Gender, union leadership and collective bargaining: Brazil and South Africa

Abstract

Links between gender politics and leadership in trade unions and how these impact collective bargaining gender agendas are explored in this study of trade unionism in Brazil and South Africa. What the International Trade Union Confederation and others refer to as ‘unexplained’ gender pay gaps are discussed in relation to the absence of women in the collective bargaining process. This examination draws on research in both countries and concludes that gender leadership gaps and gender pay gaps are related.

Introduction

Despite women’s increasing labour force participation worldwide (International Labour Organization (ILO), 2010) and the fact that union membership growth is higher among women than men (Pillinger, 2010; Yates, 2006), gender democracy in unions remains elusive and gender gaps persist in union leadership and in collective bargaining (Ledwith, 2011). Complex and shifting power relations are involved, which we explore through research with trade unions in Brazil and South Africa – countries where there are similarities as well as differences in this field and where industrial relations in each country are not infrequently compared (Coleman, 2013; Seidman, 1994).

Gender inequalities may seem paradoxical in two nations which on coming out of political oppression – military dictatorship in Brazil and apartheid in South Africa – developed strong constitutional rights to equality and for women, later becoming rapidly expanding economies.¹ Yet, although there have been corresponding moves towards gender equality in trade union constitutions in each country, differences between the rhetoric and reality of union gender democracy remain. We explore how these differences play out in the gender politics of union leaderships and what this might mean for gender bargaining and pay gaps, drawing on data from our empirical study of 2011–2012, with female (mainly) trade union leaders and activists.² We examine the relative absence of women in union positions of power and leadership and the strategies of closure and marginalisation practised by the (mainly) male senior role holders in a masculinised culture. Women’s leadership roles were predominant at the workplace, and even when they were in more senior positions were identified as without influence, and mainly absent from collective bargaining, although this was identified by them as a key aspect of trade union leadership.

The article begins by outlining the context for gender democracy in Brazil and South Africa, before overviewing gender issues in union leadership and collective bargaining in each country’s union movement. It then discusses the central issue of gender pay gaps and their relationship with collective bargaining and gender agendas, as the prelude for presenting our own research, findings and conclusions.

Gender democracy in Brazil and South Africa

At the start of the modern democratic period in each country, gender democracy was put at the heart of constitutions. Following the military dictatorship’s demise in the 1980s, Brazil’s 1988 constitution ensured a range of rights on equality and justice, with Article 5 declaring that ‘men and women have equal rights and duties’.

In 2003, under his first presidency, former trade union leader Luiz Inácio Lula da Silva created a National Conference on Women’s Policies, a National Plan for Women’s Policies and reaffirmation of the commitment to incorporate a gender perspective in all public policies. Nevertheless, a male leader in the ruling Brazilian labour party, the Partido dos Trabalhadores, is quoted as saying, ‘… in practice, internally those who define the routes of the party are men; the same ones who have been in charge for a long time. Men are those who define and write the strategies of the party’ (Sacchet, 2009: 160).

In South Africa, the 1995 constitution made gender equality a founding promise of the post-apartheid State which

… may not unfairly discriminate directly or indirectly against anyone on one or more grounds, including race, gender, sex, pregnancy, marital status, ethnic or social origin, colour, sexual orientation, age, disability, religion, conscience, belief, culture, language, and birth. (Section 9(1))

In both countries, gender quotas have been an important part of affirmative action policies and practice to move towards gender democracy, and these have been carried over into labour movements, with varying degrees of success.

Gender, union leadership and collective bargaining…

While many unions internationally now have comprehensive gender and diversity equality policies and structural rights in place, practice consistently fails to match the rhetoric. Although no longer systematically excluded altogether, women do continue to be marginalised from union power (Briskin, 2006; Colgan and Ledwith, 2002; Franzway, 2001; Parker, 2006). Gender politics in trade unions are played out at all levels, but are especially potent when it comes to leaderships (Healy and Kirton, 2000; Ledwith and Hansen, 2013).

