Decent Work for All: Rethinking Decent Work in the Context of South Africa

The authors argue that there is a need to rethink what a commitment to decent work would mean in the context of South Africa, a country with a large number of long term unemployed. Drawing on their experience of researching work in South Africa, they highlight the relevance of the agency of workers for the progressive realisation of a decent work agenda. The lukewarm response of the government to the diagnostic tool designed to measure decent work led the authors to reframe decent work in a positive way as an organising tool to recruit vulnerable workers into a broader, more inclusive movement of working people.


Introduction: The Debate on Decent Work in South Africa
The winning slogan of "decent work for all", used by the ruling African National Congress (ANC) in its campaign in the 2009 general election, marked a peak in the use of the concept of decent work in South Africa. While the demand for better working conditions is deeply rooted in the struggle for democracy in South Africa, the use of the concept of decent work in the policy discourse of the government is relatively new. It was first used at the second Growth and Employment Summit in June 2003 where, to address the challenge of employment, the government, employers and trade unions made a commitment to "work for more jobs, better jobs and decent work for all" (Mbeki 2003: 3).
Although a number of agreements were reached to improve the quantity and quality of employment at the Growth and Employment summit, post-summit progress was slow. Of 1 Edward Webster is a distinguished research professor at the Southern Centre for Inequality Studies and research associate and founder of the Society, Work and Politics Institute at Witwatersrand University Johannesburg (Edward.Web-ster@wits.ac.za).
2 Carmen Ludwig is a postdoctoral researcher at the Institute of Political Science at the Justus-Liebig-University Giessen and a former research associate at the Society, Work and Politics Institute at Witwatersrand University Johannesburg (carmen.ludwig@sowi.uni-giessen.de). particular concern for trade unions was the de facto highly flexible labour market that had emerged through the rapid growth of precarious work and the informal economy. "Liberalisation", Webster and Von Holdt (2005) argue, has polarised the labour market by increasing the resources of some 6.6 million people in the core (decent work) while at the same time reducing the resources of the 3.1. million in the intermediate category of the non-core (or atypical work) and those on the periphery, consisting of the 2.2. million workers in informal work and the 8.4 million unemployed (Webster/Von Holdt 2005: 27).
As Barchiesi (2011) argues, wage labour did not fulfil its promise of social emancipation. Instead, wage labour has declined in the face of retrenchment and has also become more precarious and insecure due to growing casualization and outsourcing of work. This has led to an increase of the working poor, workers that remain poor even though they are in formal employment (Barchiesi 2011). It is in this context of growing precarity and unemployment that President Thabo Mbeki came under attack inside the ruling party, the ANC, culminating in his ousting as president in favour of Jacob Zuma in December 2007. With strong support from the Congress of South African Trade Unions (COSATU) 3 , expectations were high for a significant shift in policy towards more and better jobs. In his inaugural State of the Nation address in June 2009, President Jacob Zuma stated: The creation of decent work will be at the centre of our economic policies and will influence our investment attraction and job creation initiatives. In line with our undertakings, we have to forge ahead to promote a more inclusive economy (quoted in Economic Development Department 2010: 10).
In 2010, the adoption of South Africa's own decent work country programme (after a series of consultations between government, employers, labour and the ILO) seemed to be further evidence of the government's commitment to the goal of decent work for all. In the same year the Member of the Executive Council (MEC) for the Gauteng Department of Economic Development (GDED) commissioned the Society, Work and Politics Institute (SWOP) at the University of Witwatersrand to develop a policy framework for the realization of decent work in Gauteng, South Africa's richest province. The Terms of Reference of the GDED decent work study was to create a diagnostic tool that could be used to measure progress towards decent work at the level of the individual worker through a questionnaire, and then aggregated to the industry or sub-industry level (Webster/Budlender/Orkin 2015). Over a period of three years, we interviewed over 3,000 employees from three selected sectors, namely agriculture, the hospitality industry and private security.
Our findings identified significant similarities and differences between scores on the decent work indicators (Webster/Budlender/Orkin 2015). The three industries we surveyed all experience decent work deficits, such as long working hours and poor wages, associated with limited benefits and a lack of an institutional voice. Another remarkable similarity across all three industries is that the following four domains are experienced as salient and interrelated, namely earnings, job security, benefits and social dialogue or voice. This suggests, we concluded, an evidence-based 'core' of decent work (Webster/Budlender/Orkin 2015: 139-141).
On the basis of this diagnostic tool we produced three sector-based reports on how decent work could be progressively realised, and a comprehensive report was submitted to the GDED in September 2012. 4 The policy framework we developed disaggregated employment creation into three phases: a short-term phase focusing on public employment schemes such as the Public Works Programme and the Community Work Programme (CWP); a medium-term phase focusing on the agricultural value chain, light manufacturing and green economy; and a longterm phase, resting on human resource development for a knowledge-intensive economy.
