Workers’ Power in Resisting Precarity: Comparing Transport Workers in Buenos Aires and Dar es Salaam

Introduction

The debate on the future of informal and precarious workers and their organization is increasingly interesting researchers in the field of industrial relations and the sociology of work. The recent publication in this journal of a special issue on precarity represents an important step in the ongoing debate on the concept of precarity and on the forms of organization and resistance of workers in precarious conditions. Most importantly, as argued by the editors and a number of articles in the special issue, there is a need to go beyond the overstretching of ‘precarity’ as a concept, and to look at precarity as a process in order to understand how structural contextual variations impact upon subjective experiences (Alberti et al., 2018; Choi, 2018; Moore and Newsome, 2018; Smith and Ngai, 2018). Attention must be paid to the role of different state and capital regulations in reconfiguring precarity and disempowering workers.

These conceptual advances and their application to empirical investigation are a promising point of departure in shaping future research on precarious workers’ organizational forms and strategies, and in overcoming what we suggest as three main limitations of existing research on precarity. The first shortcoming is an undue pessimism about the possibilities for struggles for rights at work by precarious workers. Standing (2011), for example, in his influential (and controversial) thesis about the emergence of the precariat, dismisses the possibility that trade unions, as institutions shaped by an adversarial and economistic logic tied to specific employers/workplaces, can defend the interest of precarious workers. Access to social protection, rather than workplace struggles, is instead suggested as the progressive way forward for precarious workers. Gallin, while less pessimistic about the future relevance of trade unions, similarly argues that the main agenda should be to secure ‘protection to the unprotected’, rather than ‘formalising the informal’ (Gallin, 2001: 537). The second shortcoming is a tendency to a top-down analysis of the study of precarious workers’ agency. The debate tends to revolve around the trade union as the exclusive organizational and institutional form of workers’ representation, therefore largely ignoring the formation processes of workers’ collective organization that always precede, almost by default in the case of unorganized informal and precarious workers, the existence of the union form (see, for instance, Benassi and Dorigatti, 2015; Heery, 2009; Thornley et al., 2010; Wright, 2013). In these accounts, there is little attention to informal and precarious workers’ ‘on the ground’ independent action. Rather, these workers largely appear as the passive subjects of top-down organizing strategies by trade unions in their efforts to organize and represent precarious and marginalized workers. Recent studies on precarious migrants’ collective organization in London (Alberti and Pero, 2018; Jiang and Korkzynsky, 2016; Pero, 2019), focusing on the importance of community in processes of collective identity formation, have rehabilitated a much-needed bottom-up/self-organizing perspective in industrial relations studies, and certainly represent a welcomed counter-tendency. In this article, we aim to embed this critique into variations of contexts, thus comparing organizing processes in two developing world cities. This in turn leads to the third shortcoming of these narratives, namely the lack of attention to contexts, to the labour process, to variations between and within regions, and to the different structural and political constraints and possibilities that different types of precarious workers might face.

In contrast with these narratives, in this article we reflect about the possibilities and goals of the political organization of precarious workers, and the challenges that this entails, in a way that is more attentive to the way in which global trends play themselves out in individual contexts. The comparison between the two instances of organization of precarious workers in Dar es Salaam and Buenos Aires hereby presented, and the interaction with existing union structures and industrial relations frameworks, allows appreciation of the way in which such processes are part and parcel of global trends, but, at the same time, are politically, institutionally and materially mediated in context-specific ways (Savage, 2002). With others, and following the conceptualization of precarity as a process, we therefore argue that precarious workers’ possibilities are dependent on structural conditions of precarity and exploitation but are, at the same time, spatial and contingent, thus influenced by local and contextual factors (Boampong, 2010; Chun, 2009; Gunawardana, 2014; Jenkins, 2013; Kabeer et al., 2013; Mezzadri, 2016; Ngai, 2005). The decade-long period through which we have been able to observe the development of the cases, has allowed us to focus on the ways in which different contextual political and historical processes have shaped the strategies used by workers to build power and organization.

