Abstract
This article discusses protest efforts undertaken by platform-based food delivery workers during the first wave of the pandemic. Following the lockdown, food delivery platforms were categorized as ‘essential’ to ensure that their operations continued. Several changes were made during this time to hiring practices, platforms diversified into providing grocery services and incorporating safety protocols to enhance customer ‘confidence’ in their services. The article starts by showing how the pandemic helped to strengthen the platform’s position in the market on the backs of delivery partners’ who were reliant on platform work as a means of livelihood. Though publicly glorified as ‘superheroes’, their remunerations were slashed during the pandemic, triggering a series of strikes. Since June 2020, workers across several cities have resorted to protest the worsening conditions of work. It gives an ethnographic description of two strikes that took place in June and Sept 2020 in Hyderabad. It then compares these two strikes to discuss workers’ motivation or the lack of it to strike, the strike tactics used by them, as well as the responses of platform companies to the strike. I then focus on the structural and contingent factors which rendered worker’s bargaining power weaker, despite them being providers of ‘essential services’.
Introduction
Following the declaration of COVID-19 as a pandemic by the World Health Organization (WHO), India enacted one of the strictest lockdowns in the world. During this period, only services categorized as essential by the state were allowed to operate. When Swiggy, a major food delivery platform, was not permitted to operate, its app informed customers: ‘As an essential service, we are working hard with local authorities to resume services’ (Ranipeta, 2020). Eventually, given the huge demand for home delivery, Swiggy managed to get itself declared an essential service. Prime Minister Narendra Modi praised ‘home delivery agents’ for ‘selflessly serving others, without caring about themselves’ (Press Information Bureau, 2020). Delivery platforms like Swiggy represented their workers as ‘delivery heroes’, ‘corona warriors’, ‘hunger saviours’ and the ‘backbone of our organization’. Though glorified as essential workers, ongoing reductions in earnings and incentives, accompanied by rising fuel and food prices, triggered a wave of strikes by delivery workers across cities in India and different parts of the world (Abilio et al., 2021; Elbert & Negri, 2021).
This article discusses the protest efforts undertaken by app-based food delivery workers during the first wave of the pandemic in India in 2020. It begins by exploring the factors that helped food delivery platforms enhance their position in the market during the pandemic. It then briefly looks at how labour is organized in the platform economy. A short note on methodology follows this. After briefly describing food delivery work in India, I provide an ethnographic account of two failed but differently organized strikes. The article then discusses the structural factors that weakened workers’ bargaining power during the pandemic, despite being providers of ‘essential services’. I conclude by showing how platforms capitalized on the pandemic to strengthen their power at the expense of vulnerable workers who relied on platforms for their livelihoods.
Background
COVID-19 led to a spike in the use of digital platforms to work, socialize, learn, be fed and be entertained (Nieborg et al., 2020). Restrictions on indoor dining were devastating for restaurants but accelerated the growth of digital food delivery platforms. Although several restaurants were forced to shut down, cloud-kitchens 1 witnessed a rise across India (Borah, 2021). Food delivery became a saving grace for restaurants during the lockdown (Ahuja et al., 2021). To instil a sense of safety amongst customers, delivery platforms introduced measures like contactless delivery, and apps provided information on the temperature and vaccination status of cooks and delivery workers. Mohit Sardana, the Chief Operating Officer of food delivery platform Zomato, remarked: ‘focus on safety, help us get consumer back…’ (MediaNama, 2021).
The platforms’ apps were programmed to ensure physical distancing at the food collection point. Facial recognition technology was introduced as a new surveillance technique to verify the worker’s identity and to ensure compliance with dress codes and face masking rules. The workers further had to enter their body temperature while logging in, though they did not get to see the customer’s body temperature. Meanwhile, the Government of India made it mandatory for workers to download its Arogya Setu contact tracing app. Swiggy provided training about these new COVID-19-related protocols and safety features through online tutorial videos and instructional messages.
As Klein (2007) argues, disasters enable capitalism to reformulate itself. In the course of the pandemic, food delivery platform businesses such as Swiggy and Zomato cut losses by laying off staff in management roles alongside diversifying into newer verticals such as grocery delivery. During the pandemic, Swiggy laid off 1,100 support staff and outsourced supervisory work to a third-party workforce management company (Vaidyanathan, 2020). The pandemic accelerated the company’s efforts to do away with training instructors by switching to gamified video-based training. Further, selfie verification was introduced, and Swiggy relied on customers to verify workers’ compliance with the dress code. While these mass layoffs were made to curb losses, in April 2020 Swiggy raised $43 million to expand its foray into grocery delivery (Abrar, 2020).
Platform Work and Labour Organizing
The novelty of digital labour platforms lies in how they have been able to restructure work by combining algorithmic management with labour precarity, the use of fictitious titles and flexibility. Algorithmic management entails the extraction of minute-level, real-time data (Chan & Kwok, 2022; Rosenblat & Stark, 2016). As owners of digital means of production, platforms enable new forms of accumulation without owning assets—in this case, delivery bikes and smartphones—and without formally employing workers (Burrell & Fourcade, 2021; Srnicek, 2017). Classifying workers as either partners or independent contractors reduces the need to adhere to government regulations and enables the platform to rise as a quasi-regulatory force (Christophers, 2020; Kenney et al., 2020). Platforms provide work on a piece-rate basis, with the time workers spend waiting for work being unpaid (Dubal, 2020). Platform businesses encourage a sense of entrepreneurial spirit amongst the workers, compensating for the lack of direct human supervision (Barratt et al., 2020; Purcell & Brook, 2022). Platform work is further characterized by a tension between worker’s autonomy and dependence on the platform (Hoose & Rosenbohm, 2022). The use of algorithmic management, legal misclassification of workers and the entrepreneurial culture promoted by platforms create new challenges to labour organizing. The way platform workers organize varies by sector because different sectors enable different kinds of mobilizing strategies. In the case of remote-work platforms like Upwork (Anwar & Graham, 2020) or AmazonTurk (Kassem, 2022), workers are much more alienated from one another because they are globally dispersed. While scholarship has shown how remote workers can share solidarity with one another (Malik et al., 2021), there seem to be fewer instances of collective mobilization amongst remote platform workers (Kassem, 2022).
