1 Labour Regimes: A Comparative History

At first, UMHK and the small mines that emerged in Northern Rhodesia before the First World War drew inspiration from South African mining and established a migrant labour system to recruit African workers (Henderson 1972; Perrings 1979; Kayamba 1986; Higginson 1989; Seibert 2011). By the early 1920s, however, it became clear to UMHK that migrant labour was incompatible with its industrial development plans. While these plans required a larger and more disciplined workforce, the company was confronted with a severe labour shortage because of the growing competition from other colonial companies in both the Belgian Congo and Northern Rhodesia. It is in these circumstances that UMHK launched a policy aimed at ‘stabilising’ its workforce near production sites in the copperbelt. This policy led to the company’s withdrawal from the migrant labour regime characteristic of the early twentieth century and to the formation of a new labour regime.

To overcome the labour shortage, UMHK was authorised by the Belgian colonial administration to organise its own recruitment missions in other provinces of the Belgian Congo and Ruanda-Urundi and to extend work contracts to three years (Van Nitsen 1933; Bakajika 1993). At the same time, it encouraged workers to come to the mines with their wives and children (Dibwe 2001). It was expected that the presence of their families would encourage workers to stay for longer periods, while also enabling the company to supervise the growth and education of their offspring. In worker camps, brick houses were built for married workers and collective facilities such as nurseries, canteens and schools were created to take care of children. The management of these facilities, and the moral guidance of workers’ families more generally, was entrusted to Belgian Catholic orders, which came to play a key role in the daily social life of camps. In the aftermath of the Great Depression, the stabilisation policy came to be viewed as a success (Mottoulle 1946; Toussaint 1956). The challenge from UMHK’s perspective was no longer an issue of increasing the workforce but of making it stronger, more disciplined and ‘civilised’.

To name the labour regime that UMHK gradually set up from the 1920s onwards, we can use the concept of industrial paternalism, a term coined by historians to characterise the type of social policy that large firms in various sectors – but especially in mining and metallurgy – developed during 1870–1940 (Reid 1985; Noiriel 1988; Zahavi 1988; Tone 1997; see Rubbers 2013). The social policy of UMHK shows the main features of industrial paternalism:

The conditions that led to the emergence of this labour regime in the Congo were similar to those characterising the origin of industrial paternalism in Europe and North America. First, it was a response to a shortage of labour in a period marked by the modernisation of technology, the mechanisation of production, the construction of new processing plants, and the expansion of the company’s activities. UMHK went through two such periods, in the 1920s and after the Second World War (UMHK 1956; Vellut 1981; Brion and Moreau 2006). Each time, industrial development programmes required a larger and better trained workforce.

At the same time, UMHK’s paternalism was also a response to workers’ actions and demands. In 1919–20, white English-speaking workers went on strike to claim higher wages, job reservation for whites, and recognition of white workers’ unions (Poupart 1960). The company immediately responded by not renewing strikers’ contracts, and replacing them either with fresh recruits from Belgium or experienced African workers. One year after this strike, UMHK established its first vocational school for Africans (Mottoulle 1934). From a more general perspective, the stabilisation programme was envisaged as a means to replace white English-speaking workers with Africans in both unskilled and semi-skilled positions. Such a policy allowed the company to save costs, undermine white unionism, and draw attention to its contribution to the colony’s civilising mission.2 Although white workers, who were mostly Belgian by that time, succeeded in establishing a trade union during the Second World War, the strike actions they organised were systematically repressed by layoffs and convictions of leaders (Corneille 1945).

Although stabilisation gave African workers new advancement opportunities to the detriment of white unskilled and semi-skilled workers, they did not earn the right to strike or unionise. In 1941 the most important strike in UMHK history took place (Perrings 1979; Higginson 1989; Dibwe 2001; Seibert 2013). The African workers demanding a raise were immediately successful, but the strike ended with the shooting of at least twenty workers by soldiers during a rally. This massacre seems to have discouraged the Congolese workforce from taking further action for fear of repression; unions for African workers were not introduced at UMHK until after Congo’s independence in 1960. In 1946, however, a legislative ordinance from the Belgian colonial administration forced UMHK to create Conseils indigènes d’entreprise, that is, bodies representing African workers (Dibwe 2001: 67–85). Its members, elected by the workers, could express grievances that the employer would take into consideration if they were deemed reasonable. These councils were at the root of several improvements in workers’ living conditions, even though the company would later claim credit for these improvements as products of its own social policy.