The innovation in the research was the proposal of a strategy for the progressive realisation of decent work that replaces the current polarisation and false binary between the advocates of immediate decent work and those of a massive expansion of low-quality insecure work in order to be competitive in the age of globalisation. Instead, we suggested in the GDED study that if decent work is to be progressively realised it has to be integrated into an alternative development path (see also Selwyn 2017). 5 We did not receive an official response to the report submitted over eight years ago. There are many possible reasons for that: a lack of political will of government and employers, a lack of strategic thinking in the labour movement, and the constant change of MECs (five over the period of the study) who each had new priorities. But there was a deeper, more fundamental reason for the failure of the government to implement the proposed policy framework: the contested nature of the concept of decent work in a country with high unemployment. This is the dilemma of decent work and job creation in a context where 29.1 per cent of the workforce are unemployed (Stats South Africa 2020) 6 : do you abandon labour standards to become globally competitive or do you try to balance decent work with enterprise efficiency? (Webster 2011) Alistair Sparks, a veteran journalist, went as far as to argue that it was the government commitment to COSATU's insistence on 'decent work' which compels employers to pay industry-wide minimum wages in each sector and apply a range of specified working conditions to all workers regardless of whether they are skilled or not that has driven South Africa into a economic cul de sac (Sparks 2010: 1).
Nattrass and Seekings have taken the argument a step further, labelling decent work advocates as decent work fundamentalists. Decent work fundamentalism, they argue, is "the use of rising minimum wages to boost productivity by destroying labour intensive jobs" (Nattrass/ Seekings 2019: 162). They go on, drawing on Arthur Lewis's dualist model of economic development, to argue that 4 The diagnostic tool was also used in a study of the car components sector in the Gauteng province, where the decent work indicators and questionnaire were used to operationalize the concept of social upgrading in a value chain analysis of the sector (Mashilo 2019). 5 Selwyn (2017) wrote on what he calls a "labour centred development path" and what it could look like. 6 The unemployment rate increases to 38,7 % if one includes discouraged work seekers who did not take active steps to find work during the last four weeks (Stats South Africa 2020: 7).
in the presence of surplus labour, inclusive development requires job creation. Labour-intensive sectors such as clothing manufacturing can provide much needed jobs and have been of crucial importance in the growth paths of many successful developing economies (Nattrass/Seekings 2019: 162).
They demonstrate their argument by examining the case of the clothing manufacturing industry in South Africa, "where decent work fundamentalism not only impeded job creation but resulted in job destruction in places where they were desperately needed" (Nattrass/ Seekings 2019: 162).
On the other end of the political spectrum are those who see decent work either as a reformist strategy designed to restore capitalism (both materially and ideologically), or dismiss it as unrealistic. This latter position has been put most clearly by sociologist Lucien van der Walt, when he argues that the overall aims of the Decent work agenda -sustainable decent jobs, wages and welfare, entrenched and meaningful roles for unions, and social dialogue that results in real gains -are unrealistic in the era of neoliberal capitalist globalisation. This is a period of relatively low and unstable economic growth, sustained class war from above, and a global race to the bottom. The conditions that enabled substantial economic and social reforms like those in the 'golden age' of post-war capitalism no longer exist (Van der Walt 2019: 4).
But Van der Walt's critique is not a rejection of the concept of decent work: it is instead an argument for building organisations from below as the most effective strategy to transform the lives of working people. The most important way to win even basic reforms is serious organising and actionespecially by unions […]Not only are reforms won from below generally more farreaching but the struggle and victory are also valuable forms of empowerment and self-activity (Van der Walt 2019: 7).
The disappointing response of the government to the diagnostic tool we developed to measure decent work led us to rethink what a commitment to decent work would mean in the context of a country such as South Africa; with large numbers of long-term unemployed, where a bad job could be considered better than no job at all. The implications of long-term unemployment on workers, especially the youth, has been the subject of a number of excellent ethnographic studies (Sefalafala 2018;Dawson 2014;Dawson and Fouksman 2019). Research seems to suggest that the conventional view that a bad job is better than no job is not true, especially for the younger generation, who will turn down low-quality jobs. Zizzama (2020) finds that younger workers with no dependants and with alternative sources of support and stronger outside options, are especially likely to turn down or quit low-quality jobs. Older workers, with dependants and without alternative sources of support, are more likely to accept and persist in lowquality jobs see also Jeske 2018). In this paper we examine two case studies that reframe decent work in a positive way as an organising tool in the labour movement: first we examine the use of the concept of decent work in a NUMSA project on the working conditions of petrol attendants. We discuss how the decent work questionnaire developed in the GDED study was adapted to a research project designed to recruit petrol attendants into NUMSA. In Part two we examine the COSATU Vulnerable Workers Task Team (VWTT) and their largely unsuccessful attempt to broaden their constituency to include the unorganised and precarious sections of the working class, such as domestic workers, farm workers and street traders. We show how the 'core' indicators of the GDED survey -earnings, job security, benefits and social dialogue or voice -became the basis of a campaign to organise vulnerable workers in selected affiliates of COSATU. In the course of deliberations in the VWTT it became clear that the established trade unions were not representing the growing numbers of precarious workers and that Community Advice Offices (CAOs) were beginning to fill this representational gap. A study was commissioned on CAOs, which is the focus of Part two of the article.
