Sage Journals: Discover world-class research

Abstract

This article explores how Uber drivers in Dhaka exercise agency to earn and sustain their livelihoods. Uber drivers not only experience extortion by Uber, but also face various challenges, such as precarious working conditions and algorithmic control of their activities. In most Global South countries, the regulatory practices are not in favour of Uber drivers either. Within this context, drawing on in-depth interviews with 27 Uber drivers and one focus group discussion with members of the Dhaka Ride-Sharing Drivers’ Union, this article makes an original contribution to the discussion of the algorithmic governance of labour and how gig workers in the Global South subvert algorithms utilising everyday tactics to earn their livelihoods. While existing literature mostly demonstrates that Uber drivers use bot apps or switch off their app to protest Uber’s algorithmic control, this article shows that Uber drivers in Bangladesh utilise a different strategy known as khep (contractual ride) to subvert the algorithmic governance of labour and increase their income. Through the presentation of Uber drivers’ overt and covert resistance strategies in Dhaka, this article also advances the theories of the labour movement and industrial relations and the gig economy literature by demonstrating that overt and covert resistance strategies can complement each other if workers are unable to legally form unions to bargain with platforms and claim their rights for fair pay and fair working conditions.

Keywords

algorithmic management Bangladesh digital labour platform gig economy labour movement Uber

Introduction

The ‘gig’, ‘sharing’ or ‘platform’ economy has recently attracted increasing scholarly attention, particularly in how intermediary platforms build, connect and reconstruct the social relations among labourers, consumers and companies (Anwar & Graham, 2021; Arcidiacono et al., 2018; Lata et al., 2023; Popan et al., 2023; Rosenblat, 2018; Schor et al., 2023). The gig economy can offer ample employment opportunities for lower-income groups offering them jobs in transport and courier sectors (Schor & Vallas, 2021). However, several studies reveal social and economic inequalities are exacerbated in various features of the gig economy (Lata & Copolov, 2025; Rosenblat, 2018; Schor & Vallas, 2021). Critics have further pointed out the problematic features of algorithmic management that has shifted managerial responsibilities from humans to machines (Aloisi & De Stefano, 2022; Lata et al., 2023). Platforms such as Uber, Pathao (‘send’) – a local ridesharing app, Didi, Deliveroo, Menulog and DoorDash utilise digital surveillance to monitor workers through their smartphones and mobile devices. Within this context, this article examines the nature of gig work and gig workers’ associations in Bangladesh and how gig workers manage algorithmic control of labour in their everyday lives.

Across the Global South, governments have encouraged digitalisation and the spread of the gig economy with the hope that it would generate more employment opportunities for workers (Rani et al., 2022). This is particularly important given the fact that the vast majority of workers are engaged in the informal sector in the Global South (OECD, 2023). In the Global South, although the structural adjustment policies of the 1980s and 1990s significantly reduced government investment in technological innovation, the new technological boom seems to make up for the lost decades of neoliberalism (Rani et al., 2022). Digitalisation has also been promoted by international development agencies presenting this as a ‘policy pathway towards sustainable, inclusive and equitable economic growth’ (Rani et al., 2022, p. 421). In 2021, approximately 14 million active workers were working on the five largest English-speaking online web-based platforms alone (Kässi et al., 2021).

This article makes two key contributions to the literature on the labour movement, the sociology of work and organisation, industrial relations and the gig economy. First, this article makes an original contribution to the discussion of the algorithmic governance of labour and how gig workers in the Global South subvert algorithms utilising everyday tactics to earn their livelihoods. While existing literature mostly demonstrates that Uber drivers use bot apps or switch off their app to protest Uber’s algorithmic control, this article shows that Uber drivers in Bangladesh utilise a different strategy known as khep (contractual ride), where they tell their customers to cancel the ride and offer a lower charge than the Uber app shows to drop them off to their destination. They also make an informal contract with them by exchanging their mobile numbers so that the customer can contact them without using the Uber app. In this way, they avoid Uber’s commission rate. Second, through the presentation of Uber drivers’ overt and covert resistance strategies in Dhaka, this article advances the theories of the labour movement and industrial relations and the gig economy literature by demonstrating that overt and covert resistance strategies can complement each other if workers are unable to legally form unions to bargain with platforms and claim their rights for a fair wage and fair working conditions. Existing literature mostly ignores these complementary relationships between covert and overt resistance strategies and how that can assist workers in claiming their rights.

The remainder of the article is organised into five sections. The first section provides contextual information on the gig economy in Bangladesh. The next section outlines key ideas about algorithmic control and labour agency in the gig economy. The third section provides a detailed discussion on data collection and analysis process. The next section presents findings on Uber drivers’ resistance strategies to subvert algorithmic control and to claim fair pay and fair working conditions. The final section presents the discussion and conclusion.

Gig economy in Bangladesh

Like other Global South countries, Bangladesh has stepped into using digital platforms despite the challenges gig workers experience while working in the gig economy. Bangladesh’s platform-based gig economy got its momentum in 2016, with the arrival of Uber. The gig economy gained popularity in Bangladesh as it offered a new source of employment for the large number of informal workers. Nearly 60 million people or 84.9% of total working population in Bangladesh work in the informal sector (Bangladesh Bureau of Statistics, 2022). The informal sector, including the gig economy, contributes around 43% of Bangladesh’s GDP. The ridesharing industry has contributed the largest segments, worth USD 259 million, and this sector represents 23% of the transportation sector. The estimated market value of ridesharing start-ups in the country is likely to reach USD 1 billion within next five to seven years (Fairwork, 2022). Following Uber’s footsteps, many new platforms have started their operations in the market such as Truck Lagbe (transportation), Foodpanda, Paperfly (food delivery) and Hello Task (domestic work app). However, like many other countries, gig workers are not recognised as ‘employees’ in Bangladesh.

