Varieties of platform unionism: a view from the Global South on workers’ power in the digital economy

some cases, these self-organised worker collectives go beyond established forms of unionism in attempts to control the platform technologies. We conclude by suggesting that the experiments of platform workers with new forms of power and organisation, particularly in the Global South, are important to follow in the Global North.


Introduction: The global labour movement under pressure
For some decades, analysts and observers have been declaring the 'end of labour' and a weakening of the global labour movement.We saw this in the work of Manuel Castells (1996) on the impact of new information technologies, and now in Klaus Schwab (2016), with his argument that the widespread use of digital technology constitutes a Fourth Industrial Revolution.There is little doubt that the global labour movement is no longer the social force it was in the twentieth century.Indeed, the rise of digital labour platforms has led to widespread predictions that worker resistance would be unlikely in a digitalised world, in which geographical dispersion of isolated and individualised workers would lead to a fragmentation of old collectivities and a collapse of established forms of worker representation.The classification of platform workers as self-employed or independent contractors has already stripped them of longstanding rights and protections, fundamentally reducing their bargaining power (De Stefano, 2016;Aloisi and Gramano, 2019).Others have predicted that platform technologies such as algorithmic management, 'gamification' and the 'digital panopticon' would dramatically reduce workers' capacity to challenge the demands of capital (Srnicek, 2017).
Despite predictions that the platform model would make worker organisation increasingly improbable, the protests of platform workers have grown.The Leeds Index of Platform Labour Protest has shown the rapid development of platform worker organisation and resistance across the globe (Bessa et al, 2022), and identifies many examples of gig workers' mobilisation.Indeed, Cant (2019) argues in his account of Deliveroo in the United Kingdom (UK) that digital management methods facilitate courier mobilisation.The platform, he claims, is a breeding ground for self-organised courier associations, which are boosting efforts to organise (Cant, 2019: 16).As the Leeds Index shows, however, organisations such as these, where formal employment and collective bargaining are rare, and unionisation rates are low (Basualdo et al, 2021;ILO, 2021), do not fit easily into established frameworks of labour relations.Some platform workers are organised into established labour unions, most commonly in Europe, but others are involved in much smaller grassroots unions, and growth has been observed in new online workers' collectives (Vandaele, 2021).In the Global South, particularly, hybrid forms of collective organisation are now being explored (Webster et al, 2021).For workers in the Global South, then, an important question concerns the impact on their ability to organise of the rise of on-location platform work -which involves platform workers fulfilling tasks in a specific physical location -and the forms of organisation being created.
From a historical perspective, technological change has always been contested and has led to intense struggles between capital and labour, and to new forms of work organisation.New 'technological fixes' 1 (Silver, 2003: 66) have often led to contradictory developments, aiming to weaken organised labour in existing strongholds of the labour movement, but at the same time, creating new opportunities to organise elsewhere.For instance, the introduction of the assembly line in the automobile industry aimed at rigid Taylorist labour control, but at the same time spurred coordinated strike actions in parts of the Global South such as in Brazil and South Korea (Silver, 2003: 54ff).Digitalisation plays such a contradictory role today.Algorithmic management in the platform economy can be seen as a new technological fix (Vandaele, 2021;Schaupp, 2023), bypassing existing labour laws and institutional employment standards, but it is also leading to new struggles in the platform economies of the Global South.
This contested process, and the implications for the Global South of the making and remaking of the global working class, will be examined here in the context of on-location platform work in the transport and food delivery industries.The article is thus primarily a conceptual contribution to the emerging debate over platform unionism and new forms of labour struggle in the global platform economy (Vandaele, 2018;Cini, 2023;Joyce et al, 2023).By referring to specific experiences in Africa and Latin America and to the results of a global research project, 2 it is argued that platform workers have mobilised new sources of power, and that a variety of forms of worker organisation are emerging.These extend from more established forms of unionism and hybrid forms of organisation to self-organised networks from below.To develop this argument, we use a power resources approach (PRA) in the second section to analyse technological change and discuss the divergent responses of organised labour to these new technologies in the third section.In the fourth section, we take a historical perspective to examine the role played by digitalisation in the reshaping of the global labour movement.We argue in the fifth section that particularly in the Global South, platform workers tend not to organise into established unions but rather into ad hoc groups drawn together around specific grievances, as well as hybrid forms of organisation, informal and self-organised from below.By analysing two cases of platform unionism in Argentina and Uganda in the sixth section, we distinguish different forms of worker organisation and in the seventh section, discuss the varieties of platform unionism which we have observed.We conclude in the eighth section that analyses of platform unionism need to break with the traditional one-sided focus on established labour unions and explore the new forms of power and organisation with which workers are experimenting.

How technological change can alter workers' power
Technological change has been a key driver of capitalist development and has frequently shifted the balance of power between capital and labour.From the Luddite movement in Great Britain in the early nineteenth century opposing the introduction of the mechanised loom to the struggles over rigid Taylorist labour control at the assembly line in the twentieth century, workers have challenged technological change by organising and disrupting the capitalist production process.A useful tool in an understanding of these changes and their impact on workers' power is the power resource approach (PRA).The PRA builds on Erik Olin Wright's and Beverly Silver's concept of associational and structural power (Wright, 2000;Silver, 2003: 13-16) and the subsequent debates over workers' power (Lévesque and Murray, 2010;Brookes, 2013;Schmalz et al, 2018;Vandaele, 2018;2021;Basualdo et al, 2021;Refslund and Arnholtz, 2021;Schmalz and Webster, 2024).Our concept of workers' power builds on these discussions to add two further sources, institutional and societal.A foundational tenet of the PRA is the capacity of workers to build power through the collective mobilisation of a range of specific resources by which they can advance their interests in the asymmetric capital-labour relationship.