The article is divided into four main sections. In the first, we consider the relevance of Wright’s distinction between structural and associational workers’ power in identifying the sources of workers’ power (Wright, 2000). In the second, we outline the methodological rationale of the comparison and the methods used to collect data. In the third and fourth sections, in dialogue with Wright’s framework, the article compares transport workers’ ‘structural power’ (both workplace and market power) in the two cities by looking at the relations between labour processes and labour markets. This is followed by the analysis of workers’ ‘associational power’, highlighting the stark differences in the political contexts of workers’ organization. The last section concludes by reflecting on the insights that this comparison can contribute to broader debates on the construction of precarious workers’ collective organization.

Framing workers’ power: Theoretical insights for empirical analysis

There is a long-standing theoretical tradition in the social sciences that emphasizes the importance of both structure and agency in explaining the nature and dynamics of social and class formations (Hobsbawm, 1984; Silver, 2003; Thompson, 1963; Van der Linden, 2008). Within it, particularly useful to operationalize the study of labour possibilities, is Wright’s (2000) widely adopted (Kabeer et al., 2013; Schmalz and Thiel, 2017; Silver, 2003) conceptualization of the sources of workers’ power. As such, it offers a valid departure point in the analysis of processes of workers’ collective formations. According to Wright, workers derive their collective power from two possible sources. First is the ‘structural power’ that (some) workers command. This derives from workers’ specific ‘location . . . within the economic system’ (Wright, 2000: 962). Following this argument some economies, and some industries within them, have more potential to generate labour unrest than others. Two subtypes of ‘structural power’ are to be considered. The first, named ‘marketplace bargaining power’, is the power that workers command due to conditions in the labour market across economic industries. The second, named ‘workplace bargaining power’, relates to the degree of power workers can exert in a specific industrial location for their key position in the production process. However, workers’ ‘structural power’ does not necessarily result in workers’ collective actions. The latter rest on a second source of power, namely ‘associational power’. This derives from the political organization of workers along trade union lines, or other institutional forms, and on the limitations imposed on these forms by the system of legislation and by the historical context of employment relations existing in a certain political/geographical location. Thus, there is no straightforward correlation ‘between workers’ bargaining power and the actual use by workers of that power to struggle for better working and living conditions’ (Silver, 2003: 15). Whether the socio-economic position occupied by workers translates into political consciousness and a shared identity, however, may also depend on active efforts, by workers themselves or by outsider activists/leaders, at constructing a shared notion of injustice and exploitation (Bernstein, 2007; Cohen, 2006; Darlington, 2002; Fantasia, 1988; Kelly, 1998).

In understanding the factors leading to workers’ collective agency, attention to the time-space nexus also matters, as protests and organizations have more chances to be successful at particular moments in history, when institutions and socially established arrangements are generally contested and rules can be partly re-written (Fox Piven and Cloward, 1977). As Chun (2009) more recently argued, these are times in which marginalized groups of workers, not endowed with structural power, can use the ‘symbolic power’ of socially accepted values and concepts of injustice, fairness, equality and social cohesion as moral weapons to exert pressure on the state and institutions. The importance of space dynamics in framing collective organization are now increasingly being considered in sociological studies of work (McGrath-Champ et al., 2010; Manky, 2016). Probably the most important theoretical insight from this tradition is that while capitalist production actively produces and reproduces space, it is also contemporaneously producing a new field of struggles (Harvey, 2006; Lefebvre, 1991). These insights seem particularly useful in the case of this comparison, focusing on the informal work in the transport sector that is so central to the functioning of cities. In it, capitalist dynamics produce precariousness, atomization and individualization of the labour force, thus making workplace organizing difficult. However, dependence of the city on the continuous circulation and flow of people and commodities and the visibility that the interruption of this circulation gives to precarious workers, makes cities’ squares, crossroads and streets potential ‘battlefields’, organizing spaces for marginalized groups to resist precarity.

Methodology

This article is the result of its two authors’ intellectual exchange, which began when they were invited, in 2014, to share their research findings about the organization of precarious transport workers in Buenos Aires and Dar es Salaam at a workshop organized by the International Transport Workers Federation and the Global Labour Institute. The issues analysed and questions posed independently in both studies, and their settings, were remarkably similar: the sector in which these workers operated (transport); the urban setting (two metropolises of capital cities of developing countries: Buenos Aires and Dar es Salaam); the impetus of workers to their political organization and, at the same time, the crucial role that was played subsequently by existing trade unions to support them; and the complex and tense relationship between workers’ grassroots organizations and the established trade unions. At the same time, there were major differences between the contexts in which these workers operated. Above all, the radical approach taken by transport workers in Buenos Aires to confront the state stood in sharp contrast with the less overtly confrontational strategy adopted by transport workers in Dar es Salaam. These different mobilizing approaches and outcome of workers’ actions opened further questions on the labour process, on the labour market and on their politics.

