TRADE UNIONS IN TRANSFORMATION 4.0 LABOUR AND SOCIAL JUSTICE RESISTING ­EXPLOITATION BY ALGORITHMS Drivers’ Contestation of App-based Transport in Indonesia Fahmi Panimbang, Syarif Arifin, Sugeng Riyadi, Dina Septi Utami September 2020 Drivers in the disruptive ­transportation sector have organised in innovative ways to fight for their rights amidst worsening working conditions. Through solidarity-building and providing mutual aid among drivers, the driver organisations have built significant associational power. Combined with their strategic place in the economy, they can utilize their bargaining power to confront the platforms and government ­regulations.

FRIEDRICH-EBERT-STIFTUNG – Resisting Exploitation by Algorithms LABOUR AND SOCIAL JUSTICE RESISTING EXPLOITATION BY ALGORITHMS Drivers’ Contestation of App-based Transport in Indonesia »Trade Unions in Transformation 4.0« examines unions’ strategic actions to mobilize power resources in a»new world of work« in which capital uses digital technology to re-organize the labour process. The Global Trade Union Programme of the FES aims to understand how the power balance between capital and labour is impacted and how workers are responding to the threats of the digital rollback towards greater exploitation and precariousness of workers. Pursuing a dialogue and action-oriented approach, the project ultimately intends to contribute to trade unions’ strategic reflections, experimentation and purposeful transformation.

 CONTENT Content 1 INTRODUCTION 3 2 RESEARCH METHODOLOGY 5 3 CONTEXT AND TECHNOLOGICAL CHANGE 6 4 THE EMERGENCE OF APP-BASED ­ TRANSPORTATION AND DISCONTENT 4.1 Go-Jek, Grab, and the expansion of app-based transportation........... 7 4.2 Horizontal conflict between indigenous transport and app-based transport drivers............................................. 8 4.3 Legal controversies and power relations.................................... 8 5 IMPACT OF THE RISE OF APP-BASED TRANSPORTATION 10 5.1 Domination by app-based transport companies............................ 10 5.2 Imbalance in bargaining power.............................................. 10 5.3 Algorithmic labour control................................................... 11 5.4 Occupational safety and health problems................................... 12 6 ORGANISING STRATEGIES AND WAYS OF BUILDING ASSOCIATIONAL POWER 13 6.1 Community................................................................... 13 6.2 Association.................................................................... 14 6.3 Unions......................................................................... 15 6.4 Female drivers in app-based transport...................................... 15 7 RESISTANCE 17 7.1 Repertoire of collective action and driver mobilisation: Activating structural power.................................................. 17 7.2 Algorithmic resistance....................................................... 18 7.3 Network- and alliance-building.............................................. 18 1

FRIEDRICH-EBERT-STIFTUNG – Resisting Exploitation by Algorithms 8 ORGANISING APP-BASED DRIVERS: THE SPDT-FSPMI INITIATIVE 20 8.1 Organising initiative.......................................................... 20 8.2 Internal changes.............................................................. 20 8.3 Main objective................................................................ 20 8.4 Strategies to attain the objective............................................ 20 8.5 Specific result................................................................. 21 8.6 Attempts to build power resources.......................................... 21 8.7 Challenges ahead............................................................. 22 8.8 Success factors and lesson learnt............................................ 22 9 IMPACT OF THE COVID-19 PANDEMIC 24 9.1 Impact on drivers............................................................. 24 9.2 Insufficient response by the government and the companies............. 25 9.3 Drivers’ initiatives............................................................. 25 10 CONCLUSION 26 References.................................................................... 27 Glossary, abbreviations, and acronyms..................................... 28 2

Introduction 1 INTRODUCTION This paper discusses the different strategies of worker-organising in the app-based transport sector in Indonesia, specifically in Greater Jakarta(the cities of Jakarta, Bogor, Depok, Tangerang, and Bekasi, also referred to collectively as Jabodetabek), Banten province(Serang city) and West Java province(Bandung city), and offers an analysis of how these workers and their organisations develop and activate power resources to gain real power. The analysis throughout the paper uses the power resources approach(PRA), which focuses on innovative organising strategies and resistance, as well as efforts to develop new forms of participation among drivers in the emerging app-based transport sector in Indonesia. Originating in the late 1960s and early 1970s with the re-discovery of class as an analytical category and as a mobilising and organising principle among leftist scholars and social activists, the PRA has been established based on the principle that organised labour can successfully defend its interests by collective mobilisation of power resources in the structurally asymmetric and antagonistic relationship between capital and labour(Schmalz et.al. 2018). Today, the PRA is an established tool in trade union research and used to analyse the spaces of action of labour unions and workers under given circumstances. The power resources of organised labour include associational power, structural power, institutional power, and societal power, which we define as follows(Schmalz et.al. 2018; Schmalz and Dörre, 2017): 1. Associational power refers to the power workers can potentially wield when they come together to form a trade union and collectively commit to the goals and aspirations of their organisations. This power relies on workers’ collective actions, including strikes, protest actions, campaigns, and willingness to pay dues, among other things. Workers’ associational power emanates from union membership, union density, and efforts to organise different groups of workers in different sectors. Examples of associational power include:(a) worker committees at the workplace level;(b) trade unions at the sectoral level; and(c) workers’ political party/organisations in the political system(connected to societal power). This requires infrastructural resources, organisational efficiency, member participation, and internal cohesion. 2. Structural power denotes the position of workers in the economic system. When workers possess scarce skills or competencies that are valuable to the employer, they have high bargaining power. Structural power rests on power to cause disruption of the production, economic or political system(also known as disruptive power). Structural power may be gained if workers occupy a strategic position within production processes or the supply chain. For example, dockworkers who collectively refuse to unload cargo at a large container terminal in one place can disrupt and halt the entire cargo and logistics chain. Thus, structural power is the bargaining power of workers, both at the workplace as well as in the marketplace. A skill is therefore needed to optimally combine structural power with organisational capacities in the existing institutional setting, and to develop an effective strike strategy. 3. Institutional power is a result of struggles and negotiation processes based on structural power and associational power. Laws and regulations supporting labour are generally the result of union struggles and negotiations which are made possible by the activation of structural and associational power. A union’s representation in national or sectoral negotiation structures may also enhance its institutional power. However, this institutional power is of a dual nature, as a representation gap may emerge and it is always subject to compromise. Thus, there is a»logic of membership« as well as a»logic of influence« which may impact the negotiation result. 4. Societal power derives from a union’s collaborative and coalitional relationship with other groups, such as workers’ organisations, NGOs, faith-based organisations, academia, social movements, and other organisations that share many of their goals and interests. This societal power is effective if labour organisations are able to persuasively articulate an agenda for social change and convince others that such an agenda is possible and necessary. There are two sources of societal power: coalitional power and discursive power, which are mutually reinforcing. Coalitional power means being networked with other social actors, having common goals and mutual commitments for mobilisation and joint campaigns. At the same time, discursive power is the ability to successfully intervene in public debates, while relying on opinion 3

FRIEDRICH-EBERT-STIFTUNG – Resisting Exploitation by Algorithms leaders with regard to worker- and trade union-related issues. The paper is structured as follows. Section two describes research methodology. Section three identifies factors driving the rise of the app-based transportation sector in Indonesia. Section four discusses the emergence of transport platforms and the discontent associated with it. Section five analyses the impact of transport platforms, highlighting key issues relating to platform domination, imbalanced bargaining power, algorithmic labour control, and occupational safety and health problems. Section six highlights three different forms of worker-organising strategies: community, association, and union. Section seven analyses driver resistance, including their mobilisation, algorithmic resistance, networking and alliance-building. Section eight discusses the case study of the SPDT-FSPMI initiative in organising appbased drivers by looking at the union’s attempt to mobilise resources. Section nine discusses the impact of the Covid-19 pandemic on drivers. Finally, section ten concludes with a few reflections. 4

Research methodology 2 RESEARCH METHODOLOGY This research focuses on strategies of worker/driver organising in the app-based transportation sector(both motorbike taxies and car services) in Indonesia, especially in two major companies: Go-Jek and Grab. We use»driver« and»worker« interchangeably throughout the paper to refer to those persons who work as drivers in the transportation sector, including app-based transport. Our analysis draws on data collected through structured and semi-structured interviews and focus group discussions(FGD) with(1) indigenous transport drivers;(2) app-based transport drivers;(3) appbased driver communities, associations, and unions;(4) representatives from trade unions; and(5) researchers as well as activists at non-governmental organisations(NGOs). All the interviews were conducted in two rounds of fieldwork carried out in October–December 2019, and January-May 2020, primarily in Jakarta, Bekasi, Cikarang, Depok, and Bogor, where the greatest numbers of app-based transport workers are concentrated, but also in the provincial capitals of Banten(Serang city), and West Java(Bandung city). Those cities represent over 50 percent 1 of total app-based transport workers in the country. The research specifically analyses a drivers’ union(the SPDT-FSPMI’s) organising strategies as a case study, employing the power resources approach(PRA). The selection of drivers and driver organisations for interview and focus group discussion took into account drivers’ active organising work up until when this research carried out, and was based on a significant number of interviews of factory workers in industrial areas who worked a part-time job as app-based drivers. Very few women drivers and organisers were able to be involved in the research discussion, reflecting the gender distribution of the sector. In total, we interviewed and had discussions with 44 drivers and driver communities(40 men and 4 women), 12 driver community organisers(10 men and 2 women), 6 indigenous transport drivers(all male); and 8 labour union activists(7 men and 1 woman). In addition, we had discussions with 5 NGO activists(4 men and 1 woman), and 4 researchers(3 men and 1 woman). 1 An estimation made by researchers at Institut Studi Transportasi(Institute of Transportation Study) or Instran in an interview conducted on 31 January 2020. Also, a drivers’ association, Jabodetabek, which according to application officials covers accounts for 50 percent of the total number of drivers for each company filing an application. (Interview with driver organisers at an association, 29 January 2020). 5

FRIEDRICH-EBERT-STIFTUNG – Resisting Exploitation by Algorithms 3 CONTEXT AND TECHNOLOGICAL CHANGE The swift adoption of app-based transport services by urban populations in Indonesia is, in part, related to three key characteristics of the country’s urban society: first its widespread fondness for smartphones; second its high demand for mobility; and third, the availability of large number of indigenous means of transport that serves people’s need for mobility and commuting. Both smartphone and indigenous transport, especially motorbikes, and the nexus where the two meet, have become symbols of urbanite Indonesian lifestyle, and are increasingly driving the commercial success of app-based transport companies. Technological development and competition between mobile network providers in Indonesia have brought about a drastic fall in the price of mobile phones and connection costs over the past decade. This mobile phone market has exploded, especially in the past half-decade(Baulch, 2017); affordable and cheap handsets have started to appear in the market, and the informal repair-and-recycle economy has begun to mushroom. This has involved purchasing, maintaining, and upgrading mobile phones, also with regard to the recent smartphone boom, with these devices becoming increasingly affordable for many ordinary people across the country. Secondly, the sale of motor vehicles in Indonesia has increased. As in general the road network has barely expanded to keep pace, this has led to ever-worsening traffic congestion. Traffic jams have become the rule in many cities on Java, especially Greater Jakarta. It is important to note that over half of the entire population of Indonesia live on the island of Java, with the island having one of the world’s greatest population densities. The most popular form of indigenous transport throughout Indonesia is the motorbike taxi or ojek. Ojek is a non-licensed motorbike taxi operating randomly in most areas of Indonesia, from big cities where clogged traffic commonly inhibits other forms of transport, to rural areas that cannot be reached with four-wheeled vehicles. As an unofficial means of transport, there is no need for the freelance ojek drivers to obtain permits or licenses. Anyone, therefore, can become an ojek driver even without a driver’s license.(Kusno, 2016). As Kusno(2016) notes, ojek has been the prime symbol of»private« lower-class grassroots transport that has filled the gap in the country’s poor public transportation system. The rapid growth of ojek in the past several decades has been driven mostly by the widespread availability of cheap domestic motorbikes made by Japanese companies, and even cheaper ones recently starting to be imported from China. Easy credit schemes with which to purchase motorbikes and the ease with which driver’s licence can be obtained have also been contributing factors. This particular context and development of smartphones and motorbikes and increasing use of the two have formed the basis for the emergence of the platform economy in Indonesia. With limited jobs available in the market, a large proportion of low-income and/or low-skilled people work as drivers in the informal transport sector, mainly offering paratransit services such as motorbike taxi( ojek), minivan and microbus ( angkot), and three-wheeler motorised taxi( bajaj). The existence of such an informal sector, especially in urban areas, is in part because the capacity of formal public transport services in Greater Jakarta is still inadequate to meet the ever-increasing travel demand of the population. Paratransit is one of the unique features of the Indonesian transport system, where such types of informal transportation are referred to in the scholarly literature as»indigenous transport«(Cervero, 2000). 6