We identify four sources of workers' power: structural, associational, institutional, and societal: The first, structural power is a product of the position workers occupy within the economy (Wright, 2000: 962;Silver, 2003: 13ff).Following Silver's argument, structural power arises in two ways, the first of which is from the ability to disrupt the production process (workplace bargaining power) by work stoppages, or, more generally, labour unrest.Mobilising workplace bargaining power through strike action results in rising costs for employers and can allow wage earners to enforce concessions.The second subcategory, marketplace bargaining power, in turn, stems from a tight labour market and thus from 'having rare qualifications which are demanded by employers, little unemployment' and from the 'capability to completely withdraw from the labour market and to live on other sources of income' (Silver, 2003: 13).Workers also have capacity outside the production process, taking 'structural power outside the workplace and into the public domain' (Webster et al, 2008: 13).This resource, termed logistical power, is a specific type of workplace bargaining power, usually mobilised by self-employed informal workers through occupations or street blockades, and is particularly important in the Global South (Webster, 2015: 119).
The second, associational power, as the name suggests, is derived from collective association and hence requires the emergence of workers capable of mobilising collective action.This would include trade unions, (informal) self-organised networks at and beyond the workplace level, hybrid forms of organised labour, and political parties, with unions historically playing a crucial role in the representation of labour.Associational power is only loosely indicated by membership numbers; the strength of the collective is also determined by factors such as infrastructural resources, internal cohesion, and membership participation (Lévesque and Murray, 2010;Schmalz et al, 2018: 118ff).
Thirdly, institutional power refers to the labour law and institutional rights upon which organised labour can draw.Its strength, a product of struggle and negotiation, is relative to that of the structural and associational power that labour in that context possesses.Institutional power is a double-edged sword as it is not simply emancipatory: it may grant organised labour rights, but it also imposes a system of institutional rules that restrict labour's action (Runciman, 2019).The unique feature of institutional power is its stability over time, as institutions lay down basic social compromises that transcend short-term political changes (Dörre et al, 2009).
Finally, the societal power available to organised labour is a function of its collaboration with other social actors and public support.It has two subtypes: coalitional and discursive.Coalitional power arises from networks with social organisations and movements and is used to mobilise support for workers' actions (Frege et al, 2004:137ff;Turner, 2006).Discursive power, on the other hand, is the 'ability to intervene successfully in areas of the public sphere that are pre-structured on a hegemonic basis' (Urban, 2012: 222; see also McGuire and Scherrer, 2015) and is used essentially to assert hegemony in public dialogue in a Gramscian sense on Unauthenticated | Downloaded 10/18/23 09:59 AM UTC labour-related topics.All four power resources are interconnected and embedded in the power relations between capital and labour and their specific conjunctural societal environment.They are, consequently, influenced by technological change and the development of capitalism.
The use of new technologies has shaped global capitalism and workers' power (Figure 1).In many cases, they have not only strengthened company competitiveness, but changed the relationship between capital and labour itself.Beverly Silver has argued that capital uses 'technological fixes' to respond to labour unrest by implementing major process innovations 'to fix the problems of profitability and labor control' (Silver, 2003: 66).A striking example of a technological fix is 'containerization and dock automation in the shipping industry', which dramatically downsized 'the historically militant dock labor force in the second half of the twentieth century' (Silver, 2003: 101; see also Levinson 2006), thereby weakening organised labour in the freight industry.But the implementation of new technologies has also led to contradictory developments.The introduction of the assembly line in the early twentieth century went hand in hand with Taylorist labour control, but also facilitated coordinated strike action in the automobile industry and the development of the US labour movement in the 1930s, laying the foundation for Roosevelt's New Deal politics (Skopcol et al, 1990).Similarly, automobile workers in Brazil, Korea, and South Africa in the 1970s and 1980s became the vanguard of new labour movements, and since the late 2000s, there have been large strike waves in China's automotive sector (Silver and Zhang, 2009).New technologies can thus potentially lead to a 'creative destruction' (Schumpeter, 1994) of outdated business models and shape work organisation and labour relations.
From a PRA perspective, technological innovation thus tends to transform structural power.With changing means of production, the ability to stop the production process (workplace bargaining power) is transformed, enabling or reducing the disruptive capacity of workers.At the same time, they deskill a part of the workforce and allow new groups of qualified workers to emerge (marketplace bargaining power).This changing structural power transforms associational power as well.The restructuring of the workforce, of the membership base, and of the labour process usually results in stark challenges for organised labour, such as gaps in representation, but it can also lead to potential benefits through increasing and diversifying union membership, for example through changing gender divisions of labour.In addition, new technologies trigger struggles over regulation of the new forms of employment and work (institutional power) and in some cases are also subjects of political discussion, as in the case of data security, opening up new possibilities for networks with non-governmental organisations and social movements (societal power).

Divergent responses of organised labour to new technologies
The labour movement has historically responded to technological change in three ways that draw on diverging power resources (Table 1).The first could be called wrecking and is aimed at controlling the introduction of new labour-saving technology by damaging or destroying capital.The early predecessors of the labour movement in the eighteenth and early nineteenth centuries, such as the British Luddite movement (1811-13), for instance, were 'machine breakers' (Hobsbawm, 1952).Their destruction of machines was a form of 'collective bargaining by riot' (Hobsbawm, 1952: 59) to obtain concessions on wages and working conditions, as the protesters feared being replaced or displaced by labour-saving technology.Organised 'machine breaking' as a mobilisation of structural power spread globally and became a part of the 'repertoire of contention' of the early labour movements in the Global South (Van der Linden, 2008: 174).Defiance of new technologies has also taken less violent forms.In the late twentieth century, labour unions such as the 'Society of Civil and Public Servants' in the Thatcherite UK actively campaigned against the use of computers in public administration because public employees feared job losses due to technology-driven rationalisation.