We show how local unions appropriated the global norm of decent work and 'grounded' it locally through their organising practices. In the case of petrol attendants, the concept of decent work enabled organisers to engage with workers directly on decent work deficits. Indeed, nearly half of the workers interviewed in the study saw the need for strengthening their 'voice' and joining a union.
In the VWTT case the decent work narrative helped to build a bridge between formal unions and informal associations and enabled them to find a common language and to translate informal workers' needs into a campaign for better working conditions. Most importantly, through engagement in the activities of the VWTT, key social partners from South Africa played an active role in framing the ILO recommendation No. 204, which is designed to facilitate the transition of informal workers into the formal economy.

Methodological Reflections on Engaged Research
Both initiatives, the NUMSA project and the VWTT, took place in the same time frame between 2013 and 2014. The authors were actively involved in the unions' projects, although the role of the researchers and the form and level of engagement varied across the two studies: in the VWTT project the researcher was present as an adviser and research consultant. He designed and supervised the research project on the CAOs. In the NUMSA study, the role of researcher and a union organiser were more strongly intertwined. The purpose of the study itself was to simultaneously collect information on the sector and to recruit workers into the union. In the case of the NUMSA project it exposed the researcher to the manifold challenges organisers face when entering the workplace, which significantly deepened the researchers' understanding of the specific labour relations at petrol stations. The fact that the research aimed at providing support to workers and to identify areas for union intervention might have increased workers' trust and created a greater openness for workers to speak about problems at the workplace experiences with and expectations towards union representation.
The researchers' role in both the VWTT and the NUMSA project was based on a commitment to what could be called critical engagement. Critical engagement addresses the difficult challenge of producing knowledge in a collaborative way while advancing research that might be meaningful for both, social sciences and the constituencies with whom the research is undertaken (Lozano 2018: 107). Lozano makes a powerful argument for a more collaborative understanding of knowledge production. Instead of 'working on social movements', Lozano suggests a collaborative approach which involves working and thinking together with social movements' activists, advancing research that might be meaningful for both social science and the 'research subjects' (Lozano 2018: 103).
Such an approach, Lozano argues, requires that researchers "negotiate and/or determine what knowledge should be produced, how and for what purposes, asserting a significant degree of control during the research project to ensure that it is a non-extractive process, that benefits communities, that meets their ethical standards and protocols, and that does not reproduce epistemic violence" (Lozano 2018: 107).
In defending taking sides, the distinguished American sociologist of the 1960s Alvin Gouldner (1968) emphasized an engaged sociology's ability to discover new information often hidden from mainstream sociology. "A feeling commitment to the underdog's plight", wrote Gouldner, "enables us to do a better job as sociologists". It also, Gouldner continues, "made the suffering of the underdog 'naked and visible' to the public" (Gouldner 1968: 105).
But there are many pitfalls in the practice of critical engagement. It can lead to a lack of analytical distance from the research subjects. It can also lead to a dispersal, where the sociologist leaps from one task to another (Rodríguez-Garavito 2014). But the greatest pitfall is the threat to the autonomy of the researcher.
In the face of attempted censorship of SWOP AIDS research in the late 1980s, Webster wrote: Pressure exists on scholars to make a clear declaration that their research and teaching should be constructed as support for, and on behalf of, particular organizations.
To prevent this subordination of intellectual work to the immediate interests of these organizations, I prefer the stance of critical engagement. Squaring the circle is never easy, as it involves a difficult combination of commitment to the goals of these movements while being faithful to evidence, data and your own judgment and conscience (Webster 1995: 18).
In neither of the studies covered in this article was the autonomy of the researchers threatened. In engaging critically with the different formations of the labour movement we were being faithful to the classical tradition of sociology. As Claus Offe wrote, [t]o distance oneself from social practices, is to forget that all great social and political theorizing in the nineteenth and twentieth centuries received its problematic and inspiration from social movements and contested conditions. In the past, this existential rootedness of social and political theory has always been a source of analytical strength (Offe 1979: 15).

