Global capital and local labour. Strategies and labour relations in the Hex River Textiles factory from the 1940s to the early 1990s

ABSTRACT This article analyses labour relations and management strategies in the Hex River Textiles factory in Worcester, South Africa, from the 1940s to the early 1990s. The factory was established by a French textile manufacturer in 1946, who relocated an entire mill from Bradford in England, to exploit the low wage labour provided by primarily coloured women. The strategy also included investments in new technology. The workers who were drawn into capitalist production resisted exploitation despite government attempts to crush the trade unions. In the late 1980s, trade union activity was rekindled, not least because there was a core group of coloured workers, who carried on the tradition. The strategy in the 1980s was less militant than in the 1950s, but, arguably more successful.


Introduction
In the twenty-first century, the textile industry has largely disappeared from South Africa. International competition has caused factories to close down and workers have become unemployed. During the mid-twentieth century, textiles were, however, one of the country's leading sectors. The state supported the industry and European companies invested locally. One example of a successful company with international ties is Hex River Textiles (Hextex) in Worcester in the Western Cape. It was established in 1946 when a French textile manufacturer decided to open a factory and relocated an entire plant from the wool textile centre Bradford, England. Such relocations have been common in the global textile industry since the 1930s when factories moved "from countries or regions of higher to those of lower labour costs' and were important in the emerging industry in South Africa (ILO, 1937, p. 112;Kilvington, 1996). But relocation of production in search of cheap labour does not automatically lead to profits, since capital must consider local conditions and institutions (Collins, 2003, p. 151). The Hextex case offers an opportunity to study the impact of the global process of capital relocation at the local level. It is particularly relevant to analyse which strategies capital used when relocating to an area with little textile and manufacturing tradition. Who comprised the labour force and how were labour relations formed in a rural South African factory from the 1940s to the early 1990s?
Worcester, 120 km from Cape Town, has never been an industrial centre, but Hextex is renowned in labour historiography for a strike in 1956. It is mentioned by Betty du Toit in her inside account of the early textile industry and by Ken Luckhardt and Brenda Wall in their history of the South African Congress of Trade Unions (SACTU). During the strike, coloured and black workers joined forces despite attempts by management and the state to divide them along racial lines (Du Toit, 1978; Luckhardt & Wall, 1980). Both these accounts are, however, brief and do not further analyse relations at the factory.
During the twentieth century's first decades, textile factories were almost unheard of in South Africa, despite the country's position as one of the world's foremost wool producers. 1 By the mid-1930s, only 12 textile establishments existed in the country. But after 1945, the industry quickly expanded, not least because of international capital's investments. In the early 1950s, there were 74 establishments, and almost sixty per cent of the capital in the industry was foreign, for example, from France, England and Belgium (Du Toit, 1978, p. 33). The South African state also contributed to the industry's establishment, for example through tariff protection. Through the Industrial Development Corporation (IDC), the state also financed several textile companies, often together with international capital, and attempts were made to attract industry to 'decentralised' areas in homelands and border areas. Homelands, or Bantustans, where black people were allowed to reside, constituted the foundation of the migrant labour system. Few economic possibilities existed in homelands and primarily black men were forced to seek employment in the 'white' areas.
Through the reproductive role of homelands, Harold Wolpe argued that apartheid was functional to capitalism, since families of black migrant men were sustained in the homelands and not by wages in white areas. Merle Lipton, however, argued that manufacturing capital in general opposed 'most apartheid policies', since they impeded labour mobility and generated low productivity (Wolpe, 1972;Lipton, 1985, pp. 139-159). The role of migrant labour was different in the Western Cape, since the government regarded it as the home of whites and coloureds, meaning that employing blacks in low-wage jobs threatened the coloureds. The coloured labour preference policy was therefore implemented in the mid-1950s to reduce the influx of blacks to the region. However, as Richard Humphries argues, the policy was never fully enforced and was 'partially redundant', not least because of economic growth (Humphries, 1989, p. 179).