Even within location-based platform work such as cab driving and food delivery, there are variations in organizing collective resistance across regional contexts. However, scholarship makes evident the growth and popularity of digital spaces such as Facebook and WhatsApp as critical for solidarity formation (Tassinari & Maccarrone, 2020, p. 46) across regions, such solidarity often spilling into physical spaces and evolving into collective forms of mobilization (Parth et al., 2021). In some contexts, platform workers engaged in legal battles, fighting for employee status or demanding legal benefits accruing to regular workers (Fredman & Toit, 2019).
Logging out, which resembles stopping machines in the twentieth century, emerged as a new repertoire of resistance in the platform work context (Iazzolino, 2021; Tilly, 1993; Vandaele, 2018). Organizing mass logging out from the app and making other workers do so is much more complicated than picketing at the factory gate. We will see through the ethnographic example discussed below how workers can hinder the functioning of the platform through a combination of online activity (logging out) and offline protest (picketing outside restaurants).
Methodology
This article draws on ethnographic fieldwork carried out in the southern Indian city of Hyderabad from November 2018 till the later part of 2020 as part of my doctoral research on food delivery platform workers. At the time of the strikes in mid-2020, I already had established a close rapport and built trust with several workers in the Mehdipatnam-Tolichowki zone. 2 I further shadowed a couple of workers as they went from restaurant to restaurant to monitor strike activities.
During the strike, I developed familiarity with another 15–20 workers, through my close respondents who were involved in organizing the strike from different zones of the city. I also developed a rapport with the Indian Federation of App-Based Transport Workers (IFAT), which played a limited but significant role in the strike. I had short informal interactions with some of their members during the strike. I sought regular updates through phone calls from some of the workers with whom I had already established rapport earlier. All the names of the delivery workers mentioned here have been changed to maintain anonymity.
I also made physical visits to different restaurants to observe the protests and interact with the protestors during the strike. Some of the WhatsApp groups I joined during the strike provided access to real-time discussions and updates about the strike. After the strike, I interviewed (15–90 minutes long) 10 key organizers and participants to seek their perspectives and obtain more details about the strike. IFAT facilitated a couple of those interviews. However, I could not get interviews with some of the striking workers who had left food delivery work after the strike. I triangulated my field visits with interviews and with observations on the ground, on Twitter and on WhatsApp groups.
Food Delivery Work during the Pandemic
There are about 500,000 delivery workers in the food delivery sector who are spread across 500 cities and towns in India, roughly divided between Swiggy and Zomato (Swiggy, 2023; Zomato, 2023). They are paid on a per-delivery basis, supplemented by a range of incentives. Delivery workers have multiple work timings and work in different zones. Different categories of workers consist of those who work full-time, as well as part-time employees, including college students who work on a part-time basis. The workforce is predominantly male and young, with varying educational levels, caste, religion and regional backgrounds. The workforce is characterized by high attrition rates. Amongst my participants, there was an acute sense of stigma related to being a food delivery worker. This was primarily because this work was manual, fluctuating and low-paying. Workers faced stigma as they experienced a loss of respectability amongst their family, neighbours, friends and village acquaintances, which they strongly feared would hamper their marriage prospects.
According to the National Sample Survey Office (NSSO) job survey, just before the start of the pandemic, India’s unemployment rate was at its highest in the last 45 years, and this worsened further during the pandemic (PTI, 2019). Within platform work, food delivery was less affected than other labour platforms, such as ride-hailing (Grohmann & Qiu, 2020) or home beauty services, because it remained operational during the lockdown. Food and retail deliveries were amongst the few earning opportunities available for the new pool of unemployed workers. Having emerged from the lockdown restrictions, many jobless young men were willing to opt for any work available to meet their financial needs.
In the initial days of the lockdown, when there was no clarity about the permissibility of delivery service at the national, state and city level, delivery workers faced harassment by the police authorities and, at times, by the Resident Welfare Associations that manage gated communities. For instance, when they were fined or when the police seized their bikes for violating lockdown rules, workers had to ask local politicians to intervene to get the bikes released. With little or no on-the-ground support from the platform, delivery workers often had to resolve such issues by themselves. During this period, their incomes declined because of a reduction in orders as very few restaurants were operational. But even when the volume of orders gradually increased with the easing of the first lockdown in 2020, worker incomes continued to shrink because of cuts made by Swiggy to per-order rates and the ‘incentive’ component of pay. Consequently, workers experienced a 40% to 60% decline in monthly incomes after the easing of lockdowns in the latter half of 2020.
When asked about his experience of working during the pandemic, Imran, an MBA graduate who had been looking for a job even before the pandemic, said sarcastically, ‘We are like social workers who have volunteered with an NGO in our free time doing food deliveries to needy people in return for remuneration that compensates only for the fuel cost’. When the lockdown restrictions began to ease, I trailed another worker, Ghani, on his bike to understand delivery work during the pandemic. He found it easier to work during the lockdown as there was less traffic on the roads, and customers who usually expect delivery workers to come to their door would come down to the gate to collect their food.
Platform labour protests in developing countries such as South America, Africa and Asia increased sharply during the pandemic (Bessa et al., 2022). Between June 2020 and February 2021, strikes in 11 Indian cities were reported, and during this period, Hyderabad witnessed 2 strikes by Swiggy delivery workers (see Table 1). Other small strikes went unreported by the media. There had been protests by delivery workers before, but during COVID-19, the scale of protest in terms of the number of cities and duration of the strikes was more intense. This shows that the workers’ lived experience was different from their heroic portrayals in advertisements (Swiggy, 2020).
Overview of Strikes in Different Cities between June 2020 and February 2021.
In Hyderabad, just as lockdown restrictions were being eased, two strikes took place: a ‘relatively spontaneous’ one in June 2020 and then a planned one in September 2020. In the following two sections, I present ethnographic accounts of these two strikes and discuss the protest strategies used by delivery workers and the obstacles they faced to organize collectively.