The second response of the labour movement to technological change could be called adaption, use of the structural power granted by new technologies to strike, organise and negotiate better conditions for labour.Historically, the introduction of the mechanised loom, which the Luddites violently contested, helped to nurture and encourage the organisation of textile workers; similar processes can be observed in the assembly line and emerging automobile workers unions (Cohen, 2014) and, more recently, in 'choke points' in global logistics and networked just-in-time production (Alimahomed-Wilson and Ness, 2018;Fichter et al, 2018: 7f).New technologies have consequently contributed to the emergence of worker organisations (associational power) and often to the development of collective bargaining and codetermination institutions.Adaption thus relates to what Hobsbawm has called the 'natural time-lag' between the point when changes in production are introduced and 'the habit of industrial solidarity [that] must be [re]learned' (Hobsbawm, 1964: 144) pushes unions to experiment with new techniques of organisation, mobilisation and labour action.Adaption thus aims to boost both associational and institutional power.Strategies of adaption have been the most common and significant response of the labour movement to technological change so far in both the Global North and South (Hlatshwayo, 2017: 101ff). 3 The third response of the labour movement is a strategy that could be called worker control and requires a high degree of associational power to be effectively applied.In the late 1960s and early 1970s, when the movement in advanced capitalist countries was at the height of its power, workers' mobilisation focused on the control of technology and the production process itself (Schmalz and Weinmann, 2016).In some companies, new worker-driven socio-technical approaches were adopted.A wellknown example is the Volvo Kalmar plant in Sweden, inaugurated in 1974, where an assembly system with independent production teams replaced the conventional hierarchical assembly line. 4Another is the 1976 Lucas Plan in the UK, created in response to the company's announcement that thousands of jobs were to be cut to enable industrial restructuring in the face of technological change and international competition.Rather than being made redundant, the workforce argued for their right to develop socially useful products (Cooley, 1980).Predecessors of worker control can be found in the cooperative movement in the US and Europe.The movement continues to play an important role in South America (economia solidaria) and is largely dependent on external support, thus solid societal power.

The making and unmaking of working classes: digitalisation and algorithmic management as 'a new technological fix'
Technological change has also shaped the nature and form of labour struggle, greatly contributing to the 'unmaking and remaking of working classes' (Silver, 2003: 22).In her study of global labour unrest, Silver frames her argument in terms of 'a Marx-Polanyi dialectic' where capital overcomes impediments to accumulation through various fixes (spatial, production, financial and technological), with the technological fix being the most important to an understanding of technological change and its impact on the struggles of labour.For Silver, these fixes result in cyclical processes of capital formation that make, unmake, and remake the working class.Struggles of the old sections of the working class that resist the 'recommodification' or 'de-commodification' of their labour power through attacks on both wages and social wages (or even its employment altogether) are framed as the 'Polanyi-type'.These are 'the backlash resistances to the spread of a global selfregulating market, particularly by working classes that are being unmade by global economic transformations as well as by those workers who had benefited from established social compacts that are being abandoned from above' (Silver, 2003: 20).It is this section of the working class that made material gains during the heyday of social democracy, essentially de-commodifying their labour power by regulating its sale and ultimately taking it out of the free market altogether.Attacks on this section of the working class often result in the 'unmaking' of its original location through privatisation or offshoring, or in the case of technological fixes, through rationalisation and deindustrialisation as, for instance, in the Western European steel industry since the 1980s.
Marx-type struggles, on the other hand, are presented by Silver as emerging from new sections of the working class that are made or re-made through the various capital fixes and new capital formations.These are 'the struggles of newly emerging working classes that are successively made and strengthened as an unintended outcome of the development of historical capitalism, even as old working classes are being unmade' (Silver, 2003: 20).Marx-type struggles thus comprise the unrest in the US automobile industry in the 1930s, which was related to the introduction of the assembly line and Taylorist labour control.Silver traces the unfolding of this process -the remaking of the working class -on a global level as impediments to accumulation in the textile and auto-industry forcing relocation of production to cheaper parts of the world, thus generating new struggles against it.
The Marx-type struggles in emerging industries tend primarily to be offensive, aimed at establishing workers' power and rights from a point of extreme vulnerability, and are rooted in growing structural power.In most countries today, Polanyi-type struggles in declining industries such as coal tend in turn to be defensive, rooted in shrinking structural power, and increasingly based on existing institutional and associational power resources.Both forms of struggle are ideal types in the Weberian sense (Weber, 1978).Hybrid forms of these struggles exist as well, particularly in transforming industries like automotives.From the perspective of the Global South, their story is relatively straightforward: new industries were first established in the core countries of the capitalist world and later spread to the 'less developed' parts of the world or the Global South.This is what Harvey (2003) has called a 'spatial fix' as these successive geographical relocations of capital succeeded only in rescheduling crises in time and space.
Silver has observed a global pattern of relocation of production in lead industries in a search for cheaper labour.In the 1870s, for example, the textile industry was hit with a large wave of labour unrest first in the UK and later, in continental Europe, the US and Japan.Today, important textiles and garment production sites have been relocated to countries such as Pakistan, Cambodia, Bangladesh and Myanmar, where new labour movements have developed (Zajak, 2017;Serrano and Nuon, 2018).There is, consequently, a spatial and technological logic at play: in the Global South, Marx-type offensive struggles and labour movements in lead industries normally emerge later, after the unmaking of labour movements in the core. 5This unmaking of the working class is associated with defensive Polanyi-type conflict, as has been the case with labour struggle in the textile industry.In China in the 2010s, for example, some of the largest strikes were about social security and pension payments in the garment sector (Schmalz et al, 2017).Consequently, the Global South, once a late comer to the global pattern of labour struggle, today as a result of industrial relocation hosts about 80 per cent of the world's industrial working class (Smith, 2016: 103).This struggle takes place under the conditions of segmented and informalised labour markets and global value chains, however, with companies in the core countries commonly reaping most of the profits in the chain.