Recent findings from across developed, emerging and developing economies confirm women’s continuing lack of proportional representation in positions of power in unions (Britwum et al., 2012: 50). Even where women members are in the majority, studies show a dearth of women leaders or their relegation to subordinate roles such as treasurer or organiser. Where unions and/or confederations have reserved seats for women, these are often as secondary office-holders such as ‘second vice-chairperson’ (p. 52). Nevertheless, there has also been progress towards union gender proportionality and getting women’s concerns onto union agendas (Ledwith, 2009;Parker and Douglas, 2010). In the UK, Gill Kirton (2014) finds evidence ‘strongly suggesting’ that progress is linked to the wide range of gender equality strategies of the past 30 years, but also warns of persistent gaps, seeing the union gender democracy project as unfinished business (p. 1).

Research in many countries has identified the role of masculine homosocial cultures and discourse in unions, contributing to women’s exclusion and marginalisation (Ledwith and Hansen, 2013; McBride, 2001). The notion of leader as male hero is deeply embedded in labour movements (Kaminski and Yakura, 2008; Pocock and Brown, 2013). Such cultural norms may be especially important when it comes to collective bargaining, as there is evidence of demands for gender equality and rights being ignored, discounted or dropped off the agenda (Britwum and Ledwith, 2014). Dawson’s 2014 study of the UK printing industry shows how a masculinised, long-standing shared understanding between trade union negotiators and employers about the culture and workings of the bargaining process is at the heart of the power relations that work to exclude and marginalise gender agendas.

Conversely, there is also evidence that where women and equality champions are involved in bargaining, the likelihood is of improved pay and conditions for women. A Canadian study documents the importance of women’s membership of the negotiating committee at a large state utility in intensifying demands for equal pay and maintaining the pressure over years (Creese, 1999). In the UK, Heery and Kelly (1988) found that female full-time union officers (FTOs) were more likely to promote the interests of women in collective bargaining and to prioritise recruiting and organising women. Young, highly educated, white collar, left-wing male FTOs were also likely to support gender agendas. Heery (2006) identified a ‘vanguard, an active minority of officers (women and younger men) who have engaged in equality bargaining’ (p. 459).

These findings concur with those of Linda Dickens, that equality initiatives in agreements require the involvement of women, both as participants in the actual negotiations and as providing input behind the scenes, doing research, formulating bargaining proposals and being on working parties. Women involved in bargaining develop different agendas, and their presence can not only transform the bargaining process and the interpretation and implementation of the claims but can also help challenge and undermine the prevailing masculinised ideologies and practices underpinning collective bargaining (Dickens, 1998: 34). These claims are confirmed, with reservations, through more recent discussions (see Briskin, 2014).

… and gender pay gaps

Factors in the gender pay gap are widely debated. Orthodox theory focuses on individual-level variables (e.g. human capital), occupational gender composition, and industry and organisational characteristics. Institutional theories consider wage structures, occupational gender segregation, sectoral and contractual differentials, and the role of unions and collective bargaining. Rubery et al. (2005) argue that the extent and coverage of collective bargaining are critical in maintaining wage floors in countries without national minimum wage systems and are also significant where the level of minimum wages is low or poorly enforced. Although unionisation is generally associated with smaller gender gaps (Bryson, 2007), US research found contrary results in industries where unions had a lower commitment to pay equity issues and women were under-represented in leadership positions (Elvira and Saporta, 2001).

In Brazil, there is a national minimum wage adjusted annually, but in South Africa there is none. Coleman (2013: 1, 86), comparing collective bargaining and allied strategies in the two countries, finds South Africa wanting, with collective bargaining institutions fragmented, constantly destabilised and with no coherent wage policy. He concludes that what is needed is a national wage policy, a national minimum wage and centralised bargaining, and that South Africa has much to learn from Brazil’s development of a coherent strategy to increase the real level of minimum wages and address the plight of the working poor (pp. 11, 85).³

Meanwhile in both Brazil and South Africa, the gender pay gap remains. In Brazil’s mainstream economy, women earn 70% of men’s earnings overall (Tijdens and Van Klaveren, 2012: h). Ironically, the gap closes among informal workers, with women earning 92% of men’s earnings. According to Rosana da Silva, leader of the Women’s Colectivo (women’s committee) of the Central Única dos Trabalhadores (CUT), the gender pay gap arises from the notion that ‘a woman’s wage is just a “plus” in the household income’ (Ledwith et al., 2012). It takes time to change beliefs that are so deeply embedded. In 2009, the CUT launched a permanent campaign against gender pay inequality. In Brazil, the interplay of ethnicity⁴ and gender means that indigenous women fare the worst (Atal et al., 2009). Despite achievements in education and occupational attainment, Afro-Brazilian women continue to earn significantly less than men, both overall and within their own ethnic group (Lovell, 2000), and the gender pay gap in the informal sector where women dominate continues to be ‘severe’. Discriminatory practices in wage formation continue to have a major impact on women’s pay; Van Klaveren et al. (2009) suggest the existence of a sticky floor for women working in informal sectors, with continued glass ceilings in the formal sector where women receive lower wages. A recent view is that the gap is now widening again ( Forbes, 2012). For race, a single message emerges: pay differentials by race widen across formal and non-formal sectors.