The emergence of app-based ­transportation and discontent 4 THE EMERGENCE OF APP-BASED ­TRANSPORTATION AND DISCONTENT 4.1 GO-JEK, GRAB, AND THE EXPANSION OF APP-BASED TRANSPORTATION Indonesia has been the target of many new platform companies that adopt a strategic approach to the country’s huge market, including the growing number of millennials who represent a large proportion of the overall Indonesia’s consumer market. With a population of 264 million in 2017 and a workforce of 131.01 million in 2018, the Indonesian population is largely dominated by millennials (aged between 0–34), who account for 33.75 per cent of the total population. There are more than seven million unemployed persons in the population and over ten million with semi-employment status. 2 Moreover, the National Development Planning Agency(Bappenas) has projected Indonesia’s population to increase to up to 297 million in 2040, of which 64 per cent will be of a productive age(Adiningsih, 2019: 62–64). Needless to say, this segment of people has been the main target players seeking to gain major market shares in the growing platform economy, where mobility and transportation play a significant role. The two leading app-based transport companies in Indonesia are Go-Jek and Grab. Go-Jek is an Indonesian company founded in 2010. It began by offering an online motorbike taxi booking service, 3 now called Go-Ride, and later expanded its scope to offer a range of different services. After it launched its app in January 2015, it was downloaded by 10 million users, with this figure subsequently increasing to more than 35 million in early 2017(Ford and Honan, 2017). In Indonesia alone, there were over 2.5 million drivers actively working for Go-Jek in 2019, providing services in more than 167 cities and districts(Azka, 2019; Adiningsih, 2019). Go-Jek also operates in Viet Nam, Thailand, Singapore and the Philippines. Today, Go-Jek is valued at more than US$ 1 billion, making it one of two transport unicorns in Indonesia(Google-Cloud, 2 Semi-employment may also be referred to as disguised unemployment or hidden unemployment. It is a kind of unemployment where people appear to be employed but are actually not, or engage occasionally in casual, informal work that is not stable enough to provide them a livelihood. 3 This online motorbike taxi booking service is popularly known as »ojol«, a short form of»ojek online«, which means online motorbike taxi. 2018). Even the Indonesian government claims that Go-Jek is valued at IDR 69.4 trillion(US$ 4.97 billion). Recent prominent investors in Go-Jek include Sequoia Capital, Temasek Holdings, Google Inc, Tencent, JD.com, Meituan-Dianping, KKR, Warburg Pincus, Farallon Capital, Capital Group Markets, GDP Venture, Rakuten Ventures, Allianz, and Astra International(Badan Ekonomi Kreatif, 2018: 15, 26). The second major player, Grab, is a Singapore-based company founded in Malaysia in 2012, which entered the Indonesian market when it launched the GrabTaxi service in Jakarta in June 2014, followed by GrabBike and GrabCar in 2015(Ford and Honan, 2017). Grab operates in 100 cities in Indonesia, and many cities in other Southeast Asian countries. The number of Grab-bike and Grab-car drivers surged past 2 million when it acquired Uber in early 2018. In 2019 Grab employed over 9 million drivers across Southeast Asia, providing 14 different kinds of on-demand services(Jayani, 16 November 2019). Recent research suggests that the market share of Grab in the on-demand transport service sector has increased to become even bigger than Go-Jek, both in the Indonesian market(64 per cent) and in Vietnam (74 per cent)(Jayani, 18 September 2019). In the following years, both Go-Jek and Grab sought to expand their business to other on-demand services, including delivery services for food, convenience goods, and urban logistics. Go-Jek even integrates a wide range of lifestyle services through a secondary app called Go-Life, which offers services such as home cleaning, in-home massage, home auto service, laundry pick-up and delivery, all delivered on-demand to customers’ homes. Mobility and transportation play a vital role in this on-demand economy. Moreover, the modes of payment featured by both Go-Jek and Grab are increasingly cashless, as financial technologies become more integrated into the sector, enabling users to pay fares, remit money amounts, and even purchase products or services among their growing merchant base. Since its launch in 2016 until February 2019, Go-Jek’s digital payment method, Gopay, has registered a total transaction sum of almost US$ 6.3 billion. Meanwhile, Grab’s digital payment method (OVO) has become equally popular, being valued at US$ 2.9 billion in March 2019(Yuliawati, 2019). This»digital push« has transformed not only the way people commute – above all it has changed the way they live. 7

FRIEDRICH-EBERT-STIFTUNG – Resisting Exploitation by Algorithms 4.2 HORIZONTAL CONFLICT BETWEEN INDIGENOUS TRANSPORT AND APPBASED TRANSPORT DRIVERS Behind the popularity of Go-Jek and Grab lies a history of clashes between indigenous transport and app-based drivers. 4 These clashes were the result of fierce resistance of indigenous transport companies to app-based transport services, giving rise to protests and violence. In fact, appbased transport services have actively worked to displace the indigenous sector by actively recruiting indigenous transport drivers. Clashes between drivers in various cities have made news headlines, stirring controversies in the public debate. 5 The government was indecisive and only responded sluggishly to drivers’ protests and conflicts. Between late 2015 and early 2016, the government issued a ban against app-based transport, which was later retracted following massive protests by app-based drivers(Makki, 2015). Among others demonstrations, app-based drivers drove en masse around the Hotel Indonesia traffic circle in Central Jakarta in response to the ban 6 (see section 4.3 on legal controversies and power relations). When these horizontal conflicts between the drivers receded, companies started to recruit thousands of new drivers again, deepening the many unresolved horizontal conflicts. Go-Jek, for example, moved to recruit indigenous transport drivers in August 2015, enlisting tens of thousands of indigenous transport drivers, converting them into Go-Jek workers. At the time, Go-Jek was recruiting thousands of drivers per day(Aulia, 2015). The companies hired drivers up to 55 years of age. Drivers had to have own their own motorbikes, but initially they were provided basic training, skills, guidance, as well as two helmets, a jacket and a smartphone which had to be paid off in instalments— a daily amount was debited to drivers’ digital wallet in the apps on a daily basis(FGD with driver organisations, October 2019; see also Ford and Honan, 2017). App-based transport companies compete with each other to recruit drivers, providing incentives to migrate from a competing company, and even offering much more incen4 See Ford and Honan(2019) for more detailed account on these clashes. 5 Drivers’ clashed in Kalibata City, South Jakarta, on 6 July 2015. On 14 March 2016, thousands of indigenous transport drivers went on strikes in front of the Presidential Palace, demanding closure of appbased transport operations(Michico, 2016). In Surabaya, East Java, the city mayor stated on 27 March 2016 that the city would not provide any permit to app-based transport companies, arguing that competition between the indigenous and platform companies is not fair. In Medan, North Sumatra, the protest by indigenous transport drivers turned violent on 22 February 2017. There were massive protests by indigenous transport drivers once again in Jakarta on 3 October 2017, but this triggered acts of vengeance by hundreds of appbased drivers on 6 October 2017, who condemned the attempts at intimidation(Adiningsih, 2019: 134–135; Ford and Honan, 2019). 6 See Ford and Honan(2017) for a more detailed discussion of government responses to clashes between indigenous and app-based transport workers. tives to those brokers who could convince as many drivers as possible to migrate to their platform. This competition in recruiting drivers is especially evident in relation to motorbike taxi services(interview with app-based drivers, November 2019). Recently, more indigenous transport drivers have been joining the app-based transport sector, especially since companies continue to target them specifically for recruitment. Unlike individuals who have to deal with restrictions when attempting to join these platforms, companies considerably ease these requirements in the case of indigenous transport drivers from any opang. In areas that have experienced high levels of horizontal conflict, popularly known as»red zones«, these companies reimburse drivers for their» opang card« or» opang permit« and assist them to obtain a driving license, vehicle documents, etc. In short,»red-zone« opang drivers are warmly welcomed by platform companies to their business(Interview with former opang driver, October 2019). While conflicts have decreased significantly, resistance remains in certain areas where there are large numbers of powerful opang. However, challenges remain from the paratransit sector, such as from minibuses( metromini) and minivans( angkot). Unlike opang, paratransit services are regulated despite their informal status. As of 2015, there were 24,000 minivans operating in Greater Jakarta(Katadata, 2017). The impact on minivan drivers has been obvious, with their passengers decreased significantly, consequently threatening drivers’ livelihoods. 4.3 LEGAL CONTROVERSIES AND POWER RELATIONS Legal aspects in the operation of app-based transport have been a source of public debate since 2015, and several specific new government regulations have been issued over time. Under Indonesia’s constitutional article no. 22/2009 on Road Traffic and Transportation, as well as the Presidential Regulation no. 74/2014 on Road Transport, app-based transport companies are illegal. Both national and provincial governments have declared app-based transport to be illegal, as it does not meet several basic requirements, including requirements for the establishment of a legal entity within Indonesia and to use yellow license plates on vehicles, as is stipulated for public transport. With a view to these regulations, the Ministry of Transport announced on 17 December 2015 a ban on all app-based transport services on the grounds that they do not meet basic requirements (see also Ford and Honan, 2017). Again, in early 2016, following continued massive protests by indigenous transport drivers, the Ministry of Transport issued a decree(AJ 206/1/1 PHB 2016, dated 14 March 8

The emergence of app-based ­transportation and discontent 2016), 7 which was addressed to the Ministry of Communication and Information asking the latter authority to consider blocking the operational function of app-based transport companies, especially app-based taxis operated by Uber 8 and Grab(Ford and Honan, 2017). However, these disputes were swiftly settled. As the government was preparing to issue various regulations, the app-based transport(car service) was officially recognised. In response to continued massive protests from drivers in several cities, the Ministry of Transport issued Regulation no. 108/2017 on 1 November 2017, regulating the operation of app-based transport companies, including tariffs, the labels/stickers to be used, territories, and regular inspection of vehicles(Ford and Honan, 2017). Later in 2019, the government also issued other regulation, including no. 12/2019 on protection and safety for motorbike ride services(Adiningsih, 2019: 135–136). These regulations restrict operation of app-based transport drivers in two categories: red and green zones. While app-based drivers are allowed to pick up and drop off passengers in green areas, they are not allowed to do so in red zones. in October 2019. In addition, the President has attended several Go-Jek public events, an overt gesture endorsing the company. For instance, at one Go-Jek event involving drivers held in Jakarta on 11 April 2019, the President asked whether anyone in the audience had ever delivered his order of chicken satay to Bogor presidential palace via Gofood, Go-Jek’s food delivery service. When one driver raised his hand, he was invited to come up to the stage for a brief chat, after which the President gave him bicycle as a present(Hardiyan, 2019). The previous year, the President had also travelled all the way to Hanoi, Vietnam, when Go-Jek launched Go-Viet as its operation in Vietnam. Grab has also benefited from government patronage, albeit less prominently than Go-Jek. Luhut Panjaitan, one of the President’s most trusted assistants and the Coordinating Minister of Maritime Affairs and Investment, and the former Indonesian ambassador to Singapore(1999–2000). Grab has been officially collaborating with the government since 2019 through with app-based transport operations at seven airports in Sumatra, among other agreements(Reily 2019). These legal controversies clearly show that the government is ill-prepared to respond to the rise of app-based transportation. Moreover, despite massive protest from workers, officials have been more responsive and open to measures that favour business rather than workers. Notably, when the Ministry of Transport issued a decree in mid-December 2015 prohibiting app-based transport operations, President Joko Widodo slammed the Ministry’s move, saying that these operations’ presence filled the gap of increasing demand for reliable public transport( The Jakarta Post, 18 December 2015; Ford and Honan, 2017). Since that time, the President has publicly defended app-based transport companies on several occasions. His support for the digital transport platforms is perhaps not only based on the popular argument that the digital economy in the country needs to be fostered, but is also possibly linked to the fact that Go-Jek CEO and founder Nadiem Makarim is from a family that is part of the country’s power elite, a theory supported by the fact that he appointed Makarim Minister of Education and Culture Underlying the often-unseen policy-making processes and the power relations in which the rise of app-based transport companies is embedded is a real concern about the threat to workers’ livelihoods, not only of the indigenous transport drivers who have gradually been squeezed out of the sector, but also of many app-based drivers, who are now increasingly seeing decreasing incomes. Nonetheless, several massive strikes by drivers shows that they have recognised their own structural(disruptive) power. 7 The decree highlighted eight reasons for the ban:(1) Uber and Grabcar are not public/commercial transport entities;(2) the app-based transport companies had not established any legal bodies;(3) they had no permits to operate as transport providers;(4) they were foreign investors, but were not operating as limited corporations/companies;(5) they violated Presidential Decree no. 90/2000 regarding foreign companies mandating the presence of country office representative in Indonesia, as well as the regulation introduced by the Investment Coordinating Body(no.22/2011) pursuant to foreign investment;(6) the app-based transport venture cooperated with non-legal firms and individuals in running commercial transport;(7) the app-based transport providers have caused horizontal conflicts with official transport providers and taxi companies; and(8) the appbased transport venture promote the use of unofficial transport providers, which have brought about weaker demand for public transport(Hukumonline.com, 14 March 2016). 8 Uber was the third largest player that operated in Jakarta and several other cities between 2014–2018. In April 2016, it introduced an ojek service UberMotor, but this ended in early 2018, when its operations in Southeast Asia were acquired by Grab. One researcher has revealed that Uber threw in the towel due to costly settlement of legal disputes as well as managerial considerations. 9

