadequate.

The article uses a mixed methods approach: in-depth interviews and digital ethnography. In-depth interviews with drivers, local driver representatives and local trade unions in South Africa and Kenya were conducted between April and June 2020. It also employs a novel approach of digital ethnography of big data (Wang & Liu 2021), which involved collecting data on drivers’ communications in a WhatsApp group on a daily basis. Drawing on these two data sources, the article makes two arguments. First, it argues that the gig economy offers new paid-work opportunities to the African workforce, but it also vitiates the working conditions through the commodification and informalisation of work – the effects of which have been devastating especially for certain types of drivers, such as migrants. Second, the state-directed temporary emergency measures during the pandemic act as a veneer to capital's efforts to commodify labour via the gig economy platforms. It briefly discusses drivers’ mitigation strategies to provide a narrative of resilience and solidarities. In the Conclusion, we outline the need for a better regulatory system that holds platform companies accountable and allows collective bargaining among the gig economy workers, especially in the low- and middle-income regions, to improve their material conditions.

RIDE-HAILING IN AFRICA AND LABOUR RELATIONS

Ride-hailing has grown tremendously in Africa in the last five years with an estimated 81 platforms on the continent (Insight2Impact n.d.). Though several local platforms have emerged, e.g. Oga Taxi in Nigeria, Safe Boda in Kenya, and Hailer in South Africa, international platforms such as Uber and Bolt dominate the market share in Africa. Uber currently operates in 24 cities across nine African countries, while Bolt has operations in 64 cities across the continent. Figures on the number of drivers on these two platforms remain sketchy at best.¹ Some estimates from 2018 suggested 216,000 workers in the ride-hailing sector in just seven Africa countries (Insight2Impact 2018). In South Africa, Uber has an estimated 13,000 active drivers (Walker 2019). In Kenya there are approximately 6000 active Uber drivers (Macharia 2017). However, the number of people using platforms to seek work may be higher as drivers often rent and/or share a car between two people. The growth of ride-hailing platforms in Africa is significant for several reasons.

The structural adjustment programmes over the past few decades effectively threatened workers’ rights on the continent (Barchiesi 2012, 2017), but various forms of social welfare policies were also introduced by some governments (Webster et al. 2009). In the cases of South Africa and Kenya, the adoption of neoliberal economic policies was often complemented with various social protection measures such as cash grants (Webster 2005; Bond 2014; Mati 2014; also Ferguson 2007).² However, the situation for workers has remained dire in both countries. In South Africa, job losses have been a regular phenomenon and the unemployment rate in the first quarter of 2020 increased to 30.1%, one of the highest in the world (Statistics South Africa 2020). While unemployment rates in Kenya are lower than South Africa, over 83.6% of employment is in the informal economy (Kenya National Bureau of Statistics 2019). According to the ILO (2015), the informal economy refers to all economic activities by workers and economic units that are in law or in practice not covered or insufficiently covered by formal arrangements. This excludes illicit activities. Informal employment includes own-account workers, contributing family workers, and employees whose jobs are not covered by national labour legislation, income taxation, social protection, or entitlement to employment benefits (paid annual or sick leave, etc.) (ILO 2015). Though these definitions are juridically used both in South Africa and Kenya, the informal economy may be bigger since many economic activities remain invisible (see Rogan & Skinner 2019).

Consequently, platforms are now ostensibly seen by some as a fix for joblessness and informality on the continent (Anwar & Graham 2021, 2022). The co-founder of Lynk (a ride-hailing company), Adam Grunewald has said that ‘the potential for platforms to provide a source of consistent work and centralised governance and support is even more exciting in Africa where platforms can leapfrog informal economies’ (Grunewald 2019). However, we argue that the gig economy is based on the commodification of work (i.e. gig or tasks based) (Graham & Anwar 2018a, 2018b), and gig workers get hired on conditions similar to informal employment, hence they face poor working conditions (Carmody & Fortuin 2019; Anwar & Graham 2020a, 2021, 2022). This is not to suggest that precarious working conditions are not present in the formal sector. In fact, there is a large body of work which has argued that boundaries between the informal and formal sectors are often blurred (see Meagher 2016; also Chen & Carré 2020) and working conditions in the formal sector in Africa can be precarious (Skinner 2006; Neves & Du Toit 2012).