Finally, UMHK’s industrial paternalism was a response to the colonial state’s social policies. In the aftermath of the Second World War, laws regulating workers’ rights, family life and urban planning multiplied, and the colonial administration conceived a ten-year development plan that involved the building of new houses, hospitals, schools and community centres (Vanthemsche 1994). As part of these developments, UMHK not only complied with new social regulations, but sought to remain a step ahead of colonial policy by creating new social infrastructures in the camps or by funding large community development programmes. Its ambition was to create model company towns to win the sympathy of the public and prevent further state interference.

The concept of industrial paternalism can also be used to characterise the labour regime put in place by mining companies in Northern Rhodesia, though its introduction was more halting and reluctant. In several respects, the description Powdermaker (1962: 5–6) gives of Luanshya’s RST mine township in 1953–54 could be that of a UMHK camp at the same period. At the entrance to the camp were the township offices. Behind were rows of rectangular white houses. The most common type of house comprised two rooms with a small adjoining kitchen, but there was also a growing number of larger houses with indoor sanitation, electricity and a flower garden, for senior employees. Each section of the township was equipped with public latrines, shower stalls and a washhouse where women gathered to do the laundry. Finally, the township was equipped with various facilities such as a hospital, schools, an open-air market, a sports ground, a welfare centre and a cinema. From the outset, however, a number of differences also appear from the UMHK model. Powdermaker mentions the presence of various churches and large union meetings at the theatre, which UMHK would not have tolerated in its own workers’ camps.

To study these similarities and differences in more detail, it is necessary to return to the conditions that prevailed during the emergence of the mining industry in Northern Rhodesia. In the 1920s, going to work in the Belgian Congo, Southern Rhodesia, or South Africa was already common practice among the people of Northern Rhodesia, so they could compare the conditions of employment in these different colonial territories. The main competitor of the Northern Rhodesian mining companies at the time was UMHK, which was experimenting with its new stabilisation programme. Faced with a shortage of labour, RST and RAA were thus forced to offer African workers in Northern Rhodesia similar conditions in order to attract them – that is, to pay more or less equivalent wages to those offered by UMHK, to allow them to come with their wives and children, and to build family homes to house them (Henderson 1972; Parpart 1983). One can see in these measures the sketch of a stabilisation policy that breaks with the migrant labour system characteristic of southern Africa. However, the implementation of such policies was more hesitant in Northern Rhodesia than in the Belgian Congo.

Although mining companies in the Northern Rhodesian copperbelt faced labour shortages in the late 1920s, these were soon eased by wider economic changes. Though the Northern Rhodesian government was reluctant to place strict controls on foreign labour recruitment within the territory, as UMHK looked elsewhere for its labour, recruitment in Northern Rhodesia for the Katangese mines ceased in 1931 (Parpart 1983: 32–4; Perrings 1979: 90–8). The Great Depression and fall in the copper price caused an avalanche of layoffs on both sides of the border and a sharp reduction in employment opportunities for African workers. Thus, as production began in the Northern Rhodesian copperbelt in the 1930s, the mining companies had little difficulty finding experienced workers. Now in a strong position, they could hire migrant workers and limit their efforts to improve living conditions in the mining townships. Under these conditions, the companies’ preferred option was to employ cheap migrant labour and to transfer the cost of its social reproduction onto the rural areas (Henderson 1972; Parpart 1983).

Moreover, unlike UMHK, companies in Northern Rhodesia predominantly operated underground mines, which required more labour than open pit mines. To make profits, they had to keep an eye on labour costs (Perrings 1979). It was only after the copper price rose again in the late 1930s that mining companies increased spending on workers. There were also differences between the two groups. RAA controlled the Rhokana Corporation, whose directors had made their careers in the South African mines and were slower to move away from a migrant labour model. The directors of RST, meanwhile, were predominantly American and were more inclined to continue with the limited stabilisation policy initiated before the 1930 crisis, also because the geology of the orebody required more skilled labour (Parpart 1983: 35; Perrings 1979: 111).