Research on the organization of informal motorbike delivery transport workers in Buenos Aires was carried out in the period 2012–2015 as part of a broader EU project on the organization of precarious workers across sectors (delivery, music events technicians, textiles, public employees) in the city of Buenos Aires. A qualitative approach has been used in this research. In-depth interviews (12 in total) with delivery workers and activists about the labour process, the collective actions of SIMECA (Sindicato de mensajeros y cadets) and the organization of motoqueros (motorbike) workers have provided the first set of data. Other sources have been used to build on the interviews and triangulate information. These sources include: analysis of extracts from interviews with three former activists included in a book on the history of SIMECA by former motoquero workers (Calvo and Gorini, 2013); online YouTube videos of marches and demonstrations; ¹ written reports and notes published in different outlets, such as SIMECA’s flyers collected at the time of interviews, and a left-wing magazine, such as Sudestada; independent press reports; ² and secondary sources (Barattini and Pascual, 2011; Rodriguez, 2015). This combination of qualitative sources has made the detailed reconstruction of the organizing experience of SIMECA between the end of the 1990s and 2009 possible. This reconstruction, in particular through former activists’ oral histories and interviews, took central place in the methodology adopted owing to the fact that, at the time of the fieldwork, SIMECA no longer existed.

Research on the organization of bus public transport workers in Dar es Salaam was part of a broader study of the political economy of public transport in Dar es Salaam (Rizzo, 2017). This article draws on fieldwork carried out in 2009, 2011 and 2014 on a range of sources. A review of media coverage of the issue was central to establish the chronology and key players of workers’ organization. Documentation on the interaction between the Tanzania Transport Union and the informal workers’ association, which Rizzo was kindly allowed to study, provided records on the interaction between the two organizations over time, and of the organizing strategy that was born out of it. This consisted of hundreds of letters between trade union officers and workers’ organizers, typed speeches given by trade union officers and workers’ organizers at meetings with minibus workers, and documents outlining the budgets for organizing events for which the minibus organizers asked for trade union support. Interviews (10 in total) with the leaders of the workers’ association, of the trade union and with transport workers themselves were then carried out to further understand the picture emerging from these sources and to triangulate that with workers’ own experiences of it. ³

Workers’ ‘structural power’: The organization of work and the labour market

Comparing the organizing strategy of these two groups – and different types – of transport workers in Dar es Salaam and in Buenos Aires, requires an understanding of the contexts in which they operated and the sources of power and vulnerability that workers derived from it. This calls for attention to the way in which work was organized in each context, and to the strategies adopted by each group in an effort to challenge the uneven balance of power with employers.

In Dar es Salaam, daladala workers are public transport workers in a city with over 4 million people and a virtually defunct public sector transport company. The cheapest means of public transport is provided by around 10,000 privately owned minibuses, known as daladala. Over 90% of these workers, of which the total number is between 20,000 and 30,000, earn a living by operating buses that they do not own. A clear division between a class of bus owners and a class of transport workers therefore characterizes bus public transport. Bus owners demand a daily rent (hesabu in Swahili) from workers for operating the bus. The daily return for workers will consist of whatever remains after the daily rent to bus owners, and petrol costs, have been deducted from gross income. In other words, the modalities of remuneration by employers transfer business risks onto the workforce. At the beginning of each working day, the profit for bus owners is known; the return for the workforce, if any, is uncertain. These workers are neither waged nor piece-workers. Nor should one think of them as self-employed micro-entrepreneurs, as workers do not own the buses.

Motoqueros workers instead own the means of production – the bikes or motorbike they use in the midst of traffic to deliver parcels and food. As owners of the motorbikes they are responsible for the maintenance and repair of the machines, thus bearing on them, so to speak, the entrepreneurial costs. The work of motoqueros is organized in a way that resembles that of taxi drivers. They work on calls distributed via radio by delivery companies, called agencias, to which they normally offer their services. These agencias can be specialized in the delivery of parcels and documents or provide a delivery service to other companies (especially restaurants and food). While there are agencias which operate in the formal labour market, regularly employing workers and respecting the minimum salary level negotiated for the sector, the majority of these agencias are often very small and not registered. Furthermore, they have been increasingly suffering competition from platform-based delivery companies. These differences have implications for the work performed in terms of both the geographical area and the time of day of the delivery.