Strike 1: Spontaneous and Unplanned
On the 20th of June 2020, around 11 a.m., Ghani—one of my close respondents who had worked with Swiggy for about two years—called to inform me about a strike being planned that day in the Mehdipatnam-Tolichowki Zone. Had he not told me about the strike, I would have had no clue about it. I soon reached Shah Ghouse restaurant, a ‘high-demand’ restaurant in the city, where Ghani had asked me to meet him. There were close to 50 delivery workers already gathered there. Ghani introduced me to his colleague Junaid, who added me to the protestors’ WhatsApp group. Within a few minutes, as workers continued arriving, the number increased to 100. ‘There is no need for 100 delivery workers here’, I could hear a voice from the large group of workers. He added, ‘We need to split and send teams of 10 to other restaurants that are in demand during lunchtime as lunch peak will start shortly’. Everyone agreed with this plan and the workers began to disperse to other popular restaurants nearby. I trailed Ghani on his motorbike for the next six hours.
Let us first turn to what triggered this sudden strike action. On the previous night, around 9 p.m., a screenshot of an order whose actual earning should be ₹11, but was shown as ₹1 was captured by an unidentified delivery worker (see Figure 1). According to the Fleet Manager, the miscalculation was due to a technical error. The screenshot went viral the next morning amongst the workers, shared over WhatsApp groups. A screenshot is an important way for workers to share their work experiences on WhatsApp groups. This viral screenshot became a talking point amongst workers, leading to collective anger.
A week earlier, on 13th June 2020, Swiggy had reduced the per-order rates and incentive structure for workers. The per kilometre charges based on incremental distance, which went up to ₹18 per kilometre were reduced to a uniform rate of ₹ 6 per kilometre. A new incentive structure was introduced with lower bonuses. Further, bonuses for late-night deliveries or deliveries during rain or peak bonuses would no longer be counted for incentive calculations. Such ‘changes are not negotiated with workers but rather notified through the app’ (Woodcock, 2021, p. 4). The negative impact of this fare revision became apparent to workers over the course of the week. Immense anger built up amongst the riders, especially given that fuel prices were rising. While previous changes were gradual and piecemeal, such as a reduction in the night-order incentive from ₹20 to ₹10, the reduction in fare and incentive structure post-lockdown impacted all the workers at once.
The main demand during the strike was to revert to the previous per km rates and incentive structure. There was no prior call for the strike by any formal union, nor was it pre-planned by the workers. It was a spontaneous reaction to pent-up frustrations triggered by the viral screenshot, which became a poster for the strike.
Food delivery workers met each other every day near restaurants, albeit for a limited time between orders. Sharing their work experiences during such interactions gradually resulted in the formation of informal WhatsApp groups with 10–25 members (see also Tassinari & Maccarrone, 2020). Fleet managers also encouraged workers to form such informal groups to sort out any issues that arise during work amongst themselves. Even though the strike was spontaneous, the presence of many such informal WhatsApp groups helped to quickly circulate information about the strike and keep workers informed about developments, such as moving from one restaurant to the other and so on.
Mobilization requests were accompanied by messages like ‘This is what will happen if we remain silent’. The absence of a formal union means a lack of leadership and proper mobilization. Due to the instant nature of the strike, some workers, especially part-timers, were left unaware of it until it was over.
A group of workers formed a WhatsApp group at around 11:30 a.m. and initially added 20–30 members they knew directly. These 20–30 members either forwarded the joining link of the group to their friends or received requests from existing group members to add them to the group. Within an hour, the group had 120 members and more members were being added throughout the day. Such WhatsApp groups enabled audio, video or textual communication amongst striking workers, which was necessary for the workers to be connected and coordinate the strike. Real-time updates about the strikes, news in the print, electronic, TV and YouTube media, memes and digital posters that delivery workers made were being circulated in the WhatsApp groups.
During the strike, the striking workers prevented other workers from picking up food and made them cancel the order. In the gathering of about 100 workers at Shah Ghouse, it was decided to form teams of 5–10 members and send them to different restaurants, usually the major ones such as food chains like Mcdonald’s, KFC and BurgerKing or locally popular restaurants like Shah Ghouse, Pista House and Prince Hotel. If a worker refused to take part in the strike and collected food from the restaurant, the striking workers—although they avoided snatching the food packet from their working colleague—nevertheless tried to prevent them from leaving the restaurant with the food and forced them to cancel the order. Striking workers convinced such workers to participate in the strike and the latter mostly cancelled the order and left the premises. Thus, striking workers made several restaurants ‘unserviceable’ on the app, hampering the smooth functioning of the delivery service (see Figure 2).
Though it is hard to know how many of the estimated 700–1,000 workers in the zone took part, many workers did not log in. Logging in and then rejecting orders would lead to workers being penalized. Moreover, the company would easily identify the participants because rejections are automatically recorded, resulting in access to the app being denied for the day.
In the retail sector, goods are circulated through global supply chains and pass through transportation nodes such as ports, airports and warehouses. Companies like Amazon and Flipkart work on the just-in-time model and have fulfilment centres where the goods are checked, packed and transported further. Such nodes are potential ‘chokepoints’ in the system to exercise collective worker resistance (Bonacich & Wilson, 2008; Moody, 2019). However, in the case of hyper-local food delivery, there is no single pickup point or node because the food is picked up from multiple restaurants spread all over the city. Nor do workers have a single fixed workspace that they could picket. Swiggy’s city office was closed during this strike due to the lockdown, with the staff working from home. Any issues had to be resolved through Google forms or by calling Swiggy’s call centre. Despite the lack of a single space to protest at, the many popular restaurants spread across the city became crucial sites of the strike.
Although preventing food pickups was a crucial part of the strike strategy, it came with its own challenges. The restaurants are not only spread across the city, but several popular restaurants are often within a single zone. Workers are allocated a particular zone in the city based on their preferences or the platform’s requirement. A delivery worker is confined within one or two zones, unlike taxis that ply across the city. The spatial spread of restaurants makes it challenging to picket every restaurant in the zone unless planned very well. Besides the few high-volume restaurants, there were many other pickup points where striking workers could not prevent order pickups, or if prevented, could not be sustained for more than a couple of hours. Some pickup points included apartments in residential zones, juice centres, chat centres and food vans. As Swiggy had diversified into delivering groceries, supermarkets, chicken shops and similar stores were added to the already long list of places that needed striking workers. Moreover, a strike in this work setting has to be sustained almost 24/7, unlike in other work settings with a clear start and end time for work.
