Labor Supply and Corporate Strategy, 1926–1936
The profitable operation of the mines required large quantities of capital and labor. We have seen how and why capital became available. This chapter focuses on the manner in which a competitive wage labor market shaped the labor strategies of the mines, altering the South African system of labor mobilization and control in ways which, it is argued in later chapters, had important consequences for the development of black worker consciousness and action on the Copperbelt.
BEFORE THE DEPRESSION
Labor Strategies of South African Mining Capital
Initially, the Copperbelt mining companies, particularly those owned by Anglo-American, considered emulating South African mining capital’s system of labor mobilization and control. This system minimized labor costs in various ways by establishing a recruiting monopsony over African labor. “Servile labor measures/’ such as the closed compound system, pass laws, and legislation against breach of contract, protected South African employers against potential organization and protest by black workers. This system of “exploitation color bars”1 virtually destroyed opportunities for black workers to bargain with employers over their wages and other benefits.
The compound system was first established in the 1880s in South Africa in order to prevent diamond stealing by black labor. The mine owners enclosed workers in compounds surrounded with barbed wire. All employees were searched before and after work. By 1885, the De-Beers compound was a quasi-military enclosure, surrounded by a corrugated-iron fence ten feet high, with a single large gate as an entrance. African workers entering the DeBeers compound lost all access to the outside world for the period of their work contract. Wives and children were barred. Management regulated daily life in the compounds to maximize exploitation of the work force. Thus, the compounds minimized costs through economies of scale and uniformly low living standards, and facilitated control over the work force. This system was soon copied by many other South African industries, particularly the gold mines on the Rand, and it became one of the cornerstones of the South African system of labor exploitation.2
Labor Supply and Demand
Conditions in central Africa ruled out replicating the South African compound system. As a result, the South African pattern of labor mobilization and control had to be altered to fit specific circumstances on the Copperbelt.
Unlike gold mining, which used largely unskilled labor, copper mining demanded higher proportions of skilled and semiskilled labor. The exact proportions of each varied with the nature of the ore deposits. Roan, for example, needed more skilled and semiskilled labor (hoist drivers and blasting license holders) to work its predominantly flat (scraping) stopes. Mufulira’s orebodies (three orebodies superimposed on top of each other) called for extensive use of unskilled lashers, which increased the size of the black unskilled work force. Nkana Central and Mindolo orebodies were more accessible and permitted greater use of unskilled labor.3
Skilled white miners could find work in Europe, South Africa, and North America. The poor health record and meager facilities on the Copperbelt thus held little attraction for them. In order to attract sufficient skilled labor, the mines hired experts to eradicate health hazards, built good houses and recreational facilities, and offered high wages. Gradually, the number of European miners increased from a mere 304 in 1926 to 2,934 in January of 1931.4 Men came from all over the world. “The new white communities . . . were composed of all sorts and conditions of men—Texas drillers, hardbitten Afrikaners, Yugoslav timber-men, fitters and turners from South Africa, smeltersmen from Wales, American construction experts with experience in the copper mines of Arizona or Latin America, Cornishmen and others.”5 Their one common goal was the search for high living standards.
Since white labor was both expensive and scarce, the mines recruited Africans for semiskilled and unskilled work. Large numbers of unskilled laborers were needed for mine construction and many of these were hired by contractors for limited periods. More and more surface and underground miners were brought on in the late 1920s, when construction declined and extraction began in earnest. In 1927 the mines employed about 10,000 African construction workers, but by 1930 there were nearly 30,000 production workers.6
Before the Depression the copper companies were hard pressed to obtain sufficient African labor, especially experienced underground black labor, which was in short supply all over southern Africa. As readiness for production approached, both mining companies became increasingly concerned about their labor supply. In 1929, two shafts at Mufulira even closed temporarily because of insufficient personnel.7 A year later, visiting mining engineer Sydney Ball warned that “this labor supply . . . with certain of the Companies at least, may retard appreciably their attainment of a large copper production.”8 The general manager at Mufulira reported a “grave shortage of boys, particularly for work underground.” The situation was so desperate that he advised “any reasonable expenditure on training, better food or better living conditions in the compounds is alright if it helps get labor underground.”9
Such difficulties were not due to an insufficient labor supply. Northern Rhodesian men had worked in the mines of Katanga, Southern Rhodesia, Tanganyika, and South Africa since the early twentieth century, many travelling south in search of the higher Southern Rhodesian and South African wages.10 The Southern Rhodesian gold mines recruited labor from the north through the Rhodesian Native Labor Bureau, and the Katanga copper mines obtained Northern Rhodesian labor through the Bourse du Travail du Katanga and Robert Williams and Company. In 1920 and 1921, approximately 9,000 members (or 56.1%) of the Union Miniere work force were from Northern Rhodesia. Thousands of Northern Rhodesians voluntarily migrated to Elisabethville as well,11 and a small but regular number of Northern Rhodesians worked in the Lupa gold fields of Tanganyika.12
Northern Rhodesians employed in southern Africa had developed their own networks which conveyed information on labor conditions. News travelled rapidly along the established labor routes, and conditions on the mines of Katanga, Southern Rhodesia, Tanganyika, and Northern Rhodesia were well known. Most Northern Rhodesians understood the relative rewards available to them at various places of employment, and they made decisions based on that knowledge. The general labor shortage of the 1920s in central Africa put skilled workers in a relatively good bargaining position in the labor market.13 Jobs were plentiful, experienced workers were in short supply, and employment information networks enabled miners to compare the potential rewards for their labor. Northern Rhodesian workers knew they could change jobs easily, and they did. The Africans avoided recruitment, preferring “to find their own way to the Copperbelt and to apply for work under the particular boss or department which appeals to them at the mine they have chosen. . . . [This gives them] a much greater choice of conditions of work.” Many workers preferred working for contractors because of “the loose system of control they exercised over labor at work and in the compounds.”14 Even recruited labor could, and did, choose between working in or outside Northern Rhodesia.