Notwithstanding these differences in the employment relationship, motoqueros and daladala workers share the precarity of their work, and the harshness that derives from it. This is rooted in their low marketplace power, which in turn stems from the negative impact that an oversupplied labour market has on workers’ bargaining power. The competition in the labour market for delivery work is normally high and so is the labour turnover. The sector is particularly attractive for young workers as a first-time job, offering flexibility in terms of working hours and a relatively easy entry into the sector given that the capital necessary to buy a second hand motorbike is affordable to many (US$1000, about three times the monthly minimum wage). However, the high level of informality existing in the sector and the piece rate system used as the form of payment do not normally guarantee a dignified salary. This forces workers to increase the rhythms of work (high speed driving, long shifts) and, as a consequence, the probability of a life-threatening road accident (Basta de mensajeros muertos – Stop the killing of motoqueros! say banners in various marches organized by SIMECA). For the dirty, polluted and dangerous nature of their work, workers called themselves ‘the miners of the XXI century’, as argued by a former SIMECA activist (Atzeni, 2012c).

Differently from motoqueros, the vast majority of daladala workers, as we have seen, do not own the buses on which they work. This, in addition to the relatively unskilled nature of work on buses, and to the fact that the labour market for unskilled work is grossly oversupplied in Tanzania, subject daladala workers to fierce competition for work – with pernicious consequences on working conditions and returns from work. Over 80% of its workforce has primary level education. The existence of an oversupply of unskilled job seekers significantly tilts the balance of power between bus owners and bus workers in the former’s favour. As one worker put it:

As too many of us are jobless, if for instance a bus owner is looking for a driver, he will find more than 50 people just at this station. That is why they can ask you whatever they want and you have to accept it. I worked with the same bus for two years. He used to ask me for 50,000 shillings every day. Over time the buses became too many and the chance of making money decreased. I went to my employer and I told him 50,000 was not possible anymore. He could not understand me and he wanted his bus keys back. He gave the bus to somebody else and he is still working with it. I do not know if he manages to give him back 50,000 every day. (Rizzo, 1998b)

Meagre returns, harsh working conditions (the average working day lasts 15 hours and the working week more than 6.5 days) and occupational uncertainty (as work on a given bus lasts less than 8 months on average) are the main traits of exploitation that transport workers in Dar es Salaam share with workers at the lower end of the informal economy. Financially squeezed by bus owners, workers attempt to maximize return from work by overloading the buses, by denying boarding to passengers entitled to social fares and by speeding. As another worker explains, the latter has particularly pernicious consequences:

If I drive without speeding, I will work for the whole day to gain only the money the owner wants back at the end of the day. For these reasons we are forced to speed from 5 a.m. to 9 p.m. Then they say too many accidents – how much energy should we have? (Rizzo, 1998a)

Thus, the trademarks of the infamous work on daladala are remarkably similar to those of work on the delivery motorbikes in Buenos Aires. The cause for the occupational precarity faced by these two groups of workers is also similar: their limited ‘marketplace power’.

While both these groups of workers had low ‘marketplace power’, there were substantial differences in the ‘workplace power’ – as we have seen, the other subtype of ‘structural power’ – that they commanded. Dar es Salaam workers had considerable workplace power. As private buses have long constituted the only means of (barely affordable) motorized public transport available to the public, unrest by its workforce would seriously affect the mobility of the vast majority of Dar es Salaam’s population. In Buenos Aires, motoqueros’ workplace power was limited, as they had no control of the market for the delivery of small goods, thus ruling out strikes – or the threat of them – as a weapon to bring the city’s economy to a halt.