Like other countries, the ‘contract workers’ model is predominant in the gig economy in Bangladesh. Most digital labour platforms consider platform workers as ‘independent contractors’, which means platform workers are not their employees. Due to the independent contractor status, gig workers’ work status remains legally ambiguous in Bangladesh and they have not received any attention by regulators yet. Although the government issued the ridesharing guideline to regulate ridesharing companies, the guideline does not provide any clear legal guidelines or protections for gig workers. While the guideline deals with the question of what constitute ridesharing services and liabilities of ridesharing services, such as ridesharing companies having to pay tax to the government, the guideline does not mention anything about gig workers’ employment rights other than mentioning that ‘the ridesharing company will have to have an agreement with the vehicle owner and the driver, detailing rights of all sides, and any party can withdraw from the agreement by giving one-month prior notice’ (Lata, 2024, p. 96). The existing Bangladesh Labour Act’s ‘classification of workers’ list includes other non-typical forms of employment, for example, apprentices, seasonal workers, casual, temporary probation and permanent workers but it does not explicitly recognise informal workers including gig workers (Fairwork, 2022). The platform companies are able to deny their responsibilities to ensure a minimum wage or equivalent income or other employment benefits such as leave and pension by using the classification of independent contractors. In Bangladesh, both Uber and Pathao follow this ‘independent contractor’ model and do not provide platform workers any employment benefits.

Platforms’ classification of workers as independent contractors has further implications for their ability to collectively organise and make demands (Lata, 2024). Due to their independent contractor status, Bangladeshi gig workers are not legally allowed to form a union. According to the Bangladesh Labour Act 2006 (2024), Section 176, only workers are allowed to form trade unions. Historically, the informal sector largely operates outside the purview of the state. A recent study by the Bangladesh Institute of Labour Studies reported that nearly 88% of informal workers do not receive any letters of appointment (Khan, 2020). Due to this trend, the informal economy has not really promoted a decent work tradition and this has impacted the legal claim of workers’ rights in the gig economy in Bangladesh, which largely operates like an informal economy model.

Bangladesh has a tradition of suppressing any kind of social movement using police and ruling political party supporters (Lata, 2023). Consequently, workers are in general not very enthusiastic to form unions. Due to this factor, even in the garments sector, there are few factory-based unions. There is a Bangladesh Road Transport Workers’ Federation for truck, bus and automobile workers. However, ridesharing drivers are not part of this federation. There is no taxicab workers’ union in Bangladesh because the taxicab sector is run by private companies and the drivers have not formed any association to date. There is an Association of Taxicab Owners of Bangladesh but workers are not part of this association (FBCCI, 2024). The taxicab sector is also on the verge of extinction because of the rise of Uber and Pathao (Akhter, 2022). If no company invests in the taxicab sector by 2025, the service will be abolished (Akhter, 2022). Existing research also demonstrates that most unions, including informal workers’ unions, Bangladesh Road Transport Workers’ Federation, trade unions and National Garment Workers’ Federations, are closely allied to political parties and often face widespread accusations of corruption and self-seeking (Hossain, 2024). As protests are heavily repressed by the state, in most cases informal workers utilise covert resistance strategies in Bangladesh (Lata et al., 2019). Due to the existing socio-political context, it is difficult for gig workers to form a union and use that union to bargain with the platforms.

Recently, the ridesharing drivers have started protesting ridesharing platforms’ policies, including low wages, by forming an online workers’ union known as the Dhaka Ride-Sharing Driver’s Union (DRDU). Since Uber’s arrival in Dhaka in November 2016, Uber drivers had sporadically protested Uber’s regulatory measures. However, they were not well organised until they formed DRDU on 21 August 2019, after an Uber driver was murdered by three young men in Dhaka. The driver, Md Arman, received a ride request in June 2019 from a remote location in the outskirts of Dhaka. When he arrived there, three young men got into his expensive Toyota Allion car and tried to hijack it. When Arman resisted, they grabbed hold of his hair, slit his throat and the car collided into a tree. Uber offered a pittance after Arman’s death. The fellow Uber drivers brought litigation against Uber and demanded 5,000,000 Bangladeshi taka (BDT) as a compensation for Arman’s family. However, Uber offered just 200,000 BDT which is around USD 2,325. This is a small amount of money and this is a ‘grossly inadequate sum for someone’s life’ (App-Based Drivers Union of Bangladesh, 2022). Arman’s death was the catalyst for forming DRDU.

Following the formation of DRDU, the DRDU leaders and their members announced a set of six demands in September 2021: (1) ending police harassment; (2) recognition of app-based drivers as workers under the local labour law; (3) lowering the transition fee of rides from 25% to 10%; (4) organising parking spots for Uber drivers in Chittagong, Dhaka and Sylhet; (5) exempting registered ride-hailing vehicles from advance income tax (AIT) and (6) returning the AIT collected from registered vehicle owners previously (App-Based Drivers Union of Bangladesh, 2022). Uber drivers also pointed out that the transaction fee was an important issue for them as their financial situation would improve with a lower transaction fee. Other concerns raised by drivers included the use of gaming techniques that govern the consumers’ and workers’ ratings. This is particularly harmful for drivers’ physical and mental wellbeing (App-Based Drivers Union of Bangladesh, 2022). However, Uber has not responded at all to the drivers’ demands; rather, in some cases Uber charges up to 35% of drivers’ fare in transaction fees.

Uber also does not provide any explanation to drivers when it deactivates drivers’ accounts. This is a common complaint among Uber drivers all over the world (Holtum et al., 2022; Wiener et al., 2023). Following the Uber drivers’ strike action in Dhaka, Uber had made some changes to their app. For example, the Uber app now indicates the rough location of the next trip if drivers complete five trips out of 10 requests. This help Uber drivers to better plan their workday and navigate the traffic situation in Bangladesh. Uber has recently further updated its app and drivers can now see the ride destination and payment modes before choosing to accept a ride (The Daily Star, 2022). Previously, local Uber users grumbled about Uber drivers cancelling their rides. In their defence, the drivers complained that they could not see the destinations and payment modes before accepting a ride (The Daily Star, 2022). That’s why they cancelled their rides. The new policy has solved this problem.

Currently, Uber drivers in Dhaka are also connected with Uber drivers in Chittagong and Sylhet. Consequently, the drivers from Chittagong and Sylhet often join physical protests and sit-ins along with Uber drivers in Dhaka. DRDU has provided Uber drivers in Bangladesh a strong platform to formulate common demands as well as to respond to the particular needs and concerns of city-level unions (App-Based Drivers Union of Bangladesh, 2022). Within this context, this article explores the nature of gig work and gig workers’ associations in Bangladesh and how gig workers manage algorithmic control of labour in their everyday lives.