The political context in Northern Rhodesia also meant that, unlike UMHK, RST and RAA failed to prevent the rise of white mineworkers’ power. The Northern Rhodesia Mine Workers’ Union (NRMWU) was created in 1936 and successfully lobbied London to instruct the mining companies to recognise it (Money 2016: 85–87). During the Second World War, the union managed to convince the companies to introduce a de facto colour bar, achieved by the requirement in the union recognition agreement that ‘daily-paid employees had to be union members’ and that ‘union membership was restricted to Europeans’ (Money 2016: 129). Because of this closed shop, the companies were unable to promote a policy of African advancement following the example of UMHK until the late 1950s.

Finally, although the British colonial administration took some measures to create a migrant labour market for mining companies in Northern Rhodesia, it was less favourable to stabilisation than the Belgian colonial administration. The brief existence of mining projects before the First World War, followed by mass layoffs in the aftermath of the Great Depression, had shown that employment in the mines was by definition unstable (MacMillan 2012). Under such conditions, the colonial administration feared that a stabilisation policy would generate a large ‘detribalised’ population in mining towns, both cut from its rural roots and dependent on wage work: if copper prices fell, this proletariat would be unemployed, and the situation in mining towns would become explosive.3 It was to address these fears of detribalisation and disorder that the RLI was created in 1937 (Crehan 1997; Schumaker 2001). However, although they continued to take the idea that Africans in rural areas belonged to ‘tribes’, Gluckman and his colleagues questioned the assumption that they lost this ‘tribal’ identity by migrating to colonial towns. Their work argued, on the contrary, that tribal identities were reinvented in urban situations (Mitchell 1956; Epstein 1958; Gluckman 1960). Colonial officials and white settlers considered it wiser to continue the migrant labour system.

For these reasons, and for a long time, the length of service at RST and RAA was shorter, and the percentage of married workers lower, than at UMHK (Parpart 1983; Henderson 1972; Perrings 1979; Juif and Frankema 2018). The situation gradually changed from the end of the 1930s when the Northern Rhodesian companies increased production, which required a larger and more experienced workforce, but the commitment of companies to investment in the infrastructure needed for their stabilisation policy was limited (Parpart 1983; Kalusa 1993). It was really only in the aftermath of the Second World War that their labour policies converged with that of UMHK to take the form of a paternalistic labour regime – at a time when UMHK itself was focusing on further mechanisation and stabilisation tactics.

In contrast to the Belgian Congo, repeated strike and protest action by African workers played a significant role in the emergence of industrial paternalism in Northern Rhodesia. African workers called a strike in 1935 and then again, in a more organised fashion, in 1940. This won them concessions in some mines but in the end was brutally repressed. After the war, the British colonial administration decided – like the Belgian administration – to authorise the creation of trade unions for African workers (Berger 1974). Unlike in the Belgian Congo, however, this decision was followed by concrete measures, and in 1949 the African Mineworkers’ Union was founded. It quickly gained a large base of support among workers and organised several protest actions demanding wage increases and improvements to the mine townships. The rise of union activism encouraged mining companies to develop a paternalistic labour policy similar to that of UMHK on the other side of the border (Parpart 1983: chap. 7): an end to job reservation for Whites to allow the advancement of skilled African workers; new, larger houses in the townships; ambitious welfare programmes; finally, they tightened control over workers by controlling access to the townships and paying company spies among the inhabitants. The companies aimed to increase, for the workers, both the benefits of being loyal and the cost of being critical.

Thus, despite a late convergence, industrial paternalism provides a useful conceptual frame through which to understand the labour regime that mining companies in both copperbelts established between 1925 and 1965. As we have seen, the conditions in which this labour regime was formed differed on both sides of the border. In the 1950s, it was commonplace in British colonial policy circles to say that UMHK had implemented a policy of stabilisation while RST and RAA had opted for stabilisation without urbanisation (see, for instance, Prain 1956). However, although mining companies in Northern Rhodesia were more reluctant to stabilise their workforce, this distinction is based on a misunderstanding of UMHK’s labour policy, which never aimed to cut its workforce from their rural roots. On the contrary, UMHK took measures to encourage workers to stay in contact with their rural kin and to retire to their villages at the end of their working lives.

What distinguished UMHK from RST and RAA is the amount of money that it invested in its stabilisation policy and the degree of support it received from the colonial administration and the Catholic missions (Vellut 1983). In comparison, the mining companies of Northern Rhodesia not only exerted less control over their workers (whether for marriage, religious practices, or forming associations),4 This contrast is also reflected in the wages of African mineworkers (Juif and Frankema 2018). While wages rose in parallel in both copperbelts – in the late colonial period they were the best-paid workers in Africa – RST and RAA paid a larger portion in cash relative to UMHK, which provided a larger portion as benefits in kind. they were also unable to stifle the rise of first white, and then black, trade unionism with the same force and intransigence.