Notwithstanding these differences in ‘workplace power’, in both contexts the harshness of work and the need of mitigating its pernicious consequences, conditions that all workers experienced, provided grounds for the emergence of solidarity and for the establishment of the first associational forms. Crucially, attempts to organize to resist precarity would later draw on these pre-existing associations and networks of solidarity. As a motoquero put it:

When, after a rainy winter week, you finally arrive to a Friday afternoon to drink a mate [typical Argentine infusion] with the other guys that have suffered like you, this produces very strong, very human ties, which later on in the street get transformed into solidarity . . . our job is highly individual, you are alone in the street, the boss threaten you, cars crowd you, police ask for bribe and the only person that can help you is another delivery worker who has experienced the same situations as you did. (Interview with Lulo, quoted in Calvo and Gorini, 2013)

Streets, squares and local bars were the meeting places for the informal and spontaneous workers’ gatherings in small groups. Within these groups, workers shared beers, mate and marijuana. They also supported each other to deal with mechanical problems with their motorbikes, and aired stories and complaints about payments and working conditions, creating a motoquero identity: ‘We used to say that SIMECA could have remained without a building, since it was in every place each of us was in. Each motoquero was the union’ (Atzeni, 2012b). The consolidation of SIMECA as the trade union representing delivery workers drew on these first collective forms of self-help and self-organization. In many cases, the same persons that were part of these sharing and solidarity networks constituted in the cities’ streets were also playing an active role in the structure of the organization, which had at its peak 400 activists spread across the city (Atzeni, 2012a, 2012b).

Similarly, in Dar es Salaam, daladala workers on many routes of the city had already organized informally. On some routes, typically those with limited overlap with other routes, workers took advantage of their de facto monopoly of service. By creating a queuing system to board passengers at the beginning of the route, and by paying a small fee each time a full bus left for its ride, they generated a saving fund. Associational funds were then used at times of members’ need, such as to support the burial of workers’ nuclear family members, to pay for health expenditure and to bribe authorities when members had been arrested by the police for work-related offences. ⁴ The recruitment strategy of the informal workers’ association, named Umoja wa Madereva na Makondakta wa Mabasi ya Abiria Dar es Salaam (UWAMADAR), drew heavily on these experiences of workers’ self-organization, as its recruitment drives relied on transport workers who were part of these informal associations. Leaders were identified at individual stations/routes and educated about the association’s broad mission and more discrete goals. It was then the branch leaders’ task to recruit more members. Such a strategy provided workers with some leadership over the recruitment drive. Evidence suggests this approach raises the chances of success in organizing informal sector workers (Bonner and Spooner, 2011; Gallin, 2001).

Workers’ ‘associational power’ and the political context of workers’ organization

Understanding how workers organize requires not only a comparison between the labour markets in which they operated, but also a comparison between the associational power that these workers commanded, and more broadly between the political landscapes of the two countries. These mediated the way in which workers’ organization consolidated, the realm of possibility of their collective action and workers’ strategy of engagement with the state and employers in their struggle against precarity.

The political climate was starkly different in the two cities, and, in the case of Buenos Aires, it also radically changed over time. SIMECA started to operate as the organization of delivery workers in 1999, initially as a simple de facto association of workers, and later on as a registered (though not recognized) union. However, the Argentinean 2001 crisis contributed to boost the growth and prominence of SIMECA. From 1997, Argentina entered a deep economic recession that created high unemployment; the flexibility of labour contracts and the reduction of pensions and benefits were resisted by various social forces (state and municipal employees, the unemployed movement and territorially based organizations, left political groups and trade unions) with marches, strikes and roadblocks that grew in intensity following the worsening economic conditions of the country (Dinerstein, 2002; Grigera, 2006). Such turmoil offered a fertile environment and source of examples of action for the construction of SIMECA. The years preceding the riots of December 2001, when the crisis exploded, and soon after, was a period of permanent social protest and direct confrontation of social organizations with the state and its repressive apparatus. This helped to develop methods of struggle based on the use of direct action and bottom-up decision-making processes based on the idea of horizontality. SIMECA used actions such as roadblocks, occupations of employers’ premises and of public spaces to make visible to public authorities the conditions of exploitation of their work and to ask for employment formalization. ⁵ SIMECA also link directly to working conditions the death of workers: ‘What we say in relation to precarity is that it literally does kill us’ (Compañeros de SIMeCa, Revista Pampa, November 2009: 8; quoted in Calvo and Gorini, 2013: 7).