The issue of low wages was clearly relevant in textiles. As Anne Mager notes, in the textile industry that developed in the Eastern Cape before the Second World War, white women constituted the principal labour force. They were employed on the lowest paid jobs while white men were employed for example as mechanics. But the gender and racial composition changed along with increasing capitalisation in the 1940s and 1950s. White women were replaced by black men, who could be paid lower wages. Factories that were established after the war based production on these black men. However, increasing costs caused by labour turnover and import of machinery in the 1960s gave incentives to textile capitalists to further reduce labour expenditure. By the 1970s, the black men were replaced by black women, who could be paid even lower wages. This gendered transformation impacted worker resistance to exploitation. Many of the black women were recently proletarianised and had not developed a worker consciousness. A common form of resistance was therefore to quit work. In the larger factories, workers were, however, forming an 'industrial culture' (Mager, 1989).
In other regions, the textile industry had a tradition of strong unions, shown during the early years of apartheid in the strikes at Hextex in 1956 and at Amato Textiles in Benoni in 1958. Phil Bonner and Rob Lambert argue that worker organisation was 'favourable' in the textile factories that appeared after 1945. Mechanised production empowered black and coloured workers, who became machine operatives, especially in the semi-skilled category, which gave them more bargaining power. White workers on the other hand were displaced or found their skills devalued (Bonner & Lambert, 1987, p. 352). Eddie Webster similarly concludes that mechanised production in South African foundries 'undermined the traditional division of skill and race' and 'shifted the balance of power away from the 'labour aristocrat' and his craft union, to the production worker and his industrial union' (Webster, 1985, p. 278). In the textile factories such as Hextex there was, however, no transformation from pre-machinofacture to machinofacture, wherein white workers lost power and whites did not disappear from industry. Kilvington notes that skilled workers from Europe were employed at the factory and emphasises their importance for the textile industry (Kilvington, 1996, pp. 235, 247). However, he does not discuss them a part of the working class in the racial economy of South Africa.
The non-racial trade union movement represented by SACTU was crushed by apartheid legislation in the 1950s. But it gained renewed strength after the Durban strikes in 1973 and the Wiehahn reforms in 1980, which allowed black unions. Organisation at factory level was in that way crucial in the opposition to apartheid, but also in the transition to a democratic South Africa. As Karl von Holdt argues, workers were 'engaging in fierce struggles to establish new procedures, new notions of fair treatment, and new rights in the workplace' (Von Holdt, 2003, p. 289). The way in which worker demands in the textile industry were formed could differ, though. According to Mager, union organisation in the East London area in the 1980s was characterised by two approaches. The South African Allied Workers union (SAAWU) was more militant, while the Textile Workers Industrial Union (TWIU) was 'less confrontational' and focused on material benefits when seeking recognition in factories (Mager, 1989, p. 59). So, from the 1940s to the 1990s, factories such as Hextex constituted a central arena for both class and racial struggles. But workers' strategies must also be understood in relation to both management strategy and the local political economy, wherein their lives are formed.

A textile factory in Worcester
The Western Cape was not without a textile tradition in the 1940s. For example, one of South Africa's first blanket factories was established in Ceres in 1875. In 1912, that factory was bought by Wolf Harris, who moved part of the production to Cape Town. In Huguenot there was also a blanket factory, owned by textile pioneer Morris Mauerberger. 2 Both Harris and Mauerberger were immigrants from Eastern Europe, who purchased their factories after arriving in South Africa. The establishment of Hextex was different: the owner, Raymond Pollet, arrived in South Africa because he opened a factory.
Pollet came from a French textile family, which had established a factory in Egypt in 1938. Kilvington suggests that South Africa was regarded as a 'safer place for investment than France' after the war (Kilvington, 1996, p. 235). The communist influence in France in the late 1940s must thus be regarded as important for Pollet's decision. However, according to company historiography, his interest in South Africa began when he wanted to produce wool tops for the Egyptian factory. He had contact with a South African business partner, E. R. Keegan, who became vice-chairman at Hextex, and decided on Worcester as the site for the factory. 3 Pollet contributed most of the capital for the factory and also the textile knowledge. He established Hextex by purchasing the entire plant from H. D. Shaw and Company in Bradford and had it reassembled in Worcester. H. D. Shaw was one of the largest English wool textile firms, conducting combing and carding. 4 But Hextex was to produce all types of woollen and worsted cloths and incorporate all production processes from wool washing and combing to spinning, weaving and finally dyeing and finishing. 5 This was more than a relocation of combing and carding from Bradford to Worcester. It was also an expansion towards incorporation of several production processes in one factory. Hextex was thus more than a 'safe' investment, as Kilvington indicates. Capitalists such as Pollet regarded South Africa as a place where visions could be developed.