Gig economy, algorithmic control and resistance in the gig economy

The gig economy has positioned itself as a globalising opportunity creator for entrepreneurs across borders. Digital intermediary platforms, like Uber, fuel the capitalist narrative that hard work and perseverance will achieve a net positive outcome (Peticca-Harris et al., 2020). This type of marketing is attractive to workers that are condemned to the fringes of the labour pool, like women, young people and migrants (Van Doorn et al., 2023). However, digital labour platforms have shifted enterprise risk onto workers through labour relationships and secretive algorithms (Tassinari & Maccarrone, 2020). Digital intermediary platforms force workers to rely on the technology, with little understanding of how it works. This method allows for platforms to retain the power within the employment relationship, with workers having very little bargaining power.

Algorithmic control of workers is a key feature in emerging on-demand platform apps. These control mechanisms operate within a ‘continued trajectory of neoliberalism’ (Veen et al., 2019, p. 389) and utilise an ‘increased hybridity of control’ (p. 391), where capital has blended control mechanisms. The implementation of performance metrics also creates hierarchical and asymmetrical power relations that serve as a panopticon. Performance metrics based on client feedback are utilised as audit measures to continuously push workers to self-optimise (Heeks et al., 2021). Workers are always at a disadvantage due to information asymmetry, which reduces their collective bargaining power within platform economies (Rosenblat, 2018). On-demand platforms vary in how they utilise these control mechanisms and are spatially and temporally contingent. The nature of algorithmic management allows for little transparency between workers and customers (Rosenblat, 2018). Workers often receive negative ratings for things out of their control. Platforms also engage and expand algorithmic control across a broad spectrum, meaning that there are variations of experiences between workers and control mechanisms. Using algorithms in labour control processes also allows for decreased accountability in firms’ decision making: having essentially outsourced management to an algorithm, firms can work within legal grey areas of employment contracts. Human bosses are subject to unfair dismissal policies and algorithms are not. This further acts as an asymmetrical process in which workers have little ability to negotiate the system fairly (Walker et al., 2021). Not only are workers disadvantaged in the system as they cannot predict the algorithm, but they are also given few to no avenues to have meaningful dialogue with platform companies, as management is mostly outsourced to algorithms. Additionally, due to a lack of employee status, gig workers have little agency or bargaining power with platforms like Uber.

In general, the term ‘agency’ refers to the capacity to act, intervene or exercise power to reshape dominant structures and create alternative futures (Holloway et al., 2019; Wells et al., 2021). To gain a better understanding of Uber drivers’ political agency and outcomes, it is important to consider both overt and covert (Scott, 1985) resistance strategies employed by Uber drivers. Existing literature shows that resistance has multiple and varied forms and collective resistance is not the only form that Uber drivers employ to resist the platforms and the state; rather they utilise a mix of overt and covert resistance strategies to claim their rights to a fair wage and decent working conditions (Anwar & Graham, 2020; Wells et al., 2021). Worker agency often depends on the structural conditions or context upon and through which agency is exercised (Scott, 1985; Wells et al., 2021). In his work on peasants in Malaysia, Scott argues that the form of resistance depends on relations of power between subaltern and dominant groups. Scott (1985) has categorised two main forms of resistance: everyday resistance and a more direct open confrontation. In direct confrontation, one seeks the ‘formal de jure’ recognition, whereas in everyday resistance ‘tacit de facto gains’ are prioritised (Scott, 1989). Direct resistance evolves into a pattern of resistance that often demands drastic change of social structures, whereas everyday forms of resistance are small scale, and require little or no collective mobilisation and are often hardly recognised by the group being resisted.

Scott (1985) introduced the term ‘everyday resistance’ to describe ‘quiet, dispersed, disguised or otherwise seemingly invisible’ resistance; this is in contrast to dramatic, visible, and often collective or confrontational forms of resistance. Scott (1990) refers to this form of resistance as ‘infrapolitics’. Infrapolitics refers to a form of politics where individuals or groups adopt covert resistance strategies to go beyond the gaze of authority (Scott, 1990). This kind of resistance strategy includes sarcasm, escape, passivity, laziness, disloyalty, misunderstandings, slander, avoidance or theft. Subaltern groups employ these forms of covert resistance when they find it is too risky to confront dominant groups directly (Scott, 1985, 1990). Similar kinds of resistance strategies are explored among Uber drivers in Dhaka when they subvert algorithmic control utilising various everyday practices.

For a study of the gig economy and resistance against the gig economy, the above literature provides an excellent guideline to analyse gig workers’ agency in the gig economy and how their struggles are often shaped and reshaped by the wider socio-economic and political conditions. Recent research also shows that gig workers either employ overt resistance strategies by being part of an existing union (e.g. the Independent Workers’ Union of Great Britain [IWGB]) or forming their own union or covert resistance strategies and often their covert resistance strategies are individual instances of coping or what Scott (1985) refers to as ‘everyday resistance’ (Riordan et al., 2023; Wells et al., 2021). Although gig workers utilise covert and overt resistance strategies, these strategies and the strength of utilising covert and overt resistance strategies vary in the Global South. For example, gig workers in Indonesia, China and Lagos were able to manipulate ridesharing apps’ (Uber, Didi, Gojek and Bolt) algorithmic management using bot apps and third party apps (Arubayi, 2021; Chen, 2018; Mustika & Savirani, 2021). In Lagos, drivers also learnt to create surge areas following examples from the US Uber drivers (Arubayi, 2021). These strategies are considered as covert resistance strategies as they do not directly protest ridesharing platforms’ oppression. Within this context, in this article I examine how Uber drivers in Dhaka utilise both overt and covert resistance strategies and how these strategies complement each other to protest the platform and its algorithmic control mechanisms.

Methodology

The data for this article were drawn from 27 qualitative interviews (see Table 1) with Uber and Pathao drivers including two leaders of App-based Workers’ Federation of Bangladesh and one focus group discussion (FGD) with 10 members of the Dhaka Ride-Sharing Drivers’ Union (DRDU). I used the convenience sampling approach to recruit participants due to the nature of gig work. As gig workers are hard to reach unless we use their services, I chose to recruit them using in that way. Recruiting ridesharing drivers in this way is known as ‘ride alongs’ (Holtum et al., 2022). I recruited 25 Uber and Pathao drivers through trips taken on the Uber and Pathao platforms and two drivers were recruited using the DRDU Facebook group page. All FGD participants were part of the DRDU group. The former DRDU president invited them to participate in the FGD. I contacted the former DRDU president using their Facebook page. When I met my participants, I introduced myself as a university teacher and provided them with the participant information sheet. I also told them about my research project before riding in the car. I informed them that their ratings would not be impacted if they did not participate in this study. Teachers are trusted and respected in Bangladesh. So, my participants trusted me and provided information. Some participants chose not to answer some questions as I explained to them at the beginning of the interview that they did not have to answer all my questions if they did not feel comfortable answering them.