The political landscape changed following the stabilization of the economy and sustained economic growth since 2003, with important consequences on the possibilities for workers’ action and their attitude and relationship with employers and the state. As an activist argued:

At the beginning we used to say that we did not need state recognition, we could put 400 motorbikes in front of the Ministry of Labour and set it on fire. We were not interested in being defined as a union or not, we were the motorbikers! In 2001, we were not interested, we had our people on the street, making barricades against the bourgeois legality, we went to the front, no problem, the matter was easy. After this we started to realize that we could not sign a collective agreement, we were gaining conflicts against the employers, but we were nothing. (Atzeni, 2012a)

SIMECA was subsequently registered as a union by the Ministry of Labour. However, registration in itself does not grant to new unions the legal authorization to negotiate in collective bargaining. In Argentina, union recognition is in fact granted by the Ministry of Labour exclusively to one representative organization per economic or productive sector, thus operating under a monopoly of representation (called ‘personeria gremial’). This makes union recognition a very lengthy and disputed process. The political opportunity of the moment, the absence/presence of overlapping claims for the representation of the same groups of workers by already existing unions, and the political alignment of the new union attempting to claim recognition play an important role in it.

The changing economic and political context imposed a change in the strategies and targets of SIMECA, shifting away from violent direct action towards pressure strategies aiming to formally represent workers and more centred on dialogue with employers and the state. This imposed a reconfiguration of the relationship with the state, the need to conform to its rules and institutions, most notably the formal process of union recognition, and the acceptance of the central role of the state in the ‘political’ arbitration of labour conflicts.

However, SIMECA’s attempts at organizing the sector’s workers along the lengthy and politically mediated institutional path proved fruitless. In 2009, the government granted the right of workers’ representation to a newly formed trade union, ASIMM (Asociación Sindical de Motociclistas mensajeros y servicios). ASIMM was affiliated to CGT (Confederación General del Trabajo), the most important trade union confederation, traditionally identified with Peronism, and which was supporting the government at that time. ASIMM official recognition was arguably part of a political deal between the government and the CGT aimed at reducing the power and relevance of independent unions (Atzeni and Ghigliani, 2013). The formal recognition granted to ASIMM implied the illegality of any other existing organization attempting to take action in defence of ASIMM members. This had an influence on the disappearance of SIMECA as an active organization. However, the recognition also brought about improvements in salaries and working conditions for many workers of the sector, achieving some of the demands for which SIMECA had struggled in the previous years. Therefore, while ASIMM, thanks to the political tutelage of the CGT leadership, was formally effective in gaining rights for workers, it did so by ripping off the fruits of the decade-long existence of SIMECA, whose struggles were thus effective in producing collective consciousness, identity and public awareness of the conditions of delivery workers and in forcing the government to take action. In the words of one SIMECA activist: ‘This union [ASIMM] exists just because SIMECA existed’ (interview with Javier, quoted in Calvo and Gorini, 2013). As a result of these developments – of the formalization of the sector – in 2016, part of the delivery sector had been formalized and workers were covered by collective contracts regulating their salaries, working conditions and rights of association.

The political context was very different in Dar es Salaam and in Tanzania. There, workers’ efforts to organize were located in a politically stable landscape – as the ruling party, the Tanganyika African National Union (TANU) first and the Chama cha Mapinduzi (CCM) later, had held power within a one-party political system since 1961 and by winning each multi-party election held in the country since 1995. While not free of tensions, Tanzania’s political trajectory from the late 1990s to the present does not exhibit the patterns of social unrest and political upheaval of the scale and the intensity of the Argentinean crisis. This clearly affected the type of strategy adopted by workers to make demands on employers and the state. At the same time, as we have seen, Dar es Salaam transport workers commanded higher workplace power than their Argentinian counterparts and this presented workers with a different set of opportunities, and more leverage, to put pressure on the state, employers and the public through protest/action. As privately owned buses constitute the only means of (barely affordable) motorized public transport available to the public, a strike would seriously affect the mobility of the vast majority of Dar es Salaam commuters with immediate knock-on effects on virtually every economic activity in the city, and beyond. Furthermore, at a discursive level, workers exploited the public nature of the service provided by transport workers by linking the poor condition of public transport in Dar es Salaam to that of its workers, as part of the strategy was to frame their interests as part of a wider societal ‘common good’.

The institutional channels that the workers’ organization had to follow in Tanzania were similar to those in Argentina, as even in this country the state held a tight control over associational life. So, when, in 1997, a small group of daladala workers set out to investigate the steps required to formalize the association set up by a group of 40 workers, it found that, according to Tanzanian law, trade unions were the only institutions entitled to represent workers vis-a-vis employers or the government. It was this finding that led these UWAMADAR workers to seek a partnership with the Tanzanian Transport Workers Union. The cooperation between UWAMADAR and the transport union was thus forced by the legal framework regulating associationism in Tanzania.