In 2012 an ITUC report identified South Africa as having one of the largest unexplained gender pay gaps – 22 percent (Tijdens and Van Klaveren, 2012c), although this is partly due to class difference. Despite household incomes more than doubling during the 10 years from 2001 to 2011 there have been limited gains for the working class in post-apartheid South Africa. Ncube and Tregenna (2013) find that the gender wage gap among Black South African workers can also be located within the ‘sticky floor effect’ paradigm where the gender wage gap widens at lower levels of the wage distribution.

Black women are the lowest paid. Generally, they work in private households, do not belong to a union and are aged between 15 and 24 years (South African Institute of Race Relations (SAIRR), 2013). A key factor is less access to education and other services for Black than for White women and the majority of Black men. Violence against women is also a major issue in South Africa, where the rate of sexual violence was regarded in 2013 as the highest in the world (Faul, 2015). The United Nations South Africa (2013) reported that in Gauteng Province in 2010, over half (51.3%) of women had experienced some form of violence in their lifetime and 75.5% of men admitted to perpetrating violence against women at one point in their lives.

In July 2013, the South African Statistician General reported that the wage gap between men and women continued to cause concern, with poorly paid women still most at risk of poverty and violence and links these to the lack of access to education. The interaction between race and gender placed Black African women at a particular disadvantage (eNews Channel Africa (eNCA), 2013).

The race categories used under apartheid continue and race/ethnicity is an important identifier of earnings/wages hierarchies, with Whites as the highest earners, then Asians, followed by Coloureds and finally Blacks at the bottom (Allanson et al., 2002). Maciej and Tyrowicz (2009) found a race pay gap of 30%–55%. Median wages of White earners are four times those of Black earners; of skilled employees six times those of unskilled employees; and of unionised employees twice those of non-unionised workers (SAIRR, 2013). According to Bhorat and Goga (2010), the top 5% earn 30 times more per month than the bottom 5%, and White people earn in an hour what African workers earn in a day. The private sector continues to pay White men nine times more than Black men and White women nine times more than Black women. Current bargaining strategies reinforce the apartheid cheap labour structure. Following the huge job losses accompanying the financial crisis, unions have been pressured by employers to engage in concessional bargaining (Coleman, 2013: 3, 7).

Pay and union membership

Trade union membership confers a wage premium on members (Addison, 2014). Estimates range from 10% to 30% depending on variables, including gender and skill level. In Brazil, this premium is estimated at 17% (Arbache, 2001). Bhorat et al. (2014), discounting for gender and skill, estimate premia of 5%–12% in South Africa. According to the ‘equalisation hypothesis’, whereby unions attempt to achieve equal absolute pay increases for all workers, the result should be a decline in gender and race pay gaps (Wunnava and Peled, 1999). Unionisation has been found to lower the male–female earnings differential by increasing female earnings significantly more than male earnings. Casale and Posel (2008) discuss how studies in South Africa have been largely concerned with the impact of union membership on the wages of African and White male workers. They conclude that unions do compress the distribution of wages in South Africa and that racial inequality is lower in the union sector compared to the non-union sector. For both men and women, average hourly wages among union workers were substantially higher than among non-union workers. Male union members earned twice as much, and female union members three times as much, as their non-union counterparts. A much higher proportion of workers in unionised jobs reported having a written contract with their employer and receiving paid leave, medical aid and pension benefits (p. 11).

Nevertheless, in both countries, the gender pay gap is widening, with women – Afro-Brazilian women in Brazil and Black African women in South Africa – at the bottom. Often there are accompanying comments to the effect that this is unexplained. Yet, much of this research overlooks possible explanations. It does not tell us who is involved in the collective bargaining processes, nor what demands other than pay are on negotiating agendas. To explore this further, we first set out the gender situation in unions in each country before moving to our empirical research.