NUMSA's Decent Work Study of Petrol Attendants
The NUMSA pilot project combined research on decent work deficits at petrol stations with organising (Ludwig 2014). 7 The study was conducted between October and December 2013 and included 140 interviews with workers from petrol stations in and around Tshwane 8 , of which 102 were based on face-to-face interaction and lasted between 20 and 30 minutes each. The employers' association and the central bargaining council MIBCO 9 were informed about the research and NUMSA requested access to the workplace "for the purpose of membership recruitment". Most interviews were directly conducted at the workplace, i.e., the petrol stations. While employers had to grant union organisers access to the workplace in line with the MIBCO bargaining agreement (although this was contested in single cases), conducting the study during working hours finally depended on negotiations between the researcher and local management.
The majority of survey participants were petrol attendants (72 per cent). To a lesser extent, cashiers (9 per cent) and other occupations (e.g., supervisors, bakers, cleaners) were also included. The typical respondent in the study can be classified as a South African male petrol attendant below 40.
The aim of the study was twofold: To gain a better understanding of working conditions at petrol stations and to evaluate whether an approach that engages workers on decent work deficits could serve as a useful organising tool. Petrol attendants were selected as the subject of the study for three reasons: First, they constitute a relevant employment group in South Africa, where about 73,000 workers are employed in filling stations, the majority being petrol attendants (Godfrey 2016: 16). Second, with about 30 per cent the unionisation rate in the fuel retail sector is comparatively low (Godfrey 2016: 27). One reason is the spread of workers across small or very small workplaces (the average number of workers employed at petrol stations is 18, Godfrey 2016: 25), which makes it more difficult to organise and provide support to workers. NUMSA has emphasised the strategic relevance of organising in the motor industry in general and at petrol stations in particular at its Central Bargaining Conference in 2013. The bargaining agreement in the motor industry covers a diverse range of sectors, with petrol attendants being located at the lower end of the spectrum in terms of wages and employment conditions. Third, not much research had so far been conducted on petrol attendants' working experiences (see for example Hadland 2002;Du Toit 2012).
The questionnaire in the study was to a significant extent informed by the decent work questionnaire used in the GDED study (Webster et al. 2008). It particularly covered the core indicators of decent work: earnings, job security, benefits and voice. As the compilation of the questionnaire was a collective process in the union, discussed at regional and national motor sector meetings, the questionnaire also partly went beyond the scope of the decent work questionnaire. 10 Particularly in two areas organisers felt the need to gain a deeper understanding of the specific situation of petrol attendants: their socio-economic background and their views on unionisation. The study was conducted in the wake of a four-week long strike during wage negotiations in the motor industry in September/October 2013 and therefore also included questions on strike participation. In the GDED study, the ILO's decent work agenda was adapted to allow for a comparison between industries. The application of the concept in the decent work study of petrol attendants differed in that it was used to profile the sector but also individual workplaces. The comparison of workplaces on decent work deficits provided organisers with detailed information on the situation at each of the 21 petrol stations that were included in the study, enabling them to take action and to monitor progress.
With regard to employment security the survey revealed a mixed picture: On the one hand, a large majority of workers at petrol stations, 79 per cent, were permanent workers, in contrast to 13 per cent who were employed on a temporary basis. There also was a higher staff continuity than we initially expected: 66 per cent of participants had been working at the filling station for more than 2 years. The vast majority (84 per cent) had either not worked at another petrol station before or changed the workplace only once. On the other hand, 61 per cent of respondents did not have a written contract of employment. At least at three filling stations employees were either not registered with MIBCO or registered with delay. At three other filling stations employers made use of labour brokers. A majority of workers (54 per cent) perceived job security as rather low, particularly as they felt that that it would be easy for them to lose their jobs. The vast majority of workers (73 per cent) had been affected by unemployment before; about half of the group (35 per cent) for more than 2 years. Workers who said that they are easily replaceable usually referred to the low formal skills required for the job, the possibilities of employers to hire and fire, and the high level of structural unemployment in the country. Overall, petrol attendants are dependent on the job, which also provides them with little opportunity for skill enhancement and further development.
Earnings were the major concern for petrol attendants, as expressed in this expression by a worker: "Working many hours with little money paid." Half of the workers at filling stations took home R861,75 (around 42€) per week or less. This amount corresponded with the weekly minimum wage in the sector at the time. Data from MIBCO also indicate that 75 per cent of wages in the fuel retail sector either match or are below the minimum wage. Only 25 per cent of workers were paid an actual wage above the minimum (NUMSA 2013).
