Key messages

• Social policies of colonial concession companies were driven by labour scarcity, domestic and international pressure.

• While social reproduction was possible in company towns, other concessions had no permanent workforce.

• The quality of social services and schemes declined after independence as driving forces lost importance.

Introduction

In 1885, what is today the Democratic Republic of Congo (DRC) became the personal property of the Belgian king. He made it the only territory in Africa to financially support the European metropole. To do so, he granted concessions with vast land areas to a number of companies that forced the population to collect wild rubber and ivory. Some companies, especially early ones, were given as part of their concession de facto sovereign powers over enormous areas; most concession companies have more limited contractual rights (generally the monopoly to grow and/or extract natural resources in a defined area) and duties (generally financial compensation, sometimes obligations regarding the use or African communities). In some cases, colonial governments granted land ownership to individuals or companies according to a similar logic. Violence was already widespread before the Belgian concession period but now reached unprecedented levels. Women and children were taken as hostages. Atrocities such as rape, arson, bodily mutilation and murder were widespread (Hochschild, 2006). These atrocities were revealed to an international audience as early as 1890 by an African American historian and journalist. His open letter to the Belgian king highlighted, among other issues, the fact that there ‘were no schools and no hospitals except for a few sheds “not fit to be occupied by a horse”’ (Hochschild, 2006: 110). The concessions were based on forced labour and offered hardly any social services. It took the visual proofs provided by a British activist to build up considerable international pressure. In 1908, mainly for economic reasons, the area became a Belgian colony. However, companies granted a concession, henceforth called concession companies, continued to be a central feature of the colony.

A colony or country where vast areas are granted to concession companies, and thus an important part of resources is managed by private companies, is called a concession economy (Mkandawire, 2016: 8). The Belgian Congo was not the only colonial concession economy in Middle Africa, a region where, in contrast to other regions, concessions did not predate formal colonialism but were launched by colonial administrations. The Belgian model of concession companies was copied by the French in what later became French Equatorial Africa, comprising today’s Central African Republic (CAR), Chad, Republic of Congo (Congo) and Gabon. In 1899, the French granted vast concessions to 40 companies, covering roughly three quarters of the territory (Thompson and Adloff, 1960: 13). Soon there were media reports about abuses by colonial and concession administrators. The results of the first official inquiry were hushed up, and subsequent official inquiries changed very little on the ground. Nevertheless, there was a shift in public opinion in France. However, it was again the economic failure of many concession companies that led to France taking over most of the concession areas. Some companies were given land and continued to operate with a focus on other raw materials, but reports on abuses continued until World War II (WWII).

This article will examine what happened in these concession economies after this initial phase in terms of the social policies of the concession companies. The understanding of social policy here is broad and encompasses the ways in which people’s needs for food, health, shelter, safety, education, old age and well-being are met in and around concessions. Two research questions are at the centre: What were the driving forces of social policies in concession economies? And what is their post-colonial legacy? The article starts by presenting the growing literature on the influence of European colonialism on post-colonial social policies. The colonial legacy of concession economies, predominant in the United Nations (UN) subregion Middle Africa, is, however, barely explored in the literature. The francophone and especially iberophone countries of Middle Africa are also largely ignored in the growing literature on contemporary social policies in sub-Saharan Africa. Based on a literature-based historical sociology approach, seven case studies of concession economies have been made. The case studies revealed three interrelated key drivers of social services in concession economies: the scarcity of labour, domestic pressure and international pressure. The following section shows that the social services and schemes provided varied. They were the most extensive in the company towns that started to reproduce their workforce at the end of formal colonialism. Other concession companies either offered social services on a smaller scale and with less focus on families or failed to create a permanent workforce. The next section turns to the post-colonial legacy and shows that in all cases, in a context of population growth, unskilled labour was no longer scarce and lost its bargaining power. Other domestic and especially international pressure supported the reversal of social services through privatisation and informalisation. Thus, the quality of these services and schemes generally declined after independence. Therefore, labour scarcity is a key condition for the provision of social services by concession companies.

The legacy of colonial social policies

The growing literature on the legacy of colonial social policies is premised on three types of research design. The first is large-N comparative studies (for example Schmitt, 2015; Schmitt et al, 2015). These studies show the importance of the colonial legacy for the adoption of social policies and their focus. In particular, they carve out the remarkable differences between the Spanish, French and British colonial legacies. However, these large-N comparative studies tend to mask important similarities between different colonial powers, such as the preferential social protection of civil servants (pensions, health care). They also tend to mask important differences within Spanish, French or British colonialism (see for example Ziltener et al, 2017). This is also frequently the problem with the second type of research design: small-N comparative studies (for example MacLean, 2011). While these studies add valuable insights with their in-depth presentation of colonial and post-colonial social policies, they fail to situate the selected cases in the spectrum of colonial experiences. The same strengths and weaknesses are characteristic for the third type of design: case studies (for example Bossert, 1985; Eckert, 2004).