Digitalisation and algorithmic management can be seen as a new technological fix (Vandaele, 2018;2021), with the platform economy, as the new leading economic sector (Joyce et al, 2023).There have been several waves of algorithmic management from the computer-integrated manufacturing (automation) of the 1980s to the remote work of the COVID-19 pandemic since 2020 (Schaupp, 2023).Particularly after the financial crisis of 2008, the platform economy has been the beneficiary of huge capital flows, helping to transform tech start-ups such as Airbnb, Uber and Zoom into global players (Srnicek, 2017).By extracting, analysing, and using data as a raw material, artificial intelligence, and digital management methods (app tracking and so on) in the platform economy have transformed the organisation of work (Vandaele, 2018: 5).Similar methods have been introduced in smart manufacturing (big data and so on) and logistics (warehouse automation).In the platform economy, online technologies not only help to match demand and supply but also to 'track and discipline workers, in many cases circumventing or flouting existing labour and health and safety regulations, to the detriment of platform workers' social protection' (Vandaele, 2018: 5).The rise of digitalisation and the 'gig' or platform economy has led to a far-reaching technological disruption.Industries such as financial services, retail, transportation, hospitality, and food delivery are being transformed through the application of platform-based services.Importantly, in contrast to the logic of earlier technological and spatial fixes, the platform economy has quickly become global.Recent surveys have shown that the share of total employment occupied by platform workers in the Global South is high.In the case of China, for example, 2020 figures show a total of 84 million platform workers or almost 10 per cent of the country's overall employment (CLB, 2022).
At least two economic realities are responsible for this.First, the costs of replicating capitalist means of production in the platform economy are much lower than they are in manufacturing (Mason, 2016) because of the low level of fixed capital required to launch new businesses in that mode.This allows transnational digital companies to enter foreign markets quickly and/or local competitors to grow in domestic markets.Customers in the transport industry today, for example, rely on local gig companies such as Ola (India), Rappi (Colombia), iFood (Brazil), Didi and Meituan (China), Gojek (Indonesia) and Grab (Singapore). 6Second, in many companies the fast-developing platform economy is based on new business models where the legal regulation of work is weak, if not lacking, and self-employed and informal labour play a major role.The ILO reports that in most countries, 'the regulatory response to platform work is in flux' (ILO, 2021: 211); draft legislation is rigorously debated, and court rulings are being appealed by digital platforms.Business practices in the gig economy in Europe and to a lesser extent in the US, for instance, conflict with existing regulations, while the platform economy of many less regulated countries in the Global South perfectly harmonises with the informal sector, with its various groups of workers, for example, informal and (misclassified) self-employed.Consequently, although the most advanced digital technology such as AI research continues to be developed in the US, EU and East Asia, many applications are quickly available in the Global South with some local platform companies even having the capacity to leap-frog and conquer significant domestic market shares.

Labour unrest in the platform economy: the case of the transportation sector
What then does the rise of the platform economy mean for the labour movement in the Global South?The emergence of the platform economy has led to strikes and organising efforts across the globe reaching from Africa to Latin America and Asia (Dirksen and Herberg, 2021;Haidar and Keune, 2021;Webster et al, 2021;CLB, 2022).Between January 2017 and mid-2020, the Leeds Index of Platform Labour Protest has identified 1,271 cases of platform worker organisation and resistance (Bessa et al, 2022: 16ff).Collective mobilisation of workers in on-location digital platforms has taken place particularly in the field of transportation, in for example, food delivery and courier services, which are among the platform companies that have mushroomed in the last decade (Trappmann et al, 2020).These protests follow a tradition of labour unrest in the transportation industry, which was formerly among the sectors where workers' robust workplace bargaining and logistical power underwrote a history of continuous conflict (Silver, 2003: 97-103).
Organising in the platform economy generally takes place under difficult conditions.In the absence of a clear regulatory framework for digital platform activities, many platform companies in transportation are now characterised by non-wage forms of employment, with workers being paid through clicks or orders as at Deliveroo or Uber (Vandaele, 2023).After a short 'honeymoon period' of platform work with reasonable flat rates per ride or delivery, big bonuses, lots of incentives, and flexible working hours, among other perks, issues such as shrinking pay per ride, disguised self-employment and increasing pressure through the app's algorithmic management systems are now spurring worker organisation.Although the platform sector creates employment and a livelihood for many workers in countries of the Global South with high unemployment rates, it also thus produces 'a precarious class of dependent contractors and on-demand workers' (UNCTAD, 2021: 14).As a result, the new 'digital precariat', increasingly conscious of its insecure working conditions, is engaging in labour disputes.These offensive Marx-type struggles are driven primarily by a desire to regulate and decommodify employment relations and to increase wages. 7 They are pushed by the basic interests around which labour rallied and organised in the era of the industrial revolution, in particular, wages, guarantees of continuing wages (employment security), and working conditions (Offe and Wiesenthal, 1980: 82).The 'willingness to act' by platform workers who are not usually recognised as wage labourers by platform companies, has developed through an emerging collective identity as workers.
Platform workers in transportation normally rely on a specific set of power resources.Significant workplace bargaining and logistical power gives them solid structural power.While they can generally organise strikes and blockades, however, the marketplace bargaining power of platform drivers, riders and delivery workers is comparatively low.This is a result of the informalised and precaritised labour markets in most countries of the Global South.In addition, institutional power is most often low or non-existent, as platforms are not regulated by existing labour laws and may even be exploiting their loopholes.At the same time, many existing regulations cannot be applied to industrial action due to the (mis)classification of platform workers as self-employed.Regarding societal power, transport platform workers can usually exert discursive power by staging 'public drama[s]' (Chun, 2005) and, by meeting in public spaces, articulating their demands in the public sphere.Coalitional power, in turn, differs case by case and is dependent upon country-specific partnerships with established unions, NGOs, and social movements.Some scholars of the platform economy, who have analysed new intersections of unions and social movements, highlight the importance of gig workers' societal power (Fernández-Trujillo Moares et al, 2023;Weghmann, 2023) and identify, in that sector, a new form of 'social movement unionism' (Fernández-Trujillo Moares et al, 2023: 804). 8Associational power through collective organisation has been judged to be crucial in the platform economy, however, as the workplace and marketplace bargaining power of platform workers is considered essentially to be futile, although opportunities vary with types of platform work (Vandaele, 2018;2021;Joyce et al, 2023).