Furthermore, ride-hailing is symbolic of asymmetric power relations between drivers and platforms. To give one example, platforms algorithmically control the labour process (see Gandini 2019) via the mechanism of ratings. Drivers are rated by users after every ride and if their ratings drop below a certain threshold, they are blocked by the platform (Möhlmann & Zalmanson 2017). Algorithmic management has been shown to adversely affect job quality among workers (Woodcock 2020; Anwar & Graham 2022). Additionally, perhaps one of the most contentious issues at the heart of ride-hailing is the labelling of workers as independent contractors and not employees (De Stefano 2016). Because drivers are considered by platforms as ‘independent contractors’, they do not qualify for a minimum wage, sick pay, right to unionise or join a union.³ This affects workers’ collective bargaining power which is generally available to employees in other workplaces (e.g. factories) (Aslam & Woodcock 2020). Therefore, the gig economy's contribution towards workers’ lives and livelihoods can be contradictory, especially in the African context.

On the one hand, platforms offer economic opportunities (e.g. paid work) to some of the workers on the continent. On the other hand, these opportunities come with various constraints (i.e. lack of social protection, low pay, surveillance) (Anwar & Graham 2022). To make matters worse for African drivers, the COVID-19 pandemic has exposed them to new risks. Not only is the threat of infection real, but lockdown also threatened their livelihoods (Anwar 2020; Otieno et al. 2020; Anwar et al. 2022). Here, we would like to argue that the already existing socio-economic conditions (e.g. rampant poverty, unemployment and a lack of welfare support measures) in which African drivers are already embedded means that the pandemic has further amplified their vulnerabilities. To put it differently, the current pandemic has acted as an external shock to the system to exacerbate multiple risks for drivers. In the remainder of the article, we discuss these implications, but first we turn to methodology.

METHODOLOGY

The study is based on a mixed methods approach. We conducted in-depth interviews with 23 respondents who included drivers, local trade unions representatives and informal ride-hailing driver organisers in South Africa and Kenya, between April and June 2020 (see Table I). Drivers used multiple platforms with Uber or Bolt as their main platform, along with a few domestic platforms. Drivers we interviewed were based in Nairobi, Johannesburg and Cape Town. Nine drivers in South Africa were international migrants from various African countries, mainly Zimbabwe, Somalia and Zambia. In Kenya, all drivers were internal migrants from counties such as Siaya, Nakuru and Nyeri.

Table I. Characteristics of respondents

All interviews were conducted over the telephone due to social distancing measures. Telephonic interview is a known research method with various opportunities (e.g. flexibility and access) and also challenges (e.g. quality of the voice call and not being able to understand the gestures and emotions) (Block & Erskine 2012). Some participants were interviewed several times primarily due to time constraints and also to discuss new developments (e.g. relaxation of working hours; new initiatives from platforms). This afforded the opportunity to engage with drivers over a longer period and get a sense of how they are impacted by the pandemic. It resulted in a total of 34 interviews. The interviews lasted between 25–120 minutes. Interviews were transcribed and then coded on NVivo for key themes such as loss of livelihoods, risk mitigation strategies, drivers’ discontents, and collective mobilisation and organising efforts.

We also employed a novel approach of digital ethnography of social media data (Wang & Liu 2021). We collected collecting workers’ communications in a WhatsApp group on a daily basis. Workers’ use of digital communications channels has emerged as an important tool for them to mitigate adverse working conditions found in the gig economy (Gray & Suri 2019; Anwar & Graham 2020a). There are possibly dozens of WhatsApp groups of drivers in South Africa and Kenya. Getting into these groups as a researcher is not only difficult but also presents a moral and ethical challenge (see Barbosa & Milan 2019). A ride-hailing driver in South Africa let one of the authors into their WhatsApp group of about 200 drivers in Cape Town in early May 2020. Upon joining the group, the author sent a text to the group members about the aims of the study. The group members thought the researcher was acting as an agent of Uber and they were concerned about them being blocked from the platform. The backlash from the drivers meant that the group administrators had to remove the researcher. Later in May 2020, the same researcher also joined another WhatsApp group of Uber drivers in South Africa, facilitated by one of the participants, who organised a phone call with six administrators of the group. The administrators were informed about the study and its objectives. We informed them about the need for our participation in the group and that none of the text, audio or videos on the group will be used or quoted in the publications. They then circulated a message to the whole group introducing us. This ensured trust building with group members, and allowed the researcher to engage with drivers regularly and stay in the group. At the time of writing, the group had close to 150 members. As Barbosa & Milan (2019) have argued, obtaining consent of every participant in a group with hundreds of members is practically not feasible. A better approach, according to Barbosa & Milan, is gaining consent through an ongoing negotiation, transparency and full anonymisation. Practically, this meant that we would send messages on the group regularly as a reminder of our research, objectives of the study, use of their communication data and reiterating privacy attributes including complete anonymity.

