The Transvaal branch of Mawu began by holding general meetings in Kwesine Hostel, but unlike the rallies of Saawu and Macwusa, which were often attended by students and the unemployed, it invited workers only. Eventually, up to 9 000 workers were attending Mawu rallies on the East Rand.37 Mayekiso recalls: ‘We used to meet the guys outside the compound and explain to them about the union … After that we went into the compound secretly and signed up people. I would visit them in their rooms and managed to get 600 members in this way.’38 Thereafter migrants mainly joined in groups, recruited by fellow workers, or organisers in the hostels. In Durban, the union adopted the same tactic, although Erwin remarked dismissively: ‘We never feared them [the general unions]; they looked far too much like us in the seventies.’39
A further incentive to mass recruitment was the fear that once the government lost faith in the Wiehahn reforms controlling unions, it would move to smash them. There was safety in numbers.
Another strategic shift involved the rapid organisation of large plants by organisers who told workers they could only put demands to management with majority support. ‘We were going through factories of more than a thousand in a week,’ Erwin recalls.40 Because of its emphasis on shop-floor strength, Mawu was able to retain members in a way that the general unions could not. Recruitment and organisation acquired a life of its own, as Adler remembers:
In the seventies we were more externally based, because we didn’t have the internal organisation, people and contacts and we were fewer. So we would do a lot of work at the gates and off factory premises. In the eighties we were bigger and had more factories. You couldn’t be at twenty factories on an afternoon to recruit, and the fear factor was less, and you had more organisation, and we’d get shop stewards to meet workers at home in the township, or pop along at lunchtime to the factory next door. That happened in Benoni industrial sites. It started to get a momentum of its own, that would occur at the level of disputes, you’d have someone to help so it started to be a lot more based in the factories.
To register or not
Union registration under the Wiehahn laws deepened the divide between Fosatu and other emerging unions. In the late 1970s, when the government threatened to exclude migrants, all emerging unions united in spurning registration but after government dropped this threat and many employers still refused to recognise unregistered unions, Fosatu decided to break ranks. After successfully polling its members, Numarwosa was the first to apply for registration.41 However, registration remained anathema to some radical Cape unions, including the African Food and Canning Workers Union, Saawu, Macwusa and, most notably, the Western Province General Workers Union (WPGWU). They argued that registration meant state control and, by placing decisions in the hands of officials, it forced unions to operate bureaucratically. The WPGWU believed that registration ‘spells the death knell of workers’ control of the unions the most important element of their power’,42 and it argued that the new unions’ participation in statutory industrial councils would undermine their bargaining strength because they were not strong enough to outvote their white counterparts. Their power lay on the shop floor, where they could strike and enforce agreements. They contended that worker power, not state concessions, should force recognition on employers.43
Bitter controversy raged in the alternate media including in the South African Labour Bulletin. For Fosatu, the new dispensation was in itself a battleground, particularly the registrar’s power to register unions by race groups. Fosatu would not accept racial registration, but argued that unions should register while boycotting the official bargaining system until they were allowed to represent all races. The federation took the tactically flexible view that registration should be measured by how it bolstered the drive for recognition agreements which would entrench workers’ rights and serve as platforms for further gains.
Soon after Numarwosa registered, Mawu submitted an application emphasising its commitment to non-racial organisation. It was embarrassed and infuriated when the government registered Mawu for Africans, but the Natal Supreme Court overturned the decision in a historic ruling.44 But in any event, developments in the fast-moving labour arena soon rendered the controversy obsolete. A number of pragmatic employers granted unregistered unions factory rights, while the 1980 strike at Volkswagen highlighted that registration had not tamed the Fosatu unions.45
The Wiehahn laws and registration softened employers and brought a sharp rise in recognition agreements won by Fosatu’s metal affiliates. In September 1980, Mawu signed its first agreement with Precision Tools; agreements with Tensile Rubber and Henred Freuhauf followed a year later. By 1984, 122 of the union’s 180 organised factories had conceded recognition. The tactic among some employers of demanding more than 50 per cent union membership had the ironic spin-off of forcing organisers to recruit with greater energy, and so strengthened the union on the shop floor.
By 1983, Naawu had organised 18 000 members and had won recognition in five of nine auto companies with two further agreements imminent.46 It was formally recognised in 27 of its 28 auto, rubber and components factories by 1984.47 Recognition agreements, however, did not always include wage bargaining. The auto agreements embodied an understanding that, as registered unions, Numarwosa, and later Naawu, would negotiate wages through the Eastern Cape industrial council. Because of their history, they had an appreciation of industry-wide organisation and ‘a sense of pragmatism and power not shared by the rest of the unions’.48 In addition, by 1982 Naawu represented 40 per cent of auto workers in Port Elizabeth/Uitenhage and could bargain in the council from a position of strength.49
But Mawu lacked the numbers to wield power on the huge metal industrial council. It therefore pursued a policy of plant bargaining and the negotiation of a recognition agreement often became a battle to include a clause which would permit it to plant-bargain wages. This brought it into conflict with the thrust of the Wiehahn laws and the metal employers’ body, the Steel and Engineering Industries Federation of South Africa (Seifsa), which insisted on wage bargaining at industry level.
Nevertheless, recognition agreements enshrined important factory rights, including shop steward recognition; access to factories by union officials; stop order facilities; dismissal and grievance procedures; the right to report back to membership; and other important clauses such as retrenchment provisions. At the centre of these agreements was the role of shop stewards which located power in the hands of union membership. In consequence, these agreements advanced workplace organisation and were a significant strategy for building power. As binding contracts, they gave the union permanent workplace bargaining rights enforceable in law and employers now had to adhere to dismissal procedures which made arbitrary sackings difficult. Agreements were not viewed as a substitute for factory organisation, but as an additional weapon in the fight to humanise, democratise and balance workplace power relations. As Fanaroff observed:
I think it serves a purpose both psychologically and legally, because management is bound by certain rules. And at times when we are weak it serves as a defence because it becomes an accepted way of working. Where organisation is not developed sufficiently the document cannot be used effectively. You can’t have a recognition agreement as a substitute for organisation.50
Recognition agreements usually gave company check off facilities where employers deducted union dues from workers’ wages. This was a significant right conferred by the Wiehahn reforms. In 1975 Mawu’s monthly income was about R1 000 a month, with little money for organisers’ salaries, rent or other organisational costs. Kubheka remembers: ‘We didn’t have an administrator in early 1976. Volunteer Wits students came in now and then to do typing and admin … there was no money so organisers did not have a fixed salary or wages. It would be