To allow for differences within colonial empires and cover a broad spectrum of colonial experiences, the approach of Mkandawire (2010; 2016) is helpful. Mkandawire is a widely cited development scholar that is, however, surprisingly neglected in current mainstream social policy research. He took up the idea of Amin (1972) to look at differences in the manner in which countries were incorporated into the colonial economies and went beyond this by systematically linking these historical legacies with differences in colonial and post-independence social policies. This perspective is thus able to explain differences within British, French or Portuguese colonies. Mkandawire’s approach is more differentiated regarding colonial extraction than the mostly econometric authors that focus on the differences between settler and other colonies. The classification introduced by Amin (1972) and discussed by Mkandawire differentiates between three types: cash crop economies, labour reserves and concession economies. However, Mkandawire is not very clear about the criteria he uses to assign countries to the different forms of colonial incorporation, especially when a country has mixed features. He acknowledges some countries with mixed features (2010: 1650). In what follows, the three types are presented based on Mkandawire (2010; 2016).

In cash crop economies, peasant farmers participated directly in commodity markets by producing crops for domestic or export markets on their own land, frequently in addition to food crops and to a certain extent based on their own strategic choices. This type encompassed English, French and Portuguese colonies as it allegedly covered the whole of West Africa and Cameroon, Tanzania and Uganda. This form had a number of important implications for colonial and post-colonial social services and policies. Social policies were informal and community-based (MacLean, 2011). Generally, inequalities among peasant farmers were quite low. Peasant farmers participated in commodity markets and thus had income they could use for social services such as health care and the education of their children. While the level of participation in colonial education systems was comparatively high, a strong institutional basis for social policies and services was lacking, and out-of-pocket payments continued to characterise these areas after independence. Colonial state social policies were focused on the stabilisation of commodity prices and thus on the stabilisation of the peasant farmers’ incomes. However, this description mainly fits the famous case of the Gold Coast and neglects to mention that the participation of farmers in commodity markets was in many cases based on forced cultivation (for example for Senegal, Sow, 2015). Under forced cultivation, farmers had no choices regarding cultivation and generally struggled to survive. In an earlier work, Mkandawire (1986) acknowledges forced cultivation but sees it as a phenomenon that was soon abandoned. However, in Senegal, forced labour was officially abolished after WWII. In the regions concerned, forced cultivation was thus the ‘dominant [feature] of the colonial experience’, the vague criterion Mkandawire (1986: 72) uses to deal with the dynamics of colonialism when assigning dynamic colonies to static types.

In labour reserve economies, Africans were pushed to migrate and work on mines, plantations and farms. To secure a cheap labour force, alternative work opportunities such as cash crop farming were discouraged or prohibited for Africans, for example by the creation of ‘native reserves’. Rural communities were kept at subsistence level and responsible for the reproduction of the labour force. The colonial state was strong. This type allegedly covered the whole of Southern and much of East Africa and also includes English, French and Portuguese colonies. Social policies were generally divided along racial lines. While the rural African population was basically left to its own devices, the need to stabilise the workforce pushed colonial governments to introduce a limited amount of education, health care and housing for workers as well as specific policies for indigent workers reminiscent of the English Poor Laws. Thus, labour reserve economies had social policies that strongly relied on formal-sector employment. However, Mkandawire (2016) tends to generalise Southern African experiences and thus does not acknowledge the considerable variation among labour reserve economies (Künzler, 2020). This is also problematic in the context of his description of post-colonial social policies as broader tax-funded social policies such as cash transfers or social pensions were only introduced in some of them, generally in Southern Africa.

Remarkably, Mkandawire (2016: 8) only dedicates two sentences to the third form of incorporation into the colonial economies, concession economies. He omits them from further discussion because of a lack of data and claims that this experience of incorporation is predominant in the UN subregion Middle Africa. This clearly constitutes a research gap, also from a theoretical perspective and even more so because the francophone and especially iberophone countries of Middle Africa are also largely omitted from the growing literature on contemporary social policies in sub-Saharan Africa. While the other two types of incorporation focus on the state and extended families as the main drivers of social policies, companies are key actors in concession economies. In what follows, the focus is on the social policies provided by these private companies. This provision was of course influenced by colonial states and individuals embedded in extended families. However, social policies not related to concession companies are not discussed in what follows.