While the researcher's interaction with this WhatsApp group continues, only the messages between the months of June–July 2020 (the time period of strict lockdown in both countries) were extracted and exported to an Excel sheet. All identifying information such as mobile numbers and names were removed to completely anonymise the data. This is a well-established research practice among scholars studying data extracted from social media channels (Barbosa & Milan 2019). In all, we collected 4,826 messages which included 3,798 text messages while the rest were either voice notes, videos or pictures. We only coded the text messages but not the other forms of media circulated on the group, since some files circulated included audios, photos and videos (in one case a short 10-second clip of a dog watching a football game). Text messages were openly coded to identify a range of themes emerging in the conversations and then later grouped together to highlight the key categories of communication between drivers. No text messages are quoted in the article and names of respondents are pseudonymised for privacy reasons. We now highlight the impact of the lockdown and mitigation strategies of drivers.

COMMODIFICATION OF WORK AND THE LOSS OF LIVELIHOOD

Research on the gig economy has already highlighted the commodification of labour through platforms (Anwar & Graham 2020a, 2021, 2022). A combination of a rating system and algorithmic management allows platforms to treat drivers as a commodity that can be easily replaced (Cant 2019; Woodcock 2020). Ride-hailing platforms also do not recognise workers’ employee status and treat them as independent contractors (or self-employed) (De Stefano 2016). Hence, drivers could be considered to be part of the informal sector. It is worth noting here that there already exists an urban para-transport sector in both South Africa (e.g. minibus taxis) and Kenya (e.g. matatus and boda boda). But the ride-hailing case is somewhat peculiar. The processes of onboarding drivers on platforms include several formal processes and channels (e.g. financing for cars, identification document, driving licence) (see Carmody & Fortuin 2019; Pollio 2019). Drivers also have to pay taxes (e.g. in Kenya through digital tax) and can access banking services (e.g. Kenyan drivers are able to get loans for cars to be driven for Uber). Put differently, ride-hailing operates at the intersection of the formal and informal economies. Ride-hailing companies also threaten drivers with removal from platforms if they engage in collective bargaining (Aslam & Woodcock 2020). Drivers we interviewed told us that Uber's service-level agreement states that Uber does not recognise collective bargaining. While this makes ride-hailing work a bit risky and volatile, the pandemic has been a double whammy for African gig workers.

The pandemic-induced lockdown went into effect in Kenya on 25 March 2020 and in South Africa on 27 March 2020. Most economic activities closed down and only essential services (e.g. healthcare, pharmacies, food markets, public transport) were allowed to function, but under restricted working hours. The United Nations Development Programme (UNDP 2020) estimated a pandemic-induced income loss of more than US$220 billion in low- and middle-income countries. The losses were particularly pronounced for those who cannot work from home such as drivers.

Drivers told us that they were getting few trip requests during lockdown.⁴ One of the drivers, Jim in Nairobi, was doing about 90–95 trips on average per week between January and February 2020. But during March he only averaged about 30 trips per week. He said, ‘you might end up not getting any trips at all some days’. This narrative was repeated by others in our sample. One driver in Johannesburg told us that the lack of trips is because during the lockdown people are not travelling to places such as entertainment spots, restaurants, cafes, etc. In another example, Rob in Nairobi used to get 12–14 trips a day before the pandemic but in April–May he was averaging about 2–3 trips a day. In Cape Town, drivers in our sample were averaging 4–5 trips per day in April and May.

As a result, we found evidence of sharp declines in drivers’ incomes. Dumele in Johannesburg told us that he would previously end the week with around ZAR 7,000 (US$414). Between the week of 30 March–6 April, he earned ZAR 3,739 (US$207). On 8 April he returned home after nine hours searching for fares and told us, ‘Today, I earned nothing’. This kind of drop in income is not very uncommon in the gig economy. In their study on gig workers in five African countries (Anwar & Graham 2020a, 2021) noted that gig economy jobs can go away at any time and workers do not know when the next gig will come. In ride-hailing, daily earnings depend on ride requests – if drivers do not accept rides, they do not earn. Consequently, the livelihoods generated through these platforms can be characterised as unsustainable and the pandemic has made drivers’ lives even worse.