Being present striking outside a restaurant brought about unwanted visibility for workers. Workers were relieved that the mandatory wearing of face masks helped keep their identities anonymous on CCTV cameras at restaurants. They were also cautious in choosing pictures to upload on WhatsApp groups or social media platforms, picking only those photos in which masks covered the faces of the striking workers. This is because platform companies such as Swiggy ‘are unsurprisingly hostile’ (Prassl, 2018, p. 65) to any efforts by their ‘partners’ to collectively organize and, as a reprisal, can suspend their IDs. Blocking an ID means that access to work is denied, which could be permanent or temporary at the discretion of the platform management. Because of this fear, those who were striking alerted one another to logout from the app and turn off the GPS location on their mobile phones. Several workers recalled having their IDs suspended during a strike in 2019 for being physically located near a picketed restaurant. Thus, while the spatial spread of the points of protest made it a challenge to render most restaurants ‘unserviceable’, the fear of being ‘seen’ by digital technologies also deterred workers from striking or being physically present at protest sites, thus adversely affecting a strong show of united strength to Swiggy.
At the restaurants, the picketing workers had to encounter the restaurant management, who prevented workers from picketing by either using their security guards or by calling the police. At Shah Ghouse restaurant, the restaurant manager summoned a few delivery workers and asked them why they were picketing. Unconcerned with their issues, the striking workers were instructed not to gather in front of the restaurant. When the delivery workers refused to move, the restaurant management resorted to verbally abusing them to make them move away from the restaurant. The workers dispersed to the parking space beside the restaurant and continued preventing pickups.
The workers used their networks to invite local media channels that post content on social media platforms like YouTube and Facebook. Two YouTube channels covered the strike. At 1 p.m., a news reporter came to Shah Ghouse. A voice message was soon posted on the WhatsApp group asking workers to assemble at Shah Ghouse. The protesting workers shared their grievances with the reporter and chanted slogans against the company. Workers shared the video clip on their WhatsApp group, requesting members to share it widely and post it across social media platforms.
At 2:49 p.m., a screenshot of a blocked ID was shared on the strike WhatsApp group. After some investigation by workers, it turned out that it was an old screenshot of a rider that someone had shared to dissuade workers from striking. Giri, Vijay, Sajid and others were furious and upset with this incident because it gave the impression that the management had started to suspend workers. Striking workers could not verify the authenticity of the person who had posted that message or of the worker whose ID had been blocked. The screenshot nonetheless created fear amongst the group members that the management had started blocking IDs. Vijay began getting calls from workers enquiring whose ID had been blocked. He had to convince multiple callers that the screenshot was not genuine.
At Creamstone, a popular ice cream parlour, when around 3:30 p.m. one rider came to pick up an order, the striking workers gheraoed him and told him not to pick it up. He refused to oblige and went into the restaurant and collected the food. They encircled him again when he returned to his bike and forced him to cancel the order. ‘It is my wish that I will not participate in the strike, I want to work. Let me work. If you guys want to strike, go ahead. I am not stopping you guys’, he said. Eventually, however, he had to get his order cancelled after increased persistence by the striking worker.
By the evening, the striking workers were experiencing difficulty in stopping food pickups from restaurants. Workers began posting messages on the WhatsApp group about orders being picked up from restaurants like Creamstone and Burger King. Due to the lack of clear leadership, many workers informally led the strike resulting in multiple people giving directions. There were insufficient striking workers who could prevent pickups at all the major restaurants within a zone, and motivate workers sitting on the fence to protest. Some delivery workers chose not to turn up to work that day and some others feared the consequences of participating in the strike.
As the day went on, I witnessed striking workers getting into heated arguments with non-striking workers in some restaurants, sometimes even resulting in physical fights between these two groups. Unaware of the strike, a part-time worker picked up his order from Burger King around 4 p.m. and was about to start the bike to proceed for delivery. Giri, who is a middle-aged, full-time worker, along with other workers involved in the strike, spotted that worker and was furious with him. When they found out that the worker was working for Swiggy and not Zomato (the worker was not wearing a uniform), Giri’s polite and soft tone suddenly changed to one that became loud and angry. He shouted at the worker, ‘Are we mad here trying to strike since morning and you come and pickup the order!?’ He snatched the food from him and waited for the worker to get his order cancelled even as he continued yelling at him. Meanwhile, about 10–15 passersby surrounded the two of them. After the order got cancelled, Giri took the food as spoils of the strike. Another striking worker called Vijay told Giri to hand back the food. Giri refused and left taking the food along with him. Witnessing all this, Ghani angrily said, ‘Our own people are cheating us by working behind our back even while we were on strike. Some workers turn up at the restaurant as if they are here for the strike, but they want to see if they can work’. I met some workers who were undecided about striking but who had logged off work and were simply waiting for the opportunity to resume work.
However, Vijay, Sai and a couple of other workers who were all friends were upset with Giri’s behaviour. ‘This is not how we should be going on strike’, Vijay told me later when we discussed the incident. ‘Either we stop the worker before he goes to collect the order or we get the order cancelled and give him the food. We should not take the food. It would give the wrong impression’. There were a couple of other instances where angry, protesting workers snatched collected food packets and Swiggy used this to file police complaints on striking workers who had snatched food in a fit of emotion (Staff Reporter, 2020).
From this, we can see that striking workers face multiple difficulties in their strategy to strike by preventing food pickups. On one hand, they had to deal with the spatial dispersion of work, ensure sufficient presence of striking workers all over the city, and confront co-workers who continued to work. Such confrontations sometimes ended up in nasty fights resulting in strained relations with workers. On the other hand, they had to encounter and work around the surveillance measures used by platforms and deal with hostile restaurant management and the police, who all sought to break their protests. Amongst the striking workers, there were varying levels of involvement, like picketing at the restaurant, taking on a leadership role and taking on coordination responsibilities to ensure the presence of striking workers at all major restaurants.