The decision to locate the factory in Worcester, which was mainly known for agriculture, was a surprise to some. Farmers' Weekly questioned why it would be located so far from the wool farming areas in the Eastern Cape, but also noted that the availability of both labour and water in Worcester made the town suitable. 6 Labour was available also in other parts of the country. However, in Worcester it was much cheaper than in urban centres such as Cape Town, where wages could be twice as high. In woollen production, wages constituted a large share of production costs, while transport only made up a small share. 7 So, the distance to the raw material was not unprofitable as long as suitable labour could compensate for transport costs.
It is vital to emphasise the importance of natural resources such as water for the industry. Lower labour costs in South Africa could not compensate for insufficient water supply. The many arid areas of South Africa were therefore not suited for textile production. But the Hex River was not just a water supply, it was also 'particularly suitable for the process of manufacturing worsted cloth and yarn'. 8 Worcester also had good infrastructure. The railway connected with both Cape Town and its harbour and with the wool farming eastern parts of the country. So, the notion that industries relocate to places where labour is cheaper is obviously a simplification.

Labour force compositionrace, gender and generation
Compared to most other local textile factories, Hextex employed many people. Around 1,300 people were employed by the early 1950s; the industry average was 206 workers in 1955-1956. 9 In the 1950s, there were about 900 coloureds (67 per cent) and 400 blacks (30 per cent) in the factory. 10 The racial composition is noteworthy, especially since there were only about five thousand blacks and twenty-two thousand coloureds in Worcester. 11 Whereas the coloureds were permanent residents according to apartheid legislation, the blacks were not allowed to reside permanently in the Western Cape. So, the blacks were, at least officially, migrants, who had come to Worcester in search of work in manufacturing and were ready to work for the low wages offered. It appears that Hextex had anticipated this composition of the labour force, since the factory was located on the road to the black township Zwelethemba.
Whereas we can be certain that the blacks at Hextex were almost exclusively menthere were hardly any black women in textiles in the 1950sthe coloured workers were mainly women. Du Toit is not specific, but notes that a majority were coloured women (Du Toit, 1978, p. 36). Since 1,300 people were employed in total, we estimate that at least 700 of the 900 coloured workers, or 78 per cent, were women. This was clearly unusual in the South African textile industry in the 1950s. It was not until the 1970s that women came to constitute a majority of the textile labour force. When that occurred, black women replaced the black men, since they could be paid even lower wages (Mager, 1989). The wage difference between men and women was, however, noticeable already in the 1950s in Worcester. Average weekly wages for coloured women were £1.15, compared to £2.6 for coloured and black men. 12 This difference shows that the cheapest labour was provided by the coloured women. The Hextex case thereby adds to our understanding of gender relations in industry. While white women in the Eastern Cape textile factories were replaced by black men, factories in the Western Cape relied on coloured women from the outset. So, Hextex benefitted to some extent from the policies, which turned black men into migrants. But to a greater extent it benefitted from the availability of coloured women locally, who were ready to work for the meagre wages paid at the factory. This suggests that the hiring of black men did not imply a reduction of labour costs, but meant higher labour costs for Hextex. The company could have hired only coloured women and thereby considerably reduced wages. Possibly, there were not enough coloured women in Worcester who were prepared to work in the factory.
The hiring of women, however, initially meant that machines could not be exploited fully. According to the Factories Act (1941), women were not allowed to work nights, which meant that the factory could not work two or three shifts. But Pollet was one of the most eager advocates of this practice. In 1953, the Board of Trade and Industries gave Hextex exemption from the act, enabling a more intense exploitation of machines and labour (Du Toit, 1978, pp. 36-37). It was thus not enough to hire women if they could only work certain times during the day. As Marx notes, when workers walked out of the factory, the machines were in fact 'useless' (Marx, 1990, p. 529). Legislation thus had to be fitted to the requirements of capital in order for investments to be profitable and Pollet was active on the subject.