For low- and middle-income countries, the opportunity cost for workers to join platforms can be high (Anwar & Graham 2020a, 2021; Rani & Furrer 2020). Much of the gig work depends on having access to digital technologies and the internet, the cost of which in Africa remains the highest in the world (Anwar & Graham 2020b, 2022). In the ride-hailing sector, this relates to car ownership and access to the internet. The ownership of cars remains low in Africa and drivers often rent cars from owners on a weekly basis (Graham & Anwar 2018b).⁵ Tsietsi, in Cape Town, says the weekly rental costs for the car, paying for fuel and buying airtime and data to support the app can come to around ZAR 5,000 (US$260) a week, which far outweighed the potential income from fares during lockdown. In Nairobi, Anthony, who was earning well over KES 10,000 (US$ 100) per day before the pandemic, told us that he made KES 2000 (US$ 20) on the day we interviewed him in May 2020. As a result, those who rented their cars (all but one in our sample) could not afford to pay weekly rent and were facing the threat of losing access to the cars. In essence, drivers were facing dispossession of the means of their livelihoods. Only a handful of car owners waived weekly rents. As a result, some were willing to drive for longer hours, despite knowing full well the risks of infection.

Drivers were particularly anxious about contracting Covid-19. Dumele told us, ‘In the last two weeks, I found only one customer wearing a mask. What if I get infected?’ Drivers were reluctant to ask customers to put their mask on, in case the customer gave them poor ratings. Because drivers are rated by customers or users, there is a lot of emotional and affective labour involved to please the customers in exchange for high ratings (Stark 2016). Citing the risks of catching the virus along with the lack of ride requests, some drivers decided to not go to work at all. This is illustrated by Mohammed, in Cape Town, who said,

I was not doing any work during the lockdown, but I started this week on Monday, and I only had two trips. It was ZAR 55 for almost half of the day. I started at 8 am and worked till around 3 pm. That is when I said, it is enough, I will go home … I have been hearing from drivers who are on WhatsApp group. They were complaining about not getting even a single trip for the day. You sit there in your car and you wait and you just go back home empty-handed. I cannot waste petrol for one or two trips. (Mohammed 2020, Int.)

Some drivers we interviewed were sharing cars with another person to split the day and night shifts. This works well for both drivers as they can earn an income, and also from the ride-hailing companies’ perspective as they generate commission with every trip made. Drivers are aware that ride-hailing companies make money because of their labour, despite not owning any cars or fleet. The point here is that platforms extract value from assets that they do not own, and hence ride-hailing operations have been described as ‘virtureal accumulation’ (i.e. ‘accumulation through the internet from labour processes in real places and contexts’ (Carmody & Fortuin 2019: 197).⁶ But with lockdown, drivers were affected more than the companies.⁷ One of the most immediate impacts was on the access to basic needs such as food.

Food poverty

Lack of regular income during the pandemic meant drivers faced food poverty. The nature of ride-hailing is such that drivers simply cannot work from home. During the lockdown, they had a really difficult time meeting their daily expenses such as paying housing rent and buying food and other essential commodities. Food insecurity has become a serious concern in Africa due to the pandemic, with more than 21% of the population (more than double the world average) in Africa facing hunger in 2020 (United Nations Food and Agricultural Organization (FAO) 2021). Drivers told us they were experiencing high levels of stress due to hunger and the fear of being made homeless. They told us that they would only eat twice a day. Some were surviving on mielie-meal (maize meal) and bread. In Nairobi, drivers were going without breakfast and sometimes even lunch in order to save money for fuel costs to drive for longer hours searching for more trips so that they can earn more. There are interesting parallels here with remote gig workers in Africa who would rather pay for the internet instead of food so that they can access work on platforms such as Upwork (Anwar & Graham 2022). This change in food consumption habits may contribute to an unhealthy lifestyle. There is already evidence that though life expectancy has improved in Africa over the last two decades, people's health remains poor (Wiysonge 2018).

An important aspect to note is that workers who depend on ride-hailing as their main source of income are particularly the hardest hit. This includes migrants, who remain excluded from job opportunities locally and/or occupy the lower rungs of the local labour markets doing low-paid services work (Trimikliniotis et al. 2008; Souza 2021).