Around 7 p.m., one of the organizers of the strike reached out to the Fleet Manager on the phone. The worker subsequently shared that senior managers firmly stated that there would be no restoration of the previous earning and incentive structure. By dinnertime, the number of restaurants accepting orders increased. On the same day, a few more zones in the city went on strike. However, they were not coordinated or pre-planned. There were also zones where no strikes took place, but the zones in which the strike did happen were ones that had a high volume of orders. By night the strike had cooled down.
Workers were upset that their demand was not met. Meanwhile, in the WhatsApp group, some workers shared messages about the futility of resorting to strikes. The creator of the group (Naresh) was also angry about the fact that rather than using the group to mobilize, a few people were constantly discouraging the strike organizers, thus lowering the morale of all the members. ‘If you cannot support then don’t support. But at least don’t put discouraging messages’, he wrote. Someone texted in the group that if people were not interested in the strike, they could leave the group. Afraid of being blocked, on the 21st of June Naresh removed me and other members to close one of the WhatsApp groups.
Although some workers wanted to continue striking the next day, a Sunday, thinking that their strike would have more impact on a day which usually saw more orders, there was barely any response to the idea on the WhatsApp group. The workers resumed work with the reduced fare. In the interim period between this strike and the next (discussed below), the price of petrol increased from ₹73 in June 2020 to ₹85 in September 2020. With no restoration of the previous earning and incentive structure, the anger amongst workers continued to accumulate. With limited employment opportunities, the workers had no option but to continue working with a lower earning structure.
Though the short and spontaneous protest action did not succeed in securing its demands, it served as a primer for future collective action (Snow & Moss, 2014) and gave workers the hope that a better-organized strike might yield results. Communication with other zones was established, which paved the way for the next strike.
Strike 2: Planned Resurgence
Unlike the first strike, the second strike was planned well in advance, and workers in other zones were also informed about it. The first strike started in the morning, thus affecting the lunch-peak orders, but by dinner time it had fizzled out. The second strike was better coordinated and was a demonstration of solid, unified strength for more than a week across different zones in the city. In the first week of September, a few delivery workers from one of the major zones, Gachibowli zone, contacted Naji and a few other senior delivery workers from different zones in the city, such as Kukatpally, Himayath Nagar and Banjara Hills, to harness the existing anger amongst workers about the declining order rates. They first created a citywide WhatsApp group with representatives from different zones to establish communication between zones, which had been absent earlier. The representatives were responsible for disseminating any communication from the collective group to their zonal WhatsApp groups or to share relevant information from their zone to this group. In the Mehdipatnam-Tolichowki zone, Naji was one such worker who would post such audio or video messages. From the previous strike, workers realized that their lack of unity was the principal reason for its failure. The need for a citywide, inter-zonal WhatsApp group and a better and continuous communication strategy were lessons that were learnt from the previous protest and were now operationalized in a bid to foster a sense of unity across zones.
For the strike in September 2020, pamphlets were printed in English and the regional language, Telugu (see Figure 3). These pamphlets listed the reasons for striking and detailed the demands made to Swiggy. They were posted near restaurants and circulated on WhatsApp groups. Calculations were made of how earnings had dropped due to reduced fares, and these figures were also posted on WhatsApp groups. These calculations included expenses on bike equated monthly instalments (EMIs), maintenance, fuel and the rising cost of essential commodities as a result of skyrocketing fuel prices.
The workers were careful not to plan the strike for August because it would be challenging to organize during the heavy monsoon rains. They scheduled it for September, just a few days before the start of the Indian Premier League (IPL). IPL is a major sports event in India that was to start on 19 September 2020, after being postponed earlier in March and April 2020 because of the pandemic (Laghate, 2020). Delivery workers knew that Swiggy depended very much on their delivery fleet for this much-awaited annual sporting event and would aggressively advertise and offer discounts to customers and incentives to workers during this time.
Striking workers in each zone got a large banner printed with similar content as that of the pamphlet but mentioning their zone name on it. When workers gathered near a major restaurant during the strike, a group photo was taken along with the banner, which was shared across social media platforms. This helped ‘to amplify and strengthen offline actions’ (Pasquier & Wood, 2018, p. 1) and gave the impression of a full-fledged strike across different zones of the city to the workers. Like in the previous strike, the tactic of the striking workers was to render major restaurants ‘unserviceable’ in their respective areas by preventing food pickups. This meant that the challenges associated with picketing the restaurant, such as ensuring the presence of striking workers and dealing with the restaurant management and police, continued.
In a couple of zones, IFAT led a group of delivery workers to take a knee near the restaurant as part of the strike to highlight their issues. IFAT was founded in 2019 and initially focused on organizing Ola and Uber cab drivers, but since the beginning of the pandemic, it has also started to highlight the issue of food delivery workers. IFAT is a coalition of unions with over 36,000 members in different app-based work across India (Bansal, 2023). IFAT was not directly involved in the planning or organizing of the strike. But during the strike, it mentored the striking workers about organizing by sharing its experience of organizing cab drivers associated with the platforms Ola and Uber. Raising issues related to platform workers in the media by consistently sending press releases to print and electronic media outlets was IFAT’s most significant contribution to the strike. The workers also utilized IFAT’s networks to invite journalists to cover the strike. This ensured that the strike received ample coverage in all media spheres, be it print, electronic, television and social, unlike the previous strike, which failed to get much media visibility.
Swiggy did not acknowledge to its customers that the delivery workers were on strike. They used terms like ‘operational constraints’ and ‘currently unserviceable’ in their communication to customers to explain why orders could not be placed from some restaurants. During the strike, the following message appeared to the customers: ‘Our services are currently restricted. Please expect a slight delay. Thanks for your patience’. There was not much visible support from the customers during the strikes, even though a few people raised concerns about exploitative work practices on social media, especially Twitter. Few restaurants were outrightly shown as ‘currently unserviceable’. Similarly, on the restaurant side of the application, the message read: ‘Zone is closed. Your zone is closed and the outlet has been turned off. You will not be receiving any orders till the zone is enabled again’. Workers using Twitter shared screenshots of customer tweets complaining about not being able to order food on their WhatsApp groups, to motivate other workers to strike and highlight the impact of the strike. In its communications to workers, Swiggy portrayed the strike as futile, resulting only in a loss of earnings.