Empirical strategy and case selection

The UN subregion Middle Africa consists of Angola, Cameroon, CAR, Chad, Congo, DRC, Equatorial Guinea, Gabon and São Tomé e Príncipe (STP). The term ‘Middle Africa’ is admittedly not very common. However, it is convenient as this subregion is clearly defined, in contrast to the term ‘Central Africa’, which lacks an authoritative definition and is somewhat misleading, as it historically was used for the federation of what is nowadays Malawi, Zambia and Zimbabwe. To simplify the reading, contemporary names such as CAR are used for territories that had different names and affiliations during the colonial and post-colonial periods. Mkandawire’s classification (2016: 2) lists the CAR, the Congo, the DRC and Gabon as concession economies. Chad is the only area in Middle Africa that concession companies found unattractive (Thompson and Adloff, 1960). In his earlier publication, Mkandawire (2010: 1649) also includes two non-Middle African countries in the concession economies type: Rwanda and Burundi. He later classifies them as labour reserve economies without much explanation (Mkandawire, 2016: 2).

However, on closer examination, Equatorial Guinea and STP must be added, as these countries, not discussed by Mkandawire, clearly have a colonial past dominated by concession companies. Concession companies were also economically important in Cameroon (classified as a cash crop economy by Mkandawire) and in Angola (classified as labour reserve). These two countries are thus also added. In contrast, Congo is excluded because the social services and schemes provided by private concession companies were negligible and most of the areas were taken over by the French colonial government (Thompson and Adloff, 1960).

Case studies of the remaining seven concession economies Angola, Cameroon, CAR, DRC, Equatorial Guinea, Gabon and STP have been made. They follow a historical sociology approach. In contrast particularly to econometric approaches used in large-N and case studies, the historical sociology approach is more sensitive to the temporal context on different levels (from micro to macro) and better able to identify the processes at work. In particular, it is possible to identify and discuss phenomena that have not been directly reflected in statistical data, but reported in other sources. The case studies are based on systematic research of the existing literature on social services provided by concession companies in the colonial and post-colonial periods. The literature consulted is in English, French and – when available – Portuguese and Spanish. However, the available information varies in depth. Due to space restrictions, only the comparison of the case studies is presented but not the case studies themselves. Table 1 gives an overview of the concessions that characterise the different concession economies and the main sources used.

Scarcity of labour, labour migration and international pressure

Concession economies depended on wage labour. It is striking how across the Middle African concession economies – and elsewhere! – the local population was at first not interested in working for the concession companies. Initially, concession economies relied on forced labour. While there was also forced labour in cash crop economies and labour reserves, forced labour was clearly more severe in the concession economies and lasted longer. The colonial administration upheld it as official policy well into the 20th century, especially in the Portuguese colonies. Even after the official ending of forced labour, it continued in the form of mandatory and thus forced cultivation, for example in parts of the DRC (Likaka, 1997) and Angola (Furtado, 2017), French Cameroon (Authaler, 2018), CAR (Thompson and Adloff, 1960) and in the form of public work. Forced cultivation and public work made wage labour in concession economies comparatively attractive and also triggered labour out-migration in the case of the French Cameroon (Authaler, 2018). ‘Voluntary’  recruitment was also frequently based on coercion through intermediaries such as local authorities or missions or deceit. Thus, early-stage wage labour had to be enforced in most cases. The level of coercion varied and was somewhat smaller in the British Cameroons, Gabon and the Forminière/Bécéka concessions in the DRC than the other concessions.

It could be tempting to see low population density as a reason for the difficulty of voluntary recruitment of the local population. Indeed, STP was unpopulated, and Gabon, the CAR, DRC’s Katanga region and Angola’s Lunda region had low population densities. However, more important was the availability or not of alternatives for the local population. This varied within colonies. In the DRC’s Katanga region, small-scale agriculture remained a viable alternative, and the immediate hinterland of the Union Minière du Haut-Katanga (UMHK) mines supplied food crops to the mines and ‘provided sufficient income to enable Africans to avoid the hazards of mine labour’ (Larmer, 2017: 173). In the area of the concessions exploited by Forminière there were also ample alternatives (Derksen, 1983). This is in contrast to the Lever Brothers’ oil palm concessions, where the population was no longer allowed to harvest palm fruits independently for consumption and trade and was obliged to do the same work for the company for poor pay and under worse food and housing conditions than in their villages (Seibert, 2016). In Kilo-Moto, peasant farmers were not free to cultivate cash crops according to their own choice but were de facto subjected to the compulsory cultivation of fixed cash or food crops, a policy introduced in 1917 with a special focus on cotton (Likaka, 1997). This system of forced cultivation was criticised by international reports to no avail and basically continued under the agricultural policy called paysannat introduced in the 1930s. In Angola, rubber was an alternative for the population in Lunda (Cleveland, 2015), but there were fewer alternatives in the highland, where European settlers and plantation pushed the population out of small-scale agriculture and into wage labour (Ball, 2015). In the British Cameroons, land and thus an alternative to wage labour on plantations was available (Authaler, 2018). In the CAR and Gabon (Thompson and Adloff, 1960), and parts of Angola (Keese, 2012), smallholder peasants could cultivate cash crops such as coffee and cocoa. In STP and Equatorial Guinea, smallholder peasant production was initially an alternative that became increasingly scarce (Sundiata, 1990; Schümer, 1993). The existence of an uncaptured peasantry was made possible by the availability of alternative incomes, although this peasantry is not all that uncaptured when providing food crops to concession companies. However, there was no adequate provision of health care services and education for African cash crop farmers and their families as these kinds of social services either catered mainly for European residents or employees of larger companies.