This nationalisation of colonial mining companies led to the emergence of a new labour regime that we will call ‘state paternalism’.1 State paternalism is to be distinguished from the social or welfare state in that the social benefits are reserved for the employees of state-owned enterprises; they are not part of a social security system extended to all citizens. Enjoying such benefits depends on having a job in such enterprises, not on obtaining citizenship. This labour regime kept some of the characteristic features of industrial paternalism in the colonial era:

What distinguishes state paternalism from industrial paternalism is less the takeover of production by the state than the enlarged role of the state in industrial relations and the management of mineworkers.2 Management decisions regarding mining and processing operations remained in the hands of expatriate executives until the early 1980s. In 1967, the Congolese government merged existing trade unions in the mining sector into the overarching Union Nationale des Travailleurs Zaïrois (UNTZ), which became a fully fledged body of the Mobutu regime. In Zambia, meanwhile, the newly independent government created the Zambian Congress of Trade Unions (ZCTU), which, after some resistance, the Zambian Mineworkers’ Union joined in 1966. The government then oversaw the creation of the new Mineworkers’ Union of Zambia (MUZ) in 1967 and, with the passage of the 1971 Industrial Relations Act, sought greater control over unions through ZCTU (Larmer 2005). Despite the different political trajectories and differing intensity of control, the role assigned to unions was to serve as channels of communication and support for government policies: unions were supposed to communicate workers’ demands during official negotiations, and defend them in individual labour disputes; in return, they were expected to prevent wildcat strikes by explaining to workers the decisions taken by the company and the state for their well-being and the development of the nation. In other words, trade unionists were enjoined to act ‘responsibly’ by ‘educating’ workers on both industrial and party discipline, which constituted altogether a single national imperative.

In both countries, the control exercised on workers through trade unions was supplemented by the introduction of various political bodies in mining companies. After Congo became independent the surveillance regime that UMHK had put in place in workers’ camps had relaxed: no more curfews at night, no more guards at the entrance of camps, no more company spies, and no more inspection visits to workers’ houses by social workers. Congolese workers were free to go in town at night, to have extramarital relationships, and to house relatives at home. With the rise of the Mobutu regime, however, the state began exerting tighter control over workers. A new administrative structure and political bodies were established in the camps to monitor and mobilise the workers. This control was aimed less at imposing a certain family model, or at isolating workers from the rest of society, than at creating a climate of suspicion, denunciation and fear. The establishment of this regime contributed to silencing workers’ critiques and nipping wildcat strikes in the bud. The workers who were too vocal in their criticism of the company were quickly denounced, surveilled and repressed by security forces (Dibwe 2001: 108).

In comparison, the power exerted by the Kaunda regime in Zambia was less total, arbitrary and brutal. The government co-opted national-level union leaders, most particularly those of MUZ, de facto banned strikes, and sanctioned or arrested workers who organised protest actions. Although such measures led to a marked decline in strike action, they did not succeed in breaking mineworkers’ militancy (Rakner 1992; Larmer 2007). Workers and their families continued to attend mass meetings organised by the union in mine townships to express their demands and covertly organised localised strikes. When mineworkers’ real wages began to decline in the late 1970s, union leaders became more critical of the government and in 1981, strikes occurred at an industry-wide level (Larmer 2006). Over time, trade union leaders, most especially of ZCTU, became de facto opposition leaders with strong support in Copperbelt Province. Support from mineworkers was critical for the rise of the multi-party movement which brought the Movement for Multiparty Democracy (MMD) to power in 1991, when trade unionist Frederick Chiluba became president. In contrast to the situation in Congo, the mining industry and the state in Zambia did not succeed in containing workers’ militancy.

Under the control of the postcolonial state, mining companies’ paternalism was generally more marked by the new political leaders’ nationalist projects. The first measure Mobutu and Kaunda took after nationalising the mines was to proceed with nationalising their executive employees. Long-term programmes were adopted to hire and promote national workers in supervisory positions and gradually replace the expatriate staff. Although these programmes took centre stage in the communication of mining companies, in practice their implementation was slow. From 1969 to 1976, the proportion of ‘Europeans’ in the total workforce only decreased from 5.9 (1,458/24,416) to 3.5 per cent (1,270/35,489) in Gécamines (Gécamines 1969–99), and from 10.8 (4,727/43,500) to 7.6 per cent (4,060/53,082) in the Zambian mining industry (Money 2016: 368–9). Some Zairian and Zambian engineers were appointed to top managerial positions in the 1970s, but they were surrounded by experienced expatriate technicians who were responsible for production. It was not until the 1980s that national managers were granted access to strategic positions in production and the Africanisation of the executive personnel speeded up.