Migrant experience

Poverty and search for jobs remain two of the key reasons for migration within Africa. The UNCTAD (2018) report ‘Migration for Structural Transformation’ highlights that the majority of migration is taking place within the continent, that is to say it has regional dimensions (e.g. Mozambique and Zimbabwe to South Africa and Somalia and Uganda to Kenya). In Kenya, internal migration (i.e. from rural to urban areas) has a long history from colonial times with people moving to Nairobi to do low-paid jobs (IOM 2018). South Africa is one of the largest recipients of migrants on the African continent, along with Côte d'Ivoire, Uganda and Kenya. Labour markets in South Africa prove challenging for most undocumented migrants, who end up working either illegally, e.g. in underground mines (Nesvet 2020) or in low-paid services such as domestic help (Griffin 2011). For these migrants, platforms can theoretically eliminate some of the labour market constraints (e.g. official identity documents) and offer access to paid work.

Indeed, migrant workers play a crucial role in the gig economy (e.g. Bahar 2020). In both India and China, with large labour markets, domestic migrants are key to sustaining platforms’ operations (Chen & Qiu 2019; Prabhat et al. 2019). Yet, there is a lack of public datasets on the number of migrants in the global gig economy. Scholarly research on migrant issues in the gig economy is slowly picking up pace with important interventions from Riordan et al. (2020) and Anwar and Graham (2020a, 2021) showing why migrants work in the gig economy and how the precariousness of migrants is exploited on platforms. For migrants, the pandemic has been stressful. This is illustrated by Taro, a Zimbabwean in Cape Town. He completed his O-levels in Zimbabwe but did not do higher studies. There he worked on odd jobs such as spray painting, panel beating, printing and embroidery. He came to South Africa in 2012. We asked him if he had looked for a job here. He replied that he had unsuccessfully tried in construction and logistics. He further told us most employers were asking for a South African identification document (ID). Asylum seekers, migrants and refugees are required by law to carry an ID for work, which is hard to obtain for most and near impossible for undocumented migrants. As a result, they struggle to secure meaningful work and end up working in low-paid jobs. In the case of Taro, he started working in a petrol station before starting on Uber in 2016. He rents his car from a South African owner and pays him weekly. Because he does not have an ID he is not eligible for the limited emergency support that the South African government was providing to people during the pandemic. Narrating his experience of the pandemic, Taro said,

I am experiencing some difficulties because I cannot do anything. I do not have money. There is no source of income. The only savings that I had, I paid rent for last month. I am trapped … I have never thought to be in this situation. Financially I am struggling. I have a pregnant wife. I do not even have clothes for my child and anytime, my wife can deliver. I do not know what to do. I am just desperate. (Taro 2020, Int.)

For a migrant driver who has moved to another country, the gig economy can be a paradox. It offers them a chance to earn an income, but their migrant status also exposes their vulnerabilities. Two drivers, Muhib and Abdul, who live in South Africa told us that they left Somalia because of the conflict and feared for their lives. But for migrants like them, this does not always mean better prospects and safety. Migrant drivers in South Africa mentioned that there are tensions between South Africans and other African nationals which can manifest in violence. In South Africa, xenophobic violence has erupted from time to time against migrants. Abdul was an informal trader selling chocolate, chips, cigarettes and soft drinks from his tuck shop.⁸ His shop was looted in 2018 during the protests for housing in Cape Town. He decided to find work on Uber, because for him it was one of the easier options to earn an income (Abdul 2020, Int.). Yet, working on platforms is not as easy for them as it seems.

Migrant drivers told us stories of harassment such as getting stopped by police for documents to prove they can drive on Uber. For them, this can lead to hefty fines if they do not have proper documents to hand. Others told us stories of regular violence just because they were foreigners. Their concerns are not unfounded. A Zimbabwean ride-hailing driver was found murdered in February 2020 in Langa, Cape Town (Hlati 2020) and a Rwandan driver was murdered in 2019 in Cape Town.⁹ Drivers were aggrieved over the treatment they received from the government and they felt the ride-hailing companies have deserted them as well during the pandemic.

DRIVERS’ DISCONTENTS

Towards ride-hailing companies

Drivers we interviewed expressed deep discontent towards the seeming absence and non-commitment of ride-hailing companies to help drivers during lockdown. Jim in Nairobi said, ‘No, I have not gotten anything. For me, they (Uber) have not helped me in anything’ (Jim 2020, Int.). Uber announced financial assistance for drivers who contract Covid-19 but with strict limitations (Uber 2020). Drivers could claim only two weeks of compensation based on their weekly average earnings at the time of application. To be eligible, a driver must have a confirmed case of Covid-19 or have been individually ordered by a doctor or public health official to self-quarantine. While this may be easier for drivers in parts of the world where testing is widely done (e.g. Denmark and the USA), in Kenya the testing rate at the time of writing (September 2020) was around 9.46 tests per 1000 people and in South Africa it was about 67.69 tests per 1000 people (Our World in Data n.d.). Not one participant in our sample applied for this compensation. The policy has been suspended by Uber as of August 2021.