On 15 September 2020, one of the non-striking delivery workers whom the striking workers roughed up made a telephonic complaint to the police. N. Kalinga Rao, Banjara Hills Station House Officer (SHO), said his team had gone to the location and brought a few of the men involved to the police station. ‘But the complaint couldn’t be traced to the person who lodged it. The number from which this person made an emergency call to 100 was switched off. We haven’t been able to take it forward’, he said (Chunduru, 2020). The detained workers were released late at night after the intervention of Majid Hussain, Municipal Corporator and former Mayor of Hyderabad (Chunduru, 2020).
Having experienced Swiggy’s use of incentives to break the previous strike, there was a unanimous understanding amongst the workers to not fall into ‘the trap of incentives’ again (S.B., 2018). The use of incentives to break the strike had also been deployed during strikes in Chennai. Workers called the incentive a ‘lollipop’ used to distract them from the primary demand. Each day of the strike was vulnerable to disruption because the striking workers needed 5–10 workers at each restaurant to prevent food from being picked up. During the 12 days of the strike, protesting workers discussed the future of the strike and wondered if their demands would be accepted. Gathered outside a restaurant, workers were optimistic that earnings and incentives would be restored.
In contrast to the previous strike, restaurants were not the epicentre of this strike, even though they continued to be rendered ‘unserviceable’. The main incidents of the strike happened first at Swiggy’s city office, then at the Gachibowli police office and later at the Labour Commissioner’s office. Police were called to a gathering of workers at Swiggy’s office on the night of 15 September. The workers were dispersed with the assurance that their demands would be considered, and they were invited for talks with management at the Gachibowli police station. These talks lasted a couple of days. With nothing coming out of it, on 18 September, even as negotiators remained in the police station, about 1,500 workers assembled near the stadium to demonstrate strength. Word circulated amongst the workers that they should assemble at the same place the next morning, but the police were present to constantly patrol the area to prevent any gathering.
Swiggy has area managers who are responsible for ensuring that delivery workers are disciplined and resolve disputes or strikes that may arise in their jurisdiction. They keep track of on-the-ground activities by monitoring social media groups to avert any potential mobilization. Since anyone with access to the joining link can be part of the WhatsApp group, some workers in the management’s camp infiltrated the group and shared information with management. As one striking worker, Kiran, said, ‘A pinch of salt is enough to spoil the whole milk. We planned it all and if one DE [delivery executive] falls back, hundreds will follow him’. According to a worker who was part of the negotiation team, the management had all the audio and video recordings that were shared on the groups. The management used some of the videos collected from WhatsApp groups and presented them during talks with the police and labour officials to turn the discussion about earning and incentives into law-and-order issues so as to threaten workers with criminal cases. Although WhatsApp aided workers in coordination and collectivization, it later became a weapon in the hands of these platforms.
Like its counterparts in the UK and Italy, Deliveroo and Foodora, respectively, Swiggy resorted to communication campaigns to dissuade workers from unionizing by depicting the protesters as a small minority of troublemakers (Tassinari & Maccarrone, 2020). Further, incentives and threats were offered by Swiggy to create divisions amongst workers. One such message read: ‘Unfortunately a small set of people have been intimidating some of our delivery partners and not letting them deliver orders’. ‘Surveys’ were mentioned in the in-app messages sent to the workers: ‘We would like to thank the 1,000s of DEs who logged in over those last few days. Our surveys have shown that more than 80% are happy with the package and want to continue working’.
IFAT sent letters to the city’s Labour Commissioner, Members of Parliament and the state’s Information Technology Minister and Labour Minister requesting their intervention in the matter to help them restore the previous pay structure. Workers formed a bike rally and gave presentations to various politicians, who in turn wrote letters to the Labour Commissioner to resolve the issue. On 25 September, IFAT submitted a note to the Joint Commissioner of Labour highlighting the issues. Based on this complaint, the Joint Commissioner sent a notice to Swiggy for a meeting the next day. In its reply to the notice, Swiggy claimed that they are a ‘platform’ that enables transactions between customers and merchants/restaurants through delivery partners. They further added that delivery partners are not employees and the relationship is based on principal-to-principal agreement for utilizing the platform to provide delivery services. We can see here that the traditional claim making between workers and managers is disrupted by the platform model (Tilly, 1993). The workers on one side are classified as independent contractors, and the platform on the other side is seen as a facilitator of work rather than an employer. Therefore, the workers cannot make claims.
In its reply, Swiggy further indicated that only delivery workers should be present and that external elements, that is, individuals such as Shaik Salahuddin, the National General Secretary of IFAT, should not be part of the discussions. Talks between Swiggy’s management and workers did not yield any results because the management was unwilling to accept their demand to reverse the cuts to rates and incentives. Kiran, one of the striking workers from the Kukatpally area, whose friend was part of the team representing the workers during the negotiations at the police station, at the Labour office and also during the direct negotiation between the workers and the management, said:
At the time of talks, the company representative didn’t say anything. They will talk round and round. They are MBA guys and talk like a marketing guy. There is no truth in their words but they present it as truth. What they are doing is wrong, but they present it as if they are doing the right thing.
As the talks took place, it became increasingly difficult for the workers to sustain the strike with each passing day. Workers who were their family’s sole earners mounted pressure on the striking workers because they had EMIs to pay, families to take care of and various other financial commitments. The incentives Swiggy offered to break the strike were very tempting. 3 The negotiators realized that as the days passed, they were losing the support of workers whose daily earnings had vanished and who were being tempted back by incentives. Slowly the strike lost its hold and petered out. However, unlike during the previous strike where the WhatsApp group that coordinated the strike was closed, most of the zonal-wise WhatsApp groups created during the second strike remained active.
Some of the workers left Swiggy after the strike. Naji, who was openly organizing both the strikes at the Mehdipatnam-Tolicihowki zone, was disheartened with how Swiggy responded during the strike and left delivery work to start a mobile phone repair business. In early January 2021, about four or five main organizers, including Shaik Salahuddin and a key organizer from the Gachibowli zone, were booked under Section 269 (‘Negligent act likely to spread infection of disease dangerous to life’) Section 3 of the Epidemic Diseases Act 1897 and Section 51(B) of the Disaster Management Act 2005, by the Madhapur Police Station. Though none were imprisoned, the offence for which they were booked is punishable with imprisonment for a term, which may be less than seven years or extend to seven years with or without a fine.