Small wonder that when the mines began construction in 1926, Northern Rhodesians did not flock to them. They compared Copperbelt wages, working and living conditions with those of other employers, and many continued to work elsewhere. As late as 1929, 50,000 Northern Rhodesians were still employed outside the country, most of them in Southern Rhodesia. Some went simply because of proximity, but many more were drawn by higher wages.15 Better working and living conditions also drew Northern Rhodesians to Union Miniere, which employed some 10,500 Nothern Rhodesians in 1929. Union Miniere’s shift to stabilized labor in 1926 had been accompanied by a dramatic upgrading of living conditions for their workers. By the end of the decade African workers in Elisabethville enjoyed the best food and housing in central Africa. As a result, Union Miniere recruiters had no difficulty obtaining Northern Rhodesian labor. In fact, many Northern Rhodesians went to Katanga on their own instead of working on the Copperbelt.16
Within Northern Rhodesia, Broken Hill and Bwana Mkubwa mines competed with the Copperbelt companies for labor. These proved to be the mines of choice because they allowed workers to bring dependents to the mines and provided housing and rations for them. Broken Hill was especially popular because it offered miners five-acre plots to supplement rations and wages. On the other hand, Copperbelt mines with poor reputations, such as Nkana, had difficulty obtaining sufficient voluntary labor.17
Labor Strategies and Labor Supply
The companies adopted a number of strategies, both jointly and separately, to attract labor. Because voluntary labor remained at the mines longer and cost less than recruited labor, the mines preferred to employ voluntary labor as much as possible.18 In this period, about two-thirds of the companies’ labor force came to the mines voluntarily.19
The companies joined forces with local European farmers and pressed the Northern Rhodesian government to help them secure more control over the labor supply by placing immediate limitations on foreign recruitment. However, government officials refused to cut off completely this established pattern of worker migration for what they considered the still uncertain future of the copper industry. Katanga continued recruiting until 31 July 1931, although in slowly decreasing numbers. The flow southward into Southern Rhodesia did not diminish until the Depression, and was due then to market factors rather than legislation. Labor policies in Southern Rhodesia, Northern Rhodesia, and Nyasaland were not coordinated until 1936, when an agreement established the priority of each country’s labor needs.20 Thus, despite some efforts to help the mines acquire and control labor, during the 1920s the Northern Rhodesian government did not yet accept the importance of the mining industry to Northern Rhodesia and England, and refused to grant the copper mines exclusive control over Northern Rhodesian labor.
Lack of control over the labor supply aggravated the conflict between competitive self-interest and potentially beneficial cooperation between the Northern Rhodesian copper mines, for the companies had to struggle against each other for labor as well as against foreign competitors. However, interlocking directorates between the two companies and an appreciation of the dangers of unchecked rivalry prevented competition between the mines from getting out of control. Oppenheimer wanted to amalgamate the two Copperbelt companies because he feared that destructive competition would undercut profits. In 1937 Anglo-American suggested “a fusion where both groups have an equal say but where the control is vested . . . in the ‘Chartered’ company.” Negotiations between the companies bogged down over some misunderstandings and the Rhodesian Selection Trust’s disinterest. With improved market conditions in 1937, talk of amalgamation ended, and Rhokana reopened Nchanga as a separate company. Unlike the mines in South Africa and Southern Rhodesia, the copper companies did not establish a Chamber of Mines to enforce cooperation between the mines until 1941. Management in London, New York, and Johannesburg kept in touch over larger issues, but daily cooperation between the companies did not occur.21
The mines did establish a joint recruiting agency following the failure of voluntary labor and small private recruiting agencies. Rhodesian Selection Trust hired H. H. Field in 1928 to organize a recruiting depot at Fort Jameson (Chipata) in Eastern Province.22 Anglo-American contracted for recruits with R. W. Yule, who also worked for Union Miniere du Haut Katanga. In 1927, only 12% of Roan’s labor force was recruited, but as construction expanded, the need for labor rose dramatically, and with it the need for more effective recruiting. By 1928, 44% of Roan’s work force was recruited and Roan’s general manager called for the formation of a central recruiting bureau. A year later the two groups met with the governor and agreed to create The Native Labor Association. As an interim measure, both mines agreed to employ Yule. This agreement failed to yield much labor, and labor scarcity soon drove the mines to replace Yule with A. Stephenson, former commandant of the Northern Rhodesian police.23 In its first year, the Association sent 10,000 men to the mines and in 1931, had 27 agents stationed at strategic points throughout the territory.24 Like its South African counterpart, the Association arranged for recruits to sign work contracts for set periods of employment, and guaranteed transportation to and from the mines. Although this cooperative venture never supplied all mine labor, recruitment did provide much of the skilled black labor, frequently from outside Northern Rhodesia. In the early 1930s, the Association recruited 2,400 trained Southern Rhodesian miners to upgrade the work force and supplied over 2,000 Nyasaland workers to fill clerical jobs. According to Spearpoint, these men dramatically upgraded the efficiency and productivity of the work force.25
Still both mines complained about the quality of their laborers. Said L. Eaton of RST, “The native negro is small and unused to hard physical labor.” Mechanization seemed the logical answer, for management believed labor could “be taught to handle simple mechanical equipment or routine work. The supply [of labor] is inadequate and recourse must therefore be had to minimizing methods in which the ore is moved by mechanical or gravity means.”26 Elaborate plans were drawn up to increase mechanization in the future.
However, none of these measures succeeded in overcoming competition between the mining companies for voluntary labor. Because of their need to attract voluntary labor and their failure to monopsonize recruitment, the companies had to offer conditions of employment comparable, or nearly comparable, to those at neighboring mines. This competition led to a situation in which the mines “were all looking for recruits and making offers better than the man next door to get the native to engage.”27
Wages on the Copperbelt rose as each mine tried to outbid the other for labor. For example, at Mufulira, the general manager wrote the head office in 1929 saying that “Mufulira’s pay was a little lower and should be put up with other Copperbelt wages.”28 By this time, the average monthly wage of black workers on the Copperbelt was 18s, while many experienced men earned between 20s and 22s 6d. A few underground workers even started as high as 30s. Skilled underground workers who were in great demand did even better, with some drawing wages of 45s or more per month (see Table 1. for detail). By 1930, the average wages at Roan were 21s per month for surface work and 33s for underground work. In 1932, the average wage at Nkana was 25s for surface work and 32s 6d for underground work.29 These scales enabled the Copperbelt mines to compete with Union Miniere for skilled labor, and to outbid them for unskilled labor. Northern Rhodesian wages were still below those at the sisal estates of Tanganyika and the Wankie mines in Southern Rhodesia, which had average monthly wages of 22s 6d and 32s 6d respectively.30 However, since Katanga was the nearest competitor for Copperbelt labor, the mines concentrated on competing with Union Miniere’s wages.31
Competition for labor forced the mines to deviate from the South African migrant labor system. They had to recognize that skilled and experienced Northern Rhodesian workers were accustomed to living with their families at Katanga and other local mines. And since the companies preferred experienced black labor to costly white labor, they still had to offer accommodations and rations to attract veteran African miners.32
Source: Coleman, The Northern Rhodesian Copperbelt, pp. 179–80. Note: These specimen figures are from Rhodesian Selection Trust files.