According to our survey, tips are an unpredictable component and did not significantly supplement workers' wages (the median was R50 per week). However, there was a great variation in tips as the amount primarily depended on the location of the petrol station and the composition of its customers. Overall, higher earnings above the minimum wage usually corresponded with working hours above 45 hours per week as prescribed by the bargaining agreement. On average, workers at filling stations worked about 51 hours per week.
At all 21 filling stations, the employer deducted money from workers' salary if customers claimed that workers filled in too much petrol or if there was a problem with the till. In this way, risks are shifted from employers to employees, and this unfair practice is a cause for much dissatisfaction at petrol stations.
Workers clearly struggled to make ends meet with their income, on which on average four to five family members depend on. 11 59 per cent of workers said that they do not manage to save regularly, and for a majority of workers a real danger of indebtedness existed as they had to repay instalments for expenses of daily life.
Employers in the fuel retail sector do not provide workers with benefits like medical aid or housing allowances. In only very limited cases workers received transport allowances. With regard to those benefits employers have to provide in order to comply with labour laws and the collective agreement, namely provident fund, paid leave and an end of the year bonus. there is a serious lack of compliance. 21 per cent of workers report that they do not receive an end of the year bonus, 22 per cent have no provident fund and 29 per cent do not get sick leave.
A second major concern for workers at petrol stations are deficits concerning a safe work environment. A majority, 56 per cent of workers, said that they do not feel safe at work. This included a lack of security at filling stations, particularly in the night and the risk of becoming a victim of crime and robbery. As Du Toit (2016) has also shown in his study, workplace violence, including verbal abuse and assaults but also armed robberies, is a serious concern at petrol stations. The lack of security furthermore includes health risks posed by petrol. 12 More than 50 per cent of workers said that they did not receive health and safety training by the employer, and in only six filling stations there was a worker elected safety representative who could monitor health and safety conditions.
In the study, petrol stations had a particularly low or no union representation. Participants showed a clear interest in making their voice heard when approached by a union. 85 per cent of non-union members expressed an interest to learn more about NUMSA. Mainly, workers at petrol stations expected the union to increase their wage (62 per cent), defend them in disciplinary cases (39 per cent) and to educate them how to fight for their rights (28 per cent). This indicates that workers do not only see the need for protection but also wish to increase their agency at the workplace. 50 per cent of workers in the sample, including non-unionised workers, said that they participated in the recent strike. However, satisfaction with the outcome of negotiations was low as 47 per cent of workers perceived the increase as too low to have a significant impact on their wages.13 There is a specific vulnerability that is linked to the nature of the petrol attendants' job that poses a dilemma to trade unions: If employment standards are significantly improved in the sector, the threat of displacement through automation (which is already the norm in the Global North) poses a huge challenge.
The strong hierarchy at petrol stations and the dependency of workers on maintaining their job provides management with a significant degree of power. A clear deficit with regard to the freedom of association was the level of harassment we observed at petrol stations: At several workplaces, management intimidated workers not to speak to organisers or applied direct pressure when workers decided to join the union. This corresponds with the union's general experience, who has begun to recruit at petrol stations during the night as a result. 14 The survey indicated significant decent work deficits in all four categories: earnings, job security, benefits and voice -due to either low standards in the sector or a lack of compliance by employers. This also demonstrates that a job at the core of the South African labour market -characterised by a permanent, full time job, regulated by a bargaining agreement -does not necessarily equate with being a decent job. As a result, it remains very difficult for petrol attendants to significantly improve their living conditions. Four out of ten respondents said that they either stay with family and friends or live in a shack.
During the course of the project, 51 workers joined NUMSA, which is half of all workers with whom direct face-to-face interaction took place. Our project indicates that it is worth exploring how research and organising can be more strongly combined. It furthermore encourages thinking about ways in which the concept of decent work could serve a useful organising tool and strengthen workers agency. 13 The wage increase was 11,6 % in 2013, 9 % in 2014. 14 Vivani Shezi, SWOP Research Breakfast, Witwatersrand University, 15 March 2019 First, by assisting trade unions in a systematic assessment of workplaces, a decent work questionnaire can form part of a horizontal mapping exercise where trade unions seek to document and identify the characteristics of workers, their location and the industry (Webster/Bischoff 2011). Second, for organisers a decent work questionnaire can be a useful tool to engage with workers in a conversation about their working conditions and shows the latter that the union is interested in their views. Speaking about the working conditions can also act as a door opener: once workers start talking about the difficulties they face, they might also see the need to organise. Furthermore, the engagement on the company-level helps to link the abstract concept of decent work to practical examples and the daily issues of workers. It can initiate a conversation among workers at the shop floor about their experiences, which does not necessarily happen on its own. Creating the space and awareness can therefore facilitate a process where workers start to think about ways to initiate change and act collectively.