Independent of the nationalisation of executive personnel, the mining companies’ labour policy as a whole was justified by the greater interest taken by the new nation-states. In the second half of the 1960s, the mining industry in both countries was portrayed as spearheading national development. To increase production and turn the former colonies into modern nations, the two governments claimed to give particular importance to the human factor, that is, the advancement and the well-being of African workers, and characterised strikes as a form of betrayal. Henceforth, the rites and symbols organised by mining companies were to arouse workers’ attachment not only to the corporate community, but more broadly to the nation and the charismatic figure of its leader, the president of the republic. This nationalist discourse was used to impose sacrifices onto the workers and crush wildcat strikes as seditious acts.

The newly independent states of the Democratic Republic of Congo (known as Zaire between 1971 and 1997) and Zambia benefited from high copper prices between 1965 and 1974. In this context, the recently nationalised mining companies did not encounter too many difficulties in pursuing and extending the paternalistic policy inherited from the colonial era and, at the same time, meeting workers’ new expectations for advancement and improved living conditions. Under pressure from governments, the companies also had to take charge of public infrastructure and finance national development programmes. From 1974, however, the mining industry was faced with a series of challenges that would eventually lead to bankruptcies. In 1973, the oil shock caused an increase in production costs, a decline in copper prices, and a shortage of foreign exchange. From 1973 to 1978, the border was closed between Zambia and Rhodesia, blocking the main route through which Zambia’s imports and exports were transported, and in 1975, the start of the Angolan civil war closed the railway to Lobito, the main route for copper exports to Europe and North America. In 1977 and 1978, the wars of Shaba 1 and 2 disrupted Gécamines’ development plan for the mining and industrial complex of Kolwezi. Then, in 1979, the United States Federal Reserve decided to raise interest rates and therefore the price of the dollar. For the mining companies of the Central African Copperbelt, and the economy of Zambia and Zaire more generally, this resulted in increasing import costs and an uncontrollable surge of debt taken on to finance national development plans.

To overcome the situation, the World Bank and the International Monetary Fund imposed structural adjustment plans which involved restructuring the mining industry. In Zambia, the two mining companies merged into Zambia Consolidated Copper Mines (ZCCM) in 1982. In Zaire, Gécamines was reorganised into a large holding company. The aim of these organisational reforms was to increase production, cut costs, and reduce the misappropriation and misuse of mining revenues by the regimes. Their effects were nevertheless limited: measures taken by the Zambian government to cut labour costs came up against the resistance of mineworkers, and borrowing by ZCCM funded non-mining-related projects and political expenses (Larmer 2006). Meanwhile, the appointment of expatriates to head Gécamines on the advice of the World Bank did not prevent Mobutu from embezzling money from the State-Owned Enterprise (SOE). As the copper price on international markets continued to fall, the financial situation of the mining industry became increasingly desperate.

This series of adverse economic conditions progressively undermined the paternalistic labour policies that the Zambian and Zairian mining SOEs had developed in the late 1960s and early 1970s. From 1974 to 1989, mineworkers on both sides of the border faced a continuous decline in real wages, which pushed families, and most particularly women, to develop various informal activities. In comparison to the rest of the population, however, they remained a relatively protected social category as they continued to receive various social benefits, including a house, free access to school and hospitals and, more and more importantly, food. In Zambia, ZCCM sold mealie meal (maize flour) to its workers at a subsidised price. In Congo, managers could buy subsidised food while workers received family food rations every month. So, even though it was less and less generous, the paternalism of mining SOEs helped to protect their workers from some of the consequences of the crisis in the 1970s and structural adjustment programmes in the 1980s (Mususa 2014).