The partnership between the two institutions took considerable investment from both sides. It was fraught with tensions yet productive in advancing the struggle against precarity. From 1997 to April 2000, when the workers’ association was formally registered by the Tanzanian state, several meetings between representatives of the two institutions were devoted to build a shared understanding of the exploitation experienced by daladala workers, and to devise a strategy to demand labour rights from employers and to engage the state in the process. Union support to the workers’ association took the form of legal advice on how to draw up its constitution – through several rounds of revisions – so that the Tanzanian Registrar of Society would approve it, and consisted also in financial support to organize meetings so that UWAMADAR could begin an outreach campaign to recruit members.

Once the workers’ association had been legally registered, in the process becoming an affiliate of Communication and Transport Workers Union of Tanzania (COTWUT), the struggle against workers’ precarity gathered momentum. The two parties’ strategy to bring to an end precarious work had to reflect the fact that daladala workers’ ‘structural power’ had limits. The possibility of a strike was constrained in a context of oversupply of unskilled labourers since workers on strike without contracts could be easily victimized by employers and lose their job. Owing to these political and economic circumstances, one can understand why workers, through their own association and in partnership with the transport union, had to rely on a less confrontational form of pressure on employers to trigger the involvement of the state to mediate between the two parties. Year after year, in the period from 2008 to 2011, workers organized wild-cat strikes and walk-outs (Nipashe, 9 June 2008; The Citizen, 7 December 2009; Tanzania Daima, 29 March 2010; Habari Leo, 6 April 2011). Typically, the press would report an imminent strike by daladala workers, UWAMADAR and COTWUT leaders would distance themselves from such action, and yet on the day of the industrial action public transport would be disrupted by the withdrawal of workers from service for part of the day. Those workers who chose not to adhere to the strike were targeted by stone attacks from colleagues (The Citizen, 10 December 2009). Passengers’ complaints about travel disruptions to public authorities then put pressure on the state to facilitate the negotiations between bus owners and their workers (Rizzo, 2011). Crucially, the form of protest chosen reflected both the strength and weaknesses of the workers: it drew on their ‘structural power’ and yet it did not overexpose workers. Such actions were strong enough to force Dar es Salaam authorities to intervene in the dispute without making workers vulnerable to retaliation by employers.

The partnership between the workers’ association and the transport union rested on a clear division of labour whereby the union supported the cause of daladala workers ‘from above’. This entailed drawing on its technical expertise in labour law and on its political connections. The main role of the workers’ association was to recruit members to give credibility to unionists lobbying from above, and to lead them when direct action was deployed to trigger the need for negotiation by employers or the state. ‘Talking to drivers and conductors, one by one, “You have been doing this job for many years. Tomorrow, the day after tomorrow, how will it look?”’ (Rizzo, 2009). So there was an element of sensitizing workers to the importance of employment contracts, and of trying to break the short-term time horizon of daladala workers’ attitude to work that was both an effect and a cause of workers’ occupational precariousness. The albeit small financial support from the union to hold events at which UWAMADAR could advertise its agenda is worth noting here, as it suggests that the union was prepared to invest some of its funds to promote the organization of informal workers. This helped, in a small but significant way, to partly address UWAMADAR’s lack of funds and the lack of visibility that came with it. ⁶

UWAMADAR’s outreach drive was extremely successful: in 2003, the organization had 5236 members, or about 44% of the total (estimated) workforce of daladala (UWAMADAR, 2003: 23). ⁷ Such numbers conferred legitimacy to UWAMADAR, and allowed transport unionists to start to lobby for employment contracts for daladala workers. Over the years, the achievements of this coalition were substantive, as the public transport regulation changed from a starting point in which the existence of a public transport informal workforce was not formally recognized by the state, to one in which each bus owner had to register the contracts of its workers in order to obtain a public transport licence. Despite this, workers’ leaders were aware that the issuing of contracts would not be straightforward, as attempts to non-comply with regulation by employers were likely. Still, as the COTWUT Deputy General Secretary put it, bus owners’ room for manoeuvre in avoiding labour regulations was progressively shrinking: ‘the day that an owner gets into an argument with his driver, and is asked to produce the contract, he will be in trouble’ (Rizzo, 2011).