Table 1.

Summary of participants.

Participants	No of participants	Union membership
Interview participants
Uber drivers	12	3
Pathao drivers	7	0
Uber and Pathao drivers	6	0
Members App-based Workers’ Federation of Bangladesh	2	2
Focus group discussion (FGD) participants
FGD with DRDU members (Uber drivers)	10	10
Total	37	15

For this article, I mostly used data from drivers who either work for Uber or both Uber and Pathao. The interviews and FGD were conducted between June and July 2023 and combined with participant observations in July 2023 where I observed drivers during several ‘ride-alongs’. The semi-structured interviews lasted between 60 and 90 minutes and the FGD lasted nearly three hours. All participants were male as there were no female ridesharing drivers in Dhaka. All participants were given a gift card (valued at 1000 BDT, which is equivalent to a gig worker’s three hours of income) as an appreciation of their time and participation in the study. The ages of drivers ranged from 19 to 60. Most participants were from a lower socio-economic background. However, some participants were from a lower middle-class background. This project received ethical clearance from the author’s university Ethics Committee.

The interview checklist includes information on participants’ entry into the gig economy; motivations for joining ridesharing (Uber and Pathao) platforms; employment history; length of current employment; understanding of how ridesharing apps work; benefits and challenges of working in the gig economy; the role of social networks in accessing and navigating the gig economy; information on their involvement with DRDU or other similar organisations; migration history; age; income and expenditure; education; and strategies utilised to navigate the gig economy.

The data were analysed following an open-ended abductive approach based on iterative stages of the thematic coding of the interview. Thematic analysis (Braun & Clarke, 2006) is used to systematically identify, synthesise and organise data that offer patterns of themes or meanings across a given data set. I also used an abductive approach to analyse data, which allows researchers to navigate between theory and participants’ accounts, allowing each to inform the other to answer research questions within the project (Cunliffe, 2011). I read the interview transcripts several times and coded all interview transcripts following this approach. Pseudonyms are used for all participants except Rajesh Khan, the former president of DRDU, who provided consent to use his original name.

Algorithmic control, precarious work and everyday resistance

To explore Uber drivers’ understanding of algorithms and whether they can subvert algorithmic management, I asked them whether they knew how Uber allocate rides to drivers and whether they had any idea why some drivers got better trips than others. In response to these questions, most Uber drivers said that it depended on their ratings. If they had ratings near five, they got good trips. Sometimes, it also relied on whether the car is rated as premium quality or Uber X. As Rafi said:

When my car was in premium quality, I used to get good passengers. That’s why my rating didn’t fall below 5. Ratings depend on the passengers and how they rate the driver. If a passenger gives 1 star or 2 stars after the ride, it can reduce the driver’s rating. Sometimes, passengers give 1 or 2 stars for reasons the driver may not be aware of. My rating was never below 5, but since my car quality dropped in UberX, my rating has started to decrease. Now my rating is 4.92. Many passengers don’t rate, which has also contributed to the decrease . . . No, there is no relationship between the rating and the fare. As far as I know, cars with good ratings have higher demand. If the rating is not good, many passengers cancel the trip and get a new ride. (Rafi, Uber driver)

Due to the ‘black box’ nature of algorithmic management, most Uber drivers initially did not understand how the Uber app operated. So, they asked other drivers and got to know how the rating system worked. Existing research also showed power imbalances between workers and Uber’s management due to Uber’s algorithmic management strategy (Rosenblat, 2018). Uber drivers also commented on this general power asymmetry, pointing out the lack of transparency on how different platforms’ systems function. Specifically, they had no idea what data were used in the algorithmic decision-making process and how they were weighted:

Everything is determined by the Uber app. We just had to follow whatever it tells us. It sometimes deducts money from us because they provide bonus to customers. Once it happened with me. I didn’t understand why the app was showing lower fare than the initial fare. So, I called the Uber office and they said they would adjust my payment later. (Nasim, Uber driver)

To subvert algorithmic control, Uber drivers utilised a range of everyday or covert resistance strategies. For example, they sometimes told their customers to cancel the ride so they could pay a lower fare for their rides than the Uber app showed them. Drivers also mentioned that they sometimes made informal contracts with some customers where they shared their phone numbers with customers and customers called them when they needed a car. Locally, this practice is known as khep. In this way, they made some extra income. Although existing studies showed that Uber drivers switched off their app to increase the ride price (Wells et al., 2021) or Gojek drivers used ‘ghost accounts’ – hacked versions of the Gojek application in Indonesia (Mustika & Savirani, 2021) – none of these studies documented khep practice employed by ridesharing drivers.

Apart from khep practice, many participants mentioned that they used multiple apps as well as provided a rent-a-car service to supplement their incomes as the amount of money they earned driving Uber was not sufficient to maintain their families and housing rents. This finding aligns with existing literature that demonstrated that ridesharing drivers often used multiple apps to increase their income (Cameron, 2022).

However, algorithmic management was not the only obstacle Uber drivers experience in earning their livelihoods; they also described how Uber exploited them, charging a high rate of commission from their earnings. Initially, Uber drivers were able to earn a good amount of money as Uber offered them lucrative bonuses. Many drivers who had been working for Uber since 2016, said that they earned a good amount of money in the beginning. This is because of Uber’s marketing strategy. Uber offered them many bonuses during that time. However, Covid-19 changed this scenario.