Women and union leadership

Brazil

Overall trade union density in Brazil is around 18%,⁵ with women’s density estimated at 15%–16%. Collective bargaining coverage reaches 60% of workers in the formal sector (Hayter and Stoevska, 2011). Trade unionism and bargaining are virtually non-existent in the informal sector. Across all sectors, women are predominantly those in precarious work.

The Brazilian CUT, established in 1993, is the largest and most important trade union confederation in Brazil, representing over 7.4 million members. Following government action, the union movement has also adopted gender quotas to try to move towards gender parity. A 30% women’s quota was set in 1993. By 2012, women were approximately 40% of CUT’s members, but despite the quota, there were only six women among the top 25 leaders. In 2012, the CUT Congress agreed that parity – 50% – was to be implemented in 2015. However, the CUT quota does not officially extend to affiliates, although some unions do apply it. Also in 2012, CUT’s national Congress elected 32 members for the national board (executive), of whom 31% were women. Yet, women in those senior leadership roles tend to hold positions with little or no political or decision-making power. Of the top leadership positions in CUT, only 10% are held by women – as president, and with 24% of ‘first secretaries’ being women and 16% of treasurers. In 2006, a survey at CUT’s Ninth National Congress indicated that 20% of delegates said that a woman was the president of their union, 34% reported a female union treasurer and 43% said their general secretary was a woman (Ledwith et al., 2012: 185). Only the domestic workers and some teachers – those in feminised unions – were in unions where the quota was surpassed.

South Africa

Union density in South Africa is estimated at around 38% and declining in the formal economy, with density ranging from 25% in the private sector to 69% in the public sector (Bhorat et al., 2014). Collective bargaining coverage is only 17% – entirely in the formal sector. It is estimated, following structural shifts towards public and service sector employment, that women are now about 48%⁶ of members of the Congress of South African Trade Unions (COSATU). This is an increase from 36% a decade earlier (Bischoff and Tshoaedi, 2012: 49). A third of the COSATU leadership nationally are now women.

In 2012, of the six top leader positions in COSATU itself, two were held by women: the second deputy president and the treasurer – positions with symbolic power only. In the nine provinces things were worse, with only three women in leadership positions (18%): chair in the Northern Cape, deputy chair and treasurer, in Free State. The powerful position of general secretary was held by a man in all but three of COSATU’s 21 affiliates,⁷ giving women just 14% of significant leadership roles.

Behind these disappointing data, research has found that gender equality work in unions has for years been resisted by the men. A resolution on the sexual code of conduct was blocked at the 1989 COSATU Congress, and it took 7 years for COSATU to adopt a policy against sexual harassment (Geisler et al., 2009). It set up a National Gender Unit comprising a National Gender Coordinator supported by a National Gender Coordinating Committee with senior leaders, workers and researchers, but the gender coordinator in 2009 reported general resistance. Although COSATU has a policy which aims to increase women’s representation in leadership structures and which it expects its affiliated unions to implement, there has been resistance to calls for 50% quotas for women in union leadership roles (Tshoaedi, 2012). Despite progressive resolutions adopted at COSATU’s Third National Gender Conference in 2012, proposals for implementation, such as action against sexual harassment and the Equal Pay for Work of Equal Value declaration, have not been implemented.

Bargaining and gender

Brazil

In Brazil, there is a substantial cadre of women at various levels in their unions who are pushing for a progressive agenda through standing for, and being elected to, union positions (Ledwith et al., 2012). Two examples show what can be done. By the 1980s, as unions became the vanguard of political activism against the dictatorship, women bank workers demanded better working conditions, childcare and political participation in the country’s future and achieved the first equality clause in the relevant collective agreement – a milestone for the Brazilian working class (Portilho, 2014). This gain also paved the way for the incorporation of maternity rights into the 1988 Brazilian constitution. In a very different environment in Northeast Brazil, the agricultural workers’ union, realising the significance of its large female membership, campaigned around issues important to women workers, recruiting women into leadership and bargaining for gender. Gains have included 2-month paid maternity leave, crèche facilities, paid leave to visit doctors and to breastfeed for an hour per day. This has also had an important role in women’s personal empowerment (Selwyn, 2009: 194).