COSATU's Vulnerable Workers Task Team and Researching Community Advice Offices (CWOs)
The fact that COSATU's membership had not diversified significantly to include outsourced, casual, migrant or informal workers had long been of concern to COSATU, especially as it began to feel the pressure of large-scale retrenchments and informalisation. The labour relations system based on the traditional employer/employee system was being eroded by labour broker intermediaries and an increasing variety of casual work. COSATU's 1997 September Commission "assessed the changing nature of the labour market […] and recommended that it commit itself to the strategic objective of organising vulnerable sectors" (COSATU 1997: n.p.). This was endorsed in successive congress resolutions, although little action transpired.
COSATU's 2012 congress mandated it to explore the organisation of non-standard workers, and the Vulnerable Workers Task Team (VWTT) was subsequently launched. 15 Between 2013 and 2014, it brought together representatives from its affiliates as well as two informal worker associations (the South African Informal Traders Alliance, SAITA, and the South African Domestic Service and Allied Workers Union, SADSAWU), two international informal workers' federations (Women in Employment: Globalizing and Organizing, WIEGO, and StreetNet International), and a number of research organisations. 16 The aim was to encourage and assist the recruitment of precarious workers through a campaign around five decent work indicatorsregular wages, job security, benefits, health and safety, and organisational rights.
Through discussion in the VWTT, three sectors were chosen; farm workers, domestic workers in casual employment and street traders. 17 Three of the core decent work indicators identified in the GDED study provided the basis of the campaign running from July to October 2014: the right to make or earn a decent living, work security, and comprehensive social protection.
During discussions in the VWTT two further demands were added: safe and healthy workplaces, and full organisational rights for all workers. 18 The aim of the campaign, according to the VWTT, was to first and foremost build organisation […]. We must find a balance between collecting demands from below, and providing leadership and direction as a Task Team. A set of demands can draw workers together to stand 'shoulder to shoulder' in preparation for moving forward to deepen organisation (COSATU Vulnerable Workers Task Team, Minutes September 2013to October 2014. For the VWTT, the campaign was about recruitment as well as listening and learning. The demands "must unite, and be linked to an agreed strategy to make the links between workers who are fragmented through sub-contracting, labour broking, part time work, etc" (COSATU Vulnerable Workers Task Team, Minutes September 2013 to October 2014). The campaign, it was agreed, must also engage public perceptions of waste pickers as scavengers. The VWTT concluded by emphasizing that "[w]orker knowledge of rights and state enforcement is critical to achieving rights -if they are there on paper but not implemented they are worth nothing" (CO-SATU Vulnerable Workers Task Team, Minutes September 2013 to October 2014).
In the process of formulating the demands it became clear that many of the demands that have emerged from a context where the standard employment relationship is dominant have to be revisited for a context where non-standard employment relationships are becoming widespread. The wording of the first demand -"for a living wage" -was challenged as some precarious workers did not earn wages. Street traders, for example, are mostly self-employed and do not earn a wage but earn income from selling a product on the market. The demand was changed to "decent work and livelihoods for all" and extended to include "[r]espect for own account workers (the self-employed), including rights and access to materials, markets and space" (COSATU Vulnerable Workers Task Team, Minutes September 2013 to October 2014).
The second demand was expanded to include "direct access to work opportunities for both employed workers and own account workers -away with middle men!" -a reference to opposition to labour brokers. Regarding social protection, a demand was made for a Basic Income Grant and the Unemployment Insurance Fund to be "expanded to include all those without employment" (COSATU Vulnerable Workers Task Team, Minutes September 2013 to October 2014). It was proposed that the Unemployment Insurance Fund (UIF) be brought under the control of the unions rather than the Department of Labour.
The demand for union rights was amended to [f]ull representational and organisational rights for all workers: the right of all workers to meet and gather in every work place regardless of who the employer is, including shopping malls, farms, homes, airports, stations and other multiple employer work spaces and in the case of own account workers, public spaces; the statutory right to representation and collective bargaining for both employed workers and own account workers; the right of access of organisational representatives to all work places to recruit and represent workers (COSATU Vulnerable Workers Task Team, Minutes September 2013to October 2014. The campaign was launched in the second half of 2014 and was based on the pamphlet produced by the VWTT, Join the Drive to Organise all Vulnerable Workers! The pamphlet was distributed to selected affiliates of COSATU and built into each affiliate's recruitment drive. The campaign had limited impact. In the case of domestic workers, few were recruited and no advice office for these workers has been established. In the case of farm workers, none of the ambitious plans materialised and the farm workers union has largely failed to organise farmworkers. In the case of street traders, no organisation was undertaken by the COSATU affiliate, although a successful organisation of street traders has taken place outside of COSATU. 19 Indeed, outside of the traditional unions, innovative forms of organisation were emerging amongst precarious workers. In response to these initiatives the VWTT commissioned the Chris Hani Institute (CHI), one of the research organisations in the VWTT, to conduct a survey of the rapid growth of community and worker advice offices (CAOs) in response to the lack of representation of precarious workers .