Difficulties in holding employers to account, however, resulted in renewed and unsolvable tensions between the workers’ association and the trade union, ultimately causing the end of the partnership between the two. Leaders of the informal workers’ association grew disillusioned about the necessity of their partnership with the trade union, as they doubted its effectiveness in the continued struggle for labour rights. As the UWAMADAR General Secretary recalled: ‘the service that we were getting was small, and our needs to be looked after where not satisfied’ (Rizzo, 2014). UWAMADAR leaders, together with those of the association of upcountry bus workers (UWAMATA), exited COTWUT and established a new trade union: the Tanzania Road Transport Workers Union (TARWOTU). This was officially registered in January 2013. ⁸ COTWUT’s General Secretary had little sympathy for the argument that his union neglected the interests of daladala workers. Instead, he suggested that the real motive behind UWAMADAR leaders’ decision to start a new union was: ‘the ambition to lead. [It] sometimes drives change, the desire to be the General Secretary of a national union. Otherwise, why not use a network that is already in place?’ (Rizzo, 2011).

Conclusions

Bringing the article to a close, we reflect on the lessons that can be learned from this comparison. The lack of attention to contexts and their specificities, an excessive pessimism about the possibilities for struggles for rights at work in informalized labour markets and an excessive focus on trade unions as the only vehicle for workers’ organization, were the three shortcomings in the literature on informal and precarious labour which we highlighted at the outset. The first concluding remark is that our comparison exposes the significance of these shortcomings, as in both cases we have seen how groups of informal and unorganized workers have been able to improve their income and working conditions by collective organization and struggles for employment rights vis-a-vis their employers and the state. This goes against claims that the only way forward for precarious workers is public policies and legislation for social protection. The cases also demonstrate how the construction of collective organization, rather than a top-down process initiated by trade unions, has been instead the result of gradual processes of workers’ power formation initiated by workers, and for which trade union support was an important afterword, albeit with tensions. Importantly, in both cases, due to different and context-specific political landscapes, existing legislation on workers’ representation forced self-organized groups to adopt the trade union form and follow trade union paths of organization.

The second lesson that can be learned from our comparison concerns the relationship between workers’ self-organization and trade unions, both its importance for the effective representation of workers’ interests and the tensions that tend to characterize such a relationship. In both cases, on the one hand, workers’ impetus to the process of organization was crucial to its vibrancy and early successes; on the other hand, there were significant limitations to the gains which workers’ own organization could achieve without the support of trade unions. At the same time, trade union support – legal, financial and in terms of know-how to navigate state authorities – while important in advancing the cause of workers in both contexts, took away impetus and/or radicalism from workers’ earlier organization. The tensions that characterized the relationship between established trade unions and workers’ organization, leading SIMECA to cease to exist in Buenos Aires, and UWAMADAR to break the partnership with the transport union in Dar es Salaam, were thus a reflection of the complex and ambivalent forces linking trade unions to informal workers’ own organizations. They are two instances of the recurrent tension between institutionalization and mobilization in the construction of workers’ power.

The third and final point to learn from our comparison concerns the value of the late Wright’s framework on the sources of workers’ power. This framework helped us to understand forms of precarity and struggles to overcome it in the informal economies of cities in developing countries. Wright’s conceptualization of workers’ ‘structural’ and ‘associational’ power has been a fertile starting point and guiding framework of this article’s attempt to locate a fine-grained understanding of conditions and possibilities in time and in two contexts. The two groups of workers analysed have similarly weak ‘marketplace power’ – due to unskilled labour oversupply, and different ‘workplace’ power – due to the different type of transport work performed. These workers also experienced similar trajectories in terms of ‘associational’ power, with a transition from workers’ self-organization to trade unionism. This was both beneficial to the advance of workers’ interests and yet not void of setbacks and, to some extent, disempowered workers’ organizations. What can be learned from this comparison then? Perhaps the most important lesson here is that it would be foolish to expect a framework on workers’ power to predict the outcomes of workers’ struggles, as they are necessarily open-ended. Instead, the analytical and political value of this framework is that it can help to understand and compare the messy labour markets inhabited by precarious workers and the possibilities and pitfalls of organizing for workers’ rights in them.