However, more common is the low insertion of the women’s claims into labour standards, which are limited to pregnancy and maternity rights (present in 78% of collective agreements analysed), workplace gender equality initiatives, such as incentives for the training and qualification of women (7%), prevention of sexual harassment (2%) and prohibition of discrimination (18%) (Vieira, 2013).

South Africa

In South Africa, high union density occurs mainly in masculinised sectors such as mining and metalworking, with bargaining limited to a narrow range of issues, mainly wage increments, in agreements which are often ‘no more than a couple of sentences’ (Coleman, 2013: 78). Given women’s labour market position, they are even less likely to be covered by negotiated agreements. While we do not know the extent to which women, and especially the gender coordinators, are involved in collective bargaining, given the evidence above about the widening gender pay gap, it is reasonable to assume that they are few. Yet over a decade ago, COSATU was advocating inclusion of women in bargaining teams, developing the role of gender coordinators and structures in collective bargaining and building a strategy to ensure the involvement of women in collective bargaining demands. The retail union SACCAWU⁸ has pursued this objective since the 1980s with some success; although despite prioritising maternity and parental rights in company agreements, it has also been hampered by male-dominated negotiating teams and employers (Budlender, 2009). Just so, the bargaining demands listed by affiliates to COSATU’s 2012 national congress focused on pay and working conditions with women’s wider demands receiving little attention.

Now we explore these issues further drawing on our empirical research.

The research

The research approach was both qualitative and quantitative, and triangulated using multiple methods (Table 1). It aimed to explore a wide range of issues of gender and leadership in mainstream unions in each country in a broadly comparative research programme. In this article, data about leadership and collective bargaining are brought together to examine the relationship between them. Access was through Global Labour University (GLU)⁹ colleagues. Fieldwork was carried out in Brazil in 2011 and South Africa in 2012.

Table 1. Research programme and data collection.

CUT: Central Única dos Trabalhadores; COSATU, Congress of South African Trade Unions.

Ledwith spent 5 weeks in each country, working with a researcher¹⁰ who had previously administered questionnaires through a wide range of networks and meetings and who set up research meetings and discussions. In Brazil, the questionnaire was translated into Portuguese, and the results were translated back into English. In each country, terminology was adjusted to fit local usage. In South Africa, questions were added about collective bargaining. SPSS analysis and transcription were carried out by additional colleagues in each country.¹¹

Brazil

As shown in Table 1, data were collected mainly with women and some men. One source was a meeting of women in the national Women’s Colectivo (committee) who were discussing their forthcoming demand to CUT’s congress to increase the gender quota. The women and men respondents were active in both cases mainly at the workplace and at regional/state level (Table 2). Although three-quarters of the Brazilian women described themselves as a leader in their union, 17% said they were not, and eight did not know. In all, 90% of the men said they were a leader.

Table 2. Level at which women responding to the leadership questionnaire were active in union leaderships, Brazil, 2011, and South Africa, 2012.

Percentages sum to more than 100 for each country as some women were active at several levels.

Almost three-quarters of the respondents were White, two-thirds self-identified as Brown, 43% as Black, 3% as Yellow (see Note 4), 2% as Indian and nine said they did not know. In a macho country like Brazil, it was not surprising that only 11 people identified as anything other than heterosexual, with 17 not answering the question at all. The largest age group was 40–55 years (55%), and of the women, a quarter was aged between 36 and 45 years.

In all, 41 women described themselves as feminists (39%), with 50 (48%) saying ‘No’ and 17 (16%) did not know. Feminists were those who claimed that they especially brought to the table skills of dialogue, experience of collective bargaining and knowledge of the gender bargaining agenda, whereas these capabilities were identified less often by the not-feminists and ‘don’t knows’

South Africa

In South Africa, of the 80 COSATU women and 11 men who completed questionnaires, 52 (65%) were worker-leaders (elected representatives) and 25 were office bearers (paid officials). As in Brazil, women were mainly active at regional and provincial levels, which overlapped (Table 2). In all, 22 self-identified as Black, 35 as African, 6 as Coloured and 3 as White; 35 did not identify their race/ethnicity.