Uber and Pathao were allowed to restart business on 4 September 2020, following the suspension of ridesharing services during the Covid-19 lockdown in Bangladesh. Since then Uber has not offered frequent bonuses to their drivers. Consequently, Uber drivers found that the commission rate was too high. As not all drivers owned a car, drivers who used rented cars earned a small amount of money after providing 50% of their incomes to the owner of their car. Or similarly, some drivers drew up a contract with the owner of their car and they had to pay 30,000 BDT per month to those owners. Most Uber drivers showed me their calculations of how much they earned after paying their owners and other regular expenses they had to pay to maintain the car. The majority of the participants were not happy with the amount of commission Uber charged. Uber took 25% commission. Uber drivers told me how they earned very little once Uber deducted their commission:

I haven’t been driving Uber for a few days. I’ve decided not to use Uber anymore. I had another car, but I sold it. I do not like Uber because it caused too much trouble. I’ve calculated that they take a 27% commission. Whatever amount the customer pays me, Uber deducts 27% from it . . . If I calculate after deducting my expenses, I have 30 BDT. Uber takes 27 BDT from there. So, how much do I earn? I have nothing. Many Uber drivers stop driving after a while because they’re not left with anything. (Touhid, Uber driver)

Apart from earning low incomes, drivers mentioned a number of other harassments they experienced while driving for Uber. For example, the police regularly harassed them and charged them fines for parking their cars in places where parking was not permitted. Nahid also mentioned how a police officer had threatened him that he would file a case against him for dropping a passenger at their destination. To avoid that case, he had to pay 1000 BDT. Another Uber driver said:

Our Prime Minister says Bangladesh is a digital country, but there is no trace of digitalization in the transport sector. Traffic police trouble us when we stop on the road, even if you talk to them, they don’t listen. We can’t convey how oppressed and tortured we are from all sides . . . The police constantly harass us. When we pick up passengers or drop them off, they charge us with the violation of the parking rule and take 5000 BDT from us, or they take 1000 BDT as a bribe. (Nadim, Uber driver)

Police harassment was a recurrent theme in this study. Police asked for bribes even if Uber drivers showed them valid documents for driving Uber. This is a common problem that most drivers, including taxi and truck drivers, experience in Bangladesh. Police ask for bribes if these drivers break any road rules such as parking rules. A recent case also demonstrated that Uber did not provide any support to a driver when they experienced police harassment. Several members of DRDU mentioned a case where police had arrested an Uber driver because he had a passenger who was carrying illegal drugs. It was an intercity trip and the passenger’s destination was Dhaka airport. Uber drivers are not allowed to check their passengers’ luggage. That driver appealed to the police, saying that he did not know that his passenger was a drug dealer. However, the police did not listen to him and arrested him. He was still in jail when I spoke to DRDU members. Shahin, one of the DRDU members, said:

When he [the driver] arrived in Sonargaon from Comilla, the police stopped him at a checkpoint and his car was checked. The police arrested him along with the drug smuggler. When we heard this news, we went to the Uber office and the police station. However, the Uber staff did not provide any support for this driver. He is still in jail.

This incident demonstrates how Uber makes profits from these drivers’ labour and capital but provides almost no support when they experience this kind of challenge. Apart from police harassment, drivers also complained about Uber’s algorithmic management strategies, especially the way the algorithm penalises them by blocking their apps for not accepting all rides. Uber drivers explicitly mentioned that there were certain areas in Dhaka where they did not want to go due to heavy traffic or narrow lanes. There were many narrow streets in Dhaka where it was hard to drive a car, especially in many parts of old Dhaka. However, if Uber drivers did not accept three rides, Uber blocked their apps for a couple of hours. Also, some trips were too short and the location was problematic due to traffic congestion.

Many Uber drivers also boycotted Uber and moved to Pathao due to Uber’s high commission rate. One of the strategies Uber drivers started to utilise to increase their incomes was to use multiple apps. A number of Uber drivers started using Pathao after the pandemic in opposition to Uber’s high commission rate, as Pathao’s rate was lower than Uber. Fayaz, who recently started driving for Pathao, told me that he stopped driving for Uber due to its high commission rate:

Pathao does not take any commission for the first 15 days. Then they get 5% commission for next 30 to 45 days, and then 10% for 30 days. After that, if they see that I drive the car regularly, they charge 20% to 25% commission. Due to Pathao’s 0% commission rate for the first 15 days as well as its frequent bonus offers, many Uber drivers joined Pathao. (Fayaz, Uber driver)

The discussion so far suggests that Uber drivers are able to use their agency at an individual level, which can be characterised as everyday resistance (Scott, 1985). However, drivers also utilise a number of collective strategies to bargain with Uber and to claim their rights.

Overt resistance in the gig economy

DRDU operated on social media platforms such as Facebook and Twitter and was not formally registered as a union. The Uber drivers created an online Facebook group so that they could report each other’s whereabouts at night by sending text or voice messages:

When the exploitation of the ride-sharing companies like Uber reached its peak, we organised a meeting using our Facebook page. We organised a meeting in front of the China Friendship Conference Centre building. About 100 of us gathered there that day. From there, we decided that we would conduct a peaceful human chain. We formed a human chain in front of the Uber office. Even after the human chain, Uber didn’t pay heed to our words . . . On that day, we came up with the idea of forming an organisation and from that day, our organisational activities began. From then until now, we have been organising protests and campaigns against Uber. (Rahim, DRDU central committee member)

DRDU members also mentioned that they were successful in a protest that they organised against the government’s initial enlistment certificate issue. In 2019, three ride sharing companies, Uber, Pathao and Shohoz, got enlistment certificates from the Bangladesh Road Transport Authority (BRTA) to operate their service in Bangladesh. However, initially BRTA passed the rule that ridesharing drivers could only drive for one ridesharing company. DRDU members protested against this rule. Following their protests, BRTA changed the rules and drivers can use multiple apps now:

We received an enlistment certificate from the government, where initially it was stated that a car can only provide rides for one company . . . To change this law, we went out on the streets during the COVID-19 time. There was an assembly of almost two to three thousand people in front of the press club, which, in hindsight, was a health risk at the time. Overlooking that health risk, we were forced to go on the streets because this law was a bigger risk for us. If there is no competition, our customers will suffer from a lack of good service and we will be forced to work for a single ridesharing platform. So, the government showed some leniency here. (Helal, ex-DRDU member)

Although BRTA accepted the drivers’ demand, Uber has hardly changed any of its policies in consequence of Uber drivers’ movements. If the drivers keep protesting and stop driving Uber, Uber might have to negotiate with the drivers in the future

During the FGD, DRDU leaders mentioned that they established four zonal committees in Dhaka in 2021. The zones cover Dhaka metropolitan city’s west, east, north and south parts. Each committee had around 30 members. Members of the zonal committees were also part of the central committee. However, DRDU members also mentioned several challenges in continuing their protest activities. One of the issues is that most Uber drivers were poor, and as a result it was mostly only the central and zonal committee members who could pay the membership fees. Lack of funding made it difficult for them to organise different activities. DRDU members also mentioned that they failed to register their organisation due to the last government’s registration ban. Rajesh Khan, the former president of DRDU, explained why this happened:

The registration for our organisation primarily needs to be done under the Ministry of Labor and Employment or the Social Services Department. However, currently, the government has completely suspended the registration of any organization. For almost three years now, we have been trying to register without success. (Rajesh Khan, DRDU)

As DRDU was not formally registered, DRDU leaders also faced several obstacles from Uber’s staff when they organised protests in front of Uber’s head office in Uttora. When Uber drivers organised a peaceful protest in front of the head office on 5 January 2023, Uber officials locked their office and called the police.