Another way to reduce wage costs was to hire young workers. Children and youths were frequently employed in farming in South Africa, for example as shepherds ). In factories, child labour has not been widespread, but also refinement of the wool required young people's work. At least eight 'juveniles' participated in the famous 1956 strike. 13 Judging by figures from the strike, juveniles made up 4 per cent of the workers. In total, there could thus have been around fifty juveniles in the factory. It is possible that they were hired as apprentices and performed various duties. But despite their youth they also worked alongside adults in textile production. One girl, Rachel van Wyk, was 14 years old when she started working at Hextex. She remembered that she 'had to stand on a box to reach the machines'. 14 These young workers had never worked in a factory before, indicating that experience was not required to handle the machines and that training was done at the factory.

Importing skillswhite migrant workers
By hiring a large share of women and juveniles, the company could reduce wage costs. Such savings created space for investments in technology and hiring of skilled labour. As Kilvington notes, the machinery implemented in factories such as Hextex required white immigrant workers' expert knowledge (Kilvington, 1996, p. 235). Hextex would certainly not have become a successful enterprise without them, but we must also analyse their role in the context of a racial economy.
In the early years there were productivity problems due to the lack of a 'textile background in Worcester [. . .]'. Technical personnel from the dismantled Shaw plant in Bradford were brought to Worcester to oversee the reassembling of the mill and to train local labour. In 1947, houses had already been built for these technicians and their families in Worcester. But technicians also came from other countries and 'very soon a cosmopolitan team of French, English and Belgian technicians came with their families to settle in Worcester.' 15 Companies that were established in South Africa clearly used housing to attract labour from Europe, but there were other benefits, for example the wages. One observer noted in the 1960s that technicians from Italy could double their wages by moving to South Africa. 16 Such wage levels were apparently necessary in order to establish the industry in the country.
The hiring of skilled workers from abroad signalled the divisions in the working class in South Africa. Whites were privileged and earned relatively high wages while blacks were not allowed to settle down where they wanted and earned low wages. Below this surface, they had more in common, though. Both were exploited through wage labour, and the whites can be seen as part of a migrant labour force. Certainly, they were migrants on different terms than blacks, but nonetheless travelled to find work. In fact, they were the ones who travelled the longest distance to find work. Their relatively high wages and elevated social position has, however, meant that they have not been considered as labour migrants.
Immigrant workers were found also in other sectors, such as the steel industry (Von Holdt, 2003, p. 47). At Hextex they were not only important during the assembling of the machinery, but also in day to day operations. Even twenty years after the factory's establishment, all the technical staff were 'imported personnel'. 17 Two examples are Alfred Gosman and Hubert Dewas. Gosman came to South Africa and joined the company in 1948 as foreman and was by the early 1980s part of the management team. Dewas was hired as chief foreman in 1949 and was involved in the erection of the 'factory plant'. By 1952, he had been promoted to assistant manager of the spinning department. Ten years later he became that department's manager. 18 To make a career within a company had of course been possible in the European context, but for these workers South Africa offered opportunities of quick advancement. The whites were consequently not replaced by machines in factories such as Hextex. Instead, their skills were related to the machines and therefore necessary in order for production to occur. So, workers followed machines around the world and ensured their use to exploit new workers in a new context.

Global capital and local conditions
Hextex became the largest employer in Worcester and management wanted to maintain that position. When another textile company wanted to purchase land in Worcester and establish production, Pollet clarified to the Town Council that it would be 'detrimental to the interests of my Company' since it would use the same 'type of labour' and 'require approximately the same amount of water as has been guaranteed to us'. 19 The competitor company was never established, and Hextex continued as the only textile factory, which helped keep wages down. But local conditions and power structures also impacted wages and labour supply.
Except during harvest time, pushing up agricultural wages, Hextex had no problems finding labour. For the farmers, this competition was, however, a burden and the factory management negotiated with farmers over wage levels. Norman Daniels, representing the textile union TWIU, notes that 'there was an agreement between Hextex and the farmers in the area about wages. The farmers were worried that if Hextex paid workers too much, workers would leave the farms to go to the company.' The arrangement impacted wages in the factory, which Daniels describes as 'very, very low'. 20 Competition for labour thus created an alliance between industrial and agricultural capital. In this sense, Hextex management was forced to take the local political economy into consideration. White farmers, mostly Afrikaners, were the Nationalist Party's main constituency and had political influence. By allying with farmers, Hextex could keep wages down and avoid disturbing local power structures.