Comparing the Two Strikes
After these two strikes, both of which were unsuccessful at making Swiggy meet workers’ primary demand to increase remuneration rates and incentive pay, many workers were left demotivated, with some concluding that the strikes were ‘a waste of time’. Ghani, who was very involved in organizing and carrying out the first strike, only participated for the first two days of the second strike in September 2020. After the first two days, he stopped striking to reap the benefits of the temporary increase in incentive pay. He started working within any zone that allowed him to and ended up being the one arguing with striking workers to let him work. Being associated with Swiggy for almost two years, he had earned respect amongst the delivery workers. The younger protestors did not know how to deal with middle-aged men like Ghani. When I asked why he no longer participated, he replied:
I have expenses to take care of. Moreover, what would happen with the strike? Didn’t we do it last time? Do you think it was an ordinary strike? Do you know how big it was? What happened? I realised that in this work, you have to put zip on your mouth and work or else silently leave the work and go. I have family responsibilities, which bachelors don’t have. My hand is already tight because of the lockdown.
Echoing similar views, Vijay said:
There is no space for bargaining here. We feel like we are robots. Even though Swiggy knows about our conditions, knows about the rise in fuel prices, they simply turn blind and deaf to our conditions. Look what they have done. Reduced earnings! Tell me who will do that when petrol prices are increasing?
Other workers similarly said that while salaries increased over time at most workplaces, here the opposite happened. Further, they felt it was humiliating to work at lower wages in the same organization despite having worked for one or more years.
The first strike in June 2020 was spontaneous and was held in limited parts of the city, while the second strike was well-planned and held across the city. The participants in the first strike thought that the lack of unity and coordination amongst different zones was the reason for the strike’s failure. But after the second strike, the workers felt an acute sense of powerlessness, their morale taking a beating as they witnessed Swiggy wielding its power before state bodies and getting away with asserting their claim of merely being a mediating platform. Even though workers faced other problems, such as the lack of support from the platform during accidents, problems in claiming insurance after accidents or deaths of workers, or delays in responses by platforms when facing issues at work, their main demand was to restore previous rates and incentive structures. Despite organizing such a large protest peacefully, their demands were not met. The only demand that the management accepted was to include the value of bonuses earned, be it the ‘late night bonuses’ or the ‘peak hour bonuses’, in the calculation of daily ‘target earnings’. This would translate into workers being able to hit target thresholds (which would make them eligible for incentives) slightly sooner.
Impact of Workers’ Heterogeneity on Strike
My fieldwork, as well as the existing literature on platform labour (Dunn, 2020; Schor et al., 2020), has revealed the presence of a very diverse workforce in this sector. Low barriers to entry, flexibility in choosing work timings and places, and low skill requirements attract a diverse set of workers. Despite being associated with the same platform, because of this diversity workers experience work differently, depending on their work motivation and degree of dependence on the platform (Piasna & Drahokoupil, 2021). For some workers, this is their sole source of income; for others, it is an additional source of income, and some even work for non-economic reasons (Dunn, 2020). This heterogeneity in work experiences impinges on platform workers’ ability to forge strong solidarities, which in turn affects their motivation and willingness to participate in strikes. Despite all the delivery workers having faced problems, they did not have a uniform willingness to participate in the strike as they had diverse interests. College students were afraid to be at the forefront of protests because any legal case against them could jeopardize their future career prospects, especially amongst government job aspirants. They tend to have a temporary orientation towards this work since their aspiration is to get a government job or other regular employment. Full-timers like Ghani had more stake in the strike but were afraid that if they became too vocal about their issues, their IDs might be blocked and they would lose one of the few earning opportunities available to them during the pandemic. Ghani summarized his hesitancy to participate in the strike by saying, ‘Existing jobs are getting lost. Lene ke dene padjate [instead of receiving, we have to give]’.
There was no clear distinction in the background of strike participants—both full-timers and part-timers participated and did not participate. The findings of Veen et al. (2019) that ‘transient’ or ‘short-term’ workers tolerate poorer working arrangements were partially true in my case, as some of these workers participated in the strike and few were involved in organizing it. Part-time workers were hesitant to strike as they tended to be students who feared that visibility would negatively impact their future prospects. Many full-time workers were reluctant to strike too because this work is their sole source of income and they feared being blocked and losing it altogether. Heterogeneity is a structural characteristic of this workforce that makes building of associational power more challenging for the workers.
Stigma and Fear
The stigma that is attached to this type of work, as stated above, is an additional factor that curtails resistance in the Indian context. Several striking workers were afraid of the visibility that any kind of leadership position would give because of the stigma attached to the work as well as because of fear of being singled out by management and being blocked. Prashant, who worked part-time, provided backstage support to the strikes but was unwilling to picket at a restaurant himself as that would make him more visible. But some part-timers were sympathetic to the issues and took on leadership roles in the strike. Chaitanya was one such part-timer who organized the strike in his zone. He explained:
There are some better-off workers than me that I know who are working here who are much more capable of taking leadership of delivery workers and understand the issues, but they are hesitant to come forward as their reputation will be negatively affected in their families or if the police book some cases it will cause damage to their careers.
While protesting, the striking worker had to make sure that they were not visible to their families, restaurants, the platform or the state to avoid negative repercussions. Workers feared that when they picket at restaurants, their acquaintances could potentially identify them as delivery workers, which in turn could give them a negative image within their families. Unmarried workers were afraid that this information could hamper their marriage prospects. Restaurants could inform the platform or call the police. CCTV cameras at the restaurants became the eyes of the platform or the state and could be used as proof of indulging in strike action. Workers were also vulnerable to termination from the app by the platform or being booked under a legal case by the police for participating in strikes. These considerations created considerable fear amongst delivery workers and hampered their participation, especially for full-timers like Ghani. Stigma, coupled with fear, therefore, dissuaded workers from participating in the strikes. Workers have to navigate the tension between visibility and invisibility while participating in the strike.