FRIEDRICH-EBERT-STIFTUNG – Resisting Exploitation by Algorithms 5 IMPACT OF THE RISE OF APP-BASED TRANSPORTATION 5.1 DOMINATION BY APP-BASED TRANSPORT COMPANIES has even increased vehicle miles travelled(VMT) due to the availability of this inexpensive, accessible service. The present players dominating the app-based transport sector in Indonesia, Go-Jek and Grab, have eliminated many competitors, and are now battling it out with each other to see who is able to monopolise the market. The two companies display an ingrained desire to monopolise that market, and also exhibit anticompetitive structures. One of the reasons for this is because massive acquisition of proprietary data can give a very significant competitive advantage to a single operator, as the greater the number of interactions occurring via the app platform, the better the algorithm governing transactions and the underlying service (Smorto, 2018: 21). This tendency towards monopoly, like the practice of capital and investment in general, has had a number of consequences and posed several problems. Venture capital investments are the main drivers in the rise of digital platform companies such as Go-Jek and Grab, which are able to tap into vast resources and market aggressively to consumers. One tactic employed by these companies is price wars with their competitors. App-based transport companies, all of them well-endowed through financial investment, are able to subsidise costs and offer a much lower price to consumers than indigenous transport drivers. When an app-based means of transportation has secured a dominant position in the market, it becomes the only point of access for drivers and passengers. This clearly increases the danger of an imbalance in bargaining power in favour of app-based transport companies. In the long run, consumers are at risk of facing much higher fares and workers of experiencing exploitation due to the complete dependency of society on app-based transport. It is important to note that app-based transport services in Jakarta are now superior to, and cheaper than, other modes of transport. Moreover, research indicates that commuter rail users in Jakarta are becoming dependent on app-based transport services, with usage of app-based transport surpassing shares accounted for by other means of public transport(Bus Rapid Transit and microbus) at most commuter rail stations(Saffan and Rizki, 2018). Several studies have found that the business practices of app-based transport have not reduced the number of motorised vehicles ownership and number of trips, and In short, platform companies such as Go-Jek and Grab are attaining predominance and commercial success at the expense of labour and the environment(Retamal and Dominish 2017:6; Davidson and Infranca, 2018: 4; Nastiti, 2017; Saffan and Rizki, 2018), and potentially even consumers. Unfair competition over costs and fares has been criticised by many indigenous transport drivers, who have gained the support of a handful of government officials. These include the Minister of Transportation, who initially issued the ban on app-based transport in December 2015, and the mayor of Surabaya, who in March 2016 refused to issue any permits for app-based transport to operate in the city(Adiningsih, 2019: 134-135). Local governments have significantly more to lose from the dominant app-based transport companies. If platform companies such as Go-Jek and Grab gain complete control over transportation in a locality, they will not only squeeze out alternatives like paratransit and other indigenous forms of transport, but also reduce the number of public transit users. This, in turn, has the potential to negatively impact cities’ long-term capacity to provide essential public transport services. 5.2 IMBALANCE IN BARGAINING POWER App-based companies insist that drivers are partners or independent contractors. They are free to take orders if they so desire. They are able to work with a lot flexibility. But this claim is misleading, as in reality workers are forced to take orders or to work longer in order to earn sufficient income (Ford and Honan, 2017; Nastiti, 2017). Moreover, the‘contract’ between drivers and app-based transport companies is asymmetric and one-sidedly in favour of the company. In fact, the imbalance in bargaining position and unequal relations between the company and workers has existed since its registration. The claim of partnership is hollow and empty, as the rules of any such partnership are never discussed, let alone have they been negotiated bilaterally with workers(FGD with union organisers, October 2019; FGD with driver association leaders, February 2020). Notably, work arrangements allow app-based companies to make unilateral changes in tariffs, set up point schemes 10

Impact of the rise of app-based transportation which workers are forced to follow, deactivate or suspend accounts, and terminate work arrangements. These are usually carried out in the absence of transparent mechanism, and highlight only those cases where drivers are mostly to blame rather than customers. For instance, if customers unaware of riding rules, for example the requirement that there can be only for four passengers in a car, comment negatively on driver performance when the driver informs them of this regulation or charges them extra. Moreover, lack of contractual obligations to provide an explanation makes it very difficult to substantiate a claim against the company. In the earlier period of operation between 2015–2016, companies required workers to submit either their original education certificates, vehicle documents, or other important records as a guarantee for the work arrangement. There were protests against such requirements and they were subsequently withdrawn. Such imbalances in bargaining power are based on the enormous capacity of app-based companies to generate big data from all users using a set of algorithms. 5.3 ALGORITHMIC LABOUR CONTROL Between 2015 and 2016, when the number of app-based drivers was still low and the demand for app-based transport services was high, drivers were able to earn a healthy monthly income(Ford and Honan 2017). At that time, drivers occupied a strategic position within the production processes, which means they have increased structural power. However, over the ensuing years, the ability of drivers to accrue a large number of trips and reasonable income declined drastically. There are at least three factors that have caused a decrease in drivers’ income. First is the rapid increase in the number of drivers coming onto the market. Second is the decision by the company to drop the tariff rate by 25 per cent. Third, the company has made changes in its points-based bonus system, increasing the target number of trips to be completed before a bonus can be paid out(Ford and Honan, 2017). Amidst rapidly increasing competition from other app-based transport companies and drivers, drivers are also confronted with cuts in pay. Most of these factors are related to the company’s algorithmic management. This allows the company to gradually control and structure an unequal relationship between the company and its drivers, while maintaining the power asymmetry by exploiting technology and algorithms. Not many workers comprehend what algorithms are and how they work, but for the most part they understand that their performance is being controlled and monitored by the apps. Interestingly, algorithms are also made to sometimes favour indebted drivers, who still have to pay for their jacket and helmet in instalments, which are deducted every day from the drivers‹ accounts. These drivers are sent a notification about locations with passengers requesting a ride. However, such notifications are no longer available once payment has been settled. Labour control is based on algorithmic analysis. Once drivers activate their apps, they are constantly monitored and analysed. Nastiti(2017) has meticulously identified the methods of labour control by algorithmic management in the case of Go-Jek. Methods include mechanisms to earn points, bonuses, ratings, and suspension or deactivation. These methods constitute a carrot and stick: drivers get a bonus if they can collect many points, but they are punished if they are caught not complying. Nonetheless, disincentives are disproportionately higher than incentives. The rating system serves as another strategy to evaluate drivers. If their average rating falls below four stars, drivers are automatically suspended. They are also suspended if they refuse orders. Drivers know that they are in a vulnerable position and always feel compelled to attain a»five-star« rating and positive comments from customers. One single bad comment from a customer could ruin a good performance on the whole over the past one or two months. They are compelled to keep moving and searching for passengers proactively. When an order comes in, they cannot easily refuse it. The app system managed by algorithms compels drivers to stay active and search for orders. Otherwise, they risk deactivation from the app, ranging from a short-term suspension of their account lasting several hours to deactivation(FGD with drivers, October-December 2019). App-based companies leverage this significant control over how workers behave on the job. Rosenblat(2018) analyses how Uber uses its algorithms: rather than supervising its hundreds of thousands of workers with human supervisors, the company has built an app-based transport system using a number of algorithms that serve as a virtual»automated manager«. This algorithmic labour control was pioneered by Uber in Silicon Valley, but was very quickly copied by other companies such as Go-Jek and Grab. In the beginning, many drivers resisted complying with the exploitative control and rule by the app by using a ›fake GPS‹ bug(known as tuyul) 9 to work around the workflow system. This method allows drivers to be seen on the application map at a desirable location closer to potential passengers, so they can receive orders from customers without having to make the effort of getting to the customer, even though in reality they might be taking a rest at a different location(such as at home if it is not that far away, or at a community secretariat to charge their phone). Drivers claim this is not wrong because the bookings and trips are real, and trip orders are completed as usual. However, the company at a later point persuaded the Jakarta metropolitan police to adopt policies that deem such fake GPS to be a fraudulent act. As a result, the number of drivers who are joining in the battle against the company’s algorithm has decreased(see section 7.2 on algorithmic resistance. Companies’ drivers’ data generated through algorithms are also used to control tariffs and rates as well as monitor 9 This is a term used by drivers to denote a fake global positioning system(fake GPS). Literally, tuyul means a spirit that obtains wealth for its human master. It is a mythical spirit in Malay mythology in Southeast Asia, especially in Indonesia, Malaysia, Brunei, and Singapore. 11

FRIEDRICH-EBERT-STIFTUNG – Resisting Exploitation by Algorithms drivers’ behaviour in the workplace context. As companies keep expanding their services, promotional prices and tariff cuts for passengers are advertised on a massive scale. Lowpriced ›sales campaigns‹ are at the expense of drivers whose rates are cut from time to time(Nastiti, 2017). There is no opportunity for drivers to negotiate fares and rates with the companies, which constantly announce different rates and change policies via short messages or through apps sent unilaterally to drivers. In order to have full control over drivers, app-based transport companies have hired many more drivers, gradually increasing competition among them and weakened their bargaining power. On the other hand, since December 2019, the new apps algorithms assign only one driver to carry out two food delivery orders in one go, with drivers being paid much lower rates for the second order, rather than receiving two fares(FGD with community members, January 2020). The digital-money platforms upon which companies operate are also used as additional means of controlling and monitoring drivers. While drivers can save more money by using digital money accounts, they often face difficulty in withdrawing their balances, as they have to wait from a few days to one week for withdrawals to be approved by the companies. If drivers’ phones are determined to have loaded and used additional fake apps to increase orders, drivers cannot withdraw their cash. If for any reason drivers are deactivated, the company can confiscate the remaining balance in their digital money accounts. In one case a driver lost his remaining balance of over US$ 1000 when he was laid off(interview with union leaders, October 2019). These algorithmic practices reflect how technology is permanently altering not only how we define work but also how it is organised and used to control and monitor workers(Gandini, 2018). For the moment, app-based transport companies such as Go-Jek and Grab have succeeded in bringing the world of algorithms into the context of employment, which has a host of implications for how workers are treated and protected. While regulators and legislators are still sluggishly moving to catch up, the app-based transport companies are rapidly using data-driven algorithms to reshape the norms of employment and rewrite the rules of work(Rosenblat, 2018; Nastiti, 2017). 5.4 OCCUPATIONAL SAFETY AND HEALTH PROBLEMS Indonesia is among nine Asian countries accounting for roughly half a million road fatalities annually, making a major contribution to the growth of global traffic accidents (Jiang and Zhang, 2018: 63). 10 In the category of road accidents and fatalities, Indonesia ranks among the highest among the ASEAN countries, with motorbikes being major contributors(Dananjaya, 2019; Antara, 2017). In Cikarang, one of the busiest industrial areas of Bekasi regency in West 10 Other Asian countries are Cambodia, China, India, Japan, Malaysia, Singapore, Thailand, and Vietnam(Jiang and Zhang, 2018). Java province, at least five traffic accidents and fatalities involving app-based drivers occur each month(FGD with driver association, January 2020). The app-based transport companies do not provide any assistance with regard to such occupational accidents. 11 Rather, it relies on a voluntary team of drivers to handle them; this has triggered the formation of rapid response teams or Unit Reaksi Cepat/URC by drivers on a voluntary basis(Ford and Honan, 2019)(see section 6.2 on the drivers’ association for a more detailed description of the URC team). Many app-based drivers suffer from a number of occupational illnesses, mostly lung and respiratory-related diseases, haemorrhoids, back pain, and exhaustion-related illnesses that lead to other health issues. Air pollution in certain areas has clearly contributed to severe lung and respiratory diseases that some drivers suffer from, which should be taken into account when considering limitation of drivers’ working hours. However, drivers are compelled to take orders in order to maintain their performance levels; many drivers are pushed to work long hours, making them vulnerable to such occupational illnesses. Some drivers who used to work in factories have been laid off and now have to work in appbased transport, which suggests that working in factories is relatively healthier, and that app-based transport work has a high risk of fatal accident. No doubt there are factories with unsafe working conditions as well, equal to the risks faced by drivers(FGD with driver community leaders, October 2019). Activists at an industrial union in Bekasi are concerned about the fact that at least 30 per cent of their union members are taking side-jobs as app-based drivers. One union member died in early 2019, one year after he enlisted in app-based transport. It was observed by many of his friends that he was pushing himself to drive for long hours and was consuming food supplements to stay awake and strong so as to earn points, bonuses, and eventually additional income(interview with union leaders, October 2019). He seemingly had the freedom and flexibility needed to work, but the algorithmic management described in previous section that compelled him to work longer hours and keep his performance level high may possibly have led to his death. Interviewees said they believed that there could be a lot of more similar fatalities, on top of the obvious risks of related-road accidents. App-based drivers like him are adapting to working for a faceless boss in the gig economy. Algorithmic managers direct how app-based drivers behave, when and where they work, and apply responsive incentives and penalties that affect drivers’ wages. 11 No health and fatality insurance is available for drivers. Recently ­Go-Jek and Grab have provided traffic insurance, with the former ­using a government insurance scheme(BPJS) and the latter providing private insurance, with coverage only amounting to a maximum of IDR 25 million(US$ 1800). 12