Towards governments

Governments in both countries set up emergency grants in response to the pandemic-induced lockdown. The South African state set aside US$2.6 billion to be given in cash. It also expanded child support grants and introduced emergency social grants to those whose income is affected by the lockdown (Kazeem 2020). The Temporary Employer/Employee Scheme in South Africa replaced some of the lost income due to the Covid-19 but only for registered businesses and workers, thus excluding unregistered workers. The emergency Covid-19 Social Relief of Distress grant includes individuals who are unemployed and do not receive any other form of social grant or Unemployment Insurance Fund (UIF) payment. Migrant drivers had difficulties in applying for it. All the migrants in our sample in South Africa tried to apply for the Covid-19 Social Relief of Distress grant but did not succeed. Even South African citizens had problems applying for the emergency grant because they could not provide documents such as bank statements or proof of residential address. To make matters worse, because drivers are not considered employees, they do not generally qualify for existing national social security grants and welfare schemes (e.g. UIF in South Africa; National Social Security Fund Kenya) which are available for formal sector workers.

The Kenyan government also introduced a raft of fiscal measures (e.g. tax relief) for businesses and set aside US$93 million for squatters, street vendors, motorbike drivers (boda boda), etc. along with Inua Jamii, its flagship cash-based National Safety Net Programme (NSNP) (BBC 2020). For drivers, who are classified as independent contractors, it is still to be seen if they are covered by these programmes. At the time of the interviews, all drivers in Nairobi told us they had not received any help from the government and doubted they ever will. However, most of these measures are temporary and as a result drivers have developed their own unique mitigation strategies.

MITIGATION STRATEGIES

Interpersonal networks

Informal workers are known to depend on multiple sources of income for their livelihoods. Both Callebert's (2017) and Cooper's (1987) accounts of dock workers in Durban and Mombasa respectively highlight informal workers’ attempts to diversify their income sources, i.e. not only wages they earn from their work but also from their links to the countryside, access to land, trade and commerce. This helps them to cope with an event or a shock (e.g. the pandemic) that threatens their livelihoods.

Similarly, drivers’ strategies and mitigation tactics are also built around income diversification in the form of informal street trading, and some have even tried to do online remote work (e.g. marketing and online teaching). Some sold their livestock to cover daily expenses and the maintenance of the car (e.g. fuel, data, airtime, hand sanitiser). Others resorted to selling agricultural goods on the roadside. Drivers’ interpersonal networks run deep in the community and members often help each other in difficult times.

Respondents told us that family and community networks were extremely important for their survival during lockdown. Help within the community often takes the form of borrowing from friends, family or informal lending through ‘stokvels’ or ‘chama’ (Chidziwisano et al. 2020). Where family members are already in stable and regular jobs, they become an important source of financial support. A driver in Kenya borrowed money from a relative who is in a government job and had also asked his parents to lend some money. Drivers helped each other by buying food, airtime and medicines. Local shopkeepers also supplied food and essential commodities to drivers on credit. Beyond these immediate networks of support in the locality, digital communication channels (e.g. WhatsApp and Facebook) also helped drivers maintain a sense of community and overcome some of the adverse impacts they faced due to the pandemic.

Digital communication channels

Digital communication channels have been recognised as important sites of worker networks in the gig economy. Gig workers usually turn to digital communications (e.g. chats, phone calls, video calls) with friends and fellow workers to overcome isolation and loneliness (see Anwar & Graham 2020a). Social media networks (in particular Facebook groups) are effective tools for platform workers to exert agency which can be understood as everyday practices of resilience, reworking and resistance in the gig economy (Anwar & Graham 2020a). Building on this literature, we focus on ride-hailing workers’ regular use of digital communication channels, primarily WhatsApp. With lockdown forcing some drivers to stop driving and stay home, their connections with other fellow drivers in the physical realm were relatively reduced. But WhatsApp emerged as an important tool to keep in contact with drivers (Webster et al. 2021). Therefore, we studied drivers’ interaction on one WhatsApp group to understand how they use these channels, for what purposes. Below we outline drivers’ resilience, reworking and resistance practices using digital communication channels.