As not enough workers could be recruited from the local population, concession economies turned to migrant labour that was remarkably trans-colonial. This was not a linear process, but rather a dynamic process that adapted areas and methods of recruitment depending on the changing labour needs of the concession economies. The most extreme example was Equatorial Guinea, which initially recruited workers from areas ranging from contemporary Sierra Leone to the Congo and then turned to massive recruitment in Nigeria to the extent that Nigerian migrants constituted 85 per cent of the population of Bioko Island, thus outnumbering the temporarily declining local population and Europeans (Sundiata, 1990: 47). STP turned to other Portuguese colonies and especially Angola for recruitment (Schümer, 1993). In addition, in the CAR and Gabon (Thompson and Adloff, 1960), intercolonial labour migration was organised within the French Equatorial African Federation, based on a regional quota system. The attempted recruitment of Nigerian workers to Gabon’s forest concessions was not very successful (Messi Me Nang, 2014). In Cameroon, the early concessions relied partly on workers from other colonies. When the area became the British Cameroons, they relied on workers from the French Cameroon, which was not part of the Federation but a separate mandate territory (Konings, 1993; Authaler, 2018). In the DRC, the UMHK depended initially on labour from Nyasaland (Malawi) and Portuguese colonies in Southern Africa, and especially on workers from the neighbouring British Copperbelt and, when mining started there, turned to other Belgian Congolese provinces and the Belgian mandate territory Ruanda-Urundi (Dibwe dia Mwembu, 2001; Seibert, 2016). Kilo-Moto recruited in other provinces and temporarily in Ruanda-Urundi (Bakonzi, 1982; Roberts, 1986). This points to the fact that labour scarcity has to be seen in a regional context. The concession exploited by Forminière recruited in different provinces of the Belgian Congo (Derksen, 1983). Only Angola relied mostly on intra-colonial migrant labour, with the exception of some experienced mine workers recruited from the Belgian Congo (Cleveland, 2015). Across all concession economies, most labour migrants stayed for longer periods or even permanently, but some labour migration was seasonal or casual, for example in the CAR (Thompson and Adloff, 1960), Cameroon (Konings, 1993) or the Forminière in the DRC (Derksen, 1983). There was also out-migration, for example from Gabon to mainland Equatorial Guinea and the Congo (Messi Me Nang, 2014) and from DRC’s Kilo-Moto area to Uganda (Roberts, 1986).

To sum up, the concession faced a considerable labour shortage that certainly had periodic fluctuations. The initial solution, recruitment based on coercion, made recruitment expensive for the concession companies, not least because many recruited this way died, evaded or were unproductive when evasion was not easy, as in the case of the island concessions in STP and Equatorial Guinea. Turning to intra- and especially inter-colonial labour migration made recruitment also quite expensive, especially when workers had to be replaced frequently because they died, evaded or migrated back. In this context, the UMHK moved away from its initial idea to follow the South African example and use recruiting companies to bring unskilled workers with short-term contracts (Dibwe dia Mwembu, 2001: 50) and instead looked for more permanent workers. Most other concession companies later followed suit. This scarcity of wage labour made concession companies introduce social services and schemes, presented in more detail in the next section. The idea of getting a more permanent workforce by expanding social services and schemes emerged quite early. George Grey, who was involved in the exploration and development of mining in Katanga as the manager of a British mining company, Tanganyika Concessions Ltd, warned already in 1906 of bad working conditions that deterred Katangan workers from wage labour in the mines and saw the protection of workers from illness and death as key to keeping labour costs down (Seibert, 2016: 92). The introduction of social schemes and services took place gradually and to varying extents. The first sub-Saharan African concession company to start a policy of labour force stabilisation through social policies was the UMHK from the 1920s (Dibwe dia Mwembu, 2001; Rubbers, 2013; Larmer, 2017); other companies followed later. The last places to introduce the policy were those where forced labour was still an option (for example Cassequel and Diamang in Angola). To varying degrees, workers stayed longer with the concession companies and started to seek wage employment voluntarily. Thus, regarding the first research question that looks at the driving forces of social policies in concession economies, labour scarcity is clearly the key driving force in all concession economies. However, this is not the only one.