Organising strategies and ways of building associational power 6 ORGANISING STRATEGIES AND WAYS OF BUILDING ASSOCIATIONAL POWER Several labour unions in industrial areas have admitted that many of their union members have become app-based drivers, making it more difficult for them to get involved in union activities. One obvious factor, they claim, is the meagre wages workers earn from factory work(interview with union leaders, November 2019). In general, not many trade unions have adequately responded to labour conflicts in the app-based transport sector. There are at least three major reasons for this. The first is that app-based transport companies keep claiming that their drivers are not considered workers, but rather partners. For example, the Go-Jek CEO and founder has asserted that»drivers are micro-entrepreneurs; they work for themselves … they’re free to take orders whenever they want, or not«(Tech in Asia, 2015; Ford and Honan, 2017). Entrepreneurship, freedom, and flexibility have become the ingrained public image. Many people believe that drivers accept trip orders without pressure, which is totally erroneous. Unfortunately, this image is convincingly cultivated by many drivers, including former workers, creating barriers for labour unions to organise app-based drivers. This thus makes the societal(discursive) power of driver unions particularly important in countering such deceptive claims. Secondly, formidable challenges that have emerged, including the open horizontal conflicts between incumbent drivers and the new app-based drivers that have developed in many cities. Even a union federation in the transport sector stated that organising app-based transport drivers requires different tools, tactics, and strategies. Thirdly, labour unions are pre-occupied with their existing challenges, especially unfavourable new policies, including a draft omnibus law promoting investment and business by encouraging companies to pursue profits at the expense of labour and the environment. Organising this new terrain of the gig economy has yet to become a priority for most labour unions(FGD with driver organisers, December 2019). As Ford and Honan(2019) note, there are generally speaking three forms and strategies of organising in place: community, association, and union models. 6.1 COMMUNITY The driver community, popularly known as»komunitas ojol«, is the most popular form of organisation among app-based drivers. These communities are informal, and run flexibly. They operate at the grassroots level, with members mostly from the vicinity. The driver community often mentions the area, district, or city in its community name, highlighting the collectivity and solidarity on their logos and mottos, printed on banners, badges, and stickers. A community organisational structure usually consists of(1) an advisory board;(2) a field coordinator;(3) a vice field coordinator;(4) a secretary;(5) treasury; and(6) members. 12 Communication and coordination among members consistently take place through the WhatsApp messaging service as well as daily personal meetings at a driver’s rest area, called a»base camp«, roadside areas or at a landmark or eatery, where the drivers wait for orders or recharge their phones(Ford and Honan, 2019). Community organising strategies are mostly ad hoc, focusing on mutual support in dealing with workplace problems and issues such as suspension of accounts and sharing ideas for new tactics to earn income. Many communities provide social services, and more importantly they deal with member’s emergency and non-workplace issues such as providing support when drivers or family are sick. Some communities collect membership dues on a regular basis, usually every month, varying between IDR 10,000–30,000 (US$ 0.80–2.20). Other ones only collect funds from members occasionally, especially when they are needed for collective purpose such as having base-camp awnings to provide shade or to cover electricity costs. Community organisational structures and functions are flexible and more informal, making it easy to build connections and to network with different driver organisations across the country. Through the WhatsApp messaging service, driver communities share information with each other, 12 In some areas, driver community members are called taruna, or at least their community leaders call them so. Taruna means cadet or young novice military officer. An ethnographic study is still needed to understand this, but influences from the military organisational culture that developed during the Suharto dictatorship(1967–1998) still remain. 13

FRIEDRICH-EBERT-STIFTUNG – Resisting Exploitation by Algorithms and are in frequent contact, especially when drivers need to drop by to charge phones or to have a rest after taking a long trip order. This practice of mutual aid and support is helpful especially when a driver has an emergency(FGD with driver community leaders, October 2019). Such»solidarity« in practice, along with the collectivivity and bonds of friendship, is genuinely implemented in a community. This is not unique to the app-based transport phenomena, however, as it had been previously practised among other informal transport drivers, with truck drivers using stickers or labels on their vehicles, not only to promote mutual support between members – more importantly, such labels and stickers help prevent illegal»fee collections« by road transport authorities and police(interview with indigenous transport drivers, January 2020). There is no official data on the number of driver communities that exist, but it is estimated that over 5000 communities have been established in Greater Jakarta alone, with each community having between 10 and 100 drivers or even more(interview with researchers and driver association leaders, January 2020). A survey conducted in 2018 found that not more than 27 per cent of app-based drivers in several big cities have joined a community(Instran, 2018), which shows a lot of more drivers are still unorganised. Some driver communities are affiliated to an association of driver communities, a broader scale of organisation at district or city level. general secretary in charge of information and technology; (5) a vice general secretary in charge of public communication;(6) treasury;(7) organisational development and organising work;(8) economic resource development;(9) legal and advocacy; and(10) a rapid response team or URC (unit reaksi cepat). It is important to note that an association is a hotspot for driver mobilisation, where there is an intensive battle of ideas and influence from various actors. Platform companies rely on associations to detect any possible disruptions by drivers. Needless to say, many associations and the URC are tightly monitored by the companies, who assign a security task force called satgas 14 to co-opting drivers. These task forces regularly coordinate efforts with the police at different levels. For example, satgas collaborate with the Jakarta metropolitan police to raid fake GPS( tuyul) used by drivers, for which drivers can face criminal charges. More importantly, the key task of satgas is first of all to mediate and mitigate resistance from indigenous transport drivers against appbased transport; in some cases, satgas even mobilise drivers through associations to engage in mediating such conflicts by a show of force and an offer to support their fellow appbased-drivers. Given this, many drivers’ associations have played a significant role in mobilising protests against the government’s decision to ban the platform service that rose up between 2015–2018, in which the URC was among the leading organisers, acting in close contact with the satgas. 6.2 ASSOCIATION A drivers’ association, or wadah komunitas ojol 13 , is a broader form of drivers’ organisation, whose members come from various communities, but it may also include individuals as members. Many of these associations are informal in institutional terms and work flexibly, but some are registered with a social/mass organisational status, and have a more formal organisational structure. An association that organises work covers various locations, from sub-district, district/city, province, almost to the national level. A small association’s organisational structure usually consists of(1) an advisory board;(2) a chairperson;(3) a vice chairperson in charge of Go-Jek disputes;(4) a vice chairperson in charge of Grab disputes;(5) public relations;(6) a secretary; and(7) treasury, with women usually fill this latter post. The office of vice chairperson is assigned to individuals who have good contacts and connections and experience in settling disputes with the companies, but these disputes are settled through the»back door«, with drivers having to pay the costs in the event of suspension or deactivation of their accounts. Several larger associations have slightly more complex organisational structures, covering more locations, some being nearly nation-wide, although no association at present is genuinely national. Typically, these consist of(1) an advisory board;(2) a chairperson;(3) a general secretary;(4) a vice 13 Wadah means a meeting place, or a place to associate. A URC consists of several delegates(between 2–4 members) from each community, specifically responsible for assisting drivers if they run into conflict with indigenous transport drivers or if their vehicles break down on the road. This URC operates only at sub-district level, but easily connects on a broad scale with almost any other URC team in the country. A URC is a voluntary organisation that emanates from a need for mutual support among drivers at the grassroots level. However, although this role is vital to the business operation of platform transport, these are scarcely supported by the companies. Interestingly, one member of a URC team complained about the lack of such support from the company, stating that drivers should be creative rather than relying on the company’s benevolence; he suggested that drivers recruit food-stalls and other micro-businesses in neighbourhoods to register food delivery or other platform services, so as to increase orders for drivers(interview with URC team, December 2019-January 2020). This might be one effect of the»indoctrination« regularly attempted by platform companies through driver associations and URC. Moreover, drivers affiliated with the associations have better capacities to mobilise protest and gain access to policymakers and government. One driver association has been in contact with global app-based transport workers’ unions, 14 Satgas is a short form of Satuan Tugas, which literally means a taskforce. Here, satgas refers to a task-force to control and discipline the driver community. Satgas officers may go to communities to perform undercover surveillance. Officially, these satgas staff are called»partnership engagement staff«, a sort of public relations worker. 14

Organising strategies and ways of building associational power and attended the first-ever international convention of appbased transport workers held in January 2020. This will no doubt translate into an increased institutional capacity of this association to increase societal power in the future. 6.3 UNIONS App-based drivers’ unions are still very few. The initiative to establish unions mostly comes from the existing federation of factory unions, or from transport and dockworkers’ unions. Although there are only a handful of these, there has been a positive response from some traditional trade unions regarding infringement of existing labour rights in the app-based transport sector. Nevertheless, such organising initiatives face formidable challenges wherever the union is not popular among drivers. The app-based drivers’ union is relatively small and inactive, and not one single union has been involved in policy-making processes presided over by the Ministry of Transportation. One possible reason for the unpopularity of drivers’ unions is their organisational structure and function, which are considered inflexible and rigid, as these have been adopted from traditional trade unions, even though the new economy is fundamentally different in nature. As far as their legal position is concerned, drivers’ unions have greater institutional power, which includes a collective and organisational bargaining position for policy change. Ford and Honan(2019) argue that the organisational capacity and the institutional position of unions provide them with a greater capacity to engage with companies and government, such as by mobilising industrial action to pressure or negotiate or by using legal avenues to change policy and practice. Union organisers also contend that a union is the most reliable and effective vehicle to change policy. However, this development has been hindered by the fact that platform companies are reluctant to recognise formal unions to avoid the development of industrial relations. For instance, application for registration of a Go-Jek drivers’ union in Batam City in Riau Islands province, with a potential membership of about 1000 drivers, was repeatedly refused from mid-2019 to early 2020 by a civil servant at the local Labour Office, who claimed that Go-Jek’s head office in Jakarta had not approved the registration(interview with drivers’ union leaders, January 2020). Looking at Law no. 13/2003 pursuant to the Trade Union Act, such union registration actually does not require any company’s approval as long as basic requirements for registration are fulfilled. This case demonstrates the widespread practice of collusion between the state and capital to the detriment of labour. the Labour Act, although this is still subject to certain conditions and compromises proposed by the company. 6.4 FEMALE DRIVERS IN APP-BASED TRANSPORT Female participation in app-based transport employment is relatively low. Typically, persons who take this job are the bread-winner in the family and many are single-parents. According to a survey from 2018, female drivers involved in app-based transport can be estimated at less than three per cent of total app-based drivers(Instran, 2018). 15 However, organisers state that this figure could be higher at present, as recently women working in the sector have become increasingly visible. Some of them choose to stick to working in food delivery services, both with Go-Jek(Gofood) or Grab (Grabfood). One of the reasons for this is that many passengers, especially those who are in a hurry, prefer male drivers for a faster but safe ride, so many of these people cancel orders. The stereotype namely still prevails that women do not drive professionally. App-based women drivers are associated either in a local drivers’ community or female drivers’ community or both, actively participating in association events such as anniversaries or social services. One interesting activity is performed by a female drivers’ community in Jakarta with about 80 female members. They actively provide financial assistance to orphans of former drivers who have passed away because of occupational accidents or illness. All this financial support comes from voluntary donations by drivers, contributed every month, and rotating through all of Greater Jakarta. Female drivers’ communities have names like e.g. Srikandi, Lady Riders, Go-Jek Lady, Emas Srikandi and Mojang. Like many other communities in general, there are monthly membership dues of between IDR 10,000–30,000(US$ 0.80–2.20), collected by members each month on a rotating basis and followed by arisan 16 (interview with female drivers’ community leaders, October 2019 and February 2020). Along with app-based transport, other digital platforms in Indonesia offer on-demand cleaning and in-home massage services, which employ mostly women. These platforms include GoMaid, Mbakmu, OKHOME, TukangBersih, GoClean, and GoMassage. Some of Go-Jek’s female drivers also work for an on-demand massage service (Go-Massage) belonging to Go-Jek; these workers have to pay IDR 1,080,000(US$ 80) for training at a massage centre affiliated with Go-Jek. Go-Jek deducts 30 per cent from every worker’s earning: 20 per cent for Go-Jek, and Nevertheless, the case study on SPDT-FSPMI shows(see section 8) that the established labour union’s capacity to negotiate has helped the drivers’ union connected to it to gradually create space for negotiation with the app-based transport firm. As is illustrated in the case study, unions are able to formally register at a local Labour Office, and it is perhaps the only drivers’ union officially recognised under 15 The survey was carried out in late 2018 in four key cities of Greater Jakarta or Jabodetabek, Yogyakarta, Surabaya, and Bali. Greater Jakarta alone represents over 50 per cent of total drivers working on the platform. 16 Arisan is a regular social gathering to collect money for members, or sometimes to buy certain items that are desired and agreed by all the members. Each member gets her/his turn to receive the money or the item. 15