Incentives
Platforms, at their discretion, pay delivery workers additional earnings as incentives. Early on, incentives were used to attract delivery workers onto the platform. Though incentives have become less rewarding over time, they continue to play a crucial role in ensuring high productivity by inducing workers to log on and extend their working hours (van Doorn & Chen, 2021). Reducing incentives after an initial period of high incentives is a common feature of the platform business model. The gradual restructuring of incentive structures enhances the precariousness of workers (D’Cruz & Noronha, 2022; Fleitoukh & Toyama, 2020). During the Hyderabad strike, as was the case in the UK and Italy (Tassinari & Maccarrone, 2020), workers were offered incentives by the platform to break the strike and induce them to return to work.
Along with using threats, platforms deployed short-term incentives to further divide an already heterogeneous workforce. At the strike’s peak, incentives as high as ₹1,000 were offered merely for working for one day. During the second strike, workers were already wary of the potential use of incentives to break the strike. As the lack of income began to bite, a section of workers were lured back to work by incentives and reaped their benefits. These actions made it difficult to sustain the novel attempt at collective action. Ghani, realizing the futility of the second strike, thought it would be an excellent opportunity to make use of the incentive that was offered. While some workers feared losing their job, others were attracted by the incentives.
Limitations in Unionization Efforts
When asking Shaik Salahuddin about the unionization efforts in the delivery sector, he explained:
[Swiggy] has successfully created fear amongst workers and is causing deterrence for them to be a union member. Any worker involved in union activities is vulnerable to suspension. It is difficult to conduct membership drive or raise contributions. So how can we mobilize in this context?
During the pandemic, the already short recruitment and training process was further shortened and put online to ensure that COVID-19 restrictions did not hinder recruiting new workers willing to take the risk. Constant recruitment, along with high attrition (turnover) rates in this sector, make it difficult to organize workers: Those familiar with the work conditions leave regularly and it takes time for new workers to become aware of the platforms’ predatory practices.
During the negotiation with the Fleet Manager, striking workers (one of them was Naji) were asked to speak about their own problems rather than collective problems. As with retail workers (Gooptu, 2009), food delivery workers were offered individual solutions to collective grievances. In the face of structural problems, they were told to leave the work if they did not like it. Many other new workers proved willing to take their place. The two Hyderabad strikes showed workers that organizing a strike is possible. Still, they also experienced a sense of futility and were exposed to risks associated with strike action, like being barred from work or booked under police cases. Today, food delivery workers’ anger against exploitative work conditions remains, but it is not translated into strikes or disruption. In the IFAT WhatsApp group, workers from different parts of India sometimes discuss organizing a pan-India strike, but it is yet to materialize.
Conclusion
This article contributes to the globally emerging literature on food delivery workers by providing an ethnographic account from India of the challenges delivery workers face in fostering collective action. Although global just-in-time logistics are vulnerable to disruption at various nodes or choke points (Bonacich & Wilson, 2008; Moody, 2019), in the case of hyper-local logistics, such as food delivery, workers have little power to disrupt operations. Mobilizing a diverse and spatially dispersed workforce to prevent food pickups at several high-order volume restaurants in different localities proved challenging (Vallas & Schor, 2020).
This article demonstrates that the attempts at collective action amongst Hyderabad’s delivery workers were hindered by a combination of factors. One factor was the heterogeneity of the workforce, shaped by the diverse aspirations, commitments and interests in platform work, which made concerted action very difficult to organize. A second factor was the specific context of the pandemic, during which the need to hang on to any available job became more pressing than ever before. This made workers who might otherwise have been willing participants, like Ghani, more hesitant to participate in an action that could jeopardize their sole source of income. A third factor regards the nature of platform work itself—the spatially dispersed features of which make it particularly challenging for workers to halt work across the sector. The absence of a single site, office or factory makes it near impossible to down the tools all at once. Finally, there is the very nature of the platform model itself, which allows platforms to reject any responsibility for their ‘executives’, while using a string of tactics to break all forms of resistance, ranging from blocking IDs to offering incentives or using threats and app-based surveillance tools to onboarding new recruits. Even a homogenous labour force would struggle to organize under such conditions. Platform workers are therefore left with a weak bargaining position and poor work conditions. Workers face a dilemma between work and strike. Their attempts at collective mobilization did not bring about any lasting changes. The pandemic and the strikes that followed revealed the paradoxical reality of food delivery workers being ‘essential’ and yet unable to use their position to improve their work conditions. The support of the people was more visible during the August 2020 strike in Chennai, when a video of a delivery worker went viral on Twitter, following which the company released a statement that the workers are fairly paid and that fare reduction is misunderstood.
In the absence of a strong union, workers nonetheless seek to organize into loosely structured formations that articulate their struggle (Iazzolino, 2021). WhatsApp groups help in ‘developing shared consciousness’ (Tassinari & Maccarrone, 2020, p. 45) and ‘strengthening a sense of collective identity among dispersed’ (Pasquier & Wood, 2018, p. 1) delivery workers, but such groups are easily infiltrated by management-friendly workers who identify the main striking worker and pass on information to management. The management uses such information to threaten the striking worker with a police case.
Providing the ‘essential services’ tag to food delivery reveals the important role Swiggy played during the pandemic. This tag further enhanced the platform’s role of serving the public while obscuring its commercial interests (Dijck et al., 2018). The pandemic has further strengthened Swiggy’s market position and made visible how dependency on the platform has increased for customers and restaurants to access each other and for workers to access a basic livelihood. But the essential tag, which gave legitimacy to the companies to operate while symbolically praising workers as heroes, did not translate into better work conditions; instead, it allowed things to get worse. Swiggy successfully reduced the earning and incentive structures during the pandemic, even when fuel prices and the cost of essential commodities were increasing. Food delivery platforms engaged in exploitative work practices while charging restaurants high commissions, promoting anti-competitive practices and forcing discounts. While workers’ collective efforts to resist these changes were largely in vain in Hyderabad and other parts of India, this article has drawn on their experiences to highlight some key factors that continue to thwart collective action in the platform economy.
Footnotes
Acknowledgements
I am thankful to the anonymous reviewers and Carol Upadhya, Geert De Neve, Nandini Sundar, Kaveri Medappa and Rebecca Prentice for their feedback.
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Funding
The author disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This work was supported by the IEG-ICSSR Doctoral Fellowship 2017-19 and IJURR Writing Grant-2022.