DRDU leaders further mentioned that, due to Uber officials’ complaint to police, they recently mostly organised protests against Uber’s exploitative policies in front of the Dhaka Press Club. They got support from the media and media often covered their protests and interviewed DRDU leaders about their demands. However, Uber had not negotiated with drivers about any of their demands so far. For example, one of the protests was organised around providing Uber’s terms and conditions in Bengali. Uber had not paid attention to this demand. However, BRTA has strict guidelines that if a company comes to do business in Bangladesh the agreement should be provided in both English and Bengali. As Shaheen said:

In the guidelines of BRTA, it’s mentioned that if a company comes to do business in Bangladesh, the agreement should be provided in both English and Bengali. We’ve protested about this, but they still haven’t provided it in Bengali. They’re running it in English because they know we’re less educated and won’t understand it. What they’ve written, they say [i.e. Uber staff read out to them], and we just keep pressing one [consent] button after another. (Shaheen, DRDU member)

This quote further indicates how Uber drivers experience an additional challenge due to language barriers. This issue could be easily solved with a bit of a push from the government of Bangladesh. They could tell Uber to follow BRTA rules and provide a Bengali version of Uber’s terms and conditions. However, neither BRTA nor the Ministry of Labour has done anything so far. DRDU leaders also told me how none of the relevant government departments took any initiatives to support the claims of Uber drivers:

Since 2018 until today, I’ve only been speaking about this issue. There’s no minister, no official, no person whatsoever [who] listens to us a little and tries to work or discuss with the company. We went to BRTA and they assured us with different things. We went to the Tax Office, and they also provided assurance in the same way. In this corrupt country of ours, it will only keep getting worse. (Rajesh Khan, DRDU member)

This finding resonates with Lata et al.’s (2019) research with informal workers in Dhaka. Their study found that the government and its agencies always turned a blind eye when powerless citizens raised their voice for their rights. This finding is in direct contrast with many Global North countries, including the UK and members of the European Union (Adams-Prassl et al., 2021; OECD, 2023). Those governments made Uber follow their local rules as well as made the company recognise drivers as employees (Adams-Prassl et al., 2021), whereas the government of Bangladesh did not take the initiative even to change the language of Uber app. Within such a governance structure, Uber drivers had to rely on each other and their informal networks. Neither Uber nor the government provided any safety networks for Uber drivers. Consequently, DRDU members used Telegram – a cloud-based mobile messaging app – to provide support to each other. Uber drivers shared their live locations in Telegram. In this way, if any driver experienced any safety concerns, they could post a coded message in the Telegram group. Once the coded message was seen by other drivers, the nearest drivers gathered in that location to help the driver who was experiencing security concerns. Uber drivers told me about two cases in which their Telegram group played an important role in rescuing two drivers and their cars from hijackers who had ordered an Uber trip and then tried to steal the car. In another case, an Uber driver was arrested by police because their passenger was carrying illegal goods. Rajesh Khan went to the police station to bail the driver when they got his message. Finally, when I asked Uber drivers about their current demands, they pointed to the following:

Our first demand is to reduce the commission rate to 10%, change all the issues in the app, ensure our quality of life, enforce the labour laws and BRTA policies of Bangladesh, open all the IDs that have been blocked, and not block any ID in the future. If any ID is blocked due to any reason, then compensation for that driver’s loss during that time should be provided by the party responsible. According to the 2017 Ride-Sharing Policy, the amendments to the 2006 Labor Law of Bangladesh should be recognised as contractual labour and given legal recognition. The problems between Uber and the drivers should be resolved through BRTA. Uber should open a 24-hour support centre. (Rajesh Khan, DRDU)

Taken together, these findings suggest that DRDU is playing an important role in demanding Uber drivers’ rights. They are not only protesting Uber’s exploitative policies; rather, they are having conversations with several government departments to ensure that the public service staff are aware of the challenges they are experiencing in the gig economy sector. They are also aware of the importance of registering their organisation. However, as discussed earlier, due to legal barriers they could not register DRDU yet. Despite all these barriers, DRDU provided Uber drivers with significant labour mobility power to bargain with Uber and BRTA.

Discussion and conclusion

The gig economy has presented several challenges to labour agency through the intensified commodification of labour (Anwar & Graham, 2020; Wood & Lehdonvirta, 2021) and algorithmic control mechanisms (Aloisi & De Stefano, 2022; Dubal, 2023; Rosenblat, 2018). Resistance is particularly difficult in some Global South countries including Bangladesh due to a lack of unionisation and union power and because of the authoritarian character of the state. However, through an investigation of Uber drivers’ everyday lives and their resistance strategies, this article argues that despite the precarious nature of gig work and its algorithmic control mechanisms, gig workers are able to exercise their agency.

Through the presentation of Uber drivers’ overt and covert resistance strategies, this article advances theories of the labour movement and industrial relations and the gig economy literature by demonstrating that overt and covert resistance strategies can complement each other if workers are unable to legally form their unions to bargain with platforms and claim their rights for a fair wage and fair working conditions. By applying Scott’s (1985) everyday and overt resistance strategies, this article demonstrates that gig workers are able to exercise both covert and overt resistance strategies in Dhaka. However, it should be noted that although Scott’s concept of everyday resistance is useful to analyse Uber drivers’ covert resistance strategies in Dhaka, it does not explain how these resistance strategies are developed through an unnoticed challenge of platform power by gig workers, especially in a context where unionisation is weak and the state does not recognise Uber drivers’ demand to be considered as workers. This article also contributes to the discussion of algorithmic governance of labour, labour movement and platform studies by demonstrating that everyday resistance strategies such as the khep practice and the use of multiple apps are more useful to subvert platforms’ control than resisting collectively against ridesharing platforms like Uber. Many drivers have also left Uber and started driving for Pathao as a protest mechanism. Although existing literature suggests gig workers mostly utilise a range of covert resistance strategies (Riordan et al., 2023; Rosenblat, 2018; Wells et al., 2021), none of these studies has documented the khep strategy used by Uber drivers in Dhaka. This article also shows how Uber drivers use their informal networks to deal with police harassment.