International pressure was another driving force. The intercolonial character of labour migration partly explains why there was international pressure regarding working conditions on all concession economies even before the ILO or the League of Nations became relevant international actors. The sending colonies, especially the British, had an interest in controlling the recruitment of workers. The British influenced and inspected recruitment and working conditions in the Belgian Katanga with threats of border closures (Higginson, 1989; Vaessen, 2001). This was, however, more about sounding out the possibility of a British protectorate than concern for the fate of African workers. They also influenced and inspected recruitment and working conditions in Equatorial Guinea (Sundiata, 1990) and did not allow Nigerian migrants to work in the few mines in French Gabon, which all struggled to attract workers (Thompson and Adloff, 1960). The British mandate administration in the British Cameroons was initially quite conciliatory towards the plantations and obviously not under much pressure from the League of Nations, but later increased pressure for the adoption of international standards regarding housing and health care (Authaler, 2018). The League of Nations investigated labour migration to Equatorial Guinea (Sundiata, 1990). An early international campaign, the League of Nations and especially the ILO tried without much success to steer the labour regime in Angola away from forced labour. There were also campaigns from within the colonial empires. Angolan planters who competed for the workforce were part of the international campaign against the recruitment and labour conditions in STP (Seibert, 2006). International pressure was also driven by the commercial interests of competing companies. Public scandals in France were not very successful in changing the concession economies in the CAR (Thompson and Adloff, 1960) and Gabon (Barnes, 1992). While the motives for this early international pressure generally had little to do with concern for the workers’ well-being, it did lead to some improvements in recruitment and working conditions. Most efficient was pressure from the governments of sending colonies.

Reproducing the workforce to varying degrees

It was not only scarcity of labour that made concession companies stepwise expand social services and schemes, but also the need for more experienced and skilled workers that came for example with increasing mechanisation. There were no settlers in Middle Africa that claimed lower-skilled jobs; they were available for Africans. So, what kind of social services and schemes were provided at the end of the colonial period? It is important to bear in mind that what follows concerns the final years. There were considerably fewer social services and schemes during most of the colonial period and, especially before WWII, death, disability, illness and abject poverty were prevalent. The provision became most extensive in the company towns of the UMHK in the DRC (Dibwe dia Mwembu, 2001; Rubbers, 2013; 2020; Larmer, 2017) and of the Diamang in Angola (Varanda, 2012; Cleveland, 2015), two of the biggest concessions in terms of the size of the workforce. Here, at the end of the colonial period, not only food but also housing, health care, education, pensions and recreational activities were offered. Workers with indefinite contracts, along with their unpaid wives, could thus not only feed and house their monogamous Christian nuclear families but also educate their children, who would also become company workers, and then retire in old age. In case of work-related injuries, they would at least in principle be taken care of. In the UMHK, their salaries increased and they had the prospect of promotion. The UMHK had bigger company towns compared to Diamang, which mined scattered alluvial diamonds. While these company towns catered primarily for migrant workers, they generally also offered social services to the local population, as also stipulated in Diamang’s concession. Quite similar social services could be found in Cassequel in Angola (Ball, 2015). At the end of the colonial period, company towns and especially the UMHK were close to fully reproducing the workforce within these company towns that catered for workers from the cradle to the grave. This was a consequence rather than the primary aim of the UMHK policy.

The fewer social services that were offered, the more dependence on rural communities for reproduction increased. Several concession companies offered social services on a smaller scale in what could be termed company villages. This was the case in the scattered mining camps in the CAR (Thompson and Adloff, 1960) and in Kilo-Moto after the exhaustion of the more mobile and temporary alluvial mining deposits (Bakonzi, 1982) – also one of the biggest concessions in terms of the size of the workforce – and Forminière (Derksen, 1983) in the DRC. This kind of company village was also widespread in the scattered plantations in Cameroon (Konings, 1993; Authaler, 2018) and in STP (Seibert, 2006). The scattered plantations in Equatorial Guinea varied in the social services they offered and focused to varying degrees on families (Liniger-Goumaz, 1988; Sundiata, 1990). At the end of the colonial period, the temporary forest camps in Gabon were scattered and consequently focused generally even less on family housing, health care service provision and education (Gray and Ngolet, 1999; Messi Me Nang, 2014). The scattered European-owned plantations in the CAR offered few social services for their seasonal bachelor workers (Thompson and Adloff, 1960). Finally, the Lever Brothers concession in the DRC failed to create a permanent workforce as the social services and schemes offered to the scattered workers were too little, too late (Seibert, 2016). In concessions based on forced cultivation, such as Cotonang (Keese, 2012; Furtado, 2017) or in the two main cotton-growing areas in the DRC (Likaka, 1997), concession companies generally did not offer social services and schemes as labour was not scarce.