In the early 1990s, under pressure from opposition movements which had steadily developed in the 1980s, the two regimes resolved to open Zambia and Congo to multipartyism. Contrary to expectations, this political opening did not inaugurate a new period of prosperity. In both countries, the 1990s were marked by new austerity measures, rampant inflation, and a chronic shortage of staple food. A severe drought in 1991 to 1992 severely worsened existing food shortages in Zambia (Mulwanda 1995). In Zaire, meanwhile, the political situation rapidly became chaotic: there was large-scale urban looting in 1991 and 1993, xenophobic violence in Katanga in 1992–93, and an armed rebellion in 1996–97. Supported by Rwanda and Uganda, this rebellion resulted in Laurent-Désiré Kabila’s rise to power in 1997, and a regional war which began in 1998 and continued until 2002.

To revive the economy in both countries, international financial institutions recommended privatising SOEs and creating favourable conditions for attracting foreign investors. Mining companies, which were running large deficits at that time, were particular targets of this development strategy (Serageldin 1992). Fearing to see their mining resources sold off to foreign interests, political leaders in both countries hesitated to follow this recommendation. In Zambia, it was not until 1997 that the government decided to dismantle ZCCM and to sell its assets to foreign investors; the controversial process took another three years to come to an end. Seven asset packages were created, with the government only retaining small minority interests (10–20 per cent) in companies through ZCCM-Investment Holdings (ZCCM-IH) (Craig 2001; Fraser and Lungu 2007).

In Zaire, although the first partnership agreements that Gécamines signed with foreign investors date back to 1994–96, most were renegotiated by the new Kabila regime in 1997 before being suspended by foreign investors at the start of the 1998–2002 war. During this war, the Kabila government signed new agreements with white businessmen close to Zimbabwean President Mugabe in exchange for military support from Zimbabwe. But these agreements covered nothing more than operating Gécamines’ mines and plants. It was only after the World Bank returned to the country in the early 2000s that a new mining code aiming at liberalising the mining sector and attracting foreign investors was promulgated.

Contrary to what happened in Zambia, however, the reform of the Congolese mining sector did not involve the demise of Gécamines. A recovery plan refocusing on a limited number of mining and industrial assets and making cuts to the workforce was adopted (Rubbers 2013; Carter Center 2017). To facilitate this plan, the 2002 mining code allowed Gécamines to keep rights over its assets and to decide whether it would operate them, sell them, or transfer them to joint ventures with foreign investors. This arrangement was instrumentalised by the ruling elite to make Gécamines a gatekeeper of Congolese mining resources and thereby acquire personal advantage from contracts signed with foreign investors. Gécamines’ most promising assets were sold to foreign investors for developing joint venture projects with the SOE as a minority shareholder. As a result, although Gécamines has not disappeared, it is less and less a mining company strictly speaking; independent of the projects in which it is involved as a partner, it produces less than 20,000 tonnes of copper per annum. It has become more like a chartered company, a sovereign instrument that the ruling elite uses to extract revenues from foreign investors in the mining sector.

The mineworkers of ZCCM and Gécamines were the first victims of the restructuring of the mining sector. In the early 1990s, they faced rapid deterioration in their living conditions due to rising consumer prices, disruptions in food supplies and the withdrawal of food subsidies. As their wages had become insufficient, workers and their families relied increasingly on informal economic activities to make ends meet (Dibwe 2001; Rubbers 2013; Mususa 2014). The first waves of mass layoffs came at this time. In 1991, the Zambian government adopted a redundancy programme to reduce ZCCM’s workforce from 56,000 to 31,000 workers by 1997. Another third of the workforce was retrenched during and after the privatisation process, bringing employment down to 19,145 in 2001 (Fraser and Lungu 2007: 21). During this process, the government could not afford to pay most workers their terminal benefits; company houses were sold to their existing tenants at reduced rates as one way to cover some of what was owed to former ZCCM employees.

In Congo, the process was longer and more chaotic. In 1992–93, a popular movement to expel people originating from the Kasai region resulted in the departure of 9,000 Kasaian workers from Gécamines, reducing its staff from 32,000 to 23,000 (Dibwe 2001). In 2003–04, the World Bank organised a ‘voluntary departure programme’ within the framework of its recovery plan, reducing Gécamines personnel from 24,000 to 14,000 workers.3 On the conditions in which this redundancy programme was implemented, and the protest movement it gave rise to among the workers, see Rubbers (2010). In 2017, the staff was further reduced from 12,000 to 8,500 workers through a retirement plan. In addition to these organised departures, hundreds of Gécamines workers were transferred to joint venture projects through an employer substitution procedure.4 This procedure is generally activated with the consent of the private partner when Gécamines staff are assigned to the mine or plant that is the object of the joint-venture agreement.