FRIEDRICH-EBERT-STIFTUNG – Resisting Exploitation by Algorithms 10 per cent for the massage centre. A study has shown that female workers in this gig economy in Indonesia are more vulnerable to exploitation and sexual harassment. This recently was shown by a case in Bandung, West Java, where a masseuse working for Go-Massage was raped by a costumer. This danger calls not only for safety measures from platform companies, but also government regulations to protect these vulnerable workers in this new branch of the economy(interview with researcher, October 2019 and February 2020). 16

Resistance 7 RESISTANCE A lot of drivers understand that power is a relational concept, which is to say that drivers’ ability to satisfy their own interests may depend, in part, on their ability to counter the power of their employers. As noted earlier, when the platform companies initially needed to recruit drivers, drivers’ structural power was relatively strong. Drivers as a collective are located at strategic points in the production or distribution process of the overall transportation business. Thus they have a certain structural power within the company. But this structural power needs to be exerted collectively (Luce, 2014). This collectivity, as it was perceived by many Indonesian workers in industrial areas during general strikes in 2012–2013, is powerful, especially when workers realise that together they are ›the agent of production’, and that, when they stop working, then production stops. Since early 2018, platform drivers have increasingly realised that the»honeymoon is over«: no more bonuses and easily earned money. Many have been suspended without reason. Drivers do not receive any orders for entire days or even weeks. They have to watch while orders are given to newly recruited drivers. Competition to get orders between app-based drivers has begun, and they fear this may lead to another horizontal conflict. As a collective, drivers are aware they could do a collective »off-bid«(turning off the apps) en masse to disrupt production or transactions(Ford and Honan, 2019). To some degree, there are collective repertoires that have stimulated drivers during recent mobilisation and protest. But that is exactly the intent underlying recruitment of more drivers by the company: a large concentration of drivers weakens their bargaining power, eventually preventing any possible structural disruption by workers. 7.1 REPERTOIRES OF COLLECTIVE ACTION AND DRIVER MOBILISATION: ACTIVATING STRUCTURAL POWER Riding vehicles in parade form during labour demonstrations on a road have become part of workers’ collective repertoire. This has become one of their significant forms of political articulation. As an element of»political trafficking« in worker protests, motorbikes are used for labour organising and mobilisation at every Labour Day rally. In particular, motorbikes were very essential vehicles for workers in the »factory raid«( grebek pabrik) that took place in Indonesia’s industrial heartland of Bekasi, West Java, over a period of more than six months from May to October 2012(Mufakhir, 2014). Workers’ motorbike parades staged as political protest in several industrial estates became everyday scenes between 2011 and 2013, and such mobilisation was considered to be an effective strategy. At that time, workers were demanding an end to the widespread use of subcontracting/outsourcing employment practices in factories, which are unlawful. Worker mobilisation during the factory raids, which included on-site strikes in factories, occupations of factories, ad hoc monitoring of compliance with regulations and on-site checks inside factories, while building workers’ solidarity across factories in industrial zones clearly required high mobility among organisers, who resorted to their motorbikes. Workers’ motorbike parades were usually coupled with one or two command cars( mobil komando) equipped with loudspeakers playing militant songs between instructions and speeches by protest leaders. 17 This repertoire of collective action provides an important basis for recent mobilisation of driver protest. Drivers have protested by reaffirming workers’ political demands: they drive motorbikes and cars in massive numbers, rallying at power centres near the presidential palace, house of representatives, government offices, app-based transport companies, and several landmarks, to voice their aspirations. Although drivers’ capacity of mobilisation and the way they organise protests still exhibit deficits and are patchy compared to industrial workers, they somehow mirror the usual»political trafficking« characterising workers’ protests, using vehicles, especially motorbikes. At the moment, political speeches given during mobilisation by protest leaders are ill-prepared and badly articulated. In these speeches, nationalistic jargon is sometimes recited out of necessity, reflecting the current Indonesian political climate. Interestingly, protests in front of platform company headquarters take place less frequently compared to those staged in front of government institutions because the former are seemingly uninvolved and hold greater power that is less 17 For a broader discussion of the importance of street-based labour protests in Indonesia, see Juliawan(2011), Caraway and Ford(2020), and Panimbang and Mufakhir(2018). 17

FRIEDRICH-EBERT-STIFTUNG – Resisting Exploitation by Algorithms transparent than that of the government institutions that are supposed to hold responsibility. In mid-January 2020, protesters called on the government to meet three key demands:(1) to protect drivers’ rights;(2) to stipulate reasonable tariffs to make it possible to secure a decent income for drivers; and(3) to regulate and limit new recruitment of drivers by companies in order to decrease competition between drivers. 7.2 ALGORITHMIC RESISTANCE Given the pervasive yet faceless control by algorithms, drivers have also employed a variety of resistance tactics to improve their subsistence incomes. Rather than more open forms of collective protest against app-based companies, drivers have taken advantage of some loopholes in the apps. These tactics have included negotiating with costumers over their orders. For example, a driver asks a customer to cancel an order from the app, when in reality, the trip is still performed at the same price as shown on the app. This means the driver saves 20 per cent – the amount of the deduction from his earnings. Another tactic involves the food delivery service, where drivers make use of promo and price discounts that are available. For example, if a customer orders food from Gofood (Go-Jek) and there are promos or discounts available for the same item at Grabfood(Grab), the driver uses his/her Grab account to purchase the item. This means the driver receives extra income through the discounts, sometimes up to 50 per cent of the price. These kinds of tactics are regularly shared and discussed by driver communities, as are other tactics and strategies(interview with a driver community, October 2019). These means of everyday resistance clearly show that drivers regularly attempt to circumvent their faceless bosses. One of the most interesting ongoing types of resistance against faceless bosses of algorithms is employed by several drivers’ communities who have capacities and skills in digital technologies. The group of drivers are popularly referred to as»IT jalanan«, 18 which means street programmer. The name»street programmer« is comparable to the»streetbooks-and-library« activism popularised by activists in several cities and towns, first initiated in Bandung, West Java. Later this spread to Greater Jakarta and elsewhere in the effort to promote a spirit of resistance among youth against social injustice. These»street programmers«, or IT jalanan, are self-taught persons, some of whom have been members of industrial labour unions. Because they operate clandestinely, the community of street programmers in several cities work and coordinate efforts mostly web-based or online. They share new information and tricks relating to the new release of apps from platform companies. Their main objective is to create algorithm bugs 18 IT stands for Information and Technology and»jalanan« literally means street, a popular space for marginalised people. Here, jalanan refers to a brand of activism advocating marginalised people. in the drivers’ application changing the algorithm. This is mostly intended to reduce the drivers’ workload set by the algorithm. Street programmers argue that the bonuses, ratings, and point systems push drivers to work extremely hard and for longer hours. They challenge and resist the algorithmic pressure by helping fellow drivers modify and tweak the apps’ algorithm. Interestingly, the group of street programmers perceive their resistance to be part of the class struggle against corporate greed(FGD with driver community leaders, November 2019). While algorithms seek to maximise the amount of labour extracted from drivers, an additional app is used to reduce their use of labour power. Drivers have consistently sought to discreetly circumvent these rules underlying faceless managerial assertions of control, asserting their own control over the use of their labour power. Drivers are supposedly free and independent, but companies’ rules, enforced by algorithmic managers, significantly limit opportunities for entrepreneurial decision-making available to them. Drivers have noticed the tension between the promise of freedom and the reality of invasive algorithmic management. In fact, this tension serves as the basis for legal claims that drivers should not be classified as independent contractors. 7.3 NETWORK- AND ALLIANCE-BUILDING One of the most important capacities of driver organisations is network- and alliance-building. In networking, they use the community’s sticker and badge as a symbol of network and outreach. That is one of reasons why they place and highlight collectivity and solidarity on their community’s logos and slogans, printed on banners, badges, and stickers. These stickers and badges are used in the exchange with each other’s communities when they visit each other. The number of stickers from various communities symbolises and demonstrates the networking capacity of a community; the more the stickers they have on their own banner or base-camp wall, the larger the network and the more the friendships they make. When a driver from a community in Bekasi has problems on the road in Jakarta, a community that is close to the place of accident in Jakarta is supposed to provide help, and vice versa(see also Ford and Honan, 2019). This is communicated through WhatsApp messaging by community leaders. It appears that all drivers uphold a tacitly agreed principle: »you should help other fellow drivers if you want to be helped by them«. Such»solidarity in practice«, along with collectivity and bonds of friendship, is genuinely practised by driver communities all over the place. This underscores the cooperation among drivers— as well as everyday resistance in practice— which has helped build the social cohesion needed for collective action among drivers. This social cohesion has, in turn, helped form the basis for driver solidarity when collective campaigns and actions are carried out. 18

It is important to note that it is thanks to such networking in practice that drivers’ communities play a significant role in constructing alternative social relations, in which mobility and movement of drivers are givens, while the globe is considered to be a common space. More importantly, they treat their fellow drivers everywhere as family. This kind of solidarity and camaraderie was even evident between several drivers from various countries during a recent international meeting of app-based transport workers staged in London in January 2020. Three Indonesian leaders from a drivers’ association attended the meeting, which was aimed at developing a global strategy to challenge platform companies, and establishing a convention for the International Alliance of App-Based Transport Workers. This international alliance has the vision of bringing together workers from international app-based transport services to discuss how they can collaborate to challenge the ever-increasing power of platform companies. Indonesian delegates are keenly aware that Uber might return to Indonesia or Grab and GoJek may expand to other parts of the world; such capital movements by app-based transport companies should be regulated to make sure they respect labour rights(Interview with association leaders, February 2020). Resistance 19