They were slightly younger than the Brazilian women with the largest group, just over half, aged between 36 and 45 years; a quarter were between 26 and 35 years, and 29% were aged over 46 years. In all, 90% of the women had caring responsibilities, mainly for children (71%). Altogether, over 80% of the South African respondents defined themselves as a leader. In total, 56 (70%) were in gender-related roles, including four men. An overwhelming 89% of the women also said that they were gender activists even if not in a formally designated gender role. A total of 30% of the women identified as feminists.

Feminists

In both countries, the feminists were highly politicised, belonging to the communist party and the ruling political party, generally positioning themselves on the left. Feminists also provided deeper analysis, and especially, they wrote more about their unions and the need to open these up: ‘The union is not an exclusive place for men anymore. Our presence brings a new element to our struggle as workers and gender’, and more widely, ‘I fight for the feminist movement, against inequality in the workplace, try to change the logic of gender oppression’. ‘I believe in woman power and as woman we can change world, if we work together. I am a feminist who is a gender activist’.

Leadership and collective bargaining

We now examine whether the gender power gap already identified in union leaderships followed through into collective bargaining and its outcomes.

Brazil

In the group discussions, women emphasised the importance of preparing women for positions and leadership roles, including in collective bargaining. This involved the development, empowerment and awareness of women themselves, encouraging, supporting and mobilising women onto direction and executive boards, and especially onto collective bargaining teams. For women to work together to do so was seen as essential, and unions such as those covering bank workers had set up women’s groups and committees to pursue their goals. Others found this difficult owing to discouragement within their unions. The women interviewed frequently reported antagonism and resistance from male union colleagues – ‘men close in, leave no space for women’ and ‘women do not feel they have a right to speak’. There was little networking across unions, although some women, for example in Rio, did meet socially: ‘We go out and drink beer and discuss politics, clothes, share feelings, ideas, and barriers in the union. It is nice to do it in a free space’. A central goal of the women’s fight for equality was to develop and extend women’s employment rights. Here, the banking workers had had hard-won successes and were now working to get sexual and moral (bullying by senior staff of junior staff) harassment into their national collective agreement.

South Africa

Fewer than half the women (44%) questionnaire respondents reported being involved in collective bargaining, especially workplace shop stewards. This is important since workplace bargaining is a key aspect of collective bargaining in South Africa, with the other main level being sectoral (Godfrey et al., 2007). Of those women leaders involved in bargaining, only eight took part in drawing up the agenda, nine were lead negotiators, and only four did both. In all, 11 attended negotiations or were involved in some other capacity, but did not speak at the bargaining table. Yet, a quarter of all the women identified priorities for collective bargaining, with two-thirds putting up proposals relevant to women. Most frequently mentioned was pay equity and then came maternity benefits and rights linked with family leave, workplace childcare and time and facilities for breastfeeding, women’s safety at work and in the union, and implementation of existing agreements on sexual harassment, and also HIV and AIDS. These were all items identified at the 2012 Gender Conference referred to above.

For women, having to rely on male colleagues to get these items into collective agreements is problematic, and as discussed already, they are less likely to do so. As women leaders repeatedly told us, ‘women are capacitated but are excluded in many ways’.

Both countries

Nevertheless, in each country, the women recognised the significance of collective bargaining, identifying it as a high priority for their own union leaders (Table 3). Brazilian women also saw it as a top skill for themselves. For South African women, experience in bargaining was important in getting elected to positions but not in their existing leader roles. Also striking is that women in each country identified equality and diversity as highly important, but not as something valued by their unions. Indeed, it came fourth.

Table 3. Three most important aspects/parts of being a union activist/officer/leader: women respondents, ranked 1–3 in each category, Brazil (n = 105), 2011, and South Africa (n = 80), 2012.

These findings are closely linked to the most difficult aspects of the women’s roles (Table 3). These were as follows: getting the support of experienced leaders (mainly men), sexism and machismo, and the stress of public speaking. For Brazilian women, a mixture of exclusionary cultural practices dominated, whereas in South Africa structural aspects were more problematic: union bureaucracy, time pressures and workload. Additional problems for women were being away from home, public speaking, communicating and getting support from the union leadership.