The multiplication of worker centres
CHI's innovative study of the over five hundred CAOs that had emerged over the past two and half decades in South Africa illustrates how working people are experimenting with a variety of responses to the dislocating impact of an increasingly informalised labour market. 20 Global shifts in the organisation of work may, however, present opportunities for new models of organisation and alternative sources of worker power. In this context, the report argued, it is important for trade unions and their allies to begin to explore and understand the emerging alternative organising structures and approaches to empowering workers.
The report provides an overview of the status of CAO's in South Africa more broadly and within that context, of the presence of worker-oriented advice offices and the role they are currently playing. The report seeks to understand their approaches, strengths, and challenges with the aim of assisting trade union leaders, allied organisations, and advice offices themselves to be able to engage further and experiment with these emerging approaches to worker organising.
In order to develop this report, a number of methodologies were employed: • A literature review was conducted to explore the history, current analysis, and international experiences of worker/community advice offices.
• Interviews were conducted amongst key informants, trade union leaders, activists, and academics on their interaction with advice offices.
• A broad telephone survey was conducted amongst over 130 CAO's around the country. The research team obtained a list of all the CAO's around the country, which at the time was estimated to be in the region of 500. The team successfully interviewed 129 of those. Most of the other CAO's, the team was not able to interview for one reason or another; either no one answered, the numbers were disconnected, there was no one available to do the interview, and the list could have been out of date as many advice offices are shoestring operations with short lifespans.
• The team also did more in-depth follow-up phone interviews with a group of advice offices who had more than 50% labour issue cases.
• Six site visits were conducted of advice centres around the country which in turn formed the basis of six case studies. The site visits included semi-structured interviews with staff, clients, activists, and others involved in the advice offices.
Of the 129 advice centres surveyed by telephone, all provided labour-related services, with almost 30% spending more than half of their time dealing with workplace and labour-related issues. This shows a high demand for labour-related advice in communities where they operate. Only 17% of respondents said that less than 10% of their cases are labour-related, and another 40% dedicate between 10% and 50% of their work to labour-related issues. Among the most common labour-related issues that advice offices deal with are, first of all, unfair dismissals, followed by wage arrears, and issues with payments to UIF and pension funds. The research points to the fact that the most commonly advised sectors of the economy are farm work, domestic work, retail, construction, security, general labour, and mining. The phone survey found that there appeared to be a general lack of collaboration between advice offices and trade unions. In some cases, advice office representatives did not know anything about union activity in their area or sector of the economy. The report argues that advice offices see part of their role -and a reason for their resurgence -in filling these gaps. Sometimes this includes providing a home for casual or labour broker workers to organise and meet, taking up community campaigns around worker issues that trade unions have failed to take up, or providing basic representation where trade unions have failed to support members or organise workers. For immigrant-oriented offices, this can mean building an organisation that is more reflective of the identity and concerns of immigrant workers in its makeup, services, and staff, while at the same time avoiding the well-known fear of victimisation associated with trying to build a trade union for migrant workers Yet, the strategies of the worker-oriented advice offices which were interviewed for this report differ on how to fill these gaps. Some advice offices aim at empowering and training workers so they can go back into their unions and help unions re-engage in these struggles with competent leadership from the bottom up. Other advice offices see themselves as creating the space for something completely new to emerge, particularly for casual workers. In this vision, existing union structures and approaches are a thing of the past but, the report argues, the new forms of organisation and power can only be created through on-the-ground engagement. 21 In sum, the VWTT's impact has been limited. Few affiliates engaged with the VWTT pamphlet (Webster/Forrest 2019). This reflected the gap between rhetoric and practice which had characterised COSATU's engagement with vulnerable workers over the years. It had not 21 The report also proposed that the results of the survey be tabled at a national workshop of all stakeholders. This never took place as the VWTT did not continue beyond 2015.
fundamentally shifted its organising strategy to incorporate precarious and informal workers. A growing estrangement was evident between unions and its members as the politics of the Triple Alliance -COSATU, South African Communist Party, and African National Congresswas reproduced in COSATU. Its leadership was preoccupied with influencing Alliance politics and largely indifferent to representing more vulnerable workers (Craven 2016: 66).