FRIEDRICH-EBERT-STIFTUNG – Resisting Exploitation by Algorithms 8 ORGANISING APP-BASED DRIVERS: THE SPDT-FSPMI INITIATIVE 19 8.1 ORGANISING INITIATIVE 8.3 MAIN OBJECTIVE  The Aerospace and Transportation Workers division of the Federation of Indonesian Metal Workers’ Union(Serikat Pekerja Digantara dan Transportasi Federasi Serikat Pekerja Metal Indonesia(SPDT-FSPMI) was established in 2016(Ford and Honan, 2019). Initially, the union merely organised airport workers beginning in 2000, subsequently expanding in the past few years to other transport-related sector, including app-based transport. FSPMI, the union federation, initiated driver organisation as a response to poor working conditions and the fact that several drivers have visited the FSPMI office to discuss possible driver-organising initiatives. A series of meetings then took place to devise a strategy for organising workers, and later, in May 2017, the federation began to become seriously involved in mobilising protests against the government and companies, demanding better working conditions for drivers. 19 8.2 INTERNAL CHANGES Organisational expansion has been one of FSPMI’s main priorities, as is reflected in the special department and division created within the federation’s organisational structure to focus on organisational development. The first formal meeting of the SPDT-FSPMI took place in conjunction with the fifth congress of the FSPMI in Surabaya, East Java province, from 8–10 February 2016. Formation of the SPDT-FSPMI easily obtained the full support of leaders of federations as well as that of the confederation, the Indonesian Trade Union Congress(KSPI). Internal changes have occurred as the union federation has expanded to other unions outside the metal sector. Debates have taken place, but these have related mostly to strategies and organisational consequences, such as the need to commit more human and financial resources in order to seriously organise the new sector. The SPDT-FSPMI has set out three main objectives:(1) to promote labour rights of drivers in major transport sectors, including the newly emerged app-based transport area;(2) to fight for the recognition of drivers as workers entitled to basic rights laid down in labour laws and regulations; and (3) to create a space for negotiations between driver representatives and app-based transport companies to achieve better working conditions for drivers. 8.4 STRATEGIES TO ATTAIN THE OBJECTIVE At least four strategies are employed by the SPDT-FSPMI in driver organising: 1. Awareness-raising through Jamkes Watch(Health Protection Watch) activities: Jamkes Watch(Health Protection Watch) sees itself as a watchdog over the government health care programme. Jamkes Watch organises volunteers to help citizens in need to obtain access to health care insurance. The strategy is to mobilise union volunteers to help people when they have to deal with health care services in hospitals. This kind of help is needed since many procedures for accessing the health-care service are challenging, posing even greater problems for the majority of low-educated citizen. As union volunteers have advocated and when they work to help people, they then follow up by explaining the importance of union and unionisation to the people. In particular, volunteers specifically target app-based drivers in an effort to organise workers, not only to foster mutual aid among drivers, but also to advocate their rights. 2. Union cooperative: 19 This section draws from interviews and FGD with SPDT-FSPMI leadership between October 2019 and April 2020. See also Ford and Honan(2019) for more detailed account on SPDT-FSPMI. 20 The main purpose of the union cooperative is to collect savings from members’ income to be used later whenever needed. Although it is still small in terms of size, the cooperative has attracted members because it is useful. At the moment, the union is not collecting membership dues since drivers are still facing difficult

Organising app-based drivers: The SPDT-FSPMI initiative19 times. This strategy has been implemented with the intention of promoting cooperative activity while also helping to shape union work. 3. Recruiting redundant members from the FSPMI’s other sectors: There have been massive redundancies with many workers being targeted for this or persuaded to terminate their employment agreements. Laid-off workers, especially former members of the FSPMI, are approached by the union about joining the SPDT-FSPMI when they choose to work as app-based drivers. 4. Deploying organisers in strategic locations: 8.6 ATTEMPTS TO BUILD POWER RESOURCES 1. Associational power: The obvious strength of SPDT-FSPMI is that the driver union is affiliated with a well-established metal workers’ union, with full support from the leadership at federation(sector) as well as at confederation(national) level. Resource support for the SPDT-FSPMI from the union federation can be obtained by submitting an activity plan to the federation. With hundreds of thousands of members, whose support can be mobilised, the SPDT-FSPMI can throw in its weight to enhance power in collective bargaining. The SPDT-FSPMI deploys organisers to different strategic areas to better understand the situation and provide the support needed by drivers on the ground. This direct organising strategy advocates staying in contact with drivers and driver communities, and promotes the union(SPDT-FSPMI) as well as the alliance(KATO). By leveraging the current structure of the union federation, the SPDT-FSPMI can expand its organising work to different provinces and cities where a union structure has been established. Reflecting on the recent recognition of the drivers’ union on Batam island, it has handpicked organisers in each province to reach out to app-based drivers. 8.5 SPECIFIC RESULTS Initially, following a series of efforts to register a driver union on Batam island, Riau Islands province in September 2019, attempts at registration were repeatedly rejected by the Labour Office and the company at the local level and headquarters level for obvious reasons: they did not want budge from their‘anti-union’ stance. When the union continued to apply pressure, the company finally accepted its registration in March 2020 subject to the condition that the name of the union be changed to reflect that it was still a partnership rather than a full-fledged union subject to industrial relations law with Go-Jek. The union agreed to this, as it only meant changing the local name of the union without sacrificing the goal of creating a space for collective bargaining. Since late March 2020 the union has been formally registered with 300 drivers currently signed on as members. In sum total, the SPDT-FSPMI has 1,700 members from the app-based transport sector ranging across Jabodetabek, Lampung, Batam, Goronto, Aceh, and North Sumatra. Second, as the coronavirus outbreak has impacted many drivers, several online taxi drivers whose cars have been taken back by the company for not being able to pay instalments or rent, have contacted the SPDT-FSPMI for advice. Many of these drivers have joined the union as members. Although workers usually only join a union only when they are in trouble, this also indicates that the efforts by the SPDT-FSMPI to organise are gradually paying off, at least in terms of recognition from drivers. The union is now preparing to provide advocacy in this completely new area for the union, negotiating with a leasing company. In view of the declining capacity to mobilise workers, on 3 April 2018 the SPDT-FSPMI established an alliance called KATO(Action Committee for Online Transport/ Komite Aksi untuk Transportasi Online) encompassing broad strata of groups and various driver organisations such as communities and associations. This alliance is being coordinated by the union’s national head office (Indonesian Trade Union Congress or KSPI), where the KSPI president, who is also the president of the FSPMI, has assumed a central role in the alliance(Koran Perdjoeangan, 21 September 2018). KATO provides space at the KSPI secretariat for any group of drivers to have meeting or even just to arrange an informal gathering. A series of meeting facilitated by KATO have taken place to develop action plans, organise strategies, and mobilise protest. It is important to note that the exercise of the union’s structural power to strike has not been as effective as it used to be. Generally, the mobilisation capacity on the part of the Indonesian labour movement for protests has been in decline since 2014, when capital and the state took countermeasures in the wake of successful general strikes in 2012–2013. 20 2. Structural power: When app-based companies launched their operations in 2015 and they were struggling to find drivers to join them, the drivers had high bargaining power since they occupied a strategic position within production processes. Today, drivers’ strategic position in the produc20 For a more detailed account of labour protest(1998–2013) and general strikes(2012–2013), see Panimbang F. and A. Mufakhir(2018) ‘Labour strikes in post-authoritarian Indonesia, 1998–2013’ in Jörg Nowak et.al(eds.) Strikes and workers movements in the 21st century. London: Rowman and Littlefield. 21

FRIEDRICH-EBERT-STIFTUNG – Resisting Exploitation by Algorithms tion process has increasingly eroded, as the number of drivers has risen steeply. However, a number of protests staged by the SPDT-FSPMI and other drivers’ groups over the years – both directed at the government, demanding recognition of motorbikes as a legal form of transport, and at app-based transport companies, demanding that they increase tariffs(income) – shows that drivers have realised that companies can be disrupted by massive protest and slowing down their work or work stoppage. These disruptive actions were carried out several times between 2016–2019, when workers demanded higher income on a massive scale. This increased structural power of drivers through strikes and mobilisation of protests soon prompted the companies to react by deploying satgas to reassert tight control over drivers. 3. Institutional power: Having had successful experiences in legal battles spearheaded by the union confederation(KSPI) in the past, KATO in its capacity as the alliance of app-based drivers took an action to the Constitutional Court in 2018, in which it sought judicial review of Law no. 22/2009 on Traffic and Land Transportation. The review was aimed at attaining recognition for the motorbikes that are used for app-based motorbike taxies as part of the public transport system. This was refused by the court. The alliance subsequently took another step in the form of a citizen’s lawsuit brought before Central Jakarta Court. The citizens’ lawsuit was seeking redress in a case involving six top state officials(including the president and vice president, chair of the national parliament, and several ministers) for failure to protect citizens’ rights to employment, with workers demanding recognition of their work in the app-based transport sector. Although this attempt also failed, the union was able to gain the support of many drivers, and other legal battles may be carried out in the future to enhance drivers’ bargaining power. 4. Societal power: When the SPDT-FSPMI took to the streets to protest against the government and companies, support came from other sectors within the federation, and especially from members of the alliance(KATO). This demonstrates the union’s coalitional power, although it has still remained limited to their peer groups and networks, and has yet to reach out to broader groups and social actors. 8.7 CHALLENGES AHEAD There are three major challenges facing the SPDT-FSPMI. The first one is that driver’s employment status has largely failed to receive recognition by both company and government, and it is seen as a partnership rather than employment that entails certain rights. This reflects a lack of institutional power. Second, a majority of drivers have failed to recognise the fact that they are workers who are entitled to labour rights as stipulated in the Labour Act, resulting in a lack of associational power. Third, the threat to and intimidation of union and human rights activists are being stepped up to cow them into silence. There have been recent cases in which human rights activists have received fake food orders from Grab and Go-Jek apps along with WhatsApp message from an unknown number:»be careful when you leave your home«(Florene, 2020). These three challenges reflect the current Indonesian political situation, where driver organisations such as the SPDT-FSPMI need to attempt to find innovative driver-organising strategies and develop resistance, while developing new forms of participation by drivers in the emerging app-based transport sector in the country. The SPDT-FSPMI needs to develop the capacity of its organisers, to equip them with specific organising skills in the app-based transport sector, and to provide them with a better understanding of the different contexts and circumstances of organising work. 8.8 SUCCESS FACTORS AND LESSON LEARNT The SPDT-FSPMI’s societal power is considerable, especially in terms of discursive power, where the app-based drivers’ alliance(KATO) and the union confederation (KSPI) have been able to play an active role in public debates over the rights of app-based drivers, which in part have contributed to their gaining public support. For instance, during the partial lockdown amid the Covid-19 pandemic, a drivers’ fund was established to help drivers weather the crisis, and the union was able to mobilise logistical support for drivers, and especially its members. The economic hardship being experienced by drivers generated considerable public sympathy. This also prompted app-based companies to raise funds and distribute amounts ranging between IDR 100,000– 150,000 each for some 200,000 drivers. Certainly, this distribution of support is still insufficient to reach out to a majority of impacted drivers. The SPDT-FSMI is one of the very few labour unions that interestingly enough has made inroads into the app-based transport sector, achieving considerable progress in organising app-based drivers(Ford and Honan, 2019). Although its membership is still small in comparison to the driver associations, the current power resources it has could help the SPDT-FSPMI develop into a more effective vehicle in achieving members’ goals. The union has achieved legitimacy and recognition vis-à-vis the state and capital in the representation of drivers, articulating and voicing their aspirations and demands. With the recent success of registering the union on Batam island, Riau islands province, it has established a base for future organising work in the area and elsewhere. What has made the SPDT-FSPMI successful is that the union knows how to utilise its power resources effectively in given circumstances. In fact, it is the only resourceful drivers’ un22

Organising app-based drivers: The SPDT-FSPMI initiative19 ion in Indonesia that is capable of advancing workers’ rights in the app-based transport sector. The SPDT-FSPMI has realised that there are lessons to be learnt from what has been practised by many other driver organisations on the ground: flexibility of the organisational structure, and their great capacity for solidarity-building. Drivers are for the most part very mobile and are capable of developing networking in practice speedily among drivers throughout a territory. The SPDT-FSPMI could be more successful if it were able to build alliances with other networks of drivers. As discussed in the previous section, many of these drivers’ organisations have failed to recognise the significance of the associational and institutional power that formal driver unions such as the SPDT-FSPMI have attained. Further communication and possible joint action with these driver groups in the longer run would help defend driver interests and achieve their goals. 23