The acceptance of a stabilization policy varied with the differing labor needs and managerial traditions of the two Copperbelt companies. Because of the nature of its orebody, the Rhodesian Selection Trust mine, Roan Antelope, required a relatively high proportion of experienced labor.33 The compound manager reported that “in the early days it was practically impossible to get labor and so when a native offered himself for work, we were only too pleased to take them together with their wives and families.”34 Many Rhodesian Selection Trust managers came from America and Canada with their traditions of company towns, and consequently were more willing to encourage stabilization. David Irwin, an American mining engineer, was the general manager at Roan and Mufulira for the first five years. Another American, R. M. Peterson, was general manager during World War II. The general managers had considerable freedom to set policy.35 Irwin “set standards which were based on American mining experience. He insisted on a sensible organization, good engineering, and good conditions for Africans. He favored some industrial training for the Africans. . . . Africans were given purgatives and fed well. Malarial eradication programs were started. All this was to attract European labor as well as Africans.”36
The percentage of married stabilized workers at the Rhodesian Selection Trust mines increased steadily as management discovered that married labor was “more efficient, healthier, more contented, and remains longer than the single native,” and that women in the compounds discouraged prostitution, gambling, fights, and general disorder. In 1931 married miners at Roan stayed an average of 20.25 months, or twice as long as single workers and they could be trained to carry out the more complicated tasks of copper production.37 Management soon recognized the advantages of married labor, and the percentage of married workers rose steadily from 20% in 1927, to 37.3% in 1932, and to 65.11% in 1936. The percentages lagged at Mufulira, partly because that orebody required fewer skilled workers and partly due to the mine closure during the Depression. Still, both mines remained committed to stabilization, and Mufulira eventually reached parity with Roan (see Table 2.) On both mines, the length of employment increased. The average length of service for surface workers at Roan climbed to 18 months, with 64% of the African employees in 1932 having over a year’s employment on the mine, and 8.27% having over two years’ employment.38 Labor turnover at Roan fell from 24.09% in 1927 to 7.5% in 1933 (see Table 3.)
Because of the nature of the orebodies at Nkana and Nchanga, Anglo-American mines needed fewer experienced miners. They were able to maintain production with voluntary migrant labor or unskilled recruits. As Table 4 reveals, Nkana had significantly less long-term labor than Roan. In 1932, only 18.9% of Rhokana’s miners were married, and the average length of employment for married men was 12.9 months as opposed to Roan’s 20.25 months. Married men, however, did stay longer than single employees, who averaged 8.6 months of employment.39 This discrepancy between the mines can be partially explained by the differing managerial traditions of the two companies. Rhokana field management was appointed and closely supervised by South African consulting engineers from the Johannesburg office, who were accustomed to the South African migrant labor system.40
Source: Merle Davis, ed., Modern Industry and the African, p. 71.
Economics, not ideology, was the main issue. When necessary, Anglo-American was just as capable of stabilizing workers as the Rhodesian Selection Trust mines. In 1931 the Rhodesian-Anglo-American Company even petitioned the Colonial Office to permit permanent settlement for several thousand skilled laborers from Nyasaland at Rhokana, and advertised good housing, recreation, medical and welfare facilities, as well as social security benefits. The mines were willing to bear the cost of such benefits if they would guarantee sustained improvements in output.41 Nevertheless, they preferred lower-cost migrant labor, and retracted this offer when the Depression changed the labor market and a sufficient number of experienced miners accepted short-term employment on the mines in order to remain on the Copperbelt. Before the Depression, therefore, Anglo-American never had as much married labor as the Rhodesian Selection Trust mines. And, when the labor market loosened, Anglo-American was able to further reduce its married labor force and to rely on voluntary migrant labor for both unskilled and more skilled work. Managerial ideology could be put aside when economically necessary, but whenever possible, the Anglo-American mines tried to replicate the South African migrant labor system.
The dependence upon voluntary labor, particularly married stabilized labor, forced the copper companies to alter the control and facilities of the South African compound system in ways which, it will be argued below, significantly affected the development of class consciousness and action among the miners.
The voluntary nature of the labor force left the mines vulnerable to employee mobility. As Spearpoint complained, “It was difficult to get a native to engage for work, and once he did so it was equally difficult to keep him.”42 Unlike the South African compounds which were designed to regiment a captive labor force, the Copperbelt compound system had to attract and keep labor as well as control it. With this in mind, the companies set up programs which not only drew workers and their dependents to the mines, but also helped acclimatize them to the working environment.