This article further contributes to the labour movement, industrial relations and gig economy literature demonstrating that despite police harassment and the state’s supporting role for Uber, Uber drivers are able to organise collectively and use their informal associations to resist Uber’s oppression. This is an important contribution to the existing platform studies and labour movement literature as the existing Global North literature mostly shows that workers are able to use the longstanding transport workers’ organisations to bargain with platforms. In contrast, in most South and Southeast Asian countries including Bangladesh, informality is a norm and informal workers are often not well organised and largely rely on informal networks rather than forming formal unions to demand their rights (Ford & Honan, 2019; Lata et al., 2019; Ray, 2024). Following this historical tradition, gig workers heavily rely on their informal networks to navigate the gig economy (Ford & Honan, 2019; Mustika & Savirani, 2021; Ray, 2024). Like other South and Southeast Asian countries, the transport workers’ unions are very weak in Bangladesh and they often support the ruling political party. Hence, they hardly organise to demand transport workers’ rights. Additionally, ridesharing drivers are not welcomed by these unions. Consequently, Uber drivers have formed their own associations and used their informal associations to protest Uber’s exploitative policies. However, due to the informal nature of their associations, Uber’s management staff mostly do not negotiate with them and they hardly attend to any workers’ protests. Nonetheless, it is important to recognise that Uber drivers have been successful in making some changes, such as the victory they achieved to receive financial support from Uber for Arman’s family. Also, Uber has to boost its drivers’ payments when DRDU organises protests. However, the agency of Uber drivers in general and their associations DRDU and App-based Workers’ Federation of Bangladesh have experienced legal challenges to pursue their claims to ensure gig workers’ rights and formally register DRDU as a trade union. Although DRDU members have been trying to register DRDU as a formal organisation since 2021, they have not been able to register as a trade union yet. This is because the Industrial Relations Ordinance regulates trade union activities and prohibits certain types of formal workers, such as teachers, nurses, supervisory staff and workers in export processing zones, civil service and security force employees, from forming trade unions (Fairwork, 2022). As gig workers are not recognised as employees, legally they are not allowed to form a union.

Finally, this article emphasises the importance of exploring the local political and organisational culture as this often enables or constrains gig workers’ agency to fight against the platforms and claim their rights to fair wages and job security.

Footnotes

Acknowledgements

I am grateful to the editor Professor Silke Roth, and the three anonymous reviewers for their valuable comments and feedback, which helped strengthen this manuscript. I also sincerely thank my participants for their time and contributions to this project.

Funding

This project is funded by the University of Melbourne.

References

Adams-Prassl

Coyle

(2021). Uber and beyond: Policy implications for the UK. https://www.productivity.ac.uk/wp-content/uploads/2021/04/Uber-and-Beyond-with-cover-sheet-PDF.pdf

Akhter

(2022). Taxicab service all but disappears. New Age. https://www.newagebd.net/article/182503/taxicab-service-all-but-disappears (accessed on 15 November 2024).

Aloisi

De Stefano

(2022). Your boss is an algorithm: Artificial intelligence, platform work and labour. Bloomsbury Publishing.

Anwar

M. A.

Graham

(2020). Hidden transcripts of the gig economy: Labour agency and the new art of resistance among African gig workers. Environment and Planning A: Economy and Space, 52(7), 1269–1291.

Anwar

M. A.

Graham

(2021). Between a rock and a hard place: Freedom, flexibility, precarity and vulnerability in the gig economy in Africa. Competition & Change, 25(2), 237–258.

App-Based Drivers Union of Bangladesh. (2022). Thousands of drivers unionizing Uber in Bangladesh. https://news.techworkerscoalition.org/2022/04/12/issue-6/#:~:text=12%20Apr%202022&text=This%20nascent%20union%20demands%20ride,to%20improve%20socio%2Deconomic%20conditions

Arcidiacono

Gandini

Pais

(2018). Sharing what? The ‘sharing economy’ in the sociological debate. The Sociological Review, 66(2), 275–288.

Arubayi

(2021). Documenting the everyday hidden resistance of ride-hailing platform drivers to algorithmic management in Lagos, Nigeria. South Atlantic Quarterly, 120(4), 823–838.

Bangladesh Bureau of Statistics. (2022). Bangladesh Labour Force Survey 2022.

10.

Bangladesh Labour Act 2006. (2024). Bangladesh Labour Act. https://mccibd.org/wp-content/uploads/2021/09/Bangladesh-Labour-Act-2006_English-Upto-2018.pdf

11.

Braun

Clarke

(2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77–101.

12.

Cameron

L. D.

(2022). ‘Making out’ while driving: Relational and efficiency games in the gig economy. Organization Science, 33(1), 231–252.

13.

Chen

J. Y.

(2018). Thrown under the bus and outrunning it! The logic of Didi and taxi drivers’ labour and activism in the on-demand economy. New Media & Society, 20(8), 2691–2711.

14.

Cunliffe

A. L.

(2011). Crafting qualitative research: Morgan and Smircich 30 years on. Organizational Research Methods, 14(4), 647–673.

15.

Dubal

(2023, January 19). On algorithmic wage discrimination. UC San Francisco Research Paper No. Forthcoming. SSRN. https://doi.org/10.2139/ssrn.4331080

16.

Fairwork. (2022). Fairwork Bangladesh ratings 2022: State of work in the Bangladesh gig economy.

17.

FBCCI. (2024). The Federation of Bangladesh Chambers of Commerce & Industry. https://fbcci.org/web/profile/mst-sharmin-akter

18.

Ford

Honan

(2019). The limits of mutual aid: Emerging forms of collectivity among app-based transport workers in Indonesia. Journal of Industrial Relations, 61(4), 528–548.

19.