In addition to slightly better wages, working conditions in the factory must have been attractive compared with conditions on most farms. Western Cape farms were characterised by harsh working conditions, paternalist relations between white farmers and coloured workers, and by the infamous dop system, wherein part of the wage was paid in alcohol during the working day (Lilja, 2019; Viall, James, & Gerwel, 2011). Considering this alternative, many workers probably favoured the factory despite low wages.
The factory work also attracted people from other parts of the country. Three examples are Ben Baartman, Julius Busa and Joe Ndamoyi. They all came from Molteno in the Eastern Cape, which suggests that personal relations were important in recruitment of migrant workers. For example, Baartman had previously worked in mining around Johannesburg, but came to Hextex because he knew people there. These migrant workers played a role in organising. Ndamoyi was national vicechairman of the ATWIU, and the regional chairman. In Johannesburg, Baartman had been introduced to communism and joined in strikes. Establishment of a large-scale factory thus apparently attracted workers with experience of union activity and with ideas of resistance towards capitalist and racial exploitation. Baartman was described as 'very troublesome' by authorities. He not only led organisation at Hextex, but in other factories and on the farms locally. Attempts to organise on farms were particularly provocative and not welcomed by farmers. Their power rested on paternalist relations with coloured farm workers (Sibeko, & Leeson, 2008, pp. 12-13). But whereas trade union principles did not reach coloured farm workers, the coloured workers at Hextex were more eager to organise. Their action initiated the famous 1956 strike.

Labour unrestthe 1956 strike
The 1950s was a decade of confrontation between capital and the state on the one hand, and workers' organisations on the other. Legislation such as Native Labour (Settlement of Disputes) Act of 1953 forbade black workers to strike. The Industrial Conciliation Act of 1956 further divided the union movement along racial lines. The South African Congress of Trade Unions, where the TWIU and the African TWIU (ATWIU) were affiliated, resisted this legislation, though (Du Toit, 1978). The 1956 strike must be seen in this light. Much like the strike at Amato Textiles in Benoni, where workers were brutally beaten by police, the authorities intervened on behalf of capital also in Worcester. But whereas the Benoni strike was conducted by black workers and was therefore illegal, the Hextex strike was initiated by coloureds and therefore legal. The composition of the labour force thus meant that Hextex management could not initially turn to the authorities for help.
The strike was initiated by the coloured workers in the TWIU when no new wage agreement had been reached. Management then gave notice that all workers who were on strike were dismissed and new ones were to be employed. The black workers were up to that point not participating in the strike. But the management's decision to dismiss the coloured workers for striking led 245 black workers to stop work in solidarity. The black workers were almost immediately arrested by the police, since they were striking unlawfully. All except three of them were released on bail the following day. 21 The arrests did not discourage either the black or the coloured workers. After a week of striking, they gained increased wages, the establishment of an Industrial Council and a promise to establish a sick fund. This was regarded as an important win for the strikers. 22 The main legacy of the strike was, however, the racial unity between coloured and black workers. The action taken by the black workers was consequently regarded as crucial by the leftist newspaper New Age, which argued that '[i]t was only when the Africans came out that the bosses were forced to give in and accept the union's conditions for a return to work.' 23 One factor that contributed to racial unity among workers in Worcester was, according to Archie Sibeko, language. The coloured workers spoke Afrikaans and 'most' of the black workers also knew Afrikaans from working on farms in other parts of the country where the language dominated (Sibeko, & Leeson, 2008, p. 99). The notion of racial unity was, however, only partial. About forty per cent of black workers at Hextex did not participate in the strike. Possibly they were not members of the union, but they were in any case not prepared to challenge white authority in the form of the factory management and the state.
Although the strike was regarded as a victory by the non-racial labour movement, the role played by TWIU and ATWIU was questioned many years later by some workers. For example, Rita Fredericks noted that the workers mainly 'waited for the organiser to come out of a meeting with the bosses'. After that, 'we were told to go back to work by the union, and really I feel we lost that strike'. 24 For many, the strike also resulted in hardship rather than victory. One black inhabitant in Worcester stated that 'since the Worcester textile strike, our people here have been hounded and persecuted'. Besides charges for striking and holding meetings unlawfully, some ten families had been evicted from their homes by the municipality for not 'paying rent regularly'. 25 Baartman, one of the strike leaders, was also banished from the Western Cape and forced to go to Zululand (Du Toit, 1978, p. 100).