Today, however, Gécamines remains the largest employer in the Congolese copper-mining sector. While its staff is a quarter of what it was in the early 1990s, it has still 8,500 direct workers out of approximatively 40,000 employed by the sector as a whole. These workers are the last remaining under the ‘state paternalism’ labour regime, even though the benefits they receive today are much less generous than in the past. Compared to those in new mining projects, these workers enjoy greater job security and benefits, including family housing and free company schools for their children. On the other hand, their wages are much lower than those paid by the large private mining companies and, until recently, they suffered several months of wage arrears. For these reasons, those who are offered the opportunity leave Gécamines for the new joint venture projects without hesitation.

As mentioned in the introduction of this book, this boom in the Central African Copperbelt was characterised by the diversity of investors developing new mining projects. Far from being limited to industry majors, they included juniors, former commodity trading companies, SOEs and even family firms, from all continents. Moreover, since the early 2000s, the mining and industrial projects that these investors bought and developed have changed hands several times. In Zambia, many changes occurred between 2000 and 2005, before the start of the boom, and continue to occur at regular intervals (Gewald and Soeters 2010). In Congo, merger and acquisition operations have been more numerous in the last decade following the financial crisis of 2008 and the decline of copper prices in 2011–16, leading to the rise of Chinese capital.

In 2018, the most important industrial mining projects were Kansanshi (First Quantum), Sentinel (First Quantum), Lumwana (Barrick), Konkola (Vedanta) and Mopani (Glencore) in Zambia, and Tenke Fungurume (China Molybdenum), Mutanda (Glencore), Kamoto (Glencore) and Sicomines (China Railway, Sinohydro and Huayou) in Congo (see Map 2). In addition to these relatively large projects are a number of small and midsized mining projects: about five in Zambia and more than a dozen in Congo. Finally, Congo hosts a dozen smelters, or hydrometallurgical plants, that exclusively process the ore (malachite or heterogenite) produced by artisanal miners. In total, in 2018, Congo produced more than 1.2 million tonnes of copper and Zambia about 860,000 tonnes. This gap between the two countries may grow even wider in the forthcoming years when new mining projects in Congo enter the production phase. This is particularly the case with the Kamoa-Kalula project, whose investors (Ivanhoe and Zijin) announced a projected production of 740,000 tonnes of copper per year. Such a production level would make Kamoa-Kalula the second largest mining project in the world.

At the beginning of the mining boom, economic liberalisation in both countries gave mining companies ample room to negotiate highly favourable contracts. Under pressure from international financial institutions, the Congolese and Zambian governments had implemented legal and tax reforms to provide attractive conditions to foreign investors and revitalise the mining sector. In Zambia, important changes were also made to labour legislation, allowing mining companies to implement labour policies that suit their needs. These changes included a reduction of employers’ obligations (for example, to provide housing for workers), the introduction of shorter casual contracts, and a shift of collective bargaining from the industry to the enterprise level. Finally, the one-union system was replaced with a situation where several union organisations compete for the representation of workers in mining companies, reducing and fragmenting union power (Rakner 2003; see this book’s Chapter 4).

In addition, before the rise of copper prices, foreign investors were in a strong position to negotiate the rights over Gécamines and ZCCM’s mining and industrial assets. However, as these agreements granted considerable advantages to foreign companies and were at first not made public, strong suspicions of corruption arose regarding the way they were concluded (Fraser and Lungu 2007; Commission de revisitation des contrats miniers 2007). Moreover, not content with these tax exemptions and contractual advantages, mining companies developed various tax evasion strategies which enabled them to reach a tax rate close to zero (Marysse and Tshimanga 2012; Das and Rose 2014; Readhead 2016). As a result, when copper prices suddenly soared in 2004, it appeared that the Congolese and Zambian governments would not benefit from the mining boom but that all the profits would flow to the shareholders of the various foreign companies.