Yet, the associational power of transport platform workers is most complex to assess.Like the factory production sites in the nineteenth century, location-based service platforms bring together workers who formerly worked separately from each other.By linking platform workers technologically, the gig economy tends to boost their workplace bargaining power and thus contribute to the emergence of self-organised, hybrid forms of union-like associations (associational power) (Cant, 2019).This is particularly the case since established transport unions were initially virtually absent.Platform workers are often labour market outsiders, and many of the existing unions have no history of organising precarious workers (see Cini et al, 2022).It is, therefore -based on considerable workplace bargaining and discursive power, and the work context of the location-based platform -relatively easy for workers to organise at a grassroots level.As a result, grassroots platform associations have been mushrooming in the Global South.Such collective associations carry the 'imprint of their time', as they emerge from digital online communities and make use of digital technologies in general (Stinchcombe, 1965;Vandaele, 2021).They work with such limited infrastructural resources and hostile legal environments that collaborations with established labour unions are questioned.There is no such a thing as a universal 'platform unionism', however.Distinctive models of collective associations have emerged from the platform economy, with a variety of patterns in their relationships with 'established' unions.

Two cases of platform unionism in the Global South
The new platform unionism is thus largely dependent upon the relationship between emerging grassroots associations and established unions.This relationship can take a variety of forms, which, especially in the Global South, include new experimental trends in unionisation.In the capitalist core countries, formal labour tends to dominate, with institutional rules regulating most parts of the labour market.Organisational boundaries are also usually clear, and the role of established unions, predominant, in new offensive Marx-type labour struggles.In several European countries, however, tensions have been rising between established unions and grassroots unions or workers' collectives (Joyce and Stuart, 2021).In many countries of the Global South, the context of union organising is different: substantial proportions of the labour force are employed informally, organisational boundaries are less clear, institutional rules are less well defined, and workers' associations play a prominent role in their organisation.
In what follows, two cases of platform unionism are selected for analysis of this new experimental trend of unionising in the Global South.The first is the case of Argentina's Association of Platform Workers (APP, Asociación de Personal de Plataformas), and the second is Uganda's Amalgamated Transport and General Workers' Union (ATGWU).Each represents a different variety of platform unionism: the first is an independent and autonomous collective, and the second, a hybrid organisation that comprises associations of informal self-employed workers.These form a part of a larger project based on eight cases in the Global South, 'Trade Unions in Transformation 4.0' (TuT 4.0) funded by the Friedrich-Ebert-Foundation (see Note 2).The cases chosen here represent the most divergent forms of organisation found in the project's overall sample, and highlight the range of diversity of its forms of platform unionism (Basualdo et al, 2021: 13).
Both the Argentinian and Ugandan case studies are based on original, qualitative, research which include two data sources: primary and secondary literature sources; and semi-structured interviews.For both, the first comprises union and government documents, high-quality newspaper articles, scientific literature, and survey results.For the case of APP in Argentina, the second involves semi-structured interviews conducted between 2019 and 2020 with three platform workers affiliated with APP, two legal representatives, and an expert on trade union issues.For the case of AGTWU in Uganda, four interviews were conducted between 2019 and 2021 with representatives from the International Transport Workers' Federation (ITF), ATGWU, and an informal taxi driver from Kampala.Researchers in both studies adopted a public sociology approach (Burawoy, 2021), collaborating with trade unionists in the research and knowledge transfer process.For the case study of APP, the results of the FES-study by Perelman et al (2020) were updated with more recent information; for the study of ATGWU, the FES study by Manga et al (2020) was referred to, but the more recent work of Webster et al (2021) and Webster and Ludwig (2023) was used as a main source.

Autonomous organisation: Asociación de Personal de Plataformas in Argentina
Argentina's APP is a case of formalisation of grassroots initiatives in on-location platform work.After several months of operation, the APP was reformed as a union organisation but struggled to meet all legal requirements for formal union recognition and remained independent of the established unions (Perelman et al, 2020).Argentina has a highly regulated system of collective interest representation, which tends first to protect labour market insiders, and has been dominated by the Peronist union federation General Federation of Labour CGT (Confederación General del Trabajo).This solid -albeit somewhat rigid to outsiders -trade-union tradition seems to have played an important role in both promoting the identification of APP riders as workers and supporting their attempt at collective organisation.It thus led to a Marx-type struggle over payment and working conditions.
The Colombian startup unicorn Rappi had been present in Argentina for less than six months when discontent among its couriers began to mount.Like many other food delivery companies such as Glovos and Pedidos Ya, workers at Rappi are considered to be self-employed, and a large share of the workforce is comprised of male migrant workers (Madariaga et al, 2019).Algorithmic management and a nontransparent performance-oriented qualification rating system at Rappi affects wages and working hours.In July 2018, a sudden change in the company's order allocation algorithm led to Latin America's first 'digital strike': couriers decided to use their workplace bargaining power to stop carrying out deliveries (Perelman et al, 2020;Miguez and Menendez, 2021).Rappi had not informed the workers of the changes to the algorithm it introduced, which intensified their difficulties on the job.The changes gave them 'the feeling that they were "pedalling more and earning less", while they lost the freedom to manage the duration and intensity of their working days' (Perelman et al, 2020: 7).During their digital strike, the delivery riders met at a public space and took orders but did not carry out deliveries, confusing Rappi's algorithm and temporarily gaining them higher delivery payments.