Thus, at the core of concession companies’ social policies was the formally employed male breadwinner. However, social security was not really the focus of these social policies, although there were limited pension schemes in some cases in the DRC (UMHK, Kilo-Moto, Forminière/Bécéka), Cameroon (CDC) and in the form of invalidity pensions in Angola (Diamang). The UMHK, for example, introduced an old-age pension scheme in 1938 and later harmonised it with the general pension system of the Belgian Congo introduced in 1957 shortly before independence (Mutombo, 1996: 233). The UMHK also introduced family allowances, though only for some years and conditional on sending children to school and preventive health care visits, before they were introduced in an unconditional form by the colonial administration.

Far more important – but strangely absent in much of the newer literature on social policies in sub-Saharan Africa – were wages and the possibility to save, the regular provision of adequate food and other household goods, and social services such as housing, health care and education. The provision of food was frequently subsidised, in the sense that it was based on mandatory cultivation, forced labour on plantations or low prices paid to small-scale farmers. While the concessions covered a considerable part of the small islands of Bioko (Equatorial Guinea) and STP, their share was smaller in the DRC and Angola and very small in Gabon, CAR, and Cameroon (see Table 1). On the downside, the diseases and injuries that health care facilities treated were at least partly caused by the concession system. It is important to keep in mind that living and working conditions in the concession economies were very harsh. However, forced labour and forced cultivation made the conditions outside the concession areas frequently even more difficult.

In cash crop economies and labour reserve economies, the state and extended families are the main drivers of social policies. In contrast, private companies are key providers of social services in concession economies. This provision was not only driven by labour scarcity and international pressure but also by domestic pressure from the colonial government and workers. Colonial governments were ambivalent. On the one hand, they supported concession companies, for example in the domain of recruitment, but also by using forced labour for the development of transport infrastructure or forced cultivation of food crops such as in the DRC (Derksen, 1983; Lyons, 1992; Dibwe dia Mwembu, 2001). They had close links to the concession companies, but on the other hand they had their own interests and cannot be reduced to being simple handlers of private capitalists. In the DRC, government labour regulations shaped how food rations, medical services and housing were to be provided by concession companies (Bakonzi, 1982; Derksen, 1983; Dibwe dia Mwembu, 2001) and in some cases also pressured for improvements after investigation (Lyons, 1992: 29). However, the colonial government pressured for more reforms in STP (Keese, 2011) and partly in the mandate territory of the British Cameroons (Konings, 1993; Authaler, 2018), but less so in the other areas. As the example of the UMHK reminds us, the colonial administration was not homogeneous. To name just one of many conflicts, parts of the colonial administration sided with business interests and Katanga’s regional administration and opted for more repression, while another part feared labour unrest such as the 1919–20 strikes in South Africa and wanted working and living conditions to improve (Vaessen, 2001). The resulting Decree of 16 March 1922 was a compromise between these two factions: it legislated both repression and the nutrition and housing of workers. Furthermore, as is illustrated, for example by the forest concessions in Gabon (Messi Me Nang, 2014), concession companies went further than governmental regulations in some regards (for example food ratios) but were reluctant in others (for example health care and minimal wages), much to the disapproval of the colonial government. Colonial labour regulations were in turn fuelled by workers’ unrest. Domestic pressure by workers, protests and strikes were also very effective tools to pressure the concession companies in the DRC and Cameroon directly for wage increases and improvements of social services. While the stabilisation policy of the UMHK proved successful insofar as there were no major strikes in Katanga after 1947 (Dibwe dia Mwembu, 2001), they continued longer in the CDC (Konings, 1993). The pressure from the workers took less the form of strikes and protests in Equatorial Guinea. In the French and Portuguese concession economies, the literature consulted showed no evidence of strikes and protests being drivers of social policies. However, absenteeism and other forms of workers’ resistance did of course occur.