Resilience and reworking

Drivers mainly join WhatsApp groups in search for work. Their communication on WhatsApp primarily revolves around finding a car to rent (Table II). Some members (car owners) were looking to find a driver. Also, stories of daily help and advice are frequent in these chats. Drivers’ use of WhatsApp can be understood as everyday ‘resilience’ and ‘reworking’ in the gig economy, which practically means dealing with everyday life situations these gig workers have to put up with (Anwar & Graham 2020a). For example, finding out which roads, streets and junctions in the city to avoid accepting rides, which areas have surge pricing, how to deal with customers if they misbehave, and seeking help if their cars break down. We asked one of the drivers in Nairobi to explain the importance of joining the WhatsApp groups and he said,

For the sense of community of people, you are working with … If someone is stuck, they let us know in the group and the closest person to where that driver is goes and helps them out … Also, you might not know the areas of Nairobi. So, you have to have people who know the city. People will tell you, “Do not go during this time. These are the best times to work. These are the places where you get harassed by police or by thugs.” So, the WhatsApp group, the community, it helps because then you can talk about the problems you have together.’ (Njenga 2020, Int.)

Table II. Drivers’ engagement and communication on WhatsApp

Resistance

Digital communication channels such as WhatsApp (and others such as Twitter and Facebook) can also be critical tools for resistance practices. The use of social networking sites has been critical for political participation and wider social movements (Aouragh 2012), and also employee voice and resistance (Walker 2020). While resistance practices have been rare in the remote gig economy (Anwar & Graham 2020a), in the ride-hailing sector worker actions and resistance are well recognised (Cant 2019). Drivers are known to congregate near particular places (traffic junctions, airports, restaurants, train stations), and leverage it to organise and demonstrate against platforms (e.g. Tassinari & Maccarone 2019). Because drivers are not unionised, various online groups on social networking sites have been a useful front for worker mobilisation (Anwar & Graham 2020a).

In the UK, organisers from the United Private Hire Drivers Association organised a ‘digital picket line’ in 2018, urging customers to not cross it and drivers to not log into the app (Quinn 2018). An executive at the Transport Workers Union Nairobi we interviewed acknowledged the crucial role of digital communication channels in organising workers. However, the executive also noted drivers fear that the companies will punish them and remove them from their platforms. Platforms see mobilisation of drivers as a threat and often block and remove those who engage in mobilisation (Aslam & Woodcock, 2020; a point confirmed to us by numerous drivers in both countries who were blocked from the app). Nonetheless, we found that drivers regularly communicate on WhatsApp to express discontent and dissatisfaction against platforms’ business models, fare per ride, control of work, rentals, etc. We found encouraging discussions taking place in the WhatsApp group for the mobilisation of African drivers during lockdown. Interviews with drivers further confirmed the significance of WhatsApp for organising demonstration and strikes against platforms.

There are already some nascent resistance movements against ride-hailing companies emerging in Africa with some success. A local driver organiser we spoke to in Nairobi told us that drivers organised a strike against Uber in Nairobi in March 2017 after which Uber increased the per kilometre charges from KES 35 to KES 42. More recently, drivers in Nigeria and South Africa protested against the fares and the working conditions on platforms (see Omilana 2020). At the time of writing in October 2020, Uber had increased the fares by 10% in Nigeria on the back of a week-long boycott of the app by drivers, though this is still less than the original demand of the drivers (Babatunde 2020). Furthermore, the ride-hailing companies are also under pressure from African governments. In 2018, an Egyptian court suspended the licence of Uber, which was later lifted after Uber agreed to pay value added tax (VAT) (Middle East Monitor 2019). Recently, in the midst of the pandemic, the Nigerian government's new regulation has required ride-hailing companies to pay for the licence fee to operate. While this is seen as one of the few steps to rein in the power of the ride-hailing companies, the law also stipulates that drivers will be required to pay to obtain new permits to drive on Uber and has introduced restrictions on the age of the car, which is expected to add extra cost for drivers. The Nigerian drivers have already gone on strike against these measures that they believe will affect them adversely (Ongweso Jr. 2020). It is to this we turn now to highlight the importance of better regulatory frameworks that hold platform companies accountable but also the need for collective bargaining to improve material conditions of gig workers, especially in low- and middle-income regions such as Africa.