Northern Rhodesian and Colonial Office officials were only too happy to let the mining companies control their compounds. The perilous financial condition of the Protectorate, and reservations about the long-term prosperity of the mines, made the government reluctant to interfere in company affairs. The government had passed laws pushing Africans into wage labor and keeping them there for the duration of their contract. At the same time, the colonial state legitimized its role as protector of the African people by setting minimal standards for the compounds—standards which the mines had little difficulty meeting. Mine Safety Regulations covered work conditions.43 Ostensibly, officials were supposed to enforce the regulations through regular inspections of the mines and mine compounds, but government enforcement proved difficult. Provincial Administrators made occasional sweeps through the compounds, but these were infrequent and generally in the company of mine officials. When a District Officer visited the compounds he was more “in the position of an honored guest” than an inspector.44 In 1929, Ndola District had no European officers available for patrol work in the district or township. Sub-stations with District Officers in residence did not open until 1931 at Nkana, Nchanga, Luanshya, and Mufulira. Police were in short supply, and the only Resident Magistrate resided in Ndola.45 The Provincial Administrators focused on the problems of the local Europeans and rural Africans. Only a few officers believed urban Africans were a long-term problem. Most government officials considered the mines “exemplary employers.” When officials criticized the mines, they did so in private and politely.46
Because of these inadequacies, government played a largely passive role in regard to industrial labor in this period, and thus gave the mines a free hand to set the parameters for work and living conditions.47 The Copperbelt mine compounds developed into self-sufficient townships housing black miners and their dependents during their employment.48 “Responsibility for the housing, drainage, sanitation, health, discipline, and general welfare . . . [was] vested in the compound manager and his organization.”49
Compounds were controlled by compound managers, who were quite autocratic. The compound manager had “full authority over all employees except when actually at work; he and his staff receive the new entry of recruits, allocate accommodations, supervise the issue of rations, hear complaints, superintend arrangements on payday, devise and repair housing, washing, and sanitary arrangements, maintain order, and generally control the several thousand natives of whom they are in charge.”50 In the 1940s the compound manager disciplined the African labor force. The “changa changa,” as the black miners called him, struck fear in the hearts of the miners,51 and the general managers took every opportunity to buttress his position. Management warned government officials not to undermine the authority of the compound managers. For their part, the compound managers left no doubt to visiting officials as to who was in control. Spearpoint was even known to spy on government officials because of his “great suspicion of the administration.”52
The compound managers were assisted by European and African compound staff, which varied in size at each mine. Generally at least one European was in charge of each job division in the compounds, such as sanitation, housing, repairs, and food allocation. The European staff was assisted by African manual laborers and a staff of clerks, many of whom had been trained in mission schools in Nyasaland. They were translators for the Europeans (frequently done even when the staff knew African languages) during discussions with Africans, and they screened cases coming into the compound offices, deciding who should hear each case. The work crews carried out the daily maintenance, checking on housing, sanitation, and other compound problems. As the mines expanded, services for workers broadened, and the compound staff grew accordingly. (Growth was particularly swift in the 1950s with the development of mine welfare activities.53)
Many early compound administrators came from South Africa. This is not surprising since 60% to 70% of the white laborers on the Copperbelt were South African in the early years. They invariably retained white South African racial attitudes. Most of the compound administrators had worked in South African, Southern Rhodesian, or Katangan compounds before coming to Northern Rhodesia, and many had been in the Northern Rhodesian Police, the British South Africa Company Police, or the army. A military background was considered helpful for quelling potential disturbances. The compound staff was not as well paid as those Europeans directly involved in production, and few of them were highly educated. Experience “handling” Africans was the primary qualification for the job. Many compound officials tried to simply adopt the ideas current in South African compound adminstration. At best this resulted in paternalism which likened Africans to children.54 At worst, it involved an outright contempt and distaste for anything African. The expectations of deference and obedience common in South African compounds however, had to be modified to fit the Northern Rhodesian labor market. After all, the compound manager had to attract and hold scarce labor, and as a rule he gave veteran workers special consideration.55 Thus, from the beginning of the copper industry, the labor needs of copper production and the scarcity of experienced labor altered the South African pattern of compound management, forcing greater responsiveness to the demands of labor, or at least to the demands of the more experienced sector of the black work force.
Compound administrations varied to some degree with the personality of the compound manager. Cecil Spearpoint at Roan, who had been a manager in Southern Rhodesia before moving north, was admired for his skill in sports. Both Chris Cook (Assistant Compound Manager at Roan) and Spearpoint were said to be “very liberal. They were very good men.”56 In contrast, the first compound manager at Mufulira, Ben Schaefer, was extremely harsh and unpopular. He had been compound manager of the largest Union Miniere compound before coming to Mufulira, and was a strict disciplinarian, dismissing workers for wife-beating and other “moral” offences. He was known to “box laborers on the ears” when considered appropriate.57 H. H. Field, who followed Schaefer as compound manager, was more popular. He was a “quiet and fair man,” although his assistant, Twigg, was “easier to talk to.”58 Both Field and Spearpoint distrusted the more skilled black miners and preferred to buttress the role of traditional elders in the compounds. William Scrivener of Rhokana was more aloof and unapproachable. Having worked at Union Miniere before coming to the Copperbelt, Scrivener was an expert on “scientific” compound management, and was roundly disliked by the miners. Workers feared him, avoiding him whenever possible.59 M. Mwendapole remembered Scrivener as “a very rough man . . . a person who drives some fear into the miners.”60 Gabbitas at Nchanga was more accessible and popular. Miners recalled that “Gabbitas is liked all over the Copperbelt, even in Broken Hill. He has the love of the people, he makes no difference between Europeans and Africans. He is not jealous. Is good to his clerks and his servants. . . . Chamber of Mines listens when he speaks, all the football people like him.”61 All four of these men worked at the mines for most of the colonial period, providing important continuity and giving a distinctive flavor to their compound administrations. But, the differences were more of style than substance. On broad policy issues, the compound managers followed the dictates of upper management, which were shaped by production needs and the position of Northern Rhodesian copper in the world market.62
Mine officials were directed to create a village-like atmosphere in the compounds in order to attract labor and ease the transition of African workers and their families to the urban industrial world. By 1930, Field even complained that “it almost appeared now that competitive recruiting has been done away with, the mines were going in for a competitive Compound program.”63 Because Roan needed more experienced labor, it was most directly in competition with Union Miniere and thus had to meet Union Miniere’s standards. Between 1926 and 1927, expenditure per worker at Roan increased by 40%, primarily for housing, food, and social services. In 1932, when Union Miniere’s service cost for each African employee reached a record 22s5d per ticket, Roan spent 16s5d for each of its African employees. Management at Roan and Mufulira also brought in teams of experts to help improve health standards. Anglo-American, in contrast, spent only 11s64d per ticket on services, hired fewer experts, and was less willing to provide more than the bare essentials for its workers.64