The post-colonial fading out of the drivers

So, to turn to the second research question, what is the post-colonial legacy of these social policies? In all cases discussed, whatever social services were provided by concession companies during colonialism, their quality eventually declined after independence. This started at different times and happened for various reasons and to different extents. Resolutions or abandonments of concessions were important catalysts of the decay, for example in the cotton and sugar areas in Angola (Ball, 2015; Furtado, 2017), in Equatorial Guinea (Liniger-Goumaz, 1988) or the CAR (O’Toole, 1986). So were nationalisations of concessions or their welfare services in Cameroon (Konings, 1993), Equatorial Guinea (Klitgaard, 1990) and STP (Seibert, 2006). In Angola’s diamond concession, the civil war was a key driver of the decline (Udelsmann Rodrigues and Fahy Bryceson, 2018). The provision of social services and schemes lasted the longest in the Gécamines (former UMHK), but started to decline from the 1970s as a result of economic crisis and predation by politicians (Dibwe dia Mwembu, 2001; Rubbers, 2013) In Gabon’s forest concessions, it continued throughout on a low level (Messi Me Nang, 2014) and had a certain revival in the CAR’s forest concessions (Smith, 2015).

This development is linked to the fact that labour was no longer scarce. Further mechanisation decreased the demand for unskilled labour. Urbanisation and population growth made unskilled labour supplies far from scarce and reduced labour’s bargaining power. International pressure no longer pushed for an expansion of social services but, on the contrary, for privatisation and liberalisation (see for example Rubbers, 2013). As a result, many contemporary concession companies began to rely heavily on partly subcontracted casual labour of a fluctuating quantity, depending on the volatility of the market, and only on a small and rather skilled permanent labour force. For this labour force, contemporary concession companies are in principle required to respect minimal legal standards for wages and labour conditions for workers: for example, in the DRC, the Labour Act is generally most respected by large-scale mining companies whose wages, job security and social services vary (Rubbers, 2020). While there is not much literature on this, some companies still run limited social services such as schools and health care facilities for the benefit of their workers. The best social services are reserved for the small group of generally expatriate skilled workers. Although the industry is rather secretive and not well researched, it is safe to assume that the new concession companies extracting oil offshore in Angola, Cameroon, Equatorial Guinea and Gabon also mainly rely on skilled expatriate workers. While a few contemporary concession companies extend their limited social services to surrounding communities as part of corporate social responsibility activities, these are generally limited in extent, unsystematic and non-permanent (see for example Tallio, 2015; Udelsmann Rodrigues and Fahy Bryceson, 2018). With the exception of post-colonial diamond mining in Angola (Udelsmann Rodrigues and Fahy Bryceson, 2018), transnational immigration of unskilled workers has lost importance. Consequently, sending countries do not have a great deal of influence on labour conditions. Campaigns by international NGOs have led to a number of initiatives to improve working conditions in the mining sector, but for most workers they are still a long way from those obtained in the 1950s and 1960s.

In some cases, peasant farming has replaced larger plantations, for example to a certain extent in STP (Seibert, 2006) and Cameroon (Konings, 1993), and artisanal mining has replaced formal work at industrial mines to varying extents, for example in the DRC (Fahey, 2011; Rubbers, 2020) and CAR (Dalby, 2015). Artisanal mining provides an important source of income to a broader workforce compared to more industrial production, but the income is considered insufficient to support a family (for example Rubbers, 2020: 1569). Furthermore, there is no work-related provision of social services (schools, health centres, housing), for which workers depend on government, civil society and international donors.

Conclusions for social policies and theory

In essence, scarcity of permanent labour is a key condition for the type of provision of social services and schemes by concession companies to unskilled workers that ceased to exist after independence. International organisations, foreign governments and national governments ceased to push for an expansion of social services as did workers, as a consequence of declining bargaining power.

What is the relevance of this for social policies? Concession companies are currently rather insignificant as providers of social services in contrast to periods when labour was scarce. This scarcity is unlikely to return. However, formal employment could become more important when concession companies employ more permanent and fewer precarious day workers. It could also become more important if the focus shifted from the extraction of resources to their local transformation. The degree of mechanisation also has an effect; the best option for local employment would probably be limited and locally led mechanisation (see for example Radley and Geenen, 2021). Regarding plantations, the cash crop economies of West Africa have shown that free small-scale farming can be an efficient alternative that has the potential to generate relevant income (Cooper, 1996). For reasons beyond the scope of this article, the mostly rather authoritarian states in Middle Africa are quite unlikely to push in any of these directions. Neither are most international organisations. The most likely drivers would be international NGOs and consumer pressure. However, even though concession companies could certainly formalise their workforce and offer decent social services, this is unlikely to concern a particularly large part of the population. Instead of holding up the idea that social services and schemes should be delivered to the formally employed (male) breadwinner and their families and thus neglect the vast majority of workers outside formal employment, it might be more promising to consider how concession companies could and should contribute more generally to the provision of social services and schemes. This, however, is a complex debate that must be conducted elsewhere.