FRIEDRICH-EBERT-STIFTUNG – Resisting Exploitation by Algorithms 9 IMPACT OF THE COVID-19 PANDEMIC The Indonesian central government has been criticised publicly for its poor response to the Covid-19 pandemic in the country. The government took two months to come up with a national policy strategy on how the government and citizenry should manage the epidemic. While many countries were swiftly responding to the Covid-19 outbreak in January and February 2020, the Indonesian government only established a Task Force for Rapid Response to Covid-19 on 13 March, and issued Government Regulation no. 21/2020 on 31 March to allow large-scale social restrictions aimed at limiting the spread of Covid-19. This partial lockdown since early April marked the beginning of a relatively stable approach to the national government’s management of the pandemic in the country after three months of deception. Social activists argue that this decision came so late because the government was trying to evade having to pay the greater costs of the lockdown, and was not willing to safeguard the majority of workers in the informal economy. 9.1 IMPACT ON DRIVERS Obviously, the implementation of large-scale social restrictions has impacted drivers in the transportation sector, including app-based transport, on an unprecedented scale. The majority of these drivers perform day and informal labour, making them among the most vulnerable citizens. Since early April 2020 to date, 21 people have been prohibited from commuting and public transportation has not been allowed to operate. To comply with government policy, the app-based transport companies de-activated the apps’ feature to take rides from passengers, only allowing them to take orders for deliveries of food, convenience goods, and urban logistics. All drivers hence solely have a few delivery service orders every day with which to earn money. Many drivers did not receive any orders for many days or even weeks in April-May 2020, although the app-based company sought to ensure a fairer distribution of orders for all drivers, while disregarding their past performance levels. 22 This caused a drastic decrease in drivers’ incomes, many of whom were not able to pay their vehicle’s monthly instalment to the leasing company, or their rent, and even to survive. The drivers’ association publicly protested against the policy, but to no avail(interview with union and association organisers, April-May 2020). In order to survive, many drivers’ communities have decided to sell bottled water at intersections or traffic lights while waiting for delivery orders, while still wearing a uniform and with their motorbikes parked at the roadside nearby. Some other communities have collected donations from passers-by at different locations. The situation has generated public sympathy for the drivers and led to growing food donations, mostly every afternoon at dinnertime during the fasting month of Ramadan between mid-April and mid-May. Drivers of app-based taxis are even facing more difficult times, as their nature of organising is not as flexible or mobile as their counterparts on motorbike taxis. These app-based taxi drivers, who also work for both Go-Jek(Gocar) and Grab(Grabcar), do not have any work whatsoever, as they do not have the privilege of taking delivery service orders. One leader stated that more than half of his members were jobless, unable to pay monthly instalments to the leasing company, with their cars even having been taken back by the leasing companies despite government orders to relax credit arrangements and monthly instalments(interview with union organisers, May 2020; CNN Indonesia, 9 April 2020). The situation worsened in May 2020, when some drivers were able to return home to their villages, despite restrictions on movement instituted by the government (interview with driver organisers, May 2020). 21 As of May 2020, when this paper was being written. 22 This policy has disadvantaged many drivers who have earned good performance levels. Previously, if they had a good performance level drivers were able to generate more income from a bonus system. Several communities were planning to protest against this policy with the company on 8–9 May, but the attempt never got off the ground. 24

Impact of the Covid-19 pandemic 9.2 INSUFFICIENT RESPONSE BY THE GOVERNMENT AND THE COMPANIES The Ministry of Transportation has slightly increased the tariff for app-based transport in the area of Jabodetabek since 16 March, by between IDR 2250–2650 per km, but this has barely helped drivers. In general, the government response has been patchy and food distribution assistance has often missed the target. The comprehensive and up-to-date data that is needed has been lacking and it has been reported that the government is relying on the data from application companies(interview with driver association leader, May 2020). firms for three months. In greater Jakarta, the SPDT-FSPMI mobilised support for members impacted by the pandemic, including the drivers. In Cikarang, West Java province, a drivers’ association collected donations from its members who had other jobs apart from the app-based transport, to be distributed to other members that relied only on appbased transport(interview with community and association leaders, May 2020). As the situation got worse, the application company Go-Jek launched a crowdfunding campaign which could also be accessed and paid using the Go-Jek application until the end of June 2020. In mid-April 2020, some of the fund was distributed to 200,000 Go-Jek drivers in the form of voucher worth IDR 100,000 for each driver to be exchanged for basic foodstuffs at convenience stores(Ramli, 15 April 2020). Likewise, Grab made a donation to its drivers from a crowdfunding platform in the form of a voucher worth IDR 150,000 per driver, and extra money for those who have been infected by Covid-19(Imandiar, 28 April 2020). Certainly, the majority of drivers have not received any such assistance. Becoming aware that many of his members were not receiving any voucher, one driver leader realised that the voucher was given only being given to selected drivers such as himself who were known personally by company staff (interview with association leader, May 2020). 9.3 DRIVERS’ INITIATIVES Most drivers’ communities, associations and unions have been paralysed, unable to cope with a situation in which their members are so desperate. Some drivers have given up and sold their driver’s account 23 in order to get cash for food just to survive, while others have struggled on to earn a living. Interestingly, some drivers in areas outside Jabodetabek, such as Serang in Banten province, have become conventional ojek drivers, as it is easier to get passengers without having to use an app(interview with community organisers, May 2020). Some other associations, such as Asosiasi Drivers Online or the ADO in Manado City, North Sulawesi province, have been able to act swiftly. The ADO was able to act as advocate for its 5000 members in Manado, obtaining relaxed conditions and a waiver of instalment payments to leasing 23 Driver accounts can be sold to others, especially with Go-Jek, where facial verification to open the account is not required(only passwords are used), so it can be used by others even though this is associated with several inconveniences. People would prefer to purchase an account from people they know, and especially if their own account were deactivated or blocked by the company, or they want to get better performance. The higher the performance level of an account, the greater its cost. 25

FRIEDRICH-EBERT-STIFTUNG – Resisting Exploitation by Algorithms 10 CONCLUSION Technological change and the rise of the app-based transport business in Indonesia and globally is not inevitable. Rather, it is a result of global economic and socio-political development, that is, the re-organising power of global financial capitalism. In particular, this new economy of digital platforms is profoundly driven by the increasing domination of venture capital, offering evidence that the‘push of digital’ emanates predominantly from business interests rather than societal necessity. The digital technology that is used by app-based transport companies is held to merely be a way to connect two groups of end users(drivers and passengers), but underlying it all is a more significant feature of this technology: big data, which could be exploited to achieve even greater domination of the market. The organising initiatives of driver unions and organisations such as the SPDT-FSPMI shows that drivers have been innovative in their attempts to organise to fight for their rights in the face of adverse circumstances. As we have shown, drivers’ unions and organisations are aware that they occupy a strategic point in the production of app-based transport, and they have the capacity to disrupt the business. This structural power will be an important resource in future organising work. Solidarity-building in actual practice as well as the common practice of providing mutual aid among drivers have become well-established in most driver organisations, clearly demonstrating drivers’ great potential through their significant associational power resources. This associational power could be of significance once drivers are successful in pressuring the government to protect them through regulations, including their rights to freedom of association and collective bargaining. Moreover, several drivers’ organisations have been involved in attempts to influence policy-making within the Transportation Ministry with regard to regulation of the sector. This institutional power is a significant lever for drivers whenever they want to escalate their campaigns to pressure the government, especially the Ministry of Labour, to achieve recognition of drivers as regular workers entitled to labour rights. 26

Conclusion REFERENCES Adiningsih, S. et.al(2019) Transformasi ekonomi berbasis digital di Indonesia: Lahirnya tren baru teknololgi, bisnis, ekonomi, dan kebijakan di Indonesia(Digital-based economic transformation in Indonesia: The birth of new technological trend, business, economy, and policy in Indonesia). Jakarta: Gramedia Pustaka Utama. Antara,(2017, November 15)»Angka kecelakaan lalulintas Indonesia termasuk tinggi di ASEAN«(Indonesia’s traffic accident is among the highest in ASEAN), in Tempo. https://nasional.tempo.co/read/1033993/angka-kecelakaan-lalu-lintas-indonesia-termasuk-tinggi-di-asean/ full&view=ok(accessed 12 November 2019). Aulia, M.(2015, August 14)»Go-Jek rekrut ribuan orang per hari«(GoJek recruits thousands drivers every day), in Medcom.Id. https://m.medcom.id/amp/aNro2gEb-go-jek-rekrut-ribuan-orang-per-hari(accessed 17 September 2019). Badan Ekonomi Kreatif(2018)‘Opus: Creative economy outlook 2019’. Jakarta: BEKRAF. Baulch, E.(2017)‘Mobile phones: advertising, consumerism and class’, in E. Jurriëns et al.(eds) Digital Indonesia: Connectivity and Divergence, Indonesia Update Series. Singapore: ISEAS-Yusof Ishak Institute. Caraway, T.,& Ford, M.(2020). Labor and Politics in Indonesia(Cambridge Studies in Contentious Politics). Cambridge: Cambridge University Press. doi:10.1017/9781108777858 Cervero, R.(2000) Informal transport in the developing world. Nairobi, Kenya: United Nations Commission on Human Settlements. Dananjaya, D(2019, September 9)»Angka kecelakaan tahun 2019, truk dan sepeda motor sering terlibat«(Figure on traffic fatalities in 2019: Truck and motorbikes are the most involved in road accidents) in Kompas. com. https://otomotif.kompas.com/read/2019/09/09/160200815/angka-kecelakaan-tahun-2019-truk-dan-sepeda-motor-sering-terlibat(accessed 12 November 2019). Florene, Ursula(2020) ›This is just a warning from me‹: How Grab and Gojek are utilized to intimidate activists in Indonesia’. KrAsia, 14 July 2020. Available online: https://kr-asia.com/this-is-just-a-warning-fromme-how-grab-and-gojek-are-utilized-to-intimidate-activists-in-indonesia (accessed 22 July 2020). Ford, M. and V. Honan(2019)‘The limits of mutual aid: emerging forms of collectivity among app-based transport workers in Indonesia’, Journal of Industrial Relations 61(4): 528–48. Ford, M. and V. Honan(2017)‘The Go-Jek effect’, in E. Jurriëns et al. (eds) Digital Indonesia: connectivity and divergence, Indonesia Update Series. Singapore: ISEAS-Yusof Ishak Institute. Gandini, A.(2019). Labour process theory and the gig economy. Human Relations, 72(6), 1039–1056. https://doi. org/10.1177/0018726718790002. Google-Cloud.(2019)‘GO-JEK: Using machine learning for forecasting and dynamic pricing’. https://cloud.google.com/customers/go-jek/(accessed 24 December 2019). Hardiyan, Y.(2019, April 11)»Jokowi pernah pesan sate ayam pakai GoFood«(»Jokowi ever ordered chicken satay via Go-Food), in Bisnis.Com. https://kabar24.bisnis.com/read/20190411/79/910950/jokowi-pernah-pesan-sate-ayam-pakai-go-food-(accessed 4 January 2020). Hukumonline.com.(2016, March 14)»8‘Dosa’ UberTaxi dan GrabCar versi Kementerian Perhubungan(8‘sins’ of UberTaxi and GrabCar according to Transportation Ministry). https://www.hukumonline.com/berita/ baca/lt56e65e90d3099/8-dosa-uber-taxi-dan-grab-car-versi-kementerian-perhubungan/(accessed 13 September 2019). Imandiar, Y,(2020, April 28), Grab Bagikan Voucher Belanja Rp 150 Ribu untuk Mitra Pengemudi(Grab distributed shopping vouchers valued IDR 150,000 to its drivers), DetikNews. https://news.detik.com/berita/ d-4994617/grab-bagikan-voucher-belanja-rp-150-ribu-untuk-mitrapengemudi(accessed 12 May 2020). Instran.(2018) Survei pengemudi angkutan online: Angkutan sewa khusus dan ojek online(Survey on app-based drivers in cars and motorbikes),(unpublished report). The Jakarta Post,(2015, December 18)»Jokowi defends ride-hailing apps as Transportation Ministry withdraws ban«. https://www.thejakartapost.com/news/2015/12/18/jokowi-defends-ride-hailing-apps-transportation-ministry-withdraws-ban.html(accessed 11 October 2019). The Jakarta Post(2015, December 18).»Transport Ministry bans ride-hailing apps«. https://www.thejakartapost.com/news/2015/12/18/ transport-ministry-bans-ride-hailing-apps.html(accessed 13 September 2019). Jayani, D.H.(2019, November 16)»Para penantang Gojek dan Grab di bisnis ojek online«(Competitors of Go-Jek and Grab in the app-based transport market), in Katadata. https://katadata.co.id/infografik/2019/11/16/para-penantang-gojek-dan-grab-di-bisnis-ojek-online (accessed 3 December 2019). Jayani, DH.(2019, September 18)»Pangsa pasar Grab di Indonesia dan Vietnam lebih besar daripada Gojek«(Grab’s market shares in both Indonesia and Vietnam are larger than Go-Jek), in Katadata. https://databoks. katadata.co.id/datapublish/2019/09/20/pangsa-pasar-grab-di-indonesia-dan-vietnam-lebih-besar-daripada-gojek(accessed 25 October 2019) Jiang, Y and Zhang, J.(2018)‘Road traffic safety in Asia: An analysis based on DPSIR+C framework’ in Zhang, J. and Z. Feng(eds) Routledge Handbook of Transport in Asia. London& New York: Routledge. Juliawan, B.H(2011)»Street-level Politics: Labour Protest in Post-authoritarian Indonesia.« Journal of Contemporary Asia, 41: 3, 349-370, DOI: 10.1080/00472336.2011.582706. Katadata.(2017)»Berapa jumlah angkot yang terdaftar di Jabodetabek?«(How many minivans registered in Jabodetabek?). https://databoks.katadata.co.id/datapublish/2017/03/23/berapa-jumlah-angkot-yang-terdaftar-di-jabodetabek(accessed 24 December 2019). Koran Perjdjoeangan,(2018, 21 September),‘Perbedaan KATO dengan Komunitas Ojek Online yang Lain’(the different between KATO and other driver communities), https://www.koranperdjoeangan.com/perbedaan-kato-dengan-komunitas-ojek-online-yang-lain/(accessed 20 May 2020). Kusno, A.(2016)»The order of messiness: Notes from an Indonesian city«, in M. Chalana and J. Hou(eds) Messy Urbanism: Understanding the »Other« Cities of Asia. Hong Kong University Press. Luce, S.(2014) Labor movements: Global perspectives. Cambridge: Polity Press. Makki, S.(2015, December 18)»Larangan dicabut, Gojek dan GrabBike melenggang«(Ban on app-based transport lifted, Go-Jek and GrabBike back to operate) in CNN Indonesia. https://www.cnnindonesia.com/tekn ologi/20151218154256-187-99172/larangan-dicabut-gojek-dan-grabbike-melenggang(accessed 13 September 2019). Michico, NR,(2016, March 14).»Ribuan sopir taksi dan bajaj demo di Monas: Yang legal kita, tapi jadi rugi«(Thousands of taxi and bajaj drivers protest at National monument: We hold the permit and legal, but we got bankrupt) in News.detik.com. https://news.detik.com/berita/3163851/ribuan-sopir-taksi-dan-bajaj-demo-di-monas-yang-legal-kita-tapi-jadi-rugi(accessed 13 September 2019). Mufakhir, A.(2014)»‘Grebek pabrik’ in Bekasi: Research note on unions’ mobilisation strategy’, in Suryomenggolo, J.(ed) Worker activism after Reformasi 1998: A new phase for Indonesian unions? Hong Kong: Asia Monitor Resource Centre. Nastiti, A.D.(2017)‘Worker unrest and contentious labor practice of ride-hailing services in Indonesia’. Evanston, IL: Northwestern University. Panimbang, F. and A. Mufakhir(2018)‘Labour strikes in post-authoritarian Indonesia, 1998–2013’ in Jörg Nowak et.al(eds.) Strikes and workers movements in the 21st century. London: Rowman and Littlefield. 27