Housing on the Copperbelt mines reflected different positions in the wage labor market as well. Those mines needing more experienced workers competed directly with each other, particularly Broken Hill and Union Miniere. Of course, even the best housing for African miners was far below European standards. Most mine houses were round one-room structures, built of Kimberley brick with dirt floors, a thatched or iron roof, and often lacked even one window. Single men were often housed in barracks called 10–10’s, long buildings with ten small rooms back to back. Housing had only the simplest furniture, and no water or electricity. Overcrowding was common, with seven or eight men sleeping in a house designed for four. The Rhodesian Selection Trust mines did the most to provide conditions which would attract labor. In 1930, Mufulira boasted the best housing on the Copperbelt. Roan compensated for lower-quality housing by allowing employees to build fences around their homes, make hen coops, and keep fowl and domestic animals. Married workers liked this because it supplemented the meager rations provided for wives and children.65 In contrast, Anglo-American housing at Nkana and Nchanga reflected the need for less-skilled workers and the consequent lack of interest in married labor. Nkana compound, which housed 6,000 laborers on 159 acres in 1931, was “utterly unAfrican in its precision and mathematical regularity.”66 The single men lived in barracks, there were no trees, and the general atmosphere was barren and hostile. Nchanga was similarly stark, despite a 1931 program to build some brick houses with kitchen huts in front. Most units still had only room for either a family or a number of single men.67
Health care and accident prevention received systematic attention at the mines in this early period. Both companies had to solve the malaria problem to keep both black and white labor productive. In 1928, often 30% of the European miners were laid up with malaria. Irwin brought in a tropical medicine specialist, Dr. Watson, and within a few years the number of malarial cases dropped dramatically.68 General medical facilities for blacks improved as well, and in 1931, 17 per 1,000 Roan employees died from illness, while Nkana’s death rate was 39 per 1,000. Neither of these figures compare with Union Miniere’s rate of 8.01 per 1,000, nor the Rand’s 12 per 1,000, but the Rhodesian Selection Trust mines were clearly trying harder because of their greater need to attract and hold experienced miners. Both companies sought to minimize accident rates, for accidents interfered with productivity and scared off potential employees. Again, Roan had the best safety procedures. All African gang supervisors (boss boys) were trained in first-aid work. Candidates for blasting certificates at Nkana received similar training.69
The mines made no efforts to improve compensation payments, probably because this was a less immediate issue for workers than wages and living standards. The Workmen’s Compensation Scales set up in 1929 provided a modest sum to relatives in the event of death on the mines, and compound managers set the award for injured miners. Only a court decision could alter this; consequently few workers contested their awards. Compensation was paid in a lump sum, contrasting with Katanga’s system of monthly payments throughout life or the period of disability.70
Special advantages were awarded married miners in order to attract and keep them on the mines for minimal cost. Roan gave married miners garden plots to supplement their mine rations, and purchased their surplus produce, buying 30 tons of vegetables from employees in 1931.71 Beer production also supplemented married workers’ incomes. Each mine rotated permits for beer brewing in the compounds. Since women did the brewing, only married workers made beer, which became an important source of income.72 Brewing was usually done near paydays to maximize sales. It was so profitable that illict production became an important source of additional cash. Spearpoint claimed that “for every 10 permits issued, there were another 10 people illicitly brewing beer.” Police and clerks even participated, or at least “looked the other way” for a small fee or some beer. Illegal brewing continued even after the permit system ended in the early 1930s, when all beer was ostensibly produced in government-run halls.73
The companies set up welfare facilities in the compounds to keep miners and their dependents busy and, hopefully, content during their period of employment. Once again, Roan’s programs aimed to attract and retain adequate numbers of skilled, educated workers. Many of these men wanted to better themselves through educational programs, and to fulfill these needs Roan hired a welfare officer in 1929. The following year a large recreation hall was completed.
In the main hall provision has been made for letter writing, there are tables and benches placed conveniently, and racks for native newspapers are provided. It is also possible to obtain soft drinks, biscuits, writing paper and envelopes, pencils, tobacco and cigarettes. Hot tea is always available. . . . Another item which is popular is a gramophone with a quantity of records.74
The gramophone often did not work, and the reading material disappeared, but the hall still provided a center where more educated miners could come to relax, learn, and communicate with one another. Concerts, gramophone recitals, debates, and indoor games were held in the evenings for more educated audiences. Night school offered classes in English, first-aid, and other subjects of interest to those aiming for self-improvement. While the mines offered few programs for women, there were some activities for children. A small school tried to educate a fluctuating pupil population; a playground with swings and other equipment was fenced in during 1930. And for the mass of the work force, the welfare officer organized sports, cinemas, and traditional dancing. The mines built facilities for the popular physical training classes, football, pushball, and other games. Twice-weekly movies frequently had audiences of over 2,000 people, though films were carefully screened to keep “the native from losing his respect for European women/’ Mufulira had similar programs.75 Of course, none of these programs compared favorably with the lavish European facilities. Budgetary considerations severely limited welfare facilities on the African compounds. But, within this constraint, management at the Rhodesian Selection Trust mines supported compound welfare programs, both to attract and improve the quality of experienced labor, and to monitor the leisure time of the entire work force.
The lower percentage of experienced stabilized miners at Nkana and Nchanga held down Anglo-American’s investment in welfare programs. Because of the higher turnover, management eschewed programs demanding extended participation. Education, Pathfinder Scouts, debating clubs, and other activities were less appealing to short-term employees. Anglo-American management designed welfare activities to control and amuse workers during their leisure hours, not to keep them at the mines. Consequently, welfare concentrated on sports and cinemas. A welfare officer, H. C. Nutter, ran a small children’s school at Nchanga, and taught the miners carpentry and other skills which could supplement wages. He also policed conditions and encouraged hygiene. Nonetheless, Nutter was unpopular because of his reputation for taking advantage of miners’ wives while their husbands worked. Despite objections by miners, Anglo-American retained Nutter. Clearly his low salary and money-saving activities were more important than his effectiveness as a welfare officer. Nchanga had the worst facilities on the Copperbelt before it closed down in 1931.76
Labor Control on the Mines
Along with the establishment of programs to attract mine labor, management still worried about maintaining the discipline necessary for maximal output. The mines turned to the Northern Rhodesian government for help, which in turn passed labor legislation similar to that already established in South Africa. A Native Registration Ordinance of 1929 forced all Africans working outside the native reserves to carry identification cards, or citupas. The cards helped identify and regulate workers. They recorded a person’s work history, and had to be stamped by employers at termination. This enabled employers to punish workers by refusing to sign a citupa, or writing negative comments on it. Workers could thus be “blackballed” for uncooperative behavior. However, citupas were readily forged. In 1935, the provincial administration estimated that at least half of the 7000 certificates it issued were duplicates. Provincial administrators were understandably discouraged with the system, although the mines successfully fought a move to repeal the Registration Ordinance.77
Other laws attempted to control African labor. The Employment of Natives Ordinance of 1929 made contract-breaking a criminal offense. Contracts were defined loosely, and verbal contracts were accepted as evidence in court. As a result, employers could take unsatisfactory employees to court for breach of contract, and have them punished by fines and even jail.78 Thus, government regulations provided a legal structure by which employers could threaten and coerce black laborers to fulfill their contracts. Still, as we have seen, the colonial state was unwilling to stop the flow of African labor to neighboring countries.