Heeks

Gomez-Morantes

J. E.

Graham

Howson

Mungai

Nicholson

Van Belle

J. P.

(2021). Digital platforms and institutional voids in developing countries: The case of ride-hailing markets. World Development, 145, Article 105528. https://doi.org/10.1016/j.worlddev.2021.105528

20.

Holloway

S. L.

Holt

Mills

(2019). Questions of agency: Capacity, subjectivity, spatiality and temporality. Progress in Human Geography, 43(3), 458–477.

21.

Holtum

P. J.

Irannezhad

Marston

Mahadevan

(2022). Business or pleasure? A comparison of migrant and non-migrant Uber drivers in Australia. Work, Employment and Society, 36(2), 290–309.

22.

Hossain

(2024). Extortion in transport sector changes hands from AL to BNP. The Daily Prothom Alo. https://en.prothomalo.com/bangladesh/v52r0zwkki

23.

Kässi

Lehdonvirta

Stephany

(2021). How many online workers are there in the world? A data-driven assessment. Open Research Europe, 1, 53, https://doi.org/10.12688/openreseurope.13639.4

24.

Khan

(2020). Pension: A dream for informal sector workers. Bangladesh Institute of Labour Studies.

25.

Lata

L. N.

(2023). Spatial justice, contested governance and livelihood challenges in Bangladesh: The production of counterspace. Routledge.

26.

Lata

L. N.

(2024). Digital labour platforms, resistance, and labour protection in Bangladesh. In Rahman

Masud-All-Kamal

(Eds.), Social transformation in Bangladesh (pp. 92–104). Routledge.

27.

Lata

L. N.

Burdon

Reddel

(2023). New tech, old exploitation: Gig economy, algorithmic control and migrant labour. Sociology Compass, 17(1), Article e13028.

28.

Lata

L. N.

Copolov

(2025). Parasitic platform urbanism in Dhaka and Melbourne. Human Geography. Advance online publication. https://doi.org/10.1177/19427786251321611

29.

Lata

Walters

Roitman

(2019). A marriage of convenience: Street vendors’ everyday accommodation of power in Dhaka, Bangladesh. Cities, 84, 143–150.

30.

Mustika

Savirani

(2021). ‘Ghost accounts’, ‘Joki accounts’ and ‘account therapy’: Everyday resistance among ride-hailing motorcycle drivers in Yogyakarta, Indonesia. The Copenhagen Journal of Asian Studies, 39(1).

31.

OECD. (2023). Informality and globalisation: In search of a new social contract. OECD Publishing.

32.

Peticca-Harris

DeGama

Ravishankar

M. N.

(2020). Postcapitalist precarious work and those in the ‘drivers’ seat: Exploring the motivations and lived experiences of Uber drivers in Canada. Organization, 27(1), 36–59.

33.

Popan

Perez

Woodcock

(2023). Cards against gamification: Using a role-playing game to tell alternative futures in the gig economy. The Sociological Review, 71(5), 1058–1074.

34.

Rani

Castel-Branco

Satija

Nayar

(2022). Women, work, and the digital economy. Gender & Development, 30(3), 421–435.

35.

Ray

(2024). Coping with crisis and precarity in the gig economy: ‘Digitally organised informality’, migration and socio-spatial networks among platform drivers in India. Environment and Planning A: Economy and Space, 56(4), 1227–1244. https://doi.org/10.1177/0308518X231220296

36.

Riordan

Robinson

R. N.

Hoffstaedter

(2023). Seeking justice beyond the platform economy: Migrant workers navigating precarious lives. Journal of Sustainable Tourism, 31(12), 2734–2751.

37.

Rosenblat

(2018). Uberland: How algorithms are rewriting the rules of work. University of California Press.

38.

Schor

J. B.

Tirrell

Vallas

S. P.

(2023). Consent and contestation: How platform workers reckon with the risks of gig labor. Work, Employment and Society, 38(5), 1423–1444. https://doi.org/10.1177/09500170231199404

39.

Schor

J. B.

Vallas

S. P.

(2021). The sharing economy: Rhetoric and reality. Annual Review of Sociology, 47, 369–389.

40.

Scott

J. C.

(1985). Weapons of the weak: Everyday forms of peasant resistance. Yale University Press.

41.

Scott

J. C.

(1989). Everyday forms of resistance. Copenhagen Papers, 4, 33–62.

42.

Scott

J. C.

(1990). Domination and the arts of resistance: Hidden transcripts. Yale University Press.

43.

Tassinari

Maccarrone

(2020). Riders on the storm: Workplace solidarity among gig economy couriers in Italy and the UK. Work, Employment and Society, 34(1), 35–54.

44.

The Daily Star . (2022). Uber drivers in Bangladesh can now see destinations and payment modes. https://www.thedailystar.net/tech-startup/news/uber-drivers-bangladesh-can-now-see-destinations-and-payment-modes-3073616

45.

Van Doorn

Ferrari

Graham

. (2023). Migration and migrant labour in the gig economy: An intervention. Work, Employment and Society, 37(4), 1099–1111.

46.

Veen

Barratt

Goods

(2019). Platform-capital’s ‘app-etite’ for control: A labour process analysis of food-delivery work in Australia. Work, Employment and Society, 34(3), 388–406.

47.

Walker

Fleming

Berti

(2021). ‘You can’t pick up a phone and talk to someone’: How algorithms function as biopower in the gig economy. Organization, 28(1), 26–43.

48.

Wells

K. J.

Attoh

Cullen

(2021). ‘Just-in-place’ labor: Driver organizing in the Uber workplace. Environment and Planning A: Economy and Space, 53(2), 315–331.

49.

Wiener

Cram

W. A.

Benlian

(2023). Algorithmic control and gig workers: A legitimacy perspective of Uber drivers. European Journal of Information Systems, 32(3), 485–507.

50.

Wood

A. J.

Lehdonvirta

(2021). Antagonism beyond employment: How the ‘subordinated agency’ of labour platforms generates conflict in the remote gig economy. Socio-Economic Review, 19(4), 1369–1396.

Algorithmic control and resistance in the gig economy: A case of Uber drivers in Dhaka

Abstract

Keywords

Introduction

Gig economy in Bangladesh

Gig economy, algorithmic control and resistance in the gig economy

Methodology

Algorithmic control, precarious work and everyday resistance

Overt resistance in the gig economy

Discussion and conclusion

Footnotes

Acknowledgements

Funding

References