CONCLUSIONS

In his influential work, ‘The Great Transformation’, Polanyi (2001) wrote about the threats of the deregulated markets and the commodification of labour to society and nature. Polanyi argued that such a system is contradictory, unstable, and bound to fail. He also highlighted the emergence of a ‘double movement’ in which social forces protect society from the ravages of the free market. Building on Polanyi's insights, Nancy Fraser (2013) argued that while commodification is central to the workings of the contemporary economy, the double movement conceptualised through Polanyian political logic is hard to come by. Fraser (2013: 124) notes that because there has been a change in the accumulation model (i.e. shift away from industrial production to finance), ‘labour cannot supply the backbone for the protective pole of a double movement in the 21st century’. In other words, labour has become weaker vis-à-vis capital in the contemporary world economy. This is nowhere better demonstrated than the global gig economy, which emerged out of the 2008 global financial crisis and represents an increasing commodification and informalisation (Schor 2020). Although some platforms may involve various processes of formalisation (e.g. onboarding of drivers for ride-hailing) this ultimately serves the interest of the platforms. The overall point is that the gig economy represents a class project which ultimately shifts the power towards platforms even as risks are borne by labour.

Nonetheless, the political project, to counter the transformations brought by the gig economy platforms, had remained in the background until the onset of the current pandemic, especially in low- and middle-income countries.¹⁰ While there were broad-based social struggles against neoliberal globalisation before the pandemic, there is an emergence of a stream of social movements during the pandemic especially in the low- and middle-income regions against the injustices faced by labour as the pandemic threatened to destroy jobs and force millions into poverty (see Chattopadhyay et al. 2020). As we discussed above, there have been numerous instances of push-back from labour in these regions with varying degrees of success. Furthermore, the measures set up by governments around the world in the wake of the pandemic do offer some sort of temporary relief to the working classes. However, many such initiatives do not provide adequate protection to workers in the gig economy. In essence, the pandemic has exposed the broken employment relations commonly found in the global gig economy, which offers freedom and flexibility to some but also brings precarity and vulnerability into the lives of workers (Anwar & Graham 2021, 2022). Therefore, there is an urgent need for the radical overhauling of labour relations that characterise the global gig economy. This article puts forward a two-pronged strategy to counter the exploitation inherent in the gig economy.

First, effective regulatory systems need to be put in place that can hold platform companies accountable. Challenges to ride-hailing companies’ business models and operations in the UK provide optimism that this is possible and that states can rein in the advances of capital over labour in other parts of the world as well. However, there are concerns that some African countries facing large-scale unemployment (e.g. South Africa) and where a large segment of the employed are in the informal economy (e.g. Kenya) might not see it as a policy choice to introduce new regulatory mechanisms for the gig economy. Also, the latest developments in Africa point to a changing course among decision makers to regulate platforms (e.g. introduction of licensing for platforms), albeit with less focus on the interests of gig workers. Here, the regulatory authorities in Africa need to include gig workers’ voices into the discussion on policies relating to the new world of work including the gig economy.

Second, effective regulation will require much-needed collective mobilisation of workers (Webster 2020). Despite Polanyi's disagreement with Marx over the inherently exploitative nature of capitalism and the collective power of organised workers as a potent force to resist such exploitation (on critique of Polanyi's double movement see Munck, 2006; Selwyn and Miyamura 2014), the associational power of workers has won them several fights. Historically, important battles have been won by workers both in Africa and in other parts of the world largely due to their collective efforts (Webster et al. 2009). The working class has been at the forefront of pushing against the colonial and the Apartheid regimes on the continent (Beckman et al. 2010). In the context of the contemporary gig economy, even though workers are increasingly under threat from platforms, they have agency and have various sources of power at their disposal to push back against capital (Anwar & Graham 2020a). In ride-hailing, as we have noted, workers have organised and demonstrated against platforms with some success (e.g. increasing of fares after strikes). In future, the emphasis should be placed on both old and new ways of organising in the gig economy.

There is some evidence of new forms of worker solidarity emerging in the gig economy within Africa. As we have shown, African workers are using digital communication channels to develop solidarities during the pandemic. Self-organised groups of drivers have sprung up in different parts of the continent. Digital communication channels such as WhatsApp offer an excellent tool for gig economy workers to mobilise and organise outside the domain of established trade unions. Here, we argue that online activism of gig workers on social media groups represents self-organised spaces of action against the systemic exploitation of workers. This form of labour activism speaks to Atzeni's (2021) recent call for a move away from ‘trade union fetishism’ in the contemporary era where ‘new class identities and social alliances’ are emerging. An important question here for researchers is not just to examine what kind of struggles are emerging but also how those movements can be developed for wider socio-political resistance against both platforms and digital capitalism.