Discipline on all the mines, therefore, was limited by the employment mobility of miners before the Depression. The mines discovered that when European supervisors were “too persistent, the following day most of [the workers] had deserted and completely vanished.”79 Although the mines prosecuted deserters whenever possible, this was difficult: workers could change identities easily. A man merely reported his citupa missing at the nearest District Office, and received another. He could then take his new citupa to another employer, and be hired. Voluntary labor worked by the ticket, and was free to leave at the completion of each thirty—day period of employment. In the labor shortage, even unskilled miners could find new jobs relatively easily. Those miners with some training were always in demand, and could readily find work at any of the mines.80
This state of affairs forced the mining companies to set up a system for enforcing industrial discipline which did not completely alienate the mine work force. Management designed a daily routine to encourage regular work habits among the miners. Each compound was divided into sections, with several mine police assigned to each one. The police circulated throughout their areas twenty-four hours a day, and woke everyone for work. Scrivener reported that at Nkana “a bell is rung in three places in the compound at 4:30 a.m. and the Police boys go round their sections shouting to the people that it is time to get up and go to work. . . . The underground workers must be at the shaft head not later than 6:45.”81 All miners were checked in at work; absentees were counted, and then investigated by the police. The sick were taken to the hospital, and unexcused absences were reported to the compound manager. A system of bonuses rewarded productive work habits. Each week workers with perfect attendance and high productivity received a 1s3d bonus. If a worker wore his boots and coat all week, he received another bonus of 9d per week. Workers who did not perform satisfactorily received no bonus and were sometimes fined. Roan set up a similar system. Both Copperbelt companies used “loafer tickets” as well, whereby each day European supervisors marked the tickets of their American miners. If a supervisor was displeased with a miner’s performance for the day, he gave him a loafer ticket, which entitled the miner to rations but no pay.82
The entire compound structure was designed to control the daily life of miners and their dependents as well. A high fence surrounded each compound, and carefully guarded gates monitored the flow of persons in and out of the facility. In order to avoid fights among ethnic groups, the seventy or more groups were deliberately scattered throughout the compounds. Housing was allocated by seniority, occupational level, and family size. Only mine employees, their dependents, and temporary registered guests were allowed to live in the compounds. Mine police kept the peace. Food was supplied daily in carefully monitored feeding lines to all the employees and dependents, and only registered inhabitants could use the health facilities. Beer hall hours were strictly regulated to keep the miners from arriving drunk at work. Welfare programs tried to keep the compound inhabitants busy and out of trouble when idle.83
Most disciplining in the Northern Rhodesian mine compounds was conducted without recourse to outside institutions, for the mine police were permitted to make “arrests” on the premises. The police also regularly swept through the compounds searching for illicit beer-brewing and unauthorized visitors.84 As one informant recalled, “If you opened a door, and the police wanted to search your house, they just opened up and barged through. They took you to the compound manager if arrested.”85 Compound managers heard cases, reached a verdict, and inflicted punishment much like a District Officer. Persons committing first offenses or minor infractions were fined, but serious crimes or repeated offenses brought swift dismissal. Small variations occurred. At Mufulira, Ben Schaefer dismissed men who beat their wives. while the same offense at other compounds brought only warnings or fines. The ease with which compound officials dismissed workers also varied with the labor supply and skill of the miner involved. Scrivener dismissed workers more readily because of Nkana’s more enviable labor supply.86
In order to maintain the loyalty of the mine police, management deliberately separated them from the rest of the work force. They were given higher wages and special uniforms, were housed in separate areas with better facilities. Every effort was made to isolate them from the influence of friends and relatives in order to avoid corruption. Management emphasized the special position of the police as a favored group, a “junior partner” of the compound administration.87
Despite these efforts, the mines never fully trusted the police. Compound administrators believed that “it is not a desirable practice to rely on police boys as a medium of contact with the natives in every phase of life.”88 The companies hired other Africans to spy on the miners. These “undercover agents” went to meetings, listened to conversations in the beer halls, and generally observed daily life around the compounds. Anything deemed a threat to management was reported to the authorities.89
In 1931, Spearpoint set up tribal representatives in order to bypass the authority of the African mine police and clerks and to facilitate the solution of small problems in the compound. Each ethnic group with over twenty-five people selected its own representative. These officials were directed to “adjust minor difficulties among the Natives . . . and . . . help the compound manager to keep in touch with any grievances.”90 Spearpoint expected them to “overcome the chance of difficulties about police and/or clerks reporting problems.” The representatives received better housing, and some extra rations. They “functioned] as a contact group between White officials and Natives keeping either side informed about the other,” and according to Spearpoint, they were a great success.91 In contrast, Scrivener refused to follow suit at Rhokana, preferring to keep discipline under his direct control.92 Once again, Rhodesian Selection Trust labor strategies were shaped by the greater need to attract and keep experienced labor. This pattern holds true until the early 1930s, when the Depression drastically altered the labor market in southern Africa.
THE COPPERBELT, 1931–1936
The Copper Industry in the Depression
The onset of a severe depression in 1931 dramatically reduced the price and the demand for copper on the world market. Prices plummeted from £74 per ton in early 1931 to £24 per ton a few months later. Even Northern Rhodesia’s position as a low-cost supplier could not ensure sufficient demand, and its producers had to cut back sharply. Rhokana and Roan alone turned out small amounts of copper; Mufulira and Nchanga became virtual ghost towns.93
The cutbacks forced a parallel reduction in employment, and the mines shifted suddenly from a shortage to an overabundance of laborers. The number of European employees fell from 3,326 to 964, and African employees declined from 30,000 miners in early 1931 to 11,636 in January 1932. By December 1932, this number had dwindled to a mere 6,677. Large numbers of unemployed workers drifted around the Copperbelt in search of work, and the government reported a floating population of at least 5,000 unemployed moving between Ndola and the mines as late as 1935.94
Suddenly the African laborer lost his leverage in the labor market and the balance of power shifted to the mining companies. As employment opportunities dried up throughout Southern Africa, miners clung fiercely to their jobs. Absenteeism and desertion fell to new lows. In 1932, Roan listed only five absentees daily out of a total average work force of 2,317, a rate of 34 absentees per 1,000. This contrasts sharply with 638.2 absentees per 1,000 in 1929 and 377.4 absentees per 1,000 in 1931. Even Anglo-American mines had no trouble keeping workers.95 At Nkana mine alone, over 20,000 men applied for 6,723 places in 1935; Roan was swamped with job applicants, reporting over 500 men applying for only 150 jobs in 1936. Skilled and semiskilled miners were easier to come by, for they were more accustomed to urban life, and many remained near the Copperbelt after losing their jobs in hopes of finding employment which would enable them to continue living in town. The mines were even able to set up labor pools of potential employees by feeding and housing the unemployed in the compounds.96
The two companies quickly exploited this advantage in the labor market. The lessons of earlier competition had been well learned. Led by Rhokana in 1932, the mines agreed to coordinate a cut in African wages. In 1933, surface rates for new employees were reduced from 17s6d per 30 working days to 12s6d per 30 working days, and underground scales were similarly reduced from 30s0d to 22s6d (at Nkana it was 20s6d) per 30 working days. Increments for experience remained the same, but the maximum levels were reduced.97 The mines reassured themselves that they were only providing a fair wage in a formerly unrealistic and inflated labor market. The wage cuts were declared “reasonable” and became the foundation of a new labor policy.