Since then, the two governments have taken different measures to try and amplify state power vis-a-vis foreign investors and mining companies and to increase mining revenues.1 As the Africa Mining Vision initiative shows, this evolution is far from being limited to the Central African Copperbelt (Coderre et al. 2020). In Zambia, the legal and tax regime for the mining industry was amended in 2008, 2012, 2015 and 2016 (Manley 2013, 2015; Lundstol and Isaksen 2018). Then, in 2018, following an audit of the mining industry, the government decided to take several mining companies to court for tax fraud and announced a new tax increase. In May 2019, these steps led the government to seize control of Konkola, claiming that Vedanta (which controlled 79.4 per cent of Konkola, with the rest owned by ZCCM-IH) had breached the terms of their mining licence. In April 2020, it was the turn of Mopani, a subsidiary of Glencore (73 per cent), to face the threat of having its mining licence revoked for not giving enough notice before suspending operations due to the coronavirus pandemic. By early 2021, the Zambian government declared its investment holding company ZCCM-IH would buy back Glencore’s shares in Mopani, with a US$ 1.5 billion loan from Glencore subsidiary Carlisa Investments, to be paid off from future sales and profits. The government announced plans to find another investor, emphasising they had acted to save thousands of jobs at the company. Finally, the Zambian government enacted piecemeal reforms to labour laws in response to public anger over casualisation before producing a unified code, the Employment Code Act of 2019, in an effort to strengthen a limited number of worker protections, including lump-sum payouts for employees on fixed-term contracts.

In Congo, between 2007 and 2010, the government renegotiated all partnership agreements between Gécamines and foreign investors. Following its transformation from an SOE into a commercial company in 2011, Gécamines continued to put pressure on its most important foreign partners to have greater control over mining projects in joint ventures and to demand its share of the profits by taking legal action against them and negotiating new agreements. In 2011, a long and laborious process of consultation with the various parties involved was initiated to revise the mining code of 2002. It resulted in the adoption of a new mining code in 2018, which has considerably changed the power balance between mining companies and the Congolese state. This new code involves a significant increase in taxes on the mining industry, the obligation for mining companies to repatriate more than half of their export earnings, and greater state participation in mining projects (Lassourd 2018; Custers 2019; Unceta 2020).

The impacts of these reforms remain to be seen. In both countries, the attempts to regain control over the mining sector have at times enabled governments to increase revenues and invest in infrastructure projects. Having said this, the game is not over: each of these attempts aroused strong opposition from mining companies, which have strong weapons to oppose government measures. Companies can threaten to suspend their operations, organise mass worker layoffs or go to international arbitration. Faced with such threats and their consequences, the Zambian government is currently in a more delicate position than the Congolese government for two main reasons. First, the contribution of the mining industry to employment and local and national development has always been an important political question in Zambia, where Copperbelt Province in particular has a reputation as a political bellwether. However, employment and foreign investment became a particularly sensitive issue after the Patriotic Front put it front and centre in their campaign in the mid-2000s and subsequently came to power in 2011. The leaders of the Patriotic Front, and Zambian politicians more widely, are thus in a more difficult position than Congolese leaders when mining companies threaten to suspend their operations and to retrench thousands of workers.

Second, when the Patriotic Front came to power in Zambia, it invested in numerous infrastructure projects and social programmes, considerably increasing the country’s external debt (Siachiwena 2017; Hinfelaar and Sichone 2019). In November 2020, hard hit by the coronavirus crisis, the country was the first in Africa to default on sovereign debt. This debt makes the government particularly dependent on the mining industry, which in in the mid-2010s, directly accounted for 10 per cent of Gross Domestic Product (GDP) and provided just under 20 per cent of government revenue (EITI Zambia 2017; Dobler and Kesselring 2019). The contribution of the mining sector to state revenues is higher in Congo where the government has also increased public spending in the past decade: in 2017, the sector accounted for 17 per cent of GDP and 55 per cent of government revenues (Congo EITI 2017). But the country’s external debt has remained comparatively low since it reached the completion point of the Heavily Indebted Poor Countries process in 2010 (Unceta 2020). Though still narrow, the Congolese government’s margin for manoeuvre in its relationship with mining companies is therefore greater.

Despite differences in policy and politics on either side of the border, there are nonetheless overarching trends to the labour practices emerging amidst the diverse set of investors. Based on our research and review of available literature, it is possible to outline the principal dynamics of a new kind of labour regime. We identified five general trends in the post-privatisation labour practices of new mining companies in both copperbelts.

To understand how a new labour regime comes into being and how it functions, it is necessary to delve into the micropolitics of work, the interactions between the wide range of players involved in the politics and practice of labour. The subsequent chapters thus explore the recent trends in labour practices of new mining investors in more depth, and the way they are experienced, negotiated and contested by a wide range of actors, most especially workers, trade unionists and managers.