Investments, takeovers and apartheid
The post-strike years constituted a new phase for the factory. Labour militancy declined and capital made progress through a 'modernisation' programme. This meant investment in new machinery throughout the factory, from scouring to spinning, weaving and finishing. The machinery used since 1946 had mainly been bought second hand from England. That machinery appears to have worked well, but the company policy was to take 'advantage of every latest technological breakthrough'. The investments in new machinery were successful. Hextex made record profits and launched an expansion programme in the mid-1960s. SA Textiles concluded that 're-equipment' had placed the factory among the top 12 establishments of its kind in the world. 26 New, automatic, machinery was recommended by textile experts, since it could be operated by workers with little training and workers could more easily be replaced. The advantages of, for example, automatic looms were considerable. The number of looms per weaver could be increased from two to six on worsted suiting and from one to four on woollen cloths. 27 The relocation of machinery, which characterised the initial phase of textiles in South Africa lasted only about a decade. The old machinery was used to set up the industry in a new context and generated profits, resulting in capital accumulation. But for accumulation to continue, profits had to be invested in production. So, companies that were successful did not just transfer machinery from Europe to South Africa in order to exploit new labour. They also installed new machinery suited to the skills of the labour force. These investments also generated long-term profits: a washing machine installed in 1960 was still in use more than thirty-five years later. 28 Although Hextex was successful, there were competitors with more financial strength. The entire industry expanded in the 1960s, which sometimes resulted in expansion at the expense of others, a common theme in capitalism. In 1965, Hextex was bought by Felt and Textiles, a South African subsidiary of an Australian company, which had ties to the IDC. Felt and Textiles acquired both Hextex and Fine Wool Products in Uitenhage. 29 Through these acquisitions it became a major actor in the industry, employing nine thousand people in 15 factories throughout the country in 1967. 30 Along with the new machinery and expansion programme, Hextex hired more workers. In the 1950s, there were about 50 whites (4%), 900 coloureds (67%) and 400 blacks (30%). By the early 1980s, the labour force had increased to 1,706 people employed. The racial composition had also changed. In 1982, there were 127 whites (7%), 1,398 coloureds (82%) and 181 blacks (11%). 31 The coloureds had thus replaced many black workers and the whites had also increased. Some whites were in administrative positions, but also in positions requiring high levels of technical skill. For example, a laboratory had been established in the factory by the early 1980s. 32 The increased share of coloureds can be explained by legislation. The coloured labour preference policy favoured their employment in the Western Cape. But new legislation in the 1960s, the Physical Planning Act, stipulated that factory extensions could not be done by increasing the number of black workers unless it was located in the homelands. 33 So, Hextex did not rely on black migrant workers, but continued to recruit coloured workers.

Monopoly capitalism and retrenchments
Labour relations in South Africa slowly began to change after the 1973 Durban strikes when workers protested in the streets against low wages and apartheid. The textile industry was affected, not least because the giant Frame Textiles had its centre outside Durban. Southern Africa Textiles downplayed the strikes' impact, but still recommended textile companies to increase wages. 34 On a wider scale, the Durban strikes, and the 1976 Soweto uprisings, created a new social climate and generated a state response, notably through the Wiehahn reforms in 1980, which allowed black unions. As von Holdt notes, the reforms were essential for the 'nascent union movement', which became successful through workplace organisation (Von Holdt, 2003, p. 85).
But the era of union success was also one of crisis and transformation of the labour market. In the textile industry, the number of workers was reduced from 113,700 in 1981 to 67,800 in 1994. 35 While South Africa moved towards democracy, workers were increasingly unemployed and in insecure economic positions. The labour market transformation was also related to an overall consolidation of capital, where Hextex found itself at the centre. Felt and Textiles merged with another company in the mid-1970s and changed name to Romatex. Thereafter it was taken over by the industrial and mining giant Barlow Rand. By the 1980s, Hextex was consequently owned by one of the largest conglomerates in Southern Africa.