In order to further reduce the cost of labor, the mines disbanded the Northern Rhodesian recruiting agency and established a purely voluntary work force. This exposed them to periodic shortages of skilled labor. As a result, the mines still required a force of stabilized experienced labor in order to guarantee smooth production. Roan, with its higher component of skilled labor, continued to rely on married workers, while Rhokana reduced its percentage of such workers.98 By 1935, 57% of Roan’s work force of 4,293 men were married, while only 27% of Rhokana’s 6,606 miners lived at the mine with their wives.99 Roan also encouraged longer periods of employment, and in 1935 its turnover rate for black miners was 57.1% as opposed to Nkana’s 108.6%.100 The nature of the orebody at Rhokana allowed management to maintain higher turnover rates, which it did. Rhokana also took advantage of the labor surplus to abandon its plans for permanently settling some workers.101 However, at both mines some stabilization continued, although at differing rates.
When Mufulira resumed production in 1934, it too established a core of married stabilized workers. J.D. Tallant, the general manager, and Ben Schaefer, the compound manager, were primarily concerned with getting the mine solidly established. Initially, short-term labor needed for construction projects kept the labor turnover rate higher than Nkana’s. However, Tallant’s belief that a “definite industrialized population” would be necessary in the long run reflected the more positive attitude of the Rhodesian Selection Trust mines toward longer service employees, as well as Mufulira’s technical needs. Mufulira’s management thus raised the percentage of married miners from 32% in 1933 to 38.9% in 1935, and by the late 1930s, this figure nearly equalled that at Roan.102
Having adequate labor supplies, the mines allowed housing conditions to deteriorate. In 1932, only 220 out of a total of 2,325 houses at Roan had more than one room, and most married workers lived in a single overcrowded room with their families. One informant recalled that it was common for children to sleep in little jury-rigged shelters near the house.103 Unmarried quarters were even more overcrowded, frequently housing four to six men in one small house.104 Mufulira had the best housing on the Copperbelt, but its quarters were still cramped.105 Rhokana’s three compounds remained the worst on the Copperbelt, the married quarters being “little more than slums.”106 Even Geddes admitted that “the siting of the married quarters quite defeated him.”107 At the Kitwe compound, overcrowding and lack of privacy were endemic. The new Mindolo compound had separate houses in better condition, but the compound was small and congested.108 Visitors to the Rhokana compounds were struck by “the more rigid and mechanical atmosphere and less of the personal touch than exists at Roan.”109 After their tour of the compounds in 1935, the Russell Commission (set up to investigate the 1935 strike) declared their “general impression . . . was a lack of shade in the compound and that the huts might be described as austere quarters rather than houses.”110
Little effort was made to improve the welfare programs at the mines, for the labor supply no longer hinged on such benefits. Nutter continued running his little school, teaching carpentry and generally seeing “to the repairs in the compounds.”111 Both mines maintained small welfare halls, but attendance was low. Roan reported a daily attendance at their welfare hall of less than 10 in 1932. Only the dry canteen was well used.112 The compound managers had no incentive to improve recreation programs, and dismissed the need for change. They concluded that “you are up against the native apathy towards anything that is done for his recreation. . . . It is solely for the amusement of the Bwanas.”113 Now that labor was abundant, the companies relaxed their programs.
Mining accidents increased in this period, despite government inspections of the mines. With the labor glut, the mines no longer feared loss of labor from increased accident rates. The fatal accident rate rose from 1.33 per 1,000 in 1932 to 3.93 per 1,000 in 1933, back to 1.84 per 1,000 (this due to some efforts at reform after a strike) in 1935. Serious personal injuries also rose, reaching a peak of 13.67 per 1,000 in 1933. Most accidents were underground, and some increase can be explained by initial production problems. However, the companies admitted that more precautions could be taken. Nkana was the worst offender. In 1933, one of Rhokana’s directors admitted that Nkana was shamefully behind Roan. He maintained that of the fifteen deaths in the last five months of 1933, ten “should and could” have been prevented by the European staff.114 Despite such chastisement, no serious efforts were made at reform.
Only those facilities directly affecting productivity, such as rations and health care, remained the same. Rations continued at their pre-Depression levels, and sickness rates improved at both mines, with Roan once again in the lead. African sickness and death rates still outstripped those of the local European community, but the mines made concerted, although rarely equal, efforts to protect both communities. As in earlier years, death rates were highest among miner dependents, especially the children. Crowding encouraged epidemics which swept periodically through the compounds, leaving death and disruption in their wake.115 Fear of contamination in the European community undoubtedly spurred on company efforts at improvement, but health care for Africans focused particularly on the miners, whose health directly affected production.