The VWTT account also demonstrates that organisational experimentation around precarious work was mainly happening outside of traditional trade unions.
Despite its limitations, the VWTT brought informal and formal workers together for the first time in COSATU. What emerges from these VWTT discussions is the complex nature of precarious work, and the different demands these workers were making on those in standard employment. This is illustrated by the content of the VWTT pamphlet designed as an educative tool for unionists. A key demand was the removal of obstacles for own account workers to earn a living income rather than to claim a living wage. This new way of framing remuneration demands took into account changing workplace and employment relations.
One further unexpected outcome of the VWTT is that in 2015, it contributed directly to discussions on the ILO's Recommendation No. 204 on the "Transition from the Informal to the Formal Economy". COSATU's role in supporting the recommendation was important, as "the established unions' power, experience and strategic approaches" boosts the confidence of informal worker organisations (Webster/Forrest 2019: 67). As Webster and Forrest go on to argue, "while these organizations cannot rely on traditional unions to fight their battles, they simultaneously need to engage them and demonstrate new ways of organizing and bargaining which involve the State" (Webster/Forrest 2019: 68).

Conclusion: Rethinking Decent Work in South Africa
We have argued that there is a need to rethink the concept of decent work, not simply as a diagnostic tool to measure progress towards the realisation of decent work, but also as a way of facilitating worker organisation and their struggles for decent work. By drawing on the diagnostic tool developed in the GDED study, we have shown how the concept of decent work has been positively reframed to be a potentially valuable tool in organising precarious workers. We have also shown how trade unions and vulnerable workers appropriated the concept of decent work to organise workers and campaign for better working conditions. In the process they were forced to reframe their demands to meet the specific conditions of the South African labour market.
In the case of vulnerable workers in COSATU, the great value of the concept of decent work is that it provided organisers with a common language in their attempt to cross the divide between formal and informal work. Arising out of the South African experience of organising in the informal economy, South African representatives, including COSATU, were able to provide an input into the content of Recommendation No. 204 and pushed hard for its adoption at the ILO's 104th International Labour Conference in June 2015.
In NUMSA, the decent work questionnaire was used for the combined purpose of research and organising. The survey identified decent work deficits at petrol stations in core categories and showed that a job in the core of the labour market in South Africa does not necessarily equate with having a decent job. The approach, however, contributed to the recruitment of petrol attendants at petrol stations; a group of workers that is particularly difficult to organise. For union organisers a decent work questionnaire could be a useful tool to undertake a systematic assessment of workplaces, to engage with workers and to initiate a conversation on decent work deficits on the shop floor.
At the centre of our argument is that it is the mobilisation of workers -the agency of labour -that is crucial in the realisation of decent work. 22 First, social dialogue or -in our understanding -workers' voice is one of the indicators for the realisation of decent work and at the same time the most neglected dimension. Of the eight core labour standards of the ILO, it is No. 87 on the "Freedom of Association and Protection of the Right to Organise Convention", adopted in 1948, that has the lowest number of ratifications compared to the other conventions (ILO n.d., Prieß 2010: 158). Second, to realise decent work only on a policy level without the involvement of those who are affected at the workplace is hardly feasible. As Hauf argues, we cannot not want poor working conditions to be improved, no matter how slow and incremental. However, we must also ask whether such improvements empower workers to collectively organise for their interests and actively fight for further societal change (Hauf 2015: 151).
We have shown in this article how collaborative research processes can make a contribution to build workers' associational power and to strengthen their voice in the process of shaping a decent work agenda.
On the one hand, as Munck argues, organised labour cannot be simply written off: Organised labour is clearly part of the solution as well as being a problem at times […]But even if we are pessimistic about the prospects that trade unions might restructure and re-energise to face the new challenges to labour, we need to acknowledge that they do make a difference for those in a precarious position in the labour market and that agency really does count in terms of shaping the future (Munck 2017: 95).
On the other hand, as the experience with the VWTT indicates, attempts to organise informal and precarious workers by traditional forms of union organisation often coincides with the rise of new forms of associational power of informal workers. The demands and the form of worker organisation will be different in countries where a significant number of workers are in the informal economy. It is particularly in this context that experimentation with new forms of organisations, outside of traditional unions, is also taking place. They often become hybrid organisations, which include new forms of organisation that blur the distinction between traditional trade unionism, informal workers' associations and cooperatives. An understanding of these new forms of organisation must be at the centre of reframing decent work in Africa. In South Africa, we have identified experimentation with new forms of organising, such as 22 See a further development of this argument in Mashilo 2019. community and worker advice offices. In which way and form the agency of workers is built and how this impacts the realisation of a decent work agenda remains an open question for further research.