FRIEDRICH-EBERT-STIFTUNG – Resisting Exploitation by Algorithms Ramli, RR(2020, April 15)‘Gojek Bagikan Voucher Belanja Alfamart ke 200.000 Mitra Driver’(Go-Jek distributed shopping vouchers at Alfamart convenience store for its 200,000 drivers). Kompas. https://money.kompas.com/read/2020/04/15/191516626/gojek-bagikan-voucher-belanjaalfamart-ke-200000-mitra-driver(accessed 20 April 2020). Reily, M.(2019, July 11).»Menko Luhut Dukung Grab Sediakan Layanan di Tujuh Bandara Sumatera«(Minister Luhut supports Grab to provide app-based transport services in seven airports in Sumatera), in Katadata. https://katadata.co.id/berita/2019/07/11/menko-luhut-dukung-grab-sediakan-layanan-di-tujuh-bandara-sumatera(accessed 28 December 2019). Rosenblat, A.(2018) Uberland: How algorithms are rewriting the rules of work. California: University of California Press. Saffan, A.F. and M. Rizki(2018)‘Exploring the role of online»ojek« in public transport trips: Case of Jakarta metropolitan area rail users’, IOP Conference Series: Earth and Environmental Science. Schmalz, S, Schiller, F and Webster, E.(2018)‘The Power Resources Approach: Developments and Challenges’. Global Labour Journal, 2018, 9(2). Schmalz, S and Dörre, K(2017). The power resources approach. Working Paper prepared for the FES Project»Trade Unions in Transformation«, Berlin: FES. Available online: https://www.fes.de/index.php?eID= dumpFile&t=f&f=32816&token=ce072af9d0b7670be9047b640338 abd717f87033 Smorto, G.(2018)»The protection of the weaker parties in the platform economy«, in N.M. Davidson et al.(eds) The cambridge handbook of the law of the sharing economy, Cambridge Law Handbooks, p. 16. Cambridge: Cambridge University Press. Tech in Asia(2015, November 15)»If you work at Go-Jek, Nadiem Makarim is your ›coach, sugar daddy, and friend‹ «, Tech in Asia. https://www.techinasia.com/tech-in-asia-jakarta-2015-indonesia-go-jek (accessed 22 August 2019). Yuliawati(2019)»Gelombang besar transaksi nontunai di Indonesia«(A big wave of non-cash trasnsaction in Indonesia), in Katadata. https://katadata.co.id/berita/2019/10/07/gelombang-besar-transaksinontunai-di-indonesia(accessed 28 December 2019). Zhang, J. and Z. Feng.(2018)»Introduction: Transport in Asia« in Zhang, J. and Z. Feng(eds) Routledge handbook of transport in Asia. London& New York: Routledge. GLOSSARY, ABBREVIATIONS, AND ACRONYMS Angkot Minivan and microbus used for public transport(par(angkutan kota): atransit) Arisan: A regular social gathering to collect money for members, or sometimes to buy certain items that are desired and agreed by all the members. Each member receives her/his turn to receive the money or the item. Bajaj: Three-wheeler motorised taxi Indigenous transport: Informal transport, mainly paratransit including motorbike taxi( ojek), minivan and microbus( angkot and metromini), and three-wheeler motorised taxi( bajaj). This paratransit is a unique feature of Indonesian transport, where such types of informal transportation are often referred to(in transportation studies) as»indigenous transport«. IT jalanan:»IT« stands for Information Technology and»jalanan« literally means»street«, a popular space for marginalised people to live. In this paper, jalanan stands for an activism that advocates marginalised people. Jabodetabek: Acronym denoting the cities of Jakarta, Bogor, Depok, Tangerang, and Bekasi, also known as Greater Jakarta (Jakarta with the satellites cities of Bogor, Depok, Tangerang and Bekasi). Jamkes Watch: Health Protection Watch, a watchdog for the government’s health care program. KATO: Komite Aksi untuk Transportasi Online or Action Committee for Online Transport. KATO is an alliance of broader groups and various driver organisations initiated by Indonesian Trade Union Congress(KSPI). KSPI: Konfederasi Serikat Pekerja Indonesia(Indonesian Trade Union Congress), one of the largest national federations of Indonesian labour unions. Komunitas ojol: Community of online drivers. Community is the most popular form of organisation among app-based drivers. Mobil komando: Command cars equipped with loudspeakers used in workers’ demonstrations Ojek: Motorbike taxi Ojol: Acronym for»ojek online«, which means online motorbike taxi. Opang: Acronym for»ojek pangkalan«, the indigenous motorbike taxi service( ojek), where drivers wait for fares at a stand( pangkalan). PRA: Power Resources Approach Satgas: Acronym for Satuan Tugas, which literally means a taskforce. In this analysis, Satgas refers to a task-force deployed by platform firms to control and discipline driver communities. Satgas officers may go to driver communities to perform undercover surveillance. Officially, these Satgas staff are referred to as»partnership engagement«, a sort of public relations. SPDT-FSPMI: The Aerospace and Transportation Workers division of the Federation of Indonesian Metal Workers’ Union(Serikat Pekerja Digantara dan Transportasi Federasi Serikat Pekerja Metal Indonesia/SPDT-FSPMI) Tuyul: In this article this is a term used by drivers to denote a fake global positioning system(fake GPS). Literally, tuyul means a spirit that obtains wealth for its human master. It is a mythical spirit in Malay mythology in Southeast Asia, especially in Indonesia, Malaysia, Brunei, and Singapore. URC: The acronym for Unit Reaksi Cepat, which stands for Rapid Response Team. URC consists of several representatives(between 2–4 drivers) from each driver community, specifically responsible for assisting drivers if they have conflicts with indigenous transport drivers or if their vehicle breaks down on the road. URC operates only at sub-district level, but easily connects on a broad scale with almost all URC teams throughout Indonesia. Wadah ojol: Association of online drivers. Literally, wadah denotes a meeting place, or a place to associate. Driver associations or wadah of driver communities is a broader scale of driver organisation, whose members come from various communities, but it may also include individuals as members. 28

IMPRINT ABOUT THE AUTHORS Fahmi is a labour activist based in Indonesia. His recent publications include an edited volume“Resistance on the Continent of Labour: Strategies and Initiatives of Labour Organizing in Asia”(AMRC, 2017). Syarif is a Director of Lembaga Informasi Perburuhan Sedane(LIPS- Sedane Labour Resource Centre), Indonesia. He has widely published on the history of Indonesian labour movement, labour laws and policy, and popular labour education. Sugeng is currently involved in developing popular education for labour in several industrial areas in Indonesia. He focuses his research on labour control and labour resistance in app-based transport sector. He is also one of the editors of the Indonesian labour journal Majalah Sedane (www.majalahsedane.org). Friedrich-Ebert-Stiftung| Global Policy and Development Hiroshimastr. 28| 10785 Berlin| Germany Responsible: Mirko Herberg| International Trade Union Policy Phone:+49-30-269-35-7458| Fax:+49-30-269-35-9255 www.fes.de/gewerkschaften To order publications: Julia.Schultz@fes.de Commercial use of all media published by the FriedrichEbert-Stiftung(FES) is not permitted without the written consent of the FES. Dina Septi is currently working with women workers’ learning forum in several industrial cities in Indonesia. She focuses her study on women and capitalism and reproduction work. ACKNOWLEDGMENT The authors would like to thank Melisa Serrano, Thomas Greven, and Mirko Herberg for their meticulous comments and suggestions on the draft of this chapter, and to the FES for funding this research. Special thanks to Michele Ford, Hari Nugroho, Joanna Octavia and colleagues at LIPS and all transport drivers we met, for their comments, critique, and supports. Any errors found in this chapter would be solely authors’ mistakes. The views expressed in this publication are not necessarily those of the Friedrich-Ebert-Stiftung(or of the organization for which the author works). This publication is printed on paper from sustainable forestry. ISBN 978-3-96250-638-4

RESISTING EXPLOITATION BY ALGORITHMS Drivers’ Contestation of App-based Transport in Indonesia This paper discusses the rise of the appbased transport business in Indonesia, taking the case of two currently predominant app-based transportation companies(Go-Jek and Grab), and how workers have responded to it by devising different organisational models and organising strategies. It argues that drivers of app-based transportation in Indonesia have been innovative in their attempts to organise to fight for their rights in the face of adverse circumstances. Workers in this newly emerging sector are aware that they occupy a strategic point in the production of app-based transport, and they have the capacity to disrupt the business. Solidarity-building in actual practice as well as the common practice of providing mutual aid among drivers have become well-established in most driver organisations, clearly demonstrating drivers’ great potential through their significant associational power resources. The research specifically analyses a drivers’ union (SPDT-FSPMI)’s organising initiative in the form of a case study, employing the power resources approach, which looks at the union’s attempt to mobilise its power resources. The paper highlights the impact of appbased transportation, covering issues of digital platform domination, the imbalance in bargaining power between companies and drivers, algorithmic labour control, and occupational safety and health problems. Power resources that different workers‹ organisations use to confront app-based transport companies as well as the government are analysed. Experience with labour organising and the collective repertoire that some workers have developed on the basis of their past experience provide an important basis for the articulation of political demands. It also discusses the gender dimension of this gig economy, touching upon specific issues of female workers in the sector. The paper concludes with a few reflections on how driver organisations in Indonesia could mobilise their power resources to achieve recognition of drivers as regular workers entitled to labour rights. For further information on this topic: https://www.fes.de/lnk/transform