In order to maintain productivity in a period of lower rewards for labor, the mines tightened their discipline system. Assaults and beatings increased as miner mobility decreased. The compound manager at Mufulira regularly “boxed employees’ ears” as a mild form of punishment. This was not done as frequently at the other mines, but assaults by Europeans on African workers were still tolerated by management, bringing mild rebukes “except in exceptionally brutal cases.”116 Dismissal became a potent weapon much feared by the miners. Management could count upon large numbers of recruits eager to replace dismissed workers. The threat of dismissals kept miners efficient and disciplined. This was especially true for those with families who preferred to stay on the Copperbelt. African miners were well aware of the difficulty of being rehired at other mines after a dismissal for bad conduct. Desertion rates fell rapidly in response to the tight labor market. Scrivener even conceded to the Russell Commission that the policy of dismissal would not maintain discipline if a labor shortage developed, but as long as the labor surplus continued, it was quite effective. The compound manager at Roan also told the Commission that “we find the threat of dismissal and the bonus system fulfills our needs.”117
Conditions did not deteriorate as much at the Rhodesian Selection Trust because of its dependence upon stabilized labor. Roan, for example, continued to supplement married workers’ incomes with garden plots, and maintained facilities for women and children in order to attract and hold experienced miners. Both F. Ayer and C. Spearpoint were cited for “genuine interest in the native.”118 In 1933, Reverend R. J. B. Moore found the general atmosphere at Roan “much better than anywhere else we went. There is a genuine interest in the welfare of the native at least on the part of management.”119 The Roan administration was also more open to innovations. Its general manager even told the Chief Secretary in Livingstone in 1934 that “should you in your travels see any good native movements or schemes which would assist in building up the health and efficiency and contentment of our native organization and native families, I shall appreciate it ever so much if you would drop us a line.”120
The Colonial State in the Depression
The Northern Rhodesian government willingly acceded to these changes, partly because the Depression severely reduced revenues. In fact, the year 1932–33 left the government £177,041 in debt, and government officials saw little to gain by focusing state policy on the precarious future of the mining companies. The colonial state, after all, was concerned with maximizing the transfer of surplus to the metropole. The way to achieve that in the Depression seemed clear: limit spending on Africans and settlers in Northern Rhodesia and continue to push African laborers into the migrant labor systems to the south. To that end, Governor Sir Ronald Storrs (1932–34) continued colonial taxation and labor legislation, but disbanded his predecessor’s projects, particularly those for Africans, fired several department heads, pulled down all the wireless stations throughout the country, and stopped all construction projects. He reduced the provincial administration from 110 in 1932 to 90 persons in 1935, and the number of provinces from nine to five in 1935. Technical, medical, and educational personnel were sharply curtailed as well.121
The next governor, Sir Huburt Young (1934–38), followed Storr’s policies. Formerly Governor of Nyasaland, Young was an ardent exponent of indirect rule, and like so many colonial officials of the time, believed Africans should remain tied to the land and their traditional cultures as long as they also participated in the migrant labor system. He supported the extension of the powers of rural chiefs and their councils established in 1930. This policy was predicated on the assumption that migrant workers would maintain their traditional loyalties, and return home after employment. Even after the financial recovery of the mines, Young envisioned them as precarious, citing the wasting character of the copper ores and the uncertain future of the base metal market.122
This attitude, together with the inadequate financial and manpower position of the provincial administration, led to a poorly defined, ad hoc government policy on the Copperbelt. The much reduced provincial staff improvised policy as problems came along, their efforts shaped by the need to minimize expenditures and maximize revenues. Provincial administrators made regular sweeps through the compounds to collect taxes, but as far as possible left the daily business of running the compounds to the mines.123 They had neither the resources nor the personnel to intervene in the mine compounds. Even those officials interested in the mines were “so occupied with water-borne sewage versus types of huts, etc., and court work that they have had no time to give to the native.”124 Since the mines had not yet become the primary source of government revenue, they did not appear to warrant much government support. Before and during the Depression, conditions in the mine compounds outshone those in the compounds of smaller employers, encouraging government officials to rely on the mines to provide adequate living standards for their employees. While some government officials were critical, most government reports reflected this trust.125 For the most part, the mines had a free hand in their compounds.
When problems with the Northern Rhodesian administration did occur, the mines generally tried to resolve them through influential connections in London. Both companies had offices in London, where they maintained regular contact with the Colonial Office and high-level government officials. A number of factors led to “a certain lack of mutual comprehension in the relations between the industry and the Administration,” and “a perceptible air of suspicion in the transaction of business.”126 A few government officials, notably Charles Dundas and some Copperbelt officials, found their inability to improve mine compound conditions frustrating, and openly criticized the mines for exploiting black labor.127 On the other hand, management resented the vacillating policies of Northern Rhodesian officials,128 the continued belief that Northern Rhodesia was basically a migrant labor pool for the south, and the growing influence of the settlers, particularly the white mineworkers. The companies worried that the white miners might eventually demand a union, and the establishment of a color bar which “would so put up the cost of production that the mines would have to close down.”129 In order to counter any pressures which might threaten their labor strategy, both companies used their influence in London and their crucial role in Northern Rhodesian finances to buttress their position.
This approach worked for the most part during this period. In the early 1930s, the government was too short of money and manpower to significantly threaten the companies. But as the mines began to recover in 1934, the government increasingly came to depend upon the copper industry for revenue. The Colonial Office favored financial self-sufficiency for British colonies, and so also had a stake in the industry’s success. Numerous stockholders in Britain used their influence to support the industry. As the mines recovered financially, the colonial state became more supportive. Even when a large and violent strike by the black copper miners in 1935 raised the specter of the “black peril” among settlers and government officials, no serious change in policy occurred. The strike further galvanized settler and government opinion in Northern Rhodesia against permanent urban residence for Africans. However, except for an increase in the police force and official statements against increased stabilization, the government left the companies’ labor policies unchanged.130
The colonial government did occasionally oppose the mining companies in order to maintain social order. In the early 1930s, for example, the state established public beer halls in the urban areas. Government control over beer production and sales was motivated more by a desire indirectly to “tax” Africans for their own welfare programs than by any concern to provide better services for urban Africans, although government rhetoric would not admit such a thing. The settlers’ fear of uncontrolled drinking in the urban areas may have also contributed to the goverment’s desire to monitor African drinking. Initially the mines opposed this plan, preferring to maintain beer production as an income supplement for married workers. However, when the government promised to limit drinking hours so as not to undermine industrial efficiency and offered to subsidize some of the mines’ welfare programs with the profits, the companies agreed to the change. Government beer halls were established in Luanshya in 1932, and in Nkana in 1934. Mufulira continued on the permit system until later, but only because the government had not yet been able to set up a hall. Mine management soon discovered the beer halls neither increased drunkenness nor undermined discipline, and they came to depend on beer hall profit subsidies for their compound programs.131
The labor strategy of the Northern Rhodesian copper mines before and during the Depression was significantly affected by production needs and government policy. Unlike mining capital on the Rand, the Copperbelt mines could not convince the Northern Rhodesian government to help monopsonize their labor supply. Market factors, and the options available to black workers, modified both the structure of the work force and capital’s control over labor. The mining companies had to devise strategies which would attract and keep adequate supplies of stable and efficient labor. With the sudden onset of a massive labor surplus during the Depression, the mines reduced the returns for labor and tightened discipline without losing labor. However, the companies’ decision to abandon recruiting, and the need for some experienced reliable black labor, necessitated some stabilization on the Copperbelt, particularly at Roan. The decision to stabilize a section of the black work force and to modify the compound system had important consequences for the development of worker attitudes and behavior, consequences which will be investigated in the following chapters.