The new ownership placed the factory in a more secure position when the industry experienced setbacks in the early 1980s. But it also meant that Hextex was subject to Romatex' labour strategy. By 1978, Romatex had decreased the group's workforce by about two thousand workers. The reduction was 'due both to improved efficiencies and to reductions arising from further rationalisation'. 36 In the early 1980s, such rationalisation within the group impacted Hextex, when the Romatex weaving section became part of the Worcester factory. Some months after this expansion of the factory, 175 workers were retrenched and more were expected 'before Christmas'. 37 The 1,706 workers had been reduced to 1,360 by the end of 1982. 38 So, successful companies such as Romatex found ways forward by restructuring both production and the labour force.
Restructuring of production and rationalisation indicates that power over the production process had become more firmly entrenched in management hands. This strategy was expressed by the chairman of Hextex, Wilson, when he clarified that a 'superb team' of managers would deal with the challenges of the future. He did not specify the team's strategy, but it included massive investments. Between 1984 and 1996, more than R50 million was invested in machinery and technology. 39 The strategy included retrenchments as well, which also affected the other Romatex worsted mill, Veldspun in Uitenhage. According to the Romatex chairman, 'it was necessary to retrench 500 employees and both mills are currently working short-time.' 40 By the late 1980s, Hextex employees had been decimated to around one thousand and were replaced with new, modern machinery.
Hextex workers to the fore again While organisation at the factory level was a prerequisite for worker resistance to capital, the emergence of conglomerates such as Romatex and Barlow Rand created new challenges, but also opportunities for unions. The textile unions regarded Romatex as an enemy of the working class, not least because of its many retrenchments. In 1987, various textile unions merged into the Amalgamated Clothing and Textiles Workers Union of South Africa (ACTWUSA). ACTWUSA was affiliated with the non-racial federation COSATU and meant a more united front for textile workers. In 1988, ACTWUSA called a national council where shop stewards from Romatex factories met to discuss how to act against Romatex. A crucial observation was that workers in Romatex factories were 'isolated from each other' and that there was little knowledge about conditions in other factories in the group. 41 As a member of the Romatex group, Hextex was in the middle of a larger capital-labour struggle. In 1989, that struggle was highly visible at the factory when workers went on strike after negotiations over a new wage agreement broke down. 42 After the 1956 strike, worker militancy declined. However, the large share of coloured workers at the factory contributed to a relatively stable labour force and there was a core of workers who had experienced the 1956 strike. In 1982, 111 workers had been employed longer than 20 years and 32 people had been employed longer than 30 years. 43 For example, Rita Fredericks and Rachel van Wyk began working at Hextex in 1949 and 1953 respectively and were still employed in 1989. 44 These women stood for continuity in a workplace where many workers came and went frequently and must have played a role in revitalising strong union structures at the factory. Van Wyk was known as 'Mama Actwusa', indicating that the old workers were central in organising.
During the 1989 strike, workers in other parts of the Romatex group supported the Hextex workers, meaning that management could not use other factories to replace Hextex as they wished. At Veldspun in Uitenhage, workers told management that they would 'not handle work from Hextex'. 45 The consolidation of capital in the textile industry not only strengthened capital but also brought workers in different factories together and gave them opportunities to show solidarity.

Economic and non-economic issues
As Webster notes for the foundries, unionisation was a struggle against racial capitalism. This was seen in the Hextex strike to some extent. Workers gained May 1 st and June 16 ththe day of the Soweto uprisingsas paid holidays, connecting the workers' struggle with the anti-apartheid movement. 46 Hextex workers' main demands were, however, workplace-related, specifically focusing better wages and a service bonus. Wages at Hextex were comparatively low, which explains this focus. Thelma Thomas stated that 'the demand for a service bonus goes very deep with us. For years we sweated. For years we worked without a decent reward for our long service. Now the past is catching up with management.' 47 One reason for the many years without decent wages was government policies. In 1956, those issues were at the very centre of the strike, but in 1989 these were hardly mentioned by strikers. While ACTWUSA as a union had political aims, these were held back at the factory level in favour economic issues.
The strike lasted 47 days and the workers considered it a victory: they gained a service bonus and a R25 per week wage increase. 48 The R25 meant that wages at the factory were more in line with the average for other Romatex factories, indicating a successful cooperation between shop stewards in the group. The strike was also used to further build the union structure at the factory and throughout Romatex. The new agreement guaranteed workers paid time off from work for shop steward's training.