Given the strong views among the women about women’s capacity and abilities, and the need for more women leaders, together with the mismatch of priorities in Table 3, we asked, ‘What does it take to become a leader?’ (out of 16 items). Table 4 shows the high priority accorded to displaying commitment to trade unionism, which is a given. More interesting is that the Brazilian women displayed political savvy about where support has to come from – key activists at grassroots and in top positions, combined with being in the correct political position, and how to get elected; by being visible, listened to and well known in the union. The South African women relied more on their proven record of skills and experience including negotiation/collective bargaining.

Table 4. ‘In your union, what is necessary to get into a leadership position?’ Responses – women union leaders, Brazil (n = 105), 2011, and South Africa (n = 80), 2012.

Discussion and conclusion

We propose that this research confirms the significant contribution of women’s active involvement in collective bargaining to closing the gender pay gap. It also suggests that their absence goes some way to explaining ‘unexplained’ elements of the gap. The key conclusion, which we can draw from both our own evidence and other research reported above, is that where women are politicised, active and collectively in positions of power, they can be effective and can influence collective bargaining agendas and outcomes. However, in too many situations, while women are able and willing to take on more senior leadership roles in their unions, they are frustrated in their attempts by their union ‘brothers’.

Whereas the women in this survey were predominantly in leadership roles at the workplace, in cases where they were in more senior union positions, these roles were identified as secondary or even without influence. Conversely, the men dominated the senior roles of significance. Given the macho attitudes and culture discussed among the groups, women will continue to struggle to be elected into leadership positions when (as they wrote on their questionnaires) the most difficult thing is obtaining the backing of experienced leaders – who are mainly men. This sort of support is also linked to key aspects of trade unionism, especially collective bargaining. Both women and men identified collective bargaining as of prime importance to their own unions. Nevertheless, when it came to identifying their own skills and strengths, capability in bargaining was not high on women’s lists, but was on the men’s. In both countries, women reported that there were good women ‘ready and capacitated’ to take on leadership positions, but that they were excluded in favour of those who were not; in other words, women who were less well experienced or not prepared to challenge existing gender gaps and power relations. Simply put, women are rarely in positions where they can be involved in collective bargaining, whereas this is normal practice for men.

Nevertheless, Brazilian women trade unionists have managed to be proactive and the CUT Women’s Colectivo is a powerful force, as witnessed by the increase in the gender quota to 50% in 2015. Also, since our research in Brazil, CUT has set aside funding for a 2-year project of women’s courses and publications, indicating that the gender issue is moving towards the mainstream. In both countries, education is a significant driver of women’s progress.

In contrast, in South Africa there seems to be a stasis currently, and not just in relation to gender. As Coleman has written, COSATU has recognised that the country’s labour movement is facing a major challenge in collective bargaining and lacks a coherent wage policy (including a national minimum wage) to address poverty and inequality among workers. In relation to gender, it also appears that within many unions, similar to the situation in the national gender machinery described in the 2009 report by Geisler et al., there is a lack of political will to resource, coordinate and support women. Yet, unions potentially have an important role in fighting discrimination and empowering women at the workplace and in society. South African union women told us repeatedly that there were able women ready to take on roles and others had resigned from frustration at the lack of movement, although they remained active in other women’s spheres. Tshoaedi (2002: 17) observed about South African unions that the closer women’s challenges go to the centre of male power, the more vigorous the resistance. After 10 years, there seemed to have been little change: ‘Political power and influence is critical in the definition of politics in the labour movement. Men who occupy positions of power in COSATU leadership define and conceptualise politics and political behaviour in masculine terms’ (Tshoaedi, 2012: 105). Meanwhile, South African trade union women continue to campaign inside COSATU for the ‘constitutionalisation’ of the gender/women’s structures which have been ratified by the national gender committee and conference but not by the main confederation.

In sum then, the data discussed here indicating the gender pay gap in each country, together with our research, show that when the intersections between race, gender, class and power are examined and unpicked and the links made between leaders, leadership and the gender politics of collective bargaining, these are key in the persistence of the gender pay gap and wider democratic deficits. We suggest that far from being ‘unexplained’, in South Africa particularly, the gender pay gap is attributable to the deeply sexist and racist culture of society, strongly manifest in the union movement and flowing underneath it.

In each country, our study was limited in the area of collective bargaining, especially in Brazil, but these findings do point to the need for further, more detailed research and scrutiny of regimes of gender in the bargaining sphere and more widely in Brazilian and South African Unions.