Regarding the theoretical framework, what happened in concession economies is quite similar to what Mkandawire (2016: 6) briefly mentions happened in European-owned mines, plantations and farms where the cheap labour force recruited in the labour reserves worked. There are other indications that the country-wise separation of the three forms of incorporation is somewhat artificial. To take an example from Southern Africa, Mozambique, classified as a labour reserve by Mkandawire (2016: 2), has, according to Alexopoulou and Juif (2017: 215), labour reserve regions, cash crop regions and regions with concession companies in one single colony. This results in internal inequalities, for example with regard to access to land and economic opportunities. Regarding Middle Africa, Table 2 shows that in most colonies, there were different forms of colonial incorporation that dominated the colonial experience of specific regions within the colonies. This constitutes a basis for post-colonial inequalities within countries and needs to be acknowledged theoretically. As mentioned in the theoretical section with reference to Senegal, forced cultivation by small-scale farmers is a fourth form of incorporation that dominated the colonial experience in certain areas and should be added to the typology. In Middle Africa, this was the case in Angola, Cameroon, the CAR and the DRC.

As shown earlier, social services and schemes varied in concession economies, the first form. Only towards the end of the colonial period did company towns and Cassequel provide a non-precarious life to workers and their families and nearly full reproduction of a workforce that had lost dependence on access to land. To do so, they had to provide not only a family wage and access to food but also social services such as health care, education and social security in case of sickness and old age. The further away from company towns the concessions were, the more dependent workers were on access to land with regard to old age, sickness or food security for their families throughout the year. Full reproduction was also possible in non-mandatory cash crop economies, the second form. In contrast to the first form, this was not made possible through social services and schemes, but through access to land. How precarious or not life was varied. Among those better off were farmers that employed migrant labour, which was rare in Middle Africa compared to West Africa. The most famous example is today’s Ghana (Cooper, 1996: 48–9), where these migrant workers were not only attracted by wages in cash or kind but also by access to land to feed themselves and their families, long-term relations to landowners and, eventually, customary land ownership. However, monopolistic buyers and, especially in the post-colonial period, increasing land-scarcity made this form more vulnerable. In contrast to Mkandawire’s description of the cash crop type, ‘non-mandatory’ is added here because it is key that farmers could farm cash and food crops according to their own priorities. This was not the case in the new third type, forced cultivation. Here, small-scale farmers had no choice regarding the crops they planted and were generally forced to sell them at fixed prices to concession companies granted a buying monopoly over large areas. Reproduction was possible but generally on a rather precarious level that depended on how labour intensive the crops were, on whether their season might allow the side cultivation of food crops or not, and finally on the price that was fixed for them. Forced cultivation was introduced in some concession areas but also outside formal concessions. In some concessions (for example Cotonang or Lever Brothers), farmers had already cultivated cash crops before the concession was declared. Forced cultivation meant that they had no choice when they deemed the prices fixed for them too low. With the end of forced cultivation, this type was frequently transferred to the second type, non-mandatory cash crop production, but this happened generally quite late in the colonial period or, as in the case of the DRC, quite some time after independence (Likaka, 1997). Areas of forced cultivation also served as labour reserves, albeit in a more indirect and less systematic manner than the final fourth type, labour reserves proper. In this type, self-sustaining agriculture at the lowest level was possible and also protected against private ownership by settlers or concession companies to a certain extent (see for example Meillassoux, 1981: 117). However, and this is the difference to the second type, access to cash income inside the labour reserve was severely restricted in order to push male workers into non-permanent wage labour, for example for concession companies that preferred an unreliable supply of bachelor migrants to a stabilisation policy. This restriction was softened in the post-colonial period. However, in the post-colonial period, there are also reserves for national and international labour migrants: beside poor rural areas also urban informal settlements. In all but the first type, access to health care and education depended on the unequal offer of (post-)colonial governments and faith-based organisations. During the colonial period, wage labour seems to have been less important than access to land and the unpaid labour of women for the reproduction of the workforce over the life cycle of workers. This is remarkable given the focus of post-colonial social policies, which tended to aim at formal employment for several decades. The fact that in most colonies there were specific regions with different forms of dominant colonial incorporation also suggests the need for different social policies, tackling the specific problems of cash crop, urban and rural labour reserves and concession areas.

Author biography

Daniel Künzler is Lecturer at the Department of Social Work, Social Policy and Global Development, University of Fribourg, Switzerland. He is currently working on the legacy of colonial social policies and the dynamics and politics of social policies in East Africa and Middle Africa.

Conflict of interest

The author declares that there is no conflict of interest.