The strategy used at Hextex resembled the TWIU strategy used in East London factories. As Mager describes, the TWIU was 'less confrontational' and negotiated for maternity benefits. This same pattern was seen at Hextex. Workers gained 12 weeks paid maternity leave at 32 per cent of the basic wage. 49 This signals that women had become more important in the union. As Seidman notes, unpaid maternity leave was a major problem for women forced to leave employment to take care of children. In the 1980s, female workers pressed COSATU to include such demands with political ones. 50 But in order to understand union strategy it is important to note the active role of capital, especially after the Wiehahn reforms. Romatex policy was to include unions in negotiations and to communicate with them. In the early 1980s, the company expected that the 'experience being gained by both unions and management should lead to improved understanding in the future'. 51 It would thus be beneficial if relationships with unions were positive and constructive. The Romatex policy to discuss with union representatives could in that sense have impacted how workers' demands were formed. It is also important to consider the situation in textiles during the late 1980s. Retrenchments were commonplace and Romatex was known to dismiss workers as part of rationalisation programmes and tried to move production from Worcester to Uitenhage during the strike. From the union viewpoint it was therefore equally important to ensure that the Hextex factory and work opportunities stayed in Worcester.
The strategy was also to include workers in decision making. Some workers who had been involved in the 1956 strike pointed out that there was a big difference in the sense that the union was not making decisions for workers. The 1989 strike was initiated through a ballot where a majority voted for a strike and, according to the union press, decisions were built on the participation of members and not on officials. Nevertheless, Thelma Thomas, senior shop steward, clarified that 'our official, Ebrahim [Patel], helped us all the way. He helped us to stay united, even when things became very tough.' 52 A strong union leadership was thus important for the strike's success and for creating a sense of unity among the workers. ACTWUSA's leadership was constituted by men such as Patel, who later received a ministerial post in the ANC government. But at the factory level, women such as Rachel van Wyk and Thelma Thomas linked to the tradition of worker resistance from the 1950s. Unlike the black women elsewhere in the country who were recently proletarianised, Hextex women were active union leaders and very much involved in creating a working class consciousness among fellow workers. The 1989 strikewhile not as famous as the one in 1956was perhaps more of a gain for the workers at Hextex since it strengthened union structures and gave women a more central role.

Conclusions
The establishment of Hextex in Worcester was part of the relocation of capital from core to periphery of the world economy, which characterised the global textile industry since the interwar period. Lower labour costs were important for establishment of the factory, since wages were only half of those found in urban areas such as Cape Town. But Hextex also depended on the suitable water in the Hex River, and factory management tried to keep competitors from gaining access to both water and labour resources. The majority of the workers were coloured women who could begin working with machines at fourteen years old. But these inexperienced workers could not singlehandedly create foundations for production. Relocation of machinery thus also required relocation of skilled workers from Europe, some of whom could advance quickly in the company. While whites were replaced by machinery in some parts of industry, companies such as Hextex depended on them to be able to exploit local labour.
The workers who were drawn into capitalist production also came to resist exploitation, seen most notably in the 1956 and 1989 strikes. The militant spirit in the 1950s was much due to the influence of black migrant workers such as Ben Baartman and Joe Ndamoyi, who were also ANC activists. But the re-emergence of strong union structures in the late 1980s was instead due to coloured women such as Rachel 'Mama Actwusa' van Wyk, who functioned as a link to the militant past.
The union strategy in the 1980s can be characterised as non-confrontational. It was focused on material benefits for workers, including maternity leave for women, but mentioned nothing about the political aspect of exploitation. While this was a conscious strategy, it must also be seen in the light of the economic situation in the industry during the late 1980s when retrenchments were commonplace and factories were beginning to close; it was clearly risky for the union to make demands which could be seen as too expensive or too radical for the management. But union strategy also resounded with the Romatex group policy of communicating with unions, implemented in the early 1980s. The purpose of the Romatex policy was to separate the economic from the political struggle and show that the union and the company had a mutual interest in saving jobs in South Africa. Hextex workers could surely support this policy, at least in the short term. Political freedom would definitely be worth less if there was no work.