After the strike, an informal dialogue was initiated between Rappi and a group of spokespersons elected by the couriers.This stalled after the APP decided to apply officially to the Ministry of Labour for registration as a union and informed all local platform delivery companies of its intentions.Rappi and the rest of the delivery companies responded with union busting techniques, reducing the direction of orders to APP union members.After a protest at Rappi's headquarters, the blocking of APP's entire executive committee from access to digital mobile applications eventually led to a legal dispute and accelerated the union's formation, which occurred in October 2018 (Perelman et al, 2020: 7).The legal struggle with Rappi went through several iterations and led to a trial on the nature of the workers' employment relation.At the same time, the struggle showed the relevance of APP's lack of trade union status, and thus APP's weak institutional power so that members remained in a highly precarious position.In 2020, this vulnerability was increased by the COVID-19 pandemic, which exerted additional pressure on the delivery workers who had become 'essential' workers at a time of social isolation and distancing (Miguez and Menendez, 2021;Del Bono, 2022).The extreme working conditions involving major health risks drove the International Platform Workers to strike in April and June 2020, with a high degree of worker mobilisation in Argentina and other Latin American countries.In the long run, however, Argentinian protests became increasingly difficult due to pandemic restrictions on mass gatherings in public spaces, the high level of public pressure on the riders (decrease of societal power), and the country's general state of economic crisis.
During this period, the reach and impact of APP began to fade, and other union organisations like the new Grassroot Union of App Delivery Workers (SITRAREPA, Sindicato de Base de Trabajadores de Reparto por Aplicación) and the established Trade Union Association of Bikers, Couriers, and Service (ASIMM, Asociación Sindical de Motociclistas, Mensajeros y Servicio) gained influence with platform workers, although their representation of these workers remained fragmented and without legal recognition (Arias et al, 2022).Thereafter, no single organisation could effectively and comprehensively organise the sector.During 2021 and 2022, collective organisation proved to be increasingly hard for workers with scant legal protection, and protests focused on safety and security issues (Arias et al, 2022).A major factor discouraging labour action was the increasing acceptance of the business model, as well as the difficulties in achieving labour protection for and regulation of the platform economy (Etchemendy et al, 2022).Many projects have been initiated to regulate the sector, but no laws could be passed, and the judicial actions against the companies over the violation of rights are ongoing.
From a PRA perspective, the case of APP shows the ambiguity of structural power for workers employed by food-delivery platforms.They have a high degree of workplace bargaining power through their disruption of on-demand delivery services, but at the same time, the large 'reserve army' within Argentina's labour market and high turnover rate hamper coordinated strike actions.In addition, the APP's lack of legal union status gave it weak institutional power, both of which the organisation struggled to rectify.Weak marketplace bargaining and institutional power both influenced the APP's associational power; as it became complicated to organise new members and achieve a higher union density, the sector fragmented.The APP was able to develop strong societal power, however.The organisation was supported by academic institutions, centre-left politicians, political foundations, and positive press coverage, all questioning the business model of platform companies.However, the APP could not maintain its hold on power over time, due, among other things, to the extremely challenging circumstances of the pandemic.Paradoxically, it could be argued that the highly regulated and institutionalised system of collective interest representation in Argentina was an important element in the couriers' realisation that establishment of a union was crucial.Yet the existing power resources were sufficient to opt for an independent organisation.This approach led to resistance from existing union organisations, however, which fought against the new union initiative.The response of established unions to the new organisation of labour-market outsiders was thus mixed.Some unions representing workers in similar sectors strongly criticised the creation of a new union and rejected the process of independent organisation, with some of them trying to organise the workers themselves.Others, however, backed the APP, expressed solidarity, and supported collective organisation.

Hybrid organisations: Amalgamated Transport and General Workers' Union in Uganda
Uganda's ATGWU faced a near collapse of its membership in the 1980s when liberalisation undermined the public transport sector through the introduction of structural adjustment programs and privatisation.This led to the unmaking of the traditional sectors of the working class through the informalisation of public transport and emergence of mini-taxi drivers and motorcycle taxi riders (known locally as boda boda) throughout the country.In Kampala alone about 250,000 boda boda riders form an important transport sector component.The ATGWU thus made the strategic decision to organise the growing number of boda boda drivers.
The decision was influenced by the 2013 launch of an informal transport workers project by the ITF and their willingness to assist the ATGWU to organise informal workers (Spooner and Mwanika, 2018: 6).The ITF supported its affiliate in Uganda with research, capacity-building workshops, and resources at a time when the ATGWU was facing decline.A crucial feature of these workers was that they were already organised, not into labour unions but through credit and savings cooperatives, informal self-help groups, community-based organisations and, most importantly, associations.The ATGWU did not try to recruit individual informal workers but affiliated each association as a group into the union (Spooner and Mwanika, 2018).The union needed to find new ways of responding to informal workers whose status and identity ranged from small business holder to (self-employed) worker (Webster et al, 2021(Webster et al, : 1368)).A key part of their organising strategy involved mapping the range of organisations in the informal economy, identifying the primary issues faced by these workers and 'who holds power and influence in the industry' (ITF, 2022: 30).
One of the associations that joined the ATGWU was the Kampala Metropolitan Boda Boda Entrepreneurs (KAMBE), with a membership of 64,000 in 2019 (Manga et al, 2020: 4).The ATGWU then established an informal sector committee comprised of all leaders of the affiliated associations, and the merger process began.As the ATGWU organisers explained: Rather than attempting to recruit individual informal workers into union membership, we undertook a sequence of discussions and education events with some of these associations, eventually affiliating each association into the union as a whole.The detailed process was not pre-planned, but rather a sequence of engagements with associations, each leading to contact with the next.(Manga et al, 2020: 30) The innovative response of the ATGWU to this challenge was the adoption of digital applications as tools to provide service to informal workers, an example of unions appropriating digital technology to empower their members (Barrett, 2018).As an ATGWU representative explained in an interview (ATGWU Official, online, 3 May 2021), it was the boda boda drivers who initially pushed the union towards use of digital tools in order to solve the issue of collecting union dues and to build a membership database through mobile phone applications.In the process, members raised the issue that a digital tool was also needed to connect drivers with passengers as an alternative to increasing exploitation through new multinational platform companies such as Uber and Taxify/Bolt.
The union leadership noticed that some of the association's members had begun to join new market entrants' digital platforms -something that can be interpreted as a remaking of work through technological change in the transportation sector -and that there was a growing interest among the boda boda drivers to use such technology (Barrett, 2018: 6).In 2017, the AGTWU concluded that it needed to support KAMBE in developing its own app, which should serve multiple purposes, such as the collection of membership dues, development and maintenance of a database, and service as a hailing function and chat room.The challenge for the union was to acquire external resources, which it received from the Friedrich-Ebert-Foundation (FES).
After several setbacks, the union was finally successful in setting up two apps, both of which are now in operation.The first is the KAMBE app, which aims to serve members by facilitating the collection of union dues, managing a members' database, and enabling communication among members.Only members of the association and ATGWU have access to the app and to other benefits such as loan facilities and insurance schemes.The second is the SOT Boda App, which is purely for ride-hailing.It was developed together with another association in the union, the SAGULA Online Transporters Association.The two apps have different functions: the first is a nonprofit networking app for KAMBE members, and the second aims to generate business for riders under fair conditions.Functions are separated to ensure the protection of members' data, but also to reflect the different roles of the associations and the union.It therefore mirrors the 'hybrid' nature of a union with a large membership of informal workers (Webster et al, 2021).
The focus of the ATGWU's organising efforts was associational power, and the unionisation of cooperative associations of informal workers, resulting in a hybrid organisation representing both formal and informal workers.This was possible due to partnerships, with ITF and FES hinting at transnational associational power together with coalitional power.Facing the challenge of digitalisation, the ATGWU adopted a strategy that goes beyond mere adaption to technological change and aims, rather, at controlling platform technologies (Webster et al, 2021.With its apps, the ATGWU was able, both to digitally organise, and to protect its members from overexploitation. Several divisions continue to hamper the transition to a fully integrated union of formal and informal workers, however.At the centre of these tensions is the nonpayment of membership fees by the informal workers.It is estimated that 85 per cent of the informal workers are not paying membership fees (interview, ATGWU official, 13 March 2023).The difficulty with the informal workers is that they get paid irregularly and their income varies.According to research undertaken in 2020, the daily income of boda boda riders belonging to KAMBE fluctuated between 8,000 and 20,000 shillings (Spooner et al, 2020: 58ff). 9Interestingly the calculation of expenses included the annual cost of police bribes; in the case of one rider, it came to an average of 300,000 schillings a year.

Discussion: varieties of platform unionism
These cases represent divergent approaches to organisation in the transport sector and, thus, different forms of platform unionism (autonomous vs hybrid organisation).Each relies on a specific institutional, organisational, and structural context and is thus not easily replicable in other countries and world regions.Referring to other case studies in our project and the literature on platform unionism, we have identified several varieties of platform unionism (see Figure 2).
These varieties can be understood as a continuum of relationships between trade unions -understood as membership-based worker organisations that engage in collective bargaining and the safeguarding of workers' rights -and workers' associationsdefined as less institutionalised grassroots-networks of workers that engage in collective action for workers' interests.Along that continuum are: 1) autonomous organisation, 2) cooperation, 3) hybridisation, and 4) integration.The APP is a case of autonomous organisation with some elements of cooperation with established unions, while the ATGWU represents a mixed case of integration and hybridisation.The two cases, which could hardly be more different, are located at the poles of the possible variations.To better understand the typology, it is helpful to contrast the Global-South cases of the APP and the ATGWU with two cases in our sample from the capitalist core.In the Netherlands, for example, after a period of self-organisation, protesting couriers decided to join an existing union confederation, the Dutch Trade Union Federation (FNV, Federatie Nederlandse Vakbeweging) (Vandaele, 2020).This type of action, which can be termed integration, is completely contrary to the autonomous organisation of the APP.It also differs from the ATGWU as the rider associations of transport workers were integrated into the union but kept their original form thus leading to a hybridisation of both organisations.
In the case of Belgium, the Couriers Collective, also an association of self-organised workers, opted for a different relation to the union than that of Argentina, Netherlands, and Uganda.The Couriers Collective remained autonomous but decided to work with two of the most important union confederations, with one of them creating its own union structure for freelancers and the (misclassified) self-employed, which included the food-delivery couriers (Vandaele, 2020).A similar arrangement of 'social media unionism' holds true in Denmark, with food-delivery couriers being mostly self-organised online but connecting and partnering with the primary established union (Hau and Savage, 2022).It is important to note that these forms are also present in sectors other than transport, such as the case of cooperation between the YouTuber Union and IG Metall (Niebler, 2020) and that of the integration of the Google workers (Alphabet Workers Union) in the CWA (Communication Workers of America).It is likely that the varieties of hybridisation and autonomous organisation are more present in the Global South, a situation related to peripheral capitalism, the important role of informal labour, and labour market-insider oriented labour relations, while cooperation and integration are more frequent in the Global North, resulting from formal labour markets and clearer organisational boundaries.It is therefore possible to argue that across the globe, there exist at least four different varieties of platform unionism, ranging from autonomous organisation to integration.

Conclusion: Experimental trends and new platform cooperativism
We have argued here that platform workers in the transport sector are experimenting with new forms of power and organisation.Digitalisation, with its new types of algorithmic management in the platform economy, can thus be perceived as a 'technological fix' which acts as a breeding ground for self-organised transport associations.In contrast to the earlier cycles of new offensive technologically-induced Marx-type struggle, the Global South is not a late comer here.In many countries in Africa, Asia and Latin America protests have been mounted over payment, and the legal recognition of platform workers as 'employed'.This is also due to the fast spread of ride-hailing apps and location-based platform work in the Global South, and the transnational learning processes of digital platform workers in informal cross-border networks.
With reference to the Global North and South, several varieties of platform unionism have been identified.These range from collective associations of workers (autonomous organisation) to an effective integration of protesting workers into established unions.This experimental trend in platform unionism has also led to new forms of hybrid organisations such as the ATGWU in Uganda.Varieties of platform unionism also represent different forms of associational and coalitional power, with autonomous organisation and integration into existing unions relying primarily on associational power, and hybrid forms of organisation or cooperation effectively developing both sources of power.Autonomous associations and hybrid organisations, particularly, characterise collective action in the Global South, as illustrated by the case of the APP in Argentina and of the ATGWU in Uganda. 10 Finally, the experimental trend in organising goes beyond established unionism (see Atzeni, 2021).The stance on new technologies of some unions and platform associations, for example, is not a Luddite-like repertoire of contention set on wrecking new machinery or a union strategy of adaption to technological change.Some worker organisations are attempting to effectively control these technologies.In the case of the APP this strategy failed due to the fragility of the organisation; 11 the ATGWU in Uganda, however, was able to set up its own apps, both to organise and to effectively protect their members from exploitative market players such as Uber.Similar strategies of worker control have been followed by other unions in the transportation sector such as the Transport and Allied Workers Union in Kenya (TAWU) and the Auto Rickshaw Drivers' Union (ARDU) in Bangalore, officially Bengaluru, India.This strategy was made possible by the comparatively low costs of developing such a tool and shows the potential of 'platform cooperativism' (Scholz and Schneider, 2017) in transportation.Specific forms of platform unionism, particularly autonomous associations and hybrid forms of organisation seem to be open for such 'worker control' responses of organised labour, as they follow neither the traditional repertoire of contention nor the script of collective bargaining of established unionism.The rise of new southern tech-unions could consequently spur innovative forms of transnational learning, with organised labour in the Global North taking lessons from organising strategies and concepts developed in the Global South.

Notes
1 In her seminal work 'Forces of labor', Beverly Silver has argued that capital uses 'technological fixes' to respond to organised labour by implementing major process innovations 'to fix the problems of profitability and labour control' (Silver, 2003: 66).The term 'fix' is thus used in a double sense: to 'fix' a crisis of profitability and to 'fix' overaccumulated capital with a new physical capital stock. 2 The project 'Trade Unions in Transformation 4.0' (TuT 4.0) was funded by the Friedrich-Ebert-Foundation and represents the second funding period of a larger project entitled 'Trade Unions in Transformation' (TuT).TuT 4.0 includes eight cases of worker organisation in the platform economy across the Global North and South.All authors were members of the project's international steering committee, two of whom, Victoria Basualdo and Edward Webster, were involved in the research on the cases of Argentina and Uganda presented here.One of the authors, Victoria Basualdo, was the academic consultant for the case study on Argentina written by Laura Perelman, Marcelo Mangini, Bárbara Perrot, María Belén Fierro, and Martina Sol Garbaz (Perelman et al, 2020), while Edward Webster was a member only during the project's first funding period but has since continued to work on the case of Uganda's Amalgamated Transport and General Workers' Union (ATGWU) (Webster et al, 2021;Webster and Ludwig, 2023).While many of the arguments presented were developed from TuT 4.0 data, some of the article's conclusions were originally discussed in an earlier FES working paper (Basualdo et al, 2021).The case studies to which we refer in this analysis are : Manga et al, 2020;Perelman et al, 2020;Trappmann et al, 2020 andVandaele, 2020.Results of TuT's first funding period are published in the Global Labour Journal (see Schmalz et al, 2018).
3 Hlatshwayo (2017: 102) distinguishes between a proactive response by unions to technological change in Germany, Scandinavia, and Australia, and a rearguard approach by American, British, Asian, and African unions. 4We owe this insight to Hugo Dias. 5 As we will see later, the platform economy is less likely to follow this logic. 6Far from being better employers than their counterparts in the Global North, most of these companies operate with business models based on cheap labour and weak labour standards.Up to now, they have been able to compete with (or, in the case of Didi and Meituan, even replace) companies such as Uber and Deliveroo, but in the long-term it is likely that some of them will merge and others, disappear.As discussed by Moody (2017), such concentrations have a contradictory impact on workers' power: they usually lead to dismissals and cost-cutting, but also to a standardisation of working conditions and in the long term, to higher rates of unionisation.It remains to be seen what impact the concentration of capital will have on the platform economy in the Global South. 7For more information on a typology of ongoing offensive and defensive struggles see Table 2 in Basualdo et al, 2021: 19, drafted by Mirko Herberg.  The concept of 'social movement unionism' has multiple meanings (Waterman, 2004).
It first emerged in the eighties as description of the way in which unions in developing countries such as Brazil, South Africa, and the Philippines were moving beyond the organised working class to community organisations in a struggle for social justice (Munck, 1987: 132f).We found no examples of this kind of union in our case studies in the platform economy. 9In early 2023, 100,000 shillings was the equivalent of US$27. 10 An important question for further research is whether, in the long run, autonomous worker self-organisation tends to be eroded by bureaucratisation processes, particularly in situations where a hybrid relationship or close collaboration with an established union exists, or where a class struggle-orientation of a workers' collective allows selforganisation to be avoided (Weghmann, 2023).  1The original plan was to develop an app such as the SOT Boda App with the help of external funding.However, when the app was ready to be used, the APP was already in crisis.

Figure 1 :
Figure 1: Technological change and power resources

Table 1 :
Divergent responses of organised labour to new